Survey
* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project
* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project
Laboratory Medicine for the Physician Assistant HTH 317 James Madison University Department of Health Sciences Physician Assistant Concentration David S. Knitter, MD, FCCP Leukocytes, Plts Red Cell Incides X Anemia X Electrolytes Renal Function Glucose Mg, Ca, PO4, U/A Cardiac Enzymes Lipid Profiles Cardiac Testing EKG ABG, Acid Base Spirometry CXR Liver function Hepatitis Coagulation Thyroid Adrenal Gonad Gram Stain Culture Serology Rheumatologic markers ESR, CRP Fluids Immunology Stool Introduction: The document serves as a reference for the initial approach to the topic of laboratory medicine as a branch-point in the process of medical decision-making. While the laboratory cannot make up for a poor history and physical exam, it can help to supplement and prioritize the information obtained there in. Laboratory data rarely make a diagnosis, and more importantly almost never “rules out” a clinical consideration. The values reported as “positive” or “negative” are at best arbitrary cut-offs on captive populations and don’t always lend themselves to real world application. You may ask why did I choose to write this companion series to the lecture handouts for HTH317. The answer lies in the very complexity of the topic. There are a number of “manuals of laboratory medicine” on the market. While these will serve one as a good end-point reference, they assume a significant requisite amount of knowledge about medicine and the nature of differential diagnosis. I have found them all lacking in the basic notion of “why this test” or “what do I do with this result”? By combining information from a host of references as well as almost 20 years of clinical experience, my hope is to accomplish a framework for the “why” or “what now” questions that you will be able to better answer as your knowledge base increases. In broadest terms, I hope to start with the most basic or essential tests and build in complexity as the semester progresses. ANEMIA: AN APPROACH TO DIAGNOSIS "I want to say in a single sentence what it takes books for other philosophers to say"- Frederich Nietzsche General Principles 1. Anemia is a sign, not a disease. 2. Anemias are a dynamic process. 3. Although the elderly are more prone to anemia, being elderly is not a cause of anemia. 4. The diagnosis of iron deficiency anemia mandates further work-up. Initial Work-up 1. Good H & P-Ask about blood loss, duration of anemia, family history of anemia, medication use etc.. Exam should include a careful search for splenomegaly, blood in the stool, etc.. 2. A careful review of the peripheral smear will often reveal many diagnostic clues, especially in the complex patient. 3. The reticulocyte count provides insight into whether a marrow problem is involved or if the anemia is due to blood loss or destruction. 4. Armed with the above knowledge one can then order specific tests further to explain the etiology of the anemia. 5. Given the frequency of iron deficiency and the non-specificity of RBC indices, serum ferritin should be checked in all anemic patients. 6. A word should be mentioned about the RDW. Once touted at the diagnostic panacea for anemia, more recent studies have shown it is of little help in the diagnosis of thalassemia, iron, folate or B12 deficiency, myelodysplasia, or the anemia of chronic disease (ACD). It is a boon only for the lazy clinician and the ambitious lawyer. Reticulocyte Count To be useful the reticulocyte count must be adjusted for the patient's hematocrit. Also when the hematocrit is lower reticulocytes are released earlier from the marrow so one must adjust for this phenomenon. Thus: Absolute retic= Patients retic x (Patients Hct/45) Corrected retic= Absolute retic count/Maturation time (Maturation time= 1 for Hct=45%, 1.5 for 35%, 2 for 25%, and 2.5 for 15%.) OR Total number of reticulocytes=Absolute retic x RBC number Increased reticulocytes (greater than 2-3% or 100,000/mm3 total) are seen in blood loss and hemolytic processes, although up to 25% of hemolytic anemias will present with a normal reticulocyte count due to immune destruction of red cell precursors. Retic counts are most helpful if extremely low (<0.1%) or greater than 3% (100,000/mm3 total). Anemia: Etiologies 1. Production defects: 1. Nutritional deficiencies-Vitamin B12, folate or iron deficiency. 2. Inflammation/chronic disease. 3. Primary marrow disorders-pure red cell aplasia, myelodysplasia. 2. Sequestration (hypersplenism)-usually associated with mild pancytopenia. 3. Dilutional-common in hospitalized patients. A patient's plasma volume increases with laying down and when they quit smoking. Possibly responsible for as much as a 3-6% drop in the hematocrit in the first two days of hospitalization. 4. Blood loss. 5. Blood destruction. Iron Deficiency: Diagnosis 1. RBC indices are of little diagnostic value unless the MCV is below 70fl that is only seen in iron deficiency and thalassemia. 2. Serum iron can be decreased in a variety of states including iron deficiency, inflammation and stress. The serum iron level varies tremendously from morning to evening and from day to day. The minuscule amount of iron in a multivitamin can falsely elevate the serum iron for up to 24 hours. 3. The total iron binding capacity is very specific for iron deficiency (near 100%) but has poor sensitivity (less than 30%). 4. The iron saturation (Fe/TIBC x 100) can be decreased below sixteen percents in both anemia of chronic disease and iron deficiency and is of little help in distinguishing between the two. 5. In the normal patient the serum ferritin is directly correlated with iron stores. This relationship holds true even in inflammatory states although the curve is "shifted to the left." That is, for a given level of storage iron in a patient with an inflammatory state the serum ferritin is higher. A ferritin level of greater than 100ng/ml rules out iron deficiency anemia in any patient. The only exception are in acute hepatitis or liver necrosis (but not chronic liver disease) when the serum ferritin will be massively elevated due to release of liver stores of iron. Ferritin may be falsely elevated also in disseminated TB and Hodgkin's disease. Despite these minor exceptions, the measurement of the serum ferritin is the most useful and cost-effective test of iron stores. Microcytic Anemia: Differential Diagnosis 1. Iron Deficiency. The lack of iron results in decreased hemoglobin available to the developing red cell. Thus, the erythrocytes produced are underhemoglobinized which result in smaller cells. The earliest sign of iron deficiency is decreased iron stores. This stage has a normal CBC and indices, although one can see microcytic/hypochromic cells on the smear. The anemia gradually evolves into the classic microcytic- hypochromic anemia. Diagnosis is made by showing decreased iron stores on bone marrow examination. Biochemically the diagnosis is established by a high TIBC or a low ferritin. The major diagnostic difficulty is distinguishing iron deficiency from anemia of chronic disease. 2. Anemia of Chronic Disease. (anemia of defective iron utilization). In patients with inflammatory states iron is sequestered in the RE system and is unavailable for use by the developing red cell (defective iron utilization). Thus at the erythrocyte level the defect is identical to iron deficiency and therefore results in production of underhemoglobinized red cells. This can result in a microcytic/hypochromic anemia. Additional factors including shorten red cell survival and decreased levels of erythropoietin add to the hypoproliferative state. The inflammatory state also leads to a decreased serum iron and decreased TIBC. Recently the spectrum of diseases associated with anemia of chronic disease has expanded. Besides the classic association of temporal arteritis (may be a presenting sign), rheumatoid arthritis, cancer etc., anemia of chronic disease has been found in patients with non-inflammatory medical conditions such as congestive heart failure, COPD and diabetes. Patients with anemia of chronic disease can have hemoglobins decreased into the lower 20% range and many (20-30%)will have red cell indices in the microcytic range. One important laboratory finding in anemia of chronic disease is an inappropriately low serum erythropoietin levels for a given level of anemia. Due to inhibition by cytokines and other unknown factors, in chronic disease patients the serum erythropoietin level does not rise with anemia. For example, in patients with myelodysplasia or nutritional deficiencies who have hematocrits in the 20's, the serum erythropoietin levels may be in the 100's or 1000's whereas patients with anemia of chronic disease may have levels of (say) 30. Diagnosis is made by proving ample bone marrow iron stores with decrease sideroblasts (iron containing red cell precursors). Biochemically anemia of chronic disease is a diagnosis of exclusion. The key test is to rule out iron deficiency. The serum erythropoietin level is inappropriately low when compared with the hematocrit. The RDW is of ABSOLUTELY no value in differentiating anemia due to iron deficiency from those associated with chronic disease. The serum iron is decreased in both conditions and the TIBC is low in states where iron deficiency and chronic disease co-exists thus rendering these tests useless. The finding of an elevated ferritin over 100ng/ml is an adequate demonstration of good iron stores. In the older patient or one with back pain, one should also ruleout the presence of multiple myeloma by performing a serum protein electrophoresis. In difficult cases one can resort to assessing bone marrow stores of iron. 3. Thalassemia. In this disorder it is the defective production of hemoglobin that leads to microcytosis. The main types are the beta-thalassemia, alphathalassemia and Hemoglobin E. Patients who are heterozygotes for beta-thalassemia have microcytic indices with mild (30ish) anemias. Homozygotes have very severe anemia. Peripheral smears in heterozygotes reveal microcytes and target cells. Diagnosis is established in by hemoglobin electrophoresis that shows an increased HbA2. One should check iron stores since an elevated HbA2 will not be present in patients with both thalassemia and iron deficiency. Beta-thalassemia occurs in a belt ranging from Mediterranean countries, the Middle East, India, Pakistan to Southeast Asia. Patients with beta-thalassemia traits who are of child bearing age need to have their spouse screened for beta-thalassemia and Hemoglobin E. Alpha-thalassemia also presents with microcytosis. Patients with alphathalassemia will have normal hemoglobin electrophoresis. The diagnosis of alpha-thalassemia is made by excluding other causes of microcytosis, a positive family history of microcytic anemia, and a life-long history of a microcytic anemia. Exact diagnosis requires DNA analysis. Alpha-thalassemia is distributed is a similar pattern to beta-thalassemia except for it very high frequency in Africa (up to 40%). In patients of African descent the finding of alpha-thalassemia requires no further evaluation. In patients from Asia of child bearing age the spouse should be screened (if need be with DNA analysis) to assess the risk of bearing a child with severe thalassemia. Hemoglobin E is actually an unstable beta-hemoglobin chain that presents in a similar fashion to the thalassemia. It is believed to be the most common hemoglobinopathy in the world. Hemoglobin E occurs in Southeast Asia, especially in Cambodia, Laos and Thailand. Patients who are heterozygotes are not anemic but are microcytic. Patients who are homozygotes are mildly anemic with microcytosis and target cells. The importance of Hemoglobin E lies in the fact that patients with both genes for Hemoglobin E and beta-thalassemia have severe anemia and behave in a similar fashion to patients with homozygote betathalassemia. 4. Sideroblastic Anemia. Defective production of the heme molecule is the basis of this disorder. The deficit of heme leads to the underhemoglobinazation of the erythroid precursors and microcytosis. Sideroblastic anemia can be congenital, can be due to toxins such as alcohol, lead, INH, or can be an acquired bone marrow disorder. The peripheral smear may show basophilic stippling in lead poisoned patients, a dimorphic (macrocytic and intensely microcytic red cells) in patient with acquired sideroblastic anemia, or stigmata of a myelodysplastic syndrome. Diagnosis is made by the finding of ringed sideroblasts on the bone marrow iron stain. Iron studies in patients with sideroblastic anemia usually show signs of iron-overload. Hemolytic Anemias Hemolytic anemias are either acquired or congenital. The laboratory signs of hemolytic anemias include: 1. Increased LDH (LDH1)-sensitive but not specific. 2. Increased indirect bilirubin-sensitive but not specific. 3. Increased reticulocyte count-specific but not sensitive 4. Decreased haptoglobin-specific but not sensitive. 5. Urine hemosiderin-specific but not sensitive. The indirect bilirubin is proportional to the hematocrit, so with a hematocrit of 45% the upper limit of normal is 1.00 mg/dl and with a hematocrit of 22.5% the upper limit of normal for the indirect bilirubin is 0.5mg/dl. Since tests for hemolysis suffer from a lack of sensitivity and specificity, one needs a high index of suspicion for this type of anemia. In autoimmune hemolytic anemias (AIHA) one usually sees microspherocytes on the peripheral smear and splenomegaly may be present on exam. The diagnosis is established by the finding of a positive direct antibody test (direct Coombs). AIHA may be idiopathic or associated with malignancies, drugs or other autoimmune disorders. Not all patients with a positive direct antibody test will have AIHA. Microangiopathic hemolytic anemias are hemolytic anemias in which intravascular destruction of red cells is present. One sees schistocytes in the peripheral smear and an elevated LDH. The most common associated diseases are disseminated intravascular coagulation, thrombotic thrombocytopenic purpura, hemolytic-uremic syndrome, aortic valvular disease or the presence of an artificial heart valve. Paroxysmal nocturnal hemoglobinuria is an acquired hemolytic anemia that is due to a clonal proliferation of erythrocytes abnormally sensitive to the action of compliment. Mutations in the PIG-A gene (which encodes a protein that links membrane proteins to lipid) result in hematopoietic cells lacking a variety of surface proteins. Hemolysis may be more conspicuous at night leading to the characteristic hemoglobinuria. The routine lab abnormalities of hemolysis are present. Formerly, the diagnosis is made by doing a Ham's test (acid-serum lysis) which is based on the abnormal cells unique sensitivity to complement. The best way now is to perform flow cytometer to detect the missing PIG linked proteins such as CD59. Flow cytometry is the most sensitive test for PNH. Patients are often pancytopenic and can present with aplastic anemia. The most common congenital cause of hemolysis is hereditary spherocytosis. In this disease the red cell membrane is abnormal leading to increased splenic destruction. Spherocytes are present on the peripheral smear and splenomegaly are present on exam. Patients often have a family history of gallstones. The laboratory values are consistent with hemolysis and the MCHC is elevated. The diagnosis is established by the finding of increased osmatic fragility. Other rare cause of hereditary hemolysis includes hereditary elliptocytosis. Enzyme deficiencies such as glucose-6-phosphate dehydrogenase deficiency are also important causes of hereditary hemolytic syndromes. The same population at risk for thalassemia is also at risk for G-6-PD deficiency. It is sex linked and thus only affects males. This defect is in the hexose monophosphate shunt and renders the RBC to be unable to withstand oxidative stress. Most people with this disease have hemolysis only with such stressors as infections and intake of oxidative drugs. There are two main subtypes-African (A-) and Mediterranean that tends to be more severe. Such drugs as dapsone and sulfamethoxazole may provoke severe hemolysis in these patients. Diagnosis is established by measuring enzyme activity. Since the reticulocyte has increased G-6-PD activity one may get a false negative normal level during times of hemolysis. MACROCYTOSIS An increased MCV can be due to many reasons but careful review of the patient's history and blood smear can narrow the diagnostic possibilities. The differential can be divided into two broad categories based on RBC morphology. Round macrocytosis-due to abnormal lipid composition of the erythrocyte membrane. Common etiologies include: 1. Alcoholism. 2. Liver Disease. 3. Renal Disease. 4. Hypothyroidism ("myxedema of the red cell"). Oval macrocytosis (macroovalocytes) is a sign of problems with cell DNA replication. The developing red cell has difficulty in undergoing cell division but RNA continues to be translated and transcribed into protein leading to growth of the cytoplasm while the nucleus lags behind. Often one or more cell divisions are skipped leading to a larger than normal cell. Common causes are: 1. Drug effect including cytotoxic chemotherapy (AZT now most common etiology of increased MCV). 2. Megaloblastic Anemias-Folate Deficiency or Vitamin B12 deficiency - Patients will have hypersegmented neutrophils on review of the peripheral smear. 3. Myelodysplasia - Patients have often hyposegmented neutrophils and abnormal platelet morphology. Patients with RBC autoantibodies or cold agglutinins can have a spurious increase in the MCV due to red cell clumping in the automatic counters. Patients with increased reticulocyte counts can also have an increase MCV due to the large size of the reticulocyte (MCV = 160). ABSORPTION AND METABOLISM OF VITAMIN B12 AND FOLATE Folate-The body stores very little folate (four weeks) and maintenance of folate stores is dependent on an adequate dietary intake. Folate is found in green leafy vegetables, fruits and liver. Folate is absorbed in the small bowel and circulates in a free form or loosely bond to albumin. Vitamin B12- In contrast to folate the body stores copious amounts of vitamin B 12 (2-6 years). This is fortunate as the absorption of vitamin B12 is complex and can be interrupted by a variety of mechanisms. Vitamin B12 is synthesized by microbes and the major dietary source is animal protein. When animal protein is ingested, vitamin B12 is freed from the protein and binds to "R proteins". The R protein-vitamin B12 complex travels to the duodenum where pancreatic enzymes destroy the R protein. This allows intrinsic factor (IF) to bind to vitamin B 12. This IF-vitamin B12 complex is absorbed only in the last 1-2 feet of terminal ileum. Vitamin B12 binds to transcobalamin II and is delivered to tissues. VITAMIN B12 AND FOLATE- METABOLIC PATHWAYS Both vitamin B12 and folate are key components in the synthesis of DNA due to their role in conversion of uridine to thymidine. When methyltetrahydrofolate loses a methyl group to form tetrahyrodrofolate, vitamin B12 "shuttles" the methyl group to homocysteine converting it to methionine. Tetrahydrofolate is eventually converted to methylenetetrahydrofolate required for thymidine synthase. Vitamin B12 other role is a cofactor in the conversion of methymalonyl-CoA to succinylCoA. CONSEQUENCES OF VITAMIN B12 OR FOLATE DEFICIENCY When vitamin B12 or folate is deficient, thymidine synthase function is impaired and DNA synthesis is interrupted. As described above this leads to megaloblastic changes in all rapidly dividing cells. The inability to synthesized DNA leads to ineffectual erythropoiesis. There is often erythroid hyperplasia in the marrow but most of these immature cells die before reaching maturity. This process, intramedullary hemolysis, leads to the classic biochemical picture of hemolysisraided LDH and indirect bilirubinemia. The LDH level is often in the 1,000's in patients with megaloblastic anemia. The lack of DNA synthesis affects the neutrophils leading to nuclear hypersegmentation. The anemia is of gradual onset and is often very well tolerated despite low hematocrits. Often a mild pancytopenia is seen but thrombocytopenia can be severe. Other rapidly dividing tissues are influenced by the megaloblastic process. In the GI tract this can lead to atrophy of the luminal lining and further malabsorption. As discussed further below, only vitamin B12 deficiency leads to neurological damage. The mechanism is unknown. ETIOLOGIES OF FOLATE DEFICIENCY Decreased intake Increased requirements-Patients who are pregnant, have hemolytic anemia, or psoriasis have increased needs for folate that can cause them rapidly to develop folate deficiency if intake is not kept up. Malabsorption Drugs - Patients with underlying mild folate deficiency are more susceptible to trimethoprim/sulfa, pyrimethamine and methotrexate toxicity. Oral contraceptive and anticonvulsants lead to increase consumption of folate. Alcohol- Alcohol affects several aspects of folate metabolism. Alcoholics have a poor intake of folate. In addition, folate metabolism is interfered with leading to a functional folate deficiency. Alcoholics have an inability to mobilize folate stores and can have depleted tissue stores with normal serum levels of folate. ETIOLOGIES OF VITAMIN B12 DEFICIENCY Inadequate intake is rare but seen in very strict vegetarians (no eggs or milk). Abnormal gastric events include being unable to dissociated vitamin B12 from food due to lack of stomach acid or enzymes. This is a recently recognized group of patients that may compose a very large subset of patients with vitamin B 12 deficiency. 10-30% percent of patients with partial gastrectomy will develop vitamin B12 deficiency. The recent promiscuous use of H2 and proton pump blockers is leading to an increased incidence of patients not absorbing vitamin B12. Deficient intrinsic factor most commonly occurs due to destruction of parietal cells by autoantibodies (pernicious anemia). Abnormal small bowel events include pancreatic insufficiency, blind loops syndromes (bacterial absorbing vitamin B12-IF complexes) and patients infested with Diphyllobothrium latum. Abnormal mucosal events including malabsorption syndromes and surgical removal of the terminal ileum. APPROACH TO THE PATIENT WITH A MEGALOBLASTIC ANEMIA 1. Recognizing that a megaloblastic anemia is present. 2. Diagnosing vitamin B12 and/or folate deficiency 3. Determining the underlying cause. 4. Therapy DIAGNOSING VITAMIN B12 AND/OR FOLATE DEFICIENCY When a patient is believed to have a megaloblastic anemia or a process consistent with vitamin B12 deficiency, one should draw a serum vitamin B12 level, a serum homocysteine level (more sensitive indicative of tissue stores than serum or red cell folate) and since up to 30% of patients with megaloblastic anemia have concurrent iron deficiency, a serum ferritin. One difficulty is that vitamin B12 levels are neither specific nor sensitive for vitamin B12 deficiency. Low "normal" levels have been associated with anemia and neurological disease in up to 30% of patients. Recently there has been interest in measuring levels of serum homocysteine and methylmalonic acids. These precursors build up in vitamin B12 deficiency and are more accurate indicators of tissue vitamin B12 deficiency. Increased levels of homocysteine and methylmalonic acid also are superior to vitamin B12 levels in predicting a response to vitamin B12 therapy. In patients with vitamin B12 levels under 350 one should check a methymalonic acid level to detect if tissue deficiency of vitamin B12 is present. DETERMINING THE UNDERLYING CAUSE In most of patients with folate deficiency, one can determine the underlying cause by history. The key concern in vitamin B12 deficiency is determining at which point in the complex pathway of vitamin B12 absorption the "lesion" is. The Schilling test is a test of vitamin B12 absorption. Patients are given radio labeled vitamin B12 orally and a large dose of vitamin B12 is given intravenously. The IV dose of vitamin B12 prevents binding any absorbed labeled vitamin B12 and this is excreted. The amount of excreted vitamin B12 reflects vitamin B12 absorption. The Schilling test is NOT a test of vitamin B12 deficiency but a tool to decide the etiology of the deficiency. The tradition Schilling test is called "stage I". If less than 8% of the labeled vitamin B12 is excreted then one can perform the Schilling test with a variety of diagnostic maneuvers to pinpoint the lesion. This includes giving intrinsic factor, pancreatic enzymes, or antibiotics. The Schilling test has several shortcomings. One is it require patient cooperation in collecting the 24-hour urine sample. As noted above patients can have secondary malabsorption due to vitamin B12 deficiency. Several drugs including "slow K" will cause a false positive Schilling test. Finally the classic Schilling test will not detect abnormalities in patients with difficulties in dissociating vitamin B 12 from food. The "food" Schilling test where labeled vitamin B12 is mixed with food has been proposed to detect this common group of patient. Patients with pernicious anemia can be detected by assaying for autoantibodies but these tests can lack diagnostic specificity. Antibodies to IF are specific but not sensitive and antibodies to parietal cells are sensitive but not specific for pernicious anemia. VITAMIN B12- NEUROLOGICAL CONSEQUENCES Recently it has become clear the patients can have neurological damage due to vitamin B12 deficiency without anemia. In fact as many as 30% of patients with neurological disease due to vitamin B12 deficiency will have no or only subtle hematological symptoms. Patients with the most severe neurological manifestation often have mild hematological disease. Thus it is appearing that vitamin B12 deficiency may exhibit two different types of disease states in humans - hematological or neurological. Neurological symptoms are reversible if found early but those present for over a year slowly, if ever, improve. The neurological symptoms include: Paresthesias-most often in fingers and toes. The most common symptom of vitamin B12 deficiency. Diminished vibratory sense Gait ataxia Increases deep tendon reflexes Memory loss Personality change Orthostatic hypotension VITAMIN B12 AND THE ELDERLY On routine screening as many as 10-23% of elderly patients will have low vitamin B12 levels. One study found that 14.5% had levels below 300 pg/ml with 56% of these patients having increased levels of homocysteine and methylmalonic acid indicative of tissue vitamin B12 deficiency. The most common mechanism is inability to absorb vitamin B12 from food. It is speculated the rapid rise in the use of H2 blockers will increase this problem in this patient population. Patients with dementia have lower levels of vitamin B12 then those without but treatment with vitamin B12 is often not effective, perhaps due to the long duration of the neurological damage. Studies are underway to examine the relationship of vitamin B12 deficiency to neurological disease in the elderly and the effects of early intervention. Myelodysplasia The myelodysplastic syndromes are a group of bone marrow diseases marked by various cytopenias, morphologically abnormal blood cells, dysplastic bone marrow changes and a propensity to evolve into acute leukemia. The changes on the peripheral smear can range from very abnormal looking blood smears to subtle changes. Hallmarks on the peripheral smear are pseudo-Pelger Huet cells (two-lobed neutrophils), macroovalocytes and hyposegmented neutrophils. This is commonly a disease of older patients and manifests itself as an anemia with normal iron, vitamin B12, and folate studies. Often these patients are misdiagnosed and treated ineffectually with iron or vitamin shots. Another group of patients in whom the myelodysplastic syndromes are common is patients who have undergone therapy for malignancy. Diagnosis is by bone marrow examination. Often cytogenetic abnormalities will be present and aid in diagnosis and assessing prognosis. Aplastic Anemia/Pure Red Cell Aplasia Destruction of the hemopoietic cells of the marrow by whatever means leads to the clinical condition of aplastic anemia. The patient with this condition presents with pancytopenia. A thorough history should be taken from the patient to try to identify any possible drug or toxin exposure, although in most patients the cause is unknown. Splenomegaly is absent. The peripheral smear shows a decreased number of normal red cells. Diagnosis is established by bone marrow biopsy that reveals a hypocellular marrow. Since paroxysmal nocturnal hemoglobinuria can initially present as aplastic anemia, when marrow function recovers, a Ham's test should be preformed. In young patients with aplastic anemia, bone marrow transplantation is the treatment of choice. Since the prognosis is worse in these patients if they had received transfusion, one should not transfuse these patients unless absolutely necessary. Pure red cell aplasia is the condition where the red cell precursor are selective destroyed. Since this condition can be associated with thymomas (and responsive to its removal), this should be sought with radiographic studies. Parvovirus B19, which is selectively toxic to the developing red cell, can cause a chronic infection that leads to clinical picture resembling pure red cell aplasia in susceptible patients. Parvovirus infection is also hazardous in patients dependent on increased RBC production such as those with congenital hemolytic anemia or sickle cell anemia. In those patients parvovirus infection may lead to a dramatic "aplastic crisis". Anemia and Alcoholism Anemia of complex etiologies is often present in the alcoholic patient. Alcohol has a direct toxic effect on the bone marrow that leads to decreased red cell production. Folate metabolism is also interfered with leading to functional folate deficiency. This is aggravated by the poor dietary intake of folate by alcoholics and impaired absorption of folate in these patients. The alcohol abusing patient may also have increased blood losses due to gastrointestinal bleeding and trauma. Hypersplenism due to liver disease can be a factor to the anemia. This group of patients will often have coexistent inflammatory states that will lead to defective iron use. Heavy alcohol use can even lead to a sideroblastic anemia. Diagnosis of specific defects in the alcoholic can be difficult due to the myriad of problems. The serum ferritin is a dependable gauge of marrow iron stores. Although the MCV may be normal, most alcoholics with folate deficiency will have either macroovalocytes or hypersegmented neutrophils present on the peripheral smears. Sources for blood lost should be aggressively sought. Often a bone marrow examination is required fully to explain the etiology(s) of the anemia. Indications for a Bone Marrow Aspiration and Biopsy 1. Pancytopenia. 2. Leukoerythroblastic blood smear (presence of immature white cells and nucleated red cells). 3. Staging of the lymphomas and of small cell lung cancer (not in the diagnosis of these diseases!). 4. Unexplained anemia. 5. Blood smear suggestive of myelodysplasia or of leukemia. 6. Monoclonal gammopathy. 7. Anemia with a very low (less than 0.1%) reticulocyte count. Blood Cells and the CBC Ed Uthman, MD Diplomate, American Board of Pathology This is a document in a five-part series on blood cells and anemia: 1. Blood cells and the CBC 2. Anemia: Pathophysiologic Consequences, Classification, and Clinical Investigation 3. Nutritional Anemias and Anemia of Chronic Disease 4. Hemolytic Anemias 5. Hemoglobinopathies and Thalassemias Introduction Hematopathology is not only the study of disease of the blood and bone marrow, but also of the organs and tissues which employ blood cells as principal effectors of their physiologic functions. Such would include the lymph nodes, spleen, thymus, and the many foci of lymphoid tissue found along the aerodigestive tract. Generally two types of medical subspecialists intensively practice in this area, the hematologist and the hematopathologist. The hematologist usually is a Board-certified internist who has completed additional years of training in hematology, usually as part of a combined fellowship in hematology and oncology. The thrust of this individual's work is toward the diagnosis and medical management of patients with hematologic disease, especially neoplasms, and medical management of other nonhematologic cancer. The hematopathologist, on the other hand, is usually Board-certified in anatomic and clinical pathology and has taken additional years of training in hematopathology. His or her principal activity is the morphologic diagnosis of conditions of the hematopoietic and lymphocyte-rich tissues and in the performance of laboratory testing that assists such diagnosis. Hematopathology is somewhat unique in its approach to the patient and the disease, in that 1) many diseases are understood at the molecular level, 2) the patient's tissue is easily obtainable in large quantities (in the case of peripheral blood, at least) and easily kept viable for special studies, and 3) the function of the blood (or at least the erythroid component) is relatively simple when compared to that of other organ systems. Because it is a scientifically integrated discipline hematology/hematopathology is an area which is intellectually gratifying to the eclectic individual who is well-rounded in various biomedical endeavors, including biochemistry, physiology, pharmacology, microanatomy, morphologic diagnosis, and patient care. The Blood A few nights working in a trauma center would tend to convince one that the body is just a huge bag of blood. In fact, an "average" 70 liter human body contains only about 5 liters of blood, or 7% by volume. In the normal state, blood has no business anywhere except in the confines of the heart and blood vessels and in the sinusoids of the marrow, liver, and spleen. Of the average 5 L of blood, only 2.25 L, or 45%, consists of cells. The rest is plasma, which itself consists of 93% water (by weight) and 7% solids (mostly proteins, the greatest proportion of which is albumin). Of the 2.25 L of cells, only 0.037 L (1.6%) are leukocytes. The entire circulating leukocyte population, if purified, would fit in a bartender's jigger. The total circulating platelet volume is even less -- about 0.0065 L -- or a little over one teaspoonful. Erythrocytes Structurally the simplest cell in the body, volumes have been written about the lowly red blood cell. The basic function of the rbc is the creation and maintenance of an environment salutary to the physical integrity and functionality of hemoglobin. In the normal state, erythrocytes are produced only in the skeleton (in adults only in the axial skeleton), but in pathologic states (especially myelofibrosis, which will be covered subsequently) almost any organ can become the site of erythropoiesis. Numerous substances are necessary for creation of erythrocytes, including metals (iron, cobalt, manganese), vitamins (B12, B6, C, E, folate, riboflavin, pantothenic acid, thiamin), and amino acids. Regulatory substances necessary for normal erythropoiesis include erythropoietin, thyroid hormones, and androgens. Erythrocytes progress from blast precursors in the marrow over a period of five days. Then they are released into the blood as reticulocytes, distinguishable from regular erythrocytes only with special supravital stains. The reticulocyte changes to an erythrocyte in one day and circulates for 120 days before being destroyed in the reticuloendothelial system. Clinical laboratories measure several important parameters that reflect rbc structure and function. These measurements are used to 1) evaluate the adequacy of oxygen delivery to the tissues, at least as is related to hematologic (as opposed to cardiopulmonary) factors, and 2) detect abnormalities in rbc size and shape that may provide clues to the diagnosis of a variety of hematologic conditions. Most of these tests are performed using automated equipment to analyze a simple venipuncture sample collected in a universal lavender- (or purple-) top tube containing EDTA as an anticoagulant. Let us consider each of these tests. A. Hemoglobin concentration in whole blood Referred to simply as "hemoglobin," this test involves lysing the erythrocytes, thus producing an evenly distributed solution of hemoglobin in the sample. The hemoglobin is chemically converted mole-for-mole to the more stable and easily measured cyanmethemoglobin, which is a colored compound that can be measured colorimetrically, its concentration being calculated from its amount of light absorption using Beer's Law. The normal range for hemoglobin is highly age- and sex-dependent, with men having higher values than women, and adults having higher values than children (except neonates, which have the highest values of all). For a typical clinical lab, the young adult female normal range is 12 - 16 g/dL; for adult males it is 14 - 18 g/dL. This is an easy test to perform, as hemoglobin is present in the blood in higher concentration than that of any other measured substance in laboratory medicine. The result is traditionally expressed as unit mass per volume, specifically grams per deciliter (g/dL). Ideologues in lab medicine have been maintaining for years that this unit will be replaced by Système Internationale (SI) units of moles per liter, but this has not gained any significant acceptance in clinical medicine except in the most nerdly circles. B. Erythrocyte count Also referred to as just "rbc," this simply involves counting the number of rbcs per unit volume of whole blood. Manual methods using the hated hemocytometer have been universally replaced by automated counting. The major source of error in the rbc count is an artificially reduced result that occurs in some conditions where rbcs stick together in the sample tube, with two or more cells being counted as one. The result of the test is expressed as number of cells per unit volume, specifically cells/µL. A typical lab's normal range is 4.2 - 5.4 x 106/µL for females; for adult males it is 4.7 - 6.1 x 106 /µL. C. Hematocrit This is also called the packed cell volume or PCV. It is a measure of the total volume of the erythrocytes relative to the total volume of whole blood in a sample. The result is expressed as a proportion, either unitless (e.g., 0.42) or with volume units (e.g., 0.42 L/L, or 42 cL/L [centiliters/liter]). An archaic way of expressing hematocrit is "volumes per cent" or just "percent" (42%, in the above illustration). Small office labs and stat labs measure hematocrit simply by spinning down a whole blood sample in a capillary tube and measuring the length of the column of rbcs relative to the length of the column of the whole specimen. Larger labs use automated methods that actually measure the volume individually of each of thousands of red cells in a measured volume of whole blood and add them up. The volume of individual erythrocytes can be electronically determined by measurement of their electrical impedance or their light-scattering properties. The normal range is 0.37 - 0.47 L/L for females, and 0.42 - 0.52 L/L for males. D. Erythrocyte indices The three cardinal rbc measurements described above (hemoglobin, hematocrit, and rbc count) are used to arithmetically derive the erythrocyte indices - mean corpuscular volume, mean corpuscular hemoglobin, and mean corpuscular hemoglobin concentration. As much as we all hate memorization, it is important to know how to calculate these indices and have some idea of the normal ranges. We will consider these individually. o 1. Mean corpuscular volume (MCV) This is the mean volume of all the erythrocytes counted in the sample. The value is expressed in volume units, in this case very small ones femtoliters (fL, 10-15 liter). The normal range is 80 - 94 fL. The formula for the calculation in general terms is MCV = hematocrit ÷ rbc count When using specific units, decimal fudge factors are required; for example, MCV (in fL) = (hematocrit [in L/L] x 1000) ÷ (rbc count [in millions/µL]) I think that it is easier to forget the fudge factors, use the first formula, multiply out the values while ignoring the bothersome decimal, and reposition the decimal in the final result so as to approximate the order of magnitude of the normal range. This is safe, since you will not see an MCV of 8 fL, or one of 800 fL. When the MCV is low, the blood is said to be (strong>microcytic</STRONGmacrocytic. Normocytic refers to blood with a normal MCV. Keep in mind that the MCV measures only average cell volume. The MCV can be normal while the individual red cells of the population vary wildly in volume from one to the next. Such an abnormal variation in cell volume is called anisocytosis. Some machines can measure the degree of anisocytosis by use of a parameter called the red cell distribution width (RDW). This is simply a standardized parameter (similar to the standard deviation) for mathematically expressing magnitude of dispersion of a population about a mean. The normal range for RDW is 11.5 - 14.5 %. o 2. Mean corpuscular hemoglobin (MCH) The MCH represents the mean mass of hemoglobin in the RBC and is expressed in the mass unit, picograms (pg, 10-12 gram). The value is determined by the formula, MCH (in pg) = (hemoglobin [in g/dL] x 10 ÷ (rbc count [in millions/µL]) Again, a fudge factor is required in this equation, so it helps to get some feel for the normal range (27 - 31 pg) and gestalt the decimal point, as described for MCV, above. Since small cells have less hemoglobin than large cells, variation in the MCH tends to track along with that of the MCV. The MCH is something of a minor leaguer among the indices in that it adds little information independent of the MCV. o 3. Mean corpuscular hemoglobin concentration (MCHC) This is the mean concentration of hemoglobin in the red cell. Since whole blood is about one-half cells by volume, and all of the hemoglobin is confined to the cells, you would correctly expect the MCHC to be roughly twice the value for hemoglobin in whole blood and to be expressed in the same units; the normal range is 32 - 36 g/dL. The value is calculated using the formula, MCHC [in g/dL] = hemoglobin [in g/dL] ÷ hematocrit [in L/L] Cells with normal, high, and low MCHC are referred to as normochromic, hyperchromic, and hypochromic, respectively. Again, these terms will have importance in anemia classification. Further reading on red cell disease Anemia: Pathophysiologic Consequences, Classification, and Clinical Investigation is an introduction to anemia Nutritional Anemias and Anemia of Chronic Disease deals with anemias caused by iron, folate, and vitamin B12 deficiencies. Hemolytic Anemias is concerned with anemias caused by red cells being destroyed faster than a healthy marrow can replace them. Hemoglobinopathies and Thalassemias covers sickle cell disease, hemoglobins C and E, and alpha- and beta-thalassemias. Understanding Anemia, my first book, is now available in hardback and paper. The publisher has kindly allowed me to post the full text of Chapter 1 online. You can access it through the book outline at this link. There is also a link to buy the book from online bookstores at a substantial discount. This book is aimed at general readers and presumes a knowledge of biology at the high school level, then builds from there. Leukocytes and the leukocyte differential count To consider the leukocytes together as a group is something of a granfalloon, because each type of leukocyte has its own function and ontogeny semi-independent of the others. To measure the total leukocyte count and allow this term to mean anything to the doctor is a travesty, yet the "wbc" count has traditionally been considered a cardinal measurement in a routine laboratory workup for just about any condition. I cannot emphasize too much that to evaluate critically the hematologic status of a patient, one must consider the individual absolute counts of each of the leukocyte types rather than the total wbc count. For such a critical evaluation, the first step is to order a wbc count with differential. In many labs, the result will be reported as a relative differential, something like this: WBC 6000/µL segmented neutrophils 60% band neutrophils 2% lymphocytes 25% monocytes 8% eosinophils 3% basophils 2% Your first task is to multiply the wbc count by each of the percentages given for the cell types; this gives you an absolute differential. Now you're in business to get some idea as to the pathophysiologic status of the patient's blood and marrow. Thus, the illustration above becomes: WBC 6000/µL segmented neutrophils 3600/µL band neutrophils 120/µL lymphocytes 1500/µL monocytes 480/µL eosinophils 180/µL basophils 120/µL The total wbc count is invariably done using an automated method. Routinely, the differential count is done "by hand" (i.e., through the microscope) in smaller labs, and by automated methods in larger facilities. The automated methods are amazingly accurate, considering the fine distinctions that must often be made in discerning one type of leukocyte from the other. One manufacturer's machine can quite reliably pick out one leukemic blast cell in eight hundred or more leukocytes. Now we shall consider each of the leukocyte types individually. A. Neutrophils The most populous of the circulating white cells, they are also the most short lived in circulation. After production and release by the marrow, they only circulate for about eight hours before proceeding to the tissues (via diapedesis), where they live for about a week, if all goes well. They are produced as a response to acute body stress, whether from infection, infarction, trauma, emotional distress, or other noxious stimuli. When called to a site of injury, they phagocytose invaders and other undesirable substances and usually kill themselves in the act of doing in the bad guys. Normally, the circulating neutrophil series consists only of band neutrophils and segmented neutrophils, the latter being the most mature type. In stress situations (i.e., the "acute phase reaction"), earlier forms (usually no earlier than myelocytes) can be seen in the blood. This picture is called a "left shift." The band count has been used as an indicator of acute stress. In practice, band counts tend to be less than reliable due to tremendous interobserver variability, even among seasoned medical technologists, in discriminating bands from segs by microscopy. Other morphologic clues to acute stress may be more helpful: in the acute phase reaction, any of the neutrophil forms may develop deep blue cytoplasmic granules, vacuoles, and vague blue cytoplasmic inclusions called Döhle bodies, which consist of aggregates of ribosomes and endoplasmic reticulum. All of these features are easily seen (except possibly the Döhle bodies), even by neophytes. The normal range for neutrophil (band + seg) count is 1160 - 8300 /µL for blacks, and 1700 - 8100 /µL for other groups. Keeping in mind the lower expected low-end value for blacks will save you much time (and patients much expense and pain) over the course of your career. Obesity and cigarette smoking are associated an increased neutrophil count. It is said that for each pack per day of cigarettes smoked, the granulocyte count may be expected to rise by 1000 /µL. B. Monocytes These large cells are actually more closely related to neutrophils than are the other "granulocytes," the basophil and eosinophil. Monocytes and neutrophils share the same stem cell. Monocytes are to histiocytes (or macrophages) what Bruce Wayne is to Batman. They are produced by the marrow, circulate for five to eight days, and then enter the tissues where they are mysteriously transformed into histiocytes. Here they serve as the welcome wagon for any outside invaders and are capable of "processing" foreign antigens and "presenting" them to the immunocompetent lymphocytes. They are also capable of the more brutal activity of phagocytosis. Unlike neutrophils, histiocytes can usually survive the phagocytosis of microbes. What they trade off is killing power. For instance, mycobacteria can live in histiocytes (following phagocytosis) for years. The normal range for the monocyte count is 200 - 950 /µL. C. Eosinophils These comely cells are traditionally grouped with the neutrophils and basophils as "granulocytes," another granfalloon. Current thinking is that eosinophils and neutrophils are derived from different stem cells, which are not distinguishable from each other by currently available techniques of examination. Although the hallmark of the eosinophil is the presence of bright orange, large, refractile granules, another feature helpful in identifying them (especially on H&E-stained routine histologic sections) is that they rarely have more than two nuclear lobes (unlike the neutrophil, which usually has three or four). The normal range of the absolute eosinophil count is 0 - 450 /µL. Eosinophils are capable of ameboid motion (in response to chemotactic substances released by bacteria and components of the complement system) and phagocytosis. They are often seen at the site of invasive parasitic infestations and allergic (immediate hypersensitivity) responses. Individuals with chronic allergic conditions (such as atopic rhinitis or extrinsic asthma) typically have elevated circulating eosinophil counts. The eos may serve a critical function in mitigating allergic responses, since they can 1) inactivate slow reacting substance of anaphylaxis (SRS-A), 2) neutralize histamine, and 3) inhibit mast cell degranulation. The life span of eos in the peripheral blood is about the same as that of neutrophils. Following a classic acute phase reaction, as the granulocyte count in the peripheral blood drops, the eosinophil count temporarily rises. D. Basophils The most aesthetically pleasing of all the leukocytes, the basophils are also the least numerous, the normal range of their count in peripheral blood being 0 - 200/µL. They are easily recognized by their very large, deep purple cytoplasmic granules which overlie, as well as flank, the nucleus (eosinophil granules, by contrast, only flank the nucleus but do not overlie it). It is tempting to assume that the basophil and the mast cell are the blood and tissue versions, respectively, of the same cell type. Actually it is controversial as to whether this concept is true or whether these are two different cell types. The table below presents some of the contrasts between mast cells and basophils. Feature Basophils Mast cells Nuclear morphology segmented round or ovoid Mitotic potential no yes Peroxidase content + - Acid phosphatase - + Alkaline phosphatase - + PAS reaction ++++ + In active allergic reactions, blood basophils decrease in number, while tissue mast cells increase. This reciprocal relationship suggests that they represent the same cell type (i.e., an allergen stimulates the passage of the cells from the blood to the site of the allergen in the tissues). Some experiments with animals have also shown that mast cells are marrow-derived and are capable of differentiating into cells that resemble basophils. Conversely, some recent evidence suggests that basophils (as well as eosinophils) can differentiate from metachromatic precursor cells that reside among epithelial cells in the nasal mucosa Without invoking religion or Alexander Pope ("Whatever is, is right," An Essay on Man, 1732-34) it is hard to see any useful role of the basophil/mast cell in human physiology. The mast cell is the essential effector of immediate (Type 1) hypersensitivity reactions, which produce only misery, dysfunction, and occasionally death for the hapless host. E. Lymphocytes In the immune/inflammatory response, if the neutrophils and monocytes are the brutes, the lymphocytes are the brains. It is possible to observe the horror of life without lymphocyte function by studying the unfortunate few with hereditary, X-linked, severe combined immune deficiency. Such individuals uniformly die of systemic infections at an early age (except for the "bubble boys" of yesteryear, who lived out their short lives in antiseptic prisons). The functions of lymphocytes are so diverse and complex that they are beyond the scope of this text (and the scope of the author, it must be admitted). What follows are a few general remarks concerning examination of lymphocytes in peripheral blood. After neutrophils, lymphocytes are the most numerous of the circulating leukocytes. The normal range of the lymphocyte count is 1000 - 4800/µL. Their life span may vary from several days to a lifetime (as for memory lymphocytes). Unlike neutrophils, monocytes, and eosinophils, the lymphocytes 1) can move back and forth between the vessels and the extravascular tissues, 2) are capable of reverting to blast-like cells, and 3) when so transformed, can multiply as the immunologic need arises. In normal people, most of lymphocytes are small, innocent-looking round cells with heavily "painted-on" nuclear chromatin, scant watery cytoplasm, and no granules. A small proportion of normal lymphs are larger and have more opaque, "busy-looking" cytoplasm and slightly irregular nuclei. Some of these have a few large, dark blue granules, the so called "azurophilic granules." It has been maintained that these granulated cells are Tgamma cells (i.e., T-cells that have a surface receptor for the IgG Fc region) or natural killer (NK) null-cells. Other phenotypes of lymphocytes are not recognizable as such on the routine, Wright-stained smear and require special techniques for identification. When activated by whatever means, lymphocytes can become very large (approaching or exceeding the diameter of monocytes) and basophilic (reflecting the increased amount of synthesized cytoplasmic RNA and protein). The cytoplasm becomes finely granular (reflecting increased numbers of organelles), and the nuclear chromatin becomes less clumped (the better to transcribe you with, my dear!). Such cells are called "transformed lymphocytes," "atypical lymphocytes," or "viral lymphocytes" by various votaries of blood smears. Although such cells are classically associated with viral infection (particularly infectious mononucleosis), they may also be seen in bacterial and other infections and in allergic conditions. A morphologic pitfall is mistaking them for monocytes (a harmless mistake) or leukemic blasts (not so harmless). Platelets The main thing to remember about platelets is to look for them first! A typical tyro maneuver is to study a blood smear for an hour looking for some profound hematological abnormality, never to realize there is nary a platelet in sight. It is therefore necessary to discipline yourself to first check for a normal number of platelets when sitting down with a slide, before being seduced by the midnight beauty of the basophil's alluring granules or the monocyte's monolithic sovereignty. The normal platelet count is 133 - 333 x 103/µL. Platelets are counted by machine in most hospital labs and by direct phase microscopy in smaller facilities. Since platelets are easily mistaken for garbage (and vice versa) by both techniques, the platelet count is probably the most inaccurate of all the routinely measured hematologic parameters. Actually, you can estimate the platelet count fairly accurately (up to an absolute value of about 500 x 103/µL) by multiplying the average number of platelets per oil immersion field by a factor of 20,000. For instance, an average of ten platelets per oil immersion field (derived from the counting of ten fields) would translate to 200,000/µL (10 x 20,000). Abnormal bleeding generally does not occur unless the platelet count is less than 30,000/µL, if the platelets are functioning properly. Screening for proper platelet function is accomplished by use of the bleeding time test. Other cells in peripheral blood Plasma cells sometimes appear in the peripheral blood in states characterized by reactivity of lymphocytes. Old time hematologists often maintain that the cells that look exactly like plasma cells on the smear are really "plasmacytoid lymphs," and it is usually nonproductive to argue this point with them. Endothelial cells occasionally get scooped up into the phlebotomy needle during blood collection and show up on the slide. They are huge and tend to be present in groups. Histiocytes, complete with pseudopodia and phagocytic vacuoles, may appear in states of extreme reactivity, especially in septic neonates. Nucleated red cells may also be seen in small numbers in the peripheral blood of newborns; however, in adults, even a single nucleated rbc on the slide is abnormal, indicating some sort of serious marrow stress, from hemolytic anemia to metastatic cancer. Myeloblasts are always abnormal and usually indicate leukemia or an allied neoplastic disease. Rarely they may be seen in non-neoplastic conditions, such as recovery from marrow shutdown (aplasia). Later stages of myeloid development (promyelocyte, myelocyte, metamyelocyte) may be represented in the peripheral blood in both reactive states and leukemias. Bone marrow examination This is one of the most common biopsy procedures performed on both outpatients and the hospitalized. Two types of specimens are generally obtained, the aspirate and the core biopsy. The site of biopsy is usually the posterior iliac crest (via the posterior superior iliac spine) in adults and the anterior tibia in children, although other sites are available. After local anesthesia is applied to the periosteum and overlying skin, a small needle (usually the "University of Illinois needle") is introduced (or crunched actually) into the medullary space through a small skin incision. About 0.5 mL of marrow material is aspirated and smeared onto several glass slides and stained with a stain identical or similar to the Wright stain used on peripheral blood. Some material usually remains in the syringe where it is allowed to clot. It is then fished out of the syringe, processed like all other biopsy tissue, embedded in paraffin, sectioned, and stained with hematoxylin/eosin and other selected stains. The core biopsy, generally performed after the aspirate is done, is taken with a larger, tapered needle, typically the "Jamshidi needle." This yields a core of bone (similar to a geologic core sample) which is fixed, decalcified, processed, and sectioned. The H&E-stained core biopsy and aspirate clot sections are best for assessment of marrow cellularity and the presence of metastatic neoplasms or granulomas. The Wright-stained aspirate smears are best for studying the detailed cytology of hematopoietic cells. The bone marrow biopsy procedure produces some pain for the patient, since it is impossible to anesthetize the inside of bone. The level of pain ranges from mild discomfort to agony, depending on the individual's pain threshold and level of apprehension. Some physicians elect to precede the biopsy with a benzodiazepine or other minor tranquilizer. Generally the aspiration action produces much more pain than the core biopsy. For a procedure that involves invasion of bone, the marrow biopsy is remarkably free of complications. Bleeding and infection may occur but are rare, even in severely thrombocytopenic and immunosuppressed patients. It is highly recommended that med students learn how to perform this useful procedure during the clinical years of their training. This is a document in a five-part series on blood cells and anemia: 1. Blood cells and the CBC 2. Anemia: Pathophysiologic Consequences, Classification, and Clinical Investigation 3. Nutritional Anemias and Anemia of Chronic Disease 4. Hemolytic Anemias 5. Hemoglobinopathies and Thalassemias I. Introduction Free oxygen, the plant kingdom's unique gift to this planet, is a highly reactive, dangerous substance capable of laying waste the delicate molecules that form the basis of life. How peculiar that we, as aerobes, have traded the security of a languid existence in a reducing milieu for the high-stakes, fast-lane life of free-flowing ATP, the dear currency that gives us the strength, speed, and mental facility to profoundly alter our world. Aerobic respiration, for all the complexity of the chemical reactions of intermediary metabolism, simply boils down to the body's need to find something to do with the spare electron left over from the destruction of the glucose molecule. This orphaned lepton, bereft of binding energy by its repeated violation at the hands of the cytochrome gantlet, finds no comfort in the carbon dioxide rubble of its former hexose home. Should it not find the succor of oxygen, it would escape to a feral existence of unsavory chemical reactions, where it would find itself in the company of the opprobrious Free Radicals, miscreants whose only purpose is the steric vandalism of the macromolecular cathedrals of life. It has been said that all damage to the body from any pathologic state in the end is caused by hypoxia at some level. If this is true, the story of pathology is the story of hypoxia. Preventing or correcting hypoxia is then the ultimate goal of all medical specialties. Pulmonologists and cardiologists deal with hypoxia at the gross mechanical level, but hematologists do so at the finer cellular and molecular levels. The physicochemical properties of hemoglobin and biochemical housekeeping in the erythrocyte are both in their purview, but what hematologists contend with at the grossest level is anemia. II. Definition of anemia Anemia may be defined as any condition resulting from a significant decrease in the total body erythrocyte mass. Measurement of total body rbc mass requires special radiolabeling techniques that are not amenable to general medical diagnostic work. Measurements typically substituted for rbc mass determination take advantage of the body's tendency to maintain normal total blood volume by dilution of the depleted rbc component with plasma. This adjustment results in decrease of the total blood hemoglobin concentration, the rbc count, and the hematocrit. Therefore, a pragmatic definition of anemia is a state which exists when the hemoglobin is less than 12 g/dL or the hematocrit is less than 37 cL/L. Anemia may exist as a laboratory finding in a subjectively healthy individual, because the body can, within limits, compensate for the decreased red cell mass. One must be careful in blindly applying this practical definition of anemia in every case. As the following diagram shows, it is possible to be severely anemic and have a normal hematocrit (and hemoglobin). This occurs when there is rapid hemorrhage, with red cells and plasma being rapidly lost simultaneously, before the body can respond by hiking up the plasma volume. The final example in the above diagram illustrates that a person can have a low hematocrit and not be anemic. This occurs when a patient is overhydrated, typically as a result of overenthusiastic intravenous fluid therapy. III. Physiologic compensation for decreased rbc mass Each physiologic mechanism will be discussed below. It should be noted that, although there are many adjustments that can be made, one that cannot is decrease in the tissue requirement for oxygen. Actually, overall body oxidative metabolism increases in anemia because of the energy requirement of the compensatory activities. 1. Decreased hemoglobin oxygen affinity Increased oxygen extraction of anemic blood by the tissues produces increased concentration of deoxyhemoglobin in the rbc, which stimulates the production of 2,3-diphosphoglycerate (2,3-DPG). 2,3-DPG shifts the hemoglobin-oxygen dissociation curve to the right, thus allowing the tissues to more easily strip the hemoglobin of its precious electron-accepting cargo: 2. Redistribution of blood flow In anemia selective vasoconstriction of blood vessels subserving certain nonvital areas allows more blood to flow into critical areas. The main donor sites who sacrifice their aerobic lifestyle are the skin and kidneys. Shunting of blood away from cutaneous sites is the mechanism behind the clinical finding of pallor, a cardinal sign of anemia. Although the kidney can hardly be thought of as a nonvital area, it receives (in the normal state) much more blood flow than is needed to meet its metabolic requirements. Although (by definition) total body red cell mass is decreased in anemia, in the chronically anemic patient the total blood volume paradoxically is increased, due to increased plasma volume. It is as if the body were trying to make up in blood quantity what it lacks in quality. 3. Increased cardiac output The heart can respond to tissue hypoxia by increased cardiac output. The increased output is matched by decreased peripheral vascular resistance and decreased blood viscosity (thinner blood flows more freely than thick blood), so that cardiac output can rise without an increase in blood pressure. Generally, anemia must be fairly severe (hemoglobin < 7 g/dL) before cardiac output rises. IV. Clinical signs and symptoms of anemia When the above mechanisms are overwhelmed by the increasing magnitude of the anemia, or when the demands of physical activity or intercurrent illness overwhelm them, a clinical disease state becomes apparent to the physician and to the patient. The severity of clinical symptoms bears less relationship to the severity of the anemia than to the length of time over which the condition develops. An acute hemorrhagic condition may produce symptoms with loss of as little as 20% of the total blood volume (or 20% of the total red cell mass). Conversely, anemias developing over periods long enough to allow compensatory mechanisms to operate will allow much greater loss of rbc mass before producing symptoms. It is not terribly uncommon to see a patient with a hemoglobin of 4 g/dL (hematocrit 12 cL/L), representing a loss of 70% of the rbc mass, being reluctantly dragged into a clinic by relatives concerned that he or she is looking a bit washed out. When symptoms do develop, they are pretty much what you would expect given the precarious state of oxygen delivery to the tissues: dyspnea on exertion, easy fatigability, fainting, lightheadedness, tinnitus, and headache. In addition, the hyperdynamic state of the circulatory system can produce palpitations and roaring in the ears. Pre-existing cardiovascular pathologic conditions are, as you would expect, exacerbated by the anemia. Angina pectoris, intermittent claudication, and night muscle cramps speak to the effect of anemia on already compromised perfusion. Clinical signs of a slowly developed anemia are pallor, tachycardia, and a systolic ejection murmur. In rapidly developing anemia (as from hemorrhage and certain catastrophic hemolytic anemias), additional symptoms and signs are noted: syncope on rising from bed, orthostatic hypotension (i.e., the blood pressure falls when the patient is raised from the supine to the sitting or standing positions) and orthostatic tachycardia. Keep in mind that if anemia develops through rapid enough bleeding, the hematocrit and hemoglobin will be normal (since in hemorrhage the rbc's and plasma are lost in proportion). Because of this, your appreciation of these clinical signs will serve you better in diagnosing this type of anemia than will the laboratory. V. Classification of anemias Anemias can be classified by cytometric schemes (i.e., those that depend on cell size and hemoglobin-content parameters, such as MCV and MCHC), erythrokinetic schemes (those that take into account the rates of rbc production and destruction), and biochemical/molecular schemes (those that consider the etiology of the anemia at the molecular level. An example: sickle cell anemia Cytometric classification: normochromic, normocytic Erythrokinetic classification: hemolytic Biochemical/molecular classification: DNA point mutation producing amino acid substitution in hemoglobin beta chain A. Cytometric classification Because cytometric parameters are more easily and less expensively measured than are erythrokinetic and biochemical ones, it is most practical to work from the cytometric classification, to the erythrokinetic, and then (hopefully) to the biochemical. Your first job in working up a patient with anemia is to place the case in one of three major cytometric categories: 1. Normochromic, normocytic anemia (normal MCHC, normal MCV). These include: 1. anemias of chronic disease 2. hemolytic anemias (those characterized by accelerated destruction of rbc's) 3. anemia of acute hemorrhage 4. aplastic anemias (those characterized by disappearance of rbc precursors from the marrow) 2. Hypochromic, microcytic anemia (low MCHC, low MCV). These include: 1. iron deficiency anemia 2. thalassemias 3. anemia of chronic disease (rare cases) 3. Normochromic, macrocytic anemia (normal MCHC, high MCV). These include: 1. vitamin B12 deficiency 2. folate deficiency B. Erythrokinetic classification You would now want to proceed with classifying your case based on the rate of rbc turnover. If this is high, a normoregenerative anemia exists. Such anemias are seen in hemolysis (excess destruction of rbc's) or hemorrhage (loss of rbc's from the vascular compartment. In these cases, the marrow responds appropriately to anemia by briskly stepping up the production of rbc's and releasing them into the bloodstream prematurely. There are several lab tests that allow you to determine if increased rbc turnover exists: 1. Reticulocyte count A sample of blood is stained with a supravital dye that marks reticulocytes. An increased number of reticulocytes is seen when the marrow is churning out rbc's at excessive speed (presumably to make up for those lost to hemolysis or hemorrhage). Most labs will report the result of the reticulocyte count in percent of all rbc's counted. A typical normal range is 0.51.5 %. Making clinical decisions based on this raw count is somewhat fallacious. For instance: A normal person with an rbc count of 5,000,000 /microliter and an absolute reticulocyte count of 50,000 /microliter would have a relative retic count of 1.0%. An anemic person with 2,000,000 rbc's/microliter and the same 50,000 retics/microliter would have an apparently "abnormal" relative retic count of 2.5 % and could be misdiagnosed as having high turnover. Clearly, one needs to find some way to correct the raw retic count so as to avoid this problem. One can easily calculate the absolute retic count (in cells/microliter) by multiplying the rbc count by the relative retic count. The normal range for the absolute retic count is 50,000-90,000 /microliter. 2. Serum unconjugated bilirubin and urine urobilinogen concentration When red cells, at the end of their 120-day life-span, go to the great spleen in the sky, they are systematically dismantled. Through a series of biochemical steps too boring to go into even here, the heme is changed into bilirubin. The bilirubin is greedily scarfed up by the liver, conjugated with glucuronide, squirted into the alimentary tract in the bile, and converted to urobilinogen by evangelical colonic bacteria. The urobilinogen is excreted in the stool (most of it) or reabsorbed and excreted in the urine (very little of it). This is summarized in the next diagram. In cases of accelerated rbc destruction, the capacity of the liver to capture bilirubin is saturated, and the concentration of unconjugated bilirubin in serum increases, occasionally to the point of producing clinical jaundice. Moreover, the increased production of urobilinogen that results is reflected by increased urobilinogen concentration in the urine. Unconjugated bilirubin is not water soluble and therefore will not be excreted in the urine, despite its elevation in the serum. 3. Serum haptoglobin concentration When an rbc is destroyed, the liberated hemoglobin binds mole-for-mole with a serum protein, haptoglobin. The "purpose" of this reaction is to keep the kidneys from squandering iron (free hemoglobin is freely filtered by the glomerulus, but hemoglobin-haptoglobin complexes are too big to muscle their way through, so that they are safe to bumble their way back to the reticuloendothelial system where they can be properly disassembled). The serum haptoglobin concentration then decreases. Laboratory measurement of haptoglobin is fairly easy and yields useful information to assist in documenting decreased rbc life span. In the case of hemolysis which takes place in the bloodstream (rather than in the RES), so-called intravascular hemolysis, additional biochemical phenomena are observed (see diagram, below). Free hemoglobin in excess of that which binds haptoglobin is rapidly filtered into the urine. What remains in the plasma spontaneously degrades into metheme and globin. A portion of metheme binds albumin to produce a measurable compound, methemalbumin, while the remainder binds to a measurable serum protein, hemopexin, which then decreases in serum concentration. All of the substances whose names are boxed in the diagram are those whose laboratory measurement is feasible and helpful in documenting hemolysis. 4. Bone marrow biopsy This can be used to directly observe any accelerated production of rbc's. The ratio of the number of myeloid to erythroid precursors (the M:E ratio) tends to decrease in high-production states, and the marrow becomes hypercellular. Marrow biopsy is not usually performed just to measure the M:E ratio, but to answer other hematologic questions that have been raised. The normoregenerative anemias are in contrast to those characterized by inadequate marrow response to the degree of anemia. These are the hyporegenerative anemias. In such cases, the reticulocyte production index is decreased. The classic example is aplastic anemia, in which there is primary marrow failure to produce enough erythrocyte mass. As you have probably come to expect, the distinction of these categories is not always absolute. For instance, in thalassemia major there is a degree of hemolysis (generally associated with the normoregenerative states) and inadequate marrow response to the degree of anemia. C. Biochemical classification Finally, one should attempt to determine the etiology of the anemia as specifically as possible. In some cases (e.g., iron deficiency), etiologic classification is easily attained; in others (e.g.. aplastic anemia) the biochemical mechanism of disease may be hopelessly elusive. Generally, biochemical tests are aimed at identifying a depleted cofactor necessary for normal hematopoiesis (iron, ferritin, folate, B12), an abnormally functioning enzyme (glucose-6-phosphate dehydrogenase, pyruvate kinase), or abnormal function of the immune system (the direct antiglobulin [Coombs'] test). APPROACH TO THE PATIENT History & Physical Examination Causes of changes in body fluid volume can usually be determined by the history and physical examination. Volume overload is manifested by an increase in weight and peripheral edema or ascites. Edema from local obstruction of venous return must be differentiated from systemic processes (congestive heart failure, cirrhosis, and nephrotic syndrome). A history of increased dietary sodium intake and use of medications that affect the renin-angiotensin system (angiotensin converting enzyme inhibitors, prostaglandin synthesis inhibitors, mineralocorticoids, calcium channel blockers) should be sought. Volume depletion is characterized by weight loss, excessive thirst, and dry mucous membranes. The term dehydration, pure water deficit, should be distinguished from volume depletion, in which both water and salt are lost. There may be resting tachycardia, orthostatic hypotension, or shock. Causes include vomiting or diarrhea, diuretic use, renal disease, diabetes mellitus or diabetes insipidus, inadequate oral intake associated with altered mental status, and excessive insensible losses from sweating or fever. Further Evaluation Treatment of fluid and electrolyte disorders is based on (1) assessment of total body water and its distribution, (2) serum electrolyte concentrations, (3) urine electrolyte concentrations, and (4) serum osmolality. Serial changes in body weight constitute the best way of knowing if there has been an acute change in body water balance. A. Body Water Table 21-1 shows the sex difference in total body water and the decrease in total body water that occurs with aging. B. Serum Electrolytes Table 21-2 shows the normal values for serum electrolytes. C. Evaluation of Urine Urinalysis provides information about underlying renal disorders. Samples also should be obtained for analysis of urine electrolyte abnormalities. An electrolyte concentration in urine is a useful indicator of renal handling of water and the electrolyte, ie, whether the kidney loses or preserves the electrolyte. In addition to the total urine per day, a spot urine may be used. Simple concentration, or concentration per gram creatinine excretion, is usually sufficient for initial analysis. More precisely, fractional excretion is used. Fractional excretion (FE) of an electrolyte X is calculated using a random urine sample with simultaneously obtained serum samples for X and creatinine. D. Serum Osmolality Serum osmolality (normally 285-295 mosm/kg) can be calculated from the following formula: (1 mosm of glucose equals 180 mg/L, and 1 mosm of urea nitrogen equals 28 mg/L.) Solute concentration is usually expressed in terms of osmolality. The number of particles in solution (ie, osmolytes; either molecules or ions) determines the number of milliosmoles. Each particle has a unit value of 1, so if a substance ionizes, each ion contributes the same amount as a nonionizable molecule. More importantly, permeability of the particle across the cell membrane determines if it acts as a physiologically active osmolyte. Only impermeable particles contribute to tonicity, and it is tonicity, not osmolality, that stimulates both thirst and ADH release. Substances that easily permeate cell membranes (eg, urea, ethanol) are not effective osmolytes and therefore do not cause shifting of fluid in body fluid compartments. For example, glucose in solution is nonionizable. Therefore, 1 mmol of glucose has an osmole concentration of 1 mosm/kg H2O. One millimole of NaCl, however, forms two ions in water (one Na+ and one Cl-) and has an osmole concentration of roughly 2 mosm/kg H2O. "Osmoles per kilogram of water" is osmolality; "osmoles per liter of solution" is osmolarity. At the solute concentration of body fluids, the two measurements correspond so closely that they are interchangeable. Clinical Implications In many instances, electrolyte disorders are asymptomatic. However, patients may develop lethargy, weakness, confusion, delirium, and seizures, especially in the presence of an abnormal serum sodium concentration. Often these symptoms are mistaken for primary neurologic or metabolic disorders. Muscle weakness occurs in patients with severe hypokalemia, hyperkalemia, and hypophosphatemia; confusion, seizures, and coma may develop in those with severe hypercalcemia. Measurement of electrolytes (sodium, potassium, chloride, bicarbonate, calcium, magnesium, and phosphorus) is indicated for any patient with even vague neuromuscular symptoms. CONCENTRATION DISORDERS OF SODIUM INTRODUCTION An abnormal serum sodium concentration does not necessarily imply abnormal sodium balance but rather abnormal water balance. Thus, most instances of abnormal sodium concentration are associated with abnormal serum osmolality. An abnormal sodium balance is often associated either with volume depletion or with edema formation. HYPONATREMIA Hyponatremia (defined as a serum sodium concentration less than 130 meq/L) is the most common electrolyte abnormality observed in a general hospitalized population, seen in about 2% of patients. The initial approach to its investigation is the determination of serum osmolality (Figure 21-1). The Urine Sodium Measurement of urine sodium helps distinguish renal from nonrenal causes of hyponatremia. Urine sodium exceeding 20 meq/L is consistent with renal salt wasting (diuretics, ACE inhibitors, mineralocorticoid deficiency, salt-losing nephropathy). Urine sodium less than 10 meq/L or fractional excretion of sodium less than 1% (unless diuretics have been given) implies avid sodium retention by the kidney to compensate for extrarenal fluid losses from vomiting, diarrhea, sweating, or thirdspacing, as with ascites. Isotonic Hyponatremia In hyperlipidemia and hyperproteinemia, the marked increases in lipids (chylomicrons and triglycerides, but not cholesterol) and proteins (> 10 g/dL) occupy a disproportionately large portion of the plasma volume. Plasma osmolality remains normal because its measurement is unaffected by the lipids or proteins. A decreased volume of water results, so that the sodium concentration in total plasma volume is decreased. Because the sodium concentration in the plasma water is actually normal, hyperlipidemia and hyperproteinemia cause pseudohyponatremia. Most United States laboratories measure serum electrolytes using ion-specific electrodes and thus avoid misdiagnosis. Hypotonic Hyponatremia Hypotonic hyponatremia is true hyponatremia in a physiologic sense. Since the capacity of the kidney to excrete electrolyte-free water is potentially great—up to 20-30 L/d—in the presence of a normal GFR (100 L/d), electrolyte-free water intake must theoretically exceed 30 L/d for hyponatremia to develop. Instead, in hypotonic hyponatremia, retention of electrolyte-free water nearly always occurs because of impaired excretion (renal failure, inappropriate ADH excess, etc). A. Hypovolemic Hypotonic Hyponatremia Hyponatremia with decreased extracellular fluid volume occurs in the setting of renal or extrarenal volume loss (Figure 21-1). Total body sodium is decreased. To maintain intravascular volume, antidiuretic hormone (ADH) secretion increases, and free water is retained. The drive to replenish intravascular volume overrides the need to sustain normal osmolality; losses of salt and water are replaced by water alone. The combination of low fractional excretion of sodium (< 0.5%) and low fractional urea clearance (< 55%) is the best way to predict improvement with saline therapy. Hyponatremia has been shown to develop in patients with intracranial diseases through renal sodium wasting. Unlike those with SIADH, these patients are hypovolemic, though plasma levels of ADH are inappropriately high for the osmolality. Observations in patients with subarachnoid hemorrhage suggest that the cerebral salt-wasting syndrome is caused by increased secretion of brain natriuretic peptide with suppression of aldosterone secretion. Treatment consists of replacement of lost volume with isotonic or halfnormal (0.45%) saline or lactated Ringer's infusion. The rate of correction must be adjusted to prevent permanent cerebral damage (see below). Corticosteroids can be used empirically if hypocortisolism is considered in the differential diagnosis. See Adrenocortical Hypofunction in Chapter 26 for diagnosis (by means of the cosyntropin stimulation test) and treatment of hypocortisolism. B. Euvolemic Hypotonic Hyponatremia In this setting, determinations of urine osmolality (Figure 21-1) along with urine sodium are useful for appropriate diagnosis. 1. Clinical syndromes— a. Syndrome of inappropriate ADH secretion (SIADH)—(Table 213.) Hypovolemia physiologically stimulates ADH secretion, so the diagnosis of SIADH is made only if the patient is euvolemic. In SIADH, increased ADH release occurs without osmolality-dependent or volumedependent physiologic stimulation. Normal regulation of ADH release occurs from both the central nervous system and the chest via baroreceptors and neural input. It follows that the causes of SIADH are disorders affecting the central nervous system—structural, metabolic, psychiatric, or pharmacologic—or the lungs. Furthermore, some carcinomas, such as small cell lung carcinoma, synthesize ADH. Other states associated with SIADH include administration of drugs that either increase ADH secretion or potentiate its action. SIADH induced by selective serotonin (or epinephrine) reuptake inhibitors such as fluoxetine is fairly common in geriatric patients. (1) Patterns of abnormal ADH secretion— (a) Random secretion—ADH release is unrelated to osmoregulation. This pattern is seen in carcinomas and central nervous system diseases. (b) Reset osmostat—This variant is characterized by ADH secretion appropriately suppressed at very low serum osmolalities but with ADH osmoregulation downset to a lower level of "normal." Therefore, ADH is secreted at a subnormal serum osmolality threshold (< 280 mosm/kg). Appropriate urinary dilution can be attained but at low serum osmolalities. This pattern is seen in the elderly, in patients with pulmonary processes, tuberculosis, or malnutrition. During pregnancy, the physiologic reset osmostat may suppress osmolality by about 10 mmol/kg of water. (c) Leak of ADH—In conditions such as basilar skull fractures, low levels of ADH are "leaked" into the circulation despite hypo-osmolality. If serum osmolality rises to normal, ADH secretion increases appropriately and then continues to respond normally if osmolality further increases. (2) Clinical features—SIADH is characterized by (1) hyponatremia; (2) decreased osmolality (< 280 mosm/kg) with inappropriately increased urine osmolality (> 150 mosm/kg); (3) absence of cardiac, renal, or liver disease; (4) normal thyroid and adrenal function (see Chapter 26 for thyroid function tests and cosyntropin stimulation test); and (5) urine sodium usually over 20 meq/L. Natriuresis compensates for the slight increase in volume from ADH secretion. The mechanisms that regulate sodium excretion in response to increases in extracellular volume, such as suppression of the sympathetic nervous and renin-angiotensin systems and increased secretion of atrial natriuretic factor, are preserved and account for the increase in urinary sodium. The expansion of extracellular volume is not large enough to cause clinical hypervolemia, hypertension, or edema. Other changes frequently seen in SIADH include low blood urea nitrogen (BUN) (< 10 mg/dL) and hypouricemia (< 4 mg/dL), which are not only dilutional but result from increased urea and uric acid clearances in response to the volume-expanded state. A high BUN suggests a volume-contracted state, which excludes a diagnosis of SIADH. b. Postoperative hyponatremia—Severe postoperative hyponatremia can develop in 2 days or less after elective surgery in healthy patients, especially premenopausal women. Most have received excessive postoperative hypotonic fluid in the setting of elevated ADH levels related to pain of surgery with continuing excretion of hypertonic urine. Similar mechanisms have been suspected for hyponatremia after colonoscopy. Patients awake normally from general anesthesia but within 2 days develop nausea, headache, seizures, and even respiratory arrest. Serum sodium levels may be less than 110 meq/L. Premenopausal women who develop hyponatremic encephalopathy are about 25 times more likely than menopausal women to die or to suffer permanent brain damage, suggesting a hormonal role in the pathophysiology of this disorder. Hyponatremia may also result from direct absorption of hypotonic irrigating fluids through veins during endometrial ablation or transurethral prostate resection. These patients can be symptomatic intraoperatively, with tremulousness, hypothermia, or hypoxia, or upon awakening from anesthesia, with headache, nausea, and vomiting. c. Hypothyroidism—Hyponatremia is not commonly caused by hypothyroidism, but it can occur on occasion with serum sodium levels as low as 105 meq/L. Water retention is the cause, probably both from inappropriately elevated ADH levels and from nonhormonal alterations in the handling of water by the kidneys. d. Psychogenic polydipsia and beer potomania—Marked excess free water intake (generally > 10 L/d) may produce hyponatremia. Euvolemia is maintained through the renal excretion of sodium. Urine sodium is therefore generally elevated (> 20 meq/L), but unlike SIADH, levels of ADH are suppressed. Urine osmolality is appropriately low (< 300 mosm/kg) as the increased free water is excreted. Hyponatremia from bursts of ADH occurs in manic-depressive patients with excess free water intake. Psychogenic polydipsia is observed in patients with psychologic problems, and these patients frequently take drugs interfering with water excretion. Similarly, excessive intake of beer, which contains very small amounts of sodium (< 5 meq/L), can cause severe hyponatremia in cirrhotic patients, who have elevated ADH and often decreased GFR. e. Idiosyncratic diuretic reaction—In addition to hyponatremia developing from volume contraction due to diuretic therapy (see above), a less common diuretic-induced hyponatremia can occur in euvolemic patients, typically from thiazides. This syndrome is most often seen in healthy older women (over 70 years of age) often after a few days of therapy. The mechanism for the hyponatremia appears to be a combination of excessive renal sodium loss and water retention. f. Idiosyncratic ACE inhibitor reactions—ACE inhibitors can cause central polydipsia and increased antidiuretic hormone secretion, both of which result in severe, symptomatic hyponatremia. Patients given ACE inhibitors who develop polydipsia should have their serum Na+ levels checked. g. Endurance exercise hyponatremia—Hyponatremia after endurance exercise (eg, triathlon events) may be caused by a combination of excessive fluid overload and continued ADH secretion. Reperfusion of the exercise-induced ischemic splanchnic bed causes delayed absorption of excessive quantities of hypotonic fluid ingested during exercise. Sustained elevation of ADH prevents water excretion in this setting. The retention of hypotonic fluid may be further exacerbated by NSAIDs frequently used by athletes. 2. Treatment— a. Symptomatic hyponatremia—Symptomatic hyponatremia is usually seen in patients with serum sodium levels less than 120 meq/L. If there are central nervous system symptoms, hyponatremia should be rapidly treated at any level of serum sodium concentration. (1) Rate and degree of correction—Central pontine myelinolysis may occur from osmotically induced demyelination due to overly rapid correction of serum sodium (an increase of more than 1 meq/L/h, or 25 meq/L within the first day of therapy). Hypoxic-anoxic episodes during hyponatremia may contribute to the demyelination. A reasonable approach is to increase the serum sodium concentration by no more than 1-2 meq/L/h and not more than 25-30 meq/L in the first 2 days; the rate should be reduced to 0.5-1 meq/L/h as soon as neurologic symptoms improve. The initial goal is to achieve a serum sodium concentration of 125-130 meq/L, guarding against overcorrection. (2) Saline plus furosemide—Hypertonic (eg, 3%) saline with furosemide is indicated for symptomatic hyponatremic patients. If one administers 3% saline without a diuretic to a patient with SIADH, the serum sodium concentration increases temporarily, but euvolemic patients excrete the excess sodium. If one adds furosemide (0.5-1 mg/kg intravenously), however, the kidney cannot concentrate urine even in the presence of high levels of ADH. Infusion of 3% saline is accompanied by excretion of isotonic urine with a net loss of free water. The sodium concentration of 3% saline is 513 meq/L. In order to determine how much 3% saline to administer, a spot urinary Na+ is determined after a furosemide diuresis has begun. The excreted Na+ is replaced with 3% saline, empirically begun at 1-2 mL/kg/h and then adjusted based on urinary output and urinary sodium. For example, after administration of furosemide, urine volume may be 400 mL/h and sodium plus potassium excretion 100 meq/L. The excreted Na+ plus K+ is 40 meq/h, which is replaced with 78 mL/h of 3% saline (40 meq/h divided by 513 meq/L). Free water loss is about 1% of total body water. Therefore, an approximately 1% rise in plasma sodium concentration (1-1.5 meq/L/h) can be expected. Measurements of plasma sodium should be done approximately every 4 hours and the patient observed closely. b. Asymptomatic hyponatremia—In asymptomatic hyponatremia, the correction rate of hyponatremia need be no more than 0.5 meq/L/h. No specific treatment is needed for patients with reset osmostats. (1) Water restriction—Water intake should be restricted to 0.5-1 L/d. Gradual increase of serum sodium will occur over days. (2) 0.9% saline—0.9% saline with furosemide may be used in asymptomatic patients whose serum sodium is less than 120 meq/L. Urinary sodium and potassium losses are replaced as above. (3) Demeclocycline—Demeclocycline (300-600 mg twice daily) is useful for patients who cannot adhere to water restriction or need additional therapy; it inhibits the effect of ADH on the distal tubule. Onset of action may require 1 week, and concentrating may be permanently impaired. Therapy with demeclocycline in cirrhosis appears to increase the risk of renal failure. (4) Fludrocortisone—Hyponatremia occurring as part of the cerebral saltwasting syndrome can be treated with fludrocortisone. (5) Selective vasopressin V2 antagonist—The renal effect of ADH on water excretion is mediated by the V2 receptor. Oral selective V2 antagonists have been in clinical trials and may be available for treatment of SIADH in the near future. C. Hypervolemic Hypotonic Hyponatremia Hyponatremia with increased extracellular fluid volume is seen when hyponatremia is accompanied by edema-associated disorders such as congestive heart failure, cirrhosis, nephrotic syndrome, and advanced renal disease (Figure 21-1). In congestive heart failure, total body sodium is increased, yet effective circulating volume is sensed as inadequate by baroreceptors. Increased ADH and aldosterone results, with retention of water and sodium. The urine sodium concentration is generally less than 10 meq/L unless the patient has been taking diuretics. Treatment A. Water Restriction The treatment of hyponatremia is that of the underlying condition (eg, improving cardiac output in congestive heart failure) and water restriction (to < 1-2 L of water daily). B. Diuretics To hasten excretion of water and salt, use of diuretics may be indicated. Since diuretics may worsen hyponatremia, the patient must be cautioned not to increase free water intake. C. Hypertonic (3%) Saline Hypertonic saline administration is dangerous in volume-overloaded states and is not routinely recommended. In patients with severe hyponatremia (serum sodium < 110 meq/L) and central nervous system symptoms, judicious administration of small amounts (100-200 mL) of 3% saline with diuretics may be necessary. Emergency dialysis should also be considered. Hypertonic Hyponatremia Hypertonic hyponatremia is most commonly seen with hyperglycemia. When blood glucose becomes acutely elevated, water is drawn from the cells to the extracellular space, diluting the serum sodium. The plasma sodium level falls 1.6 meq/L for every 100 mg/dL rise when the glucose concentration is between 200 and 400 mg/dL. If the glucose concentration is above 400 mg/dL, the plasma sodium concentration falls 4 meq/L for every 100 mg/dL rise in glucose. This dilutional hyponatremia is not pseudohyponatremia, since the sodium concentration does indeed fall. Infusion of hypertonic solutions containing osmotically active osmoles (eg, mannitol) may also cause hypertonic hyponatremia by drawing water to the extracellular space. Hyponatremia in AIDS Hyponatremia is seen in up to 50% of patients hospitalized for AIDS and in 20% of ambulatory AIDS patients, often associated with pneumonia and central nervous system processes. If hyponatremia is present at the time of hospital admission, it is just as likely to be due to hypovolemic gastrointestinal loss as to euvolemic SIADH. However, if hyponatremia develops after hospital admission, most patients have euvolemic SIADH. Infrequently, hypovolemic hyponatremia is due to adrenal insufficiency, isolated mineralocorticoid deficiency with hyporeninemic hypoaldosteronism, or an HIV-specific impairment in renal sodium conservation. Hyponatremia from adrenal insufficiency may coexist with hypokalemia—rather than hyperkalemia—if the patient also has diarrhea. Adrenal function can be tested with corticotropin stimulation to determine the adequacy of serum cortisol response (see Chapter 26). Patients with isolated hyporeninemic hypoaldosteronism have normal cortisol responses to corticotropin but have decreased serum aldosterone levels. Adrogue HJ et al: Hyponatremia. N Engl J Med 2000;342:1581. [PMID: 10824078] Bouman WP et al: Incidence of selective serotonin reuptake inhibitor (SSRI) induced hyponatremia due to the syndrome of inappropriate antidiuretic hormone (SIADH) secretion in the elderly. Int J Geriatr Psychiatry 1998;13:12. [PMID: 9489575] Cadnapaphornchai MA et al: Pathogenesis and management of hyponatremia. Am J Med 2000;109:688. [PMID: 11099692] Gross P: Correction of hyponatremia. Semin Nephrol 2001;21: 269. [PMID: 11320492] Hillier TA et al: Hyponatremia: Evaluating the correction factor for hyperglycemia. Am J Med 1999;106:399. [PMID: 10225241] Kumar S et al: Sodium. Lancet 1998;352:220. [PMID: 9683227] (Up-todate review of the pathogenesis of and practical treatment recommendations for sodium imbalance.) Oster JR et al: Hyponatremia, hyposmolality, and hypotonicity: tables and fables. Arch Intern Med 1999;159:333. [PMID: 10030305] (An insightful review of the interpretation of hyponatremia.) Palmer BF: Hyponatraemia in a neurosurgical patient: syndrome of inappropriate antidiuretic hormone secretion versus cerebral salt wasting. Nephrol Dial Transplant 2000;15:262. [PMID: 10648680] Thaler SM et al: "Beer potomania" in non-beer drinkers: effect of low dietary solute intake. Am J Kidney Dis 1998;31:1028. [PMID: 9631849] (Low solute intake as a cause of free water excretion defect.) HYPERNATREMIA An intact thirst mechanism usually prevents hypernatremia (> 145 meq/L). Thus, whatever the underlying disorder (eg, dehydration, lactulose or mannitol therapy, central and nephrogenic diabetes insipidus), excess water loss can cause hypernatremia only when adequate water intake is not possible, as with unconscious patients. Rarely, excessive sodium intake may cause hypernatremia. Hypernatremia in primary aldosteronism is mild and usually does not cause symptoms. Hypernatremia in the presence of salt and water overload is uncommon but has been reported in very ill patients in the course of therapy. Clinical Findings A. Symptoms and Signs When dehydration exists, orthostatic hypotension and oliguria are typical findings. Hyperthermia, delirium, and coma may be seen with severe hyperosmolality. B. Laboratory Findings 1. Urine osmolality > 400 mosm/kg—Renal water-conserving ability is functioning. a. Nonrenal losses—Hypernatremia will develop if water ingestion fails to keep up with hypotonic losses from excessive sweating, exertional losses from the respiratory tract, or through stool water. Lactulose causes an osmotic diarrhea with loss of free water. b. Renal losses—While diabetic hyperglycemia can cause pseudohyponatremia (see above), progressive volume depletion from the osmotic diuresis of glycosuria can result in true hypernatremia. Osmotic diuresis can occur with the use of mannitol or urea. 2. Urine osmolality < 250 mosm/kg—A dilute urine with osmolality less than 250 mosm/kg with hypernatremia is characteristic of central and nephrogenic diabetes insipidus. Nephrogenic diabetes insipidus, seen with lithium or demeclocycline therapy, after relief of prolonged urinary tract obstruction, or with interstitial nephritis, results from renal insensitivity to ADH. Hypercalcemia and hypokalemia may be contributing factors when present. Treatment Treatment of hypernatremia is directed toward correcting the cause of the fluid loss and replacing water and, as needed, electrolytes. In response to increases in plasma osmolality, brain cells synthesize solutes—or idiogenic osmoles—which increase osmotic flow of water back into the brain cells to regulate their volume. This begins 4-6 hours after dehydration and takes several days to reach a steady state. If hypernatremia is too rapidly corrected, the osmotic imbalance may cause water to preferentially enter brain cells, causing cerebral edema and potentially severe neurologic impairment. Fluid therapy should be administered over a 48-hour period, aiming for a decrease in serum sodium of 1 meq/L/h (1 mmol/L/h). Potassium and phosphate may be added as indicated by serum levels; other electrolytes are also monitored frequently. A. Choice of Type of Fluid for Replacement 1. Hypernatremia with hypovolemia—Severe hypovolemia should be treated with isotonic (0.9%) saline to restore the volume deficit and to treat the hyperosmolality, since the osmolality of isotonic saline (308 mosm/kg) is often lower than that of the plasma. This should be followed by 0.45% saline to replace any remaining free water deficit. Milder volume deficit may be treated with 0.45% saline and 5% dextrose in water. 2. Hypernatremia with euvolemia—Water drinking or 5% dextrose and water intravenously will result in excretion of excess sodium in the urine. If GFR is decreased, diuretics will increase urinary sodium excretion but may impair renal concentrating ability, increasing the quantity of water that needs to be replaced. 3. Hypernatremia with hypervolemia—Treatment consists of providing water as 5% dextrose in water to reduce hyperosmolality, but this will expand vascular volume. Thus, loop diuretics such as furosemide (0.5-1 mg/kg) should be administered intravenously to remove the excess sodium. In severe renal insufficiency, hemodialysis may be necessary. B. Calculation of Water Deficit When calculating fluid replacement, both the deficit and the maintenance requirements should be added to each 24-hour replacement regimen. 1. Acute hypernatremia—In acute dehydration without much solute loss, free water loss is similar to the weight loss. Initially, 5% dextrose in water may be employed. As correction of water deficit progresses, therapy should continue with 0.45% saline with dextrose. 2. Chronic hypernatremia—Water deficit is calculated to restore normal osmolality for total body water. Total body water (TBW) (Table 21-1) correlates with muscle mass and therefore decreases with advancing age, cachexia, and dehydration and is lower in women than in men. Current TBW equals 0.4-0.6 of current body weight. Adrogue HJ et al: Hypernatremia. N Engl J Med 2000;342:1493. [PMID: 10816188] Kahn T: Hypernatremia with edema. Arch Intern Med 1999; 159:93. [PMID: 9892337] Kugler JP et al: Hyponatremia and hypernatremia in the elderly. Am Fam Physician 2000;61:3623. [PMID: 10892634] POTASSIUM CONCENTRATION DISORDERS OF HYPOKALEMIA The total potassium content of the body is 50 meq/kg, more than 95% of which is intracellular. The plasma potassium concentration is maintained in a narrow range through two main regulating mechanisms: potassium shift between intracellular and extracellular compartments and modulation of renal potassium excretion. A deficit of 4-5 meq/kg occurs for each 1 meq/L decrement in serum potassium concentration below a level of 4 meq/L. Clinical Findings A. Symptoms and Signs Muscular weakness, fatigue, and muscle cramps are frequent complaints in mild to moderate hypokalemia. Smooth muscle involvement may result in constipation or ileus. Flaccid paralysis, hyporeflexia, hypercapnia, tetany, and rhabdomyolysis may be seen with severe hypokalemia (< 2.5 meq/L). B. Laboratory Findings The ECG shows decreased amplitude and broadening of T waves, prominent U waves, premature ventricular contractions, and depressed ST segments. Hypokalemia also increases the likelihood of digitalis toxicity. Thus, in patients with heart disease, hypokalemia induced by certain drugs such as 2-adrenergic agonists and diuretics may impose a substantial risk. Pathophysiology & Diagnosis Hypokalemia can occur as a result of shifting of potassium intracellularly from the extracellular space, extrarenal potassium loss (or insufficient potassium intake), or renal potassium loss (Table 21-4). Potassium uptake by the cell is stimulated by insulin in the presence of glucose. It is also facilitated by adrenergic beta-stimulation, whereas adrenergic alphastimulation blocks it. All of these effects are transient. Self-limited hypokalemia occurs in 50-60% of trauma patients, perhaps related to enhanced release of epinephrine. Profound hypokalemia due to barium intoxication has been reported that may also be the result of transport of potassium into cells. Aldosterone, which facilitates urinary potassium excretion through enhanced potassium secretion at the distal renal tubules, is the most important regulator of body potassium content. Urinary potassium concentration is low (< 20 meq/L) as a result of extrarenal fluid loss (eg, diarrhea, vomiting) and inappropriately high (> 40 meq/L) with urinary losses (eg, mineralocorticoid excess, Bartter's syndrome, Liddle's syndrome). Various genetic mutations that affect fluid and electrolyte metabolism, including disorders of potassium metabolism, have been reported recently (Table 21-5). Licorice-induced hypokalemia results from inhibition of 11-hydroxysteroid dehydrogenase, which inactivates cortisol. Cortisol thus escapes degradation, binds to aldosterone receptors, and exerts aldosterone-like effects. Recently, the transtubular K+ concentration gradient (TTKG) has been used for simple and rapid reevaluation of net potassium secretion. TTKG is calculated as follows: Hypokalemia with a TTKG > 4 suggests renal potassium loss with increased distal K+ secretion. In such cases, plasma renin and aldosterone levels are helpful in differential diagnosis. The presence of nonabsorbed anions, including bicarbonate, also increases TTKG. Treatment The safest way to treat mild to moderate deficiency is with oral potassium, and all potassium formulations are easily absorbed. Dietary potassium is almost entirely coupled to phosphate—rather than chloride— and is therefore not effective in correcting potassium loss associated with chloride depletion, such as from diuretics or vomiting. In the setting of abnormal renal function and mild to moderate diuretic dosage, 20 meq/d of oral potassium is generally sufficient to prevent hypokalemia, but 40100 meq/d over a period of days to weeks is needed to treat hypokalemia and fully replete potassium stores. Intravenous potassium replacement is indicated for patients with severe hypokalemia and for those who cannot take oral supplementation. For severe deficiency, potassium may be given through a peripheral intravenous line in a concentration that should not exceed 40 meq/L at rates of up to 40 meq/L/h. Continuous electrocardiographic monitoring is indicated, and the serum potassium level should be checked every 3-6 hours. Magnesium is an important cofactor for potassium uptake and for maintenance of intracellular potassium levels. Loop diuretics (eg, furosemide) cause substantial renal potassium and magnesium losses. Coexisting magnesium and potassium depletion can result in refractory hypokalemia despite potassium repletion if there is no magnesium repletion. Cohn JN et al: New guidelines for potassium replacement in clinical practice. Arch Intern Med 2000;160:2429. [PMID: 10979053] Gennari FJ: Hypokalemia. N Engl J Med 1998;339:451. [PMID: 9700180] (Comprehensive review.) Halperin ML et al: Potassium. Lancet 1998;352:135. [PMID: 9672294] (Pathophysiologic review of potassium disorders.) Rastegar A et al: Hypokalaemia and hyperkalaemia. Postgrad Med J 2001;77:759. [PMID: 11723313] Scheinmann SJ et al: Genetic disorders of renal electrolyte transport. N Engl J Med 1999;340:1177. [PMID: 10202170] (A comprehensive review of the molecular pathogenesis of several genetic disorders, including Bartter's syndrome and Liddle's syndrome.) Stewart OM: Mineralocorticoid hypertension. Lancet 1999;353: 1341. [PMID: 10218547] HYPERKALEMIA Many cases of hyperkalemia are spurious or associated with acidosis (Table 21-6). The common practice of repeatedly clenching and unclenching a fist during venipuncture may raise the potassium concentration by 1-2 meq/L by causing acidosis and consequent potassium loss from cells. Intracellular potassium shifts to the extracellular fluid in hyperkalemia associated with acidosis. Serum potassium concentration rises about 0.7 meq/L for every decrease of 0.1 pH unit during acidosis. Potassium movement out of cells occurs primarily in metabolic acidosis due to the accumulation of minerals such as NH4Cl or HCl. The inability of the chloride anion to permeate the cell membrane results in the transcellular exchange of H+ for K+. Metabolic acidosis from organic acids (keto acids and lactic acid) does not induce hyperkalemia. Unlike the minerals, these organic acids easily permeate cell membranes and retard Na+-K+ ATPase. The hyperkalemia frequently observed in diabetic ketoacidosis is not due to the acidosis but to a combination of the hyperosmolality (the intracellular K+ concentration of the dehydrated cell increases and K+ diffuses extracellularly) and deficiencies of insulin, catecholamines, and aldosterone. In the absence of acidosis, serum potassium concentration rises about 1 meq/L when there is a total body potassium excess of 1-4 meq/kg. However, the higher the serum potassium concentration, the smaller the excess necessary to raise the potassium levels further. Trimethoprim is structurally related to amiloride and triamterene, and all three drugs inhibit renal potassium excretion through suppression of sodium channels in the distal nephron. Serum potassium levels rise progressively over 4-5 days in patients treated with standard or highdose trimethoprim (combined with sulfamethoxazole or dapsone), especially if there is concurrent renal insufficiency. Over half of inpatients taking this drug have potassium levels over 5 meq/L, and 20% have marked hyperkalemia (> 5.5 meq/L). The potassium concentration returns to baseline after drug discontinuation. In addition, it is of note that immunosuppressive drugs such as cyclosporine and tacrolimus can induce hyperkalemia in organ transplant recipients—and especially in kidney transplant patients. This is partly due to the suppression of basolateral Na+-K+ ATPase in principal cells. Hyperkalemia is commonly seen in patients with AIDS and has been attributed to impaired renal excretion of potassium due to the use of pentamidine or trimethoprim-sulfamethoxazole or to hyporeninemic hypoaldosteronism. An abnormality in the redistribution between the intracellular and extracellular compartments may also play a role. Clinical Findings An elevated K+ concentration interferes with normal neuromuscular function to produce muscle weakness and, rarely, flaccid paralysis; abdominal distention and diarrhea may occur. Electrocardiography is not a sensitive method for detecting hyperkalemia, since nearly half of patients with a serum potassium level greater than 6.5 meq/L will not manifest electrocardiographic changes. Electrocardiographic changes in hyperkalemia include peaked T waves of increased amplitude, widening of the QRS, and biphasic QRS-T complexes. Inhibition of atrial depolarization despite normal conduction through usual pathways may occur. This sinoventricular rhythm resembles a junctional mechanism and occurs because of greater sensitivity of atrial myocytes to hyperkalemia than is the case for ventricular muscle cells. The heart rate may be slow; ventricular fibrillation and cardiac arrest are terminal events. Treatment (Table 21-7) One first confirms that the elevated level of serum K+ is genuine. Potassium concentration can be measured in plasma rather than in serum to avoid leakage of potassium out of cells into the serum of the blood sample in the course of clotting, which may be observed in thrombocytosis. Treatment consists of withholding potassium and giving cation exchange resins by mouth or enema. Sodium polystyrene sulfonate, 40-80 g/d in divided doses, is usually effective. Emergent treatment of hyperkalemia is indicated if cardiac toxicity or muscular paralysis is present or if the hyperkalemia is severe (serum potassium > 6.5-7 meq/L) even in the absence of electrocardiographic changes. Insulin plus 10-50% glucose (510 g of glucose per unit of insulin) may be employed to deposit K+ with glycogen in the liver. Calcium may be given intravenously as an antagonist ion—but not when digoxin toxicity is suspected, since calcium may augment the deleterious effects of digoxin on the heart. Transcellular shifts of potassium can also be mediated by 2-adrenergic stimulation. Albuterol, a nebulized 2 agonist, is effective in decreasing serum potassium in patients on hemodialysis. For such patients, one or two standard doses of nebulized albuterol can reduce serum K+ 0.5-1 meq/L within 30 minutes after administration, and this effect is sustained for at least 2 hours. Albuterol and insulin are probably equally efficacious in lowering potassium in uremic patients, and the hypokalemic effects of coadministration of the two drugs (with glucose) are additive and appear not to constitute a hazard. Sodium bicarbonate can be given intravenously as an emergency measure in severe hyperkalemia; the increase in blood pH results in a shift of K+ into cells. Hemodialysis or peritoneal dialysis may be required to remove K+ in the presence of protracted renal insufficiency. Therapy of the precipitating event proceeds concurrently. Caramelo C et al: Hyperkalemia in patients infected with the human immunodeficiency virus: involvement of a systemic mechanism. Kidney Int 1999;56:198. [PMID: 10411693] Charytan D et al: Indications for hospitalization of patients with hyperkalemia. Arch Intern Med 2000;160:1605. [PMID: 10847253] Perazella MA: Drug-induced hyperkalemia: Old culprits and new offenders. Am J Med 2000;109:307. [PMID: 10996582] Reardon LC et al: Hyperkalemia in outpatients using angiotensinconverting enzyme inhibitors. How much should we worry? Arch Intern Med 1998;158:26. [PMID: 9437375] (Serious hyperkalemia is uncommon in ACE-treated patients under 70 who have normal renal function.) Schepkens H et al: Life-threatening hyperkalemia during combined therapy with angiotensin-converting enzyme inhibitors and spironolactone: An analysis of 25 cases. Am J Med 2001;110:438. [PMID: 11331054] DISORDERS OF CALCIUM CONCENTRATION INTRODUCTION Calcium constitutes about 2% of body weight, but only about 1% of the total body calcium is in solution in body fluid. In the plasma, only 50% of calcium is present as ionized calcium; the remainder forms a complex with protein—mostly albumin (40%)—or anions, including citrate, bicarbonate, and phosphate (10%). The normal total plasma (or serum) calcium concentration is 9-10.3 mg/dL. It is ionized calcium (normal: 4.75.3 mg/dL) which, under physiologic regulation, is necessary for muscle contraction and nerve function. Calcium-sensing protein, a receptor-like protein with the special function of detecting extracellular calcium ion concentrations, has been identified in parathyroid cells and in the kidney. Some diseases (eg, familial hypocalcemia and familial hypocalciuric hypercalcemia) associated with disturbed calcium metabolism are due to functional defects of this protein (Table 21-5). HYPOCALCEMIA Development of hypocalcemia implies insufficient action of PTH or active vitamin D. Important causes of hypocalcemia are listed in Table 21-8. The most common cause of hypocalcemia is renal failure, in which decreased production of active vitamin D3 and hyperphosphatemia both play a role. (See Chapter 22.) Some cases of primary hypoparathyroidism are due to mutation of calcium-sensing protein in which inappropriate suppression of PTH release leads to hypocalcemia. Hypocalcemia in pancreatitis is also a marker for severe disease (see Chapter 15). Clinical Findings A. Symptoms and Signs Hypocalcemia increases excitation of nerve and muscle cells, primarily affecting the neuromuscular and cardiovascular systems. Extensive spasm of skeletal muscle causes cramps and tetany. Laryngospasm with stridor can obstruct the airway. Convulsions can occur as well as paresthesias of lips and extremities and abdominal pain. Chvostek's sign (contraction of the facial muscle in response to tapping the facial nerve anterior to the ear) and Trousseau's sign (carpal spasm occurring after occlusion of the brachial artery with a blood pressure cuff for 3 minutes) are usually readily elicited. Prolongation of the QT interval (due to lengthened ST segment) predisposes to the development of ventricular arrhythmias. In chronic hypoparathyroidism, cataracts and calcification of basal ganglia of the brain may appear. (See Hypoparathyroidism, Chapter 26.) B. Laboratory Findings Serum Ca2+ is low (< 9 mg/dL). The depressed level of serum Ca2+ must be correlated with the simultaneous concentration of serum albumin: When albumin is low, serum Ca2+ concentration is depressed in a ratio of 0.8-1 mg of Ca2+ to 1 g of albumin. Serum phosphate is usually elevated in hypoparathyroidism or end-stage renal failure, whereas it is suppressed in early stage renal failure or vitamin D deficiency. Serum Mg2+ is commonly low, and hypomagnesemia reduces both parathyroid hormone release and tissue responsiveness to parathyroid hormone, causing hypocalcemia. In respiratory alkalosis, total serum calcium is normal but ionized calcium is low. The ECG shows a prolonged QT interval. Treatment* *See also Chapter 26 for discussion of the treatment of hypoparathyroidism. A. Severe, Symptomatic Hypocalcemia In the presence of tetany, arrhythmias, or seizures, calcium gluconate 10% (10-20 mL) administered intravenously over 10-15 minutes is indicated. Because of the short duration of action, calcium infusion is usually required. Ten to 15 milligrams of calcium per kilogram body weight, or six to eight 10-mL vials of 10% calcium gluconate (558-744 mg of calcium), is added to 1 L of D5W and infused over 4-6 hours. By monitoring the serum calcium level frequently (every 4-6 hours), the infusion rate is adjusted to maintain the serum calcium level at 7-8.5 mg/dL. B. Asymptomatic Hypocalcemia Oral calcium (1-2 g) and vitamin D preparations (Table 26-10) are used. Calcium carbonate is well tolerated and less expensive than many other calcium tablets. The low serum Ca2+ associated with low serum albumin concentration does not require replacement therapy. If serum Mg2+ is low, therapy must include replacement of magnesium, which by itself usually will correct hypocalcemia. Bushinsky DA et al: Calcium. Lancet 1998;352:306. [PMID: 9690425] (Pathophysiology, differential diagnosis, and treatment of hypocalcemia and hypercalcemia.) HYPERCALCEMIA Important causes of hypercalcemia are listed in Table 21-9. Primary hyperparathyroidism is the most common cause of hypercalcemia in ambulatory patients. Tumor production of PTH-related protein is the commonest paraneoplastic endocrine syndrome, accounting for most cases of hypercalcemia in inpatients. The neoplasm is clinically apparent in nearly all cases when the hypercalcemia is detected, and the prognosis is poor. Chronic hypercalcemia (over 6 months) or some manifestation such as nephrolithiasis suggests a benign cause. Primary hyperparathyroidism and malignancy account for 90% of all cases of hypercalcemia. Milk-alkali syndrome, which had become rare with the advent of nonabsorbable antacid therapy for ulcer disease, should be considered as a cause of hypercalcemia with the current popularity of calcium ingestion for osteoporosis prevention. In the milk-alkali syndrome, massive calcium and vitamin D ingestion can cause hypercalcemic nephropathy. Because of the decreased GFR, retention of the alkali in the calcium antacid occurs and causes metabolic alkalosis, which can be worsened by the vomiting associated with this disorder. Hypercalcemia also causes nephrogenic diabetes insipidus. Development of polyuria is mediated through activation of calcium-sensing receptors in collecting ducts. Volume depletion further worsens hypercalcemia. Clinical Findings A. Symptoms and Signs Symptoms of hypercalcemia irrespective of cause are constipation and polyuria. The focus of the history and physical examination should be on the duration of the process of hypercalcemia and evidence for a neoplasm. Symptoms usually occur if the serum calcium is above 12 mg/dL and tend to be more severe if hypercalcemia develops acutely. Interestingly, polyuria is absent in hypercalcemia in familial hypocalciuric hypercalcemia, in which function of calcium-sensing protein is altered in the kidney as well. A variety of neurologic symptoms also are observed. Stupor, coma, and azotemia may develop in severe hypercalcemia. Ventricular extrasystoles and idioventricular rhythm occur and can be accentuated by digitalis. B. Laboratory Findings A significant elevation of serum Ca2+ is seen; the level must be interpreted in relation to the serum albumin level (see Hypocalcemia). The highest serum calcium levels (> 15 mg/dL) generally occur in malignancy. More than 200 mg/d of urinary calcium excretion suggests hypercalciuria; less than 100 mg/d, hypocalciuria. Hypercalciuric patients—such as those with malignancy or those receiving oral active vitamin D therapy—may easily develop hypercalcemia in case of volume depletion. Serum phosphate may or may not be low, depending on the cause. The ECG shows a shortened QT interval. Measurements of PTH and PTHrelated protein (PTHrP) help distinguish between malignancy-associated hypercalcemia (elevated PTHrP) and hyperparathyroidism (elevated PTH). Treatment Until the primary disease can be brought under control, renal excretion of calcium with resultant decrease in serum Ca2+ concentration is promoted. Excretion of Na+ is accompanied by excretion of Ca2+. The tendency in hypercalcemia is toward volume depletion from nephrogenic diabetes insipidus. Therefore, establishing euvolemia and inducing natriuresis by giving saline with furosemide is the emergency treatment of choice. In dehydrated patients with normal cardiac and renal function, 0.45% saline or 0.9% saline can be given rapidly (250-500 mL/h). Intravenous furosemide (20-40 mg every 2 hours) prevents volume overload and enhances Ca2+ excretion. Thiazides can actually worsen hypercalcemia (as can furosemide if inadequate saline is given). In the treatment of hypercalcemia of malignancy, bisphosphonates are safe and effective in more than 95% of patients and are the mainstay of treatment. In emergency cases, dialysis with low or no calcium dialysate may be needed. See Chapter 40 for a discussion of the treatment of hypercalcemia of malignancy and Chapter 26 for a discussion of the treatment of hypercalcemia of hyperparathyroidism. Body JJ: Current and future directions in medical therapy: hypercalcemia. Cancer 2000;88(12 Suppl):3054. [PMID: 10898351] Strewler GL: The physiology of a parathyroid hormone-related protein. N Engl J Med 2000;342:177. [PMID: 10639544] (An updated review on the action of PTHrP.) Ziegler R: Hypercalcemic crisis. J Am Soc Nephrol 2001;17:S39. [PMID: 11251025] DISORDERS OF PHOSPHORUS CONCENTRATION INTRODUCTION Phosphate is important in the body as a constituent of bone and is crucial in cellular energy transfer and metabolism. Phosphate occupies 1% of body weight, and 80% of it is combined with calcium in bones and teeth. Only 10% is incorporated into a variety of organic compounds, and 10% is combined with proteins, lipids, carbohydrates, and other compounds in muscle and blood. Organic phosphate is the principal intracellular anion. In plasma, phosphate is mainly present as inorganic phosphate, and this fraction is very small (< 0.2% of total phosphate). However, body phosphate metabolism is regulated through plasma inorganic phosphate. Important determinants of plasma inorganic phosphate concentration are its intestinal absorption, renal excretion, and shift between the intracellular and extracellular spaces. Intestinal absorption of phosphate is facilitated by active vitamin D. Parathyroid hormone stimulates phosphate release from bone and suppresses proximal tubular reabsorption of phosphate, leading to hypophosphatemia and to bone phosphate store depletion if hypersecretion continues. Renal proximal tubular reabsorption of phosphate is decreased by volume expansion, glucocorticoid administration, and proximal tubular dysfunction, such as occurs in Fanconi's syndrome due to myeloma or other diseases. Growth hormone, on the other hand, augments proximal tubular reabsorption of phosphate. Recently, additional phosphaturic hormones such as fibroblast growth factor 23 have been identified. Cellular phosphate uptake is stimulated by various factors and conditions, including alkalemia, insulin, epinephrine, feeding, hungry bone syndrome, and accelerated cell proliferation. Phosphorus metabolism and homeostasis are intimately related to calcium metabolism. See sections on metabolic bone disease in Chapter 26. HYPOPHOSPHATEMIA Hypophosphatemia may occur in the presence of normal phosphate stores. Serious depletion of body phosphate stores may exist with low, normal, or high concentrations of phosphorus in serum. Leading causes of hypophosphatemia are listed in Table 21-10. In the presence of severe hypophosphatemia (1 mg/dL or less), affinity of hemoglobin for oxygen is increased through a decrease in the erythrocyte 2,3-diphosphoglycerate concentration. This impairs tissue oxygenation and thus cell metabolism, which underlies the effects of hypophosphatemia such as muscle weakness or even rhabdomyolysis. Severe hypophosphatemia is common and multifactorial in alcoholic patients. In acute alcohol withdrawal, increased plasma insulin and epinephrine along with respiratory alkalosis promote intracellular shift of phosphate. Vomiting, diarrhea, and poor dietary intake contribute to hypophosphatemia. Chronic alcohol use results in a decrease in the renal threshold of phosphate excretion. This renal tubular dysfunction reverses after a month of abstinence. Patients with chronic obstructive pulmonary disease and asthma commonly have hypophosphatemia, attributed to xanthine derivatives causing shifts of phosphate intracellularly and the phosphaturic effects of beta-adrenergic agonists, loop diuretics, xanthine derivatives, and corticosteroids. Hypophosphatemia is a potent stimulator of 1-hydroxylation of vitamin D in the kidney to form active vitamin D. However, in oncogenic osteomalacia, which accompanies various mesenchymal tumors, activation of vitamin D is suppressed in spite of hypophosphatemia. This suppression may be due to overproduction of phosphatonin, a poorly understood factor that stimulates phosphaturia. Clinical Findings A. Symptoms and Signs Acute, severe hypophosphatemia (0.1-0.2 mg/dL) can lead to acute hemolytic anemia with increased erythrocyte fragility, increased susceptibility to infection from impaired chemotaxis of leukocytes, and platelet dysfunction with petechial hemorrhages. Rhabdomyolysis, encephalopathy (irritability, confusion, dysarthria, seizures, and coma), and heart failure are uncommon but serious manifestations. Chronic severe depletion may be manifested by anorexia, pain in muscles and bones, and fractures. B. Laboratory Findings In addition to hypophosphatemia, evidence of anemia due to hemolysis may be present (eg, elevated serum lactate dehydrogenase). Rhabdomyolysis results in elevated serum creatine kinase (which contains mostly the MM fraction but also some MB fraction) and, in many cases, myoglobin in the urine. Other values vary according to the cause. Renal glycosuria and hypouricemia together with hypophosphatemia indicate Fanconi's syndrome. In chronic depletion, radiographs and biopsies of bones show changes resembling those of osteomalacia. Treatment Treatment is best directed toward prophylaxis by including phosphate in repletion and maintenance fluids. A rapid decline in calcium levels can occur with parenteral administration of phosphate; therefore, when possible, oral replacement of phosphate is preferable. For parenteral alimentation, 620 mg (20 mmol) of phosphorus is required for every 1000 nonprotein kcal to maintain phosphate balance and to ensure anabolic function. A daily ration for prolonged parenteral fluid maintenance is 6201240 mg (20-40 mmol) of phosphorus. For asymptomatic hypophosphatemia (serum phosphorus 0.7-1 mg/dL), an infusion should provide 279-310 mg (9-10 mmol)/12 h until the serum phosphorus exceeds 1 mg/dL. Since the response to phosphate supplementation is not predictable, frequent monitoring of plasma and urine phosphate is necessary. A magnesium deficit often coexists and should be treated simultaneously. In administering phosphate-containing solutions, serum creatinine and calcium must be monitored to guard against hypocalcemia. For oral use, phosphate salts are available in skim milk (approximately 1 g [33 mmol]/L). Tablets or capsules of mixtures of sodium and potassium phosphate may be given to provide 0.5-1 g (18-32 mmol) per day. Contraindications to therapy with phosphate salts include hypoparathyroidism, renal insufficiency, tissue damage and necrosis, and hypercalcemia. When hyperglycemia due to any cause is treated, phosphate accompanies glucose into cells, and hypophosphatemia may ensue. Barak V et al: Prevalence of hypophosphatemia in sepsis and infection: The role of cytokines. Am J Med 1998;104:40. [PMID: 9528718] (Hypophosphatemia observed in sepsis may be related to overproduction of cytokines.) Miller DW et al: Hypophosphatemia in the emergency department therapeutics. Am J Emerg Med 2000;18:457. [PMID: 10919539] Subramian R et al: Severe hypophosphatemia: Pathophysiologic implications, clinical presentations, and treatment. Medicine 2000;79:1. [PMID: 10670405] HYPERPHOSPHATEMIA Causes of hyperphosphatemia are given in Table 21-11. Growing children normally have serum phosphate levels higher than those of adults. Clinical Findings A. Symptoms and Signs The clinical manifestations are those of the underlying disorders (eg, chronic renal failure, hypoparathyroidism). Hyperphosphatemia in chronic renal failure leads to secondary hyperparathyroidism and renal osteodystrophy. B. Laboratory Findings In addition to elevated phosphate, other blood chemistry values are those characteristic of the underlying disease. Treatment Treatment is that of the underlying disease and of associated hypocalcemia if present. In acute and chronic renal failure, dialysis will reduce serum phosphate. Absorption of phosphate can be reduced by administration of calcium carbonate, 0.5-1.5 g three times daily with meals (500 mg tablets). This approach is preferred to the traditional use of aluminum hydroxide because of concerns about aluminum toxicity. Another phosphate binder is sevelamer hydrochloride. Since this agent does not contain calcium or aluminum, it may be especially useful for patients with hypercalcemia or uremia. Levin NW et al: Consequences of hyperphosphatemia and elevated levels of the calcium-phosphorus product in dialysis patients. Curr Opin Nephrol Hypertens 2001;10:563. [PMID: 11496047] Slatopolsky EA et al: Renagel, a nonabsorbed calcium- and aluminumfree phosphate binder, lowers serum phosphorus and parathyroid hormone. Kidney Int 1999;55:299. [PMID: 9893140] Weisinger JR et al: Magnesium and phosphorus. Lancet 1998; 352:391. [PMID: 9717944] (Review of the significance of abnormal levels of magnesium and phosphorus.) DISORDERS OF MAGNESIUM CONCENTRATION INTRODUCTION About 50% of total body magnesium exists in the insoluble state in bone. Only 5% is present as extracellular cation; the remaining 45% is contained in cells as intracellular cation. The normal plasma concentration is 1.5-2.5 meq/L, with about one-third bound to protein and two-thirds existing as free cation. Excretion of magnesium ion is via the kidney. Normally, about 3% of magnesium filtered by the glomerulus is excreted in urine. Magnesium is an important activator ion, participating in the function of many enzymes involved in phosphate transfer reactions. Magnesium exerts physiologic effects on the nervous system resembling those of calcium. Magnesium acts directly upon the myoneural junction. Altered concentration of Mg2+ in the plasma usually provokes an associated alteration of Ca2+. Hypermagnesemia suppresses secretion of parathyroid hormone with consequent hypocalcemia. Severe and prolonged magnesium depletion impairs secretion of PTH with consequent hypocalcemia. Hypomagnesemia may impair end-organ response to PTH as well. HYPOMAGNESEMIA Causes of hypomagnesemia are given in Table 21-12. Nearly half of hospitalized patients in whom serum electrolytes are ordered have unrecognized hypomagnesemia. Common causes include use of large volumes of intravenous fluids, diuretics, cisplatin in cancer patients (with concomitant hypokalemia), and administration of nephrotoxic agents such as aminoglycosides and amphotericin B. Clinical Findings A. Symptoms and Signs Common symptoms are weakness, muscle cramps, and tremor. There is marked neuromuscular and central nervous system hyperirritability, with tremors, athetoid movements, jerking, nystagmus, and a positive Babinski response. There may be hypertension, tachycardia, and ventricular arrhythmias. Confusion and disorientation may be prominent features. B. Laboratory Findings Urinary excretion of magnesium exceeding 10-30 mg/d or a fractional excretion more than 2% indicates renal magnesium wasting. In calculating fractional excretion of magnesium, since only 30% is proteinbound, it follows that 70% of circulating magnesium is filtered by the glomerulus. In addition to hypomagnesemia, hypocalcemia and hypokalemia are often present. The ECG shows a prolonged QT interval, due to lengthening of the ST segment. Parathyroid hormone secretion is often suppressed (see Hypocalcemia, above). Treatment Treatment consists of the use of intravenous fluids containing magnesium as chloride or sulfate, 240- 1200 mg/d (10-50 mmol/d) during the period of severe deficit, followed by 120 mg/d (5 mmol/d) for maintenance. Magnesium sulfate may also be given intramuscularly in a dosage of 200800 mg/d (8-33 mmol/d) in four divided doses. Serum levels must be monitored and dosage adjusted to keep the concentration from rising above 2.5 mmol/L. K+ and Ca2+ may be required as well. Magnesium oxide, 250-500 mg by mouth two to four times daily, is useful for repleting stores in those with chronic hypomagnesemia. Hypokalemia and hypocalcemia of hypomagnesemia do not recover without magnesium supplementation. Argus ZS: Hypomagnesemia. J Am Soc Nephrol 1999;10:1616. [PMID: 10405219] Dacey M: Hypomagnesemic disorders. Crit Care Clin 2001;17: 155. [PMID: 11219227] HYPERMAGNESEMIA Magnesium excess is almost always the result of renal insufficiency and the inability to excrete what has been taken in from food or drugs, especially antacids and laxatives. Clinical Findings A. Symptoms and Signs Muscle weakness, decreased deep tendon reflexes, mental obtundation, and confusion are characteristic manifestations. Weakness—even flaccid paralysis—and hypotension are noted. There may be respiratory muscle paralysis or cardiac arrest. B. Laboratory Findings Serum Mg2+ is elevated. In the common setting of renal insufficiency, concentrations of BUN and of serum creatinine, phosphate, and uric acid are elevated; serum K+ may be elevated. Serum Ca2+ is often low. The ECG shows increased PR interval, broadened QRS complexes, and peaked T waves, probably related to associated hyperkalemia. Treatment Treatment is directed toward alleviating renal insufficiency. Calcium acts as an antagonist to Mg2+ and may be given intravenously as calcium chloride, 500 mg or more at a rate of 100 mg (4.5 mmol)/min. Hemodialysis or peritoneal dialysis may be indicated. Weisinger JR et al: Magnesium and phosphorus. Lancet 1998; 352:391. [PMID: 9717944] (Review of the significance of abnormal levels of magnesium and phosphorus.) HYPEROSMOLAR DISORDERS & OSMOLAR GAPS HYPEROSMOLALITY WITH TRANSIENT OR NO SIGNIFICANT SHIFT IN WATER Urea and alcohol are two substances that readily cross cell membranes and can produce hyperosmolality. Because of its permeant nature, urea has little effect on the shift of water across the cell membrane. Alcohol quickly equilibrates between intracellular and extracellular water, adding 22 mosm/L for every 1000 mg/L. This measured hyperosmolality does not produce symptoms by itself because of the equilibrium described, but in any case of stupor or coma in which measured osmolality exceeds that calculated from values of serum Na+ and glucose and BUN, ethanol intoxication should be considered as a possible explanation of the discrepancy (osmolar gap). Toxic alcohol ingestion, particularly methanol or ethylene glycol, also causes an osmolar gap characterized by anion gap metabolic acidosis (Chapter 39). The combination of anion gap metabolic acidosis and an osmolar gap exceeding 10 mosm/kg is not specific for toxic alcohol ingestion. Nearly half of patients with alcoholic ketoacidosis or lactic acidosis have similar findings, caused in part by elevations of endogenous glycerol, acetone, and acetone metabolites. Delaney MF et al: Diabetic ketoacidosis and hyperglycemic hyperosmolar nonketotic syndrome. Endocrinol Metab Clin North Am 2000;29:683. [PMID: 11149157] Glaser OS: Utility of the serum osmol gap in the diagnosis of methanol or ethylene glycol ingestion. Ann Emerg Med 1996;27:343. [PMID: 8599495] (Limitations of the test.) Magee MF et al: Management of decompensated diabetes. Diabetic ketoacidosis and hyperglycemic hyperosmolar syndrome. Crit Care Clin 2001;17:75. [PMID: 11219236] HYPEROSMOLALITY ASSOCIATED WITH SIGNIFICANT SHIFTS IN WATER Increased concentrations of solutes that do not readily enter cells produce a shift of water from the intracellular space to effect a true intracellular dehydration. Sodium and glucose are the solutes commonly involved. In these instances, the hyperosmolality does produce symptoms. Clinical symptoms are mainly referred to the central nervous system. The severity of symptoms depends on the degree of hyperosmolality and rapidity of development. In acute hyperosmolality, symptoms of somnolence and confusion can appear when the osmolality exceeds 320330 mosm/L, and coma, respiratory arrest, and death when it exceeds 340-350 mosm/L. ACID-BASE DISORDERS About 1 meq/L/kg/d of nonvolatile acid (H+) is produced, together with volatile acid (CO2), by whole body metabolism. Body fluid pH, however, remains relatively constant at about 7.40. This constancy of body fluid pH is achieved by removal of CO2 and H+ through lung and kidney, respectively. In order to assess a patient's acid-base status, measurement of arterial pH, partial pressure of carbon dioxide (PCO2), and plasma bicarbonate (HCO3-) is needed. Blood gas analyzers directly measure pH and PCO2, and the HCO3- value is calculated from the Henderson-Hasselbalch equation: The total venous CO2 measurement is a more direct determination of HCO3-. Because of the dissociation characteristics of carbonic acid (H2CO3) at body pH, dissolved CO2 is almost exclusively in the form of HCO3-, and for clinical purposes the total carbon dioxide content is equivalent ( 3 meq/L) to the HCO3- concentration: If precise measurements of oxygenation are not needed or if oxygen saturation obtained from the pulse oximeter is adequate, venous blood gases generally provide useful information for assessment of acid-base balance and can be used interchangeably with arterial blood gases since the arteriovenous differences in pH and PCO2 are small and relatively constant. Venous blood pH is usually 0.03-0.04 units lower than that of arterial blood, and venous blood PCO2 is 7 or 8 mm Hg higher. Calculated HCO3- concentration in venous blood is at most 2 meq/L higher than that of arterial blood. An important exception to the rule of interchangeability between arterial and venous blood gases for determination of acid-base balance is during cardiopulmonary arrest. In this setting, arterial pH may be 7.41 and venous pH 7.15, and arterial blood PCO2 can be 32 mm Hg with a venous blood PCO2 of 74 mm Hg. Types of Acid-Base Disorders There are two types of acid-base disorders: respiratory and metabolic. Primary respiratory disorders affect blood acidity by causing changes in PCO2, and primary metabolic disorders are caused by disturbances in the HCO3- concentration. The primary disturbances are usually accompanied by compensatory changes; however, even though these changes attenuate a pH shift from the normal value (7.40), they do not fully compensate for the primary acid-base disorders even if the disorders are chronic. Therefore, if the pH is less than 7.40, the primary process is acidosis (either respiratory or metabolic). If the pH is higher than 7.40, the primary process is either respiratory or metabolic alkalosis. The presence of one disorder with its appropriate compensatory change is a simple disorder. Mixed Acid-Base Disorders The presence of more than one simple disorder (not compensatory) is a mixed disorder. Double or triple disorders can coexist but not quadruple ones, since simultaneous respiratory acidosis and alkalosis are not possible. Clinicians frequently find it difficult to decide if a mixed disorder is present. One useful scheme is to determine if the degree of compensation for the primary disorder is appropriate (Table 21-13). In respiratory disorders, if the magnitude of compensation in HCO3- level differs from that which is predicted, the patient has a mixed disorder. Therefore, superimposed metabolic acidosis will decrease HCO3- to lower than the predicted level, and a metabolic alkalosis will increase HCO3- over the predicted value. For example, a patient with chronic respiratory acidosis and PCO2 of 60 mm Hg should have a HCO3- of 31 meq/L (assuming that normal HCO3- is 24 meq/L). If the HCO3- is 25 meq/L, a superimposed metabolic acidosis exists, and if the HCO3- is 45 meq/L, there is a superimposed metabolic alkalosis. Using data from Table 21-13, similar calculations can be made for primary metabolic disorders. Furthermore, corrected bicarbonate (cHCO3-), calculated from measured HCO3- plus the increase in anion gap (see below), is useful to assess the superimposed metabolic alkalosis or normal anion gap metabolic acidosis. In increased anion gap acidosis, there must be a mole for mole decrease in HCO3- as anion gap decreases. Therefore, an HCO3- value higher or lower than normal (24 meq/L) indicates the concomitant presence of metabolic alkalosis or normal anion gap acidosis, respectively. STEP-BY-STEP ANALYSIS OF ACID-BASE STATUS Step 1: Determine the primary (or main) disorder--whether it is metabolic or respiratory--from blood pH, HCO3-, and Pco2 values. Step 2: Determine the presence of mixed acid-base disorders by calculating the range of compensatory responses (Table 21-13). Step 3: Calculate the anion gap (Table 21-15). Step 4: Calculate the HCO3- concentration if the anion gap is increased (see above). Step 5: Examine the patient to determine whether the clinical signs are compatible with the acid-base analysis thus obtained. Gluck SL: Acid-base. Lancet 1998;352:474. [PMID: 9708770] (A clinical approach to diagnosis and management.) Laski ME et al: Acid-base disorders in medicine. Dis Mon 1996;42(2):51. [PMID: 8631223] RESPIRATORY ACIDOSIS Respiratory acidosis results from decreased alveolar ventilation and subsequent hypercapnia. Pulmonary as well as nonpulmonary disorders can cause hypoventilation. The clinician must be mindful of readily reversible causes of respiratory acidosis, especially opioid-induced central nervous system depression. Acute respiratory failure is associated with severe acidosis and only a small increase in the plasma bicarbonate. After 6-12 hours, the primary increase in PCO2 evokes a renal compensatory response to generate more HCO3-, which tends to ameliorate the respiratory acidosis. This usually takes several days to complete. Chronic respiratory acidosis is generally seen in patients with underlying lung disease, such as chronic obstructive disease. Urinary excretion of acid in the form of NH4+ and Cl- ions results in the characteristic hypochloremia of chronic respiratory acidosis. When chronic respiratory acidosis is corrected suddenly, especially in patients who receive mechanical ventilation, there is a 2- to 3-day lag in renal bicarbonate excretion, resulting in posthypercapnic metabolic alkalosis. Clinical Findings A. Symptoms and Signs With acute onset, there is somnolence and confusion, and myoclonus with asterixis may be seen. Coma from CO2 narcosis ensues. Severe hypercapnia increases cerebral blood flow and cerebrospinal fluid pressure. Signs of increased intracranial pressure (papilledema, pseudotumor cerebri) may be seen. B. Laboratory Findings Arterial pH is low, and PCO2 is increased. Serum HCO3- is elevated, but not enough to completely compensate for the hypercapnia. If the disorder is chronic, hypochloremia is seen. Treatment Since drug overdose is an important reversible cause of acute respiratory acidosis, administration of naloxone, 0.04-2 mg intravenously (see Chapter 39) is given to all such patients if no obvious cause for respiratory depression is present. In all forms of respiratory acidosis, treatment is directed at the underlying disorder to improve ventilation. RESPIRATORY ALKALOSIS Respiratory alkalosis, or hypocapnia, occurs when hyperventilation reduces the PCO2, which increases the pH. The most common cause of respiratory alkalosis is hyperventilation syndrome (Table 21-14), but bacterial septicemia and cirrhosis are other common causes. Symptoms in acute respiratory alkalosis are related to decreased cerebral blood flow induced by the disorder. Pregnancy is another cause of chronic respiratory alkalosis, probably from progesterone stimulation of the respiratory center, with an average PCO2 of 30 mm Hg. Determination of appropriate compensatory changes in the HCO3- is useful to sort out the presence of an associated metabolic disorder (see above under Mixed Acid-Base Disorders). As in respiratory acidosis, the changes in HCO3- values are greater if the respiratory alkalosis is chronic (Table 21-13). While serum HCO3- is frequently below 15 meq/L in metabolic acidosis, it is unusual to see such a low level in respiratory alkalosis, and its presence would imply a superimposed (noncompensatory) metabolic acidosis. Clinical Findings A. Symptoms and Signs In acute cases (hyperventilation), there is light-headedness, anxiety, paresthesias, numbness about the mouth, and a tingling sensation in the hands and feet. Tetany occurs in more severe alkalosis from a fall in ionized calcium. In chronic cases, findings are those of the responsible condition. B. Laboratory Findings Arterial blood pH is elevated, and PCO2 is low. Serum bicarbonate is decreased in chronic respiratory alkalosis. Treatment Treatment is directed toward the underlying cause. In acute hyperventilation syndrome from anxiety, rebreathing into a paper bag will increase the PCO2. The processes are usually self-limited since muscle weakness caused by hyperventilation-induced alkalemia will suppress ventilation. Sedation may be necessary if the process persists. Rapid correction of chronic respiratory alkalosis may result in metabolic acidosis as PCO2 is increased in the setting of previous compensatory decrease in HCO3-. Laffey JG et al: Hypocapnia. N Engl J Med 2002;347:43. [PMID: 12097540] METABOLIC ACIDOSIS The hallmark of metabolic acidosis is decreased HCO3-, seen also in respiratory alkalosis (see above), but the pH distinguishes between the two disorders. Calculation of the anion gap is useful in determining the cause of the metabolic acidosis (Table 21-15). The anion gap represents the difference between readily measured anions and cations. In plasma, The major unmeasured cations are calcium (1 meq/L), magnesium (2 meq/L), gamma globulins, and potassium (4 meq/L). The major unmeasured anions are negatively charged albumin (2 meq/L per g/dL), phosphate (2 meq/L), sulfate (1 meq/L), lactate (1-2 meq/L), and other organic anions (3-4 meq/L). Traditionally, the normal anion gap has been 12 4 meq/L. With the current generation of autoanalyzers, the reference range may be lower (6 1 meq/L), primarily from an increase in Clvalues. Despite its usefulness, the serum anion gap can be misleading. Non-acid-base disorders that may contribute to an error in anion gap interpretation include hypoalbuminemia (see below), antibiotic administration (eg, carbenicillin is an unmeasured anion; polymyxin is an unmeasured cation), hypernatremia, or hyponatremia. Decreased Anion Gap A decreased anion gap can occur because of a reduction in unmeasured anions or an increase in unmeasured cations. A. Decreased Unmeasured Anions If the sodium concentration remains normal but HCO3- and Cl- increase, the anion gap will decrease. This is seen when there are decreased unmeasured anions, especially in hypoalbuminemia. For every 1 g/dL decline in serum albumin, a 2 meq/L decrease in anion gap will occur. B. Increased Unmeasured Cations If the sodium concentration falls because of addition of unmeasured cations but HCO3- and Cl- remain unchanged, the anion gap will decrease. This is seen in (1) severe hypercalcemia, hypermagnesemia, or hyperkalemia; (2) IgG myeloma, where the immunoglobulin is cationic in 70% of cases; and (3) lithium toxicity. Jurado RL et al: Low anion gap. South Med J 1998;91:624. [PMID: 9671832] (Differential diagnosis of the low anion gap and the importance of this clue in the diagnosis of occult myeloma or intoxications.) Increased Anion Gap Acidosis (Increased Unmeasured Anions) The hallmark of this disorder is that metabolic acidosis (thus low HCO3-) is associated with normal serum Cl-, so that the anion gap increases. Normochloremic metabolic acidosis generally results from addition to the blood of nonchloride acids such as lactate, acetoacetate, hydroxybutyrate, and exogenous toxins. An exception is uremia, with underexcretion of organic acids and anions. A. Lactic Acidosis Lactic acid is formed from pyruvate in anaerobic glycolysis. Therefore, most of the lactate is produced in tissues with high rates of glycolysis, such as gut (responsible for over 50% of lactate production), skeletal muscle, brain, skin, and erythrocytes. Normally, lactate levels remain low (1 meq/L) because of metabolism of lactate principally by the liver through gluconeogenesis or oxidation via the Krebs cycle. Furthermore, the kidneys metabolize about 30% of lactate. In lactic acidosis, lactate levels are at least 4-5 meq/L but commonly 1030 meq/L. The mortality rate exceeds 50%. There are two basic types of lactic acidosis, both associated with increased lactate production and decreased lactate utilization. Type A is characterized by hypoxia or decreased tissue perfusion, whereas in type B there is no clinical evidence of hypoxia. Type A (hypoxic) lactic acidosis is the more common type, resulting from poor tissue perfusion; cardiogenic, septic, or hemorrhagic shock; and carbon monoxide or cyanide poisoning. These conditions not only cause lactic acid production to increase peripherally but, more importantly, hepatic metabolism of lactate to decrease as liver perfusion declines. In addition, severe acidosis impairs the ability of the liver to extract the perfused lactate. Type B lactic acidosis may be due to metabolic causes, such as diabetes, ketoacidosis, liver disease, renal failure, infection, leukemia, or lymphoma; or may occur as a result of toxicity from ethanol, methanol, salicylates, or isoniazid. AIDS without AIDS-related lymphoma is associated with type B lactic acidosis. Idiopathic lactic acidosis, usually in debilitated patients, has an extremely high mortality rate. (For treatment of lactic acidosis, see below and Chapter 27.) B. Diabetic Ketoacidosis This metabolic abnormality is characterized by hyperglycemia and metabolic acidosis (pH < 7.25 or plasma bicarbonate < 16 meq/L). Anion gap metabolic acidosis is the acid-base disturbance generally ascribed to diabetic ketoacidosis: where B- is -hydroxybutyrate or acetoacetate. The anion gap should be calculated from the serum electrolytes as measured, since correction of the serum sodium for the dilutional effect of hyperglycemia will incorrectly exaggerate the anion gap. The increased anion gap is due to hyperketonemia (acetoacetate and hydroxybutyrate) and at times to an increase in serum lactate secondary to reduced tissue perfusion and increased anaerobic metabolism. If a rise in anion gap from normal is equal to a fall in HCO3-, a diagnosis of simple metabolic acidosis can be made. However, the presence of concurrent metabolic alkalosis or normal anion gap metabolic acidosis is suggested if the value of the measured HCO3- plus the increase in anion gap (cHCO3-) is higher or lower than the normal value for HCO3-, respectively. During the recovery phase of diabetic ketoacidosis, anion gap acidosis can be transformed into hyperchloremic non-anion gap acidosis. The mechanism for this is as follows: As GFR increases from NaCl therapy of diabetic ketoacidosis, the retention of Cl- causes a mild decrease in the anion gap from dilution. More importantly, the increased GFR causes the urinary excretion of ketone salts (NaB), which are formed as bicarbonate is consumed: The kidney reabsorbs ketone anions poorly but can compensate for the loss of anions (and therefore Na+) by increasing the reabsorption of Cl-. Conversely, even on presentation, patients with diabetic ketoacidosis and normal renal perfusion may have marked ketonuria, severe metabolic acidosis, and only a mildly increased anion gap. Again, the variable relationship between the rise in the anion gap and the fall in the HCO3can occur with the urinary loss of Na+ or K+ salts of -hydroxybutyrate, which will lower the anion gap without altering the H+ excretion or the severity of the acidosis. Since Ketostix reacts to acetoacetate, less to acetone, and not at all to the predominant keto acid, -hydroxybutyrate, the test may become more positive even as the patient improves owing to the metabolism of hydroxybutyrate. Thus, the patient's clinical status and the reduction of the anion gap are better markers of improvement than monitoring the serum acetone test. Conversely, in the presence of concomitant lactic acidosis, a shift in the redox state can increase -hydroxybutyrate and decrease the readily detectable acetoacetate, thus lowering the nitroprusside reaction. C. Alcoholic Ketoacidosis This is a common disorder of chronically malnourished patients who consume large quantities of alcohol daily. Most of these patients have mixed acid-base disorders (10% have a triple acid-base disorder). While decreased HCO3- is usual, half the patients may have normal or alkalemic pH. The three types of metabolic acidosis seen in alcoholic ketoacidosis are the following: (1) Ketoacidosis due to -hydroxybutyrate and acetoacetate excess. (2) Lactic acidosis: Alcohol metabolism increases the NADH:NAD ratio, causing increased production and decreased utilization of lactate. Accompanying thiamin deficiency, which inhibits pyruvate carboxylase, further enhances lactic acid production in many cases. Moderate to severe elevations of lactate (> 6 mmol/L) are seen with concomitant disorders such as sepsis, pancreatitis, or hypoglycemia. (3) Hyperchloremic acidosis from bicarbonate loss in the urine associated with ketonuria (see above). Metabolic alkalosis occurs from volume contraction and vomiting. Respiratory alkalosis results from alcohol withdrawal, pain, or associated disorders such as sepsis or liver disease. Half of the patients have either hypoglycemia or hyperglycemia. When serum glucose levels are greater than 250 mg/dL, the distinction from diabetic ketoacidosis is difficult. The diagnosis of alcoholic ketoacidosis is supported by absence of a diabetic history and by no evidence of glucose intolerance after initial therapy. D. Toxins (See also Chapter 39.) Multiple toxins and drugs can increase the anion gap by increasing endogenous acid production. Examples include methanol (metabolized to formic acid), ethylene glycol (glycolic and oxalic acid), and salicylates (salicylic acid and lactic acid), which can cause a mixed disorder of metabolic acidosis with respiratory alkalosis. E. Uremic Acidosis At glomerular filtration rates below 20 mL/min, the inability to excrete H+ with retention of acid anions such as PO43- and SO42- result in an increased anion gap acidosis, which rarely is severe. The unmeasured anions "replace" HCO3- (which is consumed as a buffer). Hyperchloremic normal anion gap acidosis may be seen in milder cases of renal insufficiency. Normal Anion Gap Acidosis (Table 21-16) The hallmark of this disorder is that the low HCO3- of metabolic acidosis is associated with hyperchloremia, so that the anion gap remains normal. The most common causes are gastrointestinal HCO3- loss and defects in renal acidification (renal tubular acidoses). The urinary anion gap can differentiate between these two common causes (see below). A. Gastrointestinal HCO3- Loss Bicarbonate is secreted in multiple areas in the gastrointestinal tract. Small bowel and pancreatic secretions contain large amounts of HCO3-. Therefore, massive diarrhea or pancreatic drainage can result in HCO3loss because of increased HCO3- secretion and decreased absorption. Hyperchloremia occurs because the ileum and colon secrete HCO3- in a one-to-one exchange for Cl- by countertransport. The resultant volume contraction causes further increased Cl- retention by the kidney in the setting of decreased anion, HCO3-. Patients with ureterosigmoidostomies can develop hyperchloremic metabolic acidosis because the colon secretes HCO3- in the urine in exchange for Cl-. B. Renal Tubular Acidosis In renal tubular acidosis, the defect is either inability to excrete H+ (inadequate generation of new HCO3-) or inappropriate reabsorption of HCO3-. Four discrete types can be differentiated by the clinical setting, urinary pH, urinary anion gap (see below), and serum K+ level. Since mutations of Cl-/HCO3- exchanger and H+-ATPase genes have been recently demonstrated in hereditary distal renal tubular acidosis, further classification depending upon molecular mechanisms may soon become possible. 1. Classic distal renal tubular acidosis (type I)—This disorder is characterized by hypokalemic hyperchloremic metabolic acidosis and is due to selective deficiency in H+ secretion in the distal nephron. Despite acidosis, urinary pH cannot be acidified and is above 5.5, which retards the binding of H+ to phosphate (H+ + HPO42- H2PO4), and thus inhibits titratable acid excretion. Furthermore, urinary excretion of NH4+Cl- is decreased, and the urinary anion gap is positive (see below). Enhanced K+ excretion occurs probably because there is less competition from H+ in the distal nephron transport system. Furthermore, as a response to renal salt wasting, hyperaldosteronism occurs. Nephrocalcinosis and nephrolithiasis frequently accompany this disorder since chronic acidosis decreases tubular calcium reabsorption. The hypercalciuria, alkaline urine, and lowered level of urinary citrate cause calcium phosphate stones and nephrocalcinosis. 2. Proximal renal tubular acidosis (type II)—Proximal renal tubular acidosis is a hypokalemic hyperchloremic metabolic acidosis due to a selective defect in the proximal tubule's ability to adequately reabsorb filtered HCO3-. Carbonic anhydrase inhibitors (acetazolamide) can cause proximal renal tubular acidosis. About 90% of filtered HCO3- is absorbed by the proximal tubule. The distal nephron has a limited ability to absorb HCO3- but becomes overwhelmed and does not function adequately when there is increased delivery. Eventually, distal delivery of filtered HCO3declines because the plasma HCO3- level has dropped as a result of progressive urinary HCO3- wastage. When the plasma HCO3- level drops to 15-18 meq/L, delivery of HCO3- drops to the point where the distal nephron is no longer overwhelmed and can regain function. At that point, bicarbonaturia disappears, and urinary pH can be acidic. Thiazide-induced volume contraction can be used to enhance proximal HCO3- reabsorption, leading to the decrease in distal HCO3- delivery and improvement of bicarbonaturia and renal acidification. The increased delivery of HCO3- to the distal nephron also increases K+ secretion, and hypokalemia results if a patient is loaded with excess HCO3- and K+ is not adequately supplemented. Proximal renal tubular acidosis often exists with other defects of absorption in the proximal tubule, resulting in glucosuria, aminoaciduria, phosphaturia, and uricaciduria. Causes include multiple myeloma with Fanconi's syndrome and nephrotoxic drugs. 3. Renal tubular acidosis of glomerular insufficiency (type III)— When GFR decreases to 20-30 mL/min, ability to generate adequate NH3 is impaired, with subsequent decreased NH4+Cl- excretion. A normokalemic, hyperchloremic metabolic acidosis ensues. Further reduction in GFR results in increased anion gap acidosis of uremia (see above). 4. Hyporeninemic hypoaldosteronemic renal tubular acidosis (type IV)—Type IV is the only type characterized by hyperkalemic, hyperchloremic acidosis. The defect is aldosterone deficiency or antagonism, which impairs distal nephron Na+ reabsorption and K+ and H+ excretion. Renal salt wasting is frequently present. Relative hypoaldosteronism from hyporeninemia is most commonly found in diabetic nephropathy, tubulointerstitial renal diseases, hypertensive nephrosclerosis, and AIDS. In patients with these disorders, caution must be taken when using drugs that can exacerbate the hyperkalemia, such as angiotensin-converting enzyme inhibitors (which will further reduce aldosterone levels), aldosterone receptor blockers such as spironolactone, and NSAIDs. C. Dilutional Acidosis Rapid dilution of plasma volume by 0.9% NaCl may cause a mild hyperchloremic acidosis. D. Recovery From Diabetic Ketoacidosis See above. E. Posthypocapnia In prolonged respiratory alkalosis, HCO3- decreases and Cl- increases from decreased renal NH4+Cl- excretion. If the respiratory alkalosis is corrected quickly, PCO2 will increase acutely but HCO3- will remain low until the kidneys can generate new HCO3-, which generally takes several days. In the meantime, the increased PCO2 with low HCO3- causes metabolic acidosis. F. Hyperalimentation Hyperalimentation fluids may contain amino acid solutions that acidify when metabolized, such as arginine hydrochloride and lysine hydrochloride. Urinary Anion Gap to Assess Hyperchloremic Metabolic Acidosis Increased renal NH4+Cl- excretion to enhance H+ removal is a normal physiologic response to metabolic acidosis. NH3 reacts with H+ to form NH4+, which is accompanied by the anion Cl- for excretion. The normal daily urinary excretion of NH4Cl of about 30 meq can be increased up to 200 meq in response to acid load. Urinary anion gap from a random urine sample ([Na+ + K+] - Cl-) reflects the ability of the kidney to excrete NH4Cl as in the following equation: where 80 is the average value for the difference in the urinary anions and cations other than Na+, K+, NH3+, and Cl-. Therefore, urinary anion gap is equal to (80 - NH3+), and thus aids in the distinction between gastrointestinal and renal causes of hyperchloremic acidosis. If the cause of the metabolic acidosis is gastrointestinal HCO3loss (diarrhea), renal acidification ability remains normal and NH4Cl excretion increases in response to the acidosis. The urinary anion gap is negative (eg, -30 meq/L). If the cause is distal renal tubular acidosis, the urinary anion gap is positive (eg, +25 meq/L), since the basic lesion in the disorder is inability of the kidney to excrete H+ and thus inability to increase NH4Cl excretion. Urinary pH may not as readily differentiate between the two causes. Despite acidosis, if volume depletion from diarrhea causes inadequate Na+ delivery to the distal nephron and therefore decreased exchange with H+, urinary pH may not be lower than 5.3. In the presence of this relatively high urine pH, however, H+ excretion continues due to buffering of NH3 to NH4+, since the pK of this reaction is as high as 9.1. Potassium depletion, which can accompany diarrhea (and surreptitious laxative abuse), may also impair renal acidification. Thus, when volume depletion is present, the urinary anion gap is a better measurement of ability to acidify the urine than urinary pH. Clinical Findings A. Symptoms and Signs Symptoms of metabolic acidosis are mainly those of the underlying disorder. Compensatory hyperventilation is an important clinical sign and may be misinterpreted as a primary respiratory disorder; when severe, Kussmaul respirations (deep, regular, sighing respirations) are seen. B. Laboratory Findings Blood pH, serum HCO3-, and PCO2 are decreased. Anion gap may be normal (hyperchloremic) or increased (normochloremic). Hyperkalemia may be seen (see above). Treatment A. Increased Anion Gap Acidosis Treatment is aimed at the underlying disorder, such as insulin and fluid therapy for diabetes and appropriate volume resuscitation to restore tissue perfusion. The metabolism of lactate will produce HCO3- and increase pH. The use of supplemental HCO3- is indicated for treatment of hyperkalemia (Table 21-7) and some forms of normal anion gap acidosis but has been controversial for treatment of increased anion gap metabolic acidosis. Administration of large amounts of HCO3- may have deleterious effects, including hypernatremia and hyperosmolality. Furthermore, intracellular pH may decrease because administered HCO3- is converted to CO2, which easily diffuses into cells. There, it combines with water to create additional hydrogen ions and worsening of intracellular acidosis. Theoretically, this could impair cellular function, but the clinical significance of this phenomenon is uncertain. In addition, alkali administration is known to stimulate phosphofructokinase activity, thus exacerbating lactic acidosis via enhanced lactate production. Ketogenesis is also augmented by alkali therapy. In salicylate intoxication, however, alkali therapy must be started unless blood pH is already alkalinized by respiratory alkalosis, since the increment in pH converts salicylate to more impermeable salicylic acid and thus prevents central nervous system damage. In alcoholic ketoacidosis, thiamine should be given together with glucose to avoid the development of Wernicke's encephalopathy. The amount of HCO3- deficit can be calculated as follows: Half of the calculated deficit should be administered within the first 3-4 hours to avoid overcorrection and volume overload. In methanol intoxication, ethanol has been used as a competitive substrate for alcohol dehydrogenase, which metabolizes methanol to formaldehyde. Recently, direct inhibition of alcohol dehydrogenase by fomepizole has been reported and may be used in near future. B. Normal Anion Gap Acidosis (Table 21-16.) In distal renal tubular acidosis, supplementation of bicarbonate is necessary since acid accumulates systemically in this disorder. However, in proximal renal tubular acidosis, correction of low serum bicarbonate is sometimes hazardous and unnecessary except in severe cases. If the blood bicarbonate concentration is elevated in response to its administration, and the concentration in the glomerular filtrate exceeds the capacity of the proximal tubule to reabsorb it, a large quantity of bicarbonate is excreted into the urine accompanied by potassium, exacerbating hypokalemia. Thus, potassium should also be given when bicarbonate therapy is indicated in proximal renal tubular acidosis. Adrogue HJ et al: Management of life threatening acid base disorders. (Part 1.) N Engl J Med 1998;338:26. [PMID: 9414329] Adrogue HJ et al: Management of life-threatening acid-base disorders. Second of two parts. N Engl J Med 1998;338:107. [PMID: 9420343] Batlle D et al: Hereditary distal renal tubular acidosis: new understandings. Annu Rev Med 2001;52:471. [PMID: 11160790] Brent J et al: Fomepizole for the treatment of methanol poisoning. N Engl J Med 2001;344:424. [PMID: 11172179] Galla JH: Metabolic alkalosis. J Am Soc Nephrol 2000;11:369. [PMID: 10665945] Hood VL et al: Protection of acid-base balance by pH regulation of acid production. N Engl J Med 1998;339:819. [PMID: 9738091] (Systemic pH regulates acid production in a negative feedback manner.) Luft FC: Lactic acidosis update for critical care clinicians. J Am Soc Nephrol 2001;12:S15. [PMID: 11251027] Salem MM et al: Gaps in the anion gap. Arch Intern Med 1992; 152:1625. [PMID: 1497396] Smulders YM et al: Renal tubular acidosis: Pathophysiology and diagnosis. Arch Intern Med 1996;156:1629. [PMID: 8694660] METABOLIC ALKALOSIS Classification Metabolic alkalosis is characterized by high HCO3-. The high HCO3- is seen also in chronic respiratory acidosis (see above), but pH differentiates the two disorders. It is useful to classify the causes of metabolic alkalosis into two groups based on "saline responsiveness" or urinary Cl-, which are markers for volume status (Table 21-17). Saline-responsive metabolic alkalosis is a sign of extracellular volume contraction, and salineunresponsive alkalosis implies a volume-expanded state. It is rare for a compensatory increase in PCO2 to exceed 55 mm Hg. A higher value implies a superimposed respiratory acidosis. A. Saline-Responsive Metabolic Alkalosis Saline-responsive metabolic alkalosis is by far the more common disorder. It is characterized by normotensive extracellular volume contraction and hypokalemia. Less frequently, hypotension or orthostatic hypotension may be seen. In vomiting or nasogastric suction, for example, loss of acid (HCl) initiates the alkalosis, but volume contraction from loss of Cl- sustains the alkalosis because the decline in GFR causes avid renal Na+ and HCO3- reabsorption. Since there is Cl- depletion from loss of HCl, NaCl, and KCl from the stomach, the available anion is HCO3-, whose reabsorption is increased proximally, and urine pH may remain acidic despite alkalemia (paradoxical aciduria). Renal Cl - reabsorption (as well as Na+) reabsorption is high, and the urinary Cl- is therefore low (< 10-20 meq/L). In alkalosis, bicarbonaturia may force Na+ excretion as the accompanying cation even if volume depletion is present. Therefore, urinary Cl- is preferred to urinary Na+ as a measure of extracellular volume. An exception to the usefulness of urinary Cl- is in patients who have recently received diuretics. Their urine may contain high Na+ and Cldespite extracellular volume contraction. If diuretics are discontinued, the urinary Cl- will decrease. Metabolic alkalosis is generally associated with hypokalemia. This is due partly to the direct effect of alkalosis per se on renal potassium excretion and partly to secondary hyperaldosteronism from volume depletion. Hypokalemia induced in this fashion further worsens the metabolic alkalosis by increasing bicarbonate reabsorption in the proximal tubule and hydrogen ion secretion in the distal tubule. Administration of KCl will correct the disorder. Repletion of KCl is important to reverse the disorder. 1. Contraction alkalosis—Diuretics can acutely decrease extracellular volume from urinary loss of NaCl and water. There is no associated bicarbonaturia, so that body HCO3- content remains normal. However, plasma HCO3- increases because of extracellular fluid contraction—the reverse of what occurs in dilutional acidosis. 2. Posthypercapnia alkalosis—In chronic respiratory acidosis, compensatory increases in HCO3- occur (Table 21-13). Hypercapnia also directly affects the proximal tubule to decrease NaCl reabsorption, which can cause extracellular volume depletion. If PCO2 is corrected rapidly, as with mechanical ventilation, metabolic alkalosis will ensue until adequate bicarbonaturia occurs. Hypovolemia will inhibit bicarbonaturia until Cl- is repleted. Many patients with chronic respiratory acidosis receive diuretics, which further exacerbates the metabolic alkalosis. B. Saline-Unresponsive Alkalosis 1. Hyperaldosteronism—Primary hyperaldosteronism causes expansion of extracellular volume with hypertension. Metabolic alkalosis with hypokalemia results from the renal mineralocorticoid effect. In an attempt to decrease extracellular volume, high levels of NaCl are excreted, and for that reason the urinary Cl- is high (> 20 meq/L, often higher). Therapy with NaCl will only increase volume expansion and hypertension and will not treat the underlying problem of mineralocorticoid excess. 2. Alkali administration with decreased GFR—Despite large ingestions of HCO3-, enhanced bicarbonaturia almost always prevents a patient with normal renal function from developing metabolic alkalosis. However, with renal insufficiency, urinary excretion of bicarbonate is inadequate. If large amounts of HCO3- or metabolizable salts of organic acids such as sodium lactate, sodium citrate, or sodium gluconate are consumed, as with intensive antacid therapy, metabolic alkalosis will occur. In milk-alkali syndrome, large and sustained ingestion of absorbable antacids and milk causes renal insufficiency from hypercalcemia. Decreased GFR prevents appropriate bicarbonaturia from the ingested alkali, and metabolic alkalosis occurs. Volume contraction from renal hypercalcemic effects further exacerbates the alkalosis. Clinical Findings A. Symptoms and Signs There are no characteristic symptoms or signs. Orthostatic hypotension may be encountered. Weakness and hyporeflexia occur if serum K+ is markedly low. Tetany and neuromuscular irritability occur rarely. B. Laboratory Findings The arterial blood pH and bicarbonate are elevated. The arterial PCO2 is increased. Serum potassium and chloride are decreased. There may be an increased anion gap. Treatment Mild alkalosis is generally well tolerated. Severe or symptomatic alkalosis (pH > 7.60) requires urgent treatment. A. Saline-Responsive Metabolic Alkalosis Therapy for saline-responsive metabolic alkalosis is aimed at correction of extracellular volume deficit. Depending on the degree of hypovolemia, adequate amounts of 0.9% NaCl and KCl should be administered. Discontinuation of diuretics and administration of H2-blockers in patients whose alkalosis is due to nasogastric suction can be useful. If impaired pulmonary or cardiovascular status prohibits adequate volume repletion, acetazolamide, 250-500 mg intravenously every 4-6 hours, can be used. One must be alert to the possible development of hypokalemia, since potassium depletion can be induced by forced kaliuresis via bicarbonaturia. Administration of acid can be used as emergency therapy. HCl, 0.1 mol/L, is infused via a central vein (the solution is sclerosing). Dosage is calculated to decrease the HCO3- level by one-half over 2-4 hours, assuming a HCO3- volume of distribution (L) of 0.5 body weight (kg). Patients with marked renal insufficiency may require dialysis. B. Saline-Unresponsive Metabolic Alkalosis Therapy for saline-unresponsive metabolic alkalosis includes surgical removal of a mineralocorticoid-producing tumor and blockage of aldosterone effect with an angiotensin-converting enzyme inhibitor or with spironolactone. Metabolic alkalosis in primary aldosteronism can be treated only with potassium repletion. Adrogue HJ et al: Management of life threatening acid base disorders. (Part 2.) N Engl J Med 1998;338:107. [PMID: 98069970] Khanna A et al: Metabolic alkalosis. Respir Care 2001;46:354. [PMID: 11262555] FLUID MANAGEMENT Most of those who require water and electrolyte intravenously are relatively normal people who cannot take orally what they require for maintenance. The range of tolerance for water and electrolytes (homeostatic limits) permits reasonable latitude in therapy provided normal renal function exists to accomplish the final regulation of volume and concentration. An average adult whose entire intake is parenteral would require for maintenance 2500-3000 mL of 5% dextrose in 0.2% saline solution (34 meq Na+ plus 34 meq Cl-/L). To each liter, 30 meq of KCl could be added. In 3 L, the total chloride intake would be 192 meq, which is easily tolerated. Guidelines for gastrointestinal fluid losses are shown in Table 21-18. In situations requiring maintenance or maintenance plus replacement of fluid and electrolyte by parenteral infusion, the total daily ration should be administered continuously over the 24-hour period in order to ensure the best utilization by the patient. If parenteral fluids are the only source of water, electrolytes, and calories for longer than a week, more complex fluids containing amino acids, lipid, trace metals, and vitamins may be indicated. (See Total Parenteral Nutrition, Chapter 29.) Copyright © 2003 by The McGraw-Hill Companies, Inc. All rights reserved. Measurement of renal function: Renal function tests (see Table 214-5) are useful in evaluating the severity of kidney disease and in following its progress. Serum creatinine can be used as an index of renal function because creatinine production and excretion are reasonably constant in the absence of muscle disease. Serum concentration of creatinine varies inversely with the GFR and therefore is a useful index of the GFR if production (related to muscle mass and age) and metabolism (increased in uremia) are considered. The upper limit of serum creatinine concentration in men with normal GFR is 1.2 mg/dL (110 µmol/L); in women, 1 mg/dL (90 µmol/L). Creatinine clearance in men is 140 to 200 L/day (70 14 mL/min/m2) and in women, 120 to 180 L/day (60 10 mL/min/m2). The creatinine clearance (Clcreat) can be calculated from the serum creatinine concentration in men as: In women, the calculated values are multiplied by 0.85. Creatinine clearance is not useful for detecting early kidney damage due to hypertrophy of residual glomeruli. After loss of 50 to 75% of the normal glomerular filtration surface, a decrease in creatinine clearance is clearly detectable. Thus, a normal creatinine clearance cannot exclude the presence of mild renal disease. BUN, in contrast to serum creatinine, is unsuitable as a single measure of renal function because it is influenced by variations in urine flow rate and by the production and metabolism of urea. The BUN:creatinine ratio often is used to differentiate prerenal, renal, or postrenal (obstructive) azotemia. A ratio > 15 is abnormal and suggests prerenal or postrenal azotemia. The BUN:creatinine ratio also is elevated whenever urea production is increased by diet, TPN, or glucocorticoid therapy; with some neoplasms and antibiotics; and with excessive protein catabolism, as seen in infections and uncontrolled diabetes mellitus. Common causes of prerenal azotemia include shock, ECF depletion, massive GI hemorrhage, severe heart and liver failure, and bilateral tight renal artery stenosis. The BUN:creatinine ratio is normal in renal azotemia. The ratio is low in pregnancy, overhydration, severe liver disease, and malnutrition. Tests of renal concentrating capacity are simple and diagnostically helpful. A loss of concentrating capacity in the presence of adequate vasopressin stimulation is associated with tubulointerstitial disease (edema, infiltrate, fibrosis), except when NDI is present. The loss of concentrating ability frequently is present long before a depression of GFR is measurable. Renal concentrating capacity is best tested by water deprivation for 12 to 14 h and by the response to exogenous vasopressin. After the patient has fasted for 12 to 14 h overnight, the osmolality of the initial morning urine and of subsequent hourly samples is measured. When hourly measurements differ < 30 mOsm/kg or sp gr < 0.001, the maximum concentrating capacity has been reached with water deprivation. Aqueous vasopressin 5 U sc or desmopressin 10 µg by nasal insufflation is given, and the urine osmolality is measured after another hour. The results of this type of testing are noted in Fig. 214-1. (Caution: In patients with renal failure, water deprivation may be harmful and usually is not useful in diagnosis; the concentrating capacity is always abnormal when the GFR is significantly reduced.) A lack of response to water deprivation or exogenous vasopressin suggests an intrinsic renal concentrating defect that may be due to one or more of the following functional tubular impairments: congenital (eg, NDI, Fanconi's syndrome) or acquired (eg, osmotic diuresis, certain diuretics [furosemide, bumetanide, ethacrynic acid], K deficiency, hypercalcemia). Otherwise, tubulointerstitial disease should be considered, as in sickle cell disease, toxic nephritis, pyelonephritis, or any renal disease severe enough to produce azotemia. For other responses to these tests and their interpretations, see Treatment under Diabetes Insipidus in Ch. 7. Measurement of the renal plasma flow is no more useful clinically than the GFR but is more difficult and costly. Additional special tests of renal tubular function usually require research laboratories and are reserved for patients with specific problems. However, tests that measure plasma phosphate and urate, urinary amino acids, and urine pH are readily available and may prove useful in screening specific clinical problems. 3. Blood testing procedures— a. Glucose tolerance test— (1) Methodology and normal fasting glucose—Plasma or serum from venous blood samples has the advantage over whole blood of providing values for glucose that are independent of hematocrit and that reflect the glucose concentration to which body tissues are exposed. For these reasons, and because plasma and serum are more readily measured on automated equipment, they are used in most laboratories. If serum is used or if plasma is collected from tubes that lack an agent to block glucose metabolism (such as fluoride), samples should be refrigerated and separated within 1 hour after collection. (2) Criteria for laboratory confirmation of diabetes mellitus— If the fasting plasma glucose level is 126 mg/dL or higher on more than one occasion, further evaluation of the patient with a glucose challenge is unnecessary. However, when fasting plasma glucose is less than 126 mg/dL in suspected cases, a standardized oral glucose tolerance test may be done (Table 27-4). For proper evaluation of the test, the subjects should be normally active and free from acute illness. Medications that may impair glucose tolerance include diuretics, contraceptive drugs, glucocorticoids, niacin, and phenytoin. Because of difficulties in interpreting oral glucose tolerance tests and the lack of standards related to aging, these tests are being replaced by documentation of fasting hyperglycemia. Since fasting plasma glucose is known to increase with aging, clinicians should be more tolerant of slight abnormalities of fasting glucose values in older people (over 70 years of age) and not deprive patients of occasional sugar-containing snacks when symptoms are not evident. However, an occasional elderly patient may benefit from the diagnosis of mild diabetes in that macular edema may be detected earlier and laser treatment initiated before vision deteriorates permanently. b. Glycated hemoglobin (hemoglobin A1) measurements—Glycated hemoglobin is abnormally high in diabetics with chronic hyperglycemia and reflects their metabolic control. It is produced by nonenzymatic condensation of glucose molecules with free amino groups on the globin component of hemoglobin. The higher the prevailing ambient levels of blood glucose, the higher will be the level of glycated hemoglobin. The major form of glycohemoglobin is termed hemoglobin A1c, which normally comprises only 4-6% of the total hemoglobin. The remaining glycohemoglobins (2-4% of the total) consist of phosphorylated glucose or fructose and are termed hemoglobin A1a and hemoglobin A1b. Some laboratories measure the sum of these three glycohemoglobins and report it as hemoglobin A1, but more laboratories are converting to the more intricate but highly specific HbA1c assay. There are now monoclonal immunoassays for measuring HbA1c. Machines based on this technology can be used in clinicians' offices. They use capillary blood and give a result in about 9 minutes, allowing immediate feedback to the patient regarding their glycemic control. Since glycohemoglobins circulate within red blood cells whose life span lasts up to 120 days, they generally reflect the state of glycemia over the preceding 8-12 weeks, thereby providing an improved method of assessing diabetic control. Measurements should be made in patients with either type of diabetes mellitus at 3- to 4-month intervals so that adjustments in therapy can be made if glycohemoglobin is either subnormal or if it is more than 2% above the upper limits of normal for a particular laboratory. In patients monitoring their own blood glucose levels, glycohemoglobin values provide a valuable check on the accuracy of monitoring. In patients who do not monitor their own blood glucose levels, glycohemoglobin values are essential for adjusting therapy. Use of glycohemoglobin for screening is controversial. Sensitivity in detecting known diabetes cases by hemoglobin A1c measurements is only 85%, indicating that diabetes cannot be excluded by a normal value. On the other hand, elevated hemoglobin A1c assays are fairly specific (91%) in identifying the presence of diabetes. Occasionally, fluctuations in hemoglobin A1 are due to an acutely generated, reversible, intermediary (aldimine-linked) product that can falsely elevate glycohemoglobins when measured with "short-cut" chromatographic methods. This can be eliminated by using specific HPLC methods that detect HbA1c or by dialysis of the hemolysate before chromatography. When hemoglobin variants are present, such as negatively charged hemoglobin F, acetylated hemoglobin from high-dose aspirin therapy, or carbamoylated hemoglobin produced by the complexing of urea with hemoglobin in uremia, falsely high "hemoglobin A1" values are obtained with commonly used chromatographic methods. In the presence of positively charged hemoglobin variants such as hemoglobin S or C, or when the life span of red blood cells is reduced by increased hemolysis or hemorrhage, falsely low values for "hemoglobin A1" result. Serum fructosamine is formed by nonenzymatic glycosylation of serum proteins (predominantly albumin). Since serum albumin has a much shorter half-life than hemoglobin, serum fructosamine generally reflects the state of glycemic control for only the preceding 2 weeks. Reductions in serum albumin (eg, nephrotic state or hepatic disease) will lower the serum fructosamine value. When abnormal hemoglobins or hemolytic states affect the interpretation of glycohemoglobin or when a narrower time frame is required, such as for ascertaining glycemic control at the time of conception in a diabetic woman who has recently become pregnant, serum fructosamine assays offer some advantage. Normal values vary in relation to the serum albumin concentration and are 1.5-2.4 mmol/L when the serum albumin level is 5 g/dL. c. Self-monitoring of blood glucose—Capillary blood glucose measurements performed by patients themselves, as outpatients, are extremely useful. In type 1 patients in whom "tight" metabolic control is attempted, they are indispensable. A portable battery-operated glucometer provides a digital readout of the intensity of color developed when glucose oxidase paper strips are exposed to a drop of capillary blood for up to 45 seconds. A large number of blood glucose meters are now available. All are accurate, but they vary with regard to speed, convenience, size of blood samples required, and cost. Popular models include those manufactured by LifeScan (One Touch), Bayer Corporation (Glucometer Elite, DEX), Roche Diagnostics (Accu-Chek), Abbott Laboratories (ExacTech, Precision), and Home Diagnostics (Prestige). One Touch Ultra, for example, requires only 0.3 mL of blood and gives a result in 5 seconds—and illustrates how there has been continued progress in this technologic area. Various glucometers appeal to a particular consumer need and are relatively inexpensive, ranging from $50.00 to $100.00 each. The more expensive models compute blood glucose averages and can be attached to printers for data records and graph production. Test strips remain a major expense, costing 50-75 cents apiece. In self-monitoring of blood glucose, patients must prick their finger with a 28-gauge lancet (Monolet, Ames Co.), which can be facilitated by a small plastic trigger device such as an Autolet (Ames Co.), SoftClix (BoehringerMannheim), or Penlet (Lifescan, Inc.). When used for multiple patients, as in a clinic, physician's office, or hospital ward, disposable finger-rest platforms are required to avoid inadvertent transmission of blood-borne viral diseases. Some meters such as the FreeStyle (Therasense) have been approved for measuring glucose in blood samples obtained at alternative sites such as the forearm and thigh. There is, however, a 5- to 20-minute lag in the glucose response on the arm with respect to the glucose response on the finger. Forearm blood glucose measurements could therefore result in a delay in detection of rapidly developing hypoglycemia. The accuracy of data obtained by glucose monitoring requires education of the patient in sampling and measuring procedures as well as in proper calibration of the instruments. Bedside glucose monitoring in a hospital setting requires rigorous quality control programs and certification of personnel to avoid errors. Noninvasive glucose monitoring is a subject of intensive research interest, and a prototype, the GlucoWatch, has been approved by the FDA. It utilizes reverse iontophoresis to measure interstitial glucose values with acceptable accuracy. Because of a lag period of 15-20 minutes, it records measurements only three times an hour. Sweating can result in skipped readings, which causes problems in warm humid climates. Further experience is needed to evaluate its clinical value. Table 27-4. The Diabetes Expert Committee criteria for evaluating the standard oral glucose tolerance test. 1 Normal Glucose Tolerance Impaired Glucose Tolerance Diabetes Mellitus2 Fasting plasma glucose (mg/dL) < 110 110-125 126 Two hours after glucose load (mg/dL) < 140 140 but < 200 200 Give 75 g of glucose dissolved in 300 mL of water after an overnight fast in subjects who have been receiving at least 150-200 g of carbohydrate daily for 3 days before the test. 1 A fasting plasma glucose 126 mg/dL is diagnostic of diabetes if confirmed on a subsequent day. 2