Download Death Associated With Rifampin and Pyrazinamide 2

Death Associated With
Rifampin and Pyrazinamide
2-Month Treatment of Latent
Mycobacterium Tuberculosis*
Ann Medinger, MD, FCCP
We present the case of an elderly patient who died of
fulminant hepatic failure in the course of receiving
2 months of treatment with pyrazinamide and rifampin for his latent tuberculosis. This 2-month
course of treatment for latent tuberculosis is one of
four options recently recommended by the Centers
for Disease Control and Prevention. We discuss the
safety of using this two-drug regimen to treat latent
tuberculosis in stable elderly patients.
(CHEST 2002; 121:1710 –1712)
Key words: drug toxicity; latent tuberculosis treatment; Mycobacterium tuberculosis; pyrazinamide; rifampin
Abbreviation: CDC ⫽ Centers for Disease Control and Prevention
his elderly patient with latent tuberculosis died of
T fulminant
hepatic failure in the course of receiving 2
months of treatment with pyrazinamide and rifampin.
Case Report
The patient was a 68-year-old, male, home health aid, who
was admitted to the Washington, DC, Veterans Affairs Medical Center in June 2001 with a history of three days of nausea
and malaise. His history included alcohol abuse, hepatic
steatosis revealed by liver biopsy 5 years prior, diabetes
mellitus, hypertension, gout, and a positive tuberculin skin test
of long standing. Two months prior to hospital admission, the
patient had requested antituberculous medication because he
wanted to avoid the continuing requirement for an annual
chest radiograph. His employer required positive tuberculin
reactors to provide a physician’s certification of freedom from
active tuberculosis. The common physician requirement that
such patients have a normal current chest radiograph to be
certified is usually waived after a course of antituberculous
treatment. Urged by the patient and supported by the 2000
Centers for Disease Control and Prevention (CDC) guidelines1 for the treatment of latent tuberculosis, the physician
agreed to prescribe preventive therapy. On March 29, treatment was initiated with pyrazinamide and rifampin, for a
2-month course. On April 13, the results of follow-up liver
function testing were normal. On May 25, the patient reported
feeling well at a clinical nutritionist appointment for the
management of his diabetes. However, on May 28, he reported
to the emergency department feeling ill. The patient had
abstained from alcohol consumption for at least 5 years since
*From the Department of Veterans Affairs Medical Center,
Washington, DC.
Manuscript received July 10, 2001; revision accepted October 4,
2001.
Correspondence to: Ann Medinger, MD, FCCP, 5605 Park St,
Chevy Chase, MD 20815-3404; e-mail: medingerbeeny@
erols.com
receiving abnormal results from his liver biopsy specimen
testing. His medications included the following: aspirin; colchicine; ibuprofen; lisinopril; metformin; pyrazinamide,
2 g qd; and rifampin, 600 mg qd.
On physical examination, the following measurements were
obtained: BP, 128/36 mm Hg; body temperature, 34.8°C; heart
rate, 89 beats/min; respiratory rate, 26 breaths/min; and body
weight, 101.8 kg. The patient was an obese, lethargic, elderly
African-American man, who was confused and disoriented. He
had icteric sclerae and asterixis but no stigmata of chronic liver
disease. A chest examination revealed rales at the right base, and
the findings of a cardiac examination were normal. His abdomen
was protuberant and nontender, with normal bowel sounds. A
rectal examination revealed brown heme-positive stool. The
patient’s extremities had no edema, clubbing, bruises, muscle
atrophy, or fasciculations. A neurologic examination showed
encephalopathy and was otherwise nonfocal. His chest radiograph was clear. An abdominal sonogram showed ascites and a
fatty liver, and was negative for biliary or urinary tract obstructions. Table 1 shows the sequential laboratory measurements for
this patient.
The patient received the following supportive care: IV
fluids, including fluid containing 20% dextrose in water with
additional fluid containing 50% dextrose in water, for recurrent hypoglycemia; blood products for GI bleeding; enteral
lactulose plus neomycin for encephalopathy; IV dopamine first
for renal perfusion and then for BP support. He received
empiric broad-spectrum, renal-friendly antibiotics. On the
second hospital day, he became comatose and anuric. He was
intubated and required mechanical ventilation. He never
regained consciousness, and his hepatic function continued to
deteriorate. A liver transplantation was denied, and the patient
died on the third hospital day.
Discussion
This elderly asymptomatic individual developed fulminant hepatic necrosis, rhabdomyolysis, and acute renal
failure, and he died within 3 days of reporting symptoms of
illness. One can infer from the fact that he had been stable
prior to pyrazinamide and rifampin therapy, that the newly
added medications probably caused or precipitated his
demise. However, it is possible that his liver failure was
due to another cause. Although his liver function measurements were within normal units, he had past biopsyproven hepatic steatosis. He was also taking other medications with potential hepatotoxicity, and, as an elderly
gentleman living alone, he may have had an additional
toxic exposure that was not reported or detected. His risk
factors for complications of treatment (ie, history of liver
disease, polypharmacy, and advanced age) may have enhanced the toxicity of the pyrazinamide and rifampin. The
case underscores the importance of frequent monitoring
of liver function in patients receiving this treatment. It also
raises questions about the merits of treating asymptomatic
elderly patients for latent tuberculosis.
The CDC has reported two similar cases of severe
hepatitis in individuals who were ⬎ 50 years of age and
who were taking pyrazinamide and rifampin for latent
tuberculosis.2 Reports within the last 5 years in the
European and Japanese literature have noted an increase
in the incidence of hepatotoxicity with the simultaneous
1710
Downloaded From: http://publications.chestnet.org/pdfaccess.ashx?url=/data/journals/chest/21977/ on 06/18/2017
Selected Reports
Table 1—Sequential Laboratory Measurements in the Case Reported*
Measurement Dates
Tests
5/31/01
WBC, cells/␮L†
HGB, g/dL
MCV, mL
PLT, cells/␮L†
Glucose, mg/dL
BUN, mg/dL
Creatine, mg/dL
NA, mmol/L
Potassium, mmol/L
CL, mmol/L
CO2, mmol/L
Protein, g/dL
Albumin, g/dL
AP, mU/mL‡
AST/GOT, mU/mL‡
ALA/GPT, mU/mL‡
T Bil, mg/dL
D Bil, mg/dL
NH3, ␮mol/L
Lactate, mmol/L
CPK, U/L‡
INR
9.2
10.6
81.6
165
42
56
11
139
4.8
107
11
6.4
2
207
2,720
1,930
19.2
12.6
181
9.4
10.1
5/28/01
417
39
5.4
139
6.5
110
8
5.9
2.1
175
758
620
16.7
9.2
135
13.2
9,130
6
4/13/01
8.5
4.0
63
33
31
0.5
3/29/01
3/16/00
Reference Values
1.0
3.2–9.5
13.2–17.3
80.2–96.7
152–375
70–121
6–23
0.7–1.5
135–147
3.5–5.3
100–109
21–31
6–8.2
3.7–5
43–130
8–40
8–45
0.2–1.2
0–0.2
11–35
0.6–2.2
20–300
1.0
6.1
12.6
80
249
163
13
1.3
141
3.8
101
26
8.3
3.7
66
39
48
0.5
0.1
*Urine tests were positive for myoglobin without cells or casts. Urine tests were negative for eosinophils, blood acetaminophen, alcohol, viral
hepatitis, and alphafetoprotein. HGB ⫽ hemoglobin; MCV ⫽ mean corpuscular volume; NA ⫽ sodium; AP ⫽ alkaline phosphatase; AST/
GOT ⫽ aspartate aminotransferase/glutamic-oxaloacetic transaminase; ALA/GPT ⫽ ␴-aminolevulinic acid/glutamic-pyruvic transaminase;
CPK ⫽ creatine phosphokinase; INR ⫽ international normalized ratio; PLT ⫽ platelet count; CL ⫽ chloride concentration; T Bil ⫽ total
bilirubin; D Bil ⫽ direct bilirubin.
†Cells signifies thousands of cells.
‡Measurement made at 37°C.
administration of isoniazid, pyrazinamide, and rifampin.
One Barcelona institution reported five cases in 1 year
resulting in three deaths.3
The risk/benefit ratio of treating healthy patients for
latent tuberculosis with potentially toxic drugs has been
discussed in the past.4 Individuals who are not immunedeficient have a ⬍ 1 in 20 chance of developing active
tuberculosis after the second year of stable latency. The
risk of toxicity from antituberculous drugs increases
with age and with the number of simultaneously administered hepatotoxic drugs.5 In the pre-AIDS era, cohort
studies showed that treating patients with latent tuberculosis with 12 months of isoniazid therapy carried a greater
risk from drug toxicity than from the risk of developing active
tuberculosis in those patients who were ⬎ 50 years of age.
Previous CDC guidelines have recommended that stable
latent tuberculosis should not be treated in individuals over
the age of 35 years with few exceptions, including those
patients with immune deficiency.
During the past 2 decades, progress in the use of
combination antituberculous drugs and a heightened
sense of danger from active tuberculosis have given us
greater enthusiasm for treating patients with latent
tuberculosis. The use of two or three powerful drugs
simultaneously has permitted shorter treatment schedules for both active and latent tuberculosis, with higher
rates of compliance. The increased concern about active
tuberculosis has stemmed from the following events: a
temporary resurgence in the incidence of active tuberculosis in the 1980s; an increased incidence of drugresistant tuberculosis; and the AIDS epidemic, with an
increased prevalence of immunodeficiency. Mycobacterium tuberculosis infection hastens the progression of
AIDS. In HIV-infected individuals, latent M tuberculosis infection may be undetectable by tuberculin test, yet
it virtually always becomes active if untreated. Hence
HIV-infected individuals have much to gain from eliminating active tuberculosis in the community. Administering antituberculous antibiotics to all individuals who
have positive tuberculin test results reduces the incidence of active tuberculosis. In its 2000 revised guidelines for treating latent tuberculosis, the CDC has recommended 2 months of rifampin and pyrazinamide therapy as
being the shortest of the four short-course options for
treatment,1 based on laboratory results in mice and a multicenter trial in an HIV-infected population.6 The new guidelines also have dropped the age restriction for the treatment
of patients with latent tuberculosis.
Our case and the other reports raise questions about
the safety of using rifampin and pyrazinamide therapy
to treat latent tuberculosis in stable non-HIV patients
who are over 50 years of age. These patients often have
risk factors for hepatotoxicity in addition to age, as our
case illustrates. If active tuberculosis develops in these
www.chestjournal.org
Downloaded From: http://publications.chestnet.org/pdfaccess.ashx?url=/data/journals/chest/21977/ on 06/18/2017
CHEST / 121 / 5 / MAY, 2002
1711
people, drug treatment is highly effective, clearly worth
the risk of toxicity, and the event precipitates aggressive
follow-up by the public health system. As clinicians, we
must take care that our eagerness to eliminate tuberculosis does no harm.
therapy, and thoracic surgery.8 We report on a patient who
suddenly developed pneumorrhachis several years after
undergoing a thoracotomy and external beam radiation
therapy for lung carcinoma.
References
Case Report
1 Centers for Disease Control and Prevention. Targeted tuberculin testing and treatment of latent tuberculosis infection.
MMWR Morb Mortal Wkly Rep 2000; 49(RR-6):1–51
2 Centers for Disease Control and Prevention. Fatal and severe
hepatitis associated with rifampin, and pyrazinamide for the
treatment of latent tuberculosis infection: New York, and Georgia, 2000. MMWR Morb Mortal Wkly Rep 2001; 50:289–291
3 Moitinho E, Salmeron JN, Mas A, et al. Severe hepatotoxicity
of tuberculostatic agents: increase in incidence. Gastroenterol
Hepatol 1996; 19:448 – 451
4 Miller B. Preventive therapy for tuberculosis. Med Clin
North Am 1993; 77:1263–1275
5 Dutt AK, Stead WW. Tuberculosis in the elderly. Med Clin
North Am 1993; 77:1353–1368
6 Gordin F, Chaisson RE, Matts JP, et al. Rifampin and
pyrazinamide vs isoniazid for prevention of tuberculosis in
HIV-infected persons: an international randomized trial.
JAMA 2000; 283:1445–1450
A 60-year-old man received a diagnosis of left upper-lobe, large
cell, undifferentiated carcinoma with chest wall invasion. He
underwent a left upper-lobe lobectomy, en bloc chest wall
resection, and resection of ribs 3 to 5 followed by 500-cGy
external beam radiation therapy through anteroposterior and
posteroanterior ports.
Three years later, after protracted severe coughing, this patient
developed sudden-onset extensive subcutaneous emphysema in
the head, neck, and chest. A chest radiograph and CT scan of the
chest identified no pneumothorax. We chose conservative management because a chest tube could not be placed strategically,
and we wished to avoid surgical intervention in view of the
presence of COPD and the patient’s limited underlying pulmonary reserve. Subcutaneous emphysema of the neck and chest
persisted for the following 8 months and then resolved abruptly.
Several days later, the patients presented with excruciating
pain in the back and left arm and leg. A spinal cord compression
syndrome was suspected clinically, and an MRI scan revealed an
epidural gas collection (5 to 10 mL) that extended from C3
through T4. A chest radiograph revealed a pneumomediastinum
and subcutaneous emphysema (Fig 1). A CT scan of the chest
revealed air in the epidural space, the mediastinum, and subcutaneous tissues (Fig 2). Air from the mediastinal and subcutaneous regions moved through the neural foramina into the epidural
space. No definite pneumothorax was identified. The precise
location of the presumed bronchopleural fistula could not be
identified by MRI, CT scan, ventilation-perfusion scan, or fiberoptic bronchoscopy. The presence of an active infection at the
time of presentation was not suspected.
The patient did not respond to 2 weeks of conservative
management, and at that time he underwent a modification of his
prior thoracoplasty with resection of ribs 2 and 6 and a combined
serratus anterior and intercostal muscle flap closure of the
bronchopleural fistula. Multiple small air leaks were found
emanating from the inflamed surfaces of the superior segment of
the left lower lobe parenchyma. The bronchopleural fistula air
leak returned 2 weeks after his muscle flap closure surgery. The
failure of the surgery was not surprising since the muscle flap
covered an area of the lung that was scarred and emphysematous.
A chest tube was placed in the area of the presumed air leak and
was left to open-air drainage. Persistent drainage of air continued
through the chest tube, without recurrence of pneumorrhachis,
for the next 2 years until his death due to end-stage COPD.
An Unusual Case of
Pneumorrhachis Following
Resection of Lung Carcinoma*
Ross L. Ristagno, MD, FCCP; Loren F. Hiratzka, MD, FCCP;
and Raymond C. Rost, Jr., MD
We report an unusual case of a patient with a
bronchopleural-subcutaneous fistula that occurred 3
years following a lobectomy and en bloc chest wall
resection for carcinoma. Following a sudden resolution of subcutaneous emphysema, an epidural air
collection developed. Spontaneous pneumorrhachis
is a previously unreported late complication after
surgical or radiation therapy for lung carcinoma.
(CHEST 2002; 121:1712–1714)
Key words: bronchial fistula complications; CT; emphysema;
emphysema complications; mediastinal emphysema; pneumocephalus; radiograph; spinal canal abnormalities; subcutaneous
emphysema
ir very rarely enters the spinal canal (pneumorrhachis).
A Pneumorrhachis
has been reported to occur following
spontaneous pneumomediastinum,1– 4 traumatic pneumothorax,5 skull fracture,6 epidural anesthesia,7 radiation
*From the Department of Pulmonary Medicine (Dr. Ristagno),
and the Department of Diagnostic Radiology (Dr. Rost), The
Christ Hospital, Cincinnati, OH; and the Department of Cardiac
Surgery (Dr. Hiratzka), TriHealth, Cincinnati, OH.
Manuscript received July 3, 2001; revision accepted November
27, 2001.
Correspondence to: Ross L. Ristagno, MD, FCCP, 7960 Greylock
Dr, Cincinnati, OH 45243; e-mail: [email protected]
Discussion
In general, pneumorrhachis may enter the epidural
space or may penetrate further into the subarachnoid
space. Air that enters the epidural space usually does not
migrate significantly, and complete resorption occurs
spontaneously. Epidural space pneumorrhachis causes no
symptoms or local pain. When pulmonary air enters the
subarachnoid space, a bronchopleural-subarachnoid fistula is present, and air easily travels cephalad to the
cranium (ie, pneumocephalus). These patients experience
severe headaches, or less often, focal neurologic symptoms
similar to a stroke.9 Pneumorrhachis is best demonstrated
on CT scanning but may be seen on MRI scanning or
1712
Downloaded From: http://publications.chestnet.org/pdfaccess.ashx?url=/data/journals/chest/21977/ on 06/18/2017
Selected Reports