Download Dr Adebanjo Final Draft Thesis 8November2010 Examiner

A.B. ADEBANJO 2010
COMPARISON OF CLINICAL AND
IMMUNOLOGICAL RESPONSES TO ZIDOVUDINE
(AZT) AND TENOFOVIR (TDF) – CONTAINING ARV
REGIMENS IN PATIENTS TAKING HAART AT
ROMA HEALTH SERVICE AREA OF LESOTHO.
BY
ADEBANJO ADEFOLARIN BABAFEMI
ADEBANJO ADEFOLARIN BABAFEMI
STUDENT NUMBER 15249484
Thesis submitted in partial fulfilment of the requirements for
the degree
M Med (Family Medicine)
At
Stellenbosch University
August 2010
Supervisor: Dr. Michael Pather
1
A.B. ADEBANJO 2010
DECLARATION
I, Adefolarin Babafemi ADEBANJO, hereby declare that the work which I
hereby submit as partial fulfilment for the degree MMed (Family Medicine), on
which this thesis is based, is original (except where acknowledgements indicate
otherwise) and that neither the whole work nor any part of it has been
submitted, or is being submitted, for another degree at this or any other
university.
Signed:
Date: 30th August 2010
2
A.B. ADEBANJO 2010
ABSTRACT
Objective: The objective of this retrospective cohort study is to assess whether
demographic and anthropometric parameters, laboratory tests, co-morbidity,
co-infection, treatment regimen, IRIS and adherence to treatment predict the
expected response to HAART and differences if any, in the pattern of response
as measured by CD4 count, weight gain and haemoglobin levels in two cohorts
of patients in Roma, The Kingdom of Lesotho.
Method: Data were collected randomly from a computerised database of the
Antiretroviral Centre of the hospital and two cohorts of 151 subjects in each of
the two arms of the study were identified from hospital records from January
2008. Each of these subjects was followed up over a period of 12 months with
data obtained for at least 2 visits within the 12 month span. Data were
obtained at baseline, 3 months and also at 6 and 12 months marks. Data on
characteristics were compared between the two arms. Variables that may be
potential confounders were identified and univariate and multivariate logistic
regression analyses were carried out to establish differences independent of
confounding factors for the combined endpoints as well as for each endpoint
separately.
Results: In all 302 patients had their records analysed and comparison of
clinical and immunological response patterns in patients taking AZT and TDFcontaining ART regimens and the possible prediction of which the regimen
would be better and within which population. Despite the perceived mismatch
between two NRTIs it can be concluded from the results of this study that,
overall, the inclusion of AZT in treatment regimen showed a modest protective
effect over the TDF counterpart as measured by the endpoints of the
discriminative powers of the Receiver Operating Curves of the explanatory
variables being 66% 77% and 66% for CD4, Haemoglobin and Weight
3
A.B. ADEBANJO 2010
respectively, and 63% 70% and 65% for the same variables in the AZT and TDF
arms of the study respectively.
Conclusion: In a population of HIV patients on treatment in resource-limited
settings AZT-containing regimens appear to show a slight improvement over
the TDF-containing ones.
4
A.B. ADEBANJO 2010
Table of contents
CHAPTER 1 ........................................................................................................................................................... 12
Study design influences ...................................................................................................................................... 23
MOTIVATION AND AIM OF THE STUDY ..................................................................................................................... 25
CHAPTER 2 ........................................................................................................................................................... 26
METHODS ................................................................................................................................................................ 26
Setting. ............................................................................................................................................................... 26
This study was conducted in a district hospital in Roma, a semi-urban region of The Kingdom of Lesotho. .... 26
Study design ....................................................................................................................................................... 26
Inclusion criteria ................................................................................................................................................ 26
Exclusion criteria ............................................................................................................................................... 26
Patient selection and sampling frame ................................................................................................................ 27
Measurement of covariates ................................................................................................................................ 27
Data analysis ..................................................................................................................................................... 28
CHAPTER 3 ........................................................................................................................................................... 30
RESULTS OF STUDY ................................................................................................................................................. 30
Conditional logistic regression analysis ............................................................................................................ 38
CHAPTER 4 ........................................................................................................................................................... 47
DISCUSSION ............................................................................................................................................................. 47
Possible limitations of the study ......................................................................................................................... 48
Bias and Confounding ........................................................................................................................................ 50
REFERENCES ............................................................................................................................................................ 53
ADDENDUM 1 .......................................................................................................................................................... 60
DATA COLLECTION FORM ........................................................................................................................................ 60
5
A.B. ADEBANJO 2010
List of tables
Table 3.1 Patient demographics for the AZT cohort at baseline.
Table 3.2 Patient demographics for the TDF cohort at baseline.
Table 3.3 Other considerations on the AZT cohort.
Table 3.4 Other considerations on the TDF cohort.
Table 3.5 Univariate predictors of CD4 response at 12 months of therapy
(AZT)
Table 3.6 Univariate predictors of weight response at 12 months of therapy
(AZT)
Table 3.7 Univariate predictors of haemoglobin response at 12 months of
therapy (AZT)
Table 3.8 Univariate predictors of CD4 response at 12 months of therapy
(TDF)
Table 3.9 Univariate predictors of weight response at 12 months of therapy
(TDF)
Table 3.10 Univariate predictors of haemoglobin response at 12 months of
therapy (TDF)
Table 3.11 Adjusted Odds Ratios for retained variables for CD4 (AZT)
Table 3.12 Adjusted Odds Ratios for retained variables for haemoglobin (AZT)
Table 3.13 Adjusted Odds Ratios for retained variables for weight (AZT)
Table 3.14 Adjusted Odds Ratios for retained variables for CD4 (TDF)
Table 3.15 Adjusted Odds Ratios for retained variables for weight (TDF)
Table 3.16 Adjusted Odds Ratios for retained variables for haemoglobin (TDF)
6
A.B. ADEBANJO 2010
List of figures
Figure 3.1 ROC curve for AZT CD4
Figure 3.2 ROC for AZT haemoglobin
Figure 3.3 ROC for AZT weight
Figure 3.4 ROC for TDF CD4
Figure 3.5 ROC for TDF weight
Figure 3.6 ROC for TDF haemoglobin
7
A.B. ADEBANJO 2010
Abbreviations
ACTG
AIDS Clinical Trial Group
AIDS
Acquired Immunodeficiency Syndrome
ANC
Antenatal Care
ARV
Antiretroviral
AZT
Zidovudine
CEO
Chief Executive Officer
ddC
Zalcitabine
d4T
Stavudine
DNA
Deoxyribonucleic Acid
EDTA
Ethylenediamine Tetraacetic Acid
GRID
Gay-Related Immune Deficiency
8
A.B. ADEBANJO 2010
HAART
HIV
Highly Active Antiretroviral Agents
Human Immune Deficiency Virus
HLTV
Human T-Lymphotrophic Virus
IRIS
Immune Reconstitution Inflammatory Syndrome
LAV
Lymphadenopathy-Associated Virus
MTCT
Mother-To-Child Transmission
NGO
Non-Governmental Organization
NNRTI Non-Nucleoside Reverse Transcriptase Inhibitor
NRTI
Nucleoside Reverse Transcriptase Inhibitor
OIP
Opportunistic Infection Prophylaxis
PCP
Pneumocystis Pneumonia
PHC
Primary Health Care
9
A.B. ADEBANJO 2010
PMTCT
Prevention of Mother-To-Child Transmission
RNA
Ribonucleic Acid
TDF
Tenofovir
UNAIDS United Nations Programme on HIV/AIDS
VCT
Voluntary Counselling and Testing
WHO
World Health Organization
10
A.B. ADEBANJO 2010
CHAPTER 1
Background
The burden that the HIV/AIDS pandemic has put on economies of countries
particularly on the developing countries in sub-Saharan Africa as well as on
infected
and
affected
people
necessitated
the
development
of
various
interventions of which antiretroviral agents are principal. Worst hit countries of
the world are still grappling with effective coverage of their various sub regions
and have developed several programs to scale up the supply of antiretroviral
agents, manage complications effectively and train various staff to perform
these functions. However, it is not just enough to have indiscriminate roll-out
of these agents to eligible individuals and have efforts made at ensuring proper
adherence and adequate follow-up. Needless to say, similar efforts are needed
to predict the possible outcomes of therapy prior to its commencement in
eligible individuals, to determine the factors that drive response in some
patients while similar responses are not seen in others. This will tend to
improve overall outcome of patients placed on antiretroviral agents.
This study will look at an HIV cohort in a district hospital in Roma, a semiurban region of Lesotho.
11
A.B. ADEBANJO 2010
Literature Review
INTRODUCTION
Human immunodeficiency virus (HIV) is a retrovirus that causes acquired
immunodeficiency syndrome (AIDS), a condition in humans in which the
immune system begins to fail, leading to life-threatening opportunistic
infections. Previous names for the virus include human T-lymphotropic virusIII (HTLV-III), lymphadenopathy-associated virus (LAV), or AIDS-associated
retrovirus (ARV).[1,2]
Infection with HIV occurs by the transfer of blood, semen, vaginal fluid, preejaculate, or breast milk. Within these bodily fluids, HIV is present as both free
virus particles and virus within infected immune cells. The three major routes
of transmission are unprotected sexual intercourse, contaminated needles, and
transmission from an infected mother to her baby during pregnancy or at birth,
or through breast milk. Screening of blood products for HIV in the developed
world has largely eliminated transmission through blood transfusions or
infected blood products in these countries.
HIV infection in humans is now pandemic in certain geographic areas . As of
January 2006, the Joint United Nations Programme on HIV/AIDS (UNAIDS)
and the World Health Organization (WHO) estimate that AIDS has killed more
than 25 million people since it was first recognized on December 1, 1981,
making it one of the most destructive pandemics in recorded history. In 2005
alone, AIDS claimed an estimated 2.4–3.3 million lives, of which more than
570,000 were children. It is estimated that about 0.6% of the world's living
population is infected with HIV.[3] A third of these deaths are occurring in subSaharanAfrica, retarding economic growth and increasing poverty.[4] According
to current estimates, HIV is set to infect 90 million people in Africa, resulting in
12
A.B. ADEBANJO 2010
a minimum estimate of 18 million orphans.[5] Antiretroviral treatment reduces
both the mortality and the morbidity of HIV infection, but routine access to
antiretroviral medication is not available in all countries.[6]
HIV primarily infects vital cells in the human immune system such as helper T
cells (specifically CD4+ T cells), macrophages and dendritic cells. HIV infection
leads to low levels of CD4+ T cells through three main mechanisms: firstly,
direct viral destruction of infected cells; secondly, increased rates of apoptosis
in infected cells; and thirdly, killing of infected CD4+ T cells by CD8 cytotoxic
lymphocytes that recognize infected cells. When CD4+ T cell numbers decline
below a critical level, cell-mediated immunity is lost, and the body becomes
progressively more susceptible to opportunistic infections. If untreated,
eventually
most
HIV-infected
individuals
develop
AIDS
(Acquired
Immunodeficiency Syndrome) and die; however about one in ten remains
healthy for many years, with no noticeable symptoms.[7] Treatment with
antiretrovirals, where available, increases the life expectancy of people infected
with HIV. It is hoped that current and future treatments may allow HIVinfected individuals to achieve a life expectancy approaching that of the general
public.
ORIGIN AND DISCOVERY
The AIDS epidemic officially began on June 5, 1981, when the U.S. Centers for
Disease Control and Prevention reported a cluster of Pneumocystis pneumonia
(PCP) caused by a form of Pneumocystis carinii, now recognized as a distinct
species Pneumocystis jirovecii, in five homosexual men in Los Angeles.[8] The
disease was originally dubbed GRID, or Gay-Related Immune Deficiency, but
health authorities soon realized that nearly half of the people identified with
the syndrome were not homosexual men. In 1982, the CDC introduced the
13
A.B. ADEBANJO 2010
term AIDS to describe the newly recognized syndrome, though it was still
casually referred to as GRID.
In 1983, scientists led by Luc Montagnier at the Pasteur Institute in France
first discovered the virus that causes AIDS.[9] They called it lymphadenopathyassociated virus (LAV). A year later a team led by Robert Gallo of the United
States confirmed the discovery of the virus, but they renamed it human T
lymphotropic virus type III (HTLV-III).[10] The dual discovery led to considerable
scientific disagreement, and it was not until President Mitterrand of France and
President Reagan of the USA met that the major issues were resolved. In 1986,
both the French and the U.S. names for the virus itself were dropped in favour
of the new term, human immunodeficiency virus (HIV).[2]
HIV was classified as a member of the genus Lentivirus,[11] part of the family of
Retroviridae.[12] Lentiviruses have many common morphologies and biological
properties.
Many
species
are
infected
by
lentiviruses,
which
are
characteristically responsible for long-duration illnesses with a long incubation
period.[13] Lentiviruses are transmitted as single-stranded, positive-sense,
enveloped RNA viruses. Upon entry of the target cell, the viral RNA genome is
converted to double-stranded DNA by a virally encoded reverse transcriptase
that is present in the virus particle. This viral DNA is then integrated into the
cellular DNA by a virally encoded integrase so that the genome can be
transcribed. Once the virus has infected the cell, two pathways are possible:
either the virus becomes latent and the infected cell continues to function, or
the virus becomes active and replicates, and a large number of virus particles
are liberated that can then infect other cells.
Two species of HIV infect humans: HIV-1 and HIV-2. HIV-1 is thought to have
originated in southern Cameroon after jumping from wild chimpanzees (Pan
troglodytes troglodytes) to humans during the twentieth century.[14][15] HIV-2
may have originated from the Sooty Mangabey (Cercocebus atys), an Old World
14
A.B. ADEBANJO 2010
monkey of Guinea-Bissau, Gabon, and Cameroon.[16] HIV-1 is more virulent. It
is easily transmitted and is the cause of the majority of HIV infections globally.
HIV-2 is less transmittable and is largely confined to West Africa.[16] HIV-1 is
the virus that was initially discovered and termed LAV.
Three of the earliest known instances of HIV-1 infection are as follows:

A plasma sample taken in 1959 from an adult male living in what is now
the Democratic Republic of Congo.[17]

HIV found in tissue samples from a 15-year-old African-American
teenager who died in St. Louis in 1969.[18]

HIV found in tissue samples from a Norwegian sailor who died around
1976.[19]
Although a variety of theories exist explaining the transfer of HIV to humans,
no single hypothesis is unanimously accepted, and the topic remains
controversial. The most widely accepted theory is so called 'Hunter' Theory
according to which transference from simian to human most likely occurred
when a human was bitten by a monkey or was cut while butchering one, and
the human became infected.[20] The Times published an article in 1987 stating
that WHO suspected some kind of connection with its vaccine program and
AIDS-epidemic. The story was almost entirely based on statements given by
one unnamed WHO advisor. The theory was supported only by weak
circumstantial evidence and is now disproven by unraveling the genetic code of
the virus and finding out that the virus dates back to the 1930s.[21]
Freelance journalist Tom Curtis discussed one controversial possibility for the
origin of HIV/AIDS in a 1992 Rolling Stone magazine article. He put forward
what is now known as the OPV AIDS hypothesis, which suggests that AIDS
was inadvertently caused in the late 1950s in the Belgian Congo by Hilary
15
A.B. ADEBANJO 2010
Koprowski's research into a polio vaccine.[22] Although subsequently retracted
due to libel issues surrounding its claims, the Rolling Stone article motivated
another freelance journalist, Edward Hooper, to probe more deeply into this
subject. Hooper's research resulted in his publishing a 1999 book, The River,
in which he alleged that an experimental oral polio vaccine prepared using
chimpanzee
kidney
tissue
was
the
route
through
which
simian
immunodeficiency virus (SIV) crossed into humans to become HIV, thus
starting the human AIDS pandemic.[23] This theory is contradicted by an
analysis of genetic mutation in primate lentivirus strains that estimates the
origin of the HIV-1 strain to be around 1930, with 95% certainty of it lying
between 1910 and 1950.[24]
EPIDEMIOLOGY
UNAIDS and the WHO estimate that AIDS has killed more than 25 million
people since it was first recognized in 1981, making it one of the most
destructive pandemics in recorded history. Despite recent improved access to
antiretroviral treatment and care in many regions of the world, the AIDS
pandemic claimed an estimated 2.8 million (between 2.4 and 3.3 million) lives
in 2005 of which more than half a million (570,000) were children.[3]
Globally, between 33.4 and 46 million people currently live with HIV.[3] In 2005,
between 3.4 and 6.2 million people were newly infected and between 2.4 and
3.3 million people with AIDS died, an increase from 2004 and the highest
number since 1981.
Sub-Saharan Africa remains by far the worst-affected region, with an estimated
21.6 to 27.4 million people currently living with HIV. Two million [1.5–
3.0 million] of them are children younger than 15 years of age. More than 64%
16
A.B. ADEBANJO 2010
of all people living with HIV are in sub-Saharan Africa, as are more than three
quarters of all women living with HIV. In 2005, there were 12.0 million [10.6–
13.6 million] AIDS orphans living in sub-Saharan Africa.[3] South & South East
Asia are second-worst affected with 15% of the total. AIDS accounts for the
deaths of 500,000 children in this region. Two-thirds of HIV/AIDS infections in
Asia occur in India, with an estimated 5.7 million infections (estimated 3.4–
9.4 million)
(0.9%
of
population),
surpassing
South
Africa's
estimated
5.5 million (4.9–6.1 million) (11.9% of population) infections, making India the
country with the highest number of HIV infections in the world.[25] In the 35
African nations with the highest prevalence, average life expectancy is 48.3
years—6.5 years less than it would be without the disease.[26]
The latest evaluation report of the World Bank's Operations Evaluation
Department assesses the development effectiveness of the World Bank's
country-level HIV/AIDS assistance defined as policy dialogue, analytic work,
and lending with the explicit objective of reducing the scope or impact of the
AIDS epidemic.[27] This is the first comprehensive evaluation of the World
Bank's HIV/AIDS support to countries, from the beginning of the epidemic
through mid-2004. Because the Bank aims to assist in implementation of
national government programmes, their experience provides important insights
on how national AIDS programmes can be made more effective.
The development of Highly Active Antiretroviral Therapy (HAART) as effective
therapy for HIV infection and AIDS has substantially reduced the death rate
from this disease in those areas where these drugs are widely available. This
has created the misperception that the disease has vanished. In fact, as the life
expectancy of persons with AIDS has increased in countries where HAART is
widely used, the number of persons living with AIDS has increased
substantially. In the United States, the number of persons with AIDS increased
from about 35,000 in 1988 to over 220,000 in 1996.[28]
17
A.B. ADEBANJO 2010
In Africa, the number of Mother-To-Child Transmission (MTCT) and the
prevalence of AIDS is beginning to reverse decades of steady progress in child
survival. Countries such as Uganda are attempting to curb the MTCT epidemic
by offering VCT (voluntary counselling and testing), PMTCT (prevention of
mother-to-child transmission) and ANC (ante-natal care) services, which
include the distribution of antiretroviral therapy.
TREATMENT
There is currently no vaccine or cure for HIV or AIDS. The only known method
of prevention is avoiding exposure to the virus. However, an antiretroviral
treatment, known as post-exposure prophylaxis is believed to reduce the risk of
infection if begun directly after exposure.[29] Current treatment for HIV infection
consists of HAART.[30] This has been highly beneficial to many HIV-infected
individuals since its introduction in 1996, when the protease inhibitor-based
HAART initially became available.[31] Current HAART options are combinations
(or "cocktails") consisting of at least three drugs belonging to at least two types,
or "classes," of anti-retroviral agents. Typically, these classes are two
nucleoside analogue reverse transcriptase inhibitors (NARTIs or NRTIs) plus
either a protease inhibitor or a non-nucleoside reverse transcriptase inhibitor
(NNRTI). Because AIDS progression in children is more rapid and less
predictable than in adults, particularly in young infants, more aggressive
treatment is recommended for children than adults.[32] In developed countries
where HAART is available, doctors assess their patients thoroughly: measuring
the viral load, how fast CD4 declines, and patient readiness. They then decide
when to recommend starting treatment.[33]
18
A.B. ADEBANJO 2010
MONITORING TREATMENT
HAART allows the stabilization of the patient’s symptoms and the suppression
of viraemia, but it does cure the patient, and high levels of HIV-1, often HAART
resistant, return once treatment is stopped.[34,35] Moreover, it would take more
than a lifetime for HIV infection to be cleared using HAART.[36] Despite this,
many HIV-infected individuals have experienced remarkable improvements in
their general health and quality of life, which has led to a large reduction in
HIV-associated morbidity and mortality in the developed world.[31,37,38] A
computer based study in 2006 projected that following the 2004 United States
treatment guidelines gave an average life expectancy of an HIV infected
individual to be 32.1 years from the time of infection if treatment was started
when the CD4 count was 350/µL.[39] This study was limited as it did not take
into account possible future treatments and the projection has not been
confirmed within a clinical cohort setting. In the absence of HAART,
progression from HIV infection to AIDS has been observed to occur at a median
of between nine to ten years and the median survival time after developing
AIDS is only 9.2 months.[40] However, HAART sometimes achieves far less than
optimal results, in some circumstances being effective in less than fifty percent
of patients. This is due to a variety of reasons such as medication
intolerance/side effects, prior ineffective antiretroviral therapy and infection
with a drug-resistant strain of HIV. However, non-adherence and nonpersistence with antiretroviral therapy is the major reason most individuals fail
to benefit from HAART.[41] The reasons for non-adherence and non-persistence
with HAART are varied and overlapping. Major psychosocial issues, such as
poor access to medical care, inadequate social supports, psychiatric disease
and drug abuse contribute to non-adherence. The complexity of these HAART
regimens, whether due to pill number, dosing frequency, meal restrictions or
other issues along with side effects that create intentional non-adherence also
contribute to this problem.[42,43,44] The side effects include lipodystrophy,
19
A.B. ADEBANJO 2010
dyslipidaemia, insulin resistance, an increase in cardiovascular risks and birth
defects.[45,46]
The timing for starting HIV treatment is still debated. There is no question that
treatment should be started before the patient's CD4 count falls below 200,
and most national guidelines say to start treatment once the CD4 count falls
below 350; but there is some evidence from cohort studies that treatment
should be started before the CD4 count falls below 350.[47,37] There is also
evidence to say that treatment should be started before CD4 percentage falls
below 15%.[48] In those countries where CD4 counts are not available, patients
with WHO stage III or IV disease[49] should be offered treatment.
Anti-retroviral drugs are expensive, and the majority of the world's infected
individuals do not have access to medications and treatments for HIV and
AIDS.[50] Research to improve current treatments includes decreasing side
effects of current drugs, further simplifying drug regimens to improve
adherence, and determining the best sequence of regimens to manage drug
resistance. Unfortunately, only a vaccine is thought to be able to halt the
pandemic. This is because a vaccine would cost less, thus being affordable for
developing countries, and would not require daily treatment.[50] However, after
over 20 years of research, HIV-1 remains a difficult target for a vaccine.[50] In
February 2007, The National Institute of Allergy and Infectious Diseases
published a report that gave details of a potential region on HIV's surface that
is a potential target for a vaccine.[51]
20
A.B. ADEBANJO 2010
PREDICTING RESPONSE TO TREATMENT
The ability to use clinical or laboratory findings to predict antiretroviral success
or failure is a singular most vital component of HIV care and management. Two
developments have spurred on this field of research. The first is the "death" of
the large clinical endpoint trial. The second is the advances in developing
surrogate markers of response.
The definition of antiretroviral response is usually based upon clinical
endpoints (i.e., infections, death, and quality-of-life) or surrogate markers (i.e.,
predictors of the clinical endpoints). For years we have used CD4+ lymphocyte
counts -- host-related factors -- as a surrogate marker. More recently, HIV-1
RNA tissue and serum levels have been employed as surrogate markers.
Currently, many new markers related to the host, the virus, or the drugs are
being evaluated for correlation between clinical status and progress, and
response to therapy. A session at this conference titled, "Predictors of Response
to Antiretroviral Therapy," highlighted some of the research currently
underway.
David Katzenstein, of Stanford University reviewed the predictive markers of
clinical endpoints in a nested case-cohort study of AIDS Clinical Trials Group
(ACTG)-175 (abstract 12124). This was a trial of zidovudine versus zidovudine
plus ddI versus zidovudine plus zalcitabine (ddC) versus ddI in asymptomatic
patients with CD4+ lymphocyte counts of between 200 to 500 cells/cu mm.
Comparison of 245 patients who progressed (ie, experienced either an AIDSdefining illness or death) versus 212 controls showed that both CD4+
lymphocyte count and HIV-1 RNA levels at week eight were predictive of
outcome.
21
A.B. ADEBANJO 2010
For example, there was a 74% reduction in risk of progression in those who
had a 1 log decline in serum HIV-1 RNA levels after eight weeks of therapy. The
researchers were able to create a table, which predicted outcomes. For
example, 95% of those who had a CD4+ count > 300 cells/cu mm and a serum
HIV-1 RNA level < 1,000 were AIDS-free after 30 months. In contrast, no
patient with a week eight serum HIV-1 RNA level > 10,000 and a CD4+
lymphocyte count < 200 cells/cu mm remained AIDS-free. Katzenstein
emphasized that HIV-1 RNA levels remained predictive at each year of
measurement. His surprising suggested application of this trend was that more
aggressive therapy may be delayed in persons with a CD4+ lymphocyte count >
200 cells/cu mm and an HIV-1 RNA < 10,000 copies/mL.
This is in contrast to what many experts recommend: that the very first
regimen should be designed to maximize suppression of HIV-1 replication in
order to circumvent the development of mutations which may lessen the
benefit of antiretroviral therapy in the future.
MEASURING RESPONSE
Regular monitoring of the CD4 count and viral load is critical to identify poor
adherence to therapy or treatment failure early. The CD4 count should be
performed every 3-6 months. The viral load should be done 6-8 weeks after
commencing antiretroviral therapy and then every 3-6 months together with
the CD4 count. The purpose of early viral load test is to detect an adequate
viral load response (more than 1 log reduction). These tests should not be done
following vaccination or if an intercurrent infection is present, as this will
transiently increase the viral load and give a falsely low value of the CD4 count.
22
A.B. ADEBANJO 2010
With HAART, at least a ten-fold (1 log) drop in the viral load can be expected
within 8 weeks and the viral load should be undetectable after 16-24 weeks of
therapy. The viral load is the most important test for monitoring response to
therapy.
[52]
The CD4 count rises rapidly within 4 weeks on starting HAART and then more
gradually. The average rise in CD4 is about 75 in the first 6 months, 150 in the
first year and 80 per annum thereafter, but this is extremely variable. In some
patients (about 10-20%) the CD4 fails to rise despite a suppressed viral load there is no point in changing their HAART regimens.
[52]
Clinical monitoring is also important, including general well-being and
sustained weight gain. Changes to therapy should not be based only on
laboratory results. It is important to note that an intercurrent clinical event
should not be an indication for changing therapy if the viral load is suppressed.
Furthermore, clinical deterioration and CD4 decline both occur after many
months of virological failure, thus the main criterion for changing initial HAART
regimen is virological failure.
Study design influences
Case definition of response
Regensberg and Whitelaw [2007] examined the trend and patterns in the viral
load at 8, 16 and 24 weeks of therapy and their findings influence the defining
of clinical, virological and immunological case response with differing extents.
According to the authors, there appears to be no gold standard for the
definition of response, and studies have reported widely varying degrees
[52].
23
A.B. ADEBANJO 2010
Bias and confounding
Selection bias is a systematic error that may occur in studies where subjects
are selected by using a procedure or by factors that may influence the
participation in a study
[53].
This type of bias comes about when the association
between exposure and disease differs between those who participate and those
who do not participate in the study. However, the use of computer-generated
case file numbers to select participants into the study and incorporating this in
the study design reduces to a large extent selection bias due to the
inappropriate selection of cases.
The misclassification of subjects can lead to information bias if the information
collected from study subjects is erroneous
[53].
When using a categorical scale,
a person may be placed in the incorrect category or misclassified as a result of
this error.
Study design choice
In the absence of randomised controlled trials (RCTs) with clinical endpoints
relevant to the individual patient, to evaluate response and the pattern of
differences of HAART non-experimental studies have been applied. Prospective
and retrospective cohort studies have been used as well as case-control
designs. Of these, the prospective cohort study has consistently been regarded
as the strongest design
[53].
Retrospective studies are usually more economical,
especially when medical records have to be reviewed to retrieve valid
information on large numbers of people. By using this study design, one can
assess the effectiveness of HAART on endpoints such as changes in CD4
pattern weight and haemoglobin. Lower costs and enhanced timelines are some
of the advantages of retrospective studies over RCTs [53].
24
A.B. ADEBANJO 2010
Motivation and aim of the study
Antiretroviral therapy is available for free in most parts of Lesotho and all
government-owned as well as religious institutions of health provide HIV allied
services such as counselling and support, health education, laboratory
investigations and treatment of opportunistic infections for free. Despite all
these measures however, not all patients respond equally to antiretroviral
agents and it is the thinking of most lay people and especially HIV-infected
persons that once antiretroviral agents have been commenced then a patient is
on his way to full recovery.
Unfortunately, health workers in the field of HIV/AIDS who sometimes
inappropriately initiate patients on treatment sometimes echo this sentiment.
Despite good measures of community based motivation and extensive
education on the concept of antiretroviral agents, there still remains that
erroneous impression that whatever time a patient is started on therapy and
regardless of duration or extent of illness, the patient was bound to recover
fully. We want to know who will respond and who wouldn't. In addition this
type of study has never been done in this part of the country.
Aim: To study the response pattern and the differences in response to
antiretrovirals in the 2 cohorts over a 12 month period.
Objectives: To assess whether anthropometric parameters, laboratory tests,
co-morbidity or co-infection with opportunistic agents, treatment regimen, IRIS
and adherence predict response to HAART as measured by CD4 count, weight
gain and functional status over a time period of one year.
25
A.B. ADEBANJO 2010
CHAPTER 2
Methods
Setting.
This study was conducted in a district hospital in Roma, a semi-urban
region of The Kingdom of Lesotho.
Study design
The study design employed for this study is the Retrospective Cohort Study
Design. Data were collected from a cohort of patients with at least 12 months
follow up and with at least 2 visits in the 12 months. The cohort was started
when the patients were put on ART and follow up for at least 12 months
starting January 2008.
Inclusion criteria

Must be HIV positive

Must be on antiretroviral therapy

Must have had a CD 4 count of 200 cells or less at initiation of treatment

Must have been on treatment for at least 12 months and have at least 2
clinic visits during the study period
Exclusion criteria

Patients with CD 4 count of > 200 cells

Patients not initiated on treatment within the study period

Defaulters of treatment (did not participate in treatment for at least 6
months).
26
A.B. ADEBANJO 2010
Patient selection and sampling frame
Sample of patients consists of a cohort of 302 adult subjects initiated on
antiretroviral therapy from January 2008. Subjects in this study are
essentially cases detected in the out-patient department, those detected
on in-patient basis and those identified during routine community-based
screening exercises. The subjects selected for the study were divided into
two arms each representing a separate group of the study based on the
HAART regimen they were placed.
Identification of cases
Subjects in the cohorts qualified as cases if they were eligible and already
started on HAART as at the time the study began and continued on treatment
regardless of hospital admission or therapy for co-morbidities during the period
from January 2008 to January 2009 as confirmed by hospital (PHC)claims
data. Patients were identified as defaulters of treatment if they did not have at
least two follow up visits during the course of this study and were not eligible
to become cases. Also excluded from the study were patients who were lost to
follow up. The ART center database was used to identify patients who died or
defaulted during this period.
Measurement of covariates
The following data was retrieved from the membership, authorisation for
medicine and hospitalisation claims databases for each of the cases:
27
A.B. ADEBANJO 2010
Demographic covariates

Age (measured on the 1st of January 2008 in years)

Gender (male or female)
Covariates indicating opportunistic infections conditions
These covariate data regarded as input variables for the purposes of this study
were determined from chronic medication authorisations and database and are
non-mutually exclusive (0 = condition not registered; 1 = condition registered).
Covariates related to therapy
This refers to the treatment modality adopted for each subject. The outcomes of
interest are a direct function of the member of the NRTI group of antiretroviral
therapy included in the patient's medication.
Data management
All patient data was captured on form designed on Microsoft Access and this
program was subsequently used to produce a data spreadsheet on Microsoft
Excel where all data cleansing and editing were done. Thereafter data was
transferred to STATA version 10 statistical computer package via STATTRANSFER version 7, for analysis.
Data analysis
Stata 10 software was used to perform the data analysis
[54].
The controls were
compared in terms of exposure status, and the variables where a significant
difference was shown were eligible to be used in the models as potential
confounders if they also differed significantly between cases and controls.
28
A.B. ADEBANJO 2010
Baseline characteristics were analysed using chi-square tests for categorical
and continuous variables. Explanatory models utilizing univariate regression
analyses were used to determine the influence of the different variables on the
odds ratios estimating response.
Multivariate
logistic
regression
analysis
was
used
to
assess
therapy
effectiveness independent of confounding factors for the combined endpoint.
To assess the fit of the models, post-estimation statistics in STATA 10 software
were used
[54].
Finally, HAART effectiveness and differences between the groups were
calculated as 1-(odds ratio) as determined by the logistic regression models.
29
A.B. ADEBANJO 2010
CHAPTER 3
Results of Study
The results of this study will be reported in the following sequence.

Patient selection

Demographic characteristics

Outcome measures in the following order; CD4 count, Weight response
and finally Haemoglobin.
The reporting of the outcome measures will essentially be done at baseline and
then at 12 months
Logistic regression modeling was used to explore the relationship between Sex,
Age, TB history, ART regimen and Opportunistic infection history (all being
explanatory variables) and CD4 count, Weight estimation and changes in
Haemoglobin concentration (being the outcome variables) at one year.
Stepwise backward binary logistic regression was carried out using STATA
version 10. Sex, TB history and Opportunistic infection history were modeled
as dichotomous variables. The variable Age was modeled first as a continuous
variable and then as a categorical variable with three categories corresponding
to the terciles for the ages. The specification for age was then selected that gave
the best goodness of fit characteristics. Variables were dropped from the model
only if the LR test was statistically significant.
The Pearson’s goodness of fit test was carried out post regression and the area
under the receiver operating characteristic (ROC) curve was calculated as well.
30
A.B. ADEBANJO 2010
PATIENT SELECTION INTO BOTH ARMS OF STUDY
Of the patients that went through the mandatory pre-HAART evaluation
leading to successful work up for HAART, 302 of them were randomly selected
from the computerized patient database to prevent selection bias and as
mentioned earlier, patients who died or did not have up to a minimum two
clinic visits during the review period were excluded from the study.
Patient enrollment
Patients randomly selected into the AZT arm
151
Patients randomly selected into the TDF arm
151
Patients who died during the review period
14*
Patients that did not have at least 2 follow up visits
9**
*Excluded from study
**Excluded and replaced in study
PATIENT DEMOGRAPHICS
At baseline there were no statistically significant differences between the two
arm of study with regards to patient demographics. (Tables 3.1 and 3.2)
Table 3.1: Patient demographics for the AZT cohort at baseline
Parameter
Mean (SD)
CD4 count
108.54 (52.71)
Haemoglobin
11.54 (1.79)
Weight
55.62 (11.96)
Median
112
11.5
54
Age
38.97
(12.16)
31
A.B. ADEBANJO 2010
Table 3.2: Patient demographics for the TDF cohort at baseline
Parameter
Mean (SD)
Median
CD4 count
117.40
(52.40)
121
Haemoglobin
11.98 (4.42)
Weight
53.92 (10.38)
11.6
54
Age
37.61
(12.36)
Table 3.3 Other considerations on the AZT cohort
Parameter
n
Gender: Female
TB history
Treatment 3: AZT+3TC+EFV
Treatment 4: AZT+3TC+NVP
Measurement n (%)
151 (100)
110 (72.85)
37 (24.5)
69 (45.70)
82 (54.30)
Table 3.4 Other considerations on the TDF cohort
Parameter
n
Gender: Female
TB history
Treatment 2: TDF+3TC+EFV
Treatment 1: TDF+3TC+NVP
Measurement n (%)
151 (100)
94 (62.25)
62 (41.06)
124 (82.12)
27 (17.88)
Treatment: All patients were placed on treatment in accordance to the
national ART guidelines which stipulates triple therapy. All patients in both
arms of the study had two NRTIs and one NNRTI. Representation by TDF 3TC
and EFV was highest (124) and 82 percent of subjects on the TDF arm of the
study, followed by AZT 3TC and NVP with 82 subjects representing 54 percent
of patients on the AZT arm of the study followed by AZT 3TC and EFV with 69
subjects (45.7%), and TDF 3TC and NVP with 27 subjects (18%), respresenting
the AZT and TDF arms respectively.
32
A.B. ADEBANJO 2010
Age: The age in the two groups were normally distributed, and the means and
standard deviations did not differ significantly (tables 3.1 and 3.2). (p = 0.05)
Gender: The majority of patients in both groups were female (tables 3.3 and
3.4). The Chi2 test confirms that the proportion of Male to Female in the two
groups does not differ significantly.
THE AZT ARM OF STUDY
The purpose of performing a univariate analysis in this study was to establish
if at the end of review period there was any statistically significant differences
between variables and to identify potential confounding factors that may
influence the results.
Firstly, it was established which variables varied significantly across the review
period in this study. Tables 3.5, 3.6 and 3.7 describe the univariate analyses in
terms of the AZT arm of the study during the review period. Variables with a
statistically significant difference (p-value ≤ 0.05) could have been potential
confounders in a statistical model if they were also associated with the
outcomes of interest.
TABLE
3.5: TABLE OF UNIVARIATE PREDICTORS OF CD4 RESPONSE AT 12 MONTHS OF
THERAPY
Variable
sex
tbhistory
n
Description
(explanation)
151
(males=41)
151
Participant’s
gender
History of TB
Odds
Ratio
CD4
RESPONSE
AT 12
MONTHS
OF
THERAPY
Log
Likelihood
Chi2
Prob
95% CI
0.37
-98.72
7.03
0.008
0.18-0.78
0.77
-101.99
0.48
0.488
0.36-1.62
33
A.B. ADEBANJO 2010
oihistory
age
art
regimen
TABLE
(present=37)
150
(present=65)
151
(3 centiles)
0-28 years
28-35 years
35-44 years
44-100 years
151
(2 groups)
artreg 3
artreg 4
History of
opportunistic
infections
Participant’s
age in groups
group 1
group 2
group 3
group 4
1.66
-100.23
2.23
0.136
0.85-3.23
Referent
0.57
1.72
0.77
Referent
-101.18
-101.28
-101.99
Referent
2.12
1.93
0.48
Referent
0.145
0.165
0.488
Referent
AZT+3TC+NVP Referent
AZT+3TC+EFV
0.66
Referent
-101.47
Referent
1.55
Referent
0.214
Referent
0.27-1.21
0.79-3.74
0.36-1.62
0.34-1.27
3.6: TABLE OF UNIVARIATE PREDICTORS OF WEIGHT RESPONSE AT 12 MONTHS
OF THERAPY
Variable
sex
tbhistory
oihistory
age
art
n
Description
(explanation)
151
(males=41)
151
(present=37)
150
(present=65)
151
(3 centiles)
0-28 years
28-35 years
35-44 years
44-100 years
151
Odds
Ratio
WEIGHT
RESPONSE
AT 12
MONTHS
OF
THERAPY
Log
Likelihood
Chi2
Prob
95% CI
Participant’s
gender
History of TB
0.42
-101.91
5.50
0.019
0.20-0.88
1.96
-103.12
3.08
0.079
0.92-4.19
History of
opportunistic
infections
Participant’s
age in groups
group 1
group 2
group 3
group 4
0.95
-103.96
0.03
0.869
0.50-1.81
Referent
0.02
0.00
0.99
Referent
0.886
0.962
0.321
Referent
Referent
0.95
1.02
1.46
Referent
-104.65
-104.66
-104.17
0.45-1.99
0.49-2.12
0.69-3.08
34
A.B. ADEBANJO 2010
regimen
TABLE
(2 groups)
artreg 3
artreg 4
AZT+3TC+NVP Referent
AZT+3TC+EFV
0.79
Referent
-104.40
Referent
0.53
Referent
0.466
Referent
0.42-1.50
3.7: TABLE OF UNIVARIATE PREDICTORS OF HAEMOGLOBIN RESPONSE AT 12
MONTHS OF THERAPY
Variable
sex
tbhistory
oihistory
age
art
regimen
n
Description
(explanation)
151
(males=41)
151
(present=37)
150
(present=65)
151
(3 centiles)
0-28 years
28-35 years
35-44 years
44-100 years
151
(2 groups)
artreg 3
artreg 4
Odds
Ratio
HB
RESPONSE
AT 12
MONTHS
OF
THERAPY
Log
Likelihood
Chi2
Prob
95% CI
Participant’s
gender
History of TB
4.04
-82.04
12.53
0.0004
1.86-8.76
1.41
-87.96
0.67
0.413
0.63-3.15
History of
opportunistic
infections
Participant’s
age in groups
group 1
group 2
group 3
group 4
0.68
-87.45
1.06
0.304
0.33-1.43
Referent
1.41
0.65
0.44
Referent
-87.96
-87.80
-86.70
Referent
0.67
0.99
3.19
Referent
0.413
0.320
0.074
Referent
AZT+3TC+NVP Referent
AZT+3TC+EFV
0.44
Referent
-85.88
Referent
4.85
Referent
0.028
Referent
0.63-3.15
0.27-1.56
0.17-1.14
0.21-0.92
35
A.B. ADEBANJO 2010
THE TDF ARM OF STUDY
At one year of treatment both arms of the study appeared similar and no
statistically significant differences between the groups (Table 3.8, 3.9 and
3.10).
TABLE
3.8: TABLE OF UNIVARIATE PREDICTORS OF CD4 RESPONSE AT 12 MONTHS OF
THERAPY
Variable
n
Description
(explanation)
sex
151
(males=57)
151
(present=62)
151
(present=60)
tbhistory
oihistory
age
art
regimen
151
(3 centiles)
0-29 years
29-37 years
37-45 years
45-100 years
151
(2 groups)
artreg 1
artreg 2
Odds
Ratio
CD4
RESPONSE
AT 12
MONTHS
OF
THERAPY
Log
Likelihood
Chi2
Prob
95% CI
Participant’s
gender
History of TB
0.87
-95.80
0.16
0.689
0.43-1.74
1.07
-95.86
0.03
0.852
0.54-2.13
History of
opportunistic
infections
Participant’s
age in groups
group 1
group 2
group 3
group 4
1.64
-94.93
1.90
0.169
0.80-3.35
Referent
1.47
1.50
0.49
Referent
-95.39
-95.43
-94.30
Referent
0.98
0.90
3.16
Referent
0.323
0.343
0.075
Referent
TDF+3TC+NVP Referent
TDF+3TC+EFV
0.40
Referent
-94.16
Referent
3.43
Referent
0.064
Referent
0.68-3.18
0.64-3.51
0.23-1.07
0.14-1.13
36
A.B. ADEBANJO 2010
TABLE
3.9: TABLE OF UNIVARIATE PREDICTORS OF WEIGHT RESPONSE AT 12 MONTHS
OF THERAPY
Variable
n
Description
(explanation)
sex
151
(males=57)
151
(present=62)
151
(present=60)
tbhistory
oihistory
age
art
regimen
151
(3 centiles)
0-29 years
29-37 years
37-45 years
45-100 years
151
(2 groups)
artreg 1
artreg 2
Odds
Ratio
WEIGHT
RESPONSE
AT 12
MONTHS
OF
THERAPY
Log
Likelihood
Chi2
Prob
95% CI
Participant’s
gender
History of TB
0.34
-91.75
8.27
0.004
0.16-0.73
1.53
-95.14
1.48
0.224
0.77-3.04
History of
opportunistic
infections
Participant’s
age in groups
group 1
group 2
group 3
group 4
1.89
-94.25
3.26
0.071
0.95-3.76
Referent
0.92
1.13
0.90
Referent
-95.86
-95.83
-95.85
Referent
0.05
0.09
0.06
Referent
0.828
0.760
0.809
Referent
TDF+3TC+NVP Referent
TDF+3TC+EFV
0.81
Referent
-95.77
Referent
0.23
Referent
0.635
Referent
0.43-1.95
0.51-2.53
0.40-2.04
0.34-1.93
37
A.B. ADEBANJO 2010
TABLE
3.10: TABLE OF UNIVARIATE PREDICTORS OF HAEMOGLOBIN RESPONSE AT 12
MONTHS OF THERAPY
Variable
n
Description
(explanation)
sex
151
(males=57)
151
(present=62)
151
(present=60)
tbhistory
oihistory
age
art
regimen
151
(3 centiles)
0-29 years
29-37 years
37-45 years
45-100 years
151
(2 groups)
artreg 1
artreg 2
Odds
Ratio
HB
RESPONSE
AT 12
MONTHS
OF
THERAPY
Log
Likelihood
Chi2
Prob
95% CI
Participant’s
gender
History of TB
2.30
-84.83
4.94
0.062
1.10-4.81
1.25
-87.12
0.35
0.556
0.60-2.58
History of
opportunistic
infections
Participant’s
age in groups
group 1
group 2
group 3
group 4
1.17
-87.21
0.17
0.678
0.56-2.44
Referent
0.63
1.45
0.95
Referent
-86.70
-86.92
-87.29
Referent
1.20
0.75
0.01
Referent
0.273
0.386
0.905
Referent
TDF+3TC+NVP Referent
TDF+3TC+EFV
5.52
Referent
-83.54
Referent
7.51
Referent
0.006
Referent
0.27-1.47
0.63-3.34
0.40-2.25
1.24-24.5
Conditional logistic regression analysis
The unadjusted odds ratio at 12 months in both arms of the study for CD4
response showed similar pattern with the exception of history of Opportunistic
Infections which appeared to suggest a modest increased risk with respect
therapy with TDF-containing ART regimen 1.64 [95% CI, 0.80 - 3.35]. There
was no difference in the pattern of response with the variable haemoglobin as
38
A.B. ADEBANJO 2010
the male gender responded significantly much less than the female gender.
Odds ratio 4.04 and 2.30 [95% CI, 1.86 - 8.76 and 1.10 - 4.81] for the AZT and
TDF arms respectively.
Then the variables that differed between the AZT and the TDF arms were
entered into each model one at a time. Variables that were not statistically
significantly different (p-value ≤ 0.05) between cases and controls were not
entered into the explanatory model.
After entering demographic variables, i.e. age and gender, the variables for TB
history, Opportunistic infection history and, finally, the variable relating to
therapy was entered into the models one by one.
AZT
Table 3.11: Adjusted odds ratios for the retained explanatory variables for
CD4 (n = 151)
CD4
response
at 1 year
sex
n
(explanation)
Description
Odds
Ratio
Standard
Error
Z
Score
Probability
95%
CI
(males=41)
0.36
0.14
-2.63
0.009
oihistory
(present=65)
1.72
0.61
1.53
0.125
0.17 –
0.77
0.86 3.45
tbhistory
(present = 37)
Participants’
gender
History of
opportunistic
infections
History of TB
0.82
0.33
-0.49
0.627
0.37–
1.81
artreg
Groups
included in
final analysis
Regimen 4
0.75
0.27
-0.80
0.424
0.371.51
ART regimen
AZT+3TC+EFV
Post-regression analysis demonstrated an area under the ROC curve of 0.66
(Figure 3.1) and the Pearson’s Goodness of fit test was statistically nonsignificant (P = 0.51) indicating satisfactory goodness of fit.
39
0.50
0.25
0.00
Sensitivity
0.75
1.00
A.B. ADEBANJO 2010
0.00
0.25
0.50
1 - Specificity
0.75
1.00
Area under ROC curve = 0.6597
Figure 3.1: ROC curve for CD4 response
Table 3.12: Adjusted odds ratios for the retained explanatory variables for
Haemoglobin (n = 151)
Haemoglobi
n response
at 1 year
sex
oihistory
n
(explanation
)
(males=41)
Description
(present=65)
History of
opportunistic
infections
Age of
participants
History of TB
age
tbhistory
(present=37)
Participants’
gender
Odds
Rati
o
6.25
Standar
d Error
Probabilit
y
95%
CI
2.86
Z
Scor
e
4.00
0.000
0.60
0.25
-1.23
0.220
0.97
0.02
-1.57
0.117
2.20
1.04
1.66
0.096
2.54
–
15.3
3
0.26
1.36
0.941.01
0.87–
5.56
40
A.B. ADEBANJO 2010
artreg
Groups
included in
final analysis
Regimen 4
ART regimen
AZT+3TC+EF
V
0.25
0.11
-3.07
0.002
0.100.60
Post-regression analysis demonstrated an area under the ROC curve of 0.77
(Figure 3.2) and the Pearson’s Goodness of fit test was statistically non-
0.50
0.25
0.00
Sensitivity
0.75
1.00
significant (P = 0.24) indicating satisfactory goodness of fit.
0.00
0.25
0.50
1 - Specificity
0.75
1.00
Area under ROC curve = 0.7703
Figure 3.2: ROC curve for Haemoglobin response.
Table 3.13: Adjusted odds ratios for the retained explanatory variables for
Weight (n = 151)
Weight
n
Description
Odds
Standard
Z
Probability
95%
41
A.B. ADEBANJO 2010
response
at 1 year
sex
(explanation)
(males=41)
Participants’
gender
History of TB
tbhistory
artreg
Ratio
Groups
included in
final analysis
Regimen 4
Error
Score
CI
0.45
0.17
-2.19
0.029
2.14
0.87
1.87
0.062
0.77
0.27
-0.75
0.451
0.20 –
0.91
0.96–
4.74
ART regimen
AZT+3TC+EFV
0.391.52
Post-regression analysis demonstrated an area under the ROC curve of 0.66
(Figure 3.3) and the Pearson’s Goodness of fit test was statistically non-
0.50
0.25
0.00
Sensitivity
0.75
1.00
significant (P = 0.83) indicating satisfactory goodness of fit.
0.00
0.25
0.50
1 - Specificity
0.75
1.00
Area under ROC curve = 0.6390
Figure 3.3: ROC curve for Weight response.
42
A.B. ADEBANJO 2010
TDF
Table 3.14: Adjusted odds ratios for the retained explanatory variables for CD4
(n = 151)
CD4
response
at 1 year
sex
n
(explanation)
Description
Odds
Ratio
Standard
Error
Z
Score
Probability
(males=57)
Participants’
gender
History of TB
0.98
0.13
-1.25
0.211
2.14
0.87
1.87
0.062
0.77
0.27
-0.75
0.451
tbhistory
artreg
Groups
included in
final analysis
Regimen 2
95%
CI
0.96 –
1.01
0.96–
4.74
ART regimen
TDF+3TC+EFV
0.391.52
Post-regression analysis demonstrated an area under the ROC curve of 0.63
(Figure 3.4) and the Pearson’s Goodness of fit test was statistically nonsignificant (P = 0.61) indicating satisfactory goodness of fit.
43
0.50
0.25
0.00
Sensitivity
0.75
1.00
A.B. ADEBANJO 2010
0.00
0.25
0.50
1 - Specificity
0.75
1.00
Area under ROC curve = 0.6284
Figure 3.4: ROC curve for TDF CD4 response
Table 3.15: Adjusted odds ratios for the retained explanatory variables for TDF
Weight response (n = 151)
Weight
response
at 1 year
sex
n
(explanation)
Description
Odds
Ratio
Standard
Error
Z
Score
Probability
(males=57)
0.27
0.12
-3.02
0.003
tbhistory
(present = 62)
Participants’
gender
History of TB
2.15
0.84
1.96
0.050
oihistory
(present=60)
History of
opportunistic
infections
ART regimen
1.84
0.66
1.67
0.095
0.90 3.78
TDF+3TC+EFV
0.87
0.41
-0.30
0.767
0.34-
artreg
Groups
included in
final analysis
Regimen 2
95%
CI
0.12 –
0.63
1.00–
4.61
44
A.B. ADEBANJO 2010
2.20
Post-regression analysis demonstrated an area under the ROC curve of 0.70
(Figure 3.5) and the Pearson’s Goodness of fit test was statistically non-
0.50
0.25
0.00
Sensitivity
0.75
1.00
significant (P = 0.25) indicating satisfactory goodness of fit.
0.00
0.25
0.50
1 - Specificity
0.75
1.00
Area under ROC curve = 0.7041
Figure 3.5: ROC curve for TDF Weight response
Table 3.16: Adjusted odds ratios for the retained explanatory variables for TDF
HB response (n = 151)
HB
response
at 1 year
sex
n
(explanation)
Description
Odds
Ratio
Standard
Error
Z
Score
Probability
(males=57)
2.08
0.80
1.91
0.056
oihistory
(present=60)
1.20
0.47
0.47
0.642
artreg
Groups
included in
Participants’
gender
History of
opportunistic
infections
ART regimen
95%
CI
0.98 –
4.43
0.60 2.57
45
A.B. ADEBANJO 2010
final analysis
Regimen 2
TDF+3TC+EFV
4.87
3.73
2.07
0.039
1.0921.87
Post-regression analysis demonstrated an area under the ROC curve of 0.65
(Figure 3.6) and the Pearson’s Goodness of fit test was statistically non-
0.50
0.25
0.00
Sensitivity
0.75
1.00
significant (P = 0.67) indicating satisfactory goodness of fit.
0.00
0.25
0.50
1 - Specificity
0.75
1.00
Area under ROC curve = 0.6551
Figure 3.6: ROC curve for TDF haemoglobin response
46
A.B. ADEBANJO 2010
CHAPTER 4
Discussion
This Study
This study was carried out to compare clinical and immunological response
patterns in patients taking AZT and TDF-containing ART regimens and to
possibly predict which of the regimen would be better and within which
population. Despite the perceived mismatch between two NRTIs it can be
concluded from the results of this study that, overall, the inclusion of AZT in
treatment regimen showed a modest protective effect over the TDF counterpart
as measured by the endpoints of the discriminative powers of the Receiver
Operating Curves of the explanatory variables being 66% 77% and 66% for
CD4, Haemoglobin and Weight respectively, and 63% 70% and 65% for the
same variables in the AZT and TDF arms of the study respectively.
After each of the a priori chosen variables, namely age and gender were tested
for effect modification; TB history in patients was identified in this study as a
potential effect modifier during statistical analysis. For this reason the analysis
was stratified and estimates of effectiveness were determined separately for
patients identified to be with or without TB during the course of the study only.
Reference must quickly be made to the fact the history of affectation with TB
was assessed at baseline only. In the population of patients with neo
tuberculosis infection during the course of this study, therapy with HAART
showed a protective effect against TB. There were 5 cases reported during this
period, 2 and 3 for the AZT and TDF arms respectively representing a mere
1.7% of the entire study population and were subsequently dropped from the
study population. The reported cases were perceived to have occurred as part
47
A.B. ADEBANJO 2010
of unmasking of latent TB infections and more remotely, the immune
reconstitution inflammatory syndrome.
From this study it seems that the role of age is vital if the expected response to
treatment is to be achieved. Patients in the age range 35 - 44 had the least
response to CD4 in the AZT arm of the study. Odds ratio 1.72 [95% CI 0.79 3.74], and this is found to be the same pattern in the TDF arm as well with
odds ratio 1.50 [95% CI 0.64 - 3.51]
Response to treatment across the study followed the expected pattern. The aim
of therapy is to achieve optimal response and the study showed this to great
deals with respect to outcomes of interest. However in the TDF arm of the
study haemoglobin response failed to show concordance with respect to
therapy. Odds ratio 5.52 [95% CI 1.24 - 24.5]. Judging by the wide Confidence
Interval the reason might be because of a low number of patients in that
category who satisfied the concept of response as defined in the study.
Possible limitations of the study
The study was done in a population of Basothos who are natives of the
Mountain Kingdom of Lesotho, a country completely embedded within South
Africa and with a relatively constant GDP and no form of economic growth over
the last 4 years. A GDP of USD1500 effectively places the country among the
poorest nations in the world. This population differ substantially from the
average South African population with respect to education level, economic
status and other socio-economic factors. It is an established fact that the
percentage of people with health insurance in the country is negligible making
holistic and comprehensive HIV care almost a non-existent feature. In addition
to this the unique terrain of the country makes it extremely difficult to achieve
good coverage with respect to HIV care. This makes facilities providing care
resource poor. The singular gold standard monitoring tool for patients on
48
A.B. ADEBANJO 2010
HAART is the viral load. However CD4 count may be used to monitor response
to therapy and alongside proper monitoring of some clinical parameters, an
effective monitoring of patients can be achieved particularly in resource-poor
settings.
In ideal settings, with regards to the study design a prospective randomised
controlled trial with possibly more patients in each arm and viral load assay
would have been better; although the cost and manpower would have been
difficult to reach with the resources that were available for this study.
49
A.B. ADEBANJO 2010
Bias and Confounding
An attempt to reduce bias as made throughout this study.
Firstly the selection of patient files into both arms of the study was by random
selection via computer-generated number thereby preventing the selection of
patients with poor care to be compared to patients with better care into the
study (Selection bias). This is evident in the absence of significant difference
between the baseline parameters.
Secondly for the AZT arm of the study, the same person audited the patient
records and entered data into the collection sheets at 12 months. The person
performing the recording for the TDF arm of the study was the same
throughout the study. Observer bias was therefore limited.
All doctors attending to patients in the clinic were blinded to which patients
were selected for the study.
Confounding by the Hawthorne effect (The non-specific beneficial effect of
taking part in research) was impossible owing to the study design.
Study results in relation to other studies
Process measures
Data from the baseline figures of this study compare well to those of studies
elsewhere in the world[55]. What is clearly different in these other studies is that
dual HIV therapy was used in some and a combination of three NRTIs was
used in some others to provide triple therapy with regards to the AZT arm.
With respect to the TDF arm, there is paucity of information regarding therapy
with NNRTIs. A study done in the USA compared responses between AZT 3TC
50
A.B. ADEBANJO 2010
and EFV with TDF FTC (Emtricitabine) and EFV. Two things readily come to
mind from these studies. No therapy with NVP and viral load assay was an
integral aspect of monitoring and final evaluation in the study. Furthermore a
CD4 cut off was never established at entry into the study by participants as
any CD4 level was acceptable at baseline. Viral load of greater than 10000
copies per ml of blood was the acceptable viral load level for entry.[56,57,58] With
respect to this study, there is evidence that the adoption of two NRTIs and one
NNRTI achieved similar response even in the event of severely compromised
immune function.
Outcome measures
Largely speaking the results of the outcomes of this study compared very
favorably with their American counterparts where resources abound and even
though the time frame for preliminary assessments differ, the end results were
quite similar. [58]
Questions arising from this study for further study.
The first question arising is: How best should the database be managed in
order to be able to generalize this study.
Secondly, with the perennial problems associated with bone marrow toxicity
leading to poor adherence to AZT-containing treatments, has this study been
able to show that the incidence is quite neglible since none of the participants
in this study developed such an event? It may be argued that this group of
patients may have been excluded owing to sampling but haemoglobin levels or
anaemia were not criteria to include or exclude patients in the study. On the
51
A.B. ADEBANJO 2010
other hand, how possible is it to determine patients that may go on to develop
anaemia secondary to AZT administration.
Thirdly, what will happen to the patients that can't even afford a hospital visit
more than once or twice in a year owing to transportation difficulties? Can this
study be employed in their care?
Conclusion
In conclusion this study succeeded in providing evidence that clinical and
immunological parameters can be combined into an effective toolkit for the
monitoring of patients and assessment of response particularly in resourcepoor settings. AZT showed a modest advantage over TDF regarding response
and outcomes of interest.
52
A.B. ADEBANJO 2010
References
1. Rick Sowadsky (1999). What is HTLV-III?. The Body. The complete HIV/AIDS
Resource [homepage on the internet]. c1996 - 2008 [cited 2008 May 7].
Available
from:
http://www.thebody.com/Forums/AIDS/SafeSex/Archive/origins/index.html
2. Coffin, J., Haase, A., Levy, J. A., Montagnier, L., Oroszlan, S., Teich, N.,
Temin, H., Toyoshima, K., Varmus, H., Vogt, P. and Weiss, R. A. (1986). "What
to call the AIDS virus?". Nature 321 (6065): 10. PMID 3010128.
3. Joint United Nations Programme on HIV/AIDS (2006). "Overview of the
global AIDS epidemic", 2006 Report on the global AIDS epidemic (PDF format).
4. Greener, R. (2002). "AIDS and macroeconomic impact", in S, Forsyth (ed.):
State of The Art: AIDS and Economics (PDF), IAEN, 49-55.
5. Joint United Nations Programme on HIV/AIDS. AIDS epidemic update, 2005
(PDF format).
6. Palella, F. J. Jr, Delaney, K. M., Moorman, A. C., Loveless, M. O., Fuhrer, J.,
Satten, G. A., Aschman and D. J., Holmberg, S. D. (1998). "Declining morbidity
and mortality among patients with advanced human immunodeficiency virus
infection. HIV Outpatient Study Investigators". N. Engl. J. Med 338 (13): 853860. PMID 9516219.
7. Buchbinder SP, Katz MH, Hessol NA, O'Malley PM, Holmberg SD. (1994).
"Long-term HIV-1 infection without immunologic progression.". AIDS 8 (8):
1123-1128. PMID 7986410.
8. CDC (1981). Pneumocystis Pneumonia --- Los Angeles. CDC. [serial on the
internet]. 2008 May 7; [cited 2008 May 7]; 30 (21): [2 screens]. Available from:
http://www.aegis.com/pubs/mmwr/1981/MM3021.html
53
A.B. ADEBANJO 2010
9. Barré-Sinoussi, F., Chermann, J. C., Rey, F., Nugeyre, M. T., Chamaret, S.,
Gruest, J., Dauguet, C., Axler-Blin, C., Vezinet-Brun, F., Rouzioux, C.,
Rozenbaum, W. and Montagnier, L. (1983). "Isolation of a T-lymphotropic
retrovirus from a patient at risk for acquired immune deficiency syndrome
(AIDS)". Science 220 (4599): 868-871. PMID 6189183.
10. International Committee on Taxonomy of Viruses. 61.0.6. Lentivirus.
National Institutes of Health. Retrieved on 2006-02-28.
11. International Committee on Taxonomy of Viruses. 61. Retroviridae.
National Institutes of Health. Retrieved on 2006-02-28.
12. Lévy, J. A. (1993). "HIV pathogenesis and long-term survival". AIDS 7 (11):
1401-1410. PMID 8280406.
13. Popovic, M., Sarngadharan, M. G., Read, E. and Gallo, R. C. (1984).
"Detection, isolation, and continuous production of cytopathic retroviruses
(HTLV-III) from patients with AIDS and pre-AIDS". Science 224 (4648): 497500. PMID 6200935.
14. Gao, F., Bailes, E., Robertson, D. L., Chen, Y., Rodenburg, C. M., Michael,
S. F., Cummins, L. B., Arthur, L. O., Peeters, M., Shaw, G. M., Sharp, P. M.,
and Hahn, B. H. (1999). "Origin of HIV-1 in the Chimpanzee Pan troglodytes
troglodytes". Nature 397 (6718): 436-441. DOI:10.1038/17130. PMID
9989410.
15. Keele, B. F., van Heuverswyn, F., Li, Y. Y., Bailes, E., Takehisa, J.,
Santiago, M. L., Bibollet-Ruche, F., Chen, Y., Wain, L. V., Liegois, F., Loul, S.,
Mpoudi Ngole, E., Bienvenue, Y., Delaporte, E., Brookfield, J. F. Y., Sharp, P.
M., Shaw, G. M., Peeters, M., and Hahn, B. H. (2006). "Chimpanzee Reservoirs
of Pandemic and Nonpandemic HIV-1". Science Online 2006-05-25.
DOI:10.1126/science.1126531.
16. Reeves, J. D. and Doms, R. W (2002). "Human Immunodeficiency Virus
Type 2". J. Gen. Virol. 83 (Pt 6): 1253-1265. PMID 12029140.
17. Zhu, T., Korber, B. T., Nahmias, A. J., Hooper, E., Sharp, P. M. and Ho, D.
D. (1998). "An African HIV-1 Sequence from 1959 and Implications for the
Origin of the Epidemic". Nature 391 (6667): 594-597. DOI:10.1038/35400.
PMID 9468138.
54
A.B. ADEBANJO 2010
18. Kolata, G.. "Boy's 1969 death suggests AIDS invaded U.S. several times",
The New York Times, 1987-10-28.
19. Hooper, E. (1997). "Sailors and star-bursts, and the arrival of HIV". BMJ
315 (7123): 1689-1691. PMID 9448543.
20. Annabel Kanabus & Sarah Allen. Updated by Bonita de Boer (2007). The
Origins of HIV & the First Cases of AIDS. AVERT (an international HIV and
AIDS charity based in the UK). Retrieved on 2007-02-28.
21. The Times, frontpage, 1987/5/11, Glenn Garvin Ana Lense-Larrauri,
Monika Leal And Megan Walters, Miami Herald Staff (2006). 25 YEARS OF
AIDS/HIV AIDS Showed The Media At Their Best, Worst And Everywhere In
Between. Media Awareness Project. Retrieved on 2007-02-27.
22. Curtis, T. (1992). "The origin of AIDS". Rolling Stone (626): 54-59, 61, 106,
108.
23. Hooper, E. (1999). The River : A Journey to the Source of HIV and AIDS, 1st,
Boston, MA: Little Brown & Co, 1-1070. ISBN 0-316-37261-7.
24. Korber B, Muldoon M, Theiler J, et al. (January 30 - February 2, 2000).
"Timing the origin of the HIV-1 pandemic". Programs and abstracts of the 7th
Conference on Retroviruses and Opportunistic Infections Abstract L5.
25. Joint United Nations Programme on HIV/AIDS (2006). "Annex 2: HIV/AIDS
estimates and data, 2005", 2006 Report on the global AIDS epidemic (PDF
format).
26. UNAIDS (2001). Special Session of the General Assembly on HIV/AIDS
Round table 3 Socio-economic impact of the epidemic and the strengthening of
national capacities to combat HIV/AIDS (PDF format).
27. World Bank (2005). Evaluating the World Bank's Assistance for Fighting
the HIV/AIDS Epidemic 2006. NEJM [serial on the internet]. 2006 May 11;
[cited
2008
May
7];
354:
[5
screens].
Available
from:
http://content.nejm.org/cgi/content/extract/354/19/1977
28. Centers for Disease Control and Prevention (1996). "U.S. HIV and AIDS
cases reported through December 1996" (PDF format). HIV/AIDS Surveillance
Report 8 (2): 1-40.
55
A.B. ADEBANJO 2010
29. (2005) in Fan, H., Conner, R. F. and Villarreal, L. P. eds: AIDS : science and
society, 4th edition, Boston, MA: Jones and Bartlett Publishers. ISBN 0-76370086-X.
30. Department of Health and Human Services (January, 2005). A Pocket
Guide to Adult HIV/AIDS Treatment January 2005 edition.
31. Palella, F. J., Delaney, K. M., Moorman, A. C., Loveless, M. O., Fuhrer, J.,
Satten, G. A., Aschman, D. J. and Holmberg, S. D. (1998). "Declining morbidity
and mortality among patients with advanced human immunodeficiency virus
infection". N. Engl. J. Med. 338 (13): 853-860. PMID 9516219.
32. Department of Health and Human Services Working Group on
Antiretroviral Therapy and Medical Management of HIV-Infected Children
(November 3, 2005). Guidelines for the Use of Antiretroviral Agents in Pediatric
HIV Infection (PDF).
33. Department of Health and Human Services Panel on Clinical Practices for
Treatment of HIV Infection (October 6, 2005). Guidelines for the Use of
Antiretroviral Agents in HIV-1-Infected Adults and Adolescents (PDF).
34. Martinez-Picado, J., DePasquale, M. P., Kartsonis, N., Hanna, G. J., Wong,
J., Finzi, D., Rosenberg, E., Gunthard, H. F., Sutton, L., Savara, A.,
Petropoulos, C. J., Hellmann, N., Walker, B. D., Richman, D. D., Siliciano, R.
and D'Aquila, R. T. (2000). "Antiretroviral resistance during successful therapy
of human immunodeficiency virus type 1 infection". Proc. Natl. Acad. Sci. U. S.
A. 97 (20): 10948-10953. PMID 11005867.
35. Dybul, M., Fauci, A. S., Bartlett, J. G., Kaplan, J. E., Pau, A. K.; Panel on
Clinical Practices for Treatment of HIV. (2002). "Guidelines for using
antiretroviral agents among HIV-infected adults and adolescents". Ann. Intern.
Med. 137 (5 Pt 2): 381-433. PMID 12617573.
36. Blankson, J. N., Persaud, D., Siliciano, R. F. (2002). "The challenge of viral
reservoirs in HIV-1 infection". Annu. Rev. Med. 53: 557-593. PMID 11818490.
37. Wood, E., Hogg, R. S., Yip, B., Harrigan, P. R., O'Shaughnessy, M. V. and
Montaner, J. S. (2003). "Is there a baseline CD4 cell count that precludes a
survival response to modern antiretroviral therapy?". AIDS 17 (5): 711-720.
PMID 12646794.
56
A.B. ADEBANJO 2010
38. Chene, G., Sterne, J. A., May, M., Costagliola, D., Ledergerber, B., Phillips,
A. N., Dabis, F., Lundgren, J., D'Arminio Monforte, A., de Wolf, F., Hogg, R.,
Reiss, P., Justice, A., Leport, C., Staszewski, S., Gill, J., Fatkenheuer, G.,
Egger, M. E. and the Antiretroviral Therapy Cohort Collaboration. (2003).
"Prognostic importance of initial response in HIV-1 infected patients starting
potent antiretroviral therapy: analysis of prospective studies". Lancet 362
(9385): 679-686. PMID 12957089.
39. Schackman BR, Gebo KA, Walensky RP, Losina E, Muccio T, Sax PE,
Weinstein MC, Seage GR 3rd, Moore RD, Freedberg KA. (2006). "The lifetime
cost of current HIV care in the United States". Med Care 44 (11): 990-997.
PMID 17063130.
40. Morgan, D., Mahe, C., Mayanja, B., Okongo, J. M., Lubega, R. and
Whitworth, J. A. (2002). "HIV-1 infection in rural Africa: is there a difference in
median time to AIDS and survival compared with that in industrialized
countries?". AIDS 16 (4): 597-632. PMID 11873003.
41. Becker SL, Dezii CM, Burtcel B, Kawabata H, Hodder S. (2002). "Young
HIV-infected adults are at greater risk for medication nonadherence".
MedGenMed. 4 (3): 21. PMID 12466764.
42. Nieuwkerk, P., Sprangers, M., Burger, D., Hoetelmans, R. M., Hugen, P. W.,
Danner, S. A., van Der Ende, M. E., Schneider, M. M., Schrey, G., Meenhorst,
P. L., Sprenger, H. G., Kauffmann, R. H., Jambroes, M., Chesney, M. A., de
Wolf, F., Lange, J. M. and the ATHENA Project. (2001). "Limited Patient
Adherence to Highly Active Antiretroviral Therapy for HIV-1 Infection in an
Observational Cohort Study". Arch. Intern. Med. 161 (16): 1962-1968. PMID
11525698.
43. Kleeberger, C., Phair, J., Strathdee, S., Detels, R., Kingsley, L. and
Jacobson, L. P. (2001). "Determinants of Heterogeneous Adherence to HIVAntiretroviral Therapies in the Multicenter AIDS Cohort Study". J. Acquir.
Immune Defic. Syndr. 26 (1): 82-92. PMID 11176272.
44. Heath, K. V., Singer, J., O'Shaughnessy, M. V., Montaner, J. S. and Hogg,
R. S. (2002). "Intentional Nonadherence Due to Adverse Symptoms Associated
With Antiretroviral Therapy". J. Acquir. Immune Defic. Syndr. 31 (2): 211-217.
PMID 12394800.
57
A.B. ADEBANJO 2010
45. Montessori, V., Press, N., Harris, M., Akagi, L., Montaner, J. S. (2004).
"Adverse effects of antiretroviral therapy for HIV infection.". CMAJ 170 (2): 229238. PMID 14734438.
46. Saitoh, A., Hull, A. D., Franklin, P. and Spector, S. A. (2005).
"Myelomeningocele in an infant with intrauterine exposure to efavirenz". J.
Perinatol. 25 (8): 555-556. PMID 16047034.
47. Wang C, Vlahov D, Galai N, et al. (2004). "Mortality in HIV-seropositive
versus seronegative persons in the era of highly active antiretroviral therapy."
J. Infect. Dis. 190: 1046–54. PMID 15319852.
48. Moore, D.M.; Hogg, R.S.; Yip, B.; et al. (2006). "CD4 percentage is an
independent predictor of survival in patients starting antiretroviral therapy
with absolute CD4 cell counts between 200 and 350 cells/μL". HIV Med 7:
383–8. PMID 16903983.
49. World Health Organisation (2006). [homepage on the internet]. WHO case
definitions of HIV for surveillance and revised clinical staging and
immunological classification; c2008 [cited 2008 May 7]. Available from:
http://www.who.int/hiv/pub/guidelines/hivstaging/en/index.html
50. Ferrantelli F, Cafaro A, Ensoli B. (2004). "Nonstructural HIV proteins as
targets for prophylactic or therapeutic vaccines". Curr Opin Biotechnol. 15 (6):
543-556. PMID 15560981.
51. BBC News (2007). [homepage on the internet]. Scientists expose HIV weak
spot; 2008 [updated 2007 Feb 15; cited 2008 May 7]. Available from:
http://news.bbc.co.uk/2/hi/health/6357287.stm
52. Regensberg L, Whitelaw C, editors. AFA clinical guidelines. South Africa:
Aid for AIDS; 2007. p. 35.
53. Rothman KJ. Epidemiology: An Introduction, 1st ed. New York: Oxford
University Press, 2002.
54. StataCorp. 2006. Statistical Software: Release 9.2. College Station, TX:
Stata Corp LP.
58
A.B. ADEBANJO 2010
55. D.A. Katzenstein, M.D. Hughes, M. Albrecht, S.H. Liou, R. Murphy, H.
Balfour, Michael Para, S. Hammer. AIDS Research and Human Retroviruses.
February 2001, 17(3): 203-210.
56. Department of Health and Human Services panel on clinical practices for
treatment of HIV infection. Guidelines for the use of antiretroviral agents in
HIV-1-infected adults and adolescents. March 23, 2004. Available at:
http://www.aidsinfo.nih.gov/guidelines/default_db2.asp?id=50
Accessed
September 18 2010.
57. BHIVA writing committee on behalf of the BHIVA executive committee.
British HIV Association (BHIVA) guidelines for the treatment of HIV-infected
adults
with
antiretroviral
therapy.
July
2003.
http://www.bhiva.org/guidelines/2003/hiv/index.html
Available
Accessed
at:
September
18 2010.
58. Robbins GK, De Gruttola V, Shafer RW, et al. Comparison of sequential
three-drug regimens as initial therapy for HIV-1 infection. N Engl J Med.
2003;349:2293-2303.
59
A.B. ADEBANJO 2010
Addendum 1
Data collection form
60
A.B. ADEBANJO 2010
DATA COLLECTION
SHEET
Baseline
3 months
6 months
12
months
Measured Parameter
CD4
HB
Clinical Stage W.H.O.
Weight
Function:
Work
61
A.B. ADEBANJO 2010
Amb.
Bed
HAART Regimen
TB History:
Pulmonary
Extrapulmonary
Past
Treatment
On Treatment
None
Co-morbidity
Opportunistic Infection History:
PCP
Cryptococcus
Others:
Bactrim
Prophylaxis
62
A.B. ADEBANJO 2010
History of Malignancy
Kaposi
Lymphoma
Adherence
IRIS
63