SUMMARY OF AVAILABLE SAFETY DATA FOR NEVIRAPINE, … · Safety review NVP, d4T, AZT, TDF 4 and whether sponsorship affects the quality and characteristics of adverse event reporting.

Safety review NVP, d4T, AZT, TDF 1

SUMMARY OF AVAILABLE SAFETY DATA FOR

NEVIRAPINE, STAVUDINE, ZIDOVUDINE AND

TENOFOVIR

OCTOBER 2009


TABLE OF CONTENTS

Introduction........................................................................................................... 3 1. Nevirapine ......................................................................................................... 5 2. Stavudine..........................................................................................................20 3. Tenofovir..........................................................................................................24 4. Zidovudine ......................................................................................................30 5. Additional reviews............................................................................................33 References............................................................................................................36 Annex 1................................................................................................................44


Introduction As part of the revision of the Who ARV Therapy (ART) Guidelines for Adults and Adolescents, a review of toxicity for four drugs was requested. The identified drugs and issues are:

1. Nevirapine (NVP) in HIV+ children, adolescents and adults with higher or unknown CD4, particularly regarding the risk of severe skin and hepatic reactions;

2. Stavudine (d4T) in HIV+ pregnant women, particularly regarding the risk of lactic acidosis;

3. Zidovudine (AZT) in HIV+ children, adolescents and adults, including pregnant women, particularly regarding anaemia and neutropenia;

4. Tenofovir (TDF) in HIV+ children, adolescents and adults, including pregnant women, particularly regarding renal toxicity (glomerular and tubular dysfunction) and osteopenia.

Medline, Embase and the Cochrane Library were searched to identify available data. The searches were reasonably broad, with the search terms including the drug name, toxicity and specific adverse events. Titles and abstracts of the articles identified by the searches were assessed to determine applicability to the issues listed above. All types of trials and studies were considered, including randomized controlled trials, observational studies, retrospective reviews, systematic reviews and case control studies. Studies addressing toxicity in resource-limited settings were a main focus of the review, however studies in other settings were included. A total of 554 potentially relevant articles were identified for nevirapine, 213 for stavudine, 333 for tenofovir and 762 for zidovudine. It was requested that the dossiers be prepared in a GRADE style, however given the characteristics of the available data it was not feasible to use GRADE tables for all available data. GRADE tables are essentially comparative in nature, requiring entry of data for an intervention and control group. This is required even for the GRADE format of “should (intervention) be used for (health problem)”. Given that many of the available studies are non-comparative (eg single arm studies) there is no clear advantage to entering such data into a GRADE table which looks to calculate either a relative or absolute effect. The wide range of available data dictated that it was not considered appropriate to meta-analyse the data given the differences in populations, outcomes, settings and comparators in the identified studies. For the same reasons there were very few cases were more than one study could be included in a GRADE table. Essentially the available data differs considerably in terms of the study populations, treatment regimens, study settings and outcomes assessed. As such it is difficult to reach general conclusions regarding the safety of each drug. Instead, the more conservative option would be to consider the safety of each drug relative to the characteristics of each study. A recent systematic review by Chowers1 assessed the quality of reporting of adverse events in publications of randomized trials of highly active antiretroviral treatment, examining whether reporting quality impacts the effect estimates for adverse events


and whether sponsorship affects the quality and characteristics of adverse event reporting. The authors found that there was a large degree of variability and a lack of standardization in the reporting of adverse events, with highly selective reporting of severity grade, occurrence threshold and attribution to study drugs. Trials sponsored by industry usually reported all or Grade 2 to 4 adverse events above an arbitrary threshold. The authors state that this tends to overrepresent common and mild adverse events while serious events are filtered by the occurrence threshold. In comparison, non-profit-sponsored trials tended to report only serious adverse events without an occurrence threshold. The review calculated relative risks for discontinuations due to adverse events and compared pooled results, finding that reporting of adverse events by intention-to-treat rather than on-treatment was significantly associated with a benefit for the experimental arm. The comparison also found that adverse event definitions were reported more in trials showing a benefit for the intervention with regard to treatment discontinuations. The authors conclude that the variability of adverse event reporting and the influence of sponsor identity obstruct the ability to chose antiretroviral treatment based on currently published data. While the evidence presented in the current report has not been assessed in terms of reporting quality or potential sponsor influence the findings of Chowers1 should be taken into consideration, particularly given the lack of homogenous data for the issues addressed in the report. While there are considerable concerns with the suitability of the available data to the GRADE format there was still an attempt made to create GRADE tables where possible. The following issues were addressed using GRADE:

- Comparison of patients with high (>250 cells/µL for women and >400 cells/µL for men) and low (<250 cells/µL for women and <400 cells/µL for men) CD4 cell count using NVP with outcomes assessed including hepatotoxicity in pregnant women, treatment-naive patients, treatment-experienced patients, stable virologically suppressed patients and occurrence of rash in stable virologically suppressed patients.

- Occurrence of serious adverse events in patients taking NVP or abacavir-containing regimens.

- Occurrence of treatment-limiting toxicity in patients taking NVP or nelfinavir-containing regimens.

- Occurrence of hepatotoxicity in treatment-naive patients taking NVP or efavirenz-containing regimens.

- Occurrence of hepatotoxicity in treatment-naive patients with CD4 cell counts >200 cells/µL taking NVP or efavirenz-containing regimens.

- Occurrence of rash in treatment-naive patients with CD4 cell counts >200 cells/µL taking NVP or efavirenz-containing regimens.

- Occurrence of rash in pregnant and non-pregnant women using NVP. - Occurrence of creatinine increase in patients taking TDF-containing regimens

compared to other regimens. - Occurrence of proteinuria in patients taking TDF-containing regimens

compared to other regimens. - Discontinuation due to serious adverse events in patients taking TDF-

containing regimens compared to other regimens. - Occurrence of Grade 3 to 4 adverse events in patients taking TDF-containing

regimens compared to other regimens.


Each table is footnoted describing the issues relevant to the included study. For AZT and d4T none of the available studies were entered into GRADE given that none provided comparative data. For these drugs, summary tables describing each study are provided along with conclusions. For NVP and TDF summary tables including all available data as well as the GRADE tables are provided. Given that the GRADE software did not provide an assessment of relative effect (relative risk) for individual studies, post-hoc analyses were conducted for studies with available data. Results of these analyses are provided in Annex 1.

1. Nevirapine A total of 554 potentially relevant articles were identified by the literature searches. Articles relevant to skin and hepatic reactions, particularly in regard to the relationship with CD4 cell count, were selected. A total of 30 articles were selected for inclusion and are summarised in Table 1.1 below. Nevirapine is associated with hypersensitivity reactions, in particular skin rash, and hepatotoxicity. Both the FDA and the EMEA in 2005 issued warnings in the nevirapine product information recommending against the use of NVP in adult men with a CD4 cell count >400 cells/µL or in adult women with CD4 >250 cells/µL, except in situations where the benefits clearly outweigh the risks. The results reported by the articles are somewhat conflicting, as some did not show differences in the occurrence of adverse events based on gender and CD4 cell count. For example, Knobel2 reported that more women and men with high CD4 cell counts experienced hepatotoxicity and skin rash compared to those with low CD4 cell counts. Knobel2 concluded that recommendations not to use NVP in drug-naïve patients on the basis of gender and CD4 cell count was not supported, however the study was based on a relatively small number of patients (n=142). Bonjoch3 reported no difference in occurrence of adverse events in females with high or low CD4 cell count and De Lazzari4 concluded that warning against NVP use in patients with high CD4 cell counts may not apply to virologically suppressed patients switching to NVP. Manfredi, in one trial with treatment naive and pre-treated patients5 (2006) and another trial with pregnant women6 (2007) found no increase in hepatotoxicity and skin rash in those with high CD4 cell count. Marazzi7 found lower occurrence of Grade 3-4 hepatotoxicity in patients with high CD4 cell count and concluded that NVP-containing regimens appear to be safe in pregnant women at all CD4 cell counts. Other trials showed increased occurrence of adverse events associated with high CD4 levels (eg, van Leth8; Mocroft9; Lyons10; Jamisse11; Hitti12). The available evidence is largely based on retrospective reviews or open-label studies, with very few randomized controlled trials providing evidence. As such, all results should be interpreted with caution, taking into consideration the potential biases of such studies and analyses. Finally, attention should be paid to studies based in resource-limited settings – while there is a good representation of studies in resource-limited settings described in the


table below, it appears that the key recommendations regarding use of NVP and CD4 cell count are based on trials in an industrialised population (eg van Leth8). Table 1.1: Summary of NVP toxicity Author Design Location/population Results Ananworanich 200513

•Post-hoc analysis of patients in randomized trial (2NN) assessing incidence, severity and treatment of rash and risk factors that may predict occurrence of rash

•Thailand •treatment-naive adult patients (n=202)

•incidence of rash was 20% with EFV, 21% with NVP twice daily, 38% with NVP once daily, 67% with EFV + NVP •multivariate analyses showed females with CD4 cell count �250 ×106 cell/l, high body mass index, rise in CD4 and ALT �34U/l at week 4 to be risk factors for rash

Bonjoch 20063 MC, OBS •Spain •patients receiving NVP-containing regimen for at least 2 years (n=613)

•no difference in occurrence of adverse events and evolution of liver enzymes in females with CD4 cell counts >250 cells/µL and <250 cells/µL •males with CD4 cell counts >400 cells/µL showed a statistically significant difference in GGT levels than those with CD4 cell counts <400 cells/µL •liver toxicity was infrequent and generally mild

Boulle 200714 RET •South Africa •treatment-naive adults receiving ART (n=2679)

•7.6% substituted off NVP due to toxicity, with most changes occurring in first few months of therapy and related to hypersensitivity (rash and transaminitis)

DART 200815 R, DB •comparison of safety and tolerability of nevirapine and abacavir in combination with zidovudine/ lamivudine

•Uganda •treatment-naive adults with CD4 <200 cells/mm3 (n=600)

•2.0% of abacavir patients and 4.7% of NVP patients experienced adverse events related to treatment •2.0% of abacavir and 4.3% of NVP patients experienced HSR 4.7% of abacavir and 10.0% of NVP-treated patients discontinued due to toxicity Authors conclude there was a trend toward lower rate of serious adverse reactions in Ugandan adults with low CD4 starting ART regimens with abacavir compared to NVP

De Lazzari 20084 MA •meta-analysis of randomized studies in which virologically suppressed patients

•4 trials with a pooled total of 410 patients included

•risk of hepatotoxicity in first 3 months was 2% in the low CD4 group and 4% in the high CD4 group (OD=1.46; 95% CI: 0.43, 4.98; p=0.54) and risk of hepatotoxicity at any point in trials was similar with HR=0.8 (95% CI:


Author Design Location/population Results switched to NVP-containing regimen with a follow-up �3 months. CD4 cell counts were considered high if >400 cells/µL for males and > 250 cells/µL for females

0.3, 2.5; p=0.80) •authors conclude that current warning against NVP use in patients with high CD4 cell counts may not apply to virologically suppressed patients switching to NVP; however sample size was small thus results should be interpreted with caution

Hitti 200412 R, OL •comparison of NVP or nelfinavir combined with zidovudine plus lamivudine

•United States •treatment-naive pregnant women (n=38)

•toxicity observed in 5% of nelfinavir-treated patients and 29% of NVP-treated patients •all adverse events among women with CD4 cell count >250 cells/µL were associated with NVP •trial suspended due to greater than expected toxicity of NVP

Ho 199816 RET •China •adult patients receiving NVP (n=8)

•5 of 8 patients (62.5%) developed NVP-related rash resulting in discontinuation of treatment •small number of patients

Jamisse 200711 OL •Mozambique •pregnant women receiving NVP, lamivudine and stavudine or zidovudine (n=146)

•severe hepatotoxicity more common in women with higher CD4 cell count, 6% vs. 0% for CD4 >250 cells/µL and <250 cells/µL, respectively •rates of skin toxicity, anaemia and peripheral neuropathy did not differ by CD4 cell count group

Joao 200617 RET •Brazil •pregnant women receiving NVP (n=197)

•toxicity occurred in 5.6% of patients (n=11) and led to discontinuation in 7 patients •no serious liver toxicity except for one case of grade 4 cholestasis •median CD4 was 344 in women without toxicities and 298 in women with toxicities •hepatitis C infection was only significant factor associated with toxicity (OR=15.61; p=0.001)

Kiertiburanakul 200818

CC •Thailand •treatment naive-patients initiated on NVP at low CD4 cell counts (n=357)

•history of drug allergy, body weight, CD4 cell count and AIDS-defining illness were statistically significantly associated with rash

Kondo 200719 RET •Brazil •pregnant women using NVP (n=133)

4 of 31 women with CD4 cell count <250 cells/µL (12.9%) and 23 of 102 women with CD4 cell counts >250 cells/µL developed adverse events •baseline CD4 cell counts, viral loads and transaminases were similar in women who developed skin rashes or liver function abnormalities •no correlation between high CD4


Author Design Location/population Results cell counts and adverse events for both cutaneous and hepatic reactions

Knobel 20082 RET •patients divided into high CD4 (women >250cells/µL; men >400 cells/µL) and low CD4 (<250 and 400 cells/µL, respectively)

•Spain •drug-naive men and women taking NVP (n=142)

•142 patients included, 62 in the high group and 81 in the low group •hepatotoxicity developed in 4.92% in high group (95% CI: 1.79%, 13.50%) and in 6.17% (955 CI: 2.73%, 13.66%) in low group (p=1.0) •skin rash developed in 6.56% in high group (95% CI: 2.67%, 15.70%) and 14.81% of low group (95% CI: 8.72%, 24.17%) with p=0.18 •authors conclude that recommendations not to use NVP in drug-naive patients on basis of gender and CD4 cell count not supported, however only a small number of patients in study

Kumarasamy 20 2007

OL •treatment experienced patients using EFV-based HAART with median CD4 cell count 463 cells/µL switch to NVP-containing HAART

•India •treatment-experienced male and female patients with elevated CD4 cell count (n=36)

•liver enzymes did not significantly change from time of initiating EFV to change to NVP and at end of 12-month follow-up

Littera 200621 OL •assessment of whether higher incidence of HSR in Sardinian patients could be linked to HLA antigens

•Sardinia •adult patients treated with NVP (n=49)

•46% of NVP-sensitive patients had HLA-Cw8 and HLA-B14(65) antigens compared with 5% of patients tolerant to NVP

Lyons 200610 RET •Ireland •pregnant women taking NVP as part of combination therapy (n=123)

•8 of 123 women developed significant hepatotoxicity •women who developed hepatotoxicity had higher pre-treatment CD4 cell counts (median 493 cells/µL compared to 310 cells/µL; p=0.01)

Manfredi 200522 OL, OBS •Italy •male and female treatment-naive patients (n=623)

•hepatotoxicity characterised by at least 2-fold increase of transaminases versus baseline was significantly linked with NVP, while there was a reduction in hepatotoxicity in EFV group •authors conclude NVP-based HAART may be more hepatotoxic than EFV-based HAART and a


Author Design Location/population Results role is played by chronic liver disorders

Manfredi 20065 OL •comparison of patients treated with either NVPO of efavirenz-based HAART

•Italy •treatment-naive; patients treated with 2-10 previous ART lines; patients on salvage therapy (n=742)

•female gender and CD4 cell count >250 cells/µL did not increase risk of hepatotoxicity

Manfredi 20076 OL •sub-group analysis of pregnant women taking NVP in 2006 study

•Italy •pregnant women who were either treatment naive, treated with 2-4 prior ART lines or on salvage therapy (n=27)

•no increase in liver toxicity, cutaneous rash in pregnant women with high CD4 levels

Manosuthi 200523 RET •Thailand •patients receiving fluconazole for cryptococcal prophylaxis (n=686)

•receiving fluconazole was not predictive of clinical hepatitis, elevated aminotransferase or skin rashes •authors conclude that initiation of NVP-based ART among Thai patients with advanced HIV receiving fluconazole is safe and well-tolerated

Marazzi 20067 RET •Africa •pregnant women treated with fixed-dose combinations of NVP-containing regimen (n=703)

•Grade 3-4 hepatotoxicity occurred in 9.4% of women with CD4 cell counts <250 cells/µL and 5.9% with CD4 >250 cells/µL •overall incidence of skin rash was 2.4% and hepatotoxicity 6.5% •authors conclude that NVP-containing regimens, in pregnant women at all CD4 cell count levels appear to be safe

Mocroft 20079 RET •Europe •treatment-experienced and treatment-naive adult patients (n=1,571)

•significantly lower risk of discontinuation of NVP in group with high CD4 cell count and treatment-experienced compared to those with high CD4 cell count and treatment-naive (RH=0.56; 95% CI: 0.34, 0.94; p=0.027) •authors conclude that NVP may be safer to initiate in treatment-experienced patients compared to treatment-naïve patients with high CD4 cell counts

Phanuphak 200724

RET •Thailand •adult patients including pregnant women receiving NVP-containing HAART (n=409)

•hepatotoxicity occurred in 15.6% of all patients and rash occurred in 16.1% of all patients •gender and baseline CD4 cell count were not associated with risk of hepatotoxicity or rash •pregnant women receiving HAART for prevention of mother-to-child transmission of HIV had significantly higher rate of hepatotoxicity than pregnant


Author Design Location/population Results women receiving HAART for therapy •authors conclude results support continued use of NVP-containing regimens as first-line treatment in developing countries, although pregnant women with high CD4 cell counts may have higher rates of hepatotoxicity

Phillips 200725 RET •assessment of hypersensitivity reaction in patients receiving NVP

•Canada •treatment-naive adult patients receiving NVP-containing regimens (n=685)

•66 of 685 patients (9.6%) met definition for HSR •no variables identified as risk factors using univariate logistic regression analysis •patients with HSR and hepatitis C co-infection had higher risk of death compared to those without HSR or hepatitis C

Pitche 200526 RET •France •adults receiving HAART regimens containing NVP (n=101)

•13.86% developed cutaneous reactions attributable to NVP •multivariate analysis showed female gender and plasma HIV RNA load >10,000 copies/mL were associated with rash

Taiwo 200627 REV •review of NVP toxicity across world

•description of occurrence of rash and hepatotoxicity with NVP •concludes that women with CD4 cell count >250 cells/mm3 have increased susceptibility to NVP toxicity

Tansuphaswadikul 200728

RET •Thailand •adult patients treated with NVP (n=206)

•incidence of NVP toxicity was 1.09/11 person-months •history of drug allergy and NVP toxicity were significantly associated (p=0.006) •concurrent anti-tuberculosis drugs significantly increased risk of NVP associated liver toxicity (p=0.001)

Timmermans 200529

RET •Netherlands •pregnant women receiving nelfinavir or NVP (n=372)

•NVP-related hepatotoxicity occurred more often in pregnant women than non-pregnant case control group •risk of NVP-associated rash was not increased in pregnant women •no major adverse events occurred and nelfinavir and NVP were well-tolerated during pregnancy

Torti 200730 RET •Italy •adults receiving NVP for first time, majority (79%) were treatment-experienced (n=905)

•hepatitis C virus reactivity was associated with 3-fold increase in risk of developing liver transaminase elevations (95% CI: 1.75, 5.3; p<0.001) while gender and CD4 cell count did not impact significantly •authors conclude most of the excess rates of raised transaminase levels in patients


Author Design Location/population Results treated with NVP could be attributed to hepatitis C co-infection

van Leth 20058 •Post-hoc analysis of patients in randomized trial (2NN) comparing regimens including NVP, EFV or both

•Asia/Australia, North America, South America, South Africa, Europe (n=1216)

•incidence of rash significantly higher in female patients with higher CD4 cell counts

Wit 200831 RET •analysis of patients who had discontinued NVP-based therapy due to hypersensitivity reactions

•Netherlands •male and female patients treated with NVP in ATHENA cohort who had discontinued NVP due to hypersensitivity reactions (n=231)

•patients with undetectable viral load at start of NVP therapy had a reduced risk of developing an HSR •pre-treated patients with low pre-ART and high CD4 cell counts and detectable VL when switching to NVP had significantly higher risk of HSR compared with treatment-naive patients who started NVP with low CD4 cell counts •pretreated patients with high CD4 cell counts and undetectable VL did not have an increased risk of developing an HSR •authors conclude that treatment-experienced patients who start NVP with low pre-ART and high CD4 cell counts and undetectable VL have similar likelihood of discontinuing NVP due to HSR as treatment-naive patients with low CD4 cell counts. •However these conclusions were based on a sample of 231 patients

NA=not applicable; R=randomized; DB=double-blind; OL=open label; RET=retrospective review; REV=review; MA=meta-analysis; CC=case control; MC=multicentre; OBS=observational; GGT=�-glutamyltransferase; HSR=hypersensitivity reaction The following issues could be addressed using the GRADE format and tables including the GRADE assessment of quality and summary of findings are provided below:

- Comparison of patients with high (>250 cells/µL for women and >400 cells/µL for men) and low (<250 cells/µL for women and <400 cells/µL for men) CD4 cell count using NVP with outcomes assessed including hepatotoxicity in pregnant women, treatment-naive patients, treatment-experienced patients, stable virologically suppressed patients and occurrence of rash in stable virologically suppressed patients.

- Occurrence of serious adverse events in patients taking NVP or abacavir-containing regimens.

- Occurrence of treatment-limiting toxicity in patients taking NVP or nelfinavir-containing regimens.


- Occurrence of hepatotoxicity in treatment-naive patients taking NVP or efavirenz-containing regimens.

- Occurrence of hepatotoxicity in treatment-naive patients with CD4 cell counts >200 cells/µL taking NVP or efavirenz-containing regimens.

- Occurrence of rash in treatment-naive patients with CD4 cell counts >200 cells/µL taking NVP or efavirenz-containing regimens.

- Occurrence of rash in pregnant and non-pregnant women using NVP.


Table G1: Question: Should NVP be used in patients with high or unknown CD4? Summary of findings Quality assessment

No of patients Effect

No of studies Design Limitations Inconsistency Indirectness Imprecision Other

considerations Low CD4

High CD4

Relative (95% CI)

Absolute Quality

Importance

hepatotoxicity (follow-up median 39 weeks; standard assays1) 2 observational

studies serious2 no serious

inconsistency no serious indirectness

serious3 none

5/79 (6.3%)

6/67 (9%)

RR 0 (0 to 0)

90 fewer per 1000 (from 90

fewer to 90 fewer)

IMPORTANT

hepatotoxicity in pregnant women 2 observational



no serious imprecision

none

12/160 (7.5%)5

40/676 (5.9%)6

RR 0 (0 to 0)


fewer to 59 fewer)

IMPORTANT

hepatotoxicity in treatment- naïve patients 1 observational




none 3/81

(3.7%)8 0/61 (0%)9

RR 0 (0 to 0)

0 fewer per 1000 (from 0 fewer to 0

fewer)

IMPORTANT

hepatotoxicity in stable virologically suppressed patients (follow-up mean 3 months) 4 randomized

trials10 no serious limitations

no serious inconsistency

no serious indirectness


none

3/133 (2.3%)8

12/277 (4.3%)9

RR 0 (0 to 0)


fewer to 43 fewer)11

ÅÅÅÅ HIGH IMPORTANT

rash in stable virologically suppressed patients 4 randomized

trials10 no serious limitations




none 17/133 (12.8%)8

28/277 (10.1%)9

RR 0 (0 to 0)




fewer to 101

fewer)12 1 for the Jamisse11 study, the hepatotoxicity was defined as ALT or AST equal to or greater than 2.5 times the upper limit of normal 2 non-randomized, non-comparative studies, hence are susceptible to bias 3 as the studies were not designed to examine differences between patients with regard to CD4 cell count and the occurrence of adverse events, the results should be interpreted with caution 4 Both the Marazzi7 and Kondo19 studies were retrospective analyses of clinical files, thus open to bias 5 This is the patient group with CD4 cell count less than or equal to 250 cells/uL 6 This is the patient group with CD4 cell count greater than or equal to 250 cells/uL 7 The Knobel 20082 study is a retrospective study and the authors note that this limits it, as well as the small number of patients (n=142) and consequent limited power 8 this is the group with low CD4 cell count, which was <250 cells/uL for women and <400 cells/uL for men 9 this is the group with high CD4 cell count, which was >250 cells/uL for women and >400 cells/uL for men 10 this reports a meta-analysis of 4 randomized trials of patients who switched to an NVP-containing regimen 11 the between-group difference was not statistically significantly different 12 combined odds ratio for hepatotoxicity, rash or death was 1.17 (95% CI: 0.56, 2.42; p=0.680) indicating no difference between low and high CD4 cell count groups Although the results vary in that in some studies there was a numerically greater occurrence of hepatotoxicity in patients with low CD4 cell count and in other studies there was a greater occurrence in patients with high CD4 cell count, the meta-analysis (De Lazzari et al., 20084) found no statistically significant difference between those with high and low CD4 cell count in occurrence of hepatotoxicity. However, these results are based on stable virologically suppressed patients and may not extend to other patient groups.


Table G.2: Occurrence of serious adverse events in patients using NVP compared to abacavir Summary of findings Quality assessment

No of patients Effect No of

studies Design Limitations Inconsistency Indirectness Imprecision Other considerations NVP abacavir Relative

(95% CI) Absolute Quality

Importance

serious adverse events (follow-up mean 24 weeks) 1 randomized

trials no serious limitations




none 14/300 (4.7%)

6/300 (2%)

RR 0 (0 to 0)

20 fewer per 1000 (from 20 fewer to 20 fewer)



Table G.3: Occurrence of treatment-limiting toxicity in pregnant women using NVP or nelfinavir Summary of findings Quality assessment



considerations NVP nelfinavir Relative

(95% CI)

Absolute Quality

Importance

treatment-limiting toxicity (follow-up median 38 weeks) 11 randomized




serious2 none

5/17 (29.4%)

1/21 (4.8%)

RR 0 (0 to 0)

48 fewer per 1000 (from 48 fewer to

48 fewer)3

ÅÅÅO MODERATE IMPORTANT

1 both NVP and nefinavir were co-administered with AZT and 3TC 2 Although the trial is randomized it was open-label and also had a very small N, with 21 patients treated with nelfinavir and 17 with NVP 3 This represents fewer adverse events associated with nelfinavir compared to NVP. The trial was halted due to greater than expected toxicity and changes in NVP prescribing information.


Table G.4: Question: Occurrence of hepatotoxicity and rash in patients using NVP or efavirenz? Summary of findings Quality assessment



considerations NVP efavirenz Relative

(95% CI)

Absolute Quality

Importance

hepatotoxicity (follow-up 18 months) 1 observational

studies1 serious2 no serious



none

147/299 (49.2%)

66/324 (20.4%)

RR 0 (0 to 0)

204 fewer per

1000 (from 204 fewer to

204 fewer)3

IMPORTANT

hepatotoxicity and CD4 cell count 1 randomized

trials4 serious5 no serious



none

148/607 (24.4%)6

79/400 (19.8%)7

RR 0 (0 to 0)

198 fewer per

1000 (from 198 fewer to

198 fewer)


rash and CD4 cell count 1 randomized

trials serious5 no serious



none

76/607 (12.5%)6

38/400 (9.5%)

RR 0 (0 to 0)


95 fewer)8


1 the publication does not indicate what other treatments were used in combination with NVP or efavirenz. In addition, some patients were completely treatment-naive and others had received previous ARV treatment but were naive to NVP or efavirenz. 2 open-label observational study thus open to bias 3 indicates there were fewer cases of hepatotoxicity (AST/ALT levels equal to or greater than 2 times the upper limit of normal) observed in patients treated with efavirenz


4 post-hoc analysis of 2NN trial 5 as this is a post-hoc analysis conducted to investigate which level of CD4 and pVL the risk for subsequent virologic failure was increased and whether there were differences in virologic efficacy between NVP and efavirenz, there is some potential for bias given the post-hoc nature. In addition, toxicity outcomes were not the primary outcomes of the analysis. 6 these results are for patients with CD4 cell count >200 7 there were fewer efavirenz-treated patients with elevated liver enzymes. The paper indicates that increased liver enzymes were associated with CD4 cell count for NVP but there was no clear relationship between this outcome and CD4 cell count for efavirenz 8 there were fewer efavirenz-treated patients reported skin rash. The paper stated that the association between CD4 cell count and rash differed significantly between NVP and efavirenz-treated patients.


Table G.5: Occurrence of rash and hepatotoxicity in pregnant and non-pregnant women using NVP1 Summary of findings Quality assessment



considerations Pregnant Non-pregnant

Relative (95% CI)

Absolute Quality

Importance

rash 1 observational




none

5/58 (8.6%)3

10/95 (10.5%)4

RR 0 (0 to 0)


105 fewer)5

IMPORTANT

hepatotoxicity 1 observational




none

11/58 (19%)3

4/95 (4.2%)4

RR 0 (0 to 0)


fewer to 42 fewer)6

IMPORTANT

1 this study compared pregnant and non-pregnant women using NVP or nelfinavir, however it did not directly compare these groups across the two drugs, hence only results for NVP are provided 2 retrospective cohort study and therefore open to bias 3 this refers to pregnant women 4 this refers to non-pregnant women 5 pregnant women experienced numerically fewer cases of rash compared to non-pregnant women, however the difference was not statistically significantly different 6 this indicates pregnant women experienced statistically significantly more cases of hepatotoxicity than non-pregnant women


2. Stavudine A total of 213 potentially relevant articles were identified by the literature searches. Articles addressing the use of d4T in women, particularly pregnant women, were selected for a review. Additional articles addressing lactic acidosis in non-pregnant women and males were also included. A total of 14 articles addressing the occurrence of lactic acidosis in patients using stavudine (d4T) are summarised in Table 2.1 below. A summary of case reports is provided in Table 2.2. As none of the studies provided data suitable to the GRADE format, ie non-comparative single-arm studies, the data was not entered into GRADE. Only 4 articles and 3 case reports are based in resource-limited settings. As such, the applicability of the results of the remaining articles should be considered. Bolhaar32, Geddes33 and Osler34, all in South African settings, report that the incidence of lactic acidosis is greater in female patients. This concurs with findings from industrialised countries. Laurent et al (2008), on the basis of a study comparing d4T and zidovudine in combination with nevirapine and lamivudine, recommended that given the observed toxicity with d4T there is support for recommendations calling for a gradual switch from d4T to zidovudine. Only one of the articles (Wade35) focused on pregnant women, as well as 2 of the case reports. As such, no conclusions can be drawn regarding the occurrence of lactic acidosis in pregnant women using d4T. Given that all articles were either retrospective reviews or open label trials, the biases associated with these types of studies should be taken into consideration when interpreting the results. Table 2.1: Summary of toxicity for d4T – lactic acidosis Author Design Location/population Results Arenas-Pinto 200336

REV •systematic review of lactic acidosis

•USA, western Europe, Australia (n=90)

•90 published cases of lactic acidosis were selected •51 of 83 patients (61%) were taking d4T •using data on numbers of HIV-infected individuals in US, risk of lactic acidosis could be 2.5 times greater for women than men •authors conclude nucleoside reverse transcriptase inhibitors and female gender appear to be risk factors for development of lactic acidosis

Blanco 200337 RET •NR •percentage of patients with lactate concentration �2mmol/L significantly higher in patients receiving d4T plus ddl (71%) compared to those receiving AZT plus 3TC (10%)


Author Design Location/population Results Bolhaar 200732 RET •South Africa

•adult patients taking HAART with lactate levels �4.5mmol/L (n=1735)

•23 of 1735 patients experienced lactic acidosis •overall incidence was 10.6 cases per 1000 patient-years, with incidence 16.1 case per 1000 patient-years in females and 1.2 cases/1000 patient-years in males •of patients experiencing lactic acidosis and hyperlactatemia, 66 of 67 were receiving d4T

Boubaker 200138 RET •Switzerland •adult patients receiving ART

•use of d4T with or without ddl was a risk factor for increased serum lactate levels

Geddes 200633 OBS •South Africa •adult patients taking HAART (n=891)

•in 891 patients taking HAART, 14 cases of lactic acidosis were reported, giving incidence rate of 19 cases per 1000 person-years •all cases were female •all cases were in patients taking d4T, 3TC and one non-NRTI

Hocqueloux 200339

RET, CC NR •patients with increased lactic acid concentrations were more likely than controls to receive d4T or ddl

Huynh 200340 OL NR •elevated lactic acid levels in 178 patients out of 848 (21%) receiving HAART •significant correlation between d4T or RTV-based regimen and occurrence of hyperlactatemia

John 200141 OL •Australia •patients taking HAART

•risk of hyperlactatemia significantly associated with d4T

Lactic Acidosis International Study Group 200742

RET, CC •10 countries (unspecified) ‘•adult patients taking ART (n=330)

•lactic acidosis associated with exposure to ddl in all categories of exposure duration but most strongly associated with exposure <12 months after adjustment for age, gender, CD4 cell count •apart from exposure to d4T or ddl, age above 40 years, female gender and advanced immunosuppression were independent associations with lactic acidosis

Laurent 200843 OL •assessment of efficacy and safety of patients receiving NVP and 3TC plus d4T or AZT

•Cameroon •adult patients receiving first-line ARV treatment (n=169)

•incidence rates of adverse events were 49.1/100 person-years in patients treated with d4T and 41.7/100 person-years in patients treated with AZT •elevated ananine aminotransferase activity more frequent in d4T group (85 vs. 2%) •5 patients developed peripheral neuropathy and 1 lipodystrophy syndrome, all were treated with d4T •authors suggest that in view of d4T toxicity, there is support for recommendations calling for


Author Design Location/population Results gradual switch from d4T to AZT-based regimens

Makinson 200844 REV •review of toxic effects of d4T, particularly in resource-limited settings

•worldwide •authors recommend initiating or switching to less toxic nucleoside analogues when possible, or lowering d4T dose to 30mg/bd

Osler 200934 CC •South Africa •Female gender (OR=23.4; 95% CI: 4.0–136.6), a baseline weight between 60 and 75kg (OR=4.5; 95% CI: 1.4–14.1) or �75kg (OR=19.4; 95% CI: 4.1–82.5) at ART initiation and gaining �6 kg in the first 3 months on therapy (OR=3.5; 95% CI: 1.3–9.5) were independent risk factors identifying patients who may subsequently develop symptomatic hyperlactatemia and lactic acidosis

Tien 200745 RET •USA •women taking d4T (n=1432)

•women demonstrated decreases in weight and body circumference measurements over time whether continuing or discontinuing d4T treatment

Wade 200435 OL •USA •pregnant women receiving d4T and lamivudine

•no toxicities in women or their infants which required discontinuation or modification of d4T

NA=not applicable; NR=not reported; OL=open label; RET=retrospective review; REV=review; CC=case control; OBS=observational; HAART=highly active antiretroviral therapy Table 2.2: Summary of case reports of lactic acidosis and hyperlactatemia associated with d4T use Author Design Location/population Results Falco 200346 CR, REV NR •female gender, pregnancy and

higher body mass index were significantly associated with greater risk of developing severe hyperlactatemia

Gerard 200047 CR NR •14 adult patients developed hyperlactatemia and all were receiving d4T

Johri 200048 CR •women receiving d4T and lamivudine (n=3)

•describes occurrence of lactic acidosis in 3 women receiving d4T and 3TC, 2 of whom died.

Lonergan 200349 CR NR •10 patients treated with d4T reported hyperlactatemia

Mandelbrot 200350 CR •France •pregnant women receiving d4T and ddl (n=2)

•case reports of two women receiving d4T who developed lactic acidosis •authors conclude that the mitochondrial tendencies of nucleoside analogues should be considered when choosing


Author Design Location/population Results treatment regimens

Nelson 200851 CR •Kenya •woman with advanced HIV taking regimen including d4T and 3TC (n=1)

•patient had Fanconi syndrome and concurrent lactic acidosis •authors conclude that d4T and 3TC should be added to list of ARVs that can induce Fanconi syndrome

Sarner 200252 CR •Uganda •pregnant woman treated with d4T and didanosine (n=2)

•patients developed lactic acidosis and died •authors conclude that it is unclear whether pregnancy is a risk factor for lactic acidosis

Songa 200753 CR •Uganda •patients receiving combination therapy who developed hyperlactatemia

•24 patients had documented lactic acidosis •all were receiving combination therapy that included d4T •835 of patients were female

Tripuraneni 200454 CR; REV •USA •patients with lactic acidosis (n=58)

•zidovudine associated with higher lactate levels and mortality than d4T and 3TC; association of zidovudine with increased mortality confounded with effect of time

REV=review; CR=case report; HAART=highly active antiretroviral therapy


3. Tenofovir A total of 333 potentially relevant articles were identified by the literature searches. For this summary 18 articles were selected for review and are provided in Table 3.1 below. There are relatively few articles addressing the renal toxicity of tenofovir used in patients in resource-limited settings, with only two relevant publications. Three of the studies were suitable for entry into GRADE tables, see below following the overall summary. A review of tenofovir safety by Sax55 stated that the cumulative incidence rate of nephrotoxicity associated with the use of tenofovir is 1% to 4%, with estimated rate of Fanconi syndrome being 0.5% to 2%. Gender, age and race are have not been demonstrated to be associated with tenofovir-induced nephrotoxicity. Roling56 et al (2006) state that severe tubular dysfunction seems to be rare and decreases in glomerular filtration rate have not been clinically significant, particularly in treatment-naive patients. However, given the relative lack of trials and/or reviews addressing toxicity of tenofovir in resource-limited settings, no conclusions can be drawn regarding whether the observed effects in industrialised settings occur in a similar manner in resource-limited settings. A 2007 report of all postmarketing adverse drug reactions up to April 2005 for 10,343 patients in developed countries (Nelson57) using tenofovir reported that renal serious adverse events were observed in 0.5% of patients and graded elevations of serum creatinine were observed in 2.2% of patients. Risk factors for increased serum creatinine were concomitant nephrotoxic medications, elevated serum creatinine, low body weight, advanced age and lower CD4 cell count. Based on post-marketing data for 455,392 person-years of exposure to tenofovir the most commonly reported serious adverse events were renal events. The authors conclude that the data demonstrate a favourable safety profile for tenofovir and that risk factors for nephrotoxicity can be identified and may be useful in managing patients at risk. A recent publication by Gallant and Moore58 (2009) described renal outcomes in patients who initiated treatment with tenofovir or an alternate NRTI. The authors report no difference between tenofovir and NRTI-treated patients in estimated glomerular filtration rate and conclude that the evidence is consistent with clinical trials which show no indication of renal toxicity when tenofovir is used as part of an initial regimen. The studies listed in the table below, with the exception of the Squires59 trial, are either open-label or observational studies. As such, there is potential for bias and results should be interpreted with caution. The Arribas60, Fux61, Moreno62 and Squires59 are reasonably-sized studies, while the remainder have small n’s. There is only one paper assessing the use of tenofovir in pregnant women (Nurutdinova et al., 200863). This paper found that tenofovir was well-tolerated in a population of pregnant women. However the population was small (n=15) thus results may not apply to a wider population. Table 3.1: Summary of articles addressing toxicity of tenofovir Author Design Location/population Results


Author Design Location/population Results Ananworanich 200864

OL •patients treated with d4T/ddl were switched to TDF/3TC while receiving saquinavir/ ritonavir

•Thailand •patients with full HIV RNA suppression treated with d4T/ddl (n=35)

•significant reduction in lipids and lactate, but no change in liver enzymes for patients switched to TDF/3TC •reversal of lipoatrophy

Arribas 200860 R, OL •comparison of TDF, FTC and EFV with 3TC/AZT and EFV

•USA •adult treatment-naive patients (n=511)

•more patients in the AZT/3TC arm discontinued therapy because of adverse events (11% vs. 5%; p=0.01) •no patients discontinued because of renal events

Cassetti 200765 OL •extension study of randomized, double-blind trial comparing TNF to d4T in combination with 3TC and EFV in treatment-naive patients

•Brazil, Argentina, Dominican Republic (n=86)

•no patient discontinued due to renal adverse events •no patient developed Fanconi syndrome or renal tubulopathy during double-blind or extension phases

Cirino 200666 RET •USA (n=40)

•40 patients identified from individual safety reports who developed hypokalemia associated with the use of TDF •risk factors included RTV or ddl use, lower weight and length of TDF use

Elion 200867 OL, MC •assessment of ATV/r and TDF/FTC

•USA •adult treatment-naive patients (n=100)

•6 patients discontinued due to adverse events •Grade 4 hyperbilirubinemia occurred in 5% of patients

Fux 200861 OL •Switzerland Adult patients initiating, re-initiating or discontinuing ARV treatment with and without TDF (n=1091)

•statistically significant association between TDF and increased serum alkaline phosphatise. This indicates stimulated bone turnover, and when considered along with TDF-related renal phosphate loss raises concerns that TDF use could result in osteomalacia with a loss of bone mineral density

Gupta 200868 RET •USA (164)

•case series of 164 patients who had Fanconi syndrome •majority (83%) received PIs plus TDF

Hazra 200569 OL 6-day course of TDF monotherapy followed by combination regimen

•USA •treatment-experienced children aged >4 and <18 years (n=18)

•>6% decrease in bone mineral density in 5 of 15 patients, for which 2 patients discontinued therapy •during monotherapy 2 patients developed Grade 3 hepatic transaminase elevations and were


Author Design Location/population Results including TDF removed from the study

Irizarry-Alvarado 200970

CR USA (n=3)

•report of 3 patients who developed Fanconi syndrome and nephrogenic diabetes insipidus with an ARV regimen including TDF and ddl

Mauss 200571 CSect, OL •comparison of TDF-treated patients with patients on ARV therapy never treated with TDF

•Germany (n=174)

•24 patients on TDF (30%) had proteinuria >130mg/day compared to 5 (5%) of control patients (p<0.001) •patients on TDF showed lower mean glomerular filtration rate estimated by creatinine clearance or cystatin C clearance compared to control patients (p<0.05) •impaired GFR based on creatinine clearance in 46% of TDF patients compared to 32% of control patients (p=0.04)

Moreno 200662 OBS, MC •observational study to identify the most frequently-occurring toxicities with NRTI treatment

•Spain •treatment-experienced adult patients (n=1286)

•5 patients (0.39%) experienced renal dysfunction and discontinued TDF treatment Authors conclude that risk of developing nephrotoxicity related to TDF leading to discontinuation is rare and usually occurs when other risk factors exist

Nelson 200757 RET •Europe, Australia, North America (n=10,343)

•renal serious adverse events observed in 0.5% of patients •serum creatinine elevations reported in 2.2% of patients •based on post-marketing safety data the most commonly reported serious adverse event were renal events •authors conclude data demonstrates a favourable safety profile for TDF

Nurutdinova 200863

RET •USA •pregnant women using tenofovir as part of HAART (n=15)

•11 women (73%) had abnormal laboratory results •GFR remained above 90mL/min/1.73m2 in all women except one •no mother-to-child transmission of HIV

Padilla 200572 RET, CC •US (n=316)

•serum creatinine elevations occurred in 4.1% of patients treated with a tenofovir-containing regimen compared to 0.5% treated with other ARV regimens •2 patients discontinued TDF treatment due to creatinine elevation

Pozniak 200673 R, OL •comparison of TDF, FTC and EFV versus 3TC, AZT and EFV

•US •treatment-naive patients (n=517)

•two treatment groups had comparable renal safety •decreases in GFR in both treatment groups at week 96 •no confirmed creatinine abnormalities and no cases of


Author Design Location/population Results Fanconi syndrome

Riordan 200974 RET UK and Ireland •HIV-infected children (n=159)

•12 (7.5%) experienced serious adverse events and stopped TDF treatment •5 had renal toxicity

Squires 200359 R, DB, PC, MC •North America, Europe, Australia •adult patients with detectable viral replication despite current therapy (n=552)

•incidence of Grade 3 or 4 adverse events was similar for TDF (13%) and placebo (14%) groups. •laboratory abnormalities higher in placebo group •no evidence of TDF-related toxicity •no withdrawals due to serum creatinine or phosphorous level

Young 200975 CS •change in renal profile in patients with pre-existing renal dysfunction treated with TDF

•USA (n=19) •median 12-month change in estimated creatinine clearance from baseline of -0.3 mL/min (range, -32.2 to +23.6) and the median change in GFR of -0.1 mL/min/1.73 m2 (range, -49.8 to +29.5) •confirmed worsening of kidney disease in 5 of 19 patients (26.3%) •authors conclude TDF use can be considered in patients with pre-existing or current renal dysfunction and can be closely monitored for worsening of renal function

NA=not applicable; R=randomized; DB=double-blind; OL=open label; RET=retrospective review; REV=review; CC=case control; CS=cohort study; CSect=cross-sectional; MC=multicentre; OBS=observational; PI=protease inhibitor; GFR=glomerular filtration rate The following issues could be addressed using the GRADE format and tables including the GRADE assessment of quality and summary of findings are provided below:

- Occurrence of creatinine increase in patients taking TDF-containing regimens compared to other regimens.

- Occurrence of proteinuria in patients taking TDF-containing regimens compared to other regimens.

- Discontinuation due to serious adverse events in patients taking TDF-containing regimens compared to other regimens.

- Occurrence of Grade 3 to 4 adverse events in patients taking TDF-containing regimens compared to other regimens.


Table G.6: Occurrence of adverse events in patients using TDF-containing regimens compared to other regimens Summary of findings Quality assessment



considerations

TDF-containing regimens

other ARV

regimens

Relative (95% CI)

Absolute Quality

Importance

creatinine increase 1 observational




none 5/122 (4.1%)

1/194 (0.5%)2

RR 0 (0 to 0)


fewer to 5 fewer)

IMPORTANT

proteinuria 1 observational



serious4 none 24/82

(29.3%) 5/92

(5.4%) RR 0 (0

to 0)


fewer to 54 fewer)

IMPORTANT

discontinuation due to adverse events 15 randomized

trials no serious limitations6




none 12/231 (5.2%)

29/232 (12.5%)

RR 0 (0 to 0)

125 fewer per 1000 (from

125 fewer to 125 fewer)


Grade 3 or 4 adverse events 17 randomized





none 49/368 (13.3%)

25/182 (13.7%)

RR 0 (0 to 0)

137 fewer per 1000 (from

137 fewer to 137 fewer)8


1 retrospective case control study (Padilla et al., 2005), thus open to bias. Also, a relatively small n, thus results should be interpreted with caution 2 statistically significant difference in occurrence of elevated creatinine levels between tenofovir -treated patients and those on other regimens (p=0.018) 3 The Mauss et al (2005) trial was a cross-sectional study pharmacokinetic comparing patients treated with TDF to those never treated with TDF 4 As the Mauss study was a PK study with assessments taken after 24 hours, it may not represent the effects of TDF over a longer time period 5 Comparison of TDF with FTC and EFV versus AZT plus 3TC and EFV in treatment-naive patients (Arribas et al., 2008) 6 The Arribas et al (2008) trial, while randomized, was open-label and thus susceptible to bias


7 Squires et al (2003), a randomized, double-blind trial comparing TDF and placebo in patients with detectable viral replication despite current ARV therapy 8 There was no significant difference between patients receiving TDF and those receiving placebo in the occurrence of serious adverse events While the Arribas et al (2008)60 study indicated that there was greater discontinuation in tenofovir-treated patients, the Squires et al (2003)59 trial found no significant difference in the occurrence of Grade 3 or 4 adverse events between patients taking tenofovir as part of their ARV regimen or other regimens. However, it should be noted that these trials are in different populations (Arribas et al (2008)60 in treatment-naive patients and Squires et al (2003)59 in treatment-experienced patients).


4. Zidovudine A total of 762 potentially relevant articles were identified by the literature searches and of these, 14 were selected for review and are included in the table below. Lower body mass and lower CD4 cell count appear to be factors related to risk of developing anaemia (Ssali76; Isaakidis77). A trial assessing the prevention of mother-to-child HIV transmission (Sperling et al., 199878) reported that anaemia was significantly more common in infants receiving AZT compared to those receiving placebo (19% versus 8%; p=0.001) however the authors concluded that the trial did not identify any problems that would alter the recommendations for the routine use of AZT for the prevention of mother-to-child HIV transmission. No other trials specifically addressed the occurrence of anaemia in pregnant women using AZT. A greater decrease in haemoglobin levels in patients using AZT and also being treated for hepatitis C has resulted in the recommendation that AZT is not used in patients receiving treatment for hepatitis C79. Table 4.1: Summary of occurrence of adverse events associated with zidovudine in resource-limited settings Author Design Location/population Results Alvarez 200580 RET •USA

•HIV and hepatitis C co-infected patients (n=217)

•mean decrease in haemoglobin was significantly greater for those using AZT compared to those not taking AZT (p<0.0001) •use of AZT associated with an increased risk of anaemia

Aurpibul 200881 RET •children aged 2-15 years receiving HAART with NVP or EFV plus 3TC and d4T and had switched from d4T to AZT

•Thailand •children receiving HAART (n=78)

•following switch from d4T to AZT in children there were statistically significant decreases in haemoglobin level, WBC count and ANC, but decreases were not clinically significant as no patient had clinical symptoms of anaemia, leukopenia or neutropenia

Bae 200882 NC •nested cohort within randomized controlled trial to compare toxicities in infants who had or had not been exposed to maternal HAART

•Botswana •infants who had or had not been exposed to maternal HAART in Mashi study (n=69)

•21.7% of HAART-exposed infants were neutropenic compared to 5.5% of infants exposed to AZT (p<0.01) indicating that in-utero exposure to HAART was associated with increased risk of neutropenia •neutropenia no longer associated with antenatal exposure to HAART after 1 month of age •postnatal exposure to HAART not associated with haematologic or hepatic toxicities

Curkendall 200783 RET •USA •AZT-naive patients (n=1649)

•incidence of anaemia was 24.3 per 100 person-years for patients initiating AZT compared to 8.1 per 100 person-years for those using a regimen not including AZT •patients initiating treatment with


Author Design Location/population Results AZT at greater risk of anaemia than those initiating treatment excluding AZT

Huffam 200784 RET, OBS •analysis of factors associated with anaemia and time to discontinuation of treatment in patient commencing AZT or switching from d4T

•Asia-Pacific region •adult patients in TAHOD (TREAT Asia HIV Observational Database) cohort taking ART containing 3 or more agents (n=433)

•anaemia occurred in 57 of 433 (13%) of patients treated with AZT in first 6 months of treatment •baseline anaemia was strongest predictive factor for subsequent anaemia and anaemia reported less frequently in those on AZT who had prior ART experience •patients who commenced treatment with AZT more likely to stop within first 9 months of treatment than those who commenced d4T

Idoko 200285 OL •non-comparative trial of triple regimen including nelfinavir, zalcitabine and zidovudine

•Nigeria •adult patients with CD4 cell count between 100-500 cells/mm3 (n=40)

•two patients (5%) withdrew due to adverse events •one patient reported anaemia, but did not withdraw

Isaakidis 200877 RET, OBS •adults systematically switched from d4T to AZT-based HAART

•Cambodia •adults receiving d4T-based HAART (n=527)

•within one year of switch, 21.9% of patients developed anaemia Grades 1 to 4 and 7.1% developed severe anaemia Grades 3 to 4 •low body mass index and low CD4 cell count <200 cells/µL were factors associated with anaemia •authors conclude switch to AZT led to satisfactory clinical outcomes however there was increased burden of anaemia

Moh 200586 R, OL •pre-randomisation phase of trial in which patients received continuous standard HAART regimen including AZT. All patients also received cotrimoxazole

•Cote d’Ivoire •adult treatment-naive patients (n=498)

•118 of 498 patients had Grade 3-4 neutropenia (56.3/100 person-years), 23 had Grade 3-4 anaemia (9.6/100 person-years) •one patient discontinued AZT due to neutropenia and 11 discontinued due to anaemia •authors note that almost all of the severe neutropenia disappeared after cotrimoxazole was stopped, which suggests an accentuated interaction between cotrimoxazole and AZT

Moyle 200487 MA •randomized trials comparing AZT and d4T in triple-therapy regimens (n=6 trials)

•haemoglobin levels decreased with AZT treatment and increased with d4T treatment •all grades of anaemia and neutropenia were more common with AZT-based regimens compared to d4T-based regimens •the occurrence of Grade 1 to 4


Author Design Location/population Results neutropenia ranged from 26% to 43% in the AZT trials and 15% to 31% in the d4T trials

Nunez 200879 OL •Spain •patients co-infected with HIV and hepatitis C (n=389)

•mean decrease in haemoglobin from baseline to week 4 significantly greater in patients receiving AZT compared to other drugs (-3.09 vs. -2.3g/dL; p<0.001) •lower baseline haemoglobin and greater haemoglobin drops during first 4 weeks of therapy were independent predictors of moderate anaemia •authors advise that given the availability of other antiretrovirals, it is advised to avoid AZT treatment while receiving hepatitis C treatment

Sperling 199878 R, DB •assessment of safety of maternal-infant AZT regimen

•USA, France •pregnant women with CD4 cell counts >200×106/l and no maternal indications for antiretroviral therapy (n=424)

•anaemia significantly more common in infants receiving AZT (19%) compared to those receiving placebo (8%), p=0.001 •no association between AZT use and newborn structural abnormalities

Ssali 200676 •post-hoc analysis of patients using AZT in DART trial

•Uganda •treatment-naive adult patients with CD4 cell count <200 cells/mm3 (n=3,314)

•219 patients (6.6%) developed Grade 4 anaemia by week 48 •female patients and those with lower haemoglobin, lower CD4 cell count and lower body mass index at baseline were at significantly higher risk (p<0.05) for anaemia •authors conclude rate of anaemia is higher than that observed in studies from industrialised countries, likely due to population characteristics and higher rate of concurrent HIV-related clinical events

Violari 200888 R, OL •infants were randomized to receive lopinavir-ritonavir, zidovudine and lamivudine therapy as early therapy for 96 weeks or 40 weeks or as deferred therapy

•South Africa •infants 6 to 12 weeks of age with HIV infection (n=377)

•22 of 377 children (total) had neutropenia (13%) and 7 (4%) had anaemia •4 children switched from AZT to d4T due to neutropenia or anaemia

Willig 200989 CS •determine factors associated with

•Peru •treatment-naive adult patients (n=546)

•baseline weight <60kg was found to more than double risk of early discontinuation of AZT-containing regimens (HR=2.52; 95% CI:


Author Design Location/population Results abbreviated first regimen durability •determine factors associated with toxicity-related discontinuation of AZT

1.46, 4.35) •discontinuation rates increased with lower baseline weights, with 14% discontinuation for 70kg, 16% for 60-70kg, 26% for 50-59k and 44% for 50kg •AZT rather than ddl/d4T as part of an NRTI backbone was found to increase risk of discontinuation up to 120 days (adjusted HR=2.09; 95% CI: 1.22, 3.57) however after 120 days use of AZT was found to decrease risk of regimen discontinuation compared to ddl/d4T (adjusted HR=0.52; 95% CI: 0.28, 0.95)

NA=not applicable; NR=not reported; MA=meta-analysis; OL=open label; R=randomized; DB=double-blind; RET=retrospective review; REV=review; CC=case control; OBS=observational; CR=case report; CS=cohort study; HAART=highly active antiretroviral therapy; NC=nested cohort

5. Additional reviews There are a number of reviews of the use of ARVs and associated toxicity in limited-resource settings. These reviews are not particular to a specific drug, but instead cover a range of drugs. They are relevant in that they address the occurrence of adverse events in populations in resource-limited settings. The key results and conclusions of the reviews are summarised in the table below. Table 5.1: Additional reviews Author Design Key results/conclusions Calmy 200690 •assessment of efficacy

and safety of generic fixed dose combination ARV treatment in resource-poor settings •observational cohort using fixed dose combination of 3TC, d4T and NVP

•NVP responsible for discontinuation for 42 of 655 patients, due to initiation of rifampicin-based antituberculosis treatment (n=15), skin toxicity (n=12), liver toxicity (n=11) or unknown (n=4). •9 patients discontinued due to d4T toxicity – 5 for severe neuropathy, 2 for lipodystrophy and 2 for unknown reason

Forna 200791 •assessment of clinical toxicity in patients receiving ART in Uganda as part of a home-based AIDs care program •treatment consisted of d4T plus 3TC plus NVP (96% of patients) or EFV (4%).

•toxicities developed in 411 (40%) of patients •peripheral neuropathy 36%, rash 6%, hypersensitivity reaction 2%, acute hepatitis, anaemia, pancreatitis or lactic acidosis �0.5% and other 13% •no patients discontinued treatment due to toxicity

Kumarasamy 200892

•report of adverse events occurring with generic HAART treatment in a population in southern India

•most common adverse events following 3 months treatment were rash (15.2%), peripheral neuropathy (9.0%), clinically significant anaemia (5.4%), hepatitis (3.4%). •women were significantly more likely to experience lactic acidosis •in patients with 1 year follow-up, NVP was


Author Design Key results/conclusions associated with development of rash and d4T associated with development of peripheral neuropathy

Manosuthi 200893

•assessment of mortality rate and risk factors after hyperlactatemia in patients receiving ART •retrospective review of patients who were diagnosed with symptomatic hyperlactatemia

•hyperlactatemia was associated with d4T in 114 (91.2%) of patients, 5 (4.0%) using d4T+ddl, 4 (3.2%) using ZDV+ddl and 2(1.6%) using ddl •patients with low body weight and high serum lactate level were at a higher risk of mortality

Murphy 200794 •review of adverse events associated with ART and how to deal with these events in resource-limited settings

•key events identified are lactic acidosis, lipoatrophy, pancreatitis, peripheral neuropathy, hyperlipidaemia, anaemia, hypersensitivity, hepatotoxicity, fat accumulation, insulin resistance,

Nuesch 200695 •retrospective review of randomized controlled trials of ART in Thailand, assessing toxicity of ART

•over 3.7 years there were 142 Grade 3-4 toxicities in 101 patients (24.2%). Hepatic toxicity occurred in 33 (7.9%), hypercholesterolaemia in 57 (13.7%) and anaemia in 16 (3.8%) •ARVs used included ZDV, d4T, ddl, EFV, SQR, RTV, IDV

Palacios 200596

•observational study of toxicity of ARV treatment including d4T, AZT, 3TC, ddl and ABC

•all patients (n=1391) had developed �adverse events that justified discontinuation of treatment with the associated drug •drugs with highest discontinuation rates were d4T (65.7%) and AZT (12.7%) •most frequent adverse events were lipoatrophy and peripheral neuropathy •lipoatrophy most commonly associated with d4T (480 of 550 cases) and anaemia most commonly associated with AZT (70 of 77 cases)

Siegfried 200697

•systematic review of d4T, 3TC and NVP combination therapy (Cochrane review)

•rash leading to temporary or permanent withdrawal of drug reported to be 12.3% for NVP once daily, 6.5% for NVP twice daily, 3.8% with EFV, 13.9% for NVP+EFV Grade 3-4 liver toxicity occurred in 13.6% on NVP once daily, 8.3% on NVP twice daily, 4.5% on EFV, 9.1% on NVP+EFV

Subbaraman 200798

•review of research on adverse effects of drugs used in resource-constrained regions

•spectrum of adverse events related to HAART in developing countries may differ from that in developed countries due to high prevalence of conditions such as anaemia, malnutrition and tuberculosis and initial presentation with advanced disease. The severity of adverse effects may vary due to host genetics and delays attributable to inadequate laboratory monitoring. •article reviews current knowledge about toxicities related to HAART in resource-limited regions •authors conclude that initiating HAART before advance immunosuppression, titrating doses in fixed dose combinations to differences in patient weight, providing more


Author Design Key results/conclusions laboratory monitoring and providing access to less toxic drugs may decrease incidence of toxicities related to HAART in resource-limited settings

van Oosterhout 200599

•evaluation of ART results in resource-poor setting in Malawi •drug used was fixed dose combination of d4T, 3TC and NVP

•76% of patients reported at least one adverse event •rash occurred in 45 (26%) of patients •no statistically significant difference in occurrence of adverse events between men and women


References 1. Chowers MY GB, Leibovici L et al. Reporting of adverse events in randomized

controlled trials of highly active antiretroviral therapy: systematic review. J Antimicrob Chemotherapy 2009;64:239-50.

2. Knobel H, Guelar A, Montero M, Carmona A, Luque S, Berenguer N, et al. Risk of side effects associated with the use of nevirapine in treatment-naive patients, with respect to gender and CD4 cell count. HIV Medicine 2008;9(1):14-8.

3. Bonjoch A, Paredes R, Domingo P, Cervantes M, Pedrol E, Ribera E, et al. Long-term safety and efficacy of nevirapine-based approaches in HIV type 1-infected patients. AIDS Research & Human Retroviruses 2006;22(4):321-9.

4. De Lazzari E, Leon A, Arnaiz JA, Martinez E, Knobel H, Negredo E, et al. Hepatotoxicity of nevirapine in virologically suppressed patients according to gender and CD4 cell counts. HIV Medicine 2008;9(4):221-6.

5. Manfredi R, Calza L. Nevirapine versus efavirenz in 742 patients: no link of liver toxicity with female sex, and a baseline CD4 cell count greater than 250 cells/microl. AIDS 2006;20(17):2233-6.

6. Manfredi R, Calza L. Safety issues about nevirapine administration in HIV-infected pregnant women. Journal of Acquired Immune Deficiency Syndromes: JAIDS 2007;45(3):365-8.

7. Marazzi MC, Germano P, Liotta G, Guidotti G, Loureiro S, da Cruz Gomes A, et al. Safety of nevirapine-containing antiretroviral triple therapy regimens to prevent vertical transmission in an African cohort of HIV-1-infected pregnant women. HIV Medicine 2006;7(5):338-44.

8. van Leth F, Andrews S, Grinsztejn B, Wilkins E, Lazanas MK, Lange JMA, et al. The effect of baseline CD4 cell count and HIV-1 viral load on the efficacy and safety of nevirapine or efavirenz-based first-line HAART. AIDS 2005;19(5):463-71.

9. Mocroft A, Staszewski S, Weber R, Gatell J, Rockstroh J, Gasiorowski J, et al. Risk of discontinuation of nevirapine due to toxicities in antiretroviral-naive and -experienced HIV-infected patients with high and low CD4+ T-cell counts.[erratum appears in Antivir Ther. 2007;12(4):577]. Antiviral Therapy 2007;12(3):325-33.

10. Lyons F, Hopkins S, Kelleher B, McGeary A, Sheehan G, Geoghegan J, et al. Maternal hepatotoxicity with nevirapine as part of combination antiretroviral therapy in pregnancy. HIV Medicine 2006;7(4):255-60.

11. Jamisse L, Balkus J, Hitti J, Gloyd S, Manuel R, Osman N, et al. Antiretroviral-associated toxicity among HIV-1-seropositive pregnant women in Mozambique receiving nevirapine-based regimens. Journal of Acquired Immune Deficiency Syndromes: JAIDS 2007;44(4):371-6.

12. Hitti J, Frenkel LM, Stek AM, Nachman SA, Baker D, Gonzalez-Garcia A, et al. Maternal toxicity with continuous nevirapine in pregnancy: results from PACTG 1022. Journal of Acquired Immune Deficiency Syndromes: JAIDS 2004;36(3):772-6.

13. Ananworanich J, Moor Z, Siangphoe U, Chan J, Cardiello P, Duncombe C, et al. Incidence and risk factors for rash in Thai patients randomized to regimens with nevirapine, efavirenz or both drugs. AIDS 2005;19(2):185-92.

14. Boulle A, Orrel C, Kaplan R, Van Cutsem G, McNally M, Hilderbrand K, et al. Substitutions due to antiretroviral toxicity or contraindication in the first 3


years of antiretroviral therapy in a large South African cohort. Antiviral Therapy 2007;12(5):753-60.

15. Dart Trial T. Twenty-four-week safety and tolerability of nevirapine vs. abacavir in combination with zidovudine/lamivudine as first-line antiretroviral therapy: a randomized double-blind trial (NORA). Tropical Medicine & International Health 2008;13(1):6-16.

16. Ho TT, Wong KH, Chan KC, Lee SS. High incidence of nevirapine-associated rash in HIV-infected Chinese. AIDS 1998;12(15):2082-3.

17. Joao EC, Calvet GA, Menezes JA, D'Ippolito MM, Cruz MLS, Salgado LAT, et al. Nevirapine toxicity in a cohort of HIV-1-infected pregnant women. American Journal of Obstetrics & Gynecology 2006;194(1):199-202.

18. Kiertiburanakul S, Sungkanuparph S, Charoenyingwattana A, Mahasirimongkol S, Sura T, Chantratita W. Risk factors for nevirapine-associated rash among HIV-infected patients with low CD4 cell counts in resource-limited settings. Current HIV Research 2008;6(1):65-9.

19. Kondo W, Carraro EA, Prandel E, Dias JM, Perini J, Macedo RLd, et al. Nevirapine-induced side effects in pregnant women: experience of a Brazilian university hospital. Brazilian Journal of Infectious Diseases 2007;11(6):544-8.

20. Kumarasamy N, Venkatesh KK, Devaleenal B, Palanivel V, Cecelia AJ, Muthu S, et al. Safety of switching to nevirapine-based highly active antiretroviral therapy at elevated CD4 cell counts in a resource-constrained setting. Journal of Acquired Immune Deficiency Syndromes: JAIDS 2007;45(5):598-600.

21. Littera R, Carcassi C, Masala A, Piano P, Serra P, Ortu F, et al. HLA-dependent hypersensitivity to nevirapine in Sardinian HIV patients.[see comment]. AIDS 2006;20(12):1621-6.

22. Manfredi R, Calza L, Chiodo F. Prospective, open-label comparative study of liver toxicity in an unselected population of HIV-infected patients treated for the first time with efavirenz or nevirapine. HIV Clinical Trials 2005;6(6):302-11.

23. Manosuthi W, Chumpathat N, Chaovavanich A, Sungkanuparph S. Safety and tolerability of nevirapine-based antiretroviral therapy in HIV-infected patients receiving fluconazole for cryptococcal prophylaxis: a retrospective cohort study. BMC Infectious Diseases 2005;5:67.

24. Phanuphak N, Apornpong T, Teeratakulpisarn S, Chaithongwongwatthana S, Taweepolcharoen C, Mangclaviraj S, et al. Nevirapine-associated toxicity in HIV-infected Thai men and women, including pregnant women.[see comment]. HIV Medicine 2007;8(6):357-66.

25. Phillips E, Gutierrez S, Jahnke N, Yip B, Lima VD, Hogg RS, et al. Determinants of nevirapine hypersensitivity and its effect on the association between hepatitis C status and mortality in antiretroviral drug-naive HIV-positive patients.[see comment]. AIDS 2007;21(12):1561-8.

26. Pitche P, Drobacheff-Thiebaut C, Gavignet B, Mercier M, Laurent R. [Cutaneous drug-reactions to nevirapine: study of risk factors in 101 HIV-infected patients]. Annales de Dermatologie et de Venereologie 2005;132(12 Pt 1):970-4.

27. Taiwo BO. Nevirapine toxicity. International Journal of STD & AIDS 2006;17(6):364-9; quiz 370.

28. Tansuphaswadikul S, Aung SE, Phonrat B, Kaewkungwal J, Pitisuttithum P, Maek-a-nantawat W. Predisposing factors for nevirapine toxicity among AIDS


patients with low baseline CD4 count. Asian Pacific Journal of Allergy & Immunology 2007;25(2-3):147-54.

29. Timmermans S, Tempelman C, Godfried MH, Nellen J, Dieleman J, Sprenger H, et al. Nelfinavir and nevirapine side effects during pregnancy. AIDS 2005;19(8):795-9.

30. Torti C, Costarelli S, De Silvestri A, Quiros-Roldan E, Lapadula G, Cologni G, et al. Analysis of severe hepatic events associated with nevirapine-containing regimens: CD4+ T-cell count and gender in hepatitis C seropositive and seronegative patients. Drug Safety 2007;30(12):1161-9.

31. Wit FWNM, Kesselring AM, Gras L, Richter C, van der Ende ME, Brinkman K, et al. Discontinuation of nevirapine because of hypersensitivity reactions in patients with prior treatment experience, compared with treatment-naive patients: the ATHENA cohort study. Clinical Infectious Diseases 2008;46(6):933-40.

32. Bolhaar MG, Karstaedt AS. A high incidence of lactic acidosis and symptomatic hyperlactatemia in women receiving highly active antiretroviral therapy in Soweto, South Africa.[see comment]. Clinical Infectious Diseases 2007;45(2):254-60.

33. Geddes R, Knight S, Moosa MYS, Reddi A, Uebel K, Sunpath H. A high incidence of nucleoside reverse transcriptase inhibitor (NRTI)-induced lactic acidosis in HIV-infected patients in a South African context. South African Medical Journal 2006;Suid-Afrikaanse Tydskrif Vir Geneeskunde. 96(8):722-4.

34. Osler M SD, Rebe K et al. Risk factors for and clinical characteristics of severe hyperlactataemia in patients receiving antiretroviral therapy: a case-control study. HIV Medicine 2009;online.

35. Wade NA, Unadkat JD, Huang S, Shapiro DE, Mathias A, Yasin S, et al. Pharmacokinetics and safety of stavudine in HIV-infected pregnant women and their infants: Pediatric AIDS Clinical Trials Group protocol 332.[see comment]. Journal of Infectious Diseases 2004;190(12):2167-74.

36. Arenas-Pinto A, Grant AD, Edwards S, Weller IVD. Lactic acidosis in HIV infected patients: a systematic review of published cases. Sexually Transmitted Infections 2003;79(4):340-3.

37. Blanco F G-BT, de la Cruz JJ et al. First-line therapy and mitochonrial damage: Different nucleosides, different findings. HIV Clinical Trials 2003;4(1):11-19.

38. Boubaker K, Flepp M, Sudre P, Furrer H, Haensel A, Hirschel B, et al. Hyperlactatemia and antiretroviral therapy: the Swiss HIV Cohort Study.[see comment]. Clinical Infectious Diseases 2001;33(11):1931-7.

39. Hocqueloux L, Alberti C, Feugeas JP, Lafaurie M, Lukasiewicz E, Bagnard G, et al. Prevalence, risk factors and outcome of hyperlactataemia in HIV-infected patients. HIV Medicine 2003;4(1):18-23.

40. Huynh TK, Luttichau HR, Roge BT, Gerstoft J. Natural history of hyperlactataemia in human immunodeficiency virus-1-infected patients during highly active antiretroviral therapy. Scandinavian Journal of Infectious Diseases 2003;35(1):62-6.

41. John M, Moore CB, James IR, Nolan D, Upton RP, McKinnon EJ, et al. Chronic hyperlactatemia in HIV-infected patients taking antiretroviral therapy.[see comment]. AIDS 2001;15(6):717-23.


42. Lactic Acidosis International Study G. Risk factors for lactic acidosis and severe hyperlactataemia in HIV-1-infected adults exposed to antiretroviral therapy. AIDS 2007;21(18):2455-64.

43. Laurent C, Bourgeois A, Mpoudi-Ngole E, Ciaffi L, Kouanfack C, Mougnutou R, et al. Tolerability and effectiveness of first-line regimens combining nevirapine and lamivudine plus zidovudine or stavudine in Cameroon. AIDS Research & Human Retroviruses 2008;24(3):393-9.

44. Makinson A, Moing VL, Kouanfack C, Laurent C, Delaporte E. Safety of stavudine in the treatment of HIV infection with a special focus on resource-limited settings. Expert Opinion on Drug Safety 2008;7(3):283-93.

45. Tien PC, Schneider MF, Cole SR, Justman JE, French AL, Young M, et al. Relation of stavudine discontinuation to anthropometric changes among HIV-infected women. Journal of Acquired Immune Deficiency Syndromes: JAIDS 2007;44(1):43-8.

46. Falco V, Crespo M, Ribera E. Lactic acidosis related to nucleoside therapy in HIV-infected patients. Expert Opinion on Pharmacotherapy 2003;4(8):1321-9.

47. Gerard Y, Maulin L, Yazdanpanah Y, De La Tribonniere X, Amiel C, Maurage CA, et al. Symptomatic hyperlactataemia: an emerging complication of antiretroviral therapy. AIDS 2000;14(17):2723-30.

48. Johri S, Alkhuja S, Siviglia G, Soni A. Steatosis-lactic acidosis syndrome associated with stavudine and lamivudine therapy.[see comment]. AIDS 2000;14(9):1286-7.

49. Lonergan JT, Barber RE, Mathews WC. Safety and efficacy of switching to alternative nucleoside analogues following symptomatic hyperlactatemia and lactic acidosis. AIDS 2003;17(17):2495-9.

50. Mandelbrot L, Kermarrec N, Marcollet A, Lafanechere A, Longuet P, Chosidow D, et al. Case report: nucleoside analogue-induced lactic acidosis in the third trimester of pregnancy. AIDS 2003;17(2):272-3.

51. Nelson M, Azwa A, Sokwala A, Harania RS, Stebbing J. Fanconi syndrome and lactic acidosis associated with stavudine and lamivudine therapy. AIDS 2008;22(11):1374-6.

52. Sarner L, Fakoya A. Acute onset lactic acidosis and pancreatitis in the third trimester of pregnancy in HIV-1 positive women taking antiretroviral medication.[erratum appears in Sex Transm Infect. 2003 Oct;79(5):429]. Sexually Transmitted Infections 2002;78(1):58-9.

53. Songa PM, Castelnuovo B, Mugasha EB, Ocama P, Kambugu A. Symptomatic hyperlactatemia associated with nucleoside analogue reverse-transcriptase inhibitor use in HIV-infected patients: a report of 24 cases in a resource-limited setting (Uganda).[see comment]. Clinical Infectious Diseases 2007;45(4):514-7.

54. Tripuraneni NS, Smith PR, Weedon J, Rosa U, Sepkowitz D. Prognostic factors in lactic acidosis syndrome caused by nucleoside reverse transcriptase inhibitors: report of eight cases and review of the literature. AIDS Patient Care & Stds 2004;18(7):379-84.

55. Sax PE, Gallant JE, Klotman PE. Renal safety of tenofovir disoproxil fumarate.[see comment]. AIDS Reader 2007;17(2):90-2.

56. Roling J SH, Fischereder M et al. HIV-associated renal diseases and highly active antiretroviral therapy-induced nephropathy. Clin Infect Dis 2006;42:1488-95.


57. Nelson MR, Katlama C, Montaner JS, Cooper DA, Gazzard B, Clotet B, et al. The safety of tenofovir disoproxil fumarate for the treatment of HIV infection in adults: the first 4 years. AIDS 2007;21(10):1273-81.

58. Gallant JE MR. Renal function with use of a tenofovir-containing initial antiretroviral regimen. AIDS 2009;23.

59. Squires K, Pozniak AL, Pierone G, Jr., Steinhart CR, Berger D, Bellos NC, et al. Tenofovir disoproxil fumarate in nucleoside-resistant HIV-1 infection: a randomized trial.[summary for patients in Ann Intern Med. 2003 Sep 2;139(5 Pt 1):I22; PMID: 12965954]. Annals of Internal Medicine 2003;139(5 Pt 1):313-20.

60. Arribas JR, Pozniak AL, Gallant JE, Dejesus E, Gazzard B, Campo RE, et al. Tenofovir disoproxil fumarate, emtricitabine, and efavirenz compared with zidovudine/lamivudine and efavirenz in treatment-naive patients: 144-week analysis. Journal of Acquired Immune Deficiency Syndromes: JAIDS 2008;47(1):74-8.

61. Fux CA, Rauch A, Simcock M, Bucher HC, Hirschel B, Opravil M, et al. Tenofovir use is associated with an increase in serum alkaline phosphatase in the Swiss HIV Cohort Study. Antiviral Therapy 2008;13(8):1077-82.

62. Moreno S, Domingo P, Palacios R, Santos J, Falco V, Murillas J, et al. Renal safety of tenofovir disoproxil fumarate in HIV-1 treatment-experienced patients with adverse events related to prior NRTI use: data from a prospective, observational, multicenter study. Journal of Acquired Immune Deficiency Syndromes: JAIDS 2006;42(3):385-7.

63. Nurutdinova D, Onen NF, Hayes E, Mondy K, Overton ET. Adverse effects of tenofovir use in HIV-infected pregnant women and their infants. Annals of Pharmacotherapy 2008;42(11):1581-5.

64. Ananworanich J, Nuesch R, Cote HCF, Kerr SJ, Hill A, Jupimai T, et al. Changes in metabolic toxicity after switching from stavudine/didanosine to tenofovir/lamivudine--a Staccato trial substudy. Journal of Antimicrobial Chemotherapy 2008;61(6):1340-3.

65. Cassetti I, Madruga JVR, Suleiman JMAH, Etzel A, Zhong L, Cheng AK, et al. The safety and efficacy of tenofovir DF in combination with lamivudine and efavirenz through 6 years in antiretroviral-naive HIV-1-infected patients. HIV Clinical Trials 2007;8(3):164-72.

66. Cirino CM, Kan VL. Hypokalemia in HIV patients on tenofovir. AIDS 2006;20(12):1671-3.

67. Elion R, Cohen C, Ward D, Ruane P, Ortiz R, Reddy YS, et al. Evaluation of efficacy, safety, pharmacokinetics, and adherence in HIV-1-infected, antiretroviral-naive patients treated with ritonavir-boosted atazanavir plus fixed-dose tenofovir DF/emtricitabine given once daily. HIV Clinical Trials 2008;9(4):213-24.

68. Gupta SK. Tenofovir-associated Fanconi syndrome: review of the FDA adverse event reporting system. AIDS Patient Care & Stds 2008;22(2):99-103.

69. Hazra R, Gafni RI, Maldarelli F, Balis FM, Tullio AN, DeCarlo E, et al. Tenofovir disoproxil fumarate and an optimized background regimen of antiretroviral agents as salvage therapy for pediatric HIV infection. Pediatrics 2005;116(6):e846-54.

70. Irizarry-Alvarado JM, Dwyer JP, Brumble LM, Alvarez S, Mendez JC. Proximal tubular dysfunction associated with tenofovir and didanosine causing Fanconi


syndrome and diabetes insipidus: a report of 3 cases.[see comment]. AIDS Reader 2009;19(3):114-21.

71. Mauss S BF, Schmutz G. Antiretroviral therapy with tenofovir is associated with mild renal dysfunction. AIDS 2005;19:93-99.

72. Padilla S GF, Masia M et al. Low frequency of renal function impairment during one-year of therapy with tenofovir-containing regimens in the real-wprld: a case-control study. AIDS Patient Care & Stds 2005;19:421-24.

73. Pozniak AL, Gallant JE, DeJesus E, Arribas JR, Gazzard B, Campo RE, et al. Tenofovir disoproxil fumarate, emtricitabine, and efavirenz versus fixed-dose zidovudine/lamivudine and efavirenz in antiretroviral-naive patients: virologic, immunologic, and morphologic changes--a 96-week analysis. Journal of Acquired Immune Deficiency Syndromes: JAIDS 2006;43(5):535-40.

74. Riordan A, Judd A, Boyd K, Cliff D, Doerholt K, Lyall H, et al. Tenofovir use in human immunodeficiency virus-1-infected children in the United kingdom and Ireland. Pediatric Infectious Disease Journal 2009;28(3):204-9.

75. Young B BK, Moorman A et al. Renal function in patients with preexisting renal disease receiving tenofovir-containing highly active antiretroviral therapy in the HIV Outpatient Study. AIDS Patient Care and STDs 2009;23(8):589-92.

76. Ssali F, Stohr W, Munderi P, Reid A, Walker AS, Gibb DM, et al. Prevalence, incidence and predictors of severe anaemia with zidovudine-containing regimens in African adults with HIV infection within the DART trial. Antiviral Therapy 2006;11(6):741-9.

77. Isaakidis P, Raguenaud M-E, Phe T, Khim SA, Kuoch S, Khem S, et al. Evaluation of a systematic substitution of zidovudine for stavudine-based HAART in a program setting in rural Cambodia. Journal of Acquired Immune Deficiency Syndromes: JAIDS 2008;49(1):48-54.

78. Sperling RS, Shapiro DE, McSherry GD, Britto P, Cunningham BE, Culnane M, et al. Safety of the maternal-infant zidovudine regimen utilized in the Pediatric AIDS Clinical Trial Group 076 Study. AIDS 1998;12(14):1805-13.

79. Nunez M, Ocampo A, Aguirrebengoa K, Cervantes M, Pascual A, Echeverria S, et al. Incidence of anaemia and impact on sustained virological response in HIV/HCV-coinfected patients treated with pegylated interferon plus ribavirin. Journal of Viral Hepatitis 2008;15(5):363-9.

80. Alvarez D, Dieterich DT, Brau N, Moorehead L, Ball L, Sulkowski MS. Zidovudine use but not weight-based ribavirin dosing impacts anaemia during HCV treatment in HIV-infected persons. Journal of Viral Hepatitis 2006;13(10):683-9.

81. Aurpibul L, Puthanakit T, Sirisanthana T, Sirisanthana V. Haematological changes after switching from stavudine to zidovudine in HIV-infected children receiving highly active antiretroviral therapy. HIV Medicine 2008;9(5):317-21.

82. Bae WH, Wester C, Smeaton LM, Shapiro RL, Lockman S, Onyait K, et al. Hematologic and hepatic toxicities associated with antenatal and postnatal exposure to maternal highly active antiretroviral therapy among infants. AIDS 2008;22(13):1633-40.

83. Curkendall SM, Richardson JT, Emons MF, Fisher AE, Everhard F. Incidence of anaemia among HIV-infected patients treated with highly active antiretroviral therapy. HIV Medicine 2007;8(8):483-90.

84. Huffam SE, Srasuebkul P, Zhou J, Calmy A, Saphonn V, Kaldor JM, et al. Prior antiretroviral therapy experience protects against zidovudine-related anaemia. HIV Medicine 2007;8(7):465-71.


85. Idoko JA, Akinsete L, Abalaka AD, Keshinro LB, Dutse L, Onyenekwe B, et al. A multicentre study to determine the efficacy and tolerability of a combination of nelfinavir (VIRACEPT), zalcitabine (HIVID) and zidovudine in the treatment of HIV infected Nigerian patients. West African Journal of Medicine 2002;21(2):83-6.

86. Moh R, Danel C, Sorho S, Sauvageot D, Anzian A, Minga A, et al. Haematological changes in adults receiving a zidovudine-containing HAART regimen in combination with cotrimoxazole in Cote d'Ivoire. Antiviral Therapy 2005;10(5):615-24.

87. Moyle G, Sawyer W, Law M, Amin J, Hill A. Changes in hematologic parameters and efficacy of thymidine analogue-based, highly active antiretroviral therapy: a meta-analysis of six prospective, randomized, comparative studies. Clinical Therapeutics 2004;26(1):92-7.

88. Violari A, Cotton MF, Gibb DM, Babiker AG, Steyn J, Madhi SA, et al. Early antiretroviral therapy and mortality among HIV-infected infants. New England Journal of Medicine 2008;359(21):2233-44.

89. Willig JH EJ, Westfall AO et al. Durability of initial antiretroviral therapy in a resource constrained setting and the potential need for zidovudine weight-based dosing. J Acquir Immune Defic Syndr 2009.

90. Calmy A, Pinoges L, Szumilin E, Zachariah R, Ford N, Ferradini L, et al. Generic fixed-dose combination antiretroviral treatment in resource-poor settings: multicentric observational cohort. AIDS 2006;20(8):1163-9.

91. Forna F, Liechty CA, Solberg P, Asiimwe F, Were W, Mermin J, et al. Clinical toxicity of highly active antiretroviral therapy in a home-based AIDS care program in rural Uganda. Journal of Acquired Immune Deficiency Syndromes: JAIDS 2007;44(4):456-62.

92. Kumarasamy N, Venkatesh KK, Cecelia AJ, Devaleenal B, Lai AR, Saghayam S, et al. Spectrum of adverse events after generic HAART in southern Indian HIV-infected patients. AIDS Patient Care & Stds 2008;22(4):337-44.

93. Manosuthi W, Prasithsirikul W, Chumpathat N, Suntisuklappon B, Athichathanabadi C, Chimsuntorn S, et al. Risk factors for mortality in symptomatic hyperlactatemia among HIV-infected patients receiving antiretroviral therapy in a resource-limited setting. International Journal of Infectious Diseases 2008;12(6):582-6.

94. Murphy RA, Sunpath H, Kuritzkes DR, Venter F, Gandhi RT. Antiretroviral therapy-associated toxicities in the resource-poor world: the challenge of a limited formulary. Journal of Infectious Diseases 2007;196 Suppl 3:S449-56.

95. Nuesch R SP, Ananworanich J et al. Monitoring the toxicity of antiretroviral therapy in resource limited settings: a prospective clinical trial cohort in Thailand. J Antimicrobial Chemo 2006;58:637-44.

96. Palacios R SJ, Camino X et al. Treatment-limiting toxicities associated with nucleoside analogue reverse transcriptase inhibitor therapy: A prospective, observational study. Current Therapeutic Research 2005;66(2):117-29.

97. Siegfried N vDP, Mahomed FA et al. Stavudine, lamivudine and nevirapine combination therapy for treatment of HIV infection and AIDS in adults. Cochrane Database of Systematic Reviews 2006;2.

98. Subbaraman R CK, Mayer KH et al. Adverse effects of highly active antiretroviral therapy in developing countries. Clin Infect Dis 2007;45:1093-1101.

99. van Oosterhout JJ, Bodasing N, Kumwenda JJ, Nyirenda C, Mallewa J, Cleary PR, et al. Evaluation of antiretroviral therapy results in a resource-poor setting


in Blantyre, Malawi. Tropical Medicine & International Health 2005;10(5):464-70.


Annex 1 Post-hoc relative risk analyses Given that the GRADE software did not provide an assessment of relative effect (relative risk), post-hoc analyses were conducted for studies with available data. The questions and the outcomes from the GRADE tables were used to form the basis of these comparisons. Two of the GRADE questions were excluded: 1) hepatotoxicity in treatment-naive patients using NVP – this could not be calculated given no response in patients with high CD4 cell count; and 2) hepatotoxicity and rash in stable virologically suppressed patients using NVP – did not calculate relative risk for these values as data in GRADE table was based on a meta-analysis which provided odds ratio results (see Table 1.1 above for summary of results). The relative risk results for the comparisons are provided below in Table 1 for the NVP studies and Table 2 for the TDF studies. A meta-analysis combining the Marazzi7 and Kondo19 NVP studies in pregnant women was conducted given the similarity in outcomes and patient populations. There was not statistically significant heterogeneity, which indicates it is reasonable to combine these two studies. For all analyses, the relative risk results concur with the results reported by the studies in terms of statistical significance or non-significance. However, the results presented below should be interpreted with caution, given that the analyses are post-hoc and the majority of studies are not randomised, double-blind trials but instead are retrospective reviews or of other study design which leaves them open to potential sources of bias. Table 1: Results of RR analyses – NVP studies Question Trial Comparison Should NVP be used in patients with high or unknown CD4?

Low CD4 (NVP)

High CD4 (NVP)

RR (95% CI)

•Jamisse 200711 •pregnant women receiving NVP-based regimens •open-label •Mozambique

5/79 (6.3%) 6/67 (9.0%) 0.707 (0.24, 2.10)

Hepatotoxicity

•RR=0.707 (0.24, 2.10) which indicates no significant difference in hepatotoxicity between those with low and high CD4 cell count •results correspond to publication, which found no significant difference in hepatotoxicity (p=0.55). However, 6% of patients with high CD4 cell count had severe hepatotoxicity compared to none (0%) with low CD4 cell count (p=0.02). These results should be interpreted with caution, given small n and open-label nature of the study.

Hepatotoxicity in pregnant women

•Marazzi 20067 and Kondo 200719 •pregnant women receiving NVP-based regimens •retrospective reviews •Brazil (Marazzi) and Africa (Kondo)

12/160 (*7.5%)

40/676 (5.9%)

1.04 (0.22, 4.93)

Cochran Q=1.64 p=0.20


Question Trial Comparison •RR=1.04 (0.22, 4.93) indicates no significant difference between pregnant women with low and high CD4 cell count in the occurrence of hepatotoxicity. •results of meta-analysis concur with results of both studies. Marazzi et al (2006)7 had concluded that NVP-containing regimens, in pregnant women at all CD4 cell count levels appear to be safe.

Serious AEs in patients using NVP vs. abacavir NVP Abacavir •DART 200815 •treatment-naive adults with CD4 <200 cells/mm3 •randomized, double-blind trial comparing NVP and ABC both in combination with ZDV and 3TC •Uganda

14/300 (4.7%)

6/300 (2.0%)

2.33 (0.94, 5.81)

Serious AEs

•RR=2.33 (0.94, 5.81) indicates no significant difference between patients receiving NVP or abacavir-based regimens in the occurrence of serious AEs. •results concur with publication which indicated no significant difference in serious AEs between NVP and abacavir (HR=0.42; 95% CI: 0.16, 1.09; p=0.06)

Treatment-limiting toxicity in pregnant women using NVP or nelfinavir

NVP Nelfinavir

•Hitti 200412 •treatment-naive pregnant women •open-label study comparing NVP and nelfinavir both in combination with ZDV and 3TC •USA

5/17 (29.4%) 1/21 (4.8%) 6.18 (1.08, 38.07)

Treatment-limiting toxicity

•RR=6.18 (1.08, 38.07) indicates significantly more treatment-limiting toxicity occurred in pregnant women using NVP compared to those using nelfinavir. However results should be interpreted with caution given the small n and wide confidence intervals. •the publication did not report a p value, however the study was halted due to greater than expected toxicity and changes in NVP prescribing information.

Hepatotoxicity and rash in NVP vs. efavirenz NVP Efavirenz •Manfredi 200522 •treatment-naive patients •NVP and EFV-based regimens •open-label, observational study •Italy

147/299 (49.2%)

66/324 (20.4%)

2.41 (1.90, 3.09)

Hepatotoxicity

•RR=2.41 (1.90, 3.09) indicates significantly more patients treated with NVP-based regimens developed hepatotoxicity compared o those treated with EFV-based regimens •the publication reported that hepatotoxicity characterized by at least a 2-fold increase in serum transaminase levels compared to baseline values was significantly linked with NVP compared to EFV (p<0.00001).

Hepatotoxicity •van Leth 20058 •post-hoc analysis of

148/607 (24.4%)

79/400 (19.8%)

1.23 (0.97, 1.57)


Question Trial Comparison patients in 2NN trial comparing regimens containing NVP and EFV •Asia/Australia, North America, South America, South Africa, Europe

76/607 (12.5%)

38/400 (9.5%)

1.32 (0.92, 1.90)

Rash

•RR=1.23 (0.97, 1.57) and 1.32 (0.92, 1.90) indicate no significant difference between NVP and EFV-based regimens in occurrence of hepatotoxicity or rash •publication reports that women with a CD4 cell count >200 × 106 cells/l had a significantly greater risk to develop rash compared to men with the same high CD4 cell count (OR=2.0; 95% CI: 1.2, 3.4).

Rash and hepatotoxicity in pregnant vs. non-pregnant women using NVP

Pregnant (NVP)

Non-pregnant

(NVP)

Rash 5/58 (8.6%) 10/95 (10.5%)

0.82 (0.30, 2.16)

Hepatotoxicity

•Timmermans 200529 •pregnant women receiving NVP or NFV •retrospective review •Netherlands

11/58 (19.0%)

4/95 (4.2%) 4.50 (1.59, 12.91)

•RR=0.82 (0.30, 2.16) indicate no significant difference between pregnant and non-pregnant women receiving NVP in occurrence of rash. •results concur with publication, which reported that risk of NVP-associated rash was not increased in pregnant women. •RR=4.50 (1.59, 12.91) indicates a significant difference in pregnant and non-pregnant women receiving NVP. •results concur with publication which reported that NVP-related hepatotoxicity occurred more often in pregnant women than non-pregnant case control group. •relative risk results should be interpreted with caution given retrospective nature of the study and potential for bias, as well as post-hoc nature of the analysis.

Table 2: Results of RR analyses – TDF studies Question Trial Comparison Adverse events in patients using TDF-based regimens versus other regimens

TDF Other regimens

RR (95% CI)

•Padilla 200572 •comparison of TDF-containing regimens and other ARV regimens •retrospective review •USA

5/122 (4.1%) 1/194 (0.5%) 7.95 (1.25, 50.99)

Creatinine increase

•RR=7.95 (1.25, 50.99) indicates that more patients treated with TDF-based regimens experienced creatinine increase compared to those treated with other regimens. Results should be interpreted with caution given retrospective nature of the study, post-hoc analysis and the wide confidence intervals.

Proteinuria •Mauss 200571 •comparison of TDF-treated patients and those treated with other regimens •cross-sectional •Germany

24/82 (29.3%)

5/92 (5.4%) 5.39 (2.25, 13.20)


Question Trial Comparison •RR=5.39 (2.25, 13.20) indicates a significantly greater proportion of TDF-treated patients experienced proteinuria >130mg/day compared to those treated with other regimens. •results concur with publication, which reported significant difference in proportion of TDF-treated patients with proteinuria >130mg/day (p<0.001) compared to those treated with other regimens. •results should be interpreted with caution given small n and the cross-sectional design of the study. •Arribas 200860 •comparison of TDF, FTC and EFV with 3TC/AZT and EFV in adult treatment-naive patients •randomized, open-label •USA

12/231 (5.2%)

29/232 (12.5%)

2.41 (1.28, 4.56)

Discontinuation due to AEs

•RR=2.41 (1.28, 4.56) indicates a significantly greater proportion of patients treated with 3TC/AZT and EFV compared to those treated with TDF, FTC and EFV discontinued treatment dues to AEs. •results concur with publication which reported that significantly more patients in the AZT/3TC arm discontinued therapy because of adverse events (p=0.01). •Squires 200359 •randomized, double-blind comparison of TDF or placebo added to existing ARV regimens •North America, Europe, Australia

49/368 (13.3%)

25/182 (13.7%)

0.97 (0.62, 1.52)

Grade 3 or 4 AEs

•RR=0.97 (0.62, 1.52) indicates that there was no significant difference in the proportion of patients reporting Grade 3 or 4 AEs in those adding TDF to their regimen compared to placebo. •results concur with publication, which concluded that there was no evidence of TDF-related toxicity in the trial.

SUMMARY OF AVAILABLE SAFETY DATA FOR NEVIRAPINE, … · Safety review NVP, d4T, AZT, TDF 4 and whether sponsorship affects the quality and characteristics of adverse event reporting.

Documents