Clinical Outcomes of Tenofovir Versus Zidovudine-based Regimens Among People Living with HIV/AIDS: a Two Years Retrospective Cohort Study

INTRODUCTION

Emergence of Acquired immunodeficiency syndrome (AIDS) around 1980’s, as a major public health threat leads to the introduction of potent ART. This had resulted in dramatical reduction of associated mortality, morbidity, improved quality of life,and revitalized communities [1, 2]. Before 2009, the most commonly used backbone drug in resource-limited settings was either Zidovudine (ZDV) or stavudine (D4T), which had a high rate of side-effects [3]. These side-effects have led to banning of stavudine in developed countries in favor of less toxic longer half-life, and more friendly alternatives like tenofovir disoproxil fumarate (TDF)[4-6]. Randomized clinical trials have demonstrated comparable or greater efficacy of TDF compared with ZDV or D4T [7-12] in combination therapy with regards to virological suppression, as well as a tendency for less toxicity-related discontinuations and improved adherence in both developed and resource limited settings [8-13]. Nevertheless, most of these RCTs were about efficacy of TDF in combination therapy on virological suppression and/or immunological boosting. The effect of TDF on survival benefits and factors influencing mortality and morbidity in low income nations are rarely exploited. The 2009 World Health Organization (WHO) guidelines for ART recommended the phasing out of stavudine in resource-limited settings and many African countries adapted this recommendation and revised their guidelines in which the first line regimen consisted of either AZT or TDF backbone [14].

TDF become routinely utilized in the current practice setup since the beginning of 2013 in Ethiopia. Despite, its efficacy and safety issues are unknown in an Ethiopian setting where patients generally present late, have high rates of TB and other infectious conditions [15]. As most studies are from high income settings and extrapolation might not scientifically sound,evidences through research addressing questions regarding the optimum first-line ART regimen in patients living with HIV in low income countries are necessary. There are conflicting results coming out of the current literatures regarding efficacy and safety of TDF based regimens compared with AZT based regimen. For instance, a body of literatures reported that TDF+3TC+NVP was associated with higher hazard of mortality and virologic failure when compared to ZDV +3TC+NVP [16, 17] and even TDF based regimens were less protective than AZT based regimens in HIV patients living in resource limited settings [18]. A systematic review showed that the overall mortality rate between patients who were taking either AZT or TDF based ART regimens was not significantly different [19]. On the other hand, studies reported that TDF performed better than either d4T or AZT, most notably with less drug substitution and mortality [20, 21]. Serious renal toxicity, like acute renal failure requiring dialysis, progressive decline in renal function, proximal renal tubular dysfunction, and Fanconi-syndrome were also reported by some literatures [17, 18, 22, 23]. In another study, patients given tenofovir containing regimens experienced renal stability or improvement, even if they had pre-existing mild to moderate renal dysfunction [24].

Taking together, these inconclusive results and considering the different nature of the study setting, it is important to explore the mortality and morbidity benefits of TDF and associated factors. Therefore, this study compared AZT and TDF based first line regimens in terms of their clinical effects and associated risk factors in Ethiopia, one of the low-income countries.

METHOD AND PARTICIPANTS

Sample Size and Sampling Techniques

Sample size determination was guided by the number of patients on TDF/3TC/NVP, whereby only 70 patients fulfil the inclusion criteria and included into the study. Patients from the other regimens were selected based on the above figure to make the AZT to TDF group ratio 1:1. Therefore, frequency matching was used so select a total of 280 subjects, with 140 charts of patients from each group were reviewed (Fig. 1). TDF groups (n₁ = 140) were those initiated with TDF based regimen which were identified from patient charts of hospital records. A simple random sampling technique was used to select patient charts from each regimens using computer generated random number. One from TDF exposed patient was selected for one patient exposed to AZT, resulting in 140 total patients (n₂ = 140), which were selected by similar manner as TDF group.

RESULTS

A total of 1034 patients started antiretroviral therapy (ART) and treated for 6months. Of which 352 belonged to AZT arm, 620 were from TDF arm who have complete CD4+ count at 6month of treatment. Forty eight patients were excluded initially from either regimens due to missed CD4+ count at 6month, 22(12 and 10 from AZT and TDF) because of pregnancy and 110 patients due to regimen change and adherence issue (Fig. 1).

The overall analysis time at risk was 539.39 years. The cohort contributed to a total of 2.74/100 and 2.72/100 person-years of follow-up for TDF and AZT groups respectively. The mean + standard deviation (SD) duration of follow up was 714.2 + 69.6 and 708.8 + 78.9 days (p=0.753) among TDF and AZT, respectively. Study participants retained in the cohort for different lengths of follow up time: stayed for a minimum of 7.4 and 8.9 months for TDF and AZT groups, respectively (p=0.743).

Efficacy: Clinical Outcomes

The proportion of death among TDF and AZT group was 3.68% and 4.48% (p=0.759). The survival time was, (mean + SD), 713.46 + 4.411 and 709.57 + 4.983 days (p= 0.743) respectively. When the proportion of death is stratified among individual regimens as compared to AZT/3TC/NVP, TDF/3TC/EFV based regimen carries the lowest proportion 2(2.9%); and it was almost similar i.e. 3(4.44%), for the rest of the regimens. The proportion of opportunistic infection was 14.3%and 17.9% (p=0.228), respectively among TDF and AZT groups. The mean + SD survival time to opportunistic infection was 656.574 + 14.58 and 654.793 + 14.339 days, respectively (p=0.462). Patients exposed to TDF/3TC/EFV had favorable survival experience; and the difference was marginally significant as shown by the Kaplan-Meier survival estimate (Fig. 2).

DISCUSSION

In this population with good adherence (adherence >95%)(52), a higher proportion of death was recorded among AZT groups (p=0.759). The survival time, of TDF was also shown improvement does not show any statistically significant difference (p= 0.743).

Low body mass index (<18.5kg/m²) at baseline and a unit increment in baseline CD4+ count was the independent predictors of death. Females and patients commencement AZT based regimen were also found to be at higher risk of death, although it was statistically insignificant .

A similar finding was reported by Damtew et al. [25], from Somali region, Karamara hospital. However, the proportion of death among patients groups was 29.8% and 31.9%(p=0.429), respectively in the previous study. The higher proportions of death might be due to smaller sample size (280 vs. 485 subjects), inclusion of patients only with good adherence, and exclusion of patients with follow-up less than six months, as most of the deaths occur within four months post initiation of ART [26]. Involvement of adherence supporters, improvement in the prophylactic and VCT services, might have played a role in reducing the incidence of death in current study.

The risk of death for patients with BMI<18.5Kg/m was more than two times higher, (p=0.049) compared to those with a BMI > 18.5 kg/m². In study from Malawi, individuals with BMI<16 kg/m² had six times higher risk of dying in the first three months than those with BMI>16 kg/m² [27]. Asgaire et al. [28] also estimated one year mortality nearly 50% among patients with severe malnutrition in Tanzania. In our finding, a unit increase in baseline CD4+ count, resulted in 18% risk of reduction in death (p=0.019). Study from USA [29] had also reported that in patients with higher baseline CD4+ counts(>200) the risk of death in the coming year was reduced to < 5%.The finding is also in accordance with the study conducted in South Africa [30] and Ethiopia [25].

Patients from AZT group had 33% higher hazard of death than their TDF exposed counter parts (p=0.52). Our finding is consistent with the study from South Africa where patients exposed to TDF based regimen had 40% lower risk of death than their AZT exposed counter parts (AHR=1.4 95% CI [1.3, 1.5] [20].

In our study, the overall prevalence of OIs in TDF and AZT group is 14.3% and 17.9%, respectively (p=0.496). The mean + SD survival time to opportunistic infection for TDF group was slightly improved (p=0.462). The average treatment effect was favored TDF/3TC/EFV. History of baseline prophylaxis, baseline CD4+count and low base line body mass index (<18.5kg/m²) were the independent predictors for the occurrence of OIs as identified by multivariate cox-regression. Although, it lacked statistical significance, patients randomized to AZT based regimen were 23% at higher risk of developing OIs.

This finding was concurrent with one Indian RCT in which slightly higher proportion of OIs was recorded in patients randomized to AZT group (46% vs. 31%, p=0.22) [31]. In addition to its 23% OIs risk reduction in our study, the slightly higher median survival time in TDF group may also explain the survival advantage of this regimen. A cohort study by Samuel et al. [32] conducted in Kenya indicated that patients commenced on TDF based have relatively higher mean survival than its AZT counterpart (61 vs. 56.5 months) respectively.

In addition, one extra opportunistic infection was prevented every 14 patients treated using this TDF/3TC/EFV regimen (p=0.026). On the contrary, AZT/3TC/EFV was the least protective regimen used in this set-up, where one patient will experience 9 episodes extra of opportunistic infections with similar course of treatment (p=0.049). This implies that the TDF group has a better chance of survival and increased quality of life, as described by Sowmy V [31].

Patients with no baseline had eight times higher hazards of opportunistic infections than those who have started baseline prophylaxis with Cotrimoxazole and Isoniazid preventive therapy (p=0.009). It is clinically sound that immunologic incompetent individuals are predisposed to infection [33, 34]. Also, patients with cotrimoxazole only baseline prophylaxis were at higher probability of having OIs than their counter parts with Cotrimoxazole and Isoniazid (p=0.013). This implies that the presence of TB can change the clinical spectrum of other infections in the presence of HIV/AIDS. Stephanus K et al. [35] reported that, having a TB event during the follow-up was associated with a 2.71 times higher relative risk of a subsequent other opportunistic infection compared to having no prior TB during follow-up (95% CI [1.56, 4.70]). The impact of prophylaxis on the occurrence of opportunistic infections is also reported by other studies [36].

Patients with baseline BMI less than 18.5 were two times at higher risk of having opportunistic infections (p=0.016). Yoann et al. [37] also described low baseline BMI as a significant independent predictor for development of opportunistic infection. Another study from Nigeria has also reported opportunistic infections are most frequent in patients on ART with low body mass index [38]. For baseline CD4+ count, it was revealed that a unit increase resulted in 47% of risk reduction in OIs occurrence (p=0.039). There were also similar findings from Ethiopia [35] and Nigeria [39], which reported lower baseline CD4+ count was significantly associated with the occurrence of opportunistic infections.

Our study was not without limitation. Firstly, it was underpowered to detect the intended outcome due to inclusion of minimum number of observations for clinical outcomes. Measure of adherence by health professionals that may not fit to the reality, inability to assess the occurrence of specific OIs and selection bias due to scarcity of TDF/3TC/NVP, are some of the limitations.

CONCLUSION

In current study, there was no significant difference in mortality between those exposed to TDF versus AZT based regimens. The proportion of death and OIs in the subgroup belonged to TDF/3TC/EFV was lower as compared to those belonged to other regimens under study although the difference was not statistically significant. Moreover, low BMI and absence of prophylaxis at baseline were found to be an independent risk predictors for death and OIs. Higher CD4 count was found to be protective. The study highlighted the need for paying closer attention for these patients groups over the course of treatment provision.

REFERENCES