HPTN 071 (PopART) 試驗的經驗學習
資料來源：www.thelancet.com/hiv Vol 9 November 2022 / 財團法人台灣紅絲帶基金會編譯
過去十年中 HIV 發生率逐漸下降的大部分原因是抗反轉錄病毒療法 (ART) 的覆蓋範圍擴大。越來越多的愛滋病毒感染者正在接受抗反轉錄病毒治療並且病毒受到抑制，因此無法將病毒傳染給他們的伴侶。普遍檢測和治療已被提議作為增加 ART 覆蓋率的一項關鍵策略，以實現 HIV 發生率的大幅降低。該策略涉及在社區層面廣泛提供 HIV 檢測，同時有效聯繫照護和ART服藥順從性的支持。HPTN 071 (PopART) 試驗是四項社區隨機試驗中規模最大的一項，這些試驗旨在評估普遍檢測和治療策略的效果。
PopART 試驗於 2013 年至 2018 年在甘比亞和南非西開普省的 21 個城市社區進行。共有三個研究組。A 組接受了完整的 PopART 普遍性檢測和治療介入，包括每年在整個社區進行挨家上門 的HIV 檢測，以及對所有 HIV 檢測呈陽性的個人進行全面性的 ART。B 組除根據現行國家指引給予 ART外，並接受了完整的介入。C 組是接受常規照護標準的對照組，包括根據當前國家指引進行的 ART。 2016 年，B 組和 C 組轉變為對所有 HIV 檢測呈陽性的人全面提供 ART，以響應國家指引的變化。因此，從 2016 年開始，A 組和 B 組緊密結合。
主要發現是該介入措施是可行的、可接受的，並且總體上實現了聯合國滋愛病規劃署的 90-90-90 目標，儘管並非在每個年齡和性別組中都如此。與控制組相比，B 組社區的 HIV 發生率降低了 30%，但在 A 組中只有 7%。研究人員針對這一不一致的發現列出了幾種可能的解釋，並進行了進一步分析以探索這些解釋。
首先，假設在 2016 年之前，在 A 組的國家指引之外開始 ART 的知情同意之要求可能會阻礙採用。然而，同意率非常高，並且數據不支持這一假設，這顯示 A 組的病毒抑制至少與 B 組一樣高。其次，A 組大規模提供 ART 可能導致到性方面解除抑制。然而，本期第 4 篇文章提供的性行為數據顯示，沒有證據顯示 A 組性行為的風險高於 B 組或 C 組，儘管單純皰疹病毒第 2型的發生率數據反映了 HIV 發生率，強調這兩種感染之間的聯繫非常密切。第三，在研究組之間可以觀察到遷移或流動率的差異，但數據比較顯示這些因素在兩組之間沒有明顯差異。最後，A 組的 ART 的留率或依從性可能較低，但本期的另一篇文章報告沒有證據顯示研究組之間的這些結果存在差異。在沒有其他解釋的情況下，對 A 組和 B 組不一致結果的最可能解釋是與 HIV 發生率相關的因素的機會失衡，特別是考慮到隨機分配的社區數量很少時（每組七個）。基於橫斷面發生率分析的關於 HIV 發生率的初步數據顯示，A 組的基準線上發生率高於 B 組和 C 組。這些數據將在適當的時候公佈。鑑於這些發現，並且 A 組和 B 組在試驗的主要分析期間緊密一致，我們認為介入的總體效果（A 組和 B 組與 C 組相比較）是最可靠的效果衡量標準，並且這項措施顯示 HIV 發生率顯著降低了 20%。自試驗完成以來，已在幾個領域進行了額外的分析。首先，介入的成本已經量化，顯示上門服務的年人均成本約為 7 美元，介入在兩國可能都具有成本效益。其次，來自本期還報告了適用於大量試驗數據的數學模型。這些預測顯示，如果介入持續進行，對 HIV 發生率的影響將隨著時間的推移而繼續增加，在 2030 年之前達到 50%。如果將介入措施應用於更廣泛的區域，效果會更大，因為這將避免因移徙和流動而導致效果減弱。
我們從這些調查結果中總結了政策和實施的一些重要經驗教訓。首先，為終結愛滋病而努力將要求愛滋病發生率大幅下降。考慮到這一點，聯合國愛滋病規劃署制定了到 2025 年需要實現的雄心勃勃之 95-95-95 目標。目前的 HIV 診斷和治療方法似乎不太可能實現這些目標，但 PopART 和其他普遍檢測和治療試驗的數據顯示，基於社區的普遍檢測和治療方法是可行的、可以接受的，並且可以為實現目標和降低發生率做出重要貢獻。一個重要的發現是，雖然 PopART 介入總體上達到了 90-90-90 的目標，但它並沒有在所有年齡和性別群體中都達到。需要更加努力地提高 ART 在年輕人中的覆蓋率，尤其是在年輕男性中，這可能需要創新的方法。除非在每個人口亞群中都達到了目標，95-95-95 目標不能被認為已經達到。其次，雖然有針對性的 HIV 檢測方法（例如基於指數的檢測）可能有很高的收益，但它們不太可能實現對人群的充分覆蓋，因此仍然可能需要更普遍的社區範圍的方法，特別流行是在高度普遍化的情況下。要實現對所有年齡和性別群體的如此廣泛的覆蓋立不存在單一、有效的模式，並且需要根據不同的地理和流行病學背景調整方法。重要的是，如果要採取此類介入措施，將需要適當和有效的社區參與方法才能成功。這種社區參與是 PopART 試驗的關鍵組成部分，本期的觀點討論了用於向研究社區傳達散播試驗結果的方法。第三，PopART 和其他普遍檢測和治療試驗顯示了在社區提供服務的多重共病方法上的潛力。這些方法涉及對一系列疾病情況（如 HIV、肺結核、性傳播感染和高血壓）的篩查和治療，可以在範圍上提供經濟方面重要優勢，並可能減少恥辱感。在社區全面和持續地提供 HIV 檢測也提供了機會，可以為 HIV 檢測呈陰性的人提供 HIV 預防服務，並為越來越多退出抗病毒治療的 HIV 陽性者提供一條返回 ART 的簡便途徑。關心。最後，模型預測顯示，要取得最大效果，需要維持基於社區的策略，並在可能的情況下擴大到更廣泛的人群。
我們聲明沒有競爭利益。
*Richard Hayes, Helen Ayles, Sarah Fidler richard.hayes@lshtm.ac.uk
英國倫敦，倫敦衛生和熱帶醫學院，流行病學和公共衛生學院，傳染病流行病學系 (RH) 和傳染病和熱帶病學院，臨床研究系 (HA)； 贊比亞盧薩卡，Zambart (HA); 英國倫敦，倫敦帝國理工學院醫學院傳染病系(SF)
Lessons from the HPTN 071 (PopART) trial
www.thelancet.com/hiv Vol 9 November 2022
Much of the gradual decline in HIV incidence over the past decade has been attributed to expanded coverage of antiretroviral therapy (ART). An increasing proportion of people living with HIV are on ART and are virally suppressed, therefore unable to transmit the virus to their partners. Universal testing and treatment has been proposed as a key strategy to increase ART coverage in an attempt to achieve steeper reductions in HIV incidence. This strategy involves wide-scale delivery of HIV testing at the community level, accompanied by effective linkage to care and support for ART adherence. The HPTN 071 (PopART) trial was the largest of four community-randomised trials that were conducted to evaluate the effect of the universal testing and treatment strategy.
The PopART trial was conducted in 21 urban communities in Zambia and the Western Cape province of South Africa from 2013–2018.2 There were three study arms. Arm A received the full PopART universal testing and treatment intervention, including annual door- to-door HIV testing throughout the community and universal ART for all individuals who tested positive for HIV. Arm B received the full intervention except that ART was given according to current national guidelines. Arm C was a control arm that received the usual standard of care, including ART according to current national guidelines. In 2016, arms B and C transitioned to providing ART for all individuals who tested positive for HIV in response to changes in national guidelines. Arms A and B were therefore closely aligned from 2016 onwards.
Key findings were that the intervention was feasible, acceptable, and achieved the UNAIDS 90-90-90 targets overall, although not in every age and sex group. HIV incidence at the community level was reduced by 30% in arm B compared with the control arm, but by only 7% in arm A. The investigators listed several potential explanations for this dissonant finding and have done further analyses to explore these.
First, it was hypothesised that the requirement for informed consent to start ART outside national guidelines in arm A until 2016 could have discouraged uptake. However, consent rates were very high and this hypothesis was not supported by the data, which showed that viral suppression was at least as high in arm A as in arm B. Second, wide-scale delivery of ART in arm A could have led to sexual disinhibition. However, an Article in this issue presents data on reported sexual behaviour that show no evidence of higher-risk sexual behaviour in arm A than in arms B or C, although data on the incidence of herpes simplex virus 2 mirrored those on HIV incidence, emphasising the very close association between these two infections. Third, differences in rates of migration or mobility could have been observed between the study arms, but comparisons of data showed no appreciable differences in these factors between the arms. Finally, ART retention or adherence could have been lower in arm A, but another Article5 in this issue reports no evidence of a difference in these outcomes between study arms. In the absence of alternative explanations, the most likely explanation for the dissonant findings in arms A and B is chance imbalances in factors associated with HIV incidence, especially given the small number of communities that were randomised (seven per arm). Preliminary data on HIV incidence, based on a cross-sectional incidence assay, show higher incidence at baseline in arm A than in arms B and C. These data will be published in due course. Given these findings, and that arms A and B were closely aligned during the primary analysis period of the trial, we consider that the overall effect of intervention (combined arms A and B compared with arm C) is the most reliable measure of effect, and this measure shows a significant, 20% reduction in HIV incidence.2
Since the trial was completed, additional analyses have been conducted in several areas. First, the costs of the intervention have been quantified, showing that the annual cost of the door-to-door service was about US$7 per capita, and that the intervention was probably cost-effective in both countries.6 Second, projections from a mathematical model, fitted to the extensive data from the trial, are also reported in this issue.7 These projections show that if the intervention were sustained, the effect on HIV incidence would continue to increase over time, reaching 50% well before 2030. The effect would be even greater if the intervention were applied over a wider area, as this would avert the dilution of effect due to migration and mobility.
We summarise some of the key lessons for policy and implementation from these findings. First, working towards ending AIDS will require much steeper declines in HIV incidence. With this in mind, UNAIDS has set ambitious 95-95-95 targets that need to be reached by 2025.8 Current approaches to HIV diagnosis and treatment seem unlikely to achieve these targets, but the data from PopART and other universal testing and treatment trials show that community-based approaches to universal testing and treatment are feasible, acceptable, and could make an important contribution to reaching the targets and reducing incidence. An important finding is that although the PopART intervention reached the 90-90-90 targets overall, it did not do so in all age and sex groups.3 More intensive efforts will be needed to increase ART coverage in young people and especially in young men, and this might require innovative approaches. The 95-95-95 targets cannot be considered to have been met unless they are reached in every subgroup of the population. Second, although targeted approaches to HIV testing (such as index-based testing) might have high yield, they are unlikely to achieve adequate coverage of the population, so more universal community-wide approaches are still likely to be needed, particularly in highly generalised epidemics. No single, effective model exists for achieving such widespread coverage in all age and sex groups, and approaches will need to be tailored to different geographical and epidemiological contexts. Importantly, appropriate and effective methods of community engagement will be needed if such interventions are to be successful. Such community engagement was a key component of the PopART trial, and a Viewpoint in this issue discusses the approaches used to disseminate the trial results to the study communities. Third, the PopART and other universal testing and treatment trials have shown the potential of multimorbidity approaches to service provision in the community. Such approaches, which involve screening and treatment for a range of conditions—such as HIV, tuberculosis, sexually transmitted infections, and hypertension—could offer important advantages in terms of economies of scope and, possibly, reductions in stigma. Comprehensive and sustained delivery of HIV testing in communities also provides the opportunity to offer HIV-prevention services to people who test negative for HIV, and to provide an easy route back onto ART for the increasing number of HIV-positive individuals who have dropped out of care. Finally, the model projections show that, to achieve maximal effect, community-based strategies need to be sustained and, if possible, scaled up over a wider population.
We declare no competing interests.
*Richard Hayes, Helen Ayles, Sarah Fidler richard.hayes@lshtm.ac.uk
Department of Infectious Disease Epidemiology, Faculty of Epidemiology and Public Health (RH) and Clinical Research Department, Faculty of Infectious and Tropical Diseases (HA), London School of Hygiene and Tropical Medicine, London, UK; Zambart, Lusaka, Zambia (HA); Department of Infectious Diseases, Faculty of Medicine, Imperial College London, London, UK (SF)