結束在中、低收入國家注射藥癮者中的愛滋病毒流行
資料來源：www.thelancet.com Vol 392 September 1, 2018/刺胳針醫學期刊/財團法人台灣紅絲帶基金會編譯
在「刺胳針」醫學期刊上，William C Miller及其同事描述了他們對一項隨機對照試驗的可行性研究，該試驗中針對在基輔（烏克蘭）、太原（越南）和雅加達（印度尼西亞）等地之HIV血清反應陽性之注射藥癮者（PWID），經分派為實驗組者，提供個體化系統導航、心理社會諮詢以及抗反轉錄病毒治療（ART）的整合性介入措施；而分派在對照組的參與者，則根據當地的現場指引給予標準護理。正如預期的那樣，作者發現介入措施對實驗組之參與者其病毒被抑制以及接受藥物輔助治療（因為成癮物質使用，medication-assisted treatment, MAT）上增加了比例；其與標準治療相比死亡率亦降低。以前的研究發現，系統導航者有助於改善高收入環境中HIV陽性注射藥癮者的結果；因此，良好導航可能對許多中、低收入環境中糾結的官僚體系有所助益的發現，並不令人訝異。
該研究最有趣的方面是要求指標參與者招募其注射網絡中的一個或多個HIV血清陰性成員參與該研究，並在這些注射伴侶中測量其HIV發生率。作者發現，介入組注射伴侶的HIV發生率低於標準治療組的注射伴侶（介入組，沒有新病例/每100人年，95％CI, 0.0~1.7；標準照護組1.0新病例/每100人年，0.4~2.1;發病率差異為-1.0/每100人年，-2·1比1·1）。該研究沒有為這一結果提供統計學上足夠之效力，其差異並不顯著；標準照護組的預期發病率為3例/每100人年。而研究中標準治療參與者的注射伴侶中觀察到的低發病率，亦引發了關於以HIV感染為研究結果其倫理和實務上的問題，因為對照組被拒絕提供相關介入措施。在探究標準照護組中之高度昜感群體結果，究竟於什麼時候足夠弱但在道德上又能提供可接受的照護標準？以致於其結果在嚴格的研究可被確認時可能會使研究極其昂貴。在Miller及其同事研究中，進一步研究去觀察到發生率上之差異，將需要約3500人年的追蹤。
然而，根據這項研究的結果和我們自己的研究，我們建議在以愛滋病毒發生率作為結果的研究中，應該為介入組和標準照護組提供更高的照護標準。我們建議，當發現愛滋病毒感染事件時，應立即提供抗反轉錄病毒治療和藥物輔助治療(MAT)，並應做出一切合理的嘗試，以確定潛在的傳播源和找出在該個案其注射和性網絡中可能感染的人。對於注射和性網絡的成員們，應向HIV血清反應陽性的個體提供抗反轉錄病毒治療，並應向成癮藥物使用障礙者提供MAT，並應向所有網絡成員提供足夠的無菌注射設施和保險套。這些額外的步驟可能會使數據分析複雜化，但並不致於需要非常多的資源，應該可以減少愛滋病毒從高傳染性個體向外傳播的可能，並於解釋愛滋病毒發生率上提供有價值的數據。
Miller及其同事研究中，兩組愛滋病毒發生率出乎意料地低的具體原因很難歸因。正如作者所指出的，研究要求儘量廣擴可能會導致霍桑效應，同一注射網絡至少兩個成員的參與也可以產生社會支持以達成研究所期望的行為。此外，指標參與者和其注射伴侶或伴侶之間知曉其HIV血清狀態不一致的情況下，不僅可能，且如果無菌注射設施是供應充足 – 應該會降低指標參與者與其注射伴侶之間的風險行為。
本研究中發現的介入組和標準組的愛滋病毒發病率極低，再次提出了是否有可能在低收入和中等收入國家的注射藥癮者中結束其愛滋病毒流行。在東歐、中亞和東南亞注射藥癮問題正在推動該等地區的愛滋病毒流行。愛滋病毒綜合式的預防和照護措施，已經結束了許多高收入環境中注射藥癮者的愛滋病毒流行，而亦有大量證據顯示藥物輔助治療(MAT)在中、低收入水平地區環境中非常有效。
在中、低收入環境中，要在注射藥癮者中結束愛滋病毒流行仍然存在著多重挑戰，包括對財政資源上的明顯限制，缺乏訓練有素的衛生保健工作者，對注射藥癮者和愛滋病毒的污名化，以及適得其反的政策，如將注射藥癮者收容於拘留中心。然而，透過減少愛滋病毒傳播的適當政策承諾，這些問題都不應該是不可克服的。無論Miller及其同事的研究中所發現的愛滋病發生率低其因果關係之具體情況如何，我們相信如何減少注射藥癮者中的HIV傳播已經充分被了解，因此針對中、低收入國家注射藥癮者在結束其愛滋病流行的過程中，應該努力進行多項實施方面之科學研究。面臨的挑戰是從研究之研究組別中非常低的發生率要達到注射藥癮者人群中的低發生率。大規模實施高質量的介入措施應該在注射藥癮者社區中產生規範，以便支持MAT、ART和安全注射之實踐，並且還應該去創造群體免疫力。
任何大量具歧異性的注射藥癮者人群都不太可能徹底消除愛滋病毒傳播，因此我們建議採用以下操作定義來解決注射藥癮者中的愛滋病流行問題，例如我們在DRIVE研究中採用的那樣（NCT03526939）：愛滋病病毒感染率降低至0.5例或更少/每100人年，並且在沒有病毒抑制的情況下，HIV血清陽性盛行率在總注射藥癮者人群中為5％或更低。這些標準解決了減少新感染和向已感染者提供ART等相互關聯間的問題。
Miller及其同事的研究表明，實現這些標準應該是可行的，我們相信在越南海防市實現這些標準的前景良好。政府和以社區為基礎的組織，現在應該毫不含糊地承諾將使用基於證據的介入措施列為目標，在全球範圍內結束愛滋病毒之流行。
Ending HIV epidemics among people who inject drugs in LMICs
In The Lancet, William C Miller and colleagues1 describe their feasibility study of a randomised controlled trial for an integrated intervention of individualized system navigation and psychosocial counselling and antiretroviral therapy (ART) for people who inject drugs (PWID) and are HIV seropositive in Kyiv (Ukraine), Thai Nguyen (Vietnam), and Jakarta (Indonesia). Participants in the control group were given standard of care as per their local site guidelines. As expected, the authors found that the intervention increased the proportion of participants who were virally suppressed
and on medication-assisted treatment (MAT) for substance use, and reduced mortality compared with standard of care. Previous studies2 have found that system navigators are helpful for improving outcomes for PWID who are HIV seropositive in high-income settings; therefore, the finding that good navigation could be helpful in the tangled bureaucracies of many low-income and middle-income settings is not surprising.
The most interesting aspect of the study was that the index participants were asked to recruit one or more HIV seronegative members of their injection network to participate in the study, and HIV incidence was measured among these injection partners. The authors found that the incidence of HIV was lower among injection partners in the intervention group than among injection partners in the standard of care group (intervention, no new cases per 100 person-years, 95% CI 0·0 to 1·7; standard of care 1·0 new case per 100 person-years, 0·4 to 2·1; incidence rate difference –1·0 per 100 person-years, –2·1 to 1·1). The study was not powered for this outcome, and the difference was not significant. The expected incidence rate for the standard of care group was three cases per 100 personyears.
The observed low incidence among the injection partners of the standard of care participants in the study raises ethical and practical questions about research with HIV infection as an outcome and in which an intervention is denied to a control group. When is standard of care sufficiently weak that doing rigorous research results in the exploitation of a highly vulnerable group but supplying ethically acceptable standard of care might make the study extremely expensive? A further study to examine the difference in incidence observed in Miller and colleagues’ study would require approximately 3500 person-years of follow-up.
However, on the basis of the results of this study and our own research, we would propose that enhanced standard of care should be offered to both intervention and standard of care groups in studies in which HIV incidence is an outcome. We propose that, when an incident case of HIV is identified, ART and MAT should be offered immediately, and all reasonable attempts should be made to identify potential sources of transmission and people who might have been infected within the person’s injecting and sexual networks. For members of the injection and sexual networks, ART should be offered to individuals who are HIV seropositive, MAT should be offered to those with disorders, and adequate supplies of sterile injection equipment and condoms should be provided to all network members. These additional steps would probably complicate data analysis but should not require extraordinary resources, should reduce the likelihood of transmission of HIV from highly infectious individuals, and should provide valuable data for interpretation of HIV incidence.
A specific reason for the unexpectedly low incidences of HIV in both groups in Miller and colleagues’ study is difficult to attribute. As the authors note, the extensive study requirements could have resulted in a Hawthorne effect.3 Participation of at least two members of the same injection network could also have generated social support to achieve the desired behaviours of the study. Additionally, awareness of HIV serodiscordance between the index participant and their injection partner or partners not only could have but—if supplies of sterile injection equipment were sufficient—should have reduced risk behaviour between the index participant and their injection partner.
The very low incidences of HIV in both the intervention and standard of care groups found in this study again raise the question of whether it is possible to end HIV epidemics among PWID in low-income and middle-income countries. Injecting drug use is driving HIV epidemics in eastern Europe and central and southeast Asia.4 Combined prevention and care for HIV has ended HIV epidemics among PWID in many high-income settings,5 and substantial evidence6,7