Recommendations for the Follow-Up of Study Participants with Breakthrough HIV Infections during HIV/AIDS Biomedical Prevention Studies

Abstract

Objective

To facilitate collection of cumulative data on longitudinal HIV disease outcomes during HIV prevention studies by developing recommendations for follow-up of the relatively few study participants with breakthrough infections.

Design

We formed a working group to compare and contrast the various approaches taken in recent HIV prevention trials, to summarize the advantages and disadvantages associated with each, and to explore the feasibility of developing protocols for the long-term follow-up of seroconverters.

Methods

We reviewed study designs, objectives, and assessments in 15 interventional studies that followed HIV seroconverters. Protocol team members joined discussions of the various approaches and developed recommendations.

Results

Most HIV prevention clinical trials share a core set of objectives, including the description/comparison of virological, immunological and clinical course of HIV, and sometimes a comparison of pre-and post-seroconversion behavior. Long-term follow-up of seroconverters can be conducted in separate studies if the transition from parent protocol is effectively managed.

Conclusion

We recommend the development of harmonized seroconverter protocols. Although specific research questions in the post-seroconversion period may differ depending on prevention modality, harmonizing key evaluations would create an opportunity to ask overarching questions that inform the prevention field with respect to design and implementation of future combination prevention studies. Follow-up immediately post-seroconversion should be conducted in the parent protocol before rollover into a follow-up protocol. Development of specimen repositories with ample volumes for future assays, standardized definitions of infection, diagnosis and seroconversion dates, and harmonization of study objectives and sample collections at key time points are important.

III. Introduction

The natural history of Human Immunodeficiency Virus (HIV) infection has been an area of intense study since the start of the epidemic. Initially and throughout the epidemic, HIV-1 seroconverters were followed longitudinally to describe the clinical, virologic, and immunologic trajectory of treated[1, 2] and untreated HIV disease[3–6] and transmission[7]. Early studies also focused on psychosocial and behavioral factors associated with HIV transmission and infection[8, 9]. Long-term follow-up of HIV infected individuals in natural history studies sheds light on issues such as host genetic factors of resistance and long-term non-progression[10, 11], the impact of treatment on disease progression[12], HIV-1 transmission during and after seroconversion[13], and the effects of a positive test on sexual risk behavior[14]. Some groups followed seroconverters within the context of breakthrough infection studies conducted in high-risk populations not exposed to biomedical preventions to meet a variety of study objectives[15–18], and pathogenesis work during acute infection and long-term follow-up has elucidated the concept of “founder viruses” in the establishment of HIV infection[19, 20].

As it became clear that CD4⁺ T-cell counts [21, 22] and plasma HIV RNA[22, 23] during the first six months of infection were strong predictors of disease progression[24], the need for expensive long-term follow-up of seroconverters in HIV-1 prevention clinical trials was reduced for a time. For example, the HIV Prevention Trials Network’s HPTN 052, which showed that antiretroviral therapy (ART) drastically reduces the sexual transmission of HIV in heterosexual serodiscordant couples[25], did not follow seroconverters after obtaining initial samples and assessments. Also, behavioral prevention studies have not typically needed long-term follow-up of seroconverters to obtain study endpoints, which focused on the primary endpoint – incident HIV infection[26, 27].

More recently, however, vaccine and pre-exposure prophylaxis (PrEP) studies have sought to understand the effects of biomedical interventions on correlates of disease progression[28, 29]. In addition, as HIV infected individuals age, describing development of HIV-associated non-Acquired Immune Deficiency Syndrome (AIDS) related conditions in these cohorts of seroconverters who received PrEP and/or a vaccine may become important. Elucidating the correlations of HIV-associated inflammation will also require longitudinal data from seroconverters from diagnosis through ART. Increasingly, answering these contemporary research questions requires the short- and long-term follow-up of study participants who become infected with HIV-1[30], especially in phase 3 trials of partially effective biomedical prevention modalities.

Some networks, such as the HIV Vaccine Trials Network (HVTN)[31, 32] and the Microbicide Trials Network (MTN)[33], have already developed studies for the long-term study of participants who seroconverted while in follow-up during late phase prevention trials. However, harmonization of study protocols that follow seroconverters in such a way that would allow for cross-study analysis remains a challenge. Depending on the nature of the study question, standardizing follow-up and collection of data at key post-infection time points may yield increased statistical power, but more importantly it would create an opportunity to ask overarching questions about the course of HIV infection in individuals who participated in prevention trials, such as development of drug resistance, change in risk behavior, secondary transmission, and development of non-AIDS related complications. The ability to ask these questions in a systematic fashion will inform design of future intervention studies and demonstration projects. We therefore conducted a comparative analysis of recent or current HIV biomedical prevention studies to answer the questions: 1) How can retention of participants who acquire HIV in a prevention study and enroll in a seroconverter study be maximized; 2) What study objectives can and should be answered by the short- and long-term follow-up of seroconverters; and, 3) Which biologic specimens and assessments can and should be recommended for all studies that follow seroconverters.

IV. Methods

The leaders of the National Institutes of Health (NIH)-funded HIV/AIDS clinical trials networks requested that the Office of HIV/AIDS Network Coordination (HANC) convene senior representatives of study teams of network prevention studies that follow seroconverters[31–38] to compare and contrast consent and retention approaches, objectives, evaluations and sample collections, and to prepare recommendations. The Seroconverter Study Group (SSG), comprised of eight members, held a total of nine teleconference discussions during which the protocols were reviewed and recommendations were developed. Non-network prevention studies were also considered and senior representatives of these protocols were invited to join group discussions on an ad hoc basis[39–43].

V. Results

A. Study Design & Operational Details

A total of 15 studies that follow seroconverters were reviewed and compared. Table 1 provides an overview of these studies and Table 2 compares the aspects of study design that pertain to seroconversion. The studies represented a wide variety of prevention modalities, including microbicides, oral PrEP, and vaccines in phase 1 to 3 testing. Duration of follow-up post-seroconversion differs from study to study. In some cases, seroconverters are followed for a limited time in the “parent” study (e.g. MTN-003) before or while being transferred to a separate “rollover” study that only follows seroconverters (e.g. MTN-015); in other cases, the seroconverters are followed entirely within the parent study (e.g. iPrEx) (see Table 2).

Table 1.

Description of studies: geographic setting, study size and population, status, and duration of follow-up.

Study Type	Study	Participant age	Participant population	Location(s)	Estimated study size	Study start date	Study status as of 2 July ‘12	Duration of follow–up for non- seroconverters (months)
Microbicide Studies	CAPRISA 004	18–40	WSM	ZA	889	May ‘07	Concluded	≤30; avg. 18
	TRAPS	18–40	WSM	ZA	98	Nov ‘07	Ongoing	n/a
	MTN-003	18–45	WSM	Sub–Saharan Africa	5000	Aug ‘09	Closed to accrual	14–38
	MTN-015	Varies	WSM	MW, UG, ZA, ZM, ZW	500	Aug ‘08	Enrolling	n/a
Oral PrEP Studies	FEM–PrEP	18–35	WSM	KE, TZ, ZA, MW, ZM	3900	May ‘09	Closed to Follow-Up	14
	FEM–PrEP Sxr Sub–Study	18–35	WSM	KE, TZ, ZA, MW, ZM	55	May ‘09	Closed to Follow-Up	n/a
	HPTN 067	≥ 18	WSM, MSM	TH, ZA	360	Aug ‘11	Enrolling	8.5
	HPTN 069	≥ 18	MSM	US	400	Pending	Pending	11.5
	iPrEx	Varies	MSM	PE, EC, US, BR, TH, ZA	2499	Jul ‘07	Closed to Accrual	11–33
	iPrEx OLE	Varies	MSM	PE, EC, US, BR, TH, ZA	1500	Jun ‘11	Closed to enrollment	19
	Partners PrEP	18–65	HIV neg. partners in discordant couples	KE, UG	3900	Oct ‘09	Closed to Follow-Up	25–37
Vaccine Studies	HVTN 403	Varies	W, M	BR, PE, US, ZA	54	Apr ‘02	Closed to Follow-Up	n/a
	HVTN 404	Varies	W, M	HT, PE, US, ZA	80	Jul ‘08	Enrolling	n/a
	HVTN 505	18–50	MSM*	US	2200	May ‘09	Enrolling	60
	HVTN 802	Varies	W, M	DO, HT, PE, US, ZA	230	Jul ‘08	Enrolling	n/a

Seroconverter rollover studies are shaded.

^*Ad5 nAb negative, circumcised, also includes M - > F transgender; BR Brazil; DO Dominican Republic; EC Ecuador; HT Haiti; KE Kenya; MW Malawi; PE Peru; TZ Tanzania; UG Uganda; US United States; ZA South Africa; ZM Zambia; ZW Zimbabwe; WSM Women who have sex with men; MSM Men who have sex with men; W Women; M Men; n/a Not applicable

Table 2.

Description of Studies: Seroconverter Follow-Up.

Study Type	Study	Sxr rollover study	Parent study	Associated parent study(ies)	Definitions of infxn, dx or sxn dates, as used in the study	Transition point from parent study	Number of new infxns during study	Duration of follow-up postsxn (months)
Microbicide Studies	CAPRISA 004	−	+	n/a	Sxn: Midpoint date between last negative and first positive antibody test OR 14 days prior to first positive confirmatory PCR, whichever is earlier	n/a	98	3
	TRAPS	+	−	CAPRISA 004	n/a	HIV dx	n/a	Until ART
	MTN-003	−	+	n/a	Dx: Specimen collection date of first positive antibody test	n/a	217	14–38, or until MTN 015
	MTN-015	+	−	MTN trials (e.g. MTN 003, 018, 020); HPTN 035	n/a	ASAP	n/a	Indefinite
Oral PrEP Studies	FEM–PrEP	−	+	n/a	Infxn: Midpoint between date of last negative sample and first positive sample by PCR	n/a	68	≤12
	FEM–PrEP Sxr Sub–Study	+	−	FEM–PrEP	n/a	HIV dx	n/a	12
	HPTN 067	−	+	n/a	Dx: Date of first positive RNA or antibody test	n/a	ukn	≤ 12
	HPTN 069	−	+	n/a	Dx: Date of first positive RNA or antibody test	n/a	n/a	≤ 11.5
	iPrEx	−	+	n/a	Dx: Date of first positive antibody test	n/a	100	11–33
	iPrEx OLE	−	+	n/a	Dx: Date of first positive antibody test	n/a	ukn	19
	Partners	−	+	n/a	Dx: Date of first positive antibody test	n/a	78	12
Vaccine Studies	HVTN 403	+	−	Phase 1 & 2 vaccine trials	n/a	ASAP	n/a	60–79
	HVTN 404	+	−	Phase 1 & 2a vaccine trials	n/a	ASAP	n/a	≤84; Until ART or CD4 <200
	HVTN 505	−	+	n/a	Dx: Date of visit with first detectable viral load but negative antibody test	n/a	ukn	17
	HVTN 802	+	−	HVTN 503, 504, 505	n/a	ASAP	n/a	≤ 144

Seroconverter rollover studies are shaded.

+ yes; − no; dx diagnosis; ASAP as soon as possible after HIV diagnosis; n/a Not applicable; infxn infection; sxn seroconversion; ukn Unknown

Seroconverter rollover studies typically enroll seroconverters from the parent study as soon as possible after HIV diagnosis. However, the timing is usually flexible to allow for maximum enrollment. The duration of follow-up also varies. Although perhaps for different reasons, both vaccine and biomedical studies follow seroconverters after ART initiation.

The method used to determine HIV diagnosis, infection, or seroconversion (Table 2) and frequency of HIV testing (Table 3) were compared across studies. Most studies made use of a diagnosis date, although FEM-PrEP estimated infection date and Centre for the AIDS Programme of Research in South Africa (CAPRISA 004) estimated seroconversion date.

Table 3.

Sample collections, assessments and visit frequency in protocols that follow seroconverters.

Study Type	Microbicide Studies				Oral PrEP Studies							Vaccine Studies
Protocol	CAPRISA 004	TRAPS	MTN-003	MTN-015	FEM–PrEP	FEM–PrEP Sxr Sub–Study	HPTN 067	HPTN 069	iPrEx	iPrEx OLE	Partners	HVTN 403	HVTN 404	HVTN 505	HVTN 802
Frequency of HIV tests until dx (weeks)	4	n/a	4	n/a	4	n/a	2–4	2–8	4	4–12	4	n/a	n/a	4–12	n/a
Frequency of visits after dx or screening/enrollment (weeks)	1–12	1–12	4	2–26	1–4	4	4–12	2–8	4–12	12	12	4–26	26	2–4	2–26
Max blood volume after dx (mL)	165	165	28	86	20	30	20	20	53	63	21	250	135	155	140
Genomics information collected? Separate consent needed?	−	+/+	−	−	−	−	−	−	+/+	−	+/−	−	−	+/−	+/+
Whole blood	+	+	+	+	+	+	+	+	+	+	+s	+	+	+	+
Serum/plasma	+s	+s	+s	+s	+s	+s	+s	+s	+s	+s	+s	+s	+s	+s	+s
PBMC	+s	+s	−	+s	+	+	+s	+	+s	+s	−	+s	+s	+	+s
Genital/rectal secretion and/or biopsy	+s	+s	+s	+s	+	+	−	+	+sub	+sub	+s	+	−	+	−
Urine	+	+	+	+	+	+	+s	+	+	+	+s	−	+	−	+
Dried blood spots	−	−	−	−	−	−	+s	−	−	+s	−	−	−	−	−
Hair	−	−	+sub	−	−	−	+s	+s	+sub	+sub	−	−	−	−	−
Oropharyngeal swab	−	−	−	−	−	−	−	−	+sub	+sub	−	−	−	−	−
Ongoing behavioral risk assessment	+	+	+	+	+	+	+	+	+	−	+	−	−	+	+

Seroconverter rollover protocols are shaded.

dx diagnosis; n/a not applicable; + yes; − no; s stored; sub sub-study

B. Objectives

We compared the objectives for each study (Table 4). Some interpretation and generalization was necessary to develop a meaningful comparison. The most common objectives, which were to describe/compare the virological, immunological and clinical course of disease and to assess drug resistance, are shown in Table 4 along with three other objectives pertaining to behavioral outcomes, evaluation of response to ART and development of specimen repositories, which were less common but of particular interest to the SSG. There were also a variety of virologic, immunologic, mutation/drug resistance, and other objectives that were used in only one or two studies; these additional objectives typically related to the intervention type (see Supplemental Digital Content 1 and 2, detailed objectives of each protocol).

Table 4.

Incorporation of recommended objectives.

Study Type	Microbicide Studies				Oral PrEP Studies							Vaccine Studies
Study	CAPRISA 004	TRAPS	MTN-003	MTN-015	FEM–PrEP	FEM–PrEP Sxr Sub–Study	HPTN 067	HPTN 069	iPrEx	iPrEx OLE	Partners	HVTN 403	HVTN 404	HVTN 505	HVTN 802
Describe/compare virological course of disease	+	+	−	+	+	+	+	−	+	+	+	+	+	+	+
Describe/compare immunological course of disease	+	+	−	+	+	+	−	−	+	+	+	+	+	+	+
Describe/compare clinical course of disease	+	+	−	+	−	+	−	−	+	−	+	+	+	+	+
Assess drug resistance	+	+	+	+	+	+	+	+	+	+	+	−	−	−	−
Describe/compare the psychosocial and behavioral changes	−	−	−	+	+	−	−	−	+	−	−	−	−	−	−
Evaluate virologic and/or immunologic response to ART	−	−*	−	+	−	−	−	−	−	−	−	+	−	−	+
Provide a repository of specimens that can be used for future analyses.	−	+	−	+	−	−	−	−	−	−	−	+	+	−	−

Seroconverter rollover studies are shaded.

+ yes; − no;

^*Participants initiating ART are enrolled into CAPRISA 009.

C. Assessments & Samples

Assessments and samples collected were reviewed for each study (Table 3). Whole blood, serum/plasma and peripheral blood mononuclear cells (PBMC), at least one type of genital or rectal secretion or biopsy, and urine samples were commonly collected. Although few protocols included study objectives specific to behavior, most studies conducted ongoing behavioral assessments. Only three studies included study objectives for development of specimen repositories, but all studies collected at least serum/plasma for storage.

The frequency and volume of blood draws varied across protocols, but the trends were similar. Clinic visits and blood draws are frequent in the weeks and months following HIV diagnosis and become less frequent as time passes. The most intense period of follow-up is typically the first three months. See Supplemental Digital Content 3 for more detailed information about the frequency of HIV testing and visits.

VI. Discussion

Various organizations conducting HIV prevention trials have been committed to following individuals who seroconvert while enrolled in HIV prevention studies by either incorporating 3 to 38 months of follow-up into the parent study and/or offering enrollment in a separate seroconverter rollover study. Such studies have enabled evaluation of clinical, immunologic, virologic, drug resistance, and other outcomes. A common protocol design and/or common protocol, such as a network-specific or cross-network seroconverter study for the long-term follow-up of study participants who seroconvert while participating in a prevention study, can ease aggregation of data. A common protocol design would need to consider consent, retention, study objectives, sample collections and assessments, and consistent determination and use of diagnosis, estimated infection, and/or estimated seroconversion dates. For advancing the greater HIV prevention research agenda, early negotiation of data sharing agreements among study teams would be ideal.

A. Recommendations: Study Design & Operational Details

An important consideration when planning to follow seroconverters is to determine how long to follow them within the context of the parent study and when to ask them to enroll in a seroconverter rollover study. The timing of transition into a rollover study impacts the number of participants recruited and long-term retention. Increasing retention (i.e. the proportion of all seroconverters in a parent study that roll over into the seroconverter study) is crucial to obtain sufficient sample size for comparative analysis. There is currently insufficient data on retention because many parent studies are ongoing and remain blinded, so it is premature to recommend an optimal time point for transition from parent to rollover studies. However, anecdotally, retention is higher and easier when seroconverters are followed within the context of the parent study for some period of time. Firstly, consent for remaining in an HIV prevention study in the event of seroconversion can be obtained from all participants at the beginning of the parent study, but consent for enrollment in a long-term seroconverter follow-up study must be obtained separately. Also, follow-up visits specific to seroconverters can be coordinated with regular study visits, making it more convenient for the participants and further promoting retention.

Following seroconverters within the context of the parent studies for a limited period also facilitates the capture of critical early post-diagnosis time point data; early time points, evaluations and events may pertain to the objectives of the parent study. The MTN-015 study team reported that enrollment delays have occurred while HIV diagnoses are confirmed.

Therefore, the MTN and other groups incorporated collection of early post-diagnosis samples into the parent protocols. Another strategy for increasing retention in both the parent study at the time of HIV diagnosis and enrollment into a rollover study is to have a dedicated team of counselors meet with participants who are diagnosed with HIV infection, as was done during the CAPRISA 004 study when participants were asked to enroll in Tenofovir gel Research for Advancing Prevention Science (TRAPS). HIV diagnosis is a difficult and challenging time both for the participants and the counselors/clinicians who work with them. Counseling by a trained, dedicated team provides needed support and encouragement to remain on-study.

Building short-term seroconverter follow-up into the parent study promotes retention at the point of HIV diagnosis, allows time for obtaining consent for the rollover study, and facilitates collection of clinical data and specimens during early post-diagnosis time points, but long-term follow-up of seroconverters is costly and difficult to budget. Enrolling seroconverters from multiple parent studies into a single or a few seroconverter rollover studies may be more efficient and logistically feasible. Based on these considerations the SSG makes the protocol development, implementation and timing recommendations in Table 5. When developing a seroconverter rollover study and determining the optimal point of enrollment transition, the desire to ensure retention and capture early time points post-diagnosis has to be weighed against any potential gain in efficiency and cost savings of a rollover study.

Table 5.

Summary of recommendations

Component	Recommendations
Development and implementation of seroconverter studies	Develop network-specific or cross-network seroconverter studies for the long-term follow-up of study participants who seroconvert while participating in a network prevention study
Timing of enrollment into seroconverter rollover studies	Follow seroconverters in the short-term within the context of the parent study before or concurrent with rollover into a seroconverter follow-up study. Obtain consent for short- and long-term follow-up post-seroconversion, including the collection of samples for future genomic testing, at the time of enrollment
Study objectives	Develop for all prevention trials and all seroconverter rollover studies a core set of objectives; allow study teams to add study-specific objectives as necessary. These objectives should include: Describe/compare the virological, immunological and clinical course of HIV disease among seroconverters in active compared to control arm. Assess drug resistance. Compare changes in risk behavior pre- and post-seroconversion. Evaluate response to therapy. Bank specimens in ample volumes for future assays, including genomic testing.
Other recommendations for enhanced standardization of data collection	Standardize definition, determination, and use of diagnosis dates, estimated infection date, and estimated seroconversion date.
Sample collections, biological assays, assessments and repository development	Develop a core set of sample collections and assessments for all prevention trials and all seroconverter rollover studies; allow study teams to add study-specific sample collections and assessments as necessary. Examples might include:
	Samples & Assessments	Pre-sxn:	0–3 mo. Post-sxn	6–12 mo. Post-sxn	>12 mo. Post-sxn
	HIV rapid test	M	–	–	–
	Urine/pregnancy test (women only)	M	M	Q	S
	Clinical assessment	M	M	Q	S
	Behavioral assessment	M	M	Q	S
	Blood for HIV RNA testing and storage	Q	M	Q	S

M Monthly; Q Quarterly; S Semi-annually; sxn = seroconversion

Duration of follow-up post-diagnosis would need to vary by necessity, depending on the objectives of the parent and rollover studies. For example, questions regarding acute infection and correlates of harm and protection would require a short duration of follow-up, perhaps until participants reach viral set point. Those studies looking at disease progressions might follow participants until initiation of treatment. Studies looking at response to treatment or co-morbidities such as chronic inflammation would necessitate a longer duration of follow-up. To mitigate the costs of an extended follow-up period, the frequency of visits would decrease over time. Potentially, novel and inexpensive point-of-care testing technologies could be incorporated to save cost and decrease burden on participants and clinics.

B. Recommendations: Study Objectives

The studies were compared to determine the common study objectives that would require follow-up of seroconverters, and to recommend a core set of these objectives for inclusion in all protocols that follow seroconverters (both parent and rollover studies). Delineation of a core set of objectives would facilitate protocol development and future cross-study analyses of clinical and laboratory data. The SSG found there were a few objectives that were common to many studies, and also identified some objectives that are currently included in two to four studies, but whose inclusion in future studies would be recommended (see Table 5).

The most common objectives shared across all protocols were to describe and compare the virologic, immunologic and clinical course of HIV disease among seroconverters in active and placebo arms and to assess drug resistance; these objectives are included in the core set of recommended objectives. Although most studies collect behavioral risk information, very few have objectives related to behavioral risk. While these objectives are not common, they would be recommended for inclusion in a core set of objectives. Similarly, while development of a specimen repository is a specific objective in only four studies, it is recommended as one of the core objectives. An evaluation of response to ART is also recommended, as selected drug resistance (due to oral or topical ARV-based PrEP), or disease progression modulation (via vaccine induction) may impact the natural history of treated and untreated disease. There were many objectives that were common to only a small number of studies (one to three), but would not be suitable for inclusion in a core set of objectives. These objectives would be considered to be study-specific, and are examples of objectives that a protocol team might choose to include in addition to the recommended core objectives as appropriate (see Supplemental Digital Content 1 and 2, detailed objectives of each protocol).

One might expect differences between vaccine and PrEP (microbicide and oral) protocols, but many similarities were found, indicating that a core set of seroconverter-related objectives could be used in vaccine and PrEP studies. Furthermore, as uptake of oral PrEP increases and vaccine or other biomedical prevention studies allow for or provide PrEP, rollover studies that include participants from vaccine studies may need to evaluate the effects of PrEP on seroconversion and other post-seroconversion endpoints.

C. Recommendations: Sample Collections & Assessments

To facilitate analysis of clinical and laboratory data across studies, the protocols were compared to determine the common sample collections/assessments and to develop a core set that would be recommended for inclusion in all studies that follow seroconverters. We found that the samples collected varied across protocols, but those supporting the common objectives described above (whole blood, serum/plasma and PBMCs), urine for pregnancy testing, and various genital secretions and/or biopsies are collected for most studies (Table 4) and would be recommended for inclusion in a core set of samples.

Behavioral assessments are collected in most studies both before and after diagnosis, but will pose particular challenges to cross-protocol analysis. The specific behavioral assessments, time frame for reporting, method of administration, and scales all vary across protocols, impeding cross-protocol comparison. While acknowledging that no gold standard exists for behavioral assessments, development of best practices would facilitate both cross-protocol comparison and development of future protocols.

Ideally, specimens (e.g. blood, cells, and genital secretions) should be banked in ample volumes for future assays. While only three rollover protocols (MTN-015, HVTN 403, and HVTN 404) have specific study objectives for developing a repository of specimens for future analysis, most of the protocols store samples for future testing at the site or at a central repository. Most commonly stored are serum/plasma and cryopreserved PBMCs; collection and repository storage of these samples is recommended. Acknowledging that storage of such specimens is particularly costly and poses additional challenges in international settings, such a repository would be high-yield for answering both current and future research questions, many of which may be exploratory and/or not fully developed at the time of protocol conduct. Collection and storage of other sample types could be added to meet study-specific objectives.

Maximum blood draw volumes varied widely, ranging from 20mL to 250mL. It’s unclear what factors contributed to this range, but reports from study teams suggested that local restrictions, cultural norms and ethical considerations may be a limiting factor. It has been reported that some African communities believe a loss of blood leads to a loss of energy, immune function, and/or virility, or that it might be used in magic[44–47]. Culturally sensitive training for site staff and participants may be required to support collection of blood in volumes sufficient for real-time and future testing, Some investigators report that there have been challenges to obtaining open-ended approval for genomic studies from national/local ethics committees, which may require more specific analysis plans and details for future testing than are typically needed for regulatory approval at U.S. domestic sites. Ideally, collection of genomic information would be included in a core set of sample collections/assessments, although regulatory requirements may prove to be challenging.

Standardized timing of sample collections/assessments in various protocols must also be considered if meaningful conclusions are to be drawn, especially for early infection events. For follow-up schedules and analysis, post-seroconversion timelines should be defined consistently and studies should be explicit in what primary data was acquired (e.g. diagnosis date, estimated infection date, estimated seroconversion date). Diagnosis is the most feasible trigger for postseroconversion follow-up, but depending on the frequency of testing, diagnosis could occur at different phases of actual infection, leading to impacts on analyses of timing of early infection events, for which estimated infection and seroconversion dates may be more useful.

Consistent definitions, determinations and uses of diagnosis, estimated infection, and estimated seroconversion dates, will depend on a common frequency of rapid HIV testing and storage of plasma/blood samples across parent studies. To more accurately estimate dates of infection and seroconversion, plasma samples or dried blood spots could be stored from all visits in case HIV RNA testing is needed at a later date. Such an approach was taken in FEM-PrEP, which used monthly antibody testing and collection of plasma, and HVTN 505, which employed antibody testing and storage of blood samples less frequently. During protocol development, the cost of collecting and storing plasma/blood samples for possible HIV RNA analysis would need to be balanced against the analytical needs.

Frequent (monthly) HIV testing is advantageous, facilitating an accurate estimation of seroconversion and/or infection dates and collection of samples/assessments at early time points. Furthermore, although study participants are counseled to be alert for signs of acute infection syndrome, few seroconverters are identified by clinical symptoms[48]. Most seroconverters do experience clinical symptoms attributable to acute HIV infection, but they do not always realize or report it[49]. Therefore, frequent HIV testing also has the advantage of identifying more participants during acute infection. Also, HIV testing reduces the potential for prolonged exposure to incompletely suppressive ART in the face of occult/undetected acute/primary HIV infection. We found that HIV antibody testing is usually done on a monthly or quarterly basis; testing less frequently than monthly not only complicates estimation of seroconversion date but could also increase seroconverters’ exposure to sub-optimal treatment regimens. However, the cost and feasibility of monthly testing must also be considered. The Centers for Disease Control and Prevention’s (CDC) interim guidance for the use of pre-exposure prophylaxis for prevention of HIV infection in men who have sex with men now calls for testing every two to three months[50] and some upcoming studies, such as HPTN 069[37], will use quarterly testing to reduce costs. Although monthly testing for HIV is costly, new testing technologies (e.g. combined antibody/antigen tests) might make it more feasible. Until then, the advantages of frequent HIV testing must be balanced against its costs and participant acceptability of such frequent assessments. Protocol teams might adjust the frequency of HIV testing to the intervention method. For example, vaccine studies in which participants are not taking PrEP might test quarterly; studies in which participants do take PrEP might reduce risk of prolonged exposure to sub-optimal treatments by using monthly testing.

Although considerable overlap exists among protocols in regard to sample collections/assessments, variation in timing complicates comparison across protocols. To more easily and effectively aggregate and compare findings from multiple protocols, the SSG makes the recommendations and provides examples in Table 5. These examples are based on a hypothetical harmonization of current and recent protocols. Design of a collaborative rollover protocol would require careful consideration of the study objectives, the necessary samples/assessments, and the available resources.

Due to a low rate of seroconversion in prevention studies and the need to follow seroconverters long-term to address current research questions, it would be advantageous to consider cross-sponsor data utilization agreements that would allow aggregation of data from multiple studies, and promote data usage by a broad research community. Aggregation of data can be made possible by a common seroconverter rollover protocol or protocol design that includes a core set of study objectives, samples and assessments. Common standards for consent and retention approaches and a core set of objectives, samples and assessments would require commitment and coordination across research teams. Study-specific goals could be accommodated by allowing study teams to include additional objectives, sample collections and assessments as needed.