Abstract

In a nonrandomized study of nonoccupational postexposure prophylaxis (PEP), a cross-sectional evaluation of subjects who were the source of human immunodeficiency (HIV) exposure was performed to characterize partners of index subjects seeking nonoccupational PEP against HIV. Among 401 index subjects, 64 (16%) recruited a source subject. Those in a steady relationship and those who knew that the source subject was HIV antibody positive were more likely to recruit their source subject. Source subjects reported high rates of past (78%) and current (69%) antiretroviral use; 46% of those using antiretroviral drugs had detectable plasma HIV-1 RNA levels. Antiretroviral resistance was detected in many source subjects who reported any use of antiretrovirals and was rare among source subjects who reported no history of antiretroviral use. Clinicians often make treatment decisions on the basis of incomplete knowledge of the source subject’s HIV status or antiretroviral treatment history. The treatment history, particularly nonuse of a class of antiretroviral drugs, can be used to predict drug resistance

Postexposure prophylaxis (PEP) to prevent human immunodeficiency virus (HIV) infection after occupational exposure to HIV is recommended by the US Public Health Service [1]. Although PEP is not recommended for nonoccupational exposures in the United States [2], it is prescribed both within [3, 4] and outside the United States [5,,8]. The HIV antibody status, antiretroviral treatment history, and plasma HIV-1 RNA level of the source of exposure may assist in constructing a PEP regimen to which the virus is least likely to be resistant [1]

In the nonoccupational setting, such information often will be obtained by self-report from the source subject to the exposed (index) person. Few studies report the HIV status of the potential source of exposure [5, 6], and none has described their demographic characteristics or antiretroviral medication histories. We determined the characteristics of index subjects associated with their ability to successfully recruit their source subjects shortly after seeking PEP, the source subject’s accuracy of self-reported HIV antibody status and HIV RNA levels, and their medication history and viral resistance characteristics

Methods

Study designBetween December 1997 and April 1999, HIV-uninfected persons reporting a potential sexual or injection drug use exposure to HIV in the previous 72 h were enrolled in a feasibility study of nonoccupational PEP [3]. They received 28 days of antiretroviral medications and 5 sessions of risk reduction counseling and were followed-up for 52 weeks for HIV seroconversion. PEP regimens were modified by using information obtained by direct questioning of the exposure source subject enrolled in the study or by index subject report. If the source subject’s HIV antibody status and/or antiretroviral treatment history were unknown, zidovudine plus lamivudine (Combivir; Glaxo SmithKline) were recommended. If the treatment history was known, a 2-drug nucleoside analogue combination to which the virus was least likely to be resistant was recommended. Nelfinavir was offered if the source subject reported currently having detectable plasma HIV RNA while receiving antiretroviral therapy

Identification and recruitment of source subjectsTen of the 30min allotted to counseling at each visit were devoted to source subject recruitment. Potential benefits discussed included discontinuing PEP medications if the source subject tested HIV antibody negative and altering PEP regimens to include drugs that would be more efficacious. Counselors addressed concerns regarding regret, shame, guilt, personal responsibility, reluctance to cause stress in the source subject, the possibility of a violent reaction by the source subject, and confidentiality. To help index subjects find anonymous or casual acquaintances, counselors explored where the index and source subjects had met, establishments frequented, shared friends, and like topics. Source subjects were contacted directly by the index subjects

Evaluation of source subjectsSource subjects were asked about their HIV infection status, most recent plasma HIV RNA level, and current and past use of antiretroviral medication. Blood was obtained for HIV antibody testing and plasma HIV-1 RNA determinations (Bayer Quantiplex version 2.0 [until 12/98] or 3.0). Genotypic resistance was assessed by population sequencing of the entire protease gene (pro) and the first 250 codons of the reverse-transcriptase (pol) gene (Visible Genetics) [9]. Phenotypic testing was done by a vector-based assay (Phenosense; ViroLogic) [10] that uses established criteria to define reduced susceptibility for each drug on the basis of correlations with drug activity in vivo

Statistical analysisTimely recruitment was arbitrarily defined as that occurring within 4 days of index presentation; information collected later may be less useful in altering PEP regimens. Associations between index subject characteristics and timely source subjectrecruitment were evaluated with unadjusted bivariate techniques and multiple logistic regression. All P values are 2-tailed. Analyses were conducted with Stata software (version 6.0, StataCorp)

Results

Of the 401 index subjects, 64 (16%) each recruited a source subject. Thirty-two (50%) were recruited on the day of index subject enrollment, 8 (12.5%) between days 1 and 4, 8 (12.5%) between days 5 and 7, and 16 (25%) between days 8 and 120. In determining factors associated with timely source subject recruitment, 8 (23%) of 35 women recruited their source subject within the first 4 days, compared with 31 (8.5%) of 366 men (P=.01). Of 174 index subjects who reported that their source subject was HIV infected, 35 (20%) recruited their source subject within 4 days, compared with 4 (1.8%) of 227 who were uncertain about their source subject’s HIV antibody status (P<.001). Thirty (27%) index subjects who were in a steady relationship with their source subject recruited him or her during the first 4 days, compared with 8 (7.2%) of those in a nonsteadyrelationship (P<.001). Significant associations were not identified between index subject age, ethnicity, education, income, presenting exposure type, or condom use or failure and the timely recruitment of source subjects. In a multivariable logistic regression model, knowledge of the source subject’s HIV infection status (odds ratio [OR], 7.2; 95% confidence interval [CI], 2.4–22) and a steady relationship between the index and source subjects (OR, 7.1; 95% CI, 2.9–18) were the only independent factors associated with timely recruitment of source subjects

Source subject characteristicsThe 64 source subjects were predominantly male (n=59 [92%]), white (n=35 [55%]), and 25–40 years old (n=45 [70%]). Almost one-half (49%) of the sources engaged in insertive anal intercourse with the index subject; only 3% engaged solely in insertive oral sex with ejaculation. Of the 61 source subjects involved in sexual exposures, 34 (56%) did not use a condom and 27 (44%) reported that the condom broke or slipped. The majority (67%) of source subjects were in a steady relationship with the index subject

Fifty (78%) source subjects tested HIV antibody positive, and 13 (20%) tested HIV antibody negative; 1 who was being treated for HIV infection declined an antibody test (table 1). One source subject with undiagnosed HIV infection was identified among 14 source subjects who reported that they were HIV antibody negative. The majority of HIV antibody–positive source subjects (n=28 [56%]) had a detectable plasma HIV RNA level. The HIV-1 RNA level was reported correctly (within 0.5 log of the measured HIV RNA level) by 40% of source subjects, but 30% overestimated and 20% underestimated by more than 0.5 log, and 10% did not know their HIV RNA level

Table 1

Source subject human immunodeficiency virus (HIV) infection status.

Table 1

Source subject human immunodeficiency virus (HIV) infection status.

Of the 51 source subjects who were HIV infected, 40 (78%) reported ever using and 35 (69%) were currently using antiretroviral therapy (table 2). Although only 11 (22%) were currently taking zidovudine, an additional 16 (31%) reported prior use. Likewise, 26 (51%) reported current lamivudine use; an additional 10 (20%) reported prior use. Thirty-two (63%) were currently using a protease inhibitor (PI), with 39% using nelfinavir

Table 2

Source subject antiretroviral history.

Table 2

Source subject antiretroviral history.

Genotyping was performed successfully on 23 of 25 plasma specimens with HIV RNA levels >1000 copies/mL (table 3). HIV with ⩾1 genotypic mutations associated with reduced antiretroviral activity was identified in 14 (61%) of tested source subjects. Fourteen samples (61%) had ⩾1 mutation associated with decreased susceptibility to nucleoside analogue reverse-transcriptase inhibitors. Primary mutations in the protease gene were seen in 10 specimens (43%). Nine specimens (39%) showed a wild-type genotype in reverse-transcriptase and protease genes

Table 3

Genotypic and phenotypic antiretroviral drug resistance and relationship to antiretroviral drug use.

Table 3

Genotypic and phenotypic antiretroviral drug resistance and relationship to antiretroviral drug use.

Phenotypic antiretroviral drug susceptibility testing was completed successfully on specimens from 19 subjects with HIV RNA levels >1000 copies/mL (table 3). HIV with reduced susceptibility to ⩾1 antiretroviral drug was identified in 11 (58%) of tested source subjects. Ten source subjects (53%) had HIV with reduced susceptibility to ⩾1 nucleoside reverse-transcriptase inhibitor (NRTI), 4 (21%) to ⩾1 non-NRTI, and 10 (53%) to ⩾1 PI

Thirteen (46%) of the 28 HIV-infected source subjects with detectable plasma viremia reported current antiretroviral use, placing them at risk for harboring drug-resistant virus. The prevalence of resistance was highest for each class of antiretroviral drugs in the group of subjects who reported use of the specific drug or another drug in the same class. All subjects who reported never using any drug in a specific class had wild-type genotypes for that class. Two of 10 subjects who reported no use of zidovudine and 3 of 10 who reported no use of nelfinavir had detectable genotypic mutations associated with reduced susceptibility to these drugs. These subjects had used other drugs in the respective classes

Discussion

Knowledge of the HIV infection status of the source subject is the most important consideration in the decision to initiate PEP and in determining its cost-effectiveness [11]. Such information helps exposed persons weigh the risk of HIV transmission against the potential risks and costs associated with PEP medications. When the source subject is found to be HIV antibody negative, PEP medications can be discontinued. A risk and symptom history may help to ensure that the HIV antibody negative source is not in the preseroconversion “window period” of acute HIV infection before deciding to stop PEP medication

Source subject information is also important for tailoring PEP medication regimens. Among source subjects who reported no use of a particular class of antiretroviral drugs, little or no drug resistance was observed. In contrast, the prevalence of resistance to antiretroviral drugs or drug classes that were currently or previously used by viremic source subjects was much higher, with little difference in those who reported current or past antiviral drug use. Altogether, these data suggest that self-reported denial of antiretroviral treatment with a drug or class of drugs can be used as a reasonable surrogate to suggest that detectable resistance to drugs in that class is unlikely. Because the use of HIV drug resistance testing to construct an initial PEP regimen is limited by the time required to perform the assay, the antiretroviral treatment history is the most practical tool to predict drug resistance. The prediction of resistance could be enhanced with additional clinical information that was not collected in this study, including an assessment of adherence and HIV-1 RNA levels while receiving each combination of antiretrovirals [12]

It is difficult to recommend an empirical PEP regimen for use in San Francisco on the basis of our resistance data. Although high levels of zidovudine and lamivudine resistance were seen among source subjects with detectable HIV-1 RNA levels and a history of use of these agents, the alternatives are problematic. First, there is substantial cross-resistance among the NRTI class, so those with high-level resistance to zidovudine and lamivudine are likely to be resistant to didanosine and abacavir and possibly to stavudine. Second, the potential toxicities of the different agents, including peripheral neuropathy and life-threatening pancreatitis, must be weighed against the risk of HIV transmission. Many clinicians advocate the addition of a third agent. We saw no HIV seroconversions in the first 6 months of our study [3]. Although these data do not demonstrate the efficacy of a double nucleoside–based regimen (used by 97%), they do suggest that the need for more potent PEP regimens, even in the context of significant drug resistance among the source population, remains an unresolved problem

The difficulty of recruiting source subjects most likely reflects the difficulty that will be confronted in clinical practice. Access to source subject information may increase with more assertive problem-solving when exposed persons do not know their source well, by obtaining key medical history by telephone rather than requiring an on-site visit, and by offering home test kits. Our small source subject numbers may underrepresent the proportion of index subjects who were able to provide the study clinician with useful information about their source subject’s antiretroviral use, since we did not collect data on how frequently this information was accessed by the study clinician for source subjects who did not enroll in the study

Despite the focus of our source subject recruitment protocol on the biomedical benefits for the index subject, accessing source subjects may have broader HIV prevention benefits. HIV-negative and HIV-positive source subjects who engage in high-risk behavior may be provided with prevention counseling and referrals. Discussing and practicing communication strategies with the index subject is a risk reduction tool that may help prevent future exposures [13]. Finally, establishing comprehensive postexposure prevention programs that incorporate risk reduction counseling with medication may facilitate integration of clinical and prevention services that traditionally have been separated

Acknowledgments

We thank Lawrence Marsco, Dennis James, Beth Dillon, Ruth Gasparik, Steven Torosian, Cathie Poston, Sara Schleimer-Batya, John Melichar, Lynn Selby, Larry Hanbrook, Josh Partlow, Miriam Lima, Lisa Reyes, Stefan Rowniak, the San Francisco City Clinic staff, and faculty at the Positive Health Program, for caring for study participants; Sally Liska, Mary Clancy, Tarek Elbeik, Michael Barcellos, Lynette Sawyer, David Chernoff, Sam Lee, and Yolanda Lee, for laboratory assistance; Julie Gerberding, for providing valuable organizing help and intellectual input at the initiation of the study; and Bre Holt, for assistance in the preparation of the manuscript

References

1
Updated US Public Health Service guidelines for the management of occupational exposures to HBV, HCV, and HIV and recommendations for postexposure prophylaxis
MMWR Morb Mortal Wkly Rep
 , 
2001
, vol. 
50
 (pg. 
1
-
52
)
2
Centers for Disease Control and Prevention. Management of possible sexual, injecting-drug-use, or other nonoccupational exposure to HIV, including considerations related to antiretroviral therapy. Public Health Service statement. Centers for Disease Control and Prevention
MMWR Morb Mortal Wkly Rep
 , 
1998
, vol. 
47
 
(RR-17)
(pg. 
1
-
14
)
3
Kahn
JO
Martin
JN
Roland
ME
, et al.  . 
Feasibility of postexposure prophylaxis (PEP) against human immunodeficiency virus infection after sexual or injection drug use exposure: the San Francisco PEP Study
J Infect Dis
 , 
2001
, vol. 
183
 (pg. 
707
-
14
)
4
Kunches
LM
Meehan
TM
Boutwell
RC
McGuire
JF
Survey of nonoccupational HIV postexposure prophylaxis in hospital emergency departments
J Acquir Immune Defic Syndr
 , 
2001
, vol. 
26
 (pg. 
263
-
5
)
5
Correll
P
Smith
D
Hendry
O
Grulich
A
Nonoccupational HIV post exposure prophylaxis (PEP) in Australia [abstract WePeC4420]
Program and abstracts of the XIII International AIDS Conference
 , 
2000
Stockholm
International AIDS Society
6
Derouineau
J
Dhotte
P
Buttet
P
A retrospective study of 83 postexposure prophylaxis (PEP) tolerability in health dispensary Figuier, DASES, Paris, France [abstract 2065]
Program and abstracts of the 40th Interscience Conference on Antimicrobial Agents and Chemotherapy
 , 
2000
Washington, DC
American Society for Microbiology
pg. 
348
 
7
Wiebe
ER
Comay
SE
McGregor
M
Ducceschi
S
Offering HIV prophylaxis to people who have been sexually assaulted: 16 months’ experience in a sexual assault service
CMAJ
 , 
2000
, vol. 
162
 (pg. 
641
-
5
)
8
Rabaud
C
Bevilacqua
S
Beguinot
I
, et al.  . 
Tolerability of postexposure prophylaxis with zidovudine, lamivudine, and nelfinavir for human immunodeficiency virus infection
Clin Infect Dis
 , 
2001
, vol. 
32
 (pg. 
1494
-
5
)
9
Durant
J
Clevenbergh
P
Halfon
P
, et al.  . 
Drug-resistance genotyping in HIV-1 therapy: the VIRADAPT randomised controlled trial
Lancet
 , 
1999
, vol. 
353
 (pg. 
2195
-
9
)
10
Petropoulos
CJ
Parkin
NT
Limoli
KL
, et al.  . 
A novel phenotypic drug susceptibility assay for human immunodeficiency virus type 1
Antimicrob Agents Chemother
 , 
2000
, vol. 
44
 (pg. 
920
-
8
)
11
Pinkerton
SD
Holtgrave
DR
Bloom
FR
Cost-effectiveness of post-exposure prophylaxis following sexual exposure to HIV
AIDS
 , 
1998
, vol. 
12
 (pg. 
1067
-
78
)
12
Zolopa
AR
Shafer
RW
Warford
A
, et al.  . 
HIV-1 genotypic resistance patterns predict response to saquinavir-ritonavir therapy in patients in whom previous protease inhibitor therapy had failed
Ann Intern Med
 , 
1999
, vol. 
131
 (pg. 
813
-
21
)
13
Martin
JN
Roland
ME
Bamberger
JD
, et al.  . 
Post-exposure prophylaxis (PEP) for sexual exposure to HIV does not lead to increases in high risk behavior: the San Francisco PEP Project [abstract 224]
Program and abstracts of the 8th Conference on Retroviruses and Opportunistic Infections
 , 
2001
Alexandria, VA
Foundation for Retrovirology and Human Health
pg. 
108
 
Presented in part: 7th Conference on Retroviruses and Opportunistic Infections, Chicago, February 2000 (abstract 197)
The Committee on Human Research at the University of California at San Francisco approved the study protocol. Each participant provided written informed consent
Financial support: University of California at San Francisco (UCSF) AIDS Research Institute; National Institutes of Health (AI-42523); UCSF Center for AIDS Research (P30 MH59037); Center for AIDS Prevention Studies (MH4259); University of California University-Wide AIDS Research Program (CC97-0962 and CC99-SF-001); William McCarty-Cooper Trust; Glaxo-Wellcome; Bristol-Myers Squibb; Agouron Pharmaceuticals; Chiron; Bayer; Virologic; San Francisco Department of Public Health
J.O.K. is a principal investigator and M.E.R. is a coinvestigator, on studies with Bristol Myers Squibb, Agouron, Chiron, and Glaxo-Wellcome; J.O.K. is a consultant to Virologic; R.M.G. is a coinvestigator on studies with Visible Genetics and Virologic