PHI and Transmission Risk

Home , Infectivity

through 4, transmitters infected a steady PHI and Transmission Risk sexual partner via penile-vaginal inter- Reprinted from The PRN Notebook,™ course; in couple 5, transmission was via february 2002 insertive anal and oral sex. Dr. James F. Braun, Editor-in-Chief (Part 2): Epidemiology Tim Horn, Executive Editor. As shown in Figure 1, the single re- Published in New York City by the ported exposure occurred before the trans- Physicians’ Research Network, Inc.® John Graham Brown, mitter’s onset of symptoms for couples 4 and Infectivity Executive Director (day –2) and 5 (day –7). Couple 1 had mul- For further information and other articles tiple sexual exposures, but all occurred available online, visit prior to day +2 after the transmitter’s on- Christopher D. Pilcher, md http://www.PRN.org Research Assistant Professor of Medicine, All rights reserved. set of symptoms. A single exposure oc- © february 2002 curred on day +7 after symptom onset in University of North Carolina, Chapel Hill Chapel Hill, North Carolina couple 3. Observed incubation periods for transmitter 1, infected partner 3, and infected partner 4 were 20, 12, and 17 days, eams of epidemiological and biological data are now avail- respectively, consistent with previously able to suggest that people in the primary stages of hiv are, un- published observations (Schacker, 1996). knowingly, significant contributors to the spread of hiv and, “The conclusions we came to are basic, yet very important,” Dr. Pilcher said. “We consequently, the proliferation of the aids epidemic. However, can’t say that sexual transmission is more the precise extent to which individuals with phi play into this likely to occur during phi; this five-cou- unfortunate scenario remains unclear. To help make sense of ple study can’t tell us that. What we were the data that have emerged thus far—and to comment on its relevancy able to confirm is that hiv is readily transmitted during the acute stages of infec- within the realm of public health—Dr. Christopher Pilcher shared his on- tion, as early as seven days before the on- going experiences and thoughts with prn. set of symptoms. This is definitely some- thing to bear in mind when contemplating public health initiatives, especially when Complex mathematical models, con- studies—conducted at the University of you consider that the majority of acutely structed by teams of calculus-savvy epi- North Carolina in collaboration with sev- infected patients don’t present until symp- demiological groups, have suggested that eral prominent researchers scattered toms have developed.” individuals with phi are one of the most throughout Switzerland—involving five Moving on to some recent case-clus- important populations to target therapeu- patients drawn from four university-based tering data, Dr. Pilcher reviewed the pre- tically—even though they constitute a time- hospital clinics in whom sexual transmis- liminary results of a Swiss hiv Cohort limited minority population in the much sion was suspected to have occurred be- study, reported by Professor Sabine Yerly larger hiv community (see: “phi and Trans- tween an individual with documented phi and colleagues in AIDS (Yerly, 2001). The mission Risk (Part 1): The Contribution of and a sexual partner who later developed study included all individuals with docu- Mathematical Modeling,” beginning on documented phi (Pilcher, 2001). For the mented phi identified in six aids centers of page 8). There are also epidemiological sake of this study, phi was defined as hiv university hospitals in Switzerland and data indicating an increased incidence of p24 positivity, hiv-rna or hiv-dna posi- two aids centers of a hospital close to infection in partners of recent serocon- tivity, together with eia negativity or two or Geneva. Among the total of 197 individu- verters, compared to partners of chroni- fewer bands on Western blot. Each trans- als infected between January 1996 and cally infected men and women (Leynaert, mission pair was confirmed by phyloge- January 2000, phi was documented by 1998). In recent days, chilling data have netic analysis of hiv reverse transcriptase evolving hiv antibody response and/or surfaced to support the long-standing hy- sequences. symptoms consistent with acute retroviral pothesis that acutely infected people can The results of this analysis, published in syndrome within three months in 70% of infect their sexual partners in turn. These an October 2001 issue of the Journal of the individuals and by seroconversion within new data suggest that hiv infection can American Medical Association (JAMA), are 12 months of presentation in 30%. Se- be passed on serially from one acutely in- illustrated in Figure 1 (on the next page). quence analyses were performed on plas- fected individual to uninfected partners Three of the couples consisted of men who ma samples where plasma had been col- within a matter of days, and also that such transmitted hiv to female sex partners lected before the initiation of haart (avail- events may be quite common. (couples 2, 3, and 4); couple 1 involved a able for 193 of 197 subjects); a phyloge- woman who infected a male partner, and netic tree was constructed using a neigh- couple 5 involved a male who transmit- bor-joining method on available reverse ted the virus to his male partner. Couples 1 transcriptase sequences. phi and 2 reported frequent, regular sexual The phylogenetic analysis revealed sig- and the Serial intercourse during periods of possible ex- nificant “clustering” for 56 (29%) of 193 in- Transmission of hiv posure, whereas couples 3, 4, and 5 re- dividuals, indicating that the viruses from To address the issue of serial transmis- called only single sexual contacts during the patients in each cluster were genetically re- sion, Dr. Pilcher referred to one of his own time of possible transmission. In couples 1 lated. The eighteen clusters in this study

THE PRN NOTEBOOK™ • SPECIAL EDITION • FEBRUARY 2002 • WWW.PRN.ORG 11 ﬁgure 1. Timing of Clinical Events Within Transmission Pairs

Couple ID Patient Sex Timing of Clinical Events

1 Transmitter Woman * Infected Partner Man

2 Transmitter Man Infected Partner Woman Asymptomatic Documented Single Sexual Contact with a Transmitter

3 Transmitter Man Documented Repeated Sexual Infected Partner Woman Contact with a Transmitter

Symptoms in Transmitter 4 Transmitter Man Infected Partner Woman Symptoms in Infected Partner

Antiretroviral Therapy 5 Transmitter Man Infected Partner Man Asymptomatic

-28 -14 0 14 28 42 56 70 Days from Onset of Symptoms in Transmitter id indicates identiﬁcation number; asterisk (*) indicates that sexual contact shown for the transmitter in couple 1 represents a sexual assault on her by an hiv-positive individual. Seminal hiv-rna concentrations, collected from couple 5 only, were higher than commonly seen in chronic infection for both the transmitter and infected partner (5.7 and 5.9 log, respectively). Other sexually transmitted infections were found in couple 1 (genital herpes and chlamydia) and in couple 5 (genital herpes and early syphilis); in couple 4 the transmitter had a sterile inguinal abscess.

Source: Pilcher, 2001. Sexual transmission during the incubation period of primary hiv infection. JAMA 15(7):838-45. Reprinted with perimission of the American Medical Association. ranged in size from two to 11 phi indi- phi in Context individuals with chronic infection. This viduals per cluster and involved intra- connection is logical inasmuch as high vi- venous drug use, homosexual, and het- Dr. Pilcher was careful to point out that ral loads in blood might be indicative of erosexual modes of transmission. Retro- data concerning transmission risk during high viral loads in other infectious body spective contact tracing firmly established phi must be considered in the context of fluids, such as semen and vaginal fluid. the chain of transmission to explain 17 transmission rates and risk factors among According to a study spearheaded by (9%) of the 56 clustered infections. chronically infected hiv-positive people. Dr. Thomas Quinn of Johns Hopkins Uni- “The clusters in this analysis were re- Drawing upon the results of seven studies versity School of Medicine—conducted as ally quite noticeable,” remarked Dr. Pilch- that followed chronically infected individ- part of a larger community-based std in- er. “Approximately one-third of individu- uals engaging in penile-vaginal inter- tervention study conducted in the Rakai als with recent transmission harbored course—which, Dr. Pilcher added, is the district of Uganda—415 couples discor- variants that were genetically linked to route of transmission for approximately dant for hiv were followed for an aver- variants from other individuals with re- 75% of all people now being infected with age period of 22.5 months (Quinn, 2000). cent hiv transmission. Unfortunately, this hiv worldwide—the probability of hiv The male partner was infected with hiv in doesn’t tell us if the clustering was a result transmission was quite low. Taken to- 228/415 (55%) couples, and the female of rapid, serial transmissions or the result gether, these studies estimate hiv trans- partner was infected in 187/415 (45%) of multiple, parallel infections by core mission risk, per coital act, to be 1 in 500 couples. Collateral seroconversions oc- transmitters. Either way, we cannot un- to 1000 (0.001, ranging from 0.0008 to curred in 90/415 (22%) couples studied derestimate the importance of rapid con- 0.002). “If these numbers were accurate,” during the 30-month period. Interesting- tact tracing during this window of op- Dr. Pilcher said, “I don’t think we’d have ly, there was no difference in the male-to- portunity. In the event of serial transmis- much of an epidemic. Clearly, there are female and the female-to-male transmissions, contact tracing will allow for the in- some individuals who are more likely to sion rates: Both were 12%. fected partners to receive an early diag- transmit hiv than others. The question is: The viral load of the hiv-positive mem- nosis and possibly begin treatment. If it’s is it people with acute infection, or is it ber of a serodiscordant couple turned out core transmitters, contact tracing may some other population?” to be a major factor in collateral trans- aid in their identification and permit in- Dr. Pilcher went on to review a number mission. According to Dr. Quinn’s team’s tervention to curb the spread of infection of features of phi that might make trans- report, which is summarized in the De- by these individuals.” mission especially likely. The most obvious cember 2000 issue of The PRN Notebook, factor is viral load, which is at its highest hiv-positive men and women with hiv- during the acute retroviral syndrome and rna levels around 90,000 copies/mL were is a strong risk factor for transmission by more likely to transmit the virus to their

12 THE PRN NOTEBOOK™ • SPECIAL EDITION • FEBRUARY 2002 • WWW.PRN.ORG hiv-rna levels are shown for study subjects with primary hiv infection (phi) pri- figure 2. Compartmental hiv-rna Levels in or to antiretroviral therapy (darker boxes) and chronically infected, asymptomatic historical controls naive to antiretroviral therapy (lighter boxes). Upper and Primary Versus Chronic Infection lower fences represent the range, boxes represent the 25th to 75th interquartile cvl 8 range, and horizontal lines represent median values. = cervicovaginal lavage. P < .001 P = .02 phi 7 Source: Pilcher, 2001a. hiv in body fluids during primary hiv infection: implications for pathogenesis, treatment, and public health. AIDS 15(7):837-45. Reprinted with permission of Lippincott Williams and 6 Chronic Wilkins. 5 copies per ml copies 4 hiv-negative partner(s) than those with The Biological 3 lower hiv-rna levels (~40,000 copies/mL). Evidence rna hiv-1 In fact, there were no transmissions 2 among couples in which the hiv-positive While there is no shortage 1 Log 10 partner had a viral load less than 1,500 of data concluding that 0 copies/mL. Among hiv-positive partners hiv-rna levels in periph- N = 18 69 5 49 2 19 11 11 8 20 csf cvl with viral loads greater than 50,000 eral blood are exceedingly Blood Saliva Semen copies/mL, the risk of transmitting hiv high during phi, there is compartment was approximately 23% per year. And very little in the way of with each log increase in viral load, the data regarding hiv-rna concentrations in among the phi subjects but were not sig- risk of transmission increased 2.45-fold. the genital fluids and other peripheral com- nificantly higher overall (3.96 log) than However, Dr. Pilcher noted, it may not partments of acutely infected individuals. for the chronically infected controls (3.61 simply be differences in viral load power- “One thing we really wanted to find out log). However, a correlation was significant ing apparent differences between acute was the correlation between hiv levels in between semen and blood viral loads for and chronic infection transmission risk. the blood and those in other compartments, individuals. In addition, Dr. Pilcher noted For instance, partner susceptibility to hiv including semen and cervicovaginal lavage, that two subjects—so-called “hyperexcre- infection within stable partnerships may during phi,” Dr. Pilcher explained. “High tors”—had seminal hiv-rna levels that actually decrease over time because of ac- hiv-rna levels in these fluids would be very consistently exceeded concurrent blood quired mucosal immunity (Vernazza, 2000; likely to increase the risk of transmission; plasma levels on repeated measures. “In- Mazzoli, 1999; Langlade-Demoyen, 1994; thus we wanted to confirm this and also terestingly, these two hyperexcretors had Kelker, 1992). take a look at the effects of antiretroviral STDs, which is certainly consistent with The association between phi and other therapy on viral seeding and shedding in earlier studies showing that STDs increase acute STDs may also be important in aug- the compartments.” viral load in blood and semen. As for menting transmission in this group. In a A decidedly tall order to fill: Together treatment, we were successful in our abil- paper published recently by Professor with colleagues at the Duke-unc-Emory ity to reduce viral levels in all of the com- Ronald Gray and his colleagues, also at Acute hiv Consortium, Dr. Pilcher helped partments, which is certainly good news.” Johns Hopkins University Medical Center conduct an observational cohort study in and in Uganda’s Rakai district, STDs were which 17 individuals with phi provided once again shown to significantly increase various laboratory specimens, including Genital hiv Shedding and the risk of hiv transmission among het- blood plasma, cerebrospinal fluid (csf), Increased Transmission: erosexuals (Gray, 2001). In this analysis in- seminal fluid, cervicovaginal lavage, and/or volving 174 monogamous serodiscordant saliva (Pilcher, 2001a). Once the samples Is There a Link? couples, the overall per-coital-act of hiv were collected and processed, the viral “We were actually quite disappointed at transmission was 0.0011. In the setting of load of each fluid was compared to a cor- our inability to demonstrate that seminal an std associated with genital ulceration, responding sample collected from a hand- hiv-rna levels in our phi subjects were the probability of transmission jumped to ful of chronically infected, antiretroviral- significantly higher than our historical 0.0041, compared to a rate of 0.0011 naive hiv-positive patients serving as his- controls of chronically infected patients,” among couples in which neither partner torical controls. With the baseline assess- commented Dr. Pilcher. “In fact, the num- had an ulcerative std. ment completed, the phi subjects were bers were all over the place. We had our Behavioral factors could contribute to treated for six months with an antiretro- hyperexcretors, but we also had acutely in- high rates of hiv transmission as well. viral regimen consisting of didanosine fected individuals with moderate and low Two studies reviewed by Dr. Pilcher sug- (Videx), stavudine (Zerit), and nevirapine viral levels in semen.” But it’s important to gest that individuals with acute infection (Viramune)—either with or without hy- keep in mind that in this study, like other may have a higher number of sexual part- droxyurea (Hydrea)—and then provided studies in humans and animal models, ners than their chronically infected peers follow-up samples for analysis. Dr. Pilcher’s team demonstrated that hiv- (Colfax, 2000; Sey, 2001). “If individuals The results of the pretreatment com- rna levels in genital secretions really do with phi are engaging in risky behavior parisons between the subjects with phi mimic those in peripheral blood. It is with a number of different partners,” he and a group of chronically infected controls therefore likely that peak shedding, per- added, “this would further explain clusters are illustrated in Figure 2. Interestingly, haps very early in phi, increases the risk of of new infections.” hiv-rna levels in semen were variable transmission.

THE PRN NOTEBOOK™ • SPECIAL EDITION • FEBRUARY 2002 • WWW.PRN.ORG 13 ﬁgure 3. and quantifiable se- bilities of transmission over time for dif- Estimated Male-to-Female men samples were ferent individuals.” Per-Sexual-Contact hiv Transmission available, and 24 Dr. Pilcher’s observations are reported in women in whom the Table 1. In order to determine hypothetical 0.0180 number of endocervi- transmission probabilities over time for in- 0.0160 cal CCR5 receptors was dividuals with low, moderate, and high 0.0140 Receptor Cell: 25% determined. The out- peak and steady state hiv-rna levels in se- Receptor Cell: 50% 0.0120 come of the study was men, Dr. Pilcher assumed that changes in Receptor Cell: 75% 0.0100 a ﬁnal model equation the genital tract paralleled those that have 0.0080 allowing one to predict been well described in the blood compart- 0.0060 the probability of hiv ment. For instance, an individual with a 0.0040 transmission from very high setpoint seminal viral load of 7 log

Probability of Transmission 0.0020 men to women, per may have a peak seminal viral load of 8.85 0.0000 coital act, based on the log during phi. “For that individual,” Dr. 3 3.25 3.5 3.75 4 4.25 4.5 4.75 5 5.25 5.5 absolute burden of nsi- Pilcher commented, “the per-coital-act prob-

Log10 Seminal Viral Load in One Ejaculate hiv in a given man’s ability of transmission is 1.0—he’s bound to ejaculate (volume x transmit the virus to almost everyone he Estimated male-to-female per-sexual-contact hiv transmission probability hiv-rna copies/mL has unprotected intercourse with. The in- for different seminal viral loads and for different receptor cell counts seminal plasma) and a dividuals with low and moderate hiv-rna when 100% of the isolates in the semen are nsi. The horizontal axis rep- given woman’s recep- levels, both during peak and at setpoint, resents log10 seminal viral load in one ejaculate, and the vertical axis tor-cell density. are much less likely to transmit the virus.” represents the male-to-female per-sexual-contact hiv transmission With the data col- But what is really important to consider in probability. The three lines represent different receptor cells/mm2 lected and the applica- these patients with low and moderate viral counts: 25th percentile, 50th percentile, and 75th percentile. ble numbers plugged loads, Dr. Pilcher pointed out, is the fold- into the model, the re- change in hiv-rna concentrations from the Source: Chakraborty, 2001. Viral burden in genital secretions determines male-to-female sexual transmission of hiv-1: a probabilistic empiric model. AIDS 15(5):621-7. sulting equation yields peak to setpoint. “What we saw was ap- Reprinted with permission of Lippincott Williams and Wilkins. a number of concise proximately a 20-fold decrease in the prob- predictions. For ex- ability of transmission from peak to set- ample, when semen point. Looking at this another way, what we But to what extent might ﬂuctuations in contains 100,000 copies/mL of nsi hiv- end up seeing is a 20-fold increase in the seminal viral load, for instance, affect rna, the probability of hiv transmission is probability of transmission when viral shed- transmission risk during phi? For this, 1 per 100 episodes of intercourse. Con- ding in semen is at its peak.” Dr. Pilcher turned his attention to the work versely, with 1,000 copies/mL of nsi hiv- Another key variable discussed by Dr. of Dr. Hrishikesh Chakraborty and his col- rna in semen, the probability of trans- Pilcher was what the model may predict leagues, working with Drs. Joe Eron and mission falls to 3 per 10,000 coital acts. about the contribution of phi to an indi- Myron Cohen at the University of North More specific findings are illustrated in vidual’s total cumulative individual hiv Carolina (Chakraborty, 2001). Figure 3. transmission probability. For example, in Dr. Chakraborty’s team set out to de- “We liked this model and wanted to an individual who has peak seminal viral scribe a mathematical model to help pre- plug in our own numbers of genital hiv- load of 5.43 log, a setpoint seminal viral dict hiv transmission, in this case between rna levels seen during phi,” Dr. Pilcher load of 3.85, and a duration of chronic men and women. As reiterated by Dr. explained. “More speciﬁcally, we wanted to infection of ten years—numbers, Dr. Pilch- Pilcher, this model estimates sexual trans- calculate how the change in genital hiv- er reckons, that are representative of a mission as a function of both the infec- rna levels, from their peak concentrations large percentage of hiv-positive people— tiousness of the transmitter and the sus- to their setpoints, would affect the proba- the probability of transmission during phi ceptibility of the uninfected partner. By studying the con- table 1. Predictions of Combined Model for Three Hypothetical Individuals centration and genotype (syn- cytium-inducing [si] or non-syn- Seminal % Total Cumulative Individual cytium-inducing [nsi]) of hiv in Viral Load (log) Per-Act Transmission Probability Transmission Probability male genital secretions and the Peak Setpoint Peak Setpoint Fold- Acute Chronic (day number of receptors (CCR5) for (day 23) (day 120) (day 23) (day 120) Change (day 0–120) 120–10 years)* hiv in the endocervix of women, Dr. Chakraborty’s team 3.75 2.17 .0015 .0001 22.2 0.07 0.93 amassed biological data that could be made to square with 5.43 3.85 .0308 .0018 19.7 0.06 0.94 existing epidemiological data. Enrolled in this study were 8.85 7.27 1.000 .8484 1.4 0.02 0.98 86 men—none of whom were *Observed values from chronically infected cohort. receiving antiretroviral thera- Source: Christopher Pilcher, md py—in whom CD4+ cell counts

14 THE PRN NOTEBOOK™ • SPECIAL EDITION • FEBRUARY 2002 • WWW.PRN.ORG would amount to roughly 7% of the total cumulative probability over the duration of KEY POINTS that individual’s infection. “In other words,” Dr. Pilcher elucidated, “the 20- • Biological and epidemiological evidence converge on the hypothesis that partners of fold increase in transmission probability individuals with phi are likely to be at very high risk for infection. during phi may not dominate an individual • Proof of ready transmission by individuals with early acute infection, and of extensive case patient’s overall cumulative transmission clustering in phi cohorts, justify efforts to perform rapid contact tracing. probability.” However, he emphasized, transmission to a sexual partner is still • Future directions include the improvement in strategies to expand screening, taking into much more likely to occur sometime over account both feasibility and cost effectiveness, and the study of short-course antiretroviral the ten-year span of chronic hiv infection therapy on genital shedding of hiv and its impact on serial transmissions. if the infected individual remains undiag- nosed and untreated. Dr. Pilcher cautioned clinicians against References Quinn TC, Wawer MJ, Sewankambo N, et al. Viral load and heterosexual transmission interpreting these data as downplaying Chakraborty H, Sen PK, Helms RW, et al. Viral of human immunodeficiency virus type 1. the importance of diagnosing phi. “Actu- burden in genital secretions determines Rakai Project Study Group. N Engl J Med ally,” Dr. Pilcher postulated, “those inter- male-to-female sexual transmission of HIV- 342(13): 921-9, 2000. ested in phi as a potential public health 1: a probabilistic empiric model. AIDS 15(5):621-7, 2001. opportunity should find these data ener- Schacker T, Collier AC, Hughes J, et al. Clini- gizing. What this means is that much of an cal and epidemiologic features of primary Colfax G, Mansergh G, Vittinghoff E, et al. HIV infection. Ann Intern Med 125(4):257-64, individual’s cumulative transmission risk Drug use and high-risk sexual behavior 1996. rests during the period of chronic infection, among circuit party participants [Abstract and can be prevented by early diagnosis. TuPeC3422]. XIII International Conference on Sey K and Harawa N. High-risk behavior aids, Durban, 2000. That is, even if the window of hyperinfec- among individuals diagnosed with acute/pri- tiousness associated with phi has come mary or recent HIV infection [Abstract 216]. Gray RH, Wawer MJ, Brookmeyer R, et al. and gone by the time patients are diag- 8th Conference on Retroviruses and Oppor- Probability of HIV-1 transmission per coital tunistic Infections, Chicago, 2001. nosed, we can still make a profound dif- act in monogamous, heterosexual, HIV-1- ference by identifying these individuals discordant couples in Rakai, Uganda. Lancet Vernazza PL, Kahlert C, Fierz W. Sensitization and by being aggressive about counseling 357(9263):1149-53, 2001. to human immunodeficiency virus in seroneg- and contact tracing. It’s never too late to in- ative exposed partners. J Infect Dis terrupt the spread of the epidemic.” Kelker HC, Seidlin M, Vogler M, et al. Lym- 182(6):1810-2, 2000. phocytes from some long-term seronegative heterosexual partners of HIV-infected in- Yerly S, Vora S, Rizzardi P, et al. Acute HIV individuals proliferate in response to HIV anti- fection: impact on the spread of HIV and gens. AIDS Res Hum Retroviruses 8(8):1355-9, transmission of drug resistance. AIDS 15:2287- 1992. 92, 2001.

Langlade-Demoyen P, Ngo-Giang-Huong N, et al. Human immunodeﬁciency virus (HIV) nef- speciﬁc cytotoxic T lymphocytes in nonin- fected heterosexual contact of HIV-infected patients. J Clin Invest 93(3):1293-7, 1994.

Leynaert B, Downs AM, de Vincenzi I. Hetero- sexual transmission of human immunodeﬁ- ciency virus: variability of infectivity throughout the course of infection. European Study Group on Heterosexual Transmission of HIV. Am J Epidemiol 148(1):88-96, 1998.

Mazzoli S, Lopalco L, Salvi A, et al. Human im- munodeﬁciency virus (HIV)-speciﬁc IgA and HIV neutralizing activity in the serum of exposed seronegative partners of HIV seropos- itive persons. J Infect Dis 180(3):871-5, 1999.

Pilcher CD, Eron JJ Jr., Vernazza PL, et al. Sexual transmission during the incubation period of primary HIV infection. JAMA 286(14): 1713-4, 2001

Pilcher CD, Shugars DC, Fiscus SA, et al. HIV in body fluids during primary HIV infection: implications for pathogenesis, treatment and public health. AIDS 15(7):837-45, 2001a.

THE PRN NOTEBOOK™ • SPECIAL EDITION • FEBRUARY 2002 • WWW.PRN.ORG 15