Methods

Top Methods Results Discussion Author & Article Info References

Study Design

Patients were recruited into an institutional review board-approved protocol for the study of primary HIV infection. Eligible participants were documented to be 1) simultaneously positive for HIV p24 antigen (or plasma HIV RNA) and negative for HIV antibody (measured by an enzyme immunoassay [EIA] for HIV) at recruitment; 2) serologically (that is, measured by HIV EIA plus Western blot assay) positive at recruitment; or 3) serologically positive at recruitment and serologically negative within the 12 months before study entry, with a documented retroviral syndrome developing 3 months before study entry. We placed no restrictions on sex, age, or race. All enrolled patients who had HIV seroconversion provided written documentation of a previously negative HIV EIA test result from a laboratory proficient in HIV antibody testing. Between September 1993 and January 1995, 60 persons were referred to the study clinic to be screened for primary HIV infection; 46 met the inclusion criteria. The 14 excluded persons did not provide written documentation of previous negative HIV antibody test results or of recent seroconversion. At the enrollment visit, a standardized 160-item questionnaire was administered in private to each patient to collect demographic information, recent and earlier sexual history, and details of all recent medical illnesses. History of sexual contact was corroborated on subsequent visits. Recent medical records were also reviewed, and each patient had a complete physical examination. Any stored serum or plasma samples available from previous HIV antibody testing or a recent illness were retrieved and tested for p24 antigen, plasma HIV RNA, and antibodies against HIV to determine whether the illness associated with the serum sample had been caused by HIV seroconversion. No patient received antiretroviral therapy before or during the initial evaluation period. All patients were asked to have lumbar puncture to obtain cerebrospinal fluid specimens. Twenty-four of the 46 patients agreed to have lumbar puncture.

Laboratory Analysis

All laboratory studies were done at the University of Washington. Laboratories at the university are certified by the American College of Pathology and the AIDS (acquired immunodeficiency syndrome) Clinical Trials Group for the HIV antibody test, the p24 antigen test, and HIV RNA polymerase chain reaction assays. Enzyme immunoassays for HIV were done according to the manufacturer's instructions (Genetic Systems Corp., Seattle, Washington). All positive EIA results were confirmed by Western blot assay (Epitope, Inc., Beaverton, Oregon). Levels of HIV p24 antigen were assayed by EIA according to the manufacturer's directions (Abbott Laboratories, Chicago, Illinois); p24 antigen levels less than 20 pg/mL were considered to represent a negative result. Plasma HIV RNA was detected using the Chiron (Emeryville, California) branched-chain DNA assay. The lower limit of detection for this assay is 10 x 10³ DNA copies/mL of plasma [2, 3]. Quantitative co-cultures of peripheral blood mononuclear cells and standard HIV type 1 (HIV-1) cultures of cerebrospinal fluid and plasma were done at the University of Washington Retrovirology Laboratory using previously published methods [4, 5]. T-cell subsets were measured by the Hematopathology Laboratory of the University of Washington using flow cytometry.

Statistical Analysis

Differences in demographic characteristics and sexual contact history were evaluated using the Fisher exact test or the Mann-Whitney test, as specified. For patients with symptomatic illness, we used the date of symptom onset as the date of HIV acquisition. For patients who were HIV negative on EIA within the 6 months before study entry and who had no history of symptoms, the estimated acquisition date was established as the date midway between the most recent negative and the subsequent positive HIV antibody test result. One author assigned the acquisition date for all patients.

Results

Top Methods Results Discussion Author & Article Info References

Patients

We enrolled 46 patients during a 28-month period. Patients were referred because they were at risk for HIV infection and developed symptoms of a retroviral illness or because they were found to be HIV-positive on routine testing and had documented evidence of a recent negative test. Patients were recruited from the University of Washington HIV clinics (43%), regional AIDS prevention programs (15%), King County Jail (12%), and local community physicians (30%). Forty-three of the 46 patients (93%) were men. The median age was 30 years, and 13 of 46 patients (28%) were younger than 25 years of age. Forty-two of the 43 men had had sex with men as a risk factor for HIV acquisition; the other male patient reported only heterosexual contact. Four patients (3 men and 1 woman) reported having used injection drugs. Our cohort was similar to persons reported to have HIV-1 infection in King County, Washington (Table 1).

In 17 of 46 patients (37%), either HIV-1 p24 antigen or HIV RNA was detected when the HIV EIA result was negative. The remaining 29 patients (63%) were identified with a new positive result on an HIV Western blot assay and a previous negative result on HIV EIA paired serologic assays. These 29 patients included persons who were tested because they were at high risk and thus had routinely obtained HIV serologic testing, persons who were tested for recent symptoms consistent with primary HIV infection, and persons who were entering a new personal relationship and wanted to know their HIV serologic status. In these 29 patients, the median time from the last documented negative test result was 5 months (range, 1 to 12 months), and the median time from symptom onset to study entry was 80 days (range 2 to 178 days). No differences were noted between the demographic characteristics of the 17 patients identified as positive for HIV p24 antigen or for HIV RNA at enrollment and the demographic characteristics of the 29 patients identified through paired serologic testing.

Clinical Presentation

Forty-one patients (89%) reported symptoms associated with HIV seroconversion. Sixteen of 17 patients who were positive for HIV p24 antigen or HIV RNA or were negative for HIV EIA antibody at study entry had signs or symptoms suggestive of an acute retroviral syndrome; 15 of these 16 patients (94%) sought medical assistance, and 12 of these 15 were eventually tested for HIV infection during their illness. Of the remaining 29 patients, 25 (86%) also reported symptoms consistent with an acute retroviral syndrome during the interval between their previously negative antibody test result and study entry (Table 2). Twenty-two of these 25 (88%) sought medical care for these symptoms: Nine had HIV testing at the initial medical encounter, and 5 others referred themselves for testing shortly after the illness developed. Three of 46 patients (6%) reported that they had intentionally engaged in high-risk sexual behavior in attempts to become infected with HIV.

Twenty-three patients were participating in routine surveillance programs in which HIV testing was done every 6 months. Twenty of these 23 (87%) reported symptoms consistent with an acute retroviral syndrome, and 19 of the 20 (95%) had medical evaluation for these symptoms (48% went to their own primary care provider, 31% went to an emergency department, and 21% went to a walk-in clinic). A diagnosis of primary HIV infection was considered at these medical encounters in only 5 of these 19 patients (26%).

The five most common symptoms reported during primary HIV infection were fever, sore throat, fatigue, weight loss, and myalgia (Figure 1, top). The median maximal temperature reported was 38.9 °C (range, 39 °C to 40.5 °C), and the median weight loss was 5 kg (range, 1.4 to 10 kg). Abnormalities on physical examination for symptoms of primary HIV infection were found in 23 of 35 patients (66%) seeking medical help. These abnormalities included postural hypotension, oral ulceration, exudative pharyngitis, thrush, genital or rectal ulceration, adenopathy, and signs of neuropathy. Signs and symptoms of aseptic meningitis (fever, headache, photophobia, and stiff neck) were noted in 10 of the 41 patients (24%) presenting for medical consultation. The median duration of symptoms of the acute retroviral illness was 14 days (Figure 1, bottom). Seven of the 41 patients (17%) with symptomatic illness were hospitalized. Frequency of hospitalization was the only significant difference noted in the clinical presentation of patients who were positive for p24 antigen (or for HIV RNA) and had negative EIA results compared with patients who were identified by serial testing: 29% compared with 8%.

View larger version (27K): [in this window] [in a new window]   Figure 1. Top. Signs and symptoms in the 41 patients who presented with symptomatic primary human immunodeficiency virus (HIV) infection. Bottom. Duration of signs and symptoms of primary HIV infection in 41 patients.

Virologic Studies at Study Entry

The time from seroconversion to study entry was a significant factor in the ability to detect HIV-1 RNA in plasma (Table 3). Of 16 patients enrolled within 60 days of HIV acquisition (median, 28 days), 14 had HIV RNA detected in plasma (median titer, 116 x 10³ RNA copies/mL [range, <10 to 553 x 10³ RNA copies/mL]). Only 1 of these 14 patients had HIV p24 antigenemia. In the other 30 patients, plasma HIV RNA levels were measured a median of 101 days after seroconversion; 19 of these 30 patients (63%) had detectable HIV RNA (P < 0.01). The median HIV RNA level was 26 x 10³ RNA copies/mL (range, <10 to 1074 x 10³ RNA copies/mL). Plasma HIV-1 RNA was detected significantly more often than was p24 antigen (Table 3).

In 39 patients, peripheral blood mononuclear cell cultures were done at study entry. Human immunodeficiency virus was isolated from peripheral blood mononuclear cell cultures in all 11 patients in whom assays were done within 60 days of virus acquisition and in 24 of the 28 persons (86%) in whom assays were done after day 60. All HIV isolates at study entry had the non-syncytial-inducing phenotype.

Twenty-four patients had lumbar puncture for examination of cerebrospinal fluid and HIV culture. Human immunodeficiency virus was isolated by culture in 12 of these 24 patients (50%). The cerebrospinal fluid cell count was abnormal (leukocyte count >5 cells/mm³) in 15 of these 24 patients (HIV was isolated from 9, and 6 did not have HIV). Median plasma RNA levels measured at lumbar puncture were significantly higher in patients with a cerebrospinal fluid culture positive for HIV (Table 4).

CD4 Lymphocyte Counts at Study Entry

At study entry, CD4 counts were greater than 600 cells/mm³ in 19 of 46 patients (41%); 400 to 599 cells/mm³ in 17 of 46 patients (37%); 300 to 399 cells/mm³ in 7 of 46 patients (15%); and less than 300 cells/mm³ in 3 of 46 patients (7%). In patients who entered the study within 4 weeks of acquiring HIV, the median CD4 cell count was 655 cells/mm³ (range, 244 to 1055 cells/mm³); in those who entered 5 to 8 weeks after acquisition, the CD4 count was 568 cells/mm³ (range, 311 to 866 cells/mm³); in those who entered 9 to 16 weeks after acquisition, the CD4 count was 555 cells/mm³ (range, 210 to 878 cells/mm³); and in those who entered 17 to 24 weeks after acquisition, the CD4 count was 437 cells/mm³ (range, 217 to 1256 cells/mm³).

Sexual Activity before Acquisition of HIV

All 46 patients reported having sexual activity in the 6 months before seroconversion. The 43 male patients had had a median of three sexual partners in the 6 months before HIV acquisition and a median of one sexual partner in the month before acquisition. Twenty-one of the 46 patients (46%) reported having had only one sexual partner in the month before seroconversion (or, for the 5 patients with asymptomatic seroconversion, in the month before the first positive HIV test result). Only 5 of 41 symptomatic patients (12%) reported having had more than five sexual partners in the month before HIV acquisition.

Table 5 shows the reported frequency of sexual activity for the 43 male patients in the 6 months before HIV acquisition. These patients reported having a median of 51 sexual contacts in the 6 months before acquisition (2.1/wk). Unprotected oral-genital contact (median, 20 contacts) was the most frequently reported sexual activity. Only 3.6% of oral-genital contacts and 42% of genital-rectal contacts were protected (that is, the insertive partner wore a condom). The 3 female patients reported having both protected and unprotected receptive genital contact.

Oral-Genital Contact as a Risk Factor for HIV Infection

Of the 21 persons (18 men and 3 women) who reported having had only one sexual partner in the month before diagnosis, the date of contact and the specific details of the sexual encounter that probably led to HIV seroconversion were available in 12. All 12 persons described having unprotected sexual activity at this encounter (Table 6). For 10 of the 12 patients, the unprotected activity occurred during a single sexual encounter with a source partner known to be HIV positive; for the other 2 patients, the unprotected activity occurred during one sexual encounter with a source partner of unknown HIV status in the month before clinical illness developed. The median time from the sexual encounter to the development of symptoms was 15 days (range, 5 to 29 days) (Figure 2). Four patients (33%) reported having only unprotected oral-genital contact during the sexual encounter that led to HIV infection (Table 6). We also interviewed the source partners of 2 of the 4 patients about their sexual history to corroborate the data on sexual activity at the acquisition encounter. In both cases, the source partner and the patient provided similar sexual histories. The following is a representative case history.

View larger version (13K): [in this window] [in a new window]   Figure 2. Days from sexual exposure to onset of symptoms in 12 patients who could identify the exact date and time of the sexual exposure that led to acquisition of human immunodeficiency virus.

Patient A, a 32-year-old man, reported having had only one sexual partner in the 8 months before seroconversion. He denied having had anal intercourse (either insertive or receptive) for at least 10 years before HIV seroconversion. On 16 February 1994, he performed unprotected oral intercourse on his partner (the latter was the receptive partner). He had had no sexual contact with anyone else for the previous 120 days, and this was his only reported sexual contact in the previous 30 days. His partner did not perform oral intercourse on him. Five days later, the patient presented to the emergency department with fever, diarrhea, and pharyngitis. Herpangina was diagnosed. The patient was sent home and returned 2 days later with orthostatic hypotension, temperatures as high as 104 °C, and a macular erythematous nonpruritic rash that covered his trunk and extremities. Acute HIV infection was suspected; this diagnosis was confirmed with a positive p24 antigen result (>1000 pg/mL) on the same day that an HIV EIA result was negative. The diarrhea and fever persisted for 14 days, and myalgia and fatigue lasted for several weeks. Patient A subsequently learned that his partner had HIV infection. An independent interview with his partner confirmed that patient A had had only unprotected oral-genital contact.

Discussion

Top Methods Results Discussion Author & Article Info References

We have made several important observations about the events surrounding symptomatic HIV seroconversion in this group of homosexually active men. Forty-one of the 46 patients that we identified had definable illnesses that prompted medical attention, and hospitalization was needed to manage symptoms in 15% of patients. Most patients came to the attention of a health care provider during the acute illness, but only 25% of persons received a correct diagnosis at the initial medical encounter. Nearly half of the men in our cohort were monogamous at the time of HIV acquisition, and unprotected oral-genital contact was the most commonly reported type of sexual activity in the men who had HIV seroconversion. Oral-genital transmission appeared to have been the route of acquisition in 4 patients.

Strengths of our study include the stringent criteria that we used to confirm the diagnosis of primary HIV infection and the detailed standardized interview that we conducted near the time of acquisition. These methods allowed us to obtain a consistent sexual history and information about sexual risk factors in many persons with virologically and clinically confirmed acute HIV infection. The numbers of sexual partners that our patients reportedly had were substantially lower than the numbers reported a decade ago [6-9]. Only 12% of our patients had had more than five partners in the 6 months before seroconversion, and nearly half had had only one partner. The finding that HIV seroconversion occurred in this group of homosexual men despite their contact with relatively few partners suggests the following: that many persons in our study who seroconverted currently engage in unsafe sex with partners they know to be HIV infected; that current sexual practices thought to be safe are not; or that source partners are withholding (or do not have) information on their own HIV serologic status. Our interviews indicated that all of these factors appear to have been present.

Unprotected oral-genital contact was the most commonly reported sexual activity in patients who developed primary HIV infection. In the 6 months before acquisition of HIV, patients had a median of 20 oral-genital contacts and a median of 2 anal-genital contacts. Forty-two percent of anal-genital contacts and only 3.6% of oral-genital contacts were "protected" (associated with the use of a condom). These data are similar to those showing that, in persons at risk for acquiring HIV, oral-genital contact is more common than anal-genital contact [10, 11]. We recognize that the data on sexual history were self-reported and obtained during face-to-face interviews. However, all patients were interviewed several times, and their responses to questions about the timing and frequency of sexual behaviors did not vary. We also interviewed the source partners of two of the four persons who reported only oral-genital contact as a risk factor for HIV acquisition to confirm the history reported by these patients. If collection of sexual histories had been more anonymous, we might have obtained different data on the frequency of unprotected high-risk sexual behaviors in our cohort, but the case reports and consistency of our observations suggest that the reported information is valid. Although unprotected oral-genital contact is widely believed to be a safer form of sex [6, 7, 12, 13], case reports of HIV transmission through this route have been published [14-18]. Because we did not have a control group, we could not calculate the relative risk of oral-genital contact for transmission of HIV. Our data suggest that HIV transmission through oral-genital contact is not very efficient. However, because unprotected oral-genital contact appears to be a common sexual activity among homosexual men, increased attention to risk factors for oral-genital transmission of HIV appears warranted.

Three of our 46 patients reported that they had intentionally acquired HIV. These 3 persons reported symptoms of depression and a feeling of exclusion from their peer group; this resulted in unprotected contact with persons known to be HIV infected. Although this behavior is commonly reported in the lay press, ours is the first documentation (to our knowledge) of such activity to be reported in the medical literature.

Primary HIV infection was not often diagnosed in our patients, even in high-risk persons who were enrolled in programs of routine surveillance for HIV. An acute retroviral syndrome was correctly diagnosed in only 5 of the 19 patients who had routine serologic testing for HIV every 4 to 6 months. Several factors may explain the discrepancy between ongoing serologic surveillance and poor recognition of symptoms associated with HIV seroconversion. First, the nondescript symptoms of acute HIV infection make the infection difficult to distinguish from other illnesses. Second, many patients go to a clinic that offers anonymous HIV testing for routine screening and to a different clinic for routine medical care. Third, anonymous HIV testing allows patients to have current knowledge of their HIV serologic status while keeping the history and results of repeated testing out of the formal medical record. This strategy has many advantages [19-22] and is important to the overall epidemiologic and public health approach to HIV. However, it may also leave a caregiver ignorant of the patients' risk factors for HIV infection, thus reducing the likelihood that he or she will recognize symptoms of or correctly diagnose primary HIV infection. Our data show that health care providers need to maintain a high index of suspicion for symptoms suggestive of primary HIV infection and to obtain an adequate history of sexual contact from all persons who present to emergency departments or walk-in clinics with a febrile illness, mononucleosis, or syndromes similar to aseptic meningitis.

Diagnosing recently developed primary HIV infection may be important for several reasons. The period immediately after HIV seroconversion is associated with high titers of circulating virus [23-25], making patients more likely to transmit the virus during this period. Transmission of HIV to another person during the period between HIV acquisition and diagnosis of HIV infection was documented in several persons enrolled in our cohort. In addition, as new therapies for HIV infection are developed, prompt initiation of therapy during the seroconversion syndrome may become important [26].

Detection of HIV RNA in plasma and culture was significantly more sensitive than detection of p24 antigenemia in all study patients; this finding was similar to those previously reported [27]. Thus, we would recommend HIV RNA testing for all high-risk patients who present with symptoms compatible with primary HIV infection. We found that measuring the CD4 cell count was not a useful way to test for HIV seroconversion.

Several caveats about our study should be noted. Our cohort was referral based and was therefore biased toward the enrollment of symptomatic patients. This limitation, in addition to the fact that we primarily enrolled men who had acquired HIV through sexual contact, makes it difficult to generalize our results to all persons with HIV infection. Several studies [28-31] indicate that many HIV acquisitions result in subclinical infection. Thus, our observations appear to be most relevant for persons who develop symptomatic HIV infection. It is difficult to accurately estimate how many persons will have a symptomatic seroconversion to HIV and how many will have subclinical infection, but previous reports have suggested that 50% to 90% of all seroconversions are associated with symptoms [28, 29, 32]. Moreover, whether the acquisition events leading to subclinical HIV infection differ from those leading to symptomatic HIV infection remains to be determined. At present, we can provide little insight about acquisition events in women or in persons who acquire HIV through the use of injection drugs.

In summary, although infection with sexually acquired HIV in homosexual men is often symptomatic, it is often initially misdiagnosed. Obtaining a sexual history for high-risk behavior is a necessary part of the medical evaluation of sexually active persons who present with fever, sore throat, lymph-adenopathy, or symptoms of viral meningitis. Men today who have sex with men and then acquire HIV have fewer sexual partners than did homosexual men a decade ago. However, HIV acquisition continues to occur. Increasing attention to the risks of oral-genital contact as an important means of HIV acquisition appears to be warranted. Early diagnosis of HIV infection is necessary to reduce subsequent spread of HIV and may be important for early therapeutic intervention strategies.

Dr. Collier: AIDS Clinical Trials Unit, 1001 Broadway, Suite 218, Seattle, WA 98122.

Ms. Shea: Primary Infection Clinic, 1001 Broadway, Suite 206, Seattle, WA 98122.