May 28, 1999
For this investigation, female sex partners of the putative index case-patient were considered primary contacts, male sex partners of HIV-infected primary contacts were considered secondary contacts, and female sex partners of the HIV-infected male secondary contacts were considered tertiary contacts. Medical records of contacts were reviewed for demographic information, history of HIV counseling and testing, sexually transmitted diseases (STDs) (i.e., syphilis, gonorrhea, chlamydia, herpes, and trichomonas), and drug and alcohol use. Blood specimens from consenting persons were forwarded to CDC for HIV DNA sequence analysis and for blinded serologic testing of specimens for syphilis, Chlamydia trachomatis, and herpes simplex virus type 2 (HSV-2). No blood specimen was available from the putative index case-patient.
Forty-seven primary contacts were identified and reportedly had had vaginal sex with the index patient: 13 (31%) of 42 tested had HIV infection. From these 13 primary contacts, 84 secondary contacts were identified; one of 50 tested had HIV infection. Sixty secondary contacts had sexual exposure to the primary contacts during the same period or after the primary contacts had sexual exposure to the putative index case-patient; one of 39 tested had HIV infection. Three tertiary contacts of the one positive secondary contact were identified; the one tested was HIV negative. One of three infants born to HIV-infected women was positive by polymerase chain reaction (PCR) testing for HIV DNA. There was no evidence that the putative index case-patient or the HIV-infected primary contacts had had same-sex or needle-sharing contacts.
Blood samples for HIV DNA sequence analysis were obtained from 10 of the 13 HIV-infected primary contacts, the one HIV-infected secondary contact, and two HIV-infected persons from the community who were not epidemiologically related to the cluster (community-comparison persons). A nested PCR procedure was used to amplify proviral HIV DNA sequences from peripheral blood mononuclear cells (PBMCs) from these 13 persons. A 345 nucleotide segment of the C2V3C3 region of the env gene and approximately 400 nucleotides of the p17 coding region of gag were sequenced and analyzed in a blinded fashion. Phylogenetic analysis of the 13 sequences was performed with reference sequences from HIV subtypes A-D, F, and G from the GenBank&/content/art30072.html/content/art30072.html#134; database for both the env and gag gene regions. Bootstrapping, a technique used to assess the relatedness of the viruses, demonstrated that all 13 sequences were from subtype B viruses (1). Sequences from the 10 HIV-infected primary contacts -- but not from the infected secondary contact, the two community-comparison persons, or subtype B reference strains -- clustered strongly together in both gene regions. The phylogenetic analyses indicated a high degree of relatedness among the viruses infecting the 10 tested primary contacts and suggest that the infected secondary contact was probably infected by a source not related to this cluster.
The 13 HIV-infected primary contacts reportedly had their last sexual exposure to the putative index case-patient during February 1996-January 1997 (Figure 1); 25 of the 29 primary contacts who were not HIV infected had last contact with him during January 1995-August 1997; data were missing for four. The median number of vaginal sexual exposures to the putative index case-patient was higher, although not significantly, for the HIV-infected women (six exposures; range: two-190 exposures) than for the uninfected women (three exposures; range: one-90 exposures) (data were missing for six) (Wilcoxon rank sum test, p=0.07). Median ages at first exposure to the putative index case-patient were similar for HIV-infected women (17.8 years; range: 13-22 years) (data were missing for one) and uninfected women (17.7 years; range: 14-24 years) (data were missing for 14). Among exposed women, HIV infection was not associated significantly with a history of STDs (10 of 22), cocaine use (three of 22), alcohol use (two of 16), or serologic markers for STDs (15 of 25). When analyses were limited to seven HIV-infected and eight uninfected women with exposures only after September 1996 (Figure 1), HIV-infected women had significantly more exposures to the putative index case-patient (median: three exposures; range: two-six exposures) than the uninfected women (median: one exposure; range: one-two exposures (data were missing for two) (Wilcoxon rank sum test, p=0.005).
Reported by: FB Coles, DO, GS Birkhead, MD, P Johnson, PF Smith, MD, State Epidemiologist, New York State Dept of Health; R Berke, MD, P Allenson, M Clark, Chautauqua County Dept of Health, Mayville, New York. Div of HIV/AIDS Prevention-Surveillance and Epidemiology, and Intervention Research Svcs, National Center for HIV, STD, and TB Prevention; Div of AIDS, STD, and TB Laboratory Research, National Center for Infectious Diseases, CDC.
The findings in this report suggest a common source of HIV infection for at least 10 of 13 HIV-infected women who independently reported contact with the same partner. The high rate of HIV infection among sexual contacts of the putative index case-patient over a period of many months raises the possibility that efficient transmitters of HIV exist and may contribute disproportionately to HIV transmission.
Reasons for the apparently high attack rate (31%) among primary sex contacts in this cluster are unclear. Persons with primary HIV infection (i.e., within several weeks after infection [2,3]) or those in the late stage of HIV infection (4) may be especially infectious because these periods are usually associated with high HIV viral loads (viremia). If the putative index case-patient was the common sex partner of these women, he probably was infected by or during February 1996 because the earliest date of last exposure for an HIV-infected primary contact was during February 1996. However, seven of 15 women whose first sexual exposure to the putative index case-patient was after September 1996 were HIV infected. These contacts probably would have been infected after the presumed period of primary HIV infection but before the late stage of HIV infection in the putative index case-patient (4). Thus, at least some HIV-infected persons, such as the putative index case-patient, may be highly infectious at times other than the primary or late stage of HIV infection.
Other characteristics may be critical in determining the likelihood of HIV transmission. Host susceptibility or infectiousness may increase as a result of inflammation or ulceration associated with STDs (5). For the susceptible partner, genital ulcerative infections (e.g., syphilis and HSV-2) are cofactors that facilitate transmission (5), but STDs were not significantly associated with being HIV-infected among the primary contacts in this cluster.
This cluster occurred despite other prevention successes in the county among youth (6). Discovery and evaluation of this cluster were possible, in part, because of the low background prevalence of HIV infection in the county (6) (i.e., relatively few new cases of HIV infection could be detected and followed by public health personnel) and a coordinated response by health officials enabled prompt epidemiologic and laboratory investigations.
This cluster of infection has implications for HIV intervention and prevention. Unrecognized social and sexual networks of youth at high risk for HIV and other STDs exist even in rural areas where HIV prevalence is relatively low, and these networks can facilitate the rapid spread of HIV infection. It is important for public health programs to provide effective HIV prevention services to youth in rural areas.
* Single copies of this document will be available until May 27, 2000, from the National Prevention Information Network (NPIN) (operators of the National AIDS Clearinghouse), P.O. Box 6003, Rockville, MD 20850; telephone (800) 458-5231 or (301) 519-0459.
Use of trade names and commercial sources is for identification only and does not imply endorsement by CDC or the U.S. Department of Health and Human Services.