July 29, 2011
Human immunodeficiency virus (HIV) is categorized into two types, HIV-1 and HIV-2. Worldwide, most HIV infections are HIV-1, whereas HIV-2 largely has been confined to persons in or from West Africa.1,2 HIV-1 and HIV-2 have the same routes of transmission, and both can cause acquired immunodeficiency syndrome (AIDS)3; however, HIV-2 infections should be differentiated from HIV-1 infections because they are less likely to cause AIDS and their clinical management differs.4,5 CDC's current surveillance case definition for HIV infection applies to both variants of HIV6 but lacks criteria for differentiating between HIV-1 and HIV-2. To enumerate and describe HIV-2 cases reported in the United States, a working case definition was developed. During 1988-June 2010, a total of 242 HIV-2 cases were reported to CDC. Of these, 166 met the working definition. These HIV-2 cases were concentrated in the Northeast (66%, including 46% in New York City) and occurred primarily among persons born in West Africa (81%). Ninety-seven of the HIV-2 cases also had a positive HIV-1 immunoblot antibody test result (e.g., Western blot). Immunoblot antibody tests currently used to confirm HIV reactive screening tests do not contain reagents specific to HIV-2 and thus are not reliable for identification of HIV-2 infections.7 Additional testing specific to HIV-2 should be considered if HIV-1 test results are atypical or inconsistent with clinical findings, especially for persons from West Africa. If an HIV case is reported to the health department but subsequently identified as HIV-2, health-care providers should update the case report to reflect the correct type.
During 2009-2010, CDC convened a workgroup to develop the working HIV-2 definition used in this report. To meet this working definition, cases had to satisfy one or more of the following three criteria: 1) HIV-1/HIV-2 type-differentiating antibody immunoassay (e.g., Bio-Rad Multispot HIV1/HIV-2 Rapid Test) positive for HIV-2 but negative for HIV-1, 2) positive HIV-2 nucleic acid test (DNA or RNA), 3) positive HIV-2 immunoblot and negative or indeterminate HIV-1 immunoblot. In addition, one case reported in 1991 was accepted based only on a positive radioimmunoprecipitation assay (a now obsolete test). Neither the nucleic acid tests nor the immunoblots have been approved by the Food and Drug Administration for diagnosis of HIV-2 infection, but the Bio-Rad Multispot HIV-1/HIV-2 Rapid Test has been approved for differentiation of HIV-2 from HIV-1.
During 1988-June 2010, health departments of the 50 states and the District of Columbia reported to CDC a total of 242 HIV-2 cases, based on a variety of criteria with no formal HIV-2 infection case definition. From that total, 47 reported cases were excluded because they had insufficient identifying information to discern whether or not they were duplicate reports. An additional 29 cases were excluded because they did not meet any of the three working definition criteria.
The remaining 166 cases met one or more of the criteria of the working definition and were analyzed by diagnostic test results, reason for suspecting HIV-2 infection, region of report, country of birth, race/ethnicity, sex, age, and transmission risk factor. Poisson regression was used to assess trends in the annual number of diagnoses. The year of diagnosis was defined as the year of the first positive HIV test, which in some cases was years before recognition that the HIV type was HIV-2.
Of the 166 HIV-2 cases, 113 (68%) met the first criterion of a result positive for HIV-2 but negative for HIV-1 on a type-differentiating antibody immunoassay, 66 (40%) met the second criterion of a positive HIV-2 nucleic acid test, and 58 (35%) met the third criterion of a positive HIV-2 immunoblot and negative or indeterminate HIV-1 immunoblot. Seventy-one (43%) of the 166 cases met more than one of the three criteria. HIV-1 immunoblot results were available for 163 of the HIV-2 cases; 97 (60%) were positive for HIV-1, 63 (39%) were indeterminate, and three (2%) were negative. Of the 97 HIV-2 cases with positive HIV-1 immunoblot results, the diagnosis of HIV-2 was established by a type-differentiating antibody immunoassay in 46 cases, by an HIV-2 nucleic acid test in 22 cases, and by both in 29 cases (Table 1).
Nucleic acid test evidence of HIV-1 coinfection was present in 19 (11%) of the 166 HIV-2 cases, including two with positive qualitative HIV-1 nucleic acid tests and 17 with detectable HIV-1 viral loads, of which the highest level was 1,000-1,999 copies/mL in six cases, 2,000-4,999 copies/mL in seven cases, and >10,000 copies/mL in four cases. HIV-1 nucleic acid test results were negative in 66 (40%) HIV-2 cases and were missing in 81 (49%) HIV-2 cases.
Data were available for 116 HIV-2 cases regarding why HIV-2 might have been suspected initially. Of these, 65 (56%) patients had an indeterminate HIV-1 immunoblot despite a positive HIV-1 or HIV-1/HIV-2 antibody screening test, 47 (41%) had an undetectable HIV-1 viral load despite a positive HIV-1 immunoblot, and four patients (three born in West Africa and one who had visited West Africa) had a negative HIV-1 immunoblot despite a positive HIV-1 or HIV-1/HIV-2 antibody screening test.
Of the 164 HIV-2 patients for whom birthplace was known, 132 (81%) were born in West Africa. Seven (4%) were born in other parts of Africa, and six (4%) in unspecified parts of Africa. Nine (6%) were born in India, five in the United States, three in Europe, and two in Mexico (Table 2). Of the 166 cases of HIV-2 infection, 77 (46%) were reported from New York City, 33 (20%) from elsewhere in the Northeast, 24 (15%) from the South, 18 (11%) from the Midwest, and 14 (8%) from the West. Among the patients, 89% were non-Hispanic blacks, 58% were men, and the median age at diagnosis of HIV infection was 39 years (range: 21-76 years).
No transmission risk factor was identified in 120 (72%) of the 166 cases, including 78 cases among persons with a history of heterosexual contact but whose sex partners had unknown infection status or were known to be uninfected. The reported risk factors for the remaining 88 cases were heterosexual contact with a sex partner known to be HIV-infected (38 patients, 23%), male-to-male sexual contact (four, 2%), and injection-drug use (four, 2%). Of the 50 women aged 15-44 years at diagnosis, 24 (48%) were pregnant at or after HIV-2 diagnosis. No children born to these 24 women were reported to be HIV-infected, but follow-up information was missing for six of the children born after their mother's diagnosis.
Poisson regression indicated that the annual number of HIV-2 diagnoses in the United States increased significantly from 1987 to 2009; however, the increase might be the result of surveillance artifact. No significant trends in HIV-2 diagnoses were observed during 1990-1999 (mean: 4.3 diagnoses per year; range: 1-8), or during 2000-2009 (mean: 12.0 per year; range: 8-19). The annual number increased abruptly from two in 1999 to 12 in 2000, the year that New York began confidential name-based reporting of HIV infection cases in addition to AIDS reporting. Similar results were obtained when the trend analyses included all suspected cases that were excluded from other analyses (Figure).
The 166 HIV-2 cases constituted only 0.01% of the more than 1.4 million U.S. cases of HIV infection diagnosed during 1987-2009 (unadjusted for reporting delay). Of the 5,284 HIV infections reported in the United States among persons born in West Africa, 132 (3%) were HIV-2 infections. Among the HIV-infected persons born in West Africa, the percentage identified as HIV-2 cases varied significantly (p<0.05, chi square test) by reporting region: 5% in New York City, 2% in the rest of the Northeast, 2% in the Midwest, 2% in the West, and 1% in the South.
Reported by: Lucia V. Torian, Ph.D., Bur of HIV/AIDS Prevention and Control, New York City Dept of Health and Mental Hygiene. Richard M. Selik, M.D., Bernard Branson, M.D., S. Michele Owen, Ph.D., Timothy Granade, M.S., R. Luke Shouse, M.D., M. Patricia Joyce, M.D., Danuta Pieniazek, Ph.D., Richard Kline, M.S., Div of HIV/AIDS Prevention, National Center for HIV/AIDS, Viral Hepatitis, STD, and TB Prevention, CDC. Corresponding contributor: Richard M. Selik, email@example.com, 404-639-4495.
The results of the analyses described in this report indicate that HIV-2 infections in the United States are rare, concentrated in the Northeast, and limited mainly to persons born in West Africa. Regional differences in the percentage of reported HIV cases caused by HIV-2 might be, in part, a result of the nonuniform geographic distribution of U.S. residents born in West Africa. New York is a major gateway for African immigrants to the United States.8
However, regional differences in the percentage of HIV cases caused by HIV-2 also could be the result of variations in completeness of diagnosis and reporting of HIV-2 by laboratories and state HIV surveillance programs. In particular, the large percentage of HIV-2 cases reported from New York City might have resulted, in part, from an increased focus on ascertainment of HIV-2 cases by the New York City Department of Health and Mental Hygiene, which has conducted active investigations to identify HIV-2 infections9 and issued an advisory to clinicians regarding diagnostic testing for HIV-2. The percentage of HIV-2 diagnoses based on an HIV-2 DNA test was several times higher among cases reported from New York City (68%), where many cases were diagnosed by the city's public health laboratory, than among cases from the four other regions (range: zero to 26%).
The findings in this report are subject to at least three limitations. First, the surveillance case definition for HIV infection and the working HIV-2 definition used in this analysis are intended primarily for analysis of epidemiologic trends and associations and might be inappropriate for other purposes, such as the clinical management of patients, which might require diagnostic considerations beyond the criteria used for the case definition. CDC currently is reviewing the HIV surveillance case definition to identify areas for revision, including the addition of specific criteria for HIV-2. Second, the 166 HIV-2 infections described in this report are likely an underestimate of HIV-2 cases in the United States. Many of the 76 suspected cases excluded from the analysis because of inadequate data on test results or missing identifiers might be actual HIV-2 cases. Other cases might not have been diagnosed because diagnostic tests specific for HIV-2 (e.g., DNA test or type-differentiating antibody immunoassay) were not widely used, and some cases might not have been recognized because of negative or persistently indeterminate results on assays designed primarily to detect HIV-1 antibodies. Finally, although 11% of HIV-2 cases had nucleic acid evidence of possible coinfection with HIV-1, the full extent of coinfection could not be assessed because HIV-1 nucleic acid test results were missing for 49% of the HIV-2 cases. In addition, 13 of the 17 detectable HIV-1 viral loads were <5,000 copies/mL, raising the possibility that they might be falsely positive.
Additional specific testing for HIV-2 should be considered if test results for HIV-1 are inconsistent with one another, inconclusive, or imply the absence of HIV infection despite clinical evidence suggesting its presence, particularly if the patient was born in or had other associations with areas such as West Africa, where HIV-2 infection is prevalent. Suspected HIV-2 cases should be reported to state or local health departments, which can conduct supplemental diagnostic tests for HIV-2 or arrange for them to be done at the CDC laboratory. In every state, confirmed HIV infection diagnoses are required by law or regulation to be reported to the health department.
What is already known on this topic?
Of the two types of human immunodeficiency virus (HIV), HIV-1 accounts for most HIV infections worldwide and, untreated, causes acquired immunodeficiency syndrome (AIDS). HIV-2 is rarely reported in the United States, largely confined to persons from West Africa, and less likely to cause AIDS. Distinguishing between HIV-2 and HIV-1 is important because their clinical management differs; HIV-2 does not respond to some antiretroviral drugs effective on HIV-1, and HIV-2 cannot be measured by HIV-1 viral load tests.
What is added by this report?
Using a working case definition for HIV-2 infection, CDC identified 166 cases diagnosed during 1987-2009. Most cases (66%) were reported from the Northeast, particularly New York City (46%). A total of 132 (81%) cases were in persons born in West Africa. Most (60%) HIV-2 cases had positive HIV-1 immunoblot test results, indicating that the usual tests for HIV-1 infection can misclassify HIV-2 cases as HIV-1 unless other tests (e.g., DNA test or type-differentiating antibody immunoassay) specific for HIV-2 are used.
What are the implications for public health practice?
Specific testing for HIV-2 should be considered if results of tests for HIV-1 are inconsistent with one another, inconclusive, or imply the absence of HIV infection despite clinical evidence suggesting its presence, particularly if the patient is from West Africa. Suspected HIV-2 cases should be reported to state or local health departments, which can conduct supplemental tests for HIV-2 or arrange for testing at CDC. If the type of HIV is identified as HIV-2 after an HIV case is reported, the report should be updated with the type specified as HIV-2.