Identification of human T-cell lymphotropic virus infection in a semi-isolated Afro-Brazilian quilombo located in the Marajó Island (Pará, Brazil)

Antibodies to human T-cell lymphotropic virus-1 and 2 (HTLV-1 and 2) were tested in 259 inhabitants (98 males and 161 females) of four villages of the Marajó Island (Pará, Brazil) using enzyme immunoassays (ELISA and Western blot). Types and subtypes of HTLV were determined by nested polymerase chain reaction (PCR) targeting the pX, env and 5 LTR regions. HTLV-1 infection was detected in Santana do Arari (2.06%) and Ponta de Pedras (1%). HTLV-2 was detected only in Santana do Arari (1.06%). Sequencing of the 5 LTR region of HTLV-1 and the phylogenetic analysis identified the virus as a member of the Cosmopolitan Group, subgroup Transcontinental. Santana do Arari is an Afro-Brazilian community and the current results represent the first report of HTLV-1 infection in a mocambo located in the Brazilian Amazon region.

Human T-lymphotropic virus type I and II infections in First Nations alcohol and drug treatment centres in British Columbia, Canada, 1992-2000

Since 1992, prevalence data on HTLV-I and II have been collected as part of an ongoing viral seroprevalence study in clients of six First Nations alcohol and drug treatment centres in British Columbia, Canada. Prior studies indicate that the lifetime risk of clinical disease (neurologic or hematologic) resulting from HTLV-I infection is low (less than 5%) and HTLV-II to date has not been clearly associated with clinical disease. In 1993, the first cases of HTLV-I-associated myelopathy or tropical spastic paraparesis (HAM/TSP) were reported in four Aboriginal residents of British Columbia; these were the first reports of HTLV-I linked disease among Aboriginal persons in Canada. All clients of the treatment centres involved in this study were offered confidential, voluntary testing following pre-test counseling, and the results are given to participants before the residential session is complete. 1953 men and women were tested; 11 were positive for HTLV-1 (0.56%) and 33 were positive for HTLV-2 (1.8%).

HTLV-II risk factors in Native Americans in Florida

Earlier virologic studies established that human T-cell lymphotropic virus type II (HTLV-II) is the predominant retrovirus type found among Seminole Indians in southern Florida. We studied 46 members of the Seminole tribe living on 3 reservations to determine the risk factors for HTLV-II and to investigate disease association with the virus. The donors' plasma samples were evaluated with the enzyme-linked immunosorbent and Western blot assays. DNA extracted from their peripheral blood mononuclear cells were analyzed by type-specific polymerase chain reaction amplification and detection of the HTLV pol gene using the primer pair SK110/SK111, and the probes SK112 or SK188. One of 46 (2%) subjects was identified as HTLV-I positive and 11 (24%) were identified as HTLV-II positive. Restriction fragment length polymorphism and sequence analyses indicated that all of the HTLV-II strains were subtype b. Mitochondrial DNA analyses indicated that all of the HTLV-II-positive subjects had an Amerindian haplotype. HTLV-II was more prevalent in Indians who were >45 years of age or female, had multiple sex partners or had received a blood transfusion. However, only the latter risk factor was statistically significant. Three of the HTLV-II-positive Indians demonstrated signs and symptoms of an ataxic neuropathy. The data support that HTLV-IIb is endemic among the Seminoles and that they will be a key population for further virologic studies.

First seroepidemiological study and phylogenetic characterization of human T-cell lymphotropic virus type I and II infection among Amerindians in French Guiana

We investigated the serological, epidemiological and molecular aspects of human T-cell lymphotropic virus type I and II (HTLV-I/II) infection in the Amerindian populations of French Guiana by testing 847 sera. No HTLV-II antibodies were detected, but five individuals (0.59%) were seropositive for HTLV-I. Analysis of the nucleotide sequences of 522 bp of the env gene and the compete LTR showed that all of the strains from French Guiana belonged to the cosmopolitan subtype A. The similarities were greater between Amerindian and Creole strains than between Amerindian and Noir-Marron strains or than between Creole and Noir-Marron strains. Phylogenetic analysis showed two clusters: one of strains from Amerindians and Creoles, which belong to the transcontinental subgroup, and the other of strains from Noirs-Marrons, belonging to the West African subgroup. Our results suggest that the Amerindian HTLV-I strains are of African origin.

Characteristic distribution of HTLV type I and HTLV type II carriers among native ethnic groups in South America

To confirm the geographic and ethnic segregation of HTLV-I and HTLV-II carriers in native populations in South America, we have conducted a seroepidemiological study of native populations in South America, including HTLV-I carriers distributed among seven ethnic groups in the Andes highlands of Colombia, Peru, Bolivia, Argentina, and Chile, and two ethnic groups on Chiloe Island and Easter Island; and HTLV-II carriers distributed among seven ethnic groups of the lowlands along the Atlantic coast of Colombia, Orinoco, Amazon, and Patagonia, and one ethnic group on Chiloe Island. The incidence rate of HTLV-I and HTLV-II carriers varied among the ethnic groups, ranging from 0.8 to 6.8% for HTLV-I seropositivity and from 1.4 to 57.9% for HTLV-II seropositivity. A new HTLV-I focus was found among the Peruvian Aymara (1.6%), the Bolivian Aymara (5.3%) and Quechua (4.5%), the Argentine Puna (2.3%), and the Chilean Atacama (4.1%), while on HTLV-II focus was found among the Brazilian Kayapo (57.9%), the Paraguayan Chaco (16.4%), and the Chilean Alacalf (34.8%) and Yahgan (9.1%). The distribution of HTLV-I/II foci showed a geographic clustering of HTLV-I foci in the Andes highlands and of HTLV-II foci in the lowlands of South America. It was thus suggested that South American natives might be divided into two major ethnic groups by HTLV-I and HTLV-II carrier state.

Endemic infection with HTLV-IIB in Venezuelan Indians: molecular characterization

The peripheral blood of 41 Yaruro and Guahibo Indians from Venezuela was examined for HTLV antibodies and DNA. Twenty-five samples (61%) were found to be infected with HTLV-IIB. The sensitivities of the serologic and DNA polymerase chain reaction (PCR) analyses were 80% and 96%, respectively. Epidemiologic studies supported both sexual and perinatal transmission of the virus. Sequence analyses of the HTLV-IIB strains from these Indians indicate that they are unique relative to HTLV-II detected in other groups of humans. HTLV-IIB-G2 isolated from a Guahibo Indian is the most divergent HTLV-IIB strain relative to the prototype HTLV-II NRA.

[Infection with T-lymphotropic virus in Dutch blood donors, 1993-1996]

Infection with human T-lymphotrophic virus (HTLV) type 1 causes a neurological disorder or leukaemia in a minority of infected persons. Since January 1993 the Dutch blood banks screen each donation for presence of HTLV-1 infection. Approximately 4,000,000 donations from 700,000 donors have been tested. The numbers of confirmed HTLV-1 positive donors were: 1993: 15; 1994: 6; 1995: 8; 1996: 3. In 1995 one case of HTLV-2 infection was detected as well. In 26/32 (81%) of the HTLV-1 positive cases either the donor or his/her partner originated from HTLV-1 endemic areas. The introduction of HTLV screening prevents the silent spread of HTLV via blood transfusion.

Molecular phylogeny of human T-cell leukemia virus type I and II of Amerindians in Colombia and Chile

Six human T-cell leukemia virus type I (HTLV-I) and eight human T-cell leukemia virus type II (HTLV-II) cases newly isolated from the South American countries of Colombia and Chile were analyzed together with the two Amerindian HTLV-I isolates previously reported. All of the HTLV-I isolates belonged to the transcontinental subgroup of the "cosmopolitan" group, and Colombian isolates, including those from native Amerindians and Negroes, formed a single tight cluster within this subgroup. The transcontinental subgroup consisted of isolates from various regions such as the Caribbean basin, India, Iran, South Africa, Sakhalin, and Japan, and included isolates from the "Ainu" and "Okinawa" people, regarded as relatively pure Japanese descended from the prehistoric "Jomon" period which began more than 10,000 years ago. This implied a dissemination of the subgroup associated with the movement of human beings in ancient times. On the other hand, all of the HTLV-II isolates from native Amerindians in Colombia and Chile belonged to the HTLV-IIb subtype which has previously been reported to be mainly endemic in certain populations of native Amerindians. The southernmost isolate from Chile, showing wide distribution of the IIb subtype in native South Amerindians and largest heterogeneity of the subtype in Colombian isolates, supported the idea that the HTLV-IIb subtype has been endemic for a long time in native Indians of South America.

Comparative performances of enzyme-linked immunosorbent, western blot, and polymerase chain reaction assays for human T-lymphotropic virus type II infection that is endemic among Indians of the Gran Chaco region of South America

BACKGROUND

Human T-cell lymphoma/leukemia viruses types I and II (HTLV-I and HTLV-II) are related exogenous human retroviruses. The former is definitely pathogenic while disease association with the latter is unclear. There are two subtypes of HTLV-II, A and B. Currently, enzyme-linked immunosorbent assays (ELISAs) based on HTLV-I antigens are used to screen for the presence of HTLV-I and -II antibodies. Confirmation and subtyping are accomplished by Western blot (WB) or ELISAs based on HTLV-I whole viral antigens and/or HTLV-I and HTLV-IIA peptides. The sensitivity and specificity of these serologic assays were compared to those of HTLV-I and-II-specific polymerase chain reaction (PCR) assays in tests on samples from Indians from South America in whom the HTLV-IIB subtype is endemic.

STUDY DESIGN AND METHODS

Sera from 246 Gran Chaco Indians were evaluated for HTLV antibodies with the use of four ELISAs (Retrotek HTLV-I; Cambridge Biotech rgp21 enhanced HTLV-I/II; Vironostika HTLV-I/II; and Select HTLV-I/II), and a WB assay. Peripheral blood leukocyte DNA from each Indian was analyzed for HTLV-I or HTLV-II pol DNA via PCR. Fifteen of the PCR-positive samples were further subtyped via cloning and sequencing and/or oligomer restriction.

RESULTS

Ninety-seven samples (39%) were positive for HTLV-II by serologic and/or PCR assays. All 15 positive DNA samples that were further analyzed were of the HTLV-IIB subtype and were clustered as a highly conserved phylogenetic group. Comparative analyses indicate that the sensitivity and specificity of the various assays were: PCR, 97 and 100 percent; Retrotek, 70 and 91 percent; Cambridge Biotech, 74 and 96 percent; Vironostika, 73 and 99 percent; Select 72 and 98 percent; and WB, 70 and 100 percent.

CONCLUSION

The sensitivities of the tested HTLV serologic assays were comparable. However, the specificity of the Retrotek ELISA was significantly lower than that of the others. When positive, the subtyping assays were very specific. However, PCR assays would seem preferable or to be a necessary adjunct for the sensitive detection of HTLV-IIB infection.

Immunogenetics of HTLV-I/II and associated diseases

The ethnic background of human T-lymphotropic virus types I and II (HTLV-I/II) infections and associated diseases was investigated in association with human leukocyte antigens (HLA) (alleles) and haplotypes. Japanese HTLV-I carriers were characterized by two categories of HLA class I antigens (A24, A26, B7, B61, Cw1, and Cw7) and class II alleles (DRB1 *0101, 0803, 1403, 1501, and 1502 and DQB1 *0303, 0501, and 0601); one category was associated with adult T-cell leukemia (ATL) patients and the other with HTLV-I-associated myelopathy/tropical spastic paraparesis (HAM/TSP) patients. The ATL-associated haplotypes had unique DRB1-DQB1 alleles (0901-0303, 1501-0602, 1401-0503), which were correlated with a low immune responsiveness to HTLV-I, while the HAM/TSP haplotypes had different DRB1-DQB1 alleles (0101-0501, 0803-0601, 1502-0601), which were correlated with a high immune responsiveness to HTLV-I. Both ATL- and HAM/TSP-associated haplotypes were found among HTLV-I carriers and the patients from other ethnic groups (Jamaican blacks, Andes natives, South American mestizos, and Mashhadi Jews). HLA haplotypes of HTLV-II carriers were different from those of HTLV-I carriers among South American natives. These results suggested that HTLV-I/II infections and the associated diseases might be determined by immunogenetic factors segregated with HLA alleles and haplotypes.

Quantification of HTLV-II proviral copies by competitive polymerase chain reaction in peripheral blood mononuclear cells of Italian injecting drug users, central Africans, and Amerindians

To better correlate the burden of human T cell leukemia virus type I (HTLV-I) and type II (HTLV-II) infection with diagnostic and prognostic markers, we developed a new competitive polymerase chain reaction (PCR) assay for the quantitative determination of proviral copy numbers in infected cells. A competitive plasmid was constructed that carried a 112-bp fragment from a highly conserved region of the HTLV tax gene and that was further modified by inserting a sequence of 24 bp. This competitive PCR assay system can be used for the quantification of HTLV-I and HTLV-II proviral DNA as demonstrated by using HTLV-I- and HTLV-II-infected cell lines and/or patient material. We determined the HTLV-II proviral load in peripheral blood mononuclear cells (PBMCs) of 11 Italian injecting drug users (IDUs) infected by this virus and in PBMCs of 10 seropositive Amerindian and Central African individuals from endemically infected ethnic groups. A great variation was observed in the number of HTLV-II proviral sequences in the PBMCs of Italian drug abusers, ranging from 5-10 to 16,239 copies/10(5) cells. There was no clear-cut correlation between proviral load, CD8 count, stage of HIV-1 infection, and therapy. A considerable variation in HTLV-II proviral load was also observed in PBMCs of Amerindians and Central Africans with no correlation between the amount of HTLV-II provirus and the geographic origin of the infected individuals.

Prevalence of human T cell lymphotropic virus type II in American Indian populations of the southwestern United States

We investigated the seroprevalence of human T cell lymphotropic virus type II (HTLV-II) using a screening enzyme-linked immunosorbent assay (ELISA) and immunoblot confirmation among the predominant tribes (Pueblo and Athapaskan) served by two large Indian Health Service facilities in New Mexico. Among persons being treated for sexually transmitted diseases, eight (3.2%) of 250 were seropositive for HTLV II, compared with eight (2.1%) of 385 women attending prenatal clinics. In a survey of unselected patients at one of the facilities, 15 (3.4%) of 446 were seropositive. Of 31 seropositive subjects, 25 were infected with HTLV-II and six infections could not be typed. Sera from nine (29%) of the 31 infected subjects had absorbance values less than the manufacturer's cutoff in the ELISA. Both Pueblo and Athapaskan groups had similar overall seroprevalences, but women tended to have a slightly higher seroprevalence than men, and seroprevalence tended to increase with age. These data show that HTLV-II infection is present among diverse groups of American Indians in the southwestern United States. Present ELISA screening tests, such as those used in this study, lack sensitivity to HTLV-II infection unless a reduced absorbance cutoff is used.

Human T-lymphotropic virus type II in Panamanian Guaymi Indians

BACKGROUND

Currently, human T-lymphotropic virus type II (HTLV-II) is the most prevalent human retrovirus, detected in persons presenting to donate blood in the United States. Only scant information is available with which to counsel HTLV-II-seropositive deferred donors or other persons about the ways in which they may spread HTLV-II to others.

STUDY DESIGN AND METHODS

To increase understanding of the modes of transmission of HTLV-II, a seroepidemiologic study was conducted among Panamanian Guaymi Indians, a recently identified focus of endemic HTLV-II infection. Subjects were tested for serologic evidence of infection by HTLV-II, HTLV type I, hepatitis B virus, and nine other infectious agents by enzyme immunoassays and specific confirmatory tests.

RESULTS

Nine (8.3%) of 109 persons tested HTLV-II-seropositive. HTLV-II seropositivity was more likely in persons with serologic evidence of prior hepatitis B virus infection. Sexual contact with HTLV-II-seropositive partners, but neither parenteral exposure nor breast-feeding, was identified as a risk factor for HTLV-II.

CONCLUSION

In Guaymi Indians, HTLV-II appears to be spread primarily through sexual transmission.

Geographic independence of HTLV-I and HTLV-II foci in the Andes highland, the Atlantic coast, and the Orinoco of Colombia

To clarify the ethnic specificity of human T cell leukemia virus type I (HTLV-I) and type II (HTLV-II) carriers among Colombian native Indians, we investigated the geographic distribution of HTLV-I and HTLV-II seroprevalence among the isolated ethnic groups of Mongoloid origin in the Andes highlands and the Atlantic coast of Colombia. HTLV-I carriers were found in 1.6% (1/62 samples) of Inga, 8.5% (5/59) of Kamsa, and 0% (0/55) of Cumbal Indians who live in the Andes highlands at 3000 m above sea level. On the other hand, HTLV-II carriers were found in 4.1% (5/123) of Wayuu Indians, who live in the Guajira region of the Atlantic coast of Colombia at a distance of 1000 km from the Andes highlands. This ethnic specificity of HTLV-II was similarly observed among Guahibo Indians in the Orinoco. The seroprevalence of HTLV-I and HTLV-II was mutually exclusive among Inga, Kamsa, and Wayuu Indians. These results suggest that HTLV-I and HTLV-II may have evolved among Mongoloid populations and been independently transmitted among two different lineages of Colombian native Indians, Andes highlanders and Atlantic coast lowlanders.

First description of endemic HTLV-II infection among Venezuelan Amerindians

We describe for the first time the presence of human T lymphotropic virus type II (HTLV-II) infection in Venezuela, among the Pume Amerindians living in the southern plains of the country. Antibodies to HTLV-II antigens were assessed by enzyme immunoassays (Elisa), Western blot, radioimmuno-precipitation, and immunofluorescence; titration studies against HTLV-I- and HTLV-II-infected cell lines were very useful in the differentiation of HTLV-I and HTLV-II antibodies. The HTLV-II general prevalence was 5%; however, there is a striking difference in prevalence between the truly isolated villages (0%) when compared to those living along the riverside and thus in contact with outsiders (9%). Preliminary evidence suggests sexual contact as the main source of transmission. These findings might suggest that HTLV-II in Venezuela originated through contact with outsiders rather than ancient infection related to the origins of the Pume.

Human T lymphotropic virus type II (HTLV-II) infection among female prostitutes in Yucatan, Mexico

The antibodies to human T-lymphotropic virus type I/II (HTLV-I/II) were determined in non-intravenous drug-using female prostitutes from Merida Yucatan, Mexico. Serum specimens from 282 female prostitutes collected during 1990 were tested initially by enzyme immunoassay and further confirmed by western blot assays. Of these, 5 (1.8%, 95% confidence interval 0.2 to 3.3) were shown to be HTLV-I/II positive (reactivity to p24gag and gp68/r21eenv). All five specimens were shown to be infected with HTLV-II by immunoassays using type-specific synthetic peptides and recombinant proteins. Long-term cell lines developed from two individuals demonstrated active viral replication and were of CD8 phenotype. Polymerase chain reaction analysis from four of these five prostitutes demonstrated HTLV-II-specific amplification of all four specimens, of which one was subtype a (HTLV-IIa) and three were subtype b (HTLV-IIb). These data show that HTLV-II is the predominant HTLV type among female prostitutes from the Yucatan.

Endemic human T cell leukemia virus type II infection in southwestern US Indians involves two prototype variants of virus

Human T cell leukemia virus (HTLV) type II is endemic in certain American Indians, and high rates of infection occur in intravenous drug users (IVDUs). North American IVDUs are infected with two distinct variants, HTLV-IIa and -IIb. If IVDUs became infected as a result of interaction with members of an American Indian population, both viral forms should be demonstrable in such populations. Nucleotide sequence analysis of 630 bases of the env gene encoding the gp21 protein was done on DNA from 12 New Mexico Indians (8 Pueblo, 4 Navajo). All samples were typical subtype a or b viruses. Seven of the 8 Pueblo and 2 of 4 Navajo had subtype b; the rest had subtype a. The results are compatible with an indigenous New World origin for both subtypes of HTLV-II.

HTLV-II seroprevalence in pygmies across Africa since 1970

HTLV-II-specific antibodies, with patterns similar to those in the Americas, were present in sera collected about 1970 from Bambuti pygmies in Zaire (14/102; 14%) and from pygmies in Cameroon (5/214; 2.3%), and were more prevalent than HTLV-I. In the Central African Republic, 504 pygmies were HTLV negative. After finding of 4 HTLV-II seropositives among 12 Bambuti pygmies sampled in 1991, this established that HTLV-II or a related retrovirus is present as an ancient endemic in some, but not all, insulated groups of African pygmies, similar to the HTLV-II distribution in Amerindian populations. The endemic among the oldest inhabitants of central Africa, and the occasional and scattered occurrence of apparent HTLV-II among predominant HTLV-I in other Africans, fit well with an ancient African virus and not with importation from the New World. Theories on the origin and evolution of the primate T-lymphotropic viruses (PTLVs) should take into account the longstanding presence of HTLV-II-type viruses in both the Old and New World. Present serology suggests identity of the African viruses with HTLV-II, but their assignment to a new HTLV type is open should genetic analysis show strong divergence from American HTLV-II. Clinical expression, if any, remains to be studied.

Determination of human T lymphotropic virus type by polymerase chain reaction and correlation with risk factors in northern California blood donors

To determine the relative prevalence of human T lymphotropic virus (HTLV) types I and II, type-specific polymerase chain reaction (PCR) was done on seropositive northern California blood donors. From October 1988 through March 1990, 67 (0.055%) of 122,517 blood donors had confirmed HTLV antibody. Seropositive donors were more likely to be middle-aged, female, and nonwhite than the overall donor base. PCR of samples from 30 HTLV-seropositive donors yielded 19 (63%) with HTLV-II and 9 (30%) with HTLV-I; 2 (7%) were repeatedly negative by PCR. HTLV-I-infected subjects had ancestry (n = 3), sexual contact (n = 3), or paternal military service in (n = 1) Japan or the Caribbean. HTLV-II carriers reported past intravenous drug abuse (n = 3) or sex with a drug user (n = 11). Two carriers of each type reported previous blood transfusions, and 1 HTLV-II carrier was a dentist with no other risk factors for retroviral infection.

Human T-lymphotropic virus type II among Guaymi Indians--Panama

Human T-lymphotropic virus type 2 (HTLV-II) is one of four retroviruses (i.e., HTLV-I, human immunodeficiency virus type 1 [HIV-1], and HIV-2) that are known to infect humans. During 1988, investigators from the Gorgas Memorial Laboratory (GML) and the National Institutes of Health (NIH) tested serum specimens collected during 1978-1987 throughout Panama to determine the prevalence of HTLV-I infection (1), which is endemic in southern Japan and the Caribbean basin (2). HTLV seropositivity rates were low in all areas of Panama except among the Guaymi Indians in Changuinola (9%),* a city in western Panama (Figure 1). During 1989-1991, additional testing of 36 seropositive specimens from Guaymi Indians detected that the infections were due to HTLV-II (3,4). This report summarizes risk factor data for HTLV-II infection among Guaymi Indians.

Chronic neurodegenerative disease associated with HTLV-II infection

Although human T-cell leukemia virus (HTLV) type I is known to cause a number of diseases, there has been no convincing evidence of pathological changes after infection with the related virus, HTLV-II. We have found an endemic focus of HTLV-II infection among members of an American Indian population in New Mexico, USA. We set out to determine the pathological consequences of HTLV-II infection in this population and identified two sisters (aged 59 and 46 years) with a disease superficially resembling the myeloneuropathy induced by HTLV-I. These women had a syndrome similar to the olivopontocerebellar atrophy variant of multiple system atrophy, and HTLV-II infection was confirmed by western blot and the polymerase chain reaction. Thus, HTLV-II may, like HTLV-I, cause a progressive neurodegenerative disease.

Patterns of HIV-1 and HTLV-I/II in intravenous drug abusers from the middle atlantic and central regions of the USA

Seroprevalence of human immunodeficiency virus type 1 (HIV-1) and human T lymphotropic virus types I and II (HTLV-I/II) was determined among 1160 intravenous (iv) drug abusers from five drug treatment or medical centers (Manhattan, Brooklyn, New Jersey, Detroit, and New Orleans). HIV-1 infection ranged from 5% in New Orleans to 48% in New York City. Hispanics and blacks had a significantly higher rate of HIV-1 infection than whites (P less than .01), but within each group rates were similar between males and females and by age stratum. HTLV-I/II seroprevalence increased with age from 3% in the 20-29 year age group to 37% in the group greater than 50 years. New Orleans and Manhattan (24%) had the highest rate, and blacks (19%) had a higher rate than either Hispanics (6.3%) or whites (7.3%). No association between HIV-1 and HTLV-I/II infection was observed except in Manhattan. When compared with iv drug abusers infected only with HIV-1, dually infected subjects had more clinical symptoms related to immune deficiency but a lower prevalence of HIV antigenemia. These data document the frequent occurrence of retroviral infections in iv drug abusers. The contrast between the two classes of virus suggests that HIV-1 is more efficiently transmitted, while the age-dependent rise in HTLV-I/II seroprevalence suggests cumulative exposure of a less-transmissible agent.