Haplotype-dependent HLA susceptibility to nasopharyngeal carcinoma in a Southern Chinese population.

We have conducted a comprehensive case-control study of a nasopharyngeal carcinoma (NPC) population cohort from Guangxi Province of Southern China, a region with one of the highest NPC incidences on record. A total of 1407 individuals including NPC patients, healthy controls, and their adult children were examined for the human leukocyte antigen (HLA) association, which is so far the largest NPC cohort reported for such studies. Stratified analysis performed in this study clearly demonstrated that while NPC protection is associated with independent HLA alleles, most NPC susceptibility is strictly associated with HLA haplotypes. Our study also detected for the first time that A(*)0206, a unique A2 subtype to South and Southeast Asia is also associated with a high risk for NPC. HLA-A(*)0206, HLA-B(*)3802 alleles plus the A(*)0207-B(*)4601 and A(*)3303-B(*)5801 haplotypes conferred high risk for NPC showing a combined odds ratio (OR) of 2.6 (P<0.0001). HLA alleles that associate with low risk for NPC include HLA-A(*)1101, B(*)27, and B(*)55 with a combined OR of 0.42 (P<0.0001). The overall high frequency of NPC-susceptible HLA factors in the Guangxi population is likely to have contributed to the high-NPC incidence in this region.

The causes of cancer in France.

BACKGROUND: While external factors are responsible for many human cancers, precise estimates of the contribution of known carcinogens to the cancer burden in a given population have been scarce. METHODS: We estimated the proportion of cancer deaths which occurred in France in 2000 attributable to known risk factors, based on data on frequency of exposure around 1985. RESULTS: In 2000, tobacco smoking was responsible for 23.9% of cancer deaths (33.4% in men and 9.6% in women), alcohol drinking for 6.9% (9.4% in men and 3.0% in women) and chronic infections for 3.7%. Occupation is responsible for 3.7% of cancer deaths in men; lack of physical activity, overweight/obesity and use of exogenous hormones are responsible for 2%-3% of cancer deaths in women. Other risk factors, including pollutants, are responsible for <1% of cancer deaths. Thus, known risk factors explain 35.0% of cancer deaths, and 15.0% among never smokers. CONCLUSIONS: While cancer mortality is decreasing in France, known risk factors of cancer explain only a minority of cancers, with a predominant role of tobacco smoking.

Evidence for exposure to HTLV-III in Uganda before 1973.

Fifty of 75 serum samples collected in the West Nile district of Uganda between August 1972 and July 1973 contained antibodies reactive with human T-cell leukemia (lymphotropic) virus type 3 (HTLV-III; mean titer, 601), while 12 of 75 samples were positive in a similar test for HTLV type 1 (HTLV-1) antibodies (mean titer, 236). The samples were screened by enzyme-linked immunosorbent assay and positive results were confirmed by a newly developed unlabeled antibody-peroxidase procedure with enhanced sensitivity for detection of antibody binding to immunoblots of HTLV-III antigen, demonstrating antibodies to proteins with molecular weights of 24,000, 41,000, and 76,000 in nearly all positive samples. Analysis of titration data indicated enhanced titers of antibody against HTLV-III and HTLV-I when coinfection occurred. The high prevalence and relatively low titers [compared to serum from patients with acquired immune deficiency syndrome (AIDS)] of antibodies recognizing HTLV-III proteins in sera from this population at a time that may predate or coincide with the appearance or spread of the AIDS agent (HTLV-III) suggest that the virus detected may have been a predecessor of HTLV-III or is HTLV-III itself but existing in a population acclimated to its presence. It further suggests an African origin of HTLV-III.

Elevated immunofluorescence antibody titers to several herpesviruses in Burkitt's lymphoma patients: are high titers unique?

Antibody titers for viral capsid antigens of all four human herpesviruses were measured by immunofluorescence in the sera of 16 Burkitt's lymphoma (BL) PATIENTS, 16 AGE-, SEX-, AND LOCALITY-MATCHED CONTROLS, AND 136 FAMILY MEMBERS FROM THE West Nile District of Uganda. Among family members, titers greater than 1:4 were found in 98% for herpes simplex virus (HSV), 86% for varicella-zoster virus (VZV), 100% FOR CYTOMEGALOVIRUS (CMV), AND 94% FOR Epstein-Barr (EBV). Titers in patients averaged approximately equal to 2 logs (fourfold) higher than those in matched controls for EBV, VZV, and CMV (P EQUALS 0.001); titers for HSV were only slightly higher in cancer patients. The mothers of patients had someuhat higher EBV titers (0.05 smaller than or equal to P SMALLER THAN OR EQUAL TO 0.01) than the mothers of controls, but no other differences between patient and control families were found. By immunofluorescence, a method which apparently has not been used for all four human herpesviruses in BL patients, the patients had elevated antibody titers not only to EBV but also to CMV and VZV. The elevated titers to three of the four human herpesviruses were not due to serologic cross reactions.

Epithelial characteristics of tumor cells in nasophasyngeal carcinoma.

Electron microscopy examination of nasopharyngeal carcinoma grafts in nude mice showed that the tumor cells retained their epithelial characteristics, as betrayed by the presence of keratin fibrils and of functional complexes at the cell membrane. Thus the cells positive for Epstein-Barr virus-specific nuclear antigen present in such grafts do represent the progeny of human epithelial tumor cells. C-type virus particles were observed in one tumor graft, indicating that epithelial human cells could pick up a mouse virus.

Immunogenetic aspects of nasopharyngeal carcinoma. IV. Increased risk in Chinese of nasopharyngeal carcinoma associated with a Chinese-related HLA profile (A2, Singapore 2).

The results of this study of 110 Singapore Chinese with nasopharyngeal carcinoma (NPC) and 91 controls confirmed the association between the occurrence of HLA antigen Singapore 2 (Sin2) and NPC in the Chinese population, and indicated that their increased risk for NPC was confined to the joint occurrence of Sin 2 and A2 antigens. These findings suggested that the genotype of importance in susceptibility to NPC is the A2-Sin 2 haplotype.

Epstein-Barr virus genome studies in Burkitt's and non-Burkitt's lymphomas in Uganda.

Burkitt's lymphoma (BL) has been widely investigated and has attracted attention because of the possible etiologic role of the Epstein-Barr virus (EBV). To further determine the role of EBV in the causation of this tumor, we measured EBV-specific nuclear antigen (EBNA) and EBV DNA using immunofluorescence and nucleic acid hybridization techniques, respectively. Of 34 BL biopsies, 27 tissues (79%) were EBNA-positive, whereas none of the 25 non-BL biopsy tissues were EBNA-positive. Of 15 BL tumors tested, 14 (93%) were EBV DNA-positive with a mean of 39 (range, 8-86) EBV genome equivalents per cell. Each of the 15 non-BL biopsy specimens subjected to nucleic acid hybridization had less than two virus genome equivalents per cell, although all had serologic evidence of past EBV infection. The findings further supported the possible etiologic role of EBV in African BL and negated the passenger hypothesis. The EBV genome could, therefore, be used as a separating marker between African BL and non-BL lymphomas.

Epstein-Barr virus behavior in different populations and implications for control of Epstein-Barr virus-associated tumors.

The epidemiology of Epstein-Barr virus (EBV) infection in populations at different risk for EBV-associated diseases indicates significant differences between the populations. EBV infection takes place much earlier in Uganda, where all children are infected before the age of 2 to 3 years, than in Southeast Asia, where nasopharyngeal carcinoma is prevalent. It is proposed that such early infection in Equatorial Africa is related to the risk for Burkitt's lymphoma. Four possible interventions to control EBV-associated diseases are presented: (a) simple hygienic measures to delay natural primary infection by EBV; (b) EBV vaccine; (c) intervention against cofactors such as malaria in Burkitt's lymphoma; and (d) characterization of high-risk groups to allow early detection and successful treatment.

Cell surface markers on epithelial-Burkitt hybrid cells superinfected with Epstein-Barr virus.

Attempts were made to superinfect two epithelial-Burkitt hybrid cell lines, designated D98/HR-1 and D98/Raji, with Epstein-Barr virus (EBV) and to investigate the expression of some cell surfacr markers including histocompatibility antigens, and the presence of B-cell markers, such as receptors for the third complement component and for monkey red blood cells. Successful superinfection of D98/HR-1 cells with EBV was made evident by the expression of early antigen and, to a lesser extent, virus capsid antigen. Only a rare D98/Raji cell was found to be positive for early antigen. The histocompatibility antigens of the parental cell lines D98, HR-1, and Raji were expressed on the surfaces of the hybrid cells. Receptors for third complement components b and d were not detected on the hybrid cells or on the D98P OR HR-1 cell lines; they were found, however, on the Raji cells, indicating that EBV receptors and complement receptors can be separated. The significance of the infection of the hybrid cells with EBV and the expression of cell surface markers is described.

Epidemiology and immunovirology of human T-cell leukemia/lymphoma virus type I-associated adult T-cell leukemia and chronic myelopathies as seen in France.

Seventeen patients with adult T-cell leukemia (ATL) and 21 with tropical spastic paraparesis/human T-cell leukemia/lymphoma virus type I (HTLV-I)-associated myelopathy (TSP/HAM) were observed during a 3-yr survey (1986-1988) in some hospitals in Paris, France. Most of them were black, originating from high-HTLV-I-endemic areas (West Indies or Africa), but two cases of TSP/HAM occurred in French Caucasians. In one case, the patient acquired the virus from a transfusion during a cardiac transplantation. Most of the ATL cases were diagnosed as acute leukemia or lymphoma, with a proliferation of CD2+, CD3+, CD4+, CD8-, DR+, and CD25+ lymphoid cells. Only three cases were diagnosed as a smoldering ATL. All of the TSP/HAM cases exhibited a spastic paraparesis with a chronic and slow evolution and high HTLV-I antibody titers in serum and cerebrospinal fluid, with a high HTLV-I antibody index and specific HTLV-I immunoglobulin = oligoclonal bands. In TSP/HAM, a high percentage of DR-expressing cells (15 to 40%) was found, with a slightly elevated CD4/CD8 ratio. This was associated with the presence of 1 to 10% abnormally shaped nuclei in lymphoid cells and a polyclonal integration of HTLV-I proviruses in these peripheral blood mononuclear cells. On the contrary, a clonal integration was always found in the ATL malignant cells (leukemic, lymph node, and cutaneous infiltrate). Long-term interleukin 2-dependent T-cell lines (CD2+, CD3+, CD4+, and WT31+) with activated T-cell markers (CD25+ and DR+) producing HTLV-I were established from ATL and TSP/HAM peripheral blood mononuclear cells.

Comparative seroepidemiology of HTLV-I and HTLV-III in the French West Indies and some African countries.

The prevalence of antibodies detected by ELISA against human T-lymphotropic viruses, type I (HTLV-I) and type III (HTLV-III-LAV), is described in a comparative serosurvey in the French West Indies and African countries. The data confirm that the Caribbean basin is endemic for HTLV-I. In this region, HTLV-I antibody prevalence varied from 3.4% to 5.2% among blood donors and increased with age to reach a value of 33% among elderly people from the Dominica Island. In French Guyana, a South American country bordering the Caribbean sea, differences in antibody distribution across three ethnic groups (black Bonis, Indian Wayanas, and Hmongs from Asia) provide clues for investigation of the mode of HTLV-I transmission. Africa appears to be an endemic continent for HTLV-I and HTLV-III. For both viruses, the antibody prevalence exhibited an increasing gradient from northern to equatorial through Sudanic areas. These preliminary data by showing that Africa represents an endemic reservoir of HTLVs and, possibly, of other human retroviruses should stimulate further investigations on the natural history and the geographical origin of these viruses.

Lack of evidence of HTLV-I/II infection in T CD8 malignant or reactive lymphoproliferative disorders in France: a serological and/or molecular study of 169 cases.

Human T lymphotropic virus type II (HTLV-II), originally isolated in 1982 from a patient with a "T hairy cell leukemia", has not yet been proven to be the causative agent of any specific hematological disease. In order to screen for such an event, and because HTLV-II has a preferential tropism for OKT8 (CD8) T cells (both in vivo and in vitro), we searched for the presence of HTLV-II in lymphoproliferative diseases (LP) of CD8+ T cells. We report a serological and/or molecular study of 169 patients with a T CD8 LP, including 76 patients with malignant or reactive T CD8 LP (34 lymphomas, 27 large granular leukemias, three prolymphocytic leukemias, one hairy cell leukemia, 11 reactive T CD8 LP) and 93 HIV-1+ patients with a T CD8 peripheral lymphocytosis ( > 1500/mm3) from a prospective HIV cohort involving 1264 individuals. In the first series, the 40 sera available were all HTLV-I/II negative, except a 67-year-old French Guyanan man, with a cutaneous large T CD8 cell lymphoma, HTLV-I+. Furthermore, the molecular analysis of the 69 available DNA samples by PCR failed to detect any proviral HTLV-I/II sequences, except for the HTLV-I+ patient. The serological study of the 93 HIV-1+ individuals with CD8 lymphocytosis, showed that three patients were HTLV-I+, but none was HTLV-II+. Thus, in contrast to HTLV-I, whose etiological role in adult T cell leukemia is now well established, there is neither epidemiological nor molecular evidence that prototypic HTLV-II may be etiologically associated specifically with any of the CD8+ T cell LP investigated in this report.

Cutaneous T cell lymphomas: mycosis fungoides, Sezary syndrome and HTLV-I-associated adult T cell leukemia (ATL) in Mali, West Africa: a clinical, pathological and immunovirological study of 14 cases and a review of the African ATL cases.

Cutaneous T cell lymphomas (CTCL) are rare lymphoproliferative diseases, which are frequently suspected to be of viral origin. As very few data were available concerning cutaneous T cell lymphomas in tropical Africa, we undertook a clinical, histopathological, immunological and viro-molecular study of patients with a clinical diagnosis of cutaneous lymphoma, in Bamako, Mali. While prior to this study, no case of CTCL had been reported in this country, 14 patients (five women, nine men; mean age 58 years) with a diagnosis of cutaneous lymphoma were seen over a period of 30 months (1992-1994) in the only dermatological department in Mali. Clinically, the most frequent pattern was an infiltrated erythrodermia similar to Sezary syndrome. Nodular lesions and/or plaques were rarely observed. All these cutaneous tumors were T cell lymphoproliferations, only one expressing the CD8+ antigen. A comprehensive analysis of all the available data permitted characterization of three cases of adult T cell leukemia/lymphoma (ATL) associated with HTLV-I (one definitive case, of leukemic type, with demonstration of clonal integration of HTLV-I proviral genome and two probable ATL cases), three cases of Sezary syndrome (SS), two cases of mycosis fungoides (MF) and five cases of pleomorphic cutaneous lymphoma. In one case, the differentiation between MF and pleomorphic cutaneous lymphoma could not be established. HTLV-I serological and/or molecular markers were restricted to the three ATL cases. From the unique definitive ATL case, a T cell line was established from culture of peripheral blood mononuclear cells and sequence analysis of the env gene and the U3-LTR region demonstrated that the virus present in this patient belonged to the cosmopolitan subtype A. Thus, we report here the first evidence of HTLV-I infection and associated ATL in Mali. This is the second ATL case described for the whole Sahelian region (one ATL of the lymphoma type was reported previously in a Mauritanian patient). Furthermore, we demonstrate that the main types of CTCL described in Europe and North America are also present in this African area and that the prevalence of these diseases is greatly underestimated in such regions. Furthermore, no association was observed between HTLV-I/II infection and SS, MF or pleomorphic cutaneous lymphoma in Mali in contrast to other studies.

Human herpes virus 8 (Kaposi's sarcoma herpes virus) and malignant lymphoproliferations in France: a molecular study of 250 cases including two AIDS-associated body cavity based lymphomas.

The new human herpes virus 8 (HHV8) was recently detected in cases of body cavity based lymphoma (BCBL), a rare B cell lymphoma, mostly AIDS-associated. We investigated for HHV8 DNA sequences a series of 250 B or T cell lymphoproliferative malignancies, as seen in France, including 126 leukemias and 124 lymphomas (232 non-AIDS-associated and 18 AIDS-associated tumors). HHV8 sequences were detected in only three patients. The first two were homosexual males, HIV-infected since 1985 who suffered from a BCBL initially characterized in one case by a pleural lymphomatous effusion and a peritoneal one in the other case. A high level of HHV8 copies was detected in the tumoral cells of these two BCBL. In contrast, in the third positive patient who had an AIDS-associated immunoblastic lymphoma, the HHV8 sequences level was quite low. In the two BCBL patients, the HHV8-infected clonal B cells had a large immunoblastic feature with an indeterminate phenotype and were also infected by Epstein-Barr virus. In one BCBL case, a semiquantitative PCR analysis revealed that the HHV8 sequences were much more abundant in the effusion tumor cells than in the cutaneous Kaposi's biopsy while no HHV8 sequence was detectable in the peripheral blood lymphocytes. This study reports HHV8-associated BCBL in European AIDS patients and confirms that HHV8 is present at a high copy number in the tumoral B cells of this malignancy. Furthermore, HHV8 does not seem to play a pathogenic role in any of the other T or B malignant lymphoid neoplasias studied so far. This study also stresses the necessity for quantification studies in interpretation of a positive PCR analysis for HHV8 sequences, especially in patients at risk for HIV infection or Kaposi's sarcoma.

Clinical experience of AIDS in relation to HIV-1 and HIV-2 infection in a rural hospital in Ivory Coast, West Africa.

A dramatic rise in AIDS cases in a rural hospital 60 km west of Abidjan is described. Out of 212 adult cases diagnosed between January and November 1987 as having AIDS, using the World Health Organization (WHO) criteria defined in Bangui, 128 (60%) had antibodies to HIV-1 only, 61 (29%) had antibodies to both HIV-1 and HIV-2, 15 (7%) had antibodies to HIV-2 only and eight (4%) had no detectable antibodies to HIVs. These results, taken together with those of Ouattara et al. showing that both retroviruses have a similar seroprevalence in the general rural and urban populations, suggest that HIV-2 may have a lower physiopathological potential than HIV-1. This suggestion is supported by a 6-month follow-up of 60 AIDS cases, but only longer follow-up of cohorts of patients and healthy subjects with antibodies to one of the two viruses will allow definitive conclusions.

PIP: 212 adults with symptomatology indicative of acquired immunodeficiency syndrome (AIDS) presented to the Ivory Coast's Dabou Hospital between January-November 1987. 163 were males and 49 were females; the majority (151) were from rural areas. In terms of the clinical profile, 99% had experienced severe weight loss (greater than 10% of body weight), 43% had generalized pruritus, 66% reported fever exceeding 1 month's duration, 75% reported diarrhea exceeding 1 month's duration, 55% had experienced coughing for longer than 1 month, and 56% demonstrated generalized adenopathies. 128 (60%) of these 212 individuals were positive for antibodies to human immunodeficiency virus (HIV)-1, 15 (7%) were HIV-2 positive, 61 (29%) were seropositive for both HIV-1 and HIV-2, and 8 (4%) were negative for both viruses. Clinical follow-up was possible in 173 of these cases. After 6 months, those infected with HIV-1 manifested 16 unfavorable outcomes (deterioration or death) and 11 favorable outcomes (stable or improved condition). Among those infected with HIV-2, there were no unfavorable and 4 favorable cases. The group positive for HIV-1 and HIV-2 exhibited a clinical course at 6 months similar to that found among the HIV-1 seropositives: 11 unfavorable and 9 favorable outcomes. The data from the Dabou hospital attest to a steady rise in AIDS detection, from 0.21% of all adult outpatient cases in the 1st quarter of 1987 to 1.03% of cases in the last quarter. Although data from this series suggest a milder evolution for HIV-2 associated cases, a clinical follow-up of individuals seropositive for HIV-1 or HIV-2, over a 2-year period, is underway to confirm whether there is indeed a distinct symptomatology and disease pattern for each viral infection.

Chronic myelopathies associated with human T-lymphotropic virus type I. A clinical, serologic, and immunovirologic study of ten patients in France.

Chronic myelopathy associated with human T-lymphotropic virus type I (HTLV-I) has been described in HTLV-I endemic areas. In Paris, 167 neurologic patients were screened for HTLV-I by enzyme-linked immunosorbent, indirect immunofluorescent, and Western blot assays. Ten of the 11 patients with positive results had a chronic spastic paraparesis with IgG oligoclonal bands and elevated HTLV-I antibody index. Two of them had been born and were living in France, without HTLV-I risk factors. Evoked potentials were abnormal in the nine tested patients and brain magnetic resonance images in three of seven patients. No improvement was observed with steroid treatment. A retrovirus similar to HTLV-I was isolated in five cases at different periods of the disease. Hypotheses of limited endemic areas in western countries are discussed. Early presence and persistence of HTLV-I suggest that it is the etiologic agent.

Prevalence of HTLV-I compared to HIV-1 and HIV-2 antibodies in different groups in the Ivory Coast (West Africa)

Different Ivorian population groups were comparatively tested for antibodies to HTLV-1, HIV-1, and HIV-2. They included 1,334 healthy individuals, 176 unselected medical patients in regional hospitals, 149 female prostitutes, and 65 males with a sexually transmitted disease (STD) collected in 1987 from four regions (east, north, west, and central). The HTLV-1 prevalence averaged 1 to 2.7% in the different regions, without significant increase in sexually overexposed groups. Furthermore, in Dabou Protestant Hospital, 60 km west of Abidjan, 414 blood donors, 109 asymptomatic HIV-positive individuals, and 181 AIDS clinical cases were also tested. While a significant increase in the prevalence of HTLV-1 antibodies was observed only in AIDS patients. HIV-1 and HIV-2 prevalence increased, as expected, from a low rate (2.4%) in the general population to an intermediate rate (13-15%) in blood donors and unselected hospitalized patients and to a high rate (35%) in prostitutes and STD male patients.

The risk of tropical spastic paraparesis differs according to ethnic group among HTLV-I carriers in Inongo, Zaire.

Following the finding in 1990 of six HTLV-I-positive cases of tropical spastic paraparesis (TSP/HAM) is a field survey carried out in the city of Inongo (Bandundu, Zaire), the prevalence of HTLV-I infection has been studied by detection of specific antibodies on a randomized sample of the general population. Among the 1,162 subjects surveyed, 36 (3.1%) were found to be HTLV-I seropositive, giving a seroprevalence of 3.2% (CI: 2.1, 4.3) estimated by direct standardization on age and sex distribution given by the census in the city. Among the five major ethnic groups, the Bolia showed the highest prevalence rate (6.5%) but with no detected TSP/HAM cases, while all six TSP/HAM cases were found among the Ntomba, who showed a prevalence rate of only 2.2%. This finding suggests that besides HTLV-I infection, critical environmental and/or genetic cofactors play a part in the development of TSP/HAM.

Compared prevalence of infections by HIV-1 and HIV-2 during a 2-year period in suburban and rural areas of Ivory Coast.

The testing for human immune deficiency virus type 2 (HIV-2) antibodies of 794 sera collected in July-August 1985 in the four regions of Korhogo, Bondoukou, Man, and Bouaké in Ivory Coast and the collection and testing for HIV-1 and HIV-2 antibodies of 1,126 sera collected in July-August 1987 in the same regions and age groups showed a remarkable stability in the prevalence of infection by these two retroviruses (0.7 and 0.4% in 1985 vs. 0.9 and 0.2% in 1987, respectively, for HIV-1 and HIV-2) in rural areas. In contrast, the increase in prevalence of both HIV-1 (from 1 to 1.9%) and HIV-2 (from 0.8 to 1.3%) from 1985 to 1987 in the urban population was accompanied by a sharp increase during these 2 years of the number of acquired immune deficiency syndromes in regional hospitals.

PIP: The testing for human immune deficiency virus type 2 (HIV-2) antibodies of 794 sera collected in July-August 1985 in the 4 regions of Korhogo, Bondoukou, Man, and Bouake in Ivory Coast and the collection and testing for HIV-1 and HIV-2 antibodies of 1126 sera collected in July-August 1987 in the same regions and age groups showed a remarkable stability in the prevalence of infection by these 2 retroviruses (0.7 and 0.4% in 1985 vs 0.9 and 0.2% in 1987, respectively for HIV-1 and HIV-2) in rural areas. In contrast, the increase in prevalence of both HIV-1 (from 1-1.9%) and HIV-2 (from 0.8-1.3%) from 1985-87 in the urban population was accompanied by a sharp increase during these 2 years in the number of AIDS cases in regional hospitals. (author's modified).

Human retroviruses HTLV-I, HIV-1, and HIV-2 and neurological diseases in some equatorial areas of Africa.

HTLV-I is associated with tropical spastic paraparesis (TSP) in the Caribbean area and with certain chronic myelopathies termed HAM (HTLV-I-associated myelopathy) in Japan. In order to investigate the situation in Africa, we tested for HTLV-I, but also for HIV-1 and HIV-2 antibodies, 94 patients with epidemic spastic paraparesis (ESP) from Zaire and Tanzania, 26 cases of sporadic spastic paraparesis (SSP) and 21 cases of tropical ataxic neuropathy (TAN), both from Ivory Coast, and 319 unselected neurological patients from Ivory Coast, Congo, and Tanzania. While none of the 94 ESP cases nor any of the 21 TAN patients exhibited antibodies to any retrovirus, 4 of the 26 sporadic spastic paraparesis patients had high HTLV-I antibodies in their sera and cerebrospinal fluid (CSF). Three of those were clinically and immunologically identical to TSP, as observed in persons from the Caribbean region, and the fourth case, a poorly explored chronic pyramidal syndrome, could also represent a TSP. Only one of these four cases originally had HIV-1 antibodies. Among the 319 unselected patients, only 5 (1.6%) had HTLV-I antibodies, but 32 (10%) had HIV-1 antibodies and 14 (4.4%) had HIV-2 antibodies, with a number of combined infections, indicating that retroviruses represent potentially important etiological agents for African neurological diseases.