AMERICAN JOURNAL OF PHYSICAL ANTHROPOLOGY 81:465470 (1990)

Prevalence of Antibodies Interactive With HTLV-I Antigens in Selected Solomon Islands Populations

MARTINE Y.K. ARMSTRONG, DANIEL B. HRDY, JAMES R. CARLSON, AND JONATHAN S. FRIEDLAENDER Department of Epidemiology and Public Health, Yale University School of Medicine. New Haven. Connecticut 0651 0 IMYKA): Division o f hfectious and Immunological Diseases, University of California: Davis Medical Center, Sacramento, California 9581 7 IDBH, JRC); Department of Anthropology, Temple Uniuersity, Philadelphia, Pennsyluunia 19122 (JSF)

KEY WORDS Melanesia, Polynesian outlier atoll, Retroviral infection, ELISA, Epidemiology

ABSTRACT Serum samples obtained in 1986 from healthy individuals in three distinct Solomon Islands populations were screened for antibodies to human lymphotropic virus type I (HTLV-I). One of the populations tested lives on the remote Polynesian outlier atoll, Ontong Java. The other two groups, the Baegu and the Lau, are Melanesians living on Malaita, the most populous of the larger Solomon Islands. Eighty-eight of a total of 601 (14.6%) sera tested were repeatably reactive in an enzyme-linked immunosorbent assay (ELISA) that uses as antigen a lysate of HTLV-I viral particles. The prevalence of antibodies interactive with HTLV-I antigens varied among the three groups, ranging from 8.5% (16/188) in the Baegu, through 13% (764) in the Lau, to 18.1% (651359) among the Ontong Java population. The specificity of the screening ELISA was confirmed by protein immunoblot. No serum samples were obtained from children under 9 years of age. Although 121 of the 601 sera came from children between the ages of 9 and 19, none of these were reactive in the HTLV-I ELISA. Starting in the third decade, the prevalence of HTLV-I seropositivity increased with age, from 8.8% (10/113) between the ages of 20 and 29 to a peak of 25.9% (15/58) and 25% (15/60) in the sixth and seventh decade, respectively. This age-specific prevalence pattern is strikingly similar to that which is seen in populations where HTLV-I infection is endemic.

Lymphotropic virus type I (HTLV-I), a retrovirus, was first isolated by Poiesz and co-workers in 1980 from a patient with cuta- neous T-cell lymphoma. An identical virus was subsequently isolated by Yoshida et al. (1982) from patients with adult T-cell leuke- mia (ATL), an aggressive cancer that occurs in endemic clusters in southwestern coastal regions of Japan. Subsequent studies on the HTLV-I virus in Japan (Sugamura and Hi- numa, 1985) revealed that 1) ATL patients had serum antibodies against HTLV-I anti- gens; 2) healthy HTLV-I antibody-positive individuals were found at high frequency (10-37%) in ATL-endemic areas, and were absent or present only at very low frequency (less than l%), in ATL-nonendemic areas; 3) both ATL patients and healthy seropositive individuals tested harbored the HTLV-I vi-

rus in their peripheral blood lymphocytes. These findings established a strong causal relationship between the HTLV-I virus and ATL. A similar etiological association has been made in the Carribbean basin, where a majority of the lymphoreticular malignan- cies have clinicopathological features that correspond to those of Japanese ATL, and where there is co-existing evidence of HTLV- I infection (Blattner and Gallo, 1985). Since then, a significant prevalence of HTLV-I seropositivity has been reported from a vari- ety of geographically dispersed areas that include a number of African countries (De- The et al., 19851, several Arctic regions (Rob-

Received December 13,1988; May 31, 1989.

@ 1990 WILEY-LISS, INC

466 M.Y.K. ARMSTRONG ET AL

ert-Guroff et al., 19851, southern Italy (Man- zari et al., 19851, and Papua New Guinea (Kazura et al., 1987).

Recent resurveys of Solomon Islands pop- ulations included in the longitudinal So- lomon Islands Project initiated in 1966 (Friedlaender, 1988) have provided an op- portunity to look for evidence of human ret- roviral infection in this remote region of the South Pacific. The data reported here docu- ments a significant prevalence of HTLV-I seropositivity in the three groups revisited during 1986.

MATERIALS AND METHODS Study populations

A total of eight tribal groups were sur- veyed between 1966 and 1972 for inclusion in a longitudinal health study, the Solomon Islands Project (Friedlaender, 1988). Six of the original eight populations have been re- visited on two subsequent occasions, all six in 1980, three of the six in 1985 and the other three in 1986. Two of the groups resurveyed in 1986, the Baegu and the Lau, are closely situated upland and coastal communities, respectively, on Malaita, one of the larger and the most populous of the Solomon Is- lands. Both of these groups are Melanesian, Austronesian speakers, and speak dialects of the same language. However, the Lau were, and still are, substantially more accultur- ated than the Baegu. The third group resur- veyed in 1986 lives on the remote Polynesian outlier atoll, Ontong Java, some 150 miles northwest of Malaita. The Ontong Javanese are Polynesian speakers distantly related to the other Austronesian speakers of Malaita. Despite its isolation, this population has un- dergone an intensive acculturation during the past 15 years owing to the lucrative harvesting on the atoll of beche-de-mer, and effective copra production (Bayliss-Smith, 1986).

Serum collection An attempt was made to recover and ex-

amine all individuals previously surveyed, as well as some of their children. Blood sam- ples were obtained by venepuncture from all subjects, except for children under 9 years of age, who were not bled. Blood samples were processed in the field; serum was separated from clotted blood by centrifugation, ali- quoted into cryotubes, and frozen down in liquid nitrogen.

Enzyme-linked immunosorbent assay (ELISA)

Coded serum samples were screened for HTLV-I antibodies using an ELISA kit man- ufactured by Cellular Products Inc. (Buffalo, N.Y.) for research use only. This HTLV-I ELISA utilized 96-well microplates coated with a lysate of purified HTLV-I virus parti- cles harvested from the supernatants of cell cultures used to propagate the virus. A 1:21 serum dilution was added to the coated wells. Any antibodies that bound to the HTLV-I antigens were detected in a two-step process. A goat anti-human IgG antibody conjugated to alkaline phosphatase was added to the same wells and bound to the HTLV-I antigedantibody complex. Addition of the alkaline phosphatase substrate p-ni- trophenylphosphate converted the bound al- kaline phosphate conjugate to a detectable yellow substance in proportion to the amount of HTLV-I antibody bound to the well. The color change was quantitated in a spectrophotometer. Four positive controls and two negative controls were included in each plate. Each serum sample was tested once initially. Those sera considered “reac- tive” according to the manufacturer’s crite- ria were retested in duplicate. If one or both ofthe retests were “reactive,” the sample was considered “repeatably reactive” and inter- preted as positive for HTLV-I antibody.

Western blot (WB) The following Western blot (WB) proce-

dure was used to confirm the specificity of the reaction detected by the HTLV-I ELISA. Purified HTLV-I virus obtained from Hill- crest Biological (Pacific Palisades, CA) was electrophoresed through a 10% polyacryla- mide gel to separate the viral polypeptides according to their molecular weight. The polypeptides were then transferred from the gel on to a nitrocellulose membrane by blot- ting. The nitrocellulose paper was cut into parallel strips, each of which was incubated with an individual serum sample diluted 1:lOO. Binding of antibody to any one of the separated viral proteins was detected using a series of reactions that resulted in the appearance of a visible band on the nitrocel- lulose membrane. These reactions involved the use of a goat anti-human IgG antibody conjugated with horseradish peroxidase, fol- lowed by the addition of the horseradish peroxidase substrate 4-chloro-1-naphthol. The criterion for positivity was the presence

HTLV-I ANTIBODY IN SOLOMONS 467

45

of at least two of three bands representing HTLV-I viral proteins (p19, p24, and p55, all encoded by the viral gag gene).

RESULTS

Serum samples obtained from a total of 601 individuals were tested using an HTLV- I kit manufactured by Cellular Products. The 88 sera repeatably reactive according to the manufacturer’s criteria were interpreted as positive for HTLV-I antibody, giving an overall prevalence of 14.6% (Table 1). The prevalence of HTLV-I antibody seropositiv- ity varied among the three groups. It was highest in Ontong Java (65/359,18.1%), low- est in the Baegu (16/188,8.5%), and interme- diate in the Lau (7/54,13.0%).

-

TABLE 1. Prevalence of HTLV-I antibody seropositivity in 3 Solomon Islands Populations

Positive individuals/ Pouulation total tested(%)

Age-specific preualence No blood samples were obtained from chil-

dren under 9 years of age. Although 121 of the total 601 sera tested came from children and adolescents between the ages of 9 and 19, none ofthese were reactive in the HTLV-I ELISA. First detected in the third decade, the prevalence of HTLV-I seropositivity in- creased with age, from 8.8% (10/113) be- tween the ages of 20 and 29, to a peak of 25.9% (15/58) and 25% (15/60) in the sixth and seventh decade, respectively. The age- specific prevalence was plotted by decade for the Ontong Java population where the larg- est number of subjects (359) were tested and where there was the highest prevalence of HTLV-I antibody seropositivity (Fig. 1). The age-dependent rise was a linear one, peaking between ages 60 and 69. A similar age-de- pendent rise in prevalence was also observed in the Baegu, being 4.2% (5/119) in individu- als under age 40 and 15.9% (11169) in those over 40. In the Lau, however, the prevalence under age 40 (3123, 13%) was the same as that over 40 (4/31,12.9%).

Ontong Java 651359 (18.1) Baegu 16/188 (8.5) Lau 7/54 (13.0) Combined 881601 (14.6)

populations

Age in Years: 10-19 20-29 30-39

50 r

Sex-specific preualence Although there was a preponderance of

males among HTLV-I antibody seropositive individuals over age 40 in each of the three

40-49 50-59 60-69 70&over

> 20 0

0 c 15

2 10

5

0

$?

2 No. Positive Sera 0 Total No. Sera 68 67 89 69 32 24 13

- 18 17 11 10 - - - - 7 - -

Fig. 1. Age-specific prevalence of HTLV-I antibody seropositivity in Ontong Java population. Test for linear trend in proportions (ages 1 0 4 9 ) : XI2 = 34.154, P < ,001

468 M.Y.K. ARMSTRONG ET AL

30

25

populations, there was no statistically sig- nificant difference in prevalence between men and women either under 40 or 40 and over in the three populations combined (Fig. 2).

Specificity o f HTLV-I ELISA Thirty-five coded sera (23 ELISA positive

and 12 ELISA negative) were “blind” tested in an HTLV-I Western blot (WB) to check the specificity of the ELISA reactivity. Of the 23 sera repeatably reactive in the HTLV-I ELISA, 20 were scored as positive on WB. Eleven of the 12 sera negative by HTLV-I ELISA were also negative on WB; 1 of the 12 ELISA negatives was positive on WB. The HTLV-I Western blot pattern of selected sera is shown in Figure 3.

Male - 0 Female

-

DISCUSSION

Using an ELISA screening test, antibodies interactive with HTLV-I antigens were de- tected in a significant proportion of the So- lomon Islanders sampled in 1986. The prev- alence rate of such reactivity was of the order of that reported from areas in Japan where HTLV-I infection is endemic (Sugamura and Hinuma, 1985). In addition, the age-depen- dent rise in seropositivity and the absence of seropositive individuals under 20 years of age was similar to the pattern seen in areas of HTLV-I endemicity, where only 1% or so of infants and young children are seropositive and where the prevalence rises in a linear fashion from early adulthood to a maximum by the age of 65 (Murphy and Blattner, 1988). The preponderance of seropositive

Ontong Java 35 r

> (I) 0

> 30 0

+- 25 c

1- + .-

a. 35

0

8 20

15

1c

5

C

under 40 40 and over

Lau

20

15

10

5

n

35

30

25

20

15

10

under 40 40 and over

All 3 Populations Combined

5

n under 40 40 and over under 40 40 and over

Fig. 2. Age- and sex-specific prevalence of HTLV-I antibody seropositivity in three Solomon Islands populations.

HTLV-I ANTIBOI I Y IN SOLOMONS 469

Fig. 3. The HTLV-I Western blot pattern of selected Solomon Islands (SI) sera. The arrows indicate antibod- ies bound to three major viral proteins whose molecular weights correspond to the p19, p24, and p55 HTLV-I core proteins encoded by the viral gag gene. Lane 1: positive control serum known to contain HTLV-I antibodies. Lane 2: negative control serum. The other 14 lanes represent individual SI sera, some of which were positive in the ELISA and some of which were negative in the ELISA. For example, the sera in lanes 12 and 13 were ELISA positive and reacted strongly with all three pro- tein bands. The serum in lane 16 was also ELISA positive and reacted strongly with two of the major protein bands. The sera in lanes 6 and 11 were ELISA negative. Note the variation in staining intensity among individual sera.

males among Solomon Islanders over age 40 was not, however, characteristic of docu- mented HTLV-I endemic areas where a vari- able excess of seropositivity is seen in women, particularly those over 40 (Murphy and Blattner, 1988).

The seroepidemiological data presented here are consistent with a reservoir of HTLV-I infection in the Solomon Islands population tested. It should be noted, how- ever, that in the absence of a viral isolation from a seropositive individual, it is neces- sary to exclude the possibility that the reac- tion in the screening ELISA represents a false positive. There is at present no general agreement as to what constitutes satisfac- tory confirmation of the specificity of ELISAs used to screen sera for HTLV-I anti- bodies. The Western blot technique is cur- rently the standard confirmatory test for antibodies to another human retrovirus, the AIDS virus. Use of the Western blot proce- dure to confirm the presence of HTLV-I an- tibodies is more problematic in that there are many fewer virus-specific bands present and their visualization is very variable. Clearly, all the ELISA-positive Solomon Islands sera will have to be re-examined once a reliable confirmatory test has been developed.

There is no evidence that the AIDS virus is prevalent in the Solomon Islands. Ninety of the Ontong Java sera were screened for an- tibodies to HIV-1 (Human Immunodefi- ciency Virus-1). Although five of these sera were repeatedly reactive in an HIV-1 screen- ing ELISA, the confirmatory HIV-1 Western blot was negative in all five cases. Evidence of HIV-1 infection was also lacking in the three separate linguistic groups surveyed in 1985 on Bougainville, a Solomon Island that is now part of Papua-New Guinea (Hrdy et al., 1986).

The natural history and mode of transmis- sion of HTLV-I infection is largely unre- solved. An analysis of the distribution of HTLV-I antibodies among married couples and their children in an area of Japan en- demic for HTLV-I infection (Kajiyama et al., 1986) is consistent with two possible modes of transmission in the family setting, one from mother to child, the other from husband to wife, but not vice versa. The second mode suggests that the virus may be transmitted in semen. The virus has also been shown to be transmitted by blood transfusion (Sato and Okochi, 1986). The fact that most of the HTLV-I endemic areas are in tropical re- gions of high humidity has raised the addi- tional possibility of insect vector transmis- sion. Whether or not genetic factors underlie susceptibility to HTLV-I infection has not been determined.

There is good evidence that HTLV-I is the causal agent of certain forms of T-cell

470 M.Y.K. ARMSTRONG ET AL.

lymphomdeukemia (Murphy and Blattner, 1988). There is now evidence also linking HTLV-I to certain chronic neurological dis- eases such as tropical spastic paraparesis (TSP) (Sarin, 1988). The great majority of patients with TSP in the West Indies are HTLV-I-antibody positive, as are the great majority of such patients in Japan, where TSP is known as HAM (for HTLV-I associ- ated myelopathy). It appears that HTLV- I-antibody positive individuals in Japan run a relatively low risk of developing a T-cell malignancy and an even lower risk of devel- oping TSP or HAM over the course of a lifetime. What co-factors determine the de- velopment of such diseases remain to be elucidated. Similarly, predictors for disease development have yet to be defined.

The significance of HTLV-I infection for Solomon Islanders is not known in the ab- sence of data regarding the incidence of ei- ther T-cell malignancy or chronic neurologi- cal disease. Such data will undoubtedly be hard to come by, given the remote location of much of the population of these islands, and the fact that even in HTLV-I endemic areas, HTLV-I associated disease is relatively un- common.

ACKNOWLEDGMENTS

This project was supported by the Na- tional Science Foundation. The expert tech- nical assistance of Louise Camera Benson and J o Ann Yee is gratefully acknowledged.

LITERATURE CITED

Bayliss-Smith T (1986) Ontong Java: Population, econ- omy and society, 1970-1986. Occasional paper no. 9. South Pacific Smallholder Project, University of New England, Australia.

Blattner WA, and Gallo RC (1985) Human T-cell leukemidymphoma viruses: Clinical and epidemio- logic features. Curr. Top. Microbiol. Immunol. 115: 67-88.

De-The G, Gessain A, Gazzalo L, Robert-Guroff M, Naj- berg G, Calender A, MPangi Peti, Brubaker G, Bensli- man A, Fabry F, Stobel M, Robin Y, and Fortune R (1985) Comparative sero-epidemiology of HTLV-I and HTLV-111 in the French West Indies and some African countries, Cancer Res. (Suppl.) 45:4633s4636s.

Friedlaender JS, (1988) The Solomon Islands Project: A Long-Term Study of Health, Human Biology and Cul- tural Change. Oxford, Clarendon Press.

Hrdy DB, Carlson JR, and Friedlaender JS (1986) Ab- sence of antibodies to HTLV-I11 on Bougainville (Papua New Guinea). Med. J. Aust. 144:723.

Kajiyama W, Kashiwagi S, Ikematsu H, Hayashi J, Nomura H, and Okochi K (1986) Intrafamilial trans- mission of adult T cell leukemia virus. J. Infect. Dis.

Kazura JW, Saxinger WC, Wenger J, Forsyth K, Leder- man MM, Gillespie JA, Carpenter CCJ, and Alpers MA (1987) Epidemiology of human T-cell leukemia virus type I infection in East Sepik province, Papue New Guinea. J. Infect. Dis. 155:llOO-1107.

Manzari V, Gradilone A, Barillari G, Zani M, Collalti E, Pandolfi F, DeRossi G, Liso V, Babbo P, Robert-Guroff M, and Frati L (1985) HTLV-I is endemic in Southern Italy: Detection ofthe first infectious cluster in a white population. Int. J . Cancer 36.557-559.

Murphy EL, and Blattner WA (1988) HTLV-I associated leukemia: A model for chronic retroviral diseases. Ann. Neurol. (Suppl.) 23:S174-S180.

Poiesz BJ, Ruscetti FW, Gazdar AF, Bunn PA, Minna JD, and Gallo RC (1980) Detection and isolation of type C retrovirus particles from fresh and cultured lympho- cytes of a patient with cutaneous T-cell lymphoma. Proc. Natl. Acad. Sci. U.S.A. 77:7415-7419.

Robert-Guroff M, Clark J, Lanier AP, Beckman G, Mel- bye M, Ebbesen P, Blattner WA, and Gallo RC (1985) Prevalence of HTLV-I in arctic regions. Int. J. Cancer 36:651-655.

Sarin PS (1988) HTLV-I, adult T-cell leukemia and tropical spastic paresis. Ann. Neurol. (Suppl.) 23:

Sat0 H, and Okochi K (1986) Transmission of human T-cell leukemia virus (HTLV-I) by blood transfusion: Demonstration of proviral DNA in recipients' blood lymphocytes. Int. J. Cancer 37:395400.

Sugamura K, and Hinuma Y (1985) Human retrovirus in adult T-cell leukemidymphoma. Immunol. Today

Yoshida M, Miyoshi I, and Hinuma Y (1982) Isolation and characterization of retrovirus from cell lines of human adult T-cell leukemia and its implication in the disease. Proc. Natl. Acad. Sci. U.S.A. 79:2031-2035.

15435 1-85 7.

S181-Sl84.

6: 83-88.