HLA and nasopharyngeal carcinoma in Malays.

HLA associations were observed in unrelated Malay patients with nasopharyngeal carcinoma (NPC). HLA-B18 was observed in 18/45 (40%) Malay NPC patients compared to 22/167 (13%) Malay normals (P = 0.0001; Pc = 0.0027, RR = 4.4). The frequency of HLA-B17, one of the antigens associated with Chinese NPC, was also increased among Malay NPC (13/45 29%) compared to controls (18/167 11%; P = 0.003, Pc = 0.07 RR = 3.4). Similar to the findings among Chinese NPC, the frequency of B17 was higher in early onset (less than or equal to 30 years) Malay NPC resulting in a higher relative risk (RR = 5.0).

Nasopharyngeal carcinoma (NPC) is a geographically restricted tumour that has attracted considerable attention recently because of the unique opportunity to study the genetic and environmental factors in the pathogenesis of this tumour. Implicated environmental factors include the Epstein Barr virus, nitrosamine and plant extracts (Prasad et al., 1983). The data of different incidences among different populations, incidence among migrants and family clustering suggested the involvement of genetic factors in addition to environmental factors. HLA associations have been reported in NPC. With the highest incidence populations, similar HLA associations have been reported among the Chinese from Singapore (Simons et al., 1976;Chan et al., 1983) Malaysia, Hong Kong (Simons et al., 1977), China (Cui & Lin, 1982) and California (Jing et al., 1977). In Singapore, the associations among Chinese NPC were with A2 BW46 among older onset NPC and AW19 B17 among total NPC but this association was especially strong among younger onset NPC (Chan et al., 1983). Among the mid-incidence populations, reports of HLA associations have been inconsistent. HLA-A29 was associated with NPC among East African Blacks (RR= 19.5;Hall et al., 1982). Among North African NPC, the frequencies of A3, B5 and B15 were increased and AW33, B14 and DR4 decreased (Herait et al., 1983) and yet in another North African population, the Tunisians, no HLA association was found (Betuel et al., 1975). Among low incidence populations, reports of HLA associations in NFC were equally mixed. An Correspondence: S.H. Chan. Received 3 October 1984;and in revised form 5 December 1984. increase in the frequencies of A3 and B5 was observed among German NPC (Kruger et al., 1981), A3 among NPC from Australia (Simons & Shanmugaratnam, 1982) while no associations was reported in other Caucasian populations (Jing et al., 1977;Moore et al., 1983;Beigel et al., 1983). In South East Asia, besides the high NPC incidence Chinese population, there are ethnic groups with mid range incidence of NPC and here we report that results of HLA typing in NPC from one such ethnic group, the Malays.

Materials and methods
Subjects A total of 45 unrelated Malay NPC patients and 167 unrelated Malay normal controls was studied. The patients attended the Ear, Nose and Throat Departments of the Singapore General Hospital, Singapore and the University Hospital, Kuala Lumpur. Most of the patients were bled within 1.5 years after diagnosis but some had survived for varying periods (1.5 to 8 years) at the time of bleeding. The patients could be divided into early (<30 years of age) onset or late (>30 years) onset groups at the time of diagnosis. There were 8 early and 37 late onset patients. The controls were normal Malays and consisted of 106 Malays reported previously (Chan et al., 1978) and an additional 61 who were the normal spouses of Malay renal patients (live transplantation program) from Singapore and Malaysia.

HLA typing
Lymphocyte separation and HLA typing were performed from freshly drawn venous blood as discussed previously (Simons et al., 1976). The panel of antisera used to define the 26 locus A and B antigens consisted of over 200 sera of Malay, Chinese, Japanese, Filipino and Caucasian in origin. Antigens of the C and DR locus were not typed for. The significance of the difference in antigen frequencies between patients and controls was calculated by the Fisher's exact P-value and this value was multiplied by the number of antigens typed for (26) to give the corrected P(Pc) value. The relative risk (RR) was calculated by the odds ratio in the classical 2x2 table (Breslow & Day, 1980). If a, b, c, d represented patients with marker, patients without marker, controls with marker and controls without marker respectively, the relative risk was calculated as a xd divided by b xc.

Results
The HLA locus A and B antigen frequencies of the total NPC patients, patient subgroups and normal controls are shown in Table I However the corrected P value just failed to reach significance. The frequencies of B40 and B15 were significantly lower among NPC patients compared to controls but both these differences were not significant when corrected.
Old vs young onset NPC In Chinese NPC, there was differences in the HLA associations in early and late onset patients. BW46 was associated only with late onset (> 30 years) patients whereas B 17 was associated in total patients but particularly in early onset (_ 30 years) patients (Chan et al., 1983). In this study, early onset (_ 30 years) Malay patients had higher frequencies of A2, A9 and B 17 and lower frequencies of All compared to late onset patients or to normals (Table I). It is interesting to note that the finding of a higher frequency of B17 among early onset patients observed among the Chinese was also present in the Malays. The frequency of B17 was higher in early onset patients and this reflected in the higher relative risk (5.0; Table III) compared to total patients (RR= 3.4). HLA-A1 1 was not detected among early onset patients and the decreased frequency of B40 seen in total patients was even more marked among the late onset patients (P = 0.009; RR = 0.2; Table III).

Discussion
This study indicates HLA associations in Malay NPC. The associations were with B18 and B17. The association with B 18 was significant even after correction for the number of antigens typed for but with Bl7 the corrected P value just failed to be significant. B1 7 is also one of the antigens associated with Chinese NPC. Among Chinese NPC the frequency of B17 was lower in long term survivors compared to newly diagnosed patients (Chan et al., 1981). Unfortunately in this study we did not have sufficient patients to make this comparison and the frequency of B17 could be even higher if only newly diagnosed patients were studied. Among the Chinese, the frequency of B17 was higher in early onset patients and this study showed a similar relationship in Malay NPC. The frequency of B40 and B15 were lower among the patients compared to controls and the frequency of B40 was even slightly lower in late onset patients. However it could not be determined whether there was a specific decrease in these antigen frequencies or whether the decrease was a compensation for the increases in the frequencies of B18 and B17.
The present findings supported the hypothesis that genes are important in the pathogenesis of NPC and some of these genes are associated with the major histocompatibility complex. There are certain similar findings between Malay and Chinese NPCthe association with B17, the high frequency of B17 among early onset compared to late onset antigen frequencies in NPC patient subgroups and controls patients and the decrease frequency of Al 1. This closeness may be due in part to a strong belief that the Malays originated from the Yunan Province in China between 2500-1500BC as evidenced by their quadrangular adze culture and unglazed cordmarked pottery which have been traced from China southwards (Omar, 1983). However these findings will have to be confirmed by a larger study.