Expression of beta-2-microglobulin by nasopharyngeal carcinoma.

Serum beta-2-microglobulin (beta 2M) levels of 274 Chinese patients with different stages of nasopharyngeal carcinoma at presentation and that of 35 patients who developed distant metastases post-treatment were assayed. beta 2M level was found to increase with advancing stage of disease, with statistically significant differences among early-stage, advanced-stage, and metastatic disease. Elevated pre-treatment beta 2M levels were expressed more frequently by tumours with lower degree of histological differentiation. The sensitivity of serum beta 2M for diagnosis of nasopharyngeal carcinoma, however, is low.

Beta-2-microglobulin (132M) is a protein of low molecular weight (11,800 daltons). It was first isolated from urine in patients with Wilson's disease in 1968 (Berggard et al., 1968). It is found on the cell membrane of all nucleated cells and platelets and it forms the light chain moiety of the major histocompatibility antigens. Cell membrane turnover is the principle source of P2M in blood, plasma and body fluids (Cresswell et al., 1974;Forman, 1982). Elevated serum levels has been found to be associated with increasing age, renal impairment (Bailey et al., 1978) and a variety of malignancies and appears to be a reflection of tumour load in patients with lymphoma (Anderson et al., 1983;Hagberg et al., 1983), myeloma (Child et al., 1983;Norfolk et al., 1980;Alexanian et al., 1985), lung cancer (Schweiger & Tocsanyi, 1978), breast cancer (Teasdal et al., 1977;Rashid et al., 1980), and squamous cell carcinoma of the head and neck (Wennerberg et al., 1984). Nasopharyngeal carcinoma (NPC) is a commonly-occurring tumour in Hong Kong. The value of P2M as a diagnostic marker for NPC and its correlation with tumour load is the subject of the present investigation.

Method and materials
Serum samples of 73 healthy volunteers were used to establish a normal reference range for P2M. Serum samples of 274 Chinese patients with NPC were collected at presentation and/or at follow-up and stored at -70°until assayed. The characteristics of the patients are shown in Table I, and the distribution of histologies in Table II. Serum P2M level was measured by radioimmunoassay (Pharmacia beta-2-micro, RIA, Sweden). Serum creatinine level was routinely checked to exclude renal impairment. It was found to be normal for all the patients at the time of sample collection. The Student t-test was used for statistical analysis.

Results
Normal reference range of P2M A reference range of serum P2M levels in normal subjects was established at 0.96-1.88 mg 1-l (mean +/-2 S.D.). For the purpose of the present study, an arbitrary cut-off value of 2 mg 1' was adopted ( Figure 1).

Serum P2M levels and staging of NPC
There was a trend for progressive increase in mean serum P2M levels and in the percentage of patients with elevated P2M levels with advancing stage of disease ( Figure 1 and Table I). By grouping patients into an 'early-stage' group, an 'advanced-stage' group and a 'metastatic' group, significant differences in mean levels and percentages of patients with elevated levels were obtained. The greatest difference in mean levels was found between metastatic NPC and other stages combined (p <0.001).
Sensitivity of P2Mfor diagnosis of NPC Using a reference upper normal limit of 2mgl 1, the sensitivity of serum P2M for detection of NPC for the whole group of NPC patients is 37%, the sensitivities for the subgroups are: Stage I 23%, Stage II 17%, Stage III 25%, Stage IV 36%, metastatic disease 96%. Serum 1B 2M levels and histological differentiation Table II shows that the percentage of patients with elevated pre-treatment levels of P2M increased with lower degree of histological differentiation, though the difference has not reached statistically significant levels.

Serum 132M levels and site of distant metastatic disease
The elevation in P2M level was expressed by metastases at a variety of sites including bone, lung, liver, bone-marrow, breast and skin, and thus was not dependent on the site of distant metastases.

Discussion
The mean serum P2M levels of the 73 Chinese healthy normals is very close to that reported by Lai et al. (Lai et al., 1986) and is significantly different from that of the NPC population. However, serum 132M appears to have little diagnostic value in view of the very low sensitivity of the test and availability of more sensitive serological markers such as the IgA titre to the viral capsid antigen of Epstein-Barr virus (Ho et al., 1976). Although P2M has a high sensitivity in metastatic disease, the clinical relevance in this situation is limited.
The expression of elevated P2M levels was found to be related to the histological differentiation in NPC. This may be due to more active cellular proliferation in tumours with poorer histological differentiation. Similar findings however were not evident in non-Hodgkin's lymphoma (Anderson et al., 1983).
The general increase of mean P2M levels from stage I to IV and metastatic disease is probably a reflection of increasing tumour load, although increasing intensities of an immune response and a direct effect of Epstein-Barr virus infection are alternative explanations. Stage-dependency of P2M levels has also been reported in non-Hodgkin's lymphoma (Ander-Correspondence: S.F. Leung, Department of Clinical Oncology,    (Teasdal et al., 1977) and myeloma (Hagberg et al., 1983;Alexanian et al., 1985), and is thought to be due to increased cell turnover (Karlsson et al., 1980) and/or augmented immune response by lymphocytes to the neoplasm (Forman, 1982;Rashid et al., 1980;Karlsson et al., 1980;Shuster et al., 1976).
In another study, however, the P2M level was lower in early and advanced stages than in intermediate stages (Lotzniker et al., 1988): the decrease with more advanced stages was attributed to a weakened immunologic response in advanced disease. NPC is known to be associated with certain immunologic alterations in the patient, including lymphopenia, reduced T4/T8 lymphocyte subset ratio (Cheng et al., 1989), impaired cell-mediated immune functions , and elevated antibody titres to the Epstein-Barr virus (Ho et al., 1976). Stage-dependency is not a feature of most of these immunological alterations except lymphopenia (Cheng et al., 1989) and the IgA titre to the viral capsid antigen of Epstein-Barr virus (Henle et al., 1973). There is no evidence, except on the contrary, to suggest an absolute increase in the total lymphocyte population or its subset in advancing stages of NPC. Thus the possibility of an in- levels in different stages of NPC is unlikely. The association between NPC and Epstein-Barr virus is well-established and demonstrable at serological, histopathological and genetic levels (Ho et al., 1976;Huang et al., 1974;Lung et al., 1990).
Raised P2M levels has also been found in patients with infectious mononucleosisan Epstein-Barr virus-related conditionand other herpes virus infections (Lamelin et al., 1983;Cooper et al., 1984;Norfolk et al., 1987). However, there is no known common mechanism to account for the raised levels of P2M in different viral infections. Neither is there evidence to prove that increased T-cell activation, which occurs in infectious mononucleosis, occurs in a com parable manner in NPC. A direct effect of the virus accounting for increasing P2M levels in advancing stages of NPC is thus difficult to substantiate. The segregation of three groups of NPC patients with significant differences in P2M levels in our study may provide a basis for staging patients based on tumour burden. It may assist the selection of subsets of patients with advanced-stage disease for more aggressive treatment with adjuvant chemotherapy. The validity of these statements would require proof of pre-treatment P2M level as an independent prognostic factor, and follow-up assessment of a larger patient population would be required for this purpose.