70-80% NPC diagnosed in Singapore is WHO-Type III (undifferentiated carcinoma), pathogenesis of this is intimately related to the Epstein-Barr virus (EBV). This relationship allows potential early detection of the tumour by tracing the virus. Although EBV IgA serology has been established as an effective marker for NPC, it remains unclear how useful or cost-effective it is as a screening test.

In local studies (Low et al. Otolaryngol Head Neck Surg 2000;123:505-7), positive early antigen (EA) serology result was found in 81.2% of NPC patients and in none of the controls. Negative EA and viral capsid antigen (VCA) serology results were present in 2.7% of NPC patients and in 46.8% of controls. Negative EA and positive VCA serology results were found in 30.0% of NPC patients with early disease, 7.8% of NPC patients with advanced disease, and 53.2% of controls. Serology for VCA, although highly sensitive, has an unacceptably high false-positive rate, and its cost-efficacy as an universal screening test is questionable.

Given its high specificity, serology for EA may be claimed as a useful screening test. In a large study of nearly 10,000 people in Taiwan (Chien et al. NEJM 2001;345:1877), the presence of VCA-IgA and anti-EBV DNAase antibodies were found to be strong predictors of the risk of NPC. The study claimed a risk was 32.8 for subjects with both markers and 4.0 for subjects with one marker as compared with subjects with neither marker. However these tumour may develop up to 5 years later or not develop at all (only 0.6% of those with either marker developed NPC). Furthermore, 60% of the detected cancer in this large group did not have any EBV marker present. Hence such tests if done are likely to cause unwarranted concern and cost in a huge population especially as 12% of the population in this study had EBV-DNAase and 1.2% were VCA +ve.

Improved testing based on virus genetic imprint in the blood or nasal secretions or biopsy specimen have been looked at in small studies in Singapore, Taiwan, Hong Kong, China and Canada. Though preliminary small studies look very interesting, a large population-based screening study may prove these too to be false harbingers. Hence, till large studies confirm definite utility of any screening test, various screening tests for NPC should only be offered in a research setting after adequate counselling -- as negative results may be falsely re-assuring; and positive results may add a needless health care cost and anxiety.