1211 Incidence of NPC 1976-85 : Auckland, NZ, men. adjusted incidence rate for NPC intermediate between the rates for Europeans and Chinese. Using data from the Auckland regional cancer registry for 1976-85 and the Auckland population census of 1981 we computed age-specific incidence rates for each racial group. NPC patients not normally resident in New Zealand were excluded. There were 66 cases of NPC: The age-specific incidence rates for men are shown in the figure. There is a gradient from Europeans through Pacific Island Polynesian to Chinese throughout the age range. Further analyses are reported elsewhere .5 Few of the Polynesians in this report had been born in New Zealand, the migrants generally retaining the social and dietary patterns of the Pacific Islands. Exposure to salted fish in the diet in childhood is considered important in the genesis of NPC in Chinese;7 this could also hold true for the Polynesians since this is popular food in the islands. An early study of NPC in New Zealand yielded, as expected, crude incidence rates much higher for Chinese than Europeans, but the number of Polynesians with NPC was far greater than expected, given the small Polynesian population in New Zealand at the time. In retrospect, it appears that evidence for a high risk of NPC in Polynesians has been around for some time. The migrant Polynesian in New Zealand possibly provides us with a unique opportunity to study racial and environmental factors in the genesis of this intriguing disease. Department of Otolaryngology, Green Lane Hospital, Auckland 3, New Zealand RANDALL P. MORTON Department of Oncology and Radiology, Auckland Hospital C. S. BENJAMIN 1. Green RC. Adaption and change in Maori culture. In: Kuschel G, ed Ecology and biogeography in New Zealand. The Hague: W. Junk, 1974: 1-13. 2. Muir CS. Nasopharyngeal carcinoma in non-Chinese populations. In: Biggs PM, de-Thé G, Payne LN, eds. Oncogenesis and herpes viruses. (IARC Sci Publ no 2). Lyon: IARC, 1972: 367-71. 3. Waterhouse J, Muir C, Shanmugaramam K, Powell J, eds. Cancer incidence in five continents. Vol IV. (IARC Sci Publ no 42.) Lyon: IARC, 1982. 4. Boyd JT, Doll R, Gurd CH. Cancer incidence in Fiji. Int J Epidemiol 1973; 2: 177-87. 5 Morton RP, Benjamin CS. Nasopharyngeal carcinoma in Auckland: a study of racial factors. Aust Radiol (in press). 6 Ho HC Current knowledge of the epidemiology of nasopharyngeal carcinoma a review. In: Biggs PM, de-Thé G, Payne LN, eds. Oncogenesis and herpesviruses. (IARC Sci Publ no 2). Lyon: IARC, 1972: 357-65. 7. Salas J. A survey of post-nasal cancer in New Zealand. NZ Med J 1962; 75: 156-59. CROSS-REACTIVE PROTEIN IN DUCHENNE MUSCLE SIR,—Increased knowledge of the genetic and biochemical basis of Duchenne/Becker muscular dystrophy has led to improvements in diagnosis through DNA and protein-based analyses. Indeed, a prognosis can now be accurately given in the absence of a family history by assays of dystrophin on small amounts of muscle coupled with an astute clinical assessment.l-6 So far, published studies have used antisera to the mid-portion (rod domain) of the large dystrophin protein, and all have indicated that dystrophin deficiency (less than 3%) is specific for, and diagnostic of, severe Duchenne dystrophy. We have raised five additional polyclonal antisera against non-overlapping regions of human dystrophin (Koenig M, unpublished). These antisera, together with our well-characterised mouse 30 kD and 60 kD antisera,1-4.6 recognise antigenic sites representative of most of the entire 400 kD dystrophin protein. To extend our analysis of dystrophin abnormalities in patients with neuromuscular disease, six of the seven antisera (1-2,30 kD, 60 kD, 9, 10, and 11) were affinity purified with the respective antigen covalently coupled to ’Affigel 10’ (Biorad). Anti-trpE antibodies were then immunoabsorbed from each affinity-purified antibody preparation, such that each preparation was specific for a distinct region of the dystrophin protein. These preparations were then tested on muscle with both immunofluorescent and immunoblot techniques. When used for immunofluorescent detection of dystrophin in cryostat sections of muscle, all antibody preparations replicated the previously published results with the 30 kD and 60 kD antibodies: the characteristic peripheral immunostaining of dystrophin was deficient in Duchenne muscular dystrophy (DMD) (fig 1). However, immunoblot studies gave equivocal results: a protein of 400 kD was detected in muscle from DMD patients by antibodies 1-2 and 10 but not by 30 kD, 60 kD, 9, and 11. Two possible explanations for his paradoxical finding are that the 400 kD protein represents a hitherto uncharacterised product of the DMD gene or that it is an unrelated protein of the same apparent molecular weight as dystrophin recognised specifically by antibodies 1-2 and 10. To distinguish between these two possibilities, we identified a DMD patient who had a deletion which resulted in the absence of all detectable exons of the dystrophin gene (fig 2). The muscle of this patient was tested by immunoblot analysis with the six antibody preparations. As with other Duchenne muscle specimens, a protein which co-migrated with dystrophin was detected with antibodies 1-2 and 10, but not with 60 kD, 30 kD, 9, or 11 (fig 3). Since this patient lacks the entire dystrophin gene, yet can still produce the 400 kD protein, we conclude that this protein is not a product of the dystrophin gene Fig 1-Immunofluorescence analysis of muscle from a limb-girdle patient (A) and patient with DMD (B). Peripheral dystrophin imrnunostaming is detected in (A) but not (B). The antibody preparation used was affmity-purified "anti-10", raised against carboxy terminus of dystrophin (DMD muscle in [B] is from same patient illustrated in figs 2 and 3.)