Lymphoma presenting as "Idiopathic" juvenile osteoporosis.

two cases of recurrent ulcer within a few months of H.S.V. in a small series, and 22% of Kronborg and Madsen's (1975) 50 patients developed recurrent ulceration within four years of H.S.V. The early postoperative insulin test, however, showed that 58% of Kronborg and Madsen's patients and 28% of those of Wastell et al., had incomplete vagotomy. Since there is a good correlation between insulin positivity soon after vago-tomy and subsequent recurrent ulceration (Bell et al. it is perhaps not surprising that the incidence of recurrent ulcera-tion after H.S.V. in these series was high. Hence our conclusion-that the addition of antrectomy to vagotomy is unnecessary when patients with hypersecretion of acid are treated by H.S.V.-is valid only if the H.S.V. is shown to be complete either at the time of operation or on insulin testing soon after operation. In view of the greater risk to life and the long-term disadvantages of antrectomy we think it best to avoid antrectomy but to spare no effect to ensure that the vagotomy of the parietal cell mass is complete at operation. Using operative techniques described previously (Johnston and Wilkinson, 1970; Goligher, 1974) we kept the incidence of incomplete vagotomy at 7%. The ability of junior members of staff, who nowperformmanyoftheH.S.V. operations, to achievea complete vagotomy differs little from that of the consultants. Hence the technical aspects of this operation are well within the capability of any surgeon who is prepared to study the technique and begin using H.S.V. cautiously in slim patients. It is important that the completeness of the parietal cell vagotomy should be monitored. Such a "quality control" is provided after operation by the insulin test. If this test gives positive results in no more than 10-15% of patients probably nothing more need be done because the incidence of recurrent ulceration should be low. Ifthe incidence ofincomplete vagotomy is higher, however, the operative technique should be reassessed and consideration given to the use of some intraoperative test. Such tests have not secured wide acceptance, but they are unquestionably a more rational form of quality control than the insulin test, because if the vagotomy is incomplete a determined effort can be made to complete it, and the outcome of the sur-geon's endeavours can be assessed once more. Such intraopera-tive tests are the Burge test (Burge and Vane, 1958; Burge and Frohn, 1969) and the Grassi test (Grassi, 1971). It is better to …