To determine the indications for postoperative radiotherapy after surgical resection of a nonfunctioning pituitary macroadenoma. A retrospective chart review of 72 patients with histologically proven chromophobe adenoma who presented for pituitary surgery between January 1985 and June 1998, with a minimum follow-up period of 12 months. The study endpoint was tumour recurrence or progression detected either by routine follow-up imaging or by clinical progression with subsequent confirmation by imaging. A proportional hazards model was used to determine independent prognostic factors. Mean follow-up was 64 months. In the radiotherapy group 13 of 50 recurred (or progressed) (26%), while in the nonradiotherapy group 10 of 22 recurred (46%), logrank test, P = 0.025. In patients assessed as having complete excision of tumour (n = 20) only two recurred (10%), both in patients without radiotherapy. No further treatment has been required in either case to date. In patients with residual tumour (n = 52), 41 had radiotherapy with 13 recurrences (32%), while 11 patients had no radiotherapy with eight subsequent recurrences (73%); logrank test, P = 0.007. Further treatment has been required in the majority of these cases. Cox's proportional hazards model analysis showed that only complete tumour removal and postoperative radiotherapy were independent favourable prognostic factors. The goal of surgery should be complete surgical excision where possible. The risk of recurrence in patients with no residual tumour on postoperative imaging is low enough to justify withholding routine postoperative radiotherapy in this group. In patients with residual tumour, conventional external beam radiotherapy administered within 12 months of surgery is effective at reducing recurrence or progression.

A new test of olfactory function, the combined olfactory test, has been designed to assess odours easily recognizable by the test population. The test consists of an odour recognition test of nine odours, where an odour in a bottle is chosen from a list of four possible odours in a forced choice manner. This is followed by a threshold test using a series of three-fold dilutions of 1-butanol. The mean of the two scores is the combined olfactory score. The test was subjected to a validation study. It was performed on 133 participants with a normal sense of smell and a normal rhinological examination and on 94 participants who said that they did not have a sense of smell. There was a highly significant difference between the combined olfactory score in the normal and 'anosmic' groups (P < 0.001). This significant difference was the same between the two groups for the threshold and odour recognition arms of the test. There was a highly significant difference (P < 0.001) between the two subgroups of 'completely anosmic' and 'almost anosmic' participants, indicating that the test could grade the degree of olfactory dysfunction.