A retrospective study was performed to evaluate the diagnostic yield for lung cancer from histological biopsy specimens and from washings and brushings for cytological examination taken at fibreoptic bronchoscopy. The Any patient with an inadequate follow up (because of lost records, for example) or with insufficient data was excluded from the study.The two most common reasons for bronchoscopy were an abnormal chest radiograph (493 cases) and haemoptysis (104 cases). Other reasons for bronchoscopy included dyspnoea, stridor, chronic cough, and hypercalcaemia of unknown cause.Because this was a restrospective study, all combinations of cytological and histological procedures were found to have been used. In most cases, however, washings, brushings, and biopsy specimens were taken, especially when a tumour was visible. The sequence when all three procedures were performed was always washing, biopsy, and then brushing immediately before extraction of the bronchoscope. Washings were obtained by lavage with 20-40 ml of normal saline and aspiration into a trap. No set number of biopsy specimens was taken.When no lesion was seen endoscopically, "blind" cytology was performed by lavaging as described and brushing the appropriate segment as determined by the posteroanterior and lateral chest radiographs. Occasionally "blind" biopsy was also performed, the biopsy forceps being directed into the appropriate segment.Brushings were smeared on to two to four slides and immediately fixed in 9500 alcohol. Washings were taken to the cytology laboratory and centrifuged at 1500 rev/min for five minutes, the supernatant was poured off, and the sediment of material was pipetted on to several slides and fixed with 95% alcohol.Cytological specimens were stained routinely by the Papanicolaou technique. Specimens were interpreted by the cytopathologist with-

SUMMARY Phaeochromocytomas were diagnosed in 72 patients in Hammersmith, Beljiast and Newcastle between 1955 and 1976. Fourteen

The control of severe hypoglycemic symptoms in patients with insulin‐secreting islet tumors (insulinoma) of the pancreas is an important aspect of management. A series of 24 patients with insulinomas whose ages ranged from 5 months to 73 years seen at one center over a period of 17 years is described. The presenting symptoms and the effect of these on their lifestyles are discussed. Diazoxide has been used to control hypoglycemic symptoms in 18 of the 24 patients. Fourteen patients have received diazoxide for varying periods of time prior to successful surgery and 4 have received long‐term therapy. Three of these patients are highlighted to show various aspects of benefit and problems from the use of diazoxide. The mean dosage of diazoxide was 400 mg/day and the mean duration of therapy was 10 months. The control of symptoms was good in 8, fair in 6, and poor in 4. The main side effects observed were hirsutism (56%), ankle edema (50%), weight gain (38%), and nausea (11%). These side effects were well tolerated and the benefits of therapy were felt to outweigh disadvantages. Side effects only necessitated discontinuation of diazoxide in 1 patient who developed a severe hypersensitivity reaction. Our results show that diazoxide has a role to play in the management of patients with insulinoma during preoperative tumor localization and can also be used when surgery is contraindicated or when laparotomy is unsuccessful.

No abstract

1. Twenty-four patients with primary hyperparathyroidism were studied before and 18 restudied 6.5 months (mean) after parathyroidectomy, to investigate the pathogenesis of the hypertension which may accompany this condition. Comparison was made with age-matched patients with essential hypertension and with normotensive control subjects. 2. There was a significant inverse relationship between mean arterial pressure and 51Cr-labelled ethylene-diaminetetra-acetate (51Cr-EDTA) clearance in patients with hyperparathyroidism both before and after parathyroidectomy, but not in patients with essential hypertension. 3. Creatinine clearance appeared to overestimate glomerular filtration rate in some patients with hyperparathyroidism, falling significantly after surgery while 51Cr-EDTA clearance was unchanged. This observation may explain the failure of some previous studies to relate hypertension to impairment of renal function. 4. Plasma renin activity, plasma aldosterone and whole-body exchangeable sodium did not differ between normotensive and hypertensive patients with primary hyperparathyroidism and were unchanged after surgery. 5. Parathyroidectomy did not result in any change in blood pressure or in glomerular filtration rate measured by 51Cr-EDTA clearance.