Blood levels of prolactin are consistently raised in women who have undergone mastectomy for breast cancer, probably because of the stress of surgery. Since this increase in concentration of a known breast epithelial growth promoter might stimulate proliferation in micrometastatic cells shed at the time of surgery, a pilot study was conducted to try to abolish this effect. 38 patients with suspected operable breast cancer were given either bromocriptine (18) or placebo (20) tablets for 5 days preoperatively and thereafter for 3-10 days. Bromocriptine-treated patients showed significant reductions in prolactin levels and in S-phase fraction of tumour cells within the primary infiltrating carcinoma. There were no major side-effects. Perioperative bromocriptine may provide another approach to adjuvant therapy, possibly combined with endocrine or cytotoxic treatment for patients with axillary nodal metastases.

In 473 consecutive febrile patients a sensitive and rapid chromogenic limulus assay was used to assess the value of endotoxaemia versus bacteraemia for predicting development of the syndrome of septicaemia. In each patient three blood specimens for culture and endotoxin testing were obtained at the onset of fever. Blood pressure, urinary output, and the occurrence of thrombocytopenia and metabolic acidosis were recorded prospectively during three days of follow-up. Septicaemia developed in 19 patients (4%). The sensitivity, specificity, and likelihood ratio for a positive result with the endotoxin assay were 79%, 96%, and 20, respectively. The corresponding indices for bacteraemia were 89%, 78%, and only 4. The results suggest that endotoxaemia is a clinically valid indicator for impending gram-negative septicaemia (positive predictive value 48%) and that the absence of endotoxaemia virtually rules out the risk that septicaemia will ensue (negative predictive value 99%).

Patients with migraine who believed that red wine but not alcohol in general had a headache-provoking effect on them were challenged either with red wine or with a vodka and diluent mixture of equivalent alcohol content, both consumed cold out of dark bottles to disguise colour and flavour. The red wine, which had a negligible tyramine content, provoked a typical migraine attack in 9 of 11 such patients, whereas none of the 8 challenged with vodka had an attack. Neither red wine nor vodka provoked such episodes in other migrainous subjects or controls. These findings show that red wine contains a migraine-provoking agent that is neither alcohol nor tyramine.

To investigate whether cessation of regular beta-agonist treatment results in an increase in bronchial responsiveness, two double-blind, randomised crossover studies were done. Subjects with mild asthma were investigated to determine the course of change in bronchial responsiveness, measured as the provocative dose (PD20) of histamine that caused a 20% fall in forced expiratory volume in 1 s after short-term and longer term treatment with an inhaled beta-agonist. In the first study in 8 subjects, 500 and 2000 micrograms terbutaline thrice in 1 day protected against histamine-induced bronchoconstriction, and the increase in PD20 compared with placebo remained high throughout the day and overnight. In the second study 8 subjects received placebo or terbutaline 750 micrograms thrice daily for 14 days. The protection afforded by terbutaline against histamine-induced bronchoconstriction on day 14 was less than that on day 1 by 40% in the morning and 82% in the afternoon. On day 15 PD20 was lower after stopping terbutaline than placebo, with a maximum difference of 1.5 (95% CI 0.6-2.5) doubling-doses of histamine 23 h after the end of treatment. Thus treatment with terbutaline for 1 day did not result in any rebound increase in bronchial responsiveness. Treatment for 2 weeks impaired the ability of terbutaline to protect against histamine-induced bronchoconstriction, and was followed by a rebound increase in bronchial responsiveness after cessation of treatment.

A randomised double-blind placebo-controlled clinical trial was undertaken to assess survival after a single 30 unit intravenous dose of anisoylated plasminogen streptokinase activator complex (APSAC) within 6 h of onset of acute myocardial infarction. A planned interim analysis of mortality data was undertaken after 1004 patients had been followed for at least 30 days after randomisation. 61 (12.2%) of 502 patients on placebo died within 30 days compared with 32 (6.4%) of 502 patients on APSAC (p = 0.0016). Because of this 47% reduction in 30-day mortality (95% confidence interval 21-65%) patient entry to the trial was terminated. Preliminary 1 year mortality data show a similar trend. Percentage mortality reduction with APSAC was similar whether time since onset of symptoms was 0-4 h (660 patients) or 4-6 h (344 patients). Adverse events were few. These findings add to evidence that intravenous thrombolytic therapy after myocardial infarction offers a substantial reduction in mortality.

A double-blind comparison of clonidine and lofexidine in the management of the opioid withdrawal syndrome encountered several difficulties of a non-scientific nature which eventually forced termination of the trial. There were, for example, problems with research beds, self-discharges, staff attitudes, access to haematological monitoring, and worries about exposure to staff to human immunodeficiency virus. The trial collapsed when the manufacturers stopped making lofexidine on commercial grounds and the charitable foundation which was backing the trial withdrew its support. Some of the insecurities of medical research may have little to do with scientific merit.

433 patients aged 29-80 with mild to moderate heart failure entered a multicentre double-blind randomised between-patient comparison of xamoterol 200 mg twice daily, digoxin 0.125 mg twice daily, and placebo. Patients were assessed at baseline and after three months. Of 349 who completed the double-blind phase, 300 had valid exercise tests. Compared with placebo, xamoterol significantly increased exercise duration and work done on a bicycle ergometer and improved breathlessness and tiredness during daily life as assessed by visual analogue scale and by Likert scale. Digoxin showed no statistically significant advantage over placebo on any of the measures except the Likert scale. Exercise performance and work done were significantly higher with xamoterol than with digoxin.

Growth was assessed in 82 children with acute lymphoblastic leukaemia (ALL) who achieved complete continuous first remission following treatment. 34 received prophylactic cranial irradiation at a total dose over 2000 cGy (group I) and 48 at a dose of 1800 cGy (group II). Chemotherapy was given over two or three years and was more intense in group II. Both groups showed a similar significant decrease in height standard deviation score (SDS) over four years (group I -0.31, group II -0.39). 15 children in group I were followed to ten years and continued to show restricted growth with a mean height SDS decrease of -0.84 (range 0 to -1.7). The greatest reduction in yearly decrements in height SDS occurred in the first year after diagnosis. In both groups height SDS increased significantly on completion of chemotherapy. Thus chemotherapy protocols contribute to growth retardation in ALL. In most of these children the mean loss of height over ten years was not sufficiently great to justify long-term growth hormone (GH) therapy.

11 patients having major gastrointestinal surgery were allocated at random to receive either biosynthetic human growth hormone (BSHGH) 0.1 mg/kg or placebo daily for the first 7 postoperative days. All patients received the same intravenous feeding regimen, which contained 2.09 MJ glucose, 1.88 MJ fat, and 7 g N daily. Patients receiving BSHGH were in positive nitrogen balance throughout the study (mean 1.8 [SEM 0.4] g N/day) and those receiving placebo were in negative nitrogen balance (mean -0.9 [0.7] g N/day). Resting energy expenditure progressively increased in the patients receiving BSHGH (115.7% [14.8] on day 7) but remained unchanged in patients receiving placebo (99.35% [1.4]). Fat oxidation was nearly three times higher in the patients on BSHGH (4.09 [0.38] MJ/day) than in controls (1.38 [0.50]). Carbohydrate oxidation remained about the same in both groups. Whole-body protein turnover, synthesis, and breakdown were increased in the patients receiving growth hormone.

The effect of a 7-day course of intravenous penicillin (6 million units/day) on severe, advanced leptospirosis was examined in a randomised, placebo-controlled, double-blind trial involving 42 patients. Every measurable aspect of the disease was favourably affected by penicillin. Fever lasted more than twice as long in the placebo group (11.6 [SD 8.34] days vs 4.7 [4.19] days, p less than 0.005), and by the fourth day after starting penicillin more than half the treatment group, but only 1 of 19 in the placebo group, were afebrile (p less than 0.005). Creatinine rises persisted more than thrice as long in the patients receiving only placebo (8.3 [8.46] days vs 2.7 [1.90] days; p less than 0.01). Penicillin also shortened the hospital stay and prevented leptospiruria. Intravenous penicillin should be given to patients with severe leptospirosis, even if therapy can be begun only late in the course of their disease.

In February to July, 1986, an outbreak of type 3 poliomyelitis occurred in north-east Brazil that was linked to type-specific failure of trivalent oral polio vaccine (TOPV). To see if alternative vaccines would improve seroconversion to type 3, 441 children less than 5 years of age who had previously received no or up to four doses of TOPV were randomly assigned to receive one dose of standard TOPV (1,000,000, 100,000, and 300,000 median tissue culture infection doses [TCID50] of types 1, 2, and 3, respectively); a new formulation of TOPV containing twice the dosage of type 3 (600,000 TCID50); or a monovalent vaccine containing 300,000 TCID50 of type 3. While rates of seroconversion to types 1 or 2 were equivalent following vaccination with either formulation of TOPV, children who received the new formulation were 2.7 times more likely to seroconvert to type 3. Similar differences for type 3 were observed when monovalent vaccine was compared with standard TOPV, though both groups had received the same dose of type 3 antigen. The low rate of seroconversion to type 3 in the standard TOPV group was associated with a higher rate of reinfection with type 2, which also appeared to interfere to some extent with seroconversion to type 1. These findings extend earlier observations that interference from Sabin type 2 virus may be an important contributory cause of type-specific TOPV failure, and suggest that interference can be overcome with alterations in the formulation.