In 50 consecutive patients with Graves' disease treated with PTU, 7 (group 1) developed increasing goitre in spite of unmeasurable TSH. Thyroid variables were compared with those from 10 controls with an ordinary response to PTU (group 2). Serum T4 decreased in group 1 from 246 +/- 47 nmol/l (mean +/- SD) to 40 +/- 9 nmol/l after 6 weeks of PTU treatment and continued to be below the normal range during the next 4 months. In group 2 serum T4 decreased from 190 +/- 35 to 88 +/- 47 nmol/l and stayed in the normal range. Serum T3 was normalized in both groups after 6 weeks but increased to values above the normal range in group 1 after that time. In spite of unmeasurable TSH during the 6 months of treatment in group 1, thyroid volume, determined ultrasonically, increased significantly from 60 +/- 29 to 93 +/- 68 ml (P less than 0.05), but was unaltered in group 2 about 25 ml. Thyroid stimulating antibodies (TSAb) measured by adenylate cyclase activation (normal below 109%) decreased in group 2 from 117 +/- 23 to 90 +/- 17% (P less than 0.01) (6 months of therapy), but increased significantly in group 1, from 201 +/- 47% to a maximum value of 234 +/- 69% (P less than 0.05). TSH binding inhibitory immunoglobulins (TBII) (given as per cent inhibition, normal below 26%) decreased in group 2 from 43 +/- 29 to 29 +/- 27% (P less than 0.05) but were unaltered high in group 1, 66 +/- 25% before therapy and 57 +/- 26% after 6 months of therapy.(ABSTRACT TRUNCATED AT 250 WORDS)

Brain scintigraphy with Tc99-pertechnetate (Tc99-scan) was performed 4 times in 95 consecutive stroke patients: on average 5 days, 18 days, 103 days and 194 days after the stroke. The type (infarct, hematoma), size and localisation of the lesion was evaluated by CT-scan performed 3 times in all: on average 3 days, 18 days and 207 days after the stroke. The detection rate of Tc99-scan was 31% on day 5, 47% on day 18, 18% on day 103 and 9% on day 194. The detection rate was however highly dependent upon: (i) the localisation of the lesion, i.e. superficially localised lesions were nearly always identified (90%) while infarcts localised deep in the hemisphere were identified in only 20% of the patients; (ii) the size of the lesion, i.e. large deep infarcts were seen with a much higher frequency than small deep infarcts. The detection rate of the CT-scan was practically not dependent upon the time of investigation: 63% on day 3, 70% on day 18, 69% on day 207. The detection rate of the Tc99-scan and CT-scan was nearly identical for large superficially localised lesions 2-3 weeks after the stroke. In this period, however, CT-scan identified 30 deep infarcts. Only 12 of these were identified by Tc99-scan.