The occurrence of bone marrow carcinosis was investigated in 380 patients at the time of first recurrence of breast cancer. Results were related to results from radiographic bone survey, 99mTc MDP bone scintigraphy, clinical examination and serum alkaline phosphatase and serum calcium levels. Eighty-seven patients (23%) had tumor cells in the bone marrow. X-rays showed metastases in 78% of the patients with and in 16% of the patients without bone marrow carcinosis. The diagnostic efficiency of x-rays with bone marrow biopsy as the key diagnostic factor was 83%, and it was superior to that of other investigation methods. Bone tissue biopsies were positive alone in 15 patients (17%) and marrow aspirations were positive alone in seven patients (8%). Imprint preparations were positive alone in 7% of the patients and bone tissue biopsy in 5% of the patients. Heavy tumor infiltration (250%) of the bone marrow was associated with the occurrence of numerous regions of radiographically involved bone lesions and with histopathologic evidence of bone destruction. Furthermore, pronounced bone formation and marrow fibrosis were more commonly seen in patients with osteosclerotic bone metastases than in patients with osteolytic bone metastases. This study provides evidence that the primary soil of meta-static bone disease in human breast cancer is the bone marrow and that radiographic evidence of bone metastases is a result of an invasion and destruction of the bone tissue matrix by tumor cells from the marrow cavity. Cancer 60~306-1312, 1987. TAGING PROCEDURES for patients with first recurS rence of breast cancer are important prerequisites both for determining the clinical course to be followed and for establishing a thorough baseline evaluation before the onset of treatment.'-3 Moreover, the fact that most agents used in the treatment of recurrent breast cancer are myelosuppressive stresses the importance of evaluating the condition of the bone marrow before treatment is started.' Usually, x-rays and/or bone scin-tigraphy are used to evaluate the occurrence of bone metastases. However, histopathologic evidence of bone marrow carcinosis is not always associated with radio-graphic and scintigraphic abn~rmalities.~-' Previous clinical studies have shown a variation of 3% to 40% in the frequency of positive bone marrow exami-From the Departments of *Oncology ONA, ?Pathology, $Radiol-ogy, and §Clinical Physiology,

Estrogen replacement therapy retards the development of cardiovascular disease and osteoporosis in postmenopausal women. However, long-term unopposed use increases the risk of cancer in endometrium and possibly in breast. The racemic compound ormeloxifene, widely used in India as an antifertility agent, is a partial estrogen receptor agonist with antiosteoporotic properties. The present study was undertaken to investigate the effect of the L-enantiomer (levormeloxifene) and the d-enantiomer (d-ormeloxifene) on the development of atherosclerosis. In a short-term experiment (6 weeks), 4 x 10 ovariectomized female rabbits were fed a 0.25% cholesterol-enriched diet and the effect on plasma cholesterol levels was studied. In a long-term experiment (13 weeks), 4 x 15 ovariectomized female and 4 x 15 shamoperated male rabbits were maintained at a similar plasma cholesterol level of 25 mmol/L and the effect on undamaged and balloon-injured arterial wall was studied. In both experiments, the rabbits were treated with levormeloxifene, d-ormeloxifene, 17 beta-estradiol, or placebo, respectively. In the short-term experiment, levormeloxifene, in contrast to d-ormeloxifene, significantly reduced plasma cholesterol by 30% compared with the placebo group. In the long-term experiment, levormeloxifene, in contrast to d-ormeloxifene, significantly reduced atherosclerosis by 50% in the undamaged arterial wall of both female and male rabbits. Because these rabbits were cholesterol-clamped, the antiatherogenic effect was not mediated via plasma cholesterol lowering. Like estrogen, levormeloxifene did not inhibit atherosclerosis in the endothelium-denuded site of aorta. The antiatherogenic effects of levormeloxifene were thus similar to those of estrogen, but produced in the absence of any noticeable estrogenic effect on uterine or testicular tissue.

No abstract