mm, but not even by in vitro examination of 84 resected lymph nodes could we find reliable signs for malignancy. Infiltration of the bron- chial wall and involvement of large vessels by mediastinal masses was frequently diagnosed. Involvement of the airways by vascular anoma- lies, pleural etliision or solid masses could be differentiated, Conclu- sions: In view of the fact that the procedure costs around $20. it seems feasible and we are currently examining its value in a prospective study.

Pulmonary thrombosis induced by anticancer chemotherapy in a patient with lung cancer Imae N, Suzuki Y, Fukuoka K. Second Departmenf Internal Medicine, NaraMedical Universi&, Kashihara. Jpn J Chest Dis 1996;55:368-73.

A 66-year-old woman was admitted to our hospital with a coin le- sion in the right upper lung field. The diagnosis of lung cancer (adeno- carcinoma) was established by transcutaneous lung biopsy guided by the computed tomography. A combined chemotherapy with cisplatin, vindesine and mitomycin C was started. On the 13th day, peripheral platelet count fell down and plasma level oftibrin degradation products (FDP) increased. On the 22nd day, a thrombus in the right main pul- monary artery was found accidentally by chest CT. Because the diagno- sis of pulmonary thrombosis was established by the pulmonary perfusion scintigraphy, the anticoagulant and thrombolytics (heparin and uroki- nase) were administered. Following the treatment complete disappear- ance of the thrombus was confirmed by chest CT and pulmonary perfusion scintigraphy. This cast report suggests that the anticancer chemotherapy for lung cancer may induce the hypercoagulopathy.

Prognostic value of angiogenesis in operable non-small cell lung cancer Giatromanolaki A, Koukourakis M, O’Byme K et al. Departmenf of Cellular Science, John Radclr@ Hospital, OxJbrd OX3 9DU. J Pathol 1996;179:80-8.

Tumour angiogenesis is an important factor for tumour growth and metastasis. Although some recent reports suggest that microvessel counts in non-small cell lung cancer are related to a poor disease outcome, the results were not conclusive and were not compared with other molecu- lar prognostic markers. In the present study, the vascular grade was assessed in 107 (T1,2-NO, I) operable non-small cell lung carcinomas, using the JC70 monoclonal antibody to CD3 1. Three vascular grades were defined with appraisal by eye and by Chalkley counting: high (Chalkley score 7-12). medium (5-6), and low (2-4). There was a sig- nificant correlation between eye appraisal and Chalkley counting (P < 0.0001). Vascular grade was not related to histology, grade, prolifera- tion index (Ki67), or EGFR or ~53 expression. Tumours from younger patients had a higher grade of angiogenesis (P = 0.05). Apart from the vascular grade, none of the other factors examined was statistically related to lymph node metastasis (P < 0.0001). A univariate analysis of survival showed that vascular grade was the most significant prognos- tic factor (P = 0.0004), followed by N-stage (P = 0.00 1). In a multivariate analysis, N-stage and vascular grade were not found to be independent prognostic factors, since they were strongly related to each other. Ex- cluding N-stage, vascular grade was the only independent prognostic factor (P = 0.007). Kaplan-Meier survival curves showed a statistically significant worse prognosis for patients with high vascular grade, but no difference was observed between low and medium vascular grade These data suggest that angiogenesis in operable non-small cell lung cancer is a major prognostic factor for sutvival and, among the param- eters tested, is the only factor related to cancer cell migration to lymph nodes. The integration ofvascular grading in clinical trials on adjuvant chemotherapy and/or radiotherapy could substantially contribute in defining groups of operable patients who might benetit from cytotoxic treatment.

Prediction of survival and recurrence by serum and cytosolic kvels of CPA, CA125 and SCC antigens in resectable non- small-cell lung cancer Diez M, TORTE A, Maestro ML et al. Clrugra General, HospitalPrincipe de Asturias. Universidad de Alcala de Henares, 28805 Madrid. Br J Cancer 1996;73: 1248-54.

Risk ofdeath and risk of recurrence in 108 potentially curable non- small-cell lung cancer patients were analysed with respect to TNM stage, histological type and carcinoembryonic antigen (CEA), CA125 anti- gen and squamous cell carcinoma antigen (SCC) levels in serum and CytosOl. CA125 and CEA levels were closely related to outcome fig- ures. Multivariate analyses indicated that TNM stage and histological type had the best predictive power, but serum and cytosolic CA125 and serum CEA contained additional, independent prognostic information. Predictive information drawn from serum and cytosolic levels proved mutually complementary. We conclude that CA125 and CEA comple- ment TNM classification and histological type for the purpose of quan- tifying risk of death or recurrence.

Lung cancer with visual loss Ing EB, Augsburger JJ, Eagle RC. Depf ojOphfhahnology. Toronto Hospital, 399 Bathurst St., Toronto, Onl. bf57’ 2S8. Sun, Ophthamol 1996;40:505-IO.

A 58-year-old man with primary large cell carcinoma of the lung presented with rapidly progressive, bilateral visual loss. The patient was left and oriented, had no complaints of headache, and was found to have full ocular motility. The optic disks and fundi appeared normal. No visual pathway lesions or other CNS abnormalities were detected on neuro-imaging. The authors discuss their differential diagnosis, clini- cal diagnosis approach, and subsequent management of this unusual patient.

Squamous cell carcinoma antigen (SccAg) in the diagnosis of lung cancer Chen Q, Zhou X, Huang HR Respirafoy Deparftnenf, Shanghai hrsl People k Hosprfal, Shanghai. Chin J Clin Oncol 1996;23:240-2.

The serum SccAg levels were measured by immunoradiometric as- say in 45 patients with lung cancer, 4 1 with other lung diseases. Serum SccAg was above 1.5 rig/ml in 60% of patients with lung cancer and 14.6% of those with other lung diseases (P < 0.01); 89.5% with lung

squamous cell carcinoma and 16.7% with lung adenocarcinoma (P < 0.05). In cases of lung cancer, the sensitivity rate was 60% (89 5% for squamous cell carcinoma) and specificity rate 85.4%. In lung cancer patients, the positivity rate of SccAg increased with progression of clini- cal stages (50%, 64.7%. 82.4% for stage II, III. IV respectively). It is concluded that SccAg is useful in diagnosis of lung cancer, especially of squamous cell carcinoma and could be even more helpful provided being used in combination with other tumour markers and clinical lind- ings as well.

Analysis of serum CYFRZl-1 by immunoradiometric assay in patients with non-small cell lung cancer Guo HB, Klingmuller D, Bidlingmaier F. Deparm!mf ojClin!col Bfo- chemrstyM Bengbu Medical College, Bengbu 233003. Chin J Clin Oncol 1996;23 1236-9.

The fragments of serum cytokeratin 19 subunit, referred to as CYFRAZI-1 were determined using immunoradiometric assay in I33 patients with lung cancer. CYFRA2 I-1 was significantly higher in lung cancer patients than that ofbenign pulmonary disease and healthy adults (P < 0.001). Compared to serum NSE. CYFRAZ l- I was more frequently elevated in patients with non-small cell cancer (NSCLC) (Sensitivity 52.8%). especially squamous cell carcinoma (SQC) (72.1%). than in