RESUMO
BACKGROUND: Intraoperative crush smear is an adjuvant in diagnosing central nervous system (CNS) lesions on tissue sent for frozen section. Besides rapid decision-making, it also ensures that minimum injury is caused to the normal brain structures surrounding the intracranial neoplasm. A rapid intraoperative diagnosis helps the surgeon in planning the appropriate surgery. OBJECTIVE: Our objective is to review all the discordant cases between intraoperative and histopathological diagnosis and also to study the crush smear slides for morphological clues that could have been helpful in minimizing such errors, especially for an inexperienced neuropathologist/general pathologist. The surgeon's perspective on the impact of these errors on management is also discussed. METHOD: A prospective study of six years from 2013 to 2019 was conducted. Crush smears were made and stained with rapid hematoxylin and eosin (H&E). The rest of the tissue was processed for permanent tissue sections. Slides in which there was discordance between the intraoperative and permanent paraffin sections were reviewed to ascertain the reasons thereof. RESULTS: A total of 81 specimens of CNS tumors were sent for intraoperative consultation. Out of these, discordance was seen in 13 (16%) cases. CONCLUSION: To minimize diagnostic errors, it is important to do regular analyses of the misinterpreted cases. Knowledge of the pre-operative radiological differential diagnosis is mandatory. Discussion with the surgeon regarding the clinical impact of the errors made will give a clearer picture to the pathologists regarding clinically relevant reporting during intraoperative consultation.
RESUMO
Stage of liver fibrosis is critical for treatment decision and prediction of outcomes in chronic hepatitis C (CHC) patients. We evaluated the diagnostic accuracy of transient elastography (TE)-FibroScan and noninvasive serum markers tests in the assessment of liver fibrosis in CHC patients, in reference to liver biopsy. One-hundred treatment-naive CHC patients were subjected to liver biopsy, TE-FibroScan, and eight serum biomarkers tests; AST/ALT ratio (AAR), AST to platelet ratio index (APRI), age-platelet index (AP index), fibrosis quotient (FibroQ), fibrosis 4 index (FIB-4), cirrhosis discriminant score (CDS), King score, and Goteborg University Cirrhosis Index (GUCI). Receiver operating characteristic curves were constructed to compare the diagnostic accuracy of these noninvasive methods in predicting significant fibrosis in CHC patients. TE-FibroScan predicted significant fibrosis at cutoff value 8.5 kPa with area under the receiver operating characteristic (AUROC) 0.90, sensitivity 83%, specificity 91.5%, positive predictive value (PPV) 91.2%, and negative predictive value (NPV) 84.4%. Serum biomarkers tests showed that AP index and FibroQ had the highest diagnostic accuracy in predicting significant liver fibrosis at cutoff 4.5 and 2.7, AUROC was 0.8 and 0.8 with sensitivity 73.6% and 73.6%, specificity 70.2% and 68.1%, PPV 71.1% and 69.8%, and NPV 72.9% and 72.3%, respectively. Combined AP index and FibroQ had AUROC 0.83 with sensitivity 73.6%, specificity 80.9%, PPV 79.6%, and NPV 75.7% for predicting significant liver fibrosis. APRI, FIB-4, CDS, King score, and GUCI had intermediate accuracy in predicting significant liver fibrosis with AUROC 0.68, 0.78, 0.74, 0.74, and 0.67, respectively, while AAR had low accuracy in predicting significant liver fibrosis. TE-FibroScan is the most accurate noninvasive alternative to liver biopsy. AP index and FibroQ, either as individual tests or combined, have good accuracy in predicting significant liver fibrosis, and are better combined for higher specificity.