Prognostic factors in pulmonary fibrohistiocytic lesions
A A Gal,M N Koss,W F McCarthy,L Hochholzer
DOI: https://doi.org/10.1002/1097-0142(19940401)73:7<1817::aid-cncr2820730708>3.0.co;2-k
IF: 6.9209
1994-04-01
Cancer
Abstract:Background: The spectrum of pulmonary fibrohistiocytic lesions (PFL) ranges from benign fibrohistiocytoma or inflammatory pseudotumor, fibrohistiocytic type (IP), to fibrohistiocytic lesions of borderline malignant appearance (B), to unequivocal malignant fibrous histiocytoma (MFH). The authors reviewed a spectrum of PFL in an attempt to define their natural history and establish independent factors of prognostic significance. Methods: Thirty-one fibrohistiocytic lesions were analyzed by multiple clinical, gross, and microscopic factors. Prognostic factors were evaluated by univariate and multivariate Cox proportional hazards models. Results: After histopathologic review, the cases were classified into IP (15 cases), B (3 cases), and MFH (13 cases). IP tended to occur in younger patients (median age, 30 years; range, 11-64 years) than did MFH (median age, 54 years; range, 12-81 years) (P = 0.03). All 15 patients with IP and 3 patients with B survived. Local recurrence developed in 2 patients with IP and 6 patients with MFH. Distant metastases occurred in 7 patients with MFH. Eight of 13 patients with MFH died of disease, with a median survival time of 16 months (range, 6-36). For all pulmonary fibrohistiocytic lesions, univariate Cox regression analysis disclosed several statistically significant prognostic factors associated with mortality: metastasis (P = 0.0005), significant necrosis (P = 0.0015), local recurrence (P = 0.003), bizarre giant cells (P = 0.0282), mitoses greater than or equal to 3 per 50 high-power fields (P = 0.0286), stage at surgery (P = 0.002), high cellularity (P = 0.0267), and poor circumscription (P = 0.0294). Other parameters (age, sex, symptoms, laterality, location, atypical mitoses, nuclear pleomorphism, degree of fibrosis, and size) were not significant prognostic factors. By multivariate Cox regression analysis, metastasis was the most significant individual parameter (P = 0.002), overriding all other prognostic factors. With metastasis removed from the model, local recurrence (P = 0.0199) and necrosis (P = 0.037) were the other significant independent prognostic factors. Conclusions: Certain individual clinical, gross, and microscopic features have prognostic relevance in determining survival time in pulmonary fibrohistiocytic lesions.