Visceral pleura invasion by non-small cell lung cancer: An underrated bad prognostic factor

Citation
D. Manac'H et al., Visceral pleura invasion by non-small cell lung cancer: An underrated bad prognostic factor, ANN THORAC, 71(4), 2001, pp. 1088-1093
Citations number
27
Categorie Soggetti
Cardiovascular & Respiratory Systems","Medical Research Diagnosis & Treatment
Journal title
ANNALS OF THORACIC SURGERY
ISSN journal
00034975 → ACNP
Volume
71
Issue
4
Year of publication
2001
Pages
1088 - 1093
Database
ISI
SICI code
0003-4975(200104)71:4<1088:VPIBNC>2.0.ZU;2-J
Abstract
Background. Visceral pleura invasion (VPI) by nonsmall cell lung cancer is a factor of poor prognosis. A tumor of any size that invades the visceral p leura is classified as T2. Few studies have been conducted concerning the p rognostic significance of VPI relative to other staging factors. Methods. Between April 1984 and December 1996, 1,281 patients with T1 (n = 430) and T2 (n = 851) non-small cell lung cancer underwent curative surgica l resection. Adjuvant radiation therapy was performed in 455 patients. Ther e were 176 women and 1,105 men aged 30 to 86 years (mean, 60.9 years). Five hundred nineteen pneumonectomies, 742 lobectomies, and 20 segmentectomies were performed. In all patients, a complete mediastinal lymph node dissecti on was performed. International staging was stage IA and B (n = 697); stage II A and B (n = 247), and stage III A (n = 337). The patients were divided into two groups according to the existence of VPI (group I without, group II with). Both groups were compared with regard to the size of the tumors, histology, associated lymph node involvement, survival rates, and cause of death. Univariate and multivariate analyses were conducted. Results. VPI (group II) was identified in 19.1% of the resected specimens: group I, n = 1036; group II, n = 245. The VPI was present in only 10% of no n-small cell lung cancer 3 cm or less in size, reaching 33% of patients wit h non-small cell lung cancer larger than 5 cm (p = 0.0001). Squamous non-sm all cell lung canter were significantly less accompanied by VPI (13.5%) tha n the other histologic categories. The VPI was associated with a higher fre quency of N2 involvement (group I = 24.6%, group II = 33.4%, p = 0.01) and N2 involvement was more extensive (two or more N2 involved stations: group I = 8.2%, group II = 15.6%, p = 0.003). Actuarial survival rates were 51.8% at 5 years and 33.8% at 10 years in group I (median, 66 months), and 34.6% at 5 years and 27.9% at 10 years in group II (median, 30 months) (p = 0.00 0002). Long-term survival rates significantly decreased for larger tumors. Even in patients with N2 stage tumors, the difference of survival curves be tween the two groups was statistically significant. Cancer-related deaths w ere more frequent in group II and were mainly caused by distant metastases. By multivariate analysis, visceral pleura invasion proved to be a signific ant independent fatter of poor prognosis. Conclusions. The VPI is a factor of poor prognosis. Its frequent associatio n with extensive N2 involvement supports the hypothesis that exfoliated tum or cells are drained through the pleural lymphatics by the mediastinal lymp hatic pathways and then into the bloodstream. The VPI is an important progn ostic factor and, as such should stimulate more studies to better select th e patients who could benefit from adjuvant therapy. (C) 2001 by The Society of Thoracic Surgeons.