Prognostic significance of microvascular invasion and microlymphatic permeation in non-small-cell lung cancer

We read with interest the recently published retrospective study by Hishida et al. [1]. The authors reviewed the data of 1039 patients with stage T1a-3N0M0 non-small-cell lung cancer (NSCLC) and concluded that microvascular invasion and microlymphatic permeation had different impact on the outcome. We are interested with the methodology of pathological microvascular invasion and microlymphatic permeation. In this article, the authors used haematoxylin and eosin and Victoria blue-van Gieson (VVG) staining to evaluate microvascular invasion. The presence of microvascular invasion was determined by identifying conspicuous clusters of intravascular cancer cells surrounded by a VVG-stained elastic layer. Microlymphatic permeation was considered to be positive if tumour cells were present inside VVG-negative, thin endothelial-lined channels without supporting smooth muscles or elastic fibres. It is essentially impossible by haematoxylin and eosin staining alone to reliably distinguish lymphatic channels from small blood vessels including capillaries and post-capillary venules or from artefactual stromal retraction. The authors noted this problem and considered that lymphatic permeation was concluded to be present when floating tumour cells were found in lumens within the bronchovascular bundle, subpleural or interlobular pleural spaces. Elastic stains are routinely used to help identify larger blood vessels, but they are less useful in excluding capillaries and are not useful in making the distinction between stromal retraction and true lymphatic spaces. Thus, this methodological issue could decrease the rationale and reliability of the results and conclusions. More recently, some studies used immunohistochemical markers such as CD34 to specifically identify blood vessels [2] and D2-40 to identify lymphatic endothelium [3]. These specific markers might be of value in improving the accuracy of detecting and distinguishing between microvascular invasion and microlymphatic permeation.

Hishida et al. found that microvascular invasion (hazard ratio [HR]: 1.65, P = 0.001), but not microlymphatic permeation (HR: 1.14, P = 0.588), had an important effect on the prognosis of patients. In a meta-analysis involving 16 535 cases from 52 studies [4], we concluded that the multivariate estimate for overall survival was 1.90 (P = 0.0001), although in Stage I NSCLC patients, the meta-risk of recurrence (HR: 6.93, P = 0.0001) and death (HR: 2.15, P = 0.0001) remained highly statistically significant. The observations of Hishida et al. correspond with our results in which NSCLC patients with microvascular invasion can benefit adjunct systematic chemotherapy, and this pathological marker should be incorporated in the new edition of the TNM classification. However, we suggest that the prognostic significance not be restricted to stage T1a-3N0M0 NSCLC patients. In addition, through assessing these factors separately, the authors' results showed that microvascular invasion had more impact on survival than microlymphatic permeation. More comparison studies using similar methods should be presented in other centres, and a meta-analysis also be considered to verify their conclusions.

We believe that, in NSCLC patients, microvascular invasion may have led to an increase in the risk of recurrence and death. Large, properly designed studies that employ standard methodology and quality control to assess lymphvascular invasion are required in demonstrating whether lymphovascular invasion and microlymphatic permeation can provide prognostic information.

Conflict of interest: none declared.

REFERENCES