Table 3.

Representative studies that emphasise the importance of lymphovascular invasion in surgical resections listed chronologically

Authors, setting and date of study (reference number)	Number and type of biopsy	Principle findings	Comments
Alexander-Sefre et al. London, UK 2003 [72]	108 patients with stage 1 endometrial adenocarcinoma	Substantial increase in the detection of vascular invasion with pan-cytokeratin and CD31 immunohistochemistry (from 21 to 58 cases). No distinction between blood and lymphatic vessels	Vascular invasion has been underestimated in early endometrial adenocarcinoma
Vass et al. Glasgow, UK 2004 [73]	75 Colonic adenocarcinomas	Elastin staining increased detection of venous invasion from 18 to 32 cases for extramural invasion and from 8 to 30 cases for intramural invasion	Elastic staining should be part of standard protocols
Pawlik et al. China, France, Japan, USA 2006 [74]	Multinational registry of 1073 resections for hepatocellular carcinoma	41 % of tumours >5 cm had LBVI in comparison with 27 % <5 cm. Multicentricity, histological grade and high AFP levels also associated with LBVI	No comments on histological methods for identifying LBVI
Chen et al. South Australia 2010 [75]	110 Whipple’s resections for pancreatic carcinoma between 1998 and 2008	5 year survival 77 % in patients negative for both LBVI and perineurial invasion but only 15 % in patients positive for both	Poor differentiation, size >3 cm and nodal involvement also poor prognostic features
Storr et al. Nottingham, UK 2012 [76]	202 cutaneous melanomas	Lymphatic invasion more common than venous invasion (27 vs 4 %). Immunohistochemistry with CD34 and D2-40 increases detection rate of blood vessel and lymphatic invasion.	Lymphatic invasion is associated with adverse factors but lymphatic characteristics do not predict outcome
Kirsch et al. Canada 2013 [77]	Sections of 40 colorectal carcinomas circulated to specialist and non-specialist GI pathologists	GI pathologists detected venous invasion more frequently than non-GI with both H&E and Movat’s stain. Detection of venous invasion was >2 fold higher with Movat’s stain (46 vs 20 %)	Venous invasion is under-detected with H&E, even by specialist pathologists
Gujam et al. Glasgow, UK 2014 [78]	Review of 59 reports of 62514 patients with breast carcinoma	19/21 studies demonstrated lymphatic vessel invasion predicted poorer prognosis. Improvement of lymphatic detection using immunohistochemistry	Guidelines for the use of immunohistochemistry in mammary carcinoma should be followed
Castonguay et al. Canada 2014 [79]	103 oesophageal adenocarcinoma resections	Venous invasion detected in 8 cases with H&E but in an additional 66 cases with Movat’s pentachrome	Venous invasion, stage, size and grade prognostically significant on univariate analysis

Expressed in haemopoetic and vascular associated tissues, D2-40—an antibody directed against a glycoprotein selectively expressed on lymphatic endothelium; Movat’s pentachrome—a modification of the trichrome method incorporating specific staining of elastic tissue, widely used by cardiovascular pathologists especially in North America

CD 31 platelet endothelial cell adhesion molecule, CD 34 a cell surface glycoprotein of uncertain function, GI gastrointestinal, BLVI lymphovascular invasion, H&E haematoxylin and eosin