Peripheral lymphovenous communication in lymphoedema

AIM To investigate whether in lower extremity lymphoedema, lymph proteins enter blood before they do the thoracic duct. METHODS Retrospective analysis of routine lymphoscintigraphy in 69 adults imaged at 5, 45 and 150 min following bilateral subcutaneous web space injection of Tc-nanocolloid was carried out. Regions of interest were placed over the liver and ilioinguinal lymph nodes bilaterally on the anterior images at 45 and 150 min. Individual minor (0.5 point for each) and major (1 point for each) criteria of abnormal scintigraphy were applied to each limb and summed to give a lymphoscintigraphic abnormality score. An abnormal limb had a score =1. RESULTS The ratio of hepatic counts per pixel to total bilateral ilioinguinal counts (L/N ratio) was higher in patients with abnormal results on lymphoscintigraphy (median 6.2; interquartile range 4.0-15.6 pixels×10; n=48) compared with that in patients with normal lymphoscintigraphic results (2.5 [1.5-5.0] pixels×10; n=21; P<0.0002). In the abnormal group, the lymphoscintigraphic score (two limbs summed) correlated with the 150-min L/N ratio (r s=0.42; P<0.005). L/N ratios at 45 and 150 min correlated in the abnormal group (r s=0.44; P<0.005) but not in the normal group (r s=0.3; P>0.05). The 45-min activity, as a percentage of the 150-min activity, was higher in lymph nodes than in the liver in both the abnormal (35.0 [8.2-50.0] vs. 10.6 [5.8-30.0]%; P<0.0001]) and normal groups (38.3 [18.4-63.5] vs. 23.3 [12.4-33.1]%; P<0.05), and, with respect to the liver, was higher in the normal group (P<0.01). CONCLUSION In lymphoedema, more lymph proteins enter blood proximal to the thoracic duct. The time courses of nodal and hepatic activities suggest that access may occur within nodes themselves.