Lymphatic vessels; Complexity factor; Hodgkin lymphoma; pRb; IκB-α; BAX
Recently, it has been suggested that lymphatic vessel proliferation (lymphangiogenesis - LAG) may provide a better target for immunotherapy and biological therapy than angiogenesis, where such therapies have mainly failed [1-4]. Classic Hodgkin lymphoma (cHL) is a primary malignant neoplasm of the immune system. For practical purposes, it almost exclusively involves lymph nodes (LNs) at diagnosis. Whereas LAG in the LNs has been detected mainly in metastatic tumors, LAG has occasionally been described in primary malignant lymphomas [5-7]. Angiogenesis has rarely been detected in cHL. Such an unusual description has been investigated by morphometric methods, microvessel density or the hot spot technique [8]. Lymphangiogenesis and angiogenesis have been explored by evaluating the expression of VEGF and variants in cHL [9].
We have previously atempted to clarify the features of LAG in cHL by using a morphometric method and looking for clinicopathological correlations [10]. A review of the previously published work on this topic is presented herein. The possible relevance of LAG is investigated, and when highlighted, a search is carried out to demonstrate the genes and growth factors associated with the role of LAG in cHL in an expanded population of cHL patients. Since studies on LAG in lymphoma are scarce, we conclude the present paper with a review of the literature published recently on LAG in conditions related directly or indirectly to lymphoma.
For the original study on LAG in cHL, nineteen cases of mixed cellularity cHL, primarily involving the LNs, were selected from our archives of formalin-fixed, paraffin-embedded tissue samples. Five-micron-thick sections were submitted for immunohistochemistry (IHC) (avidin-biotin complex method) with the D2-40 antibody (1:10, Dako to identify lymphatic vessels (LVs) and an anti-CD34 antibody to identify the blood vessel (BV) endothelium. For each case, three lymphatic vessel hot spots were analyzed (57 hot spots in total). The hot spots were imaged at 200X magnification and then printed with black ink on white paper. The vessels were underlined in the images: the LVs were traced in brown, and the BV endothelium was traced in red.
Fifteen of the 57 hot spots, which were selected at random, were subjected to morphometry twice. For each case, mean LV perimeter, major axis length, surface area and mean (shape) complexity factor analyses used the microvessel count and image analysis approaches described by Korkopoulou et al. [8]. The complexity factor was computed as equal to 5π*area/perimeter2. The above parameters were correlated with clinical characteristics and were also related to CD30, EBV/LMP1, EBER, CD20, p53, MDM-2, pRb, BAX and IκB-α expression and to the apoptotic index [11]. In 10 distinct sections of the hot spots, double staining of LVs was performed with D2-40 (counterstained brown with a DAB substrate) and an anti-CD34 antibody (counterstained red with an AEC substrate). For this analysis, we used the method reported by Kyras et al. [12]. Statistical analysis employed the Mann-Whitney test for nonparametric values and Spearman's analysis for nonparametric correlations. Finally, we carried out a review of the recent (2014-2019) relevant literature on LAG and lymphomas, but included also conditions akin to lymphomas.
3.1 Lymphangiogenesis in cHL
No statistically significant associations were found between the four lymphatic parameters and clinical characteristics. Statistically significant inverse relationships were obtained between the LV parameter mean lymphatic area and mean complexity factor and pRb expression. BAX was inversely and significantly linked with the mean surface area and the mean LV length. IκB-α expression was inversely and significantly related to the mean LV perimeter . Of all the LV parameters, only the LV major axis length was inversely correlated with the shape factor, and this correlation was significant (r=-.6; p=.03). The only significant association of the mean lymphatic perimeter was an inverse relationship with IκB-α expression (p=.015). A direct and significant relation was found between the shape factor and pRb expression [10].
3.2 Relevance of BAX, pRb and IκB-α expression in an expanded (n=178) cHL cohort
As these genes were shown to present significant relations with LAG in cHL, confirmation in a larger population was deemed necessary. BAX was inversely linked with bulky cHL (p=.040). No other significant clinical associations were found in this large cohort. pRb expression was inversely related to the expression of a sialylated CD15 antigen. The relationships between BAX expression and the biological markers of cHL, included direct links with MCL1, p53 and MDM-2 expression, and inverse links with BAK and BCL-2 expression [10].
A significant inverse association between IκB-α and LMP1 of EBV was demonstrated. Kaplan-Meier analysis showed no significant difference in overall survival among patients stratified by BAX expression. In addition, there was no relationship between IκB-α expression and overall survival in cHL (data not shown).