Detection of lymphovascular invasion in urothelial carcinoma of the bladder through D2-40 immunostaining

Lymphovascular invasion (LVI) represents a surrogate marker for micrometastatic urothelial carcinoma of the bladder (UCB). We evaluated whether D2-40 immunhistochemistry (IHC) alters detection of LVI when compared to conventional HE (hematoxylin-eosin) staining of UCB specimens in a blinded fashion. HE- and D2-40-IHC-stained representative sections of 80 patients after radical cystectomy (RC) were re-reviewed. LVI detection rates were recorded and compared after blinded evaluation. LVI was present in 53 patients (66.3%) in HE-stained sections and in 44 patients (55%) in D2-40 stainings. In 13 patients, LVI (16.3%) was found in HE stained sections but not confirmed when IHC was applied (false positive when using IHC as a reference standard). D2-40 IHC identified LVI in 4 additional patients (5%) who were classified as LVI negative in conventional HE staining (false negative). 52 patients (65%) were lymph node negative (pN0), 21 of whom (40.4%) were LVI positive in conventional HE sections and 16 of whom (30.8%) were LVI positive in IHC. In 9 pN0 patients (17.3%), LVI was diagnosed in HE sections but not confirmed by IHC (false positive). D2-40 IHC identified LVI in 4 additional patients (7.7%) who were node negative and classified as LVI negative in conventional HE staining (false negative). In patients who experienced recurrence (n=35) and who were classified as pN0 at the time of RC, HE staining resulted both in false-positive (n=2; 5.7%) and false-negative (n=3; 8.6%) findings. Different detection rates of LVI were observed when using IHC with D2-40 in UCB patients compared to conventional HE staining. The routine use of D2-40 IHC should be considered in clinical trial design to improve risk stratification of pN0 patients after RC.