ABSTRACT: A comparative morphological analysis of parenchyma adjacent to testicular germ cell tumours (TGCT) was performed in a series of 181 orchidectomy specimens: 86 with seminomas (Se), 72 with nonseminomatous germ cell tumours (NS) and 23 with combined tumours (CT, which have a Se and a NS component). The following morphological features were semiquantitatively scored: spermatogenesis (modified Johnsen score); amount of tubular atrophy; amount of carcinoma in situ (CIS); amount of intertubular tissue. Absence and presence was scored for the following features: lymphocytic infiltrate surrounding and invading CIS; intratubular seminoma (ISe); intratubular nonseminoma (INS); microlithiasis; diffuse and nodular hyperplasia of Leydig cells; angioinvasiveness; testicular angiopathy. Using non-parametric statistics these features were correlated with each other and with tumour type, tumour size and age of the patient. Se-patients presented at significantly higher age than NS-patients (36 vs 29 years, p=0.001). The age of patients with CT (32 years) was in between that of Se- and NS-patients. No correlation was found between patient age and tumour size. Parenchyma adjacent to Se, compared to parenchyma adjacent to NS had the following significant differences: a lower Johnsen score (5.6 vs 7.2, p=0.005); less frequent (85% vs 97% of specimens, p=0.016) and a lesser amount of CIS (26% vs 32% of tubules, p=0.015); more frequent peri- (80% vs 60% of specimens, p=0.001) and intratubular (68% vs 30% of specimens, p=0.001) lymphocytic infiltrates; more extensive tubular atrophy (36% vs 15% of tubules, p=0.001); and a larger area of intertubular tissue (42% vs 34% of parenchyma area, p=0.016). The pooled Se and CT had a significantly higher frequency of ISe than the NS (31% vs 17% of specimens, p=0.036). With one exception INS was only found adjacent to NS or CT, with a frequency of 16%, and 20% of the specimens, respectively. It was significantly associated with angio-invasiveness. In specimens lacking angio-invasion the frequency of INS was 6%. The correlation of INS with tumour size and patient age was studied in a series of 145 NS and CT (95 from the original series supplemented by 50 newer cases). In this series INS was significantly associated with smaller tumours and younger patients. Extensive tubular atrophy was significantly correlated with higher age, the diagnosis of Se, a low Johnsen score, and the presence of angiopathy. The more tubular atrophy, the less CIS (both in incidence and amount). Inversely, a higher Johnsen score is associated with smaller tumours, the diagnosis of NS or CT, a higher incidence and a larger amount of CIS, and little tubular atrophy. Tubules with mature spermatogenesis were found in 42% of the specimens regardless of tumour type. We conclude that ISe and INS are probably frequent intermediate stages between CIS and Se and NS, respectively. The features of parenchyma adjacent to Se are probably due to the host response elicited by the invasive Se, which secondarily also affects CIS. The long time to clinical presentation allows the host to eradicate most of the CIS by the time the tumour is surgically removed. The much less extensive morphological features of a host response in parenchyma adjacent to NS support the contention that NS originates as INS, behind the blood/testis barrier, without exposure of the host to tumour cells with a seminomatous phenotype (CIS- or Se cells). Microlithiasis and testicular angiopathy are frequent, but not specific findings in parenchyma next to TGCT. Their relationship with the development with TGCT is unexplained.
Apmis 02/2003; 111(1):32-40; discussion 41-2. · 1.99 Impact Factor