Novel insights into the mixed germ cell-sex cord stromal tumor of the testis: detection of chromosomal aneuploidy and further morphological evidence supporting the neoplastic nature of the germ cell component

Abstract

The existence of a true mixed germ cell-sex cord stromal tumor (MGSCT) of the testis remains controversial. Based on our experience with rare testicular tumors in this spectrum, we sought to perform a detailed clinicopathologic and molecular study of MGCSCT. Eight cases of testicular MGSCT were morphologically reviewed, screened for chromosomal aberrations (using array comparative genomic hybridization (aCGH) and low pass genomic sequencing), and analyzed by next generation sequencing (The Illumina TruSight Tumor 170). Immunohistochemistry for OCT3/4, Nanog, SALL4, DMRT1, and inhibin was performed on the cohort. Clinical data and follow-up were assessed by medical record review. All patients were karyotypically normal men aged 27–74 years (median 41). All tumors had a similar biphasic morphology characterized by various proportions of the sex cord component resembling granulosa cell tumor of adult type and the germ cell component cytomorphologically akin to spermatocytic tumor. Germ cells were haphazardly scattered throughout the tumor or arranged in larger groups, without tubular formation. In 4 cases, atypical mitoses were found within the germ cells. Additionally, in 2 cases there was invasion into the spermatic cord, adjacent hilar soft tissue and into the tumor capsule, which contained both tumor components. Immunohistochemically, focal nuclear expression of DMRT1 was found in the germ cell component in 7/7 analyzable tumors, while SALL4 was positive in 6 cases and negative in one case. All tumors were negative with OCT3/4 and Nanog. The sex cord stromal component had immunoreactivity for inhibin in 7/7 analyzable cases. Four of 8 cases were cytogenetically analyzable: 4/8 by low pass genomic sequencing and 2/8 by aCGH. The results of both methods correlated well, revealing mostly multiple chromosomal losses and gains. One case revealed loss of chromosome 21; 1 case had loss of chromosomes 21 and 22 and partial gain of 22; 1 case had loss of chromosomes 22 and Y, partial loss of X, and gain of chromosomes 20, 5, 8, 9, 12, and 13; and the remaining one gain of chromosomes 20, 3, 6, 8, 2x(9), 11, 2x(12), 13, 14, 18, and 19. Three cases were analyzable by NGS; clinically significant activating mutations of either FGFR3 or HRAS were not detected in any case. Follow-up was available for 4 patients (12, 24, 84, and 288 months) and was uneventful in all 4 cases. The identification of extratesticular invasion of both the germ cell and sex cord stromal components, the DMRT1 expression, and the presence of atypical mitoses in germ cells argue for the neoplastic nature of the germ cell component. The molecular genetic study revealing multiple chromosomal losses and gains in a subset of the cases provides the first evidence that molecular abnormalities occur in testicular MGSCT. Multiple chromosomal aneuploidies, namely, recurrent losses of chromosomes 21 and 22 and gains of 8, 9, 12, 13, and 20, indicate that the germ cell component might be related to the morphologically similar spermatocytic tumor, which is characterized by extensive aneuploidies including recurrent gains of chromosomes 9 and 20 and loss of chromosome 7. In summary, our data support that rare examples of true MGSCT of the testis do exist and they represent a distinct tumor entity with admixed adult-type granulosa cell tumor and spermatocytic tumor components.