Incidence of testicular tumor subtypes according to the updated WHO classification, North Rhine-Westphalia, Germany 2008–2013

Abstract

Background:

In 2016, the WHO introduced an updated classification for testicular tumors. The application of this updated classification to cancer registry data requires some re-coding of tumors.

Objectives:

The aim of this study was to provide up-to-date population-based incidence estimates of testicular tumor subtypes according to the updated classification.

Material and methods:

We reviewed 2,251 pathology reports (42.9%) out of 5,252 testicular tumors at the cancer registry of North Rhine-Westphalia for the years 2008–2013. We used population counts to estimate age-standardized incidence rates per million person-years (EUROSTAT revised European Standard Population).

Results:

The application of the updated WHO classification resulted in a re-coding of 8.9% of all testicular tumors. While the re-codings have no influence on the incidence of seminomatous and non-seminomatous TGCTs that include mixed TGCTs, they influence the incidence of individual histological types of seminomatous and non-seminomatous TGCTs. Among the 4,935 testicular germ cell tumors (TGCT), 23.7% were mixed TGCTs. Overall 46.9% of all mixed TGCTs included seminoma and age-standardized incidence rates were highest for the combination seminoma plus embryonal carcinoma (5.9 per million person-years) and embryonal carcinoma plus teratoma (4.9 per million person-years). The median age at diagnosis was higher for mixed TGCTs including seminoma (31 years) than those that did not include seminoma (28 years).

Discussion and Conclusions:

Population-based incidence time trends for seminomatous and non-seminomatous TGCTs that include mixed TGCTs are not distorted by the introduction of the WHO update. Trend distortions can only be expected if time trends of individual histological subtypes of the seminomatous and non-seminomatous TGCTs are examined.

Introduction

With the introduction of the updated WHO classification of testicular germ cell tumors (TGCTs), a variety of changes have occurred. The updated classification is not only based on morphological features, as was the prior WHO classification, but also on pathogenetic features: testicular germ cell tumors are now categorized as either tumors derived from germ cell neoplasia in situ (GCNIS), or as TGCTs unrelated to germ cell neoplasia in situ. GCNIS is a new term, which did not exist in previous WHO classifications (Berney et al., 2016). GCNIS-derived entities are now designated as postpubertal-type, while the non-GCNIS-related type are designated as prepubertal-type. Benign prepubertal-type teratomas are acknowledged to occur rarely in the postpubertal testis. Spermatocytic seminoma is now renamed as spermatocytic tumor and placed within the non-GCNIS related tumors. Dermoid cyst, epidermoid cyst, and well-differentiated neuroendocrine tumors are specialized forms of prepubertal-type teratomas. Monophasic choriocarcinoma is now considered to be a morphologic variant of choriocarcinoma and the non-choriocarcinomatous trophoblastic tumor group has been expanded (Moch et al., 2016; Ulbright, 2016). Furthermore, TGCTs are grouped into tumors with a single histological type (pure tumors) and tumors with more than one histological type (mixed tumors).

The update of the WHO classification requires some ICD-O re-coding of TGCT. For example, TGCTs formerly coded as teratocarcinomas (International Classification of Diseases for Oncology, ICD-O (Fritz et al., 2000), M9081/3) or coded as “choriocarcinoma combined with other germ cell elements” (M9101/3) are now coded as mixed tumors (M9085/3). Anaplastic seminoma (M9062/3) are now coded as seminoma (M9061/3). Undifferentiated malignant teratomas (anaplastic malignant teratomas) (M9082/3) are now coded as embryonal carcinomas (M9070/3) and malignant trophoblastic teratomas (M9102/3) – so termed by the British classification − are now coded as choriocarcinomas (M9100/3) (Moch et al., 2016; Ulbright, 2016).

To the best of our knowledge, reports of population-based TGCT incidence rates according to the updated WHO classification of testicular tumor subtypes have not been previously published. The cancer registry of North Rhine-Westphalia in Germany digitally stores clinical and pathology reports of reported cases so that pathology reports can be reviewed. The aim of this study was to provide an ICD-O conversion table for the application of the updated WHO classification and to provide up-to-date population-based incidence estimates of testicular tumor subtypes with a special focus on mixed TGCTs according to the updated WHO classification.

Material and Methods

Since 2007, the estimated completeness of the cancer registry of North Rhine-Westphalia (NRW), Germany has been 95% or greater (Koch-Institut, 2015). NRW is the most populated Federal State in the country (18 million people). Cancer reporting in NRW is mandatory and dominated by pathology reports. The electronic storage of pathology and clinical reports enables the review of testicular tumors (ICD-10: C62, (1992)) reports. From 2008 through 2013, overall 5,263 C62-coded tumors were reported to the registry.

First, we examined the ICD-O coding quality of the data that existed prior to the application of the new WHO classification for all reports with the exception of tumors coded as “seminoma, not otherwise specified” (9061/3, n=2,994) and “choriocarcinoma, not otherwise specified” (9100/3, n =18). This examination included the review of 2,251 (42.8%) reports and revealed that 11 cases were extragonadal germ cell tumors. These were excluded from all further analyses. ICD-O codes were revised where necessary. As even a little miscoding of mixed tumors as pure seminoma could have produced a substantial decline of the total number of mixed tumors, we reviewed the pathology reports of a random sample of 100 tumors coded as pure seminoma (M9061/3).

Second, we re-coded TGCTs by use of an ICD-O conversion table that we developed according to the updated WHO classification (Supplementary Table 1). For some mixed tumors, we could not document all histological components that were diagnosed because some tumor reports were incomplete.

We used population counts by calendar year (2008–2013) and 5-year age groups to estimate crude and age-standardized incidence rates according to the EUROSTAT revised European Standard Population of 2013. In addition, we provide median ages and the 25^th and 75^th percentiles of age at the time of diagnosis and coefficients of variation of age at diagnosis (CV), a standardized measure of the variability of age calculated as the ratio of the standard deviation to the mean. Smaller CVs indicate less variability than larger CVs. We report CVs only for histological groups with at least 10 patients.

Results

Overall 5,252 testicular tumors were reported to the cancer registry of NRW from 2008 through 2013. Among these tumors, 4,946 (94.2%) were histologically verified and 68 (1.3%) were reported as death certificate only cases (DCO). The remaining 238 testicular tumors were reported on the basis of cytology, clinical data, other or unknown sources. Figure 1 shows the total numbers of testicular tumors by histological groups. The review of 100 pathology reports of pure seminoma revealed that only one out of 100 tumors was a mixed tumor (seminoma plus embryonal carcinoma). The application of the ICD-O coding according to the updated WHO classification resulted in an ICD-O re-coding of 468 of 5,252 TGCTs (8.9%) (Supplementary Table 2). While the ICD-O re-codings have no influence on the incidence of seminomatous and non-seminomatous TGCTs that include mixed TGCTs, they influence the incidence of histological subtypes of seminomatous and non-seminomatous TGCTs.

Figure 1.

Incident testicular tumors in North Rhine-Westphalia, Germany, 2008–2013

TGCT – testicular germ cell tumor; SEM: seminoma;

Among the 4,935 TGCTs, 3,763 (76.3%) were pure TGCTs and 1,172 (23.7%) were mixed TGCTs. The distribution of pure TGCTs shows that seminoma (82.4%) and embryonal carcinoma (10.8%) were the most frequent histologies. Only 3.4% of pure TGCTs were teratoma. Spermatocytic tumors comprised only 0.8% of the total and the patients were considerably older than were men with other pure TGCTs. The variability of age at diagnosis was lower for seminoma and spermatocytic tumors than for the other pure TGCTs (Figure 2 & Supplementary Table 3).

Figure 2.

Age distributions and incidence rates (cases per million person-years) of 3,763 primary malignant testicular pure germ cell tumors, North Rhine-Westphalia, Germany, 2008–2013

Boxes indicate the interquartile range. Whiskers indicate observations within the lower and upper fence [i.e. 1.5*interquartile range (IQR) below 25th percentile and 1.5*IQR above 75th percentile]. Stars indicate observations outside the fences. Dots indicate mean age. Horizontal lines within the boxes indicate the median ages; CV%: coefficient of variation in percent; crude (SE): crude rate per million person-years with standard error; ASR (SE): age-standardized rate per million person-years (EUROSTAT revised European Standard Population 2013); SEM: seminoma, EC: embryonal carcinoma; YST postpub: yolk sac tumor, postpubertal-type; CHOR: choriocarcinoma; TER postpub: teratoma, postpubertal type; TER som-type: teratoma with somatic-type malignancy; SPC: spermatocytic tumor; YST prepub: yolk sac tumor, prepubertal-type; n=2 well-differentiated neuroendocrine tumors are not displayed.

Patients with malignant Leydig cell tumors, carcinoma, sarcoma and soft tissue tumors of the testis were considerably older than patients with other testicular tumors. Sarcoma and soft tissue tumors showed the largest variability of age at diagnosis. Overall 5.0% (262 out of 5,252) of all testicular tumors reported to the cancer registry were either histologically unspecified or only marginally specified (“epithelial neoplasia”, “carcinoma”) which was mainly due to the lack of detailed pathology reports (Supplementary Table 4).

Among the 1,172 mixed TGCTs, 550 (46.9%) had a seminoma component, while 547 (46.7%) did not. Cancer registry reports of the remaining 75 mixed TGCTs (6.4%) were too incomplete to further subtype the histological elements. Among mixed TGCTs with a seminoma component, tumors with two histological elements had the highest age-standardized incidence rates (IR), specifically the combination of seminoma and embryonal carcinoma (IR=4.1) and the combination of seminoma and teratoma (IR=1.5). The median age at diagnosis was higher for mixed TGCTs that included a seminoma component (31 years) than it was for those that did not include a seminoma component (28 years). Among mixed TGCTs without a seminoma component, the combination of embryonal carcinoma and teratoma had the highest age-standardized incidence rate (IR=4.9). Among mixed TGCTs without a seminoma component that included three histologies, the combination of embryonal carcinoma, yolk sac tumor, and teratoma had the highest incidence rate (IR=1.7). A few combinations of histological elements were associated with a very low median age at diagnosis (25 years) including teratoma plus yolk sac tumor, embryonal carcinoma plus yolk sac tumor plus choriocarcinoma, and the combination embryonal carcinoma, yolk sac tumor, teratoma, and choriocarcinoma. Age variability at diagnosis was greater in mixed TGCTs without a seminoma component than in those with a seminoma component; furthermore, this age variability decreased with increasing number of histological elements in mixed TGCT without seminoma, while this variability remained fairly constant in mixed TGCT with a seminoma component (Table 1).

Table 1.

Age distribution and incidence rates (cases per million person-years) of 1,172 primary malignant testicular germ cell tumors with more than one histology, North Rhine-Westphalia, Germany, 2008–2013

Histology	Cases		Age (years)				Crude Rate		Age-standardized
	N	%	P25	Med	P75	CV%	Rate	SE	Rate	SE
All mixed germ cell tumors	1,172		25	31	39	32	22.7	0.7	22.4	0.7
Mixed tumors with seminoma 1)	550	100	28	34	42	29	10.6	0.5	10.5	0.5
Two types	308	56.0	29	34	42	27	5.9	0.3	5.9	0.3
SEM + EC	217		29	33	41	27	4.2	0.3	4.1	0.3
SEM + TER	77		29	34	41	27	1.5	0.2	1.5	0.2
SEM + YST	10		33	43	48	19	0.2	0.1	0.2	0.1
SEM + CHOR	4
Three types	166	30.2	27	34	43	32	3.2	0.3	3.2	0.3
SEM + EC + TER	87		25	32	42	31	1.7	0.2	1.7	0.2
SEM + EC + YST	49		28	35	43	30	1.0	0.1	1.0	0.1
SEM + YST + TER	18		26	30	43	38	0.4	0.1	0.3	0.1
SEM + EC + CHOR	11		29	33	52	37	0.2	0.1	0.2	0.1
SEM + YST + CHOR	1
Four types	68	12.4	27	34	45	30	1.3	0.2	1.3	0.2
SEM + EC + YST + TER	46		28	36	44	28	0.9	0.1	0.9	0.1
SEM + EC + TER + CHOR	12		24	26	32	25	0.2	0.1	0.2	0.1
SEM + EC + YST + CHOR	8		32	48	53	28	0.2	0.1	0.1	0.1
SEM + TER + YST + CHOR	2
Five types
SEM + EC+ YST + TER+ CHOR	8	1.5	26	28	40	27	0.2	0.1	0.2	0.1

Histology	Cases		Age (years)				Crude		Age-standardized
	N	%	P25	Med	P75	CV%	Rate	SE	Rate	SE
Mixed tumors without seminoma 3)	547	100	23	28	36	35	10.6	0.5	10.4	0.5
Two types	380	69.5	23	28	37	36	7.3	0.4	7.2	0.4
EC + TER	259		22	27	37	36	5.0	0.3	4.9	0.3
EC + YST	68		25	32	38	31	1.3	0.2	1.3	0.2
TER + YST	24		20	25	33	45	0.5	0.1	0.5	0.1
EC + CHOR	15		22	28	43	47	0.3	0.1	0.3	0.1
TER + CHOR	9		24	34	40	36	0.2	0.1	0.2	0.1
YST + CHOR	4
CHOR + further nonseminoma type	1
Three types	134	24.5	23	28	35	32	2.6	0.2	2.6	0.2
EC + YST + TER	98		23	29	35	27	1.9	0.2	1.9	0.2
EC + YST + CHOR	18		21	25	33	47	0.4	0.1	0.4	0.1
EC + TER + CHOR	13		25	28	41	42	0.3	0.1	0.2	0.1
TER + YST + CHOR	5		30	32	35	32	0.1	0.04	0.1	0.04
Four types
EC + YST+ TER + CHOR	33	6.0	21	25	30	28	0.6	0.1	0.6	0.1

age standard: EUROSTAT revised European Standard Population of 2013; SEM: seminoma; EC: embryonal carcinoma; YST: yolk sac tumor; TER: teratoma; CHOR: choriocarcinoma; for combinations with less than 5 cases, the age distribution and rates are not presented;

¹⁾Histological information of 75 mixed germ cell tumors was too incomplete to further subtype the histological elements (for N=30 mixed tumors, the histological components were not specified; for N=9 mixed tumors with seminoma, the remaining histological components were not specified; for N=36 mixed tumors without seminoma, the remaining histological components were not specified).

The percentage distribution of histological types among pure TGCTs and among mixed TGCTs differed considerably. Among pure TGCTs, seminoma was the predominant histology (82.4) while all nonsemimona histologies were rare especially yolk sac tumors (0.6%) and choriocarcinoma (0.6%). Among mixed TGCTs, the percentage of seminoma components was only 46.9% whereas the percentage of nonseminoma histologies was considerably higher than among pure TGCTs with embryonal carcinoma and teratoma as the dominating histologies (Table 2).

Table 2.

Percentages of germ cell tumor histologies among all testicular germ cell tumors, pure, and mixed germ cell tumors, North Rhine-Westphalia, Germany, 2008–2013

	Percentages
Histological component	among pure TGCTs (n=3,763)	among mixed TGCTs (n=1,172)
Seminoma	82.4	46.9
Spermatocytic tumor	0.8	0
Embryonal carcinoma	10.8	80.4
Yolk sac tumor	0.6	33.4
Teratoma	3.5	59.0
Choriocarcinoma	0.6	12.3
Well-differentiated neuroendocrine tumor	0.05	0
Non-seminomatous GCT, NOS or mixed GCT NOS	1.3	6.4

NOS – not other specified; TGCT: testicular germ cell tumor; percentages among mixed TGCTs do not add up to 100% because mixed tumors have more than one histology.

Discussion

The application of the updated WHO classification resulted in a re-coding of 8.9% of all testicular tumors. Testicular germ cell tumors coded as 9081/3 (teratocarcinoma) and 9101/3 (choriocarcinoma combined with other germ cell elements) were the most frequent reasons for a re-coding according to the updated WHO classification. Both entities are now coded as mixed germ cell tumors. While the recodings have no influence on the incidence of seminomatous and non-seminomatous TGCTs, they influence the incidence of histological subtypes of seminomatous and non-seminomatous TGCTs. Trend distortions can be expected if time trends of histological subtypes of the seminomatous and non-seminomatous TGCTs are examined.

To the best of our knowledge, this is the first report of population-based incidence rates of TGCT subtypes according to the updated WHO classification of testicular tumors. Only one population-based in-depth analysis of the histological components of mixed TGCTs has been previously performed. In the “Danish Testicular Carcinoma Project (DATECA)”, 1,058 consecutive testicular TGCTs diagnosed between 1976 and 1980 were registered in a special registry and underwent review by reference pathologists (Krag Jacobsen, 1984). Overall, 34% of all TGCTs in DATECA were mixed TGCTs whereas this percentage was 24% in our analysis.

In contrast to the DATECA project, the registered TGCTs at the cancer registry of NRW did not undergo an additional pathology review of histological slides or tissue blocks by a panel of pathologists. We were only able to review testicular tumor reports in order to extract the reported histological elements of TGCTs. Therefore, DATECA achieved a higher sensitivity in the detection of more than one component. Incorrect ICD-O coding of mixed TGCTs in the NRW cancer registry is an unlikely explanation for the lower percentage of mixed TGCTs than in DATECA as the percentage of miscoding of mixed TGCTs as pure seminoma was only 1% in our study. If we would consider 1% of the 3,101 tumors erroneously coded as pure seminoma as mixed TGCTs, the percentage of mixed TGCTs would increase from 23.7% to 24.4%, still considerably lower than in the DATECA project. Interestingly, the distribution of the number of histological elements among the mixed TGCTs with and without seminoma was relatively similar in DATECA and NRW (Supplementary Figure 1). Alternatively, however less likely, the distribution of risk factors for TGCT may differ between Denmark in the 1970s and Germany in 2008–2013. The fact that Danish testicular tumor incidence rates in the 1970s were considerably higher than in Germany suggests that the TGCT risk factor distributions between these countries may have differed and could have produced different distributions of histological types of testicular tumors in these two countries.

The WHO states “after seminoma, EC is the most prevalent germ cell tumour (GCT) of the testis. Although only 2.3–16% of testicular GCTs are pure ECs, EC occurs in 40% of all GCTs and in 87% of the non-seminomatous tumours” (Ulbright, 2016) Our study reveals that embryonal carcinoma is relatively rare (10.8%) among incident pure TGCTs. However, among mixed TGCTs, embryonal carcinoma is the most frequent histological element (80.4%).

TGCTs that are diagnosed in post-pubertal adolescents or young adults arise from germ cell neoplasia in situ (GCNIS) that originate from arrested gonocytes that failed to mature to spermatogonia (Skakkebaek, 1972). GCNIS that proliferate transform to seminoma, whereas GCNIS that acquire reprogramming transform to non-seminoma (Oosterhuis and Looijenga, 2005) (Rajpert-De Meyts et al., 2016). TGCTs containing nonseminomatous components are clinically diagnosed at a substantially earlier age than pure seminoma. Pathogenetically, this observation may imply that TGCT growth associated with reprogramming is faster than the TGCT growth associated with proliferation, resulting in younger ages at diagnosis of nonseminoma (Oosterhuis et al., 1989).

The strengths of our study include the population-based approach, the study size, and the review of clinical and pathological documents from 42.9% of all registered cases. However, our study suffers from some limitations. First, our study did not include a review of histological slides by a panel of reference pathologists and our review was only based on available pathology and clinical reports. Second, our in-depth analysis of histological components as reported in pathology reports was not possible for 5% of testicular tumors. Furthermore, 48 out of 5,252 TGCTs (0.9%) could only be categorized as nonseminomatous TGCTs, not otherwise specified, as reporting of the exact histology to the cancer registry was incomplete. Third, we expect some diagnostic error of the histological subtypes as the histopathology assessment was not quality-assured by a panel of testicular tumor pathologists. Recently, Sharma et al. reviewed histological slides of 235 cases of TGCT. Discrepancies occurred in 50 (21%) cases. Histopathological discrepancies occurred most often among mixed TGCTs (17 out of 50, 34%) with discrepancies mainly in germ cell elements (15 out of 17) (Sharma et al., 2015).

In conclusion, we provided for the first time a conversion table of ICD-O codes of TGCTs in order to provide TGCT statistics according to the updated WHO classification of testicular tumors and presented detailed population-based incidence estimates of testicular tumors subtypes. Population-based incidence time trends for seminomatous and non-seminomatous TGCTs that include mixed TGCTs are not distorted by the introduction of the WHO update. Trend distortions can only be expected if time trends of individual histological subtypes of the seminomatous and non-seminomatous TGCTs are examined.