Adolescent and Young Adult Germ Cell Tumors: Epidemiology, Genomics, Treatment, and Survivorship

Abstract

Innovations in the care of adolescent and young adult (AYA) germ cell tumors (GCT) are needed for one of the most common AYA cancers for which treatment has not significantly changed for several decades. Testicular GCTs (TGCT) are the most common cancer in 15–39-year-old men and ovarian GCTs (OvGCT) are the leading gynecologic malignancy in women under age 25 years. Excellent outcomes, even in widely metastatic disease using cisplatin-based chemotherapy have been achievable since Einhorn and Donohue’s landmark 1977 study in TGCT. However, as the severity of accompanying late effects (ototoxicity, neurotoxicity, cardiovascular disease, second malignant neoplasms, nephrotoxicity, and others) has emerged, efforts to de-intensity treatment and find alternatives to cisplatin have taken on new urgency. Current innovations include the collaborative design of clinical trials that accrue GCT tumors across all ages and both sexes, including adolescents (previously on pediatric trials), and OvGCT (previously on gynecologic-only trials). Joint trials accrue larger sample sizes at a faster rate and therefore evaluate new approaches more rapidly. These joint trials also allow for biospecimen collection to further probe GCT etiology as well as underlying mechanisms of tumor growth, thus providing new therapeutic options. This AYA approach has been fostered by The Malignant Germ Cell International Consortium (MaGIC) which includes over 115 GCT disease experts from pediatric, gynecologic and genitourinary oncology in 16 countries. Trials in development incorporate, for the first time, molecular risk stratification and precision oncology approaches based on specific GCT biology. This collaborative AYA approach pioneered successfully in GCT could serve as a model for impactful research for other AYA cancer types.

INTRODUCTION

King Pyrrhus of Greece triumphed over a Roman army in a key 279 B.C. battle, but most of his forces were destroyed, and he concluded, “Another such victory and we are undone.” Since that time, a “Pyrrhic victory” refers to one that take such a destructive toll on the victor that it is almost equivalent to defeat. Prior to cisplatin’s introduction into germ cell tumor (GCT) management in the 1970s, only approximately 20% of patients with advanced disease achieved durable remissions.¹ Cisplatin’s addition to the already established regimen of vinblastine plus bleomycin increased durable remission rates to over 80% as detailed in the seminal 1977 publication by Einhorn and Donohue.² While cisplatin has since demonstrated activity in many additional tumor types, the extent and duration of response in metastatic GCT remains unparalleled. Nonetheless, as the long-term sequelae of these highly successful treatments have become apparent,^3,4 it is their prevalence and severity that has driven many of the changes in management. No new agent or regimen has emerged for these patients since the early 2000s when oxaliplatin was introduced. Thus, in the first- and second-line settings as well as post-high-dose chemotherapy, standard chemotherapy regimens have remained largely unchanged for over 20 years.

Here we provide an overview of GCT epidemiology, genomics, therapeutic approaches, ongoing clinical trials, and survivorship, citing studies representative of recent literature. GCT comprise a paradigm of innovation possible through collaborative adolescent and young adult (AYA) efforts. In fact, since the vast majority of GCTs occur in patients aged 15–39, most of what we know about GCTs and the evidence presented herein relates specifically to AYA populations with few exceptions.

EPIDEMIOLOGY AND GENOMICS

Testicular GCTs.

All GCT originate from primordial germ cells. Testicular GCTs (TGCTs) are the most common cancer in AYA males aged 15–39 year.^5,6 U.S. TCGT incidence rates per 100,000 men are highest among non-Hispanic white men (7.0), followed by American Indians/Alaska Natives (4.8), Hispanics (4.2), Asians/Pacific Islanders (2.0), and black men (1.2).⁷ A growing body of evidence reviewed elsewhere,⁷ supports the hypothesis that most TGCT are connected to testicular dysgenesis and originate in utero, but that postnatal events determine in part which men will eventually develop TGCT.

The high genetic heritability of TCGT is well-documented,⁸ with the relative risk of developing TGCT in first-degree relatives of affected men between 6 and 10.^9–11 To date, genome-wide association studies (GWAS) have identified 78 independent susceptibility loci for TGCT accounting for 44% of disease heritability.¹² These loci are noteworthy for their relatively large per allele effect sizes (odds ratios: 1.4–3.0) and because they are located in genetic regions highly relevant to germ cell development. Associations with some of these variants were recently confirmed in pediatric and adolescent GCTs,¹³ with some variants showing similar effect sizes in both pre- and post-pubertal GCTs while others were stronger in the post-pubertal age group. Notably, a sequencing study in adult TGCT did not find evidence for rare mutations with large genetic effects.¹⁴ Thus, available data support polygenic inheritance with substantial contributions from common variants, with many not yet identified.¹⁵

Additional genomic hallmarks of GCT, observed in nearly all post-pubertal tumors, include aneuploidy and gain of chromosome 12p, most commonly as isochromosome (i[12p]).¹⁶ Despite knowledge of the ubiquitous 12p gain in GCT for over 3 decades, the specific gene(s) on 12p responsible for tumorigenesis remains unclear. In fact, one hypothesis is that 12p gain represents merely a bystander effect, reflecting dysfunctional meiosis and not a pathogenic role. Nevertheless, given its specificity, 12p gain remains a useful marker for determining whether poorly differentiated malignancies are of GCT origin.¹⁷

Ovarian GCTs.

Ovarian GCT (OvGCTs) have an overall incidence rate of 1.2 cases per 100,000 women aged 15–19.¹⁸ Incidence peaks in adolescence and young adulthood, and under age 25 years, OvGCTs are the leading gynecologic cancer.¹⁹ OvGCT prevalence as a proportion of all ovarian tumors is lower in North America (2%) and Europe (1.3%) vs. Central and South America (3.9%) and Asia (4.2%)²⁰ although this differs between histological subtypes, as reviewed elsewhere.²¹ OvGCTs originate in utero,²² likely influenced by genetic alterations (e.g., cKIT mutation and amplification in dysgerminoma)²³ combined with external factors.²² They are more common in individuals with gonadal dysgenesis with excess risk from 0–60% depending on the syndrome, with prophylactic gonadectomy advised in high-risk individuals, e.g., Turner’s and Swyer Syndromes.²⁴

OvGCTs are classified by morphological appearance and specific immunohistochemical profiles, which are shared with analogous histologies at other anatomic sites. OvGCTs have low mutation rates, very few recurrent somatic mutations and stable copy number profiles²³. Ovarian yolk sac tumors, like testicular counterparts, are distinguished by recurrent KRAS mutations and PIK3CA and AKT1 amplification, while KIT mutations are commonly observed in ovarian dysgerminomas and testicular seminomas²³. OvGCTs also exhibit chromosome 12p gain, the notable exception being pure mature and immature teratomas (MT, IT).²³ Interestingly, frequent loss of heterozygosity is observed in IT and MT²⁵, but not in OvGCTs with mixed histologies that include teratomatous components²⁶. In addition to indicating common clonal origins of IT and MT²⁶, these biological features separate teratomas from other OvGCT histologies and may have relevance for clinical management.

Pediatric GCTs.

Pediatric GCTs represent 3% of childhood cancers before age 14, but incidence rises with puberty onset at ~age 8 in girls and ~age 10 in boys. In adolescence, GCTs are the most common solid tumor, besides thyroid carcinoma, representing 15% of cancer diagnoses. Incidence rates are similar in males and females diagnosed before age 5 years, but are much higher in males in the adolescent age group.²⁷ Incidence rates differ by race and ethnicity in males vs. females in the adolescent age group²⁷ and largely follow trends described above for adult TGCT and OvGCT. Annually in the U.S., approximately 900 GCTs are diagnosed in children (ages 0–19), with 5,700 testicular and 260 ovarian GCTs in young adults (ages 20–39).²⁸

THERAPEUTIC APPROACHES: OVERVIEW

Management:

GCTs are largely managed by pediatric oncologists, gynecologic oncologists or surgeons, or TGCT specialists (oncologists, urologists). Etoposide and cisplatin-based chemotherapy have anchored systemic treatment for over four decades but are associated with significant toxicity.^3,29,30 Safe approaches to reduce treatment-related toxicity are a current priority. Systematic investigation, however, has been hampered by challenges in obtaining research funding and creating global clinical trial networks for these cancers, which are rare compared with other adult malignancies. In particular, progress for OvGCT and pediatric GCT has lagged behind TGCT due to smaller populations. Nonetheless, there have been recent advances in OvGCT management, including changes in pathologic classification³¹ and emerging evidence to support different approaches for these distinct histologic subtypes.³²

Cisplatin sensitivity:

A unique, defining GCT characteristic is their exquisite cisplatin sensitivity, allowing cure to be achieved even in widely disseminated disease. Therefore, long sought-after goals have been to understand the etiology of this sensitivity and mechanisms of resistance in the minority who are not cured. While this goal remains elusive, several important clues have emerged over the last three decades. Cisplatin interacts directly with DNA to form intrastrand and interstrand adducts, disrupting DNA structure, resulting in failed recognition by DNA repair proteins, inhibition of transcription and DNA replication, and ultimately apoptosis. Proposed mechanisms of cisplatin sensitivity include low levels of DNA repair proteins, including those involved in nucleoside excision repair,³³ defective homologous recombination,³⁴ and primed apoptotic response.^35,36 These processes are largely dependent on intact TP53 pathways, which in contrast to most malignancies, are almost always present in GCT. TP53 mutations occur exclusively in cisplatin-resistant GCT and interestingly, are most frequent in primary mediastinal nonseminoma (~70%), which are characterized by poor response rates and unfavorable prognosis.³⁷ Amplification of MDM2, which downregulates TP53, has also been associated with cisplatin resistance in advanced GCT.³⁷

MicroRNAs:

An exciting development in the last decade has been the identification of circulating microRNAs in plasma or serum of testicular GCT patients; the utility in pediatric and OvGCT remains to be defined, and is part of the goal of Children’s Oncology Group (COG) AGCT1531. These small noncoding RNAs inhibit translation by binding to complementary messenger RNA sequences. Among the miRNA 302 and 371–373 clusters identified specific to GCT,³⁸ miRNA 371a-3p has emerged as the most promising new GCT marker. Among 522 TGCT patients, the sensitivity, specificity, positive predictive value, and negative predictive value of serum miRNA 371a-3p for GCT diagnosis were 92%, 96%, 97%, and 83%, far exceeding AFP, HCG, and LDH individually and combined.³⁹ Numerous potential applications of this novel biomarker exist include use as a diagnostic test, a prognostic factor, a means to assess treatment response, and detection of minimal residual disease, and as part of surveillance to potentially reduce the frequency of currently recommended axial imaging.

Unified approach:

What has become increasingly clear is the need for a unified approach to manage GCT. The future holds promise for the ability to molecularly stratify patients and go beyond clinical characteristics, histopathology, and serum tumor markers. Trials in development focus on biospecimen collection and incorporation of genomic alterations into risk stratification for treatment; other adaptations have incorporated all patients with post-pubertal GCT, including adolescents (traditionally treated on pediatric trials), and women with GCT (typically treated on gyn-oncology only trials), justified by cross-comparisons showing that GCT outcomes are similar when comparing by age/stage/risk group, independent of sex and primary tumor site. Moreover, this joint approach has facilitated more rapid accrual⁴⁰, allowing contemporary trials to be designed with larger sample sizes, thus providing more power to detect treatment differences. This contrasts to several important TGCT trials which closed prematurely due to poor accrual.⁴¹ A unified approach also allows for biospecimen collection to continue to delve into etiology and molecularly-based treatment options; for example, 3 cross-age, cross-sex National Clinical Trials Network (NCTN) trials are either underway or recently completed (NCT03067181, NCT02582697, NCT02375204), designed by the Malignant Germ Cell International Consortium (MaGIC) that together will accrue nearly 3,000 patients. Two new concepts under review at NCTN have been co-designed and co-sponsored by pediatric, gynecologic, and genitourinary oncology.

TREATMENT

First-line:

Standard of care, first-line chemotherapy management for TGCT and post-pubertal extragonadal GCT is determined by the International Germ Cell Cancer Collaborative Group (IGCCCG) prognostic model, which divides patients into good-, intermediate- and poor-prognosis with cure rates of approximately 90%, 80%, and 55%, respectively.^42,43 For good-risk patients, three cycles of bleomycin, etoposide, and cisplatin (BEP×3) or four cycles of etoposide and cisplatin (EP×4) is standard whereas for intermediate- and poor-risk patients, BEPx4 or four cycles of etoposide, ifosfamide, and cisplatin (VIP×4) (for those unable to receive bleomycin) are recommended. Efforts to decrease treatment intensity and toxicity for patients with good-risk disease, including etoposide dose reductions and substitution of carboplatin for cisplatin, led to decremental progression-free survival and in some series, overall survival.^44–46 Similarly, attempts to escalate treatment intensity to improve efficacy in intermediate- and poor-risk patients failed to improve upon BEP results, but led to excess toxicity.^47–49 Although risk stratification schemes have been devised for adult OvGCTs,^50,51 these have not been adopted in clinical practice, and NCCN and ESMO guidelines both recommend BEP for all stages of non-dysgerminoma, regardless of histology and without specifying the optimal number of cycles.

In contrast to men with TGCT, in the UK, children and adolescents with GCT had excellent outcomes with first-line carboplatin-based regimens, notably delivered at higher dose and dose-intensity than in the TGCT trial. This observation led to agreements between COG in the U.S., Canada, and Australia (where BEP was standard) and the Children’s Cancer and Leukemia Group, UK (where carboplatin-based treatment was standard) to combine past clinical data to compare outcomes between cisplatin- and carboplatin-based regimens. This collaboration started the MaGIC consortium which now includes over 115 GCT experts in 16 countries representing numerous disciplines comprising clinical care and research (Figure 1). MaGIC demonstrated no significant difference in outcome by age or risk group in children and adolescents administered carboplatin vs. cisplatin.⁵² This equipoise justified the current randomized clinical trial (RCT), AGCT1531 (NCT03067181), comparing BEP to BEC (bleomycin, etoposide carboplatin) in both pre-pubertal children and AYAs up to age 25 years.

Figure 1.

The Malignant Germ Cell International Consortium, June 2023

Similarly, MaGIC showed that adolescent males over age 11 years with TGCT had outcomes similar to those expected in adult TCGT using the IGCCCG, and that adolescent females with FIGO Stage IV OvGCT had outcomes equivalent to men with IGCCCG poor-risk tumors.⁵³ These observations rationalized the inclusion of adolescent males and females onto a clinical trial for IGCCCG intermediate- and poor-risk patients developed by the Australian New Zealand Urogenital and Prostate Clinical Trials Group (ANZUP) comparing standard BEP given every 3 weeks to compressed BEP given every 2 weeks (P3BEP, AGCT1532, NCT02582697). This study had originally been limited only to adult men with TGCT. Notably, AYA accrual from pediatric clinical trial organizations account for 30% of total enrollments to date, showing that cooperation between oncologic sub-specialties can enhance accrual.

Second- and later-line chemotherapy:

For patients progressing after first-line chemotherapy, second-line chemotherapy consists of either conventional-dose chemotherapy (CDCT) regimens such as paclitaxel, ifosfamide, and cisplatin (TIP) or vinblastine, ifosfamide, and cisplatin (VeIP), or alternatively high-dose chemotherapy with autologous stem cell transplant (HDCT/ASCT).^54–57 Due to a lack of randomized trials, no optimal CDCT or HDCT/ASCT regimen has been determined. In addition, whether HDCT or CDCT is the preferred second-line strategy remains controversial with retrospective studies consistently favoring HDCT/ASCT,^58,59 but the one RCT demonstrating no benefit to a single HDCT cycle over initial salvage CDCT.⁶⁰ International collaboration facilitated the design and successful accrual completion of the phase III TIGER trial (A031102, E1407, NCT02375204) to definitively answer this question. Patients were randomized to receive either TIPx4 or HDCT/ASCT with the TI-CE regimen (paclitaxel plus ifosfamide followed by high-dose carboplatin and etoposide), with the primary endpoint of overall survival. Based on input from MaGIC and COG, the lower study age limit was reduced from 16 to 14 years; it opened in 13 countries across 3 continents without pharmaceutical company support, sponsored by the NCTN’s adult and pediatric cooperative groups, European Organisation for Research and Treatment of Cancer (EORTC), ANZUP, and the philanthropic organization, Movember.

Patients progressing after HDCT/ASCT are generally considered incurable with most dying from their disease. GCT death typically occurs in the 20s and 30s, and as such, accounts for the greatest average number of life-years’ lost of any non-childhood malignancy.^61,62 Options in this setting are inadequate, but include gemcitabine plus oxaliplatin, gemcitabine plus paclitaxel, and oral etoposide.^63–66 No new agent or regimen has emerged for these patients since the early 2000s when oxaliplatin was introduced. In summary, in the first- and second-line settings as well as post-HDCT, standard chemotherapy regimens have remained unchanged for over 20 years.

Currently, two trials in development bring molecular risk stratification into GCT treatment for the first time. A trial under Eastern Cooperative Oncology Group (ECOG) review proposes to use post-orchiectomy serum miRNA 371a-p to counsel Stage I patients regarding retroperitoneal lymph node dissection, based on a retrospective study in which miR 371a-3p was 100% sensitive and 90% specific in predicting which patients had viable GCT in resected nodes.⁶⁷ A second trial under review by Alliance for men and women with IGCCCG intermediate- or poor-risk disease will use identification of TP53 pathway alterations³⁷ as well as slower than expected serum tumor marker decline after 1 chemotherapy cycle⁶⁸ to select patients for intensified therapy.

In the relapsed/refractory setting (progression post-HDCT), MaGIC has been fostering basic and translational science collaborations needed to develop novel rational agents and simultaneously working to develop a clinical trials consortium to facilitate more rapid drug study. Promising agents under development include demethylating agents, WNT inhibition, CAR-T strategies targeting claudin 6, and therapies targeting glypican-3.

SURVIVORSHIP

As reviewed by Fung,⁶⁹ ototoxicity, neurotoxicity, cardiovascular disease (CVD), second malignant neoplasms (SMN), pulmonary toxicity, nephrotoxicity, infertility, hypogonadism, depression, anxiety, and cognitive impairment have emerged as important survivorship concerns for GCT survivors affecting quality of life and functional status, with the majority of data generated in AYA TGCT survivors. Long-term platinum retention may contribute in part to several long-term toxicities.⁷⁰ Key research findings in recent publications from the three European cohorts of testicular cancer survivor cohorts are summarized in Table 1 and briefly described below, along with major results from the multi-institutional U.S.-based TGCT survivor study (the “Platinum Study”). The latter clinical cohort includes patients from 8 major cancer centers, including Indiana University and Memorial Sloan Kettering Cancer Center,^3,4 and importantly, a small subset is non-white. The type, number, and prevalence of a wide range of adverse health outcomes (AHOs) in 952 adult TGCT survivors given cisplatin-based chemotherapy in this cohort were described in detail.³ Over one-third of TGCT survivors (median age: 37) reported ≥3 AHOs with 12.5% reporting ≥5 AHOs. The most common AHOs were tinnitus (37.1%), hearing loss (HL) (31.5%), obesity (30.9%), and peripheral neuropathy (27.0%). After EPX4 or BEPX3, types and prevalence of individual AHOs were similar (P >0.05), except for Raynaud phenomenon (11.6% v 21.4%; P <0 .01), peripheral neuropathy (29.2% v 21.4%; P =0.02), and obesity (25.5% v 33.0%; P = 0.04). Risk factors predicting higher numbers of AHOs/patient included larger cumulative bleomycin doses, increasing age, current or former smoking, lower physical activity levels, and less than college education. Self-reported health was excellent/very good in approximately 60% of TGCT survivors, but declined as AHOs increased (P <0.001). Similarly, chemotherapy with BEP and more than one line of chemotherapy (OR 2.4 to 4.7 depending on treatment intensity, P<0.001); current smoking status (OR=1.7), and Charlson comorbidity (OR=2.1) were associated with significantly increased neurotoxicity in the Danish Testicular Cancer Late Effect Treatment Cohort.⁷¹ Within the Dutch Testicular Cancer Survivor case-cohort study, cisplatin-based chemotherapy (HR=1.9); obesity (HR=4.6) tobacco use (HR=1.7); Raynaud’s phenomenon (HR=1.9); dyslipidemia (HR=2.8); and positive CVD family history (HR=2.9) were associated with significantly higher CVD risks.⁷²

Table 1:

European Studies of Testicular Cancer Survivors (TCS)

Study Cohort	Study Methods	Treatment Dates	Number of Survivors	Age at TC diagnosis (years)	Outcome(s) in Recent Publication	Overview: All Cohort Outcomes
National Norwegian Testicular Cancer Survivor Study73,86	Longitudinal study: Identified by Norwegian Cancer Registry for survivors treated for unilateral TC at 4 Norwegian university hospitals. In 1998, eligible TCS were invited by mail to complete a survey and then invited to undergo physical examination including blood tests. Subsequently, 2 more surveys administered: 2007–2009 and 2015–2016.	1980–1994	First survey: 1,813 invited, 1,436 responded (79%) Second survey: 1,371 invited, 1,050 responded (77%) Third survey: 963 invited, 783 responded (81%)	Responders for first survey: Median: 33 SD: ± 8.9	Cumulative overall 20-year mortality: 14% (95% CI 11.8 to 16.8) compared with expected rate: 11%. Significantly increased overall mortality was related to increasing age (HR=1.08), cisplatin-based chemotherapy with cumulative dose >630 mg (HR=1.97), major comorbidity (HR=1.93), lower socioeconomic status (HR=1.97), unhealthy lifestyle (HR=1.65), and depressive disorder (HR=1.86). Cancer-related mortality. Higher cancer-related mortality was associated with cisplatin-based chemotherapy with cumulative dose >630 mg (HR=3.65) and depressive disorder (HR=2.21).	Results of physical examination, including blood tests Survey: socioeconomic status, lifestyle behaviors, major somatic co-morbidity (including cisplatin-related toxicities- neuropathy, ototoxicity, Raynaud’s phenomena, cardiovascular disease, second malignant neoplasms), hypertension, mental distress, fatigue and health-related quality of life.
The Danish Testicular Cancer Late Effect Treatment Cohort (DeTeCa Late)71,87	Cross-sectional study: Identified by applying these inclusion criteria to DeTeCa database: Germ cell cancer diagnosis, Danish citizenship, follow- up and medical treatment at oncology ward in Denmark. Postal invitation sent to all eligible TCS: November 2014. Clinical data collected using hospital files, pathology reports.	1984–2007	4,271 TCS invited to participate, 2,572 responded (60%).	Mean: 53 Range: 25–95	Neurotoxicity: Significantly increased neurotoxicity was associated with chemotherapy with BEP and more than one line of chemotherapy (OR 2.4 to 4.7 depending on treatment intensity, P < 0.001); current smoking (OR 1.7), and Charlson comorbidity score (OR 2.1). Radiotherapy and surveillance were not associated with neurotoxicity. After receiving BEP, neurotoxicity was highly associated with all indicators of worse quality of life after approximately 20 years of follow-up (P-trend:1.5 × 10−17 to 1.1 × 10−28)	Clinical data obtained from hospital files, pathology reports. Survey: quality of life, fatigue, symptoms of testosterone deficiency, erectile dysfunction, psychological distress, depression, and anxiety. Additional patient-reported outcomes: education, occupation, income, alcohol, tobacco, drug use and abuse, self-rated health, physical activity, anthropometrics, pain, infertility.a Biobank: DNA samples from sputum or blood.
Dutch Testicular Cancer Survivor Study72	Case-cohort study: Five Dutch TC treatment centers provided access to medical records of TC patients. Those developing CVD after TC diagnosis were identified via regular follow-up and compared with 15% of base cohort in 3 of 5 treatment centers and 25% of base cohort in remaining 2 centers. Treatment data abstracted from medical records; all surviving patients in case-cohort study invited to complete risk factor survey: 2015–2017.	1976–2007	Entire case-cohort:Cases with CVD: 272 Base cohort: 925 Case-cohort with completed risk factor survey:Cases with CVD: 120 Base cohort: 447	For entire 4,748 patients: Mean: Not provided Range: 12–50	Cardiovascular disease risks: Significantly higher CVD risks were associated with cisplatin combination chemotherapy (HR=1.9); obesity at TC diagnosis (HR=4.6); smoking at TC diagnosis (HR=1.7); development of Raynaud’s phenomenon (HR=1.9); dyslipidemia (HR=2.8); and positive family history for CVD (HR=2.9). Quality of life: TC survivors with CVD reported inferior quality of life on physical domains vs. survivors not developing CVD. Cardiovascular disease risk factors: Among 304 testicular cancer survivors with CVD (median age at clinical evaluation=51), 86% had dyslipidemia, 50% had hypertension, and 35% had metabolic syndrome, irrespective of treatment.	Treatment data from medical records Survey: lifestyle, medication use, family history for CVD, quality of life (SF-36), and known adverse treatment effects: Raynaud’s phenomenon, neurotoxicity, ototoxicity, fatigue. Cardiometabolic study assessment included physical examination, blood pressure measurement, and fasting blood samples.

Abbreviations: BEP, bleomycin, etoposide, cisplatin; CVD, cardiovascular disease; HR, hazard ratio; OR odds ratio; TC, testicular cancer; TCS, testicular cancer survivors.

^a.Refer to Table 2 in original manuscript by Kreiberg et al.⁸⁷ for comprehensive outline of outcomes studied. These span physical health (e.g., neurotoxicity, fatigue, infertility before and after testicular cancer diagnosis, symptoms of testosterone deficiency and erectile dysfunction) and psychosocial health (e.g. psychological distress; anxiety; alcohol and tobacco habits; substance abuse history).

Both the frequency and severity of a wide spectrum of AHOs were combined into one overall cumulative burden of morbidity (CBM) score after 1,214 TGCT survivors were accrued to The Platinum Study.⁴ About one in five patients had a high (15%) or very high/severe (4.1%) CBM score (Figure 2). Significant risks factors for higher CBM scores included VIPX4 (OR=2.0 vs BEPX3) or BEPX4 (OR=1.4 vs BEPX3), older attained age (OR=1.2 per 5 years), current disability leave (OR=3.5), less-than-college education (OR=1.4), and current or former smoking status (OR=1.3). Asian race (OR=0.41) and vigorous physical activity (OR=0.68) were inversely associated with higher CBM scores (Table 2). Worse CBM scores may translate to worse overall mortality. In the Norwegian TGCT study, patients had worse cumulative overall 20-year mortality of 14% (95% CI: 11.8 to 16.8) vs. an expected rate of 11%.⁷³ Major comorbidity (HR=1.93), increasing age (HR=1.08), cisplatin-based chemotherapy with cumulative dose >630 mg (HR=1.97), lower socioeconomic status (HR=1.97), unhealthy lifestyle (HR=1.65), and depressive disorder (HR=1.86) were associated with significantly increased overall mortality.

Figure 2.

Distribution of Cumulative Burden of Morbidity (CBM) Score Among 1,214 Participants in the Platinum Study. The above Figure is reproduced from Kerns et al.⁴ with permission granted by Wolters Kluwer Health, Inc. on May 1, 2023.

Table 2.

Multivariable Ordinal Logistic Regression^a of Factors Associated With CBM Score

	Odds Ratio (95% CI)	p-value
Age at evaluation (per 5 years)b	1.18 (1.10, 1.26)	<0.001
Time since chemotherapy completion, years  < 2  2–5  6–9  10+	Ref. 0.91 (0.68, 1.23) 0.61 (0.42, 0.89) 0.55 (0.38, 0.85)	- 0.54 0.010 0.002
Race  White  Black/African-American  Asian  Other	Ref. 1.56 (0.49, 5.03) 0.41 (0.23, 0.72) 1.05 (0.63, 1.76)	- 0.45 0.002 0.84
Education  College or post-college graduate   Less than college education	Ref. 1.44 (1.11, 1.87)	- 0.006
Current employment status  Employed  Unemployed  Retired  On disability leave	Ref. 0.90 (0.55, 1.47) 1.10 (0.36, 3.39) 3.53 (1.57, 7.95)	- 0.66 0.87 0.002
Smoking status  Never  Current or Former	Ref. 1.28 (1.02, 1.63)	- 0.037
Vigorous physical activity (≥6 METs)c   No   Yes	Ref. 0.68 (0.52, 0.89)	- 0.004
RPLNDd  No  Yes	Ref. 0.88 (0.69, 1.12)	- 0.31
Type of chemotherapye  BEPX3  EPX4  BEPX4  VIPX4	Ref. 1.09 (0.75, 1.60) 1.44 (1.04, 1.98) 1.96 (1.04, 3.71)	- 0.65 0.028 0.039

Abbreviations: BEP, bleomycin, etoposide, cisplatin; EP, etoposide, cisplatin; MET, metabolic equivalent task; RPLND, retroperitoneal lymph node dissection; VIP, etoposide, ifosfamide, cisplatin

^aOdds ratios and p-values are from an adjusted model that includes all other variables listed in the table, as well as enrollment center, with CBM score as the outcome (dependent) variable. The ‘very high’ and ‘severe’ categories were collapsed due to sparse data. Analysis includes 1,013 (83.4%) testicular cancer survivors with non-missing data for all variables in the model.

^bAge at diagnosis was not included in the model given the strong correlation with age at evaluation (r=0.81), which was included

^cSee Table 1, footnote o and Appendix-Supplemental Methods for details on assessment of physical activity.

^dRPLND was retained in the multivariable model to control for potential residual confounding, given its correlation with chemotherapy regimen (P < 0.001); approximately 34%, 55%, 66%, and 44% of TCS treated with BEPX3, EPX4, BEPX4, and VIPX4, respectively, had an RPLND.

^eDisease stage was not significantly associated with CBM score (P = 0.48), suggesting that increased scores after BEPX4 or VIPX4 were not explained by more advanced tumor status

The above table is reproduced from Kerns et al.⁴ with permission granted by Wolters Kluwer Health, Inc. on May 1, 2023.

Cisplatin-related toxicities among TGCT survivors can also negatively impact functional status. In one recent report,⁷⁴ 243 TGCT survivors were evaluated for functional impairment due to cisplatin-related ototoxicity. Approximately one-third (35.8%) of those with HL experienced clinically significant functional impairment. Severity of both HL and tinnitus were significantly associated with worsening scores on the Hearing Handicap for Adults (HHIA) and Tinnitus Primary Function Questionnaire (TPFQ), respectively (P<.0001 each). Survivors with either greater HL or more severe tinnitus were also significantly more likely to report cognitive dysfunction (OR=5.52 and 2.56), fatigue (OR=6.18 and 4.04), depression (OR=3.93 and 3.83), and lower overall health (OR=0.39 and 0.46). These results prompted a clinical recommendation that the HHIA and TPFQ be used to risk-stratify TGCT survivors at clinical follow-up who could benefit from referrals to audiologists for evaluation and management.

In order to elucidate etiopathogenetic mechanisms of AHOs among TGCT survivors, prior investigations have identified significant associations of germline genetic mutations with ototoxicity, peripheral neuropathy, and bleomycin-induced pulmonary toxicity, as reviewed elsewhere.⁶⁹ Single nucleotide polymorphisms (SNPs) in megalin (rs2075252), COMT (rs9332377), TPMT (rs12201199), ACYP2 (rs1872328), WFS1 (rs62283056), and OTOS (rs7606353) are significantly associated with platinum-related ototoxicity. Lower expression levels of RPRD1B and SNPs of both glutathione S-transferase (GST) P1 and RPRD1B are associated with cisplatin-induced peripheral neuropathy. The H63D variant of HFE is significantly associated with higher frequencies of bleomycin-induced pulmonary toxicity vs. wild-type HFE.

Late effects of treatment among children and adolescents with GCT are largely unexplored since they were not previously included in large cohort studies of childhood cancer survivors.⁷⁵ Nonetheless, GCT survivors comprise the 3^rd largest group of U.S. childhood cancer survivors due to the high survival rate, currently numbering close to 40,000 in the U.S. alone.⁷⁶ Because primary treatment regimens for pediatric GCT include the same agents used for adult TGCT (cisplatin, etoposide, and bleomycin), similar late effects are likely to occur. However, because risk may differ with either exposures at younger ages or by sex, efforts are needed to characterize these toxicities.^77,78 A cohort exclusively composed of pediatric and adolescent GCT survivors (the GCT Outcomes and Late effects Data (GOLD) Study)⁷⁹ was recently established using COG registry protocols to identify/recruit patients.⁸⁰ Initial efforts will focus on the most prevalent short-term toxicities, i.e., ototoxicity and neuropathy. Preliminary findings (Table 3) suggest that tinnitus (14%), hearing loss (20%), and neuropathy (9%) occurs less frequently in children vs. adult TGCT survivors, but are still noteworthy. The GOLD study will also permit future investigations of other potential late effects in pediatric GCT survivors with delayed onset, including infertility, CVD, and SMN.

Table 3.

Selected data from the GOLD Study (N=325)

	N (%)
Male sex	165 (51)
Median age at questionnaire completion	21yrs (range 5–30)
Median follow-up time	8yrs (range 4–14)
Tumor Location
Testis	55 (17)
Ovary	86 (26)
Intracranial	122 (38)
Extragonadal	62 (19)
Treatment
Chemotherapy +/− other treatments	244 (75)
No chemotherapy	81 (25)
Hearing lossa	59 (24)
Tinnitusa	34 (14)
Neuropathya	22 (9)

^aRestricted to 244 individuals who reported chemotherapy

CONCLUSIONS AND FUTURE DIRECTIONS

In a 2007 Journal of Clinical Oncology editorial entitled “The Graying of Testis Cancer Patients: What Have We Learned?”, Bajorin⁸¹ noted how the TGCT model established far more than curative chemotherapy combinations in the 1970’s and 1980s. Indeed “the concept of modeling response to chemotherapy using serum tumor markers emerged and risk-directed treatment strategies and risk-directed clinical trials soon followed, particularly to reduce toxicity in highly curable patients.” Bajorin stated that TGCT patients were “directly responsible for these great advances by having volunteered for clinical trials in such high numbers that important clinical advances were made at an unparalleled pace for a rare adult disease, and they helped establish randomized clinical trials as the gold standard in oncology research.” This monumental contribution of these early TGCT treatments to the field of oncology and, importantly, the sacrifices and dedication of the involved patients, should never be forgotten.

Similarly, in 2023, there is another tremendous opportunity for patients and researchers to impact GCT management and outcomes. First, the collaborative, inclusive approach exemplified by MaGIC allows advances to benefit all GCTs in all patient populations, spanning pediatric and AYA age groups, and including men and women. This unified approach also facilitates biospecimen collection allowing deeper interrogation of GCT etiology, biology and mechanisms of toxicity and treatment resistance, allowing for improved prognostication, risk stratification and development of novel treatments. Similar to MaGIC, the Global Germ Cell Cancer Group has fostered international collaborations, playing an integral role in development of the TIGER trial, updating the IGCCCG risk models, and conducting retrospective analyses on important clinical situations, e.g., patients with bone or brain metastases and those relapsing after carboplatin for stage I seminoma or BEP for stage I nonseminoma.^{42,43,82–85} Survivorship research should be built alongside these efforts. Importantly, most GCT treatments are homogeneous and allow in-depth analysis of platinating agent toxicities. Subsequently, etiopathogenetic pathways can be elucidated, which are needed to formulate risk-based, targeted prevention and intervention strategies. Optimally, patients should be administered validated baseline questionnaires at clinical trial enrollment with later surveys addressing potential AHOs, comorbidities, and lifestyle behaviors during life-long follow-up. Interactions between aging and the onset and progression of AHOs after GCT treatment should also be explored, given long patient life spans. GCT populations also provide a prime opportunity for novel interventional strategies, including incorporation of emerging technologies such as radiomics and artificial intelligence. As Bajorin⁸¹ stated in 2007, beyond its established role as a model for a curable neoplasm, TGCT also confers a unique opportunity for conducting collaborative survivorship studies with potential to impact all cancer survivors. Similarly, the synergistic efforts currently being pioneered for all AYA GCT could serve as a model for research in other malignancies. A future in which our patients’ cancer victories are no longer “Pyrrhic” may finally be within reach.