Testicular germ cell tumors (TGCT) are the most frequent solid malignancies in young adult male patients, typically affecting individuals between 15 and 40 years. Although they account for only 1% of all male cancers, their incidence has been steadily rising in Western countries over the past decades (1, 2). TGCTs are highly curable, with long-term survival exceeding 95% in the early stage of the disease and approaching 90% at 10 years, even in advanced stages, largely due to the effectiveness of cisplatin-based chemotherapy (1, 3, 4). Given the young age of the affected population and the high curability of the disease, optimizing supportive care and minimizing treatment-related morbidity has become increasingly important to preserve long-term health and quality of life (5).

Among the potential complications of treatment, thromboembolism (TE)—encompassing deep vein thrombosis (DVT), pulmonary embolism (PE), and other thrombotic events—has emerged as a clinically relevant and increasingly recognized concern in patients with TGCT undergoing systemic therapy. The incidence of TE is approximately 10% in patients with advanced TGCT, occurring mostly during chemotherapy treatment, which is notably high compared to other solid tumors, particularly accounting for the younger age and low comorbidities of these patients (6, 7). Major thrombotic risk factors include bulky retroperitoneal lymph nodes, elevated tumor markers after orchiectomy, advanced or poor-risk disease according to the International Germ Cell Cancer Collaborative Group (IGCCCG) classification, use of central venous catheters (CVC), and the thrombogenic effects of cisplatin-based chemotherapy (7–9).

The biological mechanisms underlying cancer-associated thrombosis are multifactorial and include tumor-related hypercoagulability, endothelial dysfunction, chemotherapy-induced vascular injury, and pro-inflammatory cytokine release (10, 11). In TGCT specifically, bulky retroperitoneal lymph nodes may cause mechanical compression of major vessels, further predisposing patients to thrombosis, while cisplatin independently increases the risk of both arterial and venous events through its pro-thrombotic effect driven by several distinct biological mechanisms (7, 8, 12).

Firstly, cisplatin exerts direct toxic effects on the vascular endothelium, leading to endothelial cell apoptosis and the subsequent exposure of the subendothelial matrix (13), which triggers the release of pro-coagulant ultra-large von Willebrand factor multimers and activates the extrinsic coagulation pathway. Moreover, cisplatin therapy has been associated with the induction of pro-inflammatory cytokines and the generation of microparticles that further amplify thrombin production and platelet activation (14, 15). Understanding these mechanisms is crucial, as the peak of this vascular damage often occurs early during the first cycles of treatment, coinciding with the highest risk period for clinical thromboembolic events (6).

Risk prediction in this setting remains challenging. While several predictive models, such as the Khorana Score and OnkoTEV Score, have been developed to estimate the risk of TE in cancer patients receiving chemotherapy, their performance in TGCT populations remains uncertain. These models typically incorporate variables such as tumor site, blood counts, body mass index, and other clinical parameters, yet they may lack sensitivity and specificity in younger patients with highly chemosensitive malignancies (16, 17). Furthermore, pharmacological thromboprophylaxis in high-risk populations has been evaluated in small or non-randomized studies, yielding conflicting results on efficacy and bleeding risk (6, 18). The most recent ESMO-EURACAN guidelines recommend considering thromboprophylaxis with low molecular weight heparin (LMWH) in patients with advanced TGCT under chemotherapy (excluding those with central nervous system and/or hemorrhagic lesions), especially for those who have a Khorana score of 3 or more, retroperitoneal lymph nodes that are more than 3.5 cm, stage III TCGT or IGCCCG poor prognosis group (level of evidence: IIA). Peripheral venous access should be used instead of an indwelling vascular access device (19, 20).

The clinical consequences of TE should not be neglected, as it can delay surgical interventions or chemotherapy in a highly curable disease; mobility and pain may negatively impact the quality of life (22), and it can also be a life-threatening event, being one of the leading causes of noncancer deaths among cancer patients (6, 9). Two recent large studies reported a five- to sevenfold increased risk of death from cardiovascular disease, including TE, during the first year after cisplatin-based ChT (21, 22). Data from the GCT Spanish Group point to a significant reduction in progression-free survival (PFS) (HR 2.29, p = 0.02) and OS (HR 5.14, p <.001) once TE has been documented (23).

Despite these findings, prospective randomized studies and real-world data specific to TGCT patients are scarce. In this study, we report the incidence and clinical characteristics of TE and explore its impact on survival outcomes. A better understanding of TE risk factors, including tumor burden, treatment intensity, and validated clinical scores such as the Khorana and OnkoTEV, may enhance individual risk stratification. These insights could guide the selective use of prophylactic anticoagulation, particularly in high-risk patients, and contribute to improving both oncologic and supportive care strategies in this unique and highly curable population.

Between January 2007 and December 2021, 160 patients with TGCT were screened; 136 met all of the eligibility criteria and constituted the study population. Seven of these 136 patients (5.1%) experienced a TE event during follow−up, and all events occurred in the subgroup with advanced or recurrent disease (Figure 1).

For subsequent analyses, the patients were divided into two clinical cohorts:

This retrospective multicenter study evaluated the incidence, risk factors, and prognostic impact of TE events in 136 patients with TGCT treated across three Portuguese oncology centers. The overall incidence of TE was 5.1%, with all events occurring in patients with advanced disease, corresponding to 15.2% within this subgroup. No TE events were reported in stage I patients, including those who received adjuvant chemotherapy. These findings are consistent with previously reported TE rates of approximately 10% in advanced TGCT populations, with variation depending on disease stage, treatment intensity, and baseline risk factor distribution (6–9).

The association between disease burden and TE risk was reinforced as a primary driver of these events (7–9). In our cohort, the IGCCCG prognostic category emerged as the most significant clinical predictor for TE events. Poor-risk patients showed a significantly higher incidence of TE compared to good or intermediate-risk patients (57.1% vs. 14.6%; p = 0.003). This was further supported by the significant association with metastatic patterns, where non-pulmonary visceral metastases (p = 0.005) and lung metastases (p = 0.020) were more common in the TE group. Nevertheless, no correlations were found between clinicopathological features and TE occurrence in logistic regression.

All TE cases occurred in patients with CVC, although only a minority were directly attributed to catheter thrombosis, suggesting that CVC use may be a surrogate of advanced disease or treatment intensity rather than an independent risk factor for TE. This observation may be influenced by biases, such as the small sample size and short chemotherapy duration in this population. The TE risk in patients with a CVC increases with treatment duration: it is low in the first month but rises significantly after 120 days of use (24). In TGCT patients, who typically receive only three to four cycles of platinum-based chemotherapy, this shorter exposure likely explains the low rate of catheter-related events compared to other cancers requiring prolonged regimens.

Furthermore, we analyzed patients receiving salvage therapy and HDCT, which require more complex and prolonged CVC placement, to determine whether higher treatment intensity in later lines influenced TE events. Despite this increased complexity, six of the seven TE events occurred during first-line treatment, one during second-line therapy, and not in the single patient who received third-line chemotherapy with HDCT. These findings reinforce the hypothesis that acute vascular toxicity from initial cisplatin exposure and primary bulky disease are the primary drivers of thrombosis in TGCT, surpassing the influence of subsequent treatment duration or intensity.

Unlike large cooperative studies reporting a significant negative impact of TE on survival (23), we did not observe statistically significant survival differences between TE and non-TE patients. The 5-year OS was 71.4% in TE and 86.2% in non-TE group (p = 0.22), while 5-years PFS was 47.6% in TE and 55.8% in non-TE group, with a median PFS of 31.5 and 29.3 months, respectively (p = 0.74). This discrepancy may reflect the limited number of events, which reduced the statistical power, as well as improvements in supportive care. Importantly, one death was directly attributed to TE, highlighting its clinical relevance even in a highly curable malignancy.

The timing and pattern of TE events were also consistent with prior reports, with most occurring within 30 days of chemotherapy initiation and pulmonary embolism being the most frequent manifestation. The early clustering of TE events reinforces the importance of vigilance during the first cycles of chemotherapy, when cisplatin-driven endothelial damage and inflammatory cytokines peak (6–8, 10–13). This period coincides with maximum biological stress, including direct vascular apoptosis and pro-coagulant factor release (14, 15).

Crucially, our study reveals a significant gap in current TE risk assessment tools. While the Khorana and ONKOTEV scores are validated across various malignancies, these scores failed to identify any of the patients who developed TE in our cohort, where 100% (7/7) of TE events occurred in patients classified as “low-risk” by Khorana (p = 0.643) and ONKOTEV (p = 0.603) scores. This suggests that in TGCT, the pro-thrombotic drivers are not the systemic factors measured by these scores (such as BMI or leukocytosis) but rather disease-specific factors like high tumor burden and the acute endothelial toxicity of cisplatin-based regimens. Consequently, we hypothesize that the IGCCCG risk status—which was highly predictive in our study (p = 0.003)—should be considered a more reliable indicator than traditional VTE scores when deciding on clinical monitoring or the need for thromboprophylaxis.

This study provides one of the few real-world, multicenter analyses of TE in TGCT within a European and specifically Portuguese population, offering region-specific insights to support clinical decision-making. The availability of detailed clinical annotation and standardized methodology enabled a meaningful assessment of risk factors despite the small number of events.

Clinically, our findings support current guideline recommendations to avoid CVC use whenever feasible (16, 17), highlight the limitations of cancer global thromboembolic risk scores, and suggest that routine prophylaxis is likely unnecessary in stage I disease. In contrast, patients with advanced disease, poor IGCCCG prognosis, visceral metastases, or requiring CVCs represent a higher-risk subgroup that may benefit from closer monitoring and potentially prophylactic strategies.

Several limitations must be acknowledged. The retrospective design is subject to selection and information bias, and the small number of TE events (n = 7) limited the statistical power to conduct multivariable analyses. TE incidence may also have been underestimated, as asymptomatic events were unlikely to be detected in routine clinical practice.

In conclusion, in this cohort of patients with TGCT, TE events occurred exclusively in patients with advanced disease and were strongly associated with tumor burden and central venous catheter use. Importantly, existing risk scores (Khorana and ONKOTEV) failed to predict any events, whereas the IGCCCG poor-risk category emerged as a highly significant predictor. While there were no statistically significant differences in survival between TE and non-TE patients, the recurrence rates were higher and the PFS was numerically shorter in the TE group, supporting its clinical relevance even in a highly curable malignancy. The less pronounced prognostic impact compared with previous large studies likely reflects the limited number of events and improvements in supportive care. Prospective studies are warranted to validate TGCT-specific risk stratification tools and to determine whether risk-adapted thromboprophylaxis can safely reduce TE incidence and improve outcomes in high-risk patients.