Association between the psoas muscle index and disease progression and mortality in testicular germ cell tumours

Methods

This retrospective study was undertaken by the department of Urology, Bakirkoy Dr. Sadi Konuk Training and Research Hospital, Istanbul, Turkey. The study was approved by the Institutional Review Board of the University of Health Sciences, Bakirkoy Dr. Sadi Konuk Training and Research Hospital.

This retrospective multicentre study included patients who underwent chemotherapy for testicular germ cell tumour and were followed between January 2010 and December 2023. Patients with other diagnoses, those in whom skeletal muscle measurement could not be performed on axial imaging, and those with incomplete data were excluded.

Results

Data from 159 patients who underwent chemotherapy for testicular germ cell tumour were analysed. The mean age of the cohort was 30.4±10.2 yr, and the mean body mass index was 25.8±4.00 kg/m². Of these, 113 patients (71.1%) were classified into Group 1, while 46 patients (28.9%) were classified into Group 2. The median follow-up duration for the entire cohort was 47 months (range: 8–142 months) ( Table I).

Pathological subtypes in the entire patient cohort were distributed as follows: seminoma (40.3%), non-seminoma (18.9%), and mixed germ cell tumours (40.9%). No significant differences were observed between the two groups in terms of pathological type, pathological T stage, clinical N stage, S stage, or overall TNM stage (P>0.05). Although a higher proportion of patients with M1a and M1b stages was observed in Group 2, the difference between groups was borderline insignificant (P=0.055). The number of chemotherapy cycles was significantly greater in Group 2 (P= 0.045). Progression and mortality rates were 10.6% and 1.8% in Group 1, and 26.1% and 8.7% in Group 2, respectively. These differences in progression and mortality rates were statistically significant (P=0.023 and P=0.038, respectively). A comparative analysis of pathological characteristics and staging data is presented in Table I.

Kaplan-Meier curves for progression-free survival and overall survival are shown in Supplementary Figure 2. Cox regression analysis was conducted to identify predictors of progression and mortality. Multivariable analysis for progression revealed three factors significantly associated with disease progression ( Table II). A PMI change greater than 10 % [hazard ratio (HR)=6.499, P<0.001], rete testis invasion (HR=3.459, P=0.007), and the presence of non-seminoma or mixed germ cell tumours (HR=5.777, P=0.020) were identified as independent predictors of progression. Figure presents the ROC curve analysis of PMI and SMI changes for predicting disease progression. The AUC values (95% confidence interval) were 0.985 (0.951–0.998) for PMI change and 0.891 (0.832–0.935) for SMI change. Exact cutoffs with sensitivity/specificity values ​​are given in Table III. Comparison of the AUC values revealed a marginally significant difference between PMI and SMI changes (P=0.038). In Cox regression analysis for mortality prediction, both univariate (HR=5.215, P=0.047) and multivariable (HR=5.994, P=0.049) analyses indicated that only a PMI change greater than 10% was statistically significant ( Table IV).

Close

Figure.

ROC curve analysis comparing the predictive ability of PMI and SMI changes for progressive disease. ROC, receiver operating characteristic.

Export to PPT

Discussion

In this study, reductions in PMI following first-line chemotherapy were identified as independent risk factors for both progression and mortality. PMI change was also found to be a more precise predictive tool than SMI. Mitsui et al¹⁴ analysed data from 50 patients undergoing first-line chemotherapy for testicular germ cell tumour and investigated the relationship between progression and changes in skeletal muscle indices after chemotherapy. In their study, the SMI index—also utilised in our study—was assessed alongside psoas muscle volume (PMV), measured using automated volume analysis software on CT images. Their univariate analysis revealed that poor risk classification according to the IGCCCG, metastases excluding the lungs, and PMV change were significant predictors of progression. However, in multivariate analysis, only PMV change remained significant (HR=4.681, P=0.033). PMV change was identified as a more accurate predictive factor than SMI change (AUC=0.832 vs. 0.695), consistent with our findings. Similarly, Baky et al¹⁵ evaluated patients undergoing post-chemotherapy retroperitoneal lymph node dissection and concluded that reductions in PMI were predictive of postoperative morbidity.

SMI is widely accepted in sarcopenia-related studies due to its strong correlation with total body muscle mass.⁹ However, many studies have assessed SMI alongside PMI, utilising both as indicators of sarcopenia.^13,14,18,23 Research demonstrating a strong correlation between PMI and SMI suggests that PMI can also serve as a reliable surrogate for total skeletal muscle mass.²⁴ In our study, despite the substantial correlation between SMI and PMI, PMI was selected as the primary evaluation metric owing to its practical measurement, simpler clinical applicability, and statistically superior predictive performance.

Many studies define sarcopenic patients by setting a threshold value for sarcopenia and evaluating prognosis accordingly.^10-12 However, studies similar to ours examined the kinetics of indices representing sarcopenia in relation to prognosis.^13-17,25 Fukushima et al²⁵ analysed SMI kinetics in patients with metastatic RCC undergoing cytoreductive nephrectomy and demonstrated that increasing SMI positively affected overall survival. In another study, Miyake et al¹³ examined the effects of abdominal skeletal muscle and adipose tissue on prognosis in patients undergoing radical cystectomy for urothelial carcinoma. Similar to our study, they adopted a 10 % cut-off for muscle volume loss. While sarcopenia (HR 2.2, P=0.03) and psoas volume loss (HR 2.4, P=0.02) were independent predictors of overall survival, neither was significant for disease-specific survival. Phuong et al¹⁶ evaluated changes in skeletal muscle and adipose tissue in 141 patients undergoing cisplatin-based chemotherapy for testicular germ cell tumours and investigated the association with chemotherapy-associated adverse events. While pre-chemotherapy skeletal muscle indicators and post-chemotherapy SMI were not associated with adverse events, a decrease in SMI (odds ratio: 0.89, P=0.02) was independently associated with a higher incidence of adverse events. Similarly, a multicentre study by Buxton et al¹⁷ demonstrated that reductions in SMI following cisplatin-based chemotherapy were associated with an increased risk of grade 3 adverse events in patients with testicular germ cell tumour.

Even with successful primary cancer treatment, obesity and sarcopenia can adversely affect quality of life, reduce survival, and contribute to metabolic syndrome and secondary malignancies.^23,26 One study demonstrated that chemotherapy for testicular cancer altered body composition by reducing lean mass and impairing taste and smell perception.¹⁴ Takai et al²³ retrospectively analysed body composition changes in 44 patients with testicular germ cell tumours undergoing systemic chemotherapy. While fat mass progressively increased during treatment, skeletal muscle mass initially declined but returned to baseline levels by the 12th month.²³ Their study also identified an association between the number of chemotherapy cycles and skeletal muscle loss. Another study showed that three cycles of bleomycin-etoposide-cisplatin therapy reduced muscle fibre size and strength, indicating a direct detrimental effect of chemotherapeutic agents on skeletal muscle mass.²⁷ Furthermore, significantly lower levels of free and total testosterone have been reported in patients with testicular germ cell tumour,²⁸ and the post-chemotherapy disruption in body composition may be linked to these decreased testosterone levels.⁴

Tumour progression, tumour-induced systemic inflammation, and metabolic dysregulation constitute tumour-associated factors, while deterioration of overall condition and functional status reflect host-related factors. Together, these elements position sarcopenia as a distinctive biomarker indicative of tumour aggressiveness.¹¹ Nutritional support and exercise have been shown to mitigate sarcopenia, with both interventions demonstrating potential oncological benefits. Early initiation of an exercise programme during neoadjuvant therapy has been reported to reduce skeletal muscle loss, as assessed by CT imaging.²⁹

In our study, the prognostic value of pathological characteristics was also assessed. Various pathological features have been reported as prognostic indicators for testicular tumours in the literature. For seminoma, prognostic factors typically include the size of the primary tumour and rete testis invasion, whereas lymphovascular invasion is considered the most important prognostic factor for non-seminomatous germ cell tumours.³⁰ Embryonal carcinoma histology is also recognised as a poor prognostic factor.^30,31 Although our study did not analyse embryonal carcinoma as a distinct subgroup, non-seminomatous and mixed histological types, which encompass embryonal carcinoma, were identified as risk factors for progression in multivariable analysis. Similarly, rete testis invasion emerged as an independent predictor of progression. While lymphovascular invasion was significant in univariate analysis, it did not retain significance in the multivariable model, possibly due to the limited sample size affecting statistical power.

The number of studies investigating the impact of skeletal muscle loss on progression and survival in patients with testicular germ cell tumour remains limited.¹⁴ Our study represents the first multicentre investigation in this field and is notable for its relatively larger patient cohort. Nevertheless, the sample size still constitutes a potential limitation. Additional limitations include the retrospective study design and heterogeneity in pathological subtypes. Patients received varying numbers of chemotherapy cycles and regimens, but due to insufficient patient numbers, a statistically significant subgroup analysis cannot be performed. Additionally, aging, comorbid chronic diseases, prolonged inactivity, and malnutrition can impair skeletal muscle health.⁵ These data are not consistently available for all patients, and their effects on sarcopenia could not be examined. Mortality occurred in only 6 (3.8%) patients, which reduces the statistical power of multivariable analysis, reduces the reliability of hazard ratios, and may lead to overfitting.

In conclusion, for patients with testicular germ cell tumour undergoing systemic chemotherapy, a reduction in skeletal muscle mass serves as a prognostic factor for both disease progression and mortality. PMI change, which can be measured using CT imaging, represents a simple and reliable method for assessing sarcopenia.