Prognostic value of cachexia index in HR +/HER2 − metastatic breast cancer treated with CDK4/6 inhibitors

Abstract

Purpose

The prognostic significance of body composition parameters in patients with HR + /HER2 − metastatic breast cancer treated with CDK4/6 inhibitors remains poorly defined. The cachexia index (CXI), a composite marker integrating skeletal muscle mass, nutritional status, and systemic inflammation, may provide additional prognostic information in this setting. This study aimed to evaluate the prognostic value of body composition parameters, with a particular focus on CXI, in this patient population.

Methods

This retrospective single-center study included 84 patients with HR + /HER2 − metastatic breast cancer receiving CDK4/6 inhibitor therapy. Body composition parameters were assessed using baseline imaging at the L3 vertebral level. Sarcopenia was defined as skeletal muscle index (SMI) < 38.5 cm²/m². CXI was calculated as [skeletal muscle index × serum albumin]/neutrophil-to-lymphocyte ratio. Prognostic factors were evaluated using univariate and multivariate Cox regression analyses.

Results

At a median follow-up of 42.0 months, median PFS and OS were 40.8 and 49.6 months, respectively. The median SMI was 32.1 cm²/m² (IQR 25.2–39.9). Sarcopenia was present in 50 patients (64.3%). Patients with high CXI (≥ 44.9) demonstrated significantly longer PFS (not reached vs. 26.3 months, p = 0.001) and OS (not reached vs. 33.3 months, p = 0.004) compared to those with low CXI. In multivariate analysis, high CXI independently predicted reduced risk of disease progression (HR 0.34, 95% CI 0.16–0.71, p = 0.004) and death (HR 0.26, 95% CI 0.11–0.61, p = 0.002). While sarcopenia and low skeletal muscle density were associated with inferior outcomes in univariate analysis, only CXI retained independent prognostic significance after adjustment for confounding factors including liver metastases and luminal subtype.

Conclusion

In this study, we demonstrated that CXI serves as an independent prognostic biomarker for both PFS and OS in patients with HR + /HER2 − metastatic breast cancer treated with CDK4/6 inhibitors. CXI, a cost-effective tool that can be easily obtained from routine imaging and laboratory parameters, has the clinical capacity to facilitate risk classification and guide the multidisciplinary supportive interventions.

Introduction

Breast cancer is the most common malignancy among women globally, with an estimated 2.3 million new cases and 666,000 related deaths reported in 2022 [1]. Approximately 6% of patients present with de novo metastatic disease at diagnosis, while nearly 30% of those initially diagnosed with early-stage disease eventually develop distant metastases [2]. Hormone receptor–positive/human epidermal growth factor receptor 2–negative (HR +/HER2 −) subtype represents the largest biological subgroup of breast cancer, accounting for 60–65% of metastatic cases [3, 4]. The introduction of cyclin-dependent kinase 4 and 6 (CDK4/6) inhibitors has significantly improved survival outcomes for patients with HR +/HER2 − metastatic breast cancer (MBC) [5, 6]. However, response to treatment remains heterogeneous, and several patient and tumor related factors continue to influence survival, quality of life, and treatment-related toxicities. To date, no clinically validated predictive biomarker reliably guides treatment selection for CDK4/6 inhibitors in HR +/HER2- MBC [7]. Although exploratory analyses from the MONALEESA trials have identified potential genomic alterations detected via circulating tumor DNA (ctDNA), as correlates with progression-free survival, these findings require external validation and are not yet clinically applicable in practice [8].

As the search for reliable predictive tools continues, increasing interest has shifted towards host related factors including systemic inflammatory markers and body composition parameters that may influence both therapeutic efficacy and toxicity. Parameters such as obesity, body mass index (BMI), sarcopenia, adipose tissue distribution, particularly subcutaneous and visceral adipose tissue have gained attention as potential prognostic and predictive markers [9]. Sarcopenia affects approximately 41.6% of patients with metastatic breast cancer [10]. Sarcopenia and cancer cachexia are complex conditions characterized by persistent loss of skeletal muscle mass and, leads to progressive functional disability that cannot be completely reversed with conventional nutritional support [10, 11]. Muscle quality is assessed through skeletal muscle density (SMD) on CT imaging and reflects the degree of intramuscular fatty infiltration, known as myosteatosis [12, 13]. Myosteatosis has also been linked to systemic inflammation and metabolic dysregulation, which may influence cancer progression and treatment response [14, 15].

The cachexia index (CXI) is a recently developed composite biomarker that incorporates skeletal muscle index (SMI), serum albumin, and neutrophil to lymphocyte ratio (NLR). By combining parameters reflecting body composition, nutritional status, and systemic inflammation, CXI provides a more comprehensive assessment of cancer-related cachexia than any single marker alone [16]. Previous studies have demonstrated the prognostic value of CXI across several malignancies, including lung, gastric, hepatocellular, colorectal, and renal cancers [16–21]. However, its prognostic relevance in patients treated with CDK4/6 inhibitors remains insufficiently explored. Given the lack of validated predictive biomarkers for CDK4/6 inhibitor response and the growing evidence that host-related factors influence treatment outcomes, we aimed to investigate the prognostic impact of CXI and baseline body composition parameters, including sarcopenia and myosteatosis, on progression-free survival (PFS) and overall survival (OS) in HR +/HER2 − MBC patients treated with CDK4/6 inhibitors.

Methods

Patient selection and study design

This retrospective single-center study included 84 patients with HR +/HER2 − metastatic breast cancer who received first or second-line CDK4/6 inhibitor therapy between 2020 and 2024 at the University of Health Sciences, Gulhane Training and Research Hospital, Ankara, Turkey. Eligible patients had histologically confirmed HR +/HER2 − disease, an Eastern Cooperative Oncology Group (ECOG) performance status of 0–2, available baseline CT imaging performed within four weeks prior to treatment initiation, and complete clinical and follow-up data. To minimize confounding effects on body composition and metabolic parameters, patients with uncontrolled diabetes mellitus, advanced chronic kidney disease, or severe cardiovascular disease that could substantially affect nutritional and inflammatory parameters were excluded. Furthermore, patients with inadequate baseline imaging or missing key clinical information were also excluded from the analysis.

The study was approved by the Human Research Ethics Committees of the Gulhane Training and Research Hospital on 28 June 2024 (2024–322).

Patients were treated with palbociclib or ribociclib in combination with either an aromatase inhibitor or fulvestrant. Endocrine sensitivity and resistance were classified according to ESMO international consensus guidelines [22].

Tumors were categorized into luminal A-like and luminal B-like subtypes according to the St. Gallen criteria [23].

Imaging and body composition assessment

Baseline body composition parameters were evaluated using the CT part of 18F-FDG PET/CT images. Image acquisitions were performed with an integrated PET/CT scanner (Discovery 690-GE Healthcare). Unenhanced low dose CT and PET emission data were acquired from mid-thigh to the vertex of the skull in supine position with the arms raised over the head. Computed tomography data was obtained by automated dose modulation of 120kVp (maximal 100 mA), collimation of 64 × 0.625 mm, measured field of view (FOV) of 50 cm, nose index of 20% and reconstructed to images of 0.625 mm transverse pixel size and 3.75 mm slice thickness. Body composition analysis was performed on axial CT images obtained at the level of the third lumbar vertebra (L3). All images were analyzed by a single experienced physician trained in both nuclear medicine and radiology, who was blinded to clinical outcomes. Skeletal muscle area (SMA, cm²), skeletal muscle density (SMD, Hounsfield units [HU]), and skeletal muscle index (SMI, cm²/m²) were quantified using automated image analysis software. Muscle tissue was identified using predefined attenuation thresholds ranging from − 29 to + 150 HU. The SMI was calculated as SMA divided by the square of the height. Sarcopenia was defined as SMI < 38.5 cm²/m², based on established cut-offs for female patients with cancer [14]. Myosteatosis (low muscle density) was defined using BMI-adjusted cutoffs: SMD < 33 HU for patients with BMI < 25 kg/m², and SMD < 41 HU for patients with BMI ≥ 25 kg/m² [13].

The cachexia index was calculated using the formula: CXI = [SMI × serum albumin (g/dL)]/neutrophil-to-lymphocyte ratio (NLR) [16]. Serum albumin and NLR were obtained from laboratory tests performed before baseline imaging.

Study endpoints and statistical analysis

The primary endpoints of this study were PFS and OS. PFS was defined as the interval from the initiation of CDK4/6 inhibitor therapy to the date of first radiologic or clinical disease progression or death from any cause. OS was defined as the time from treatment initiation until death from any cause. For both endpoints, patients who were alive or progression-free at the time of the last clinical contact were censored at the date of their last follow-up.

All statistical analyses were performed using IBM SPSS Statistics version 25.0 (IBM Corp., Armonk, NY, USA). Continuous variables were summarized as medians with interquartile ranges (IQR) and categorical variables as frequencies and percentages. Survival outcomes were estimated using the Kaplan–Meier method and compared via the log-rank test. To identify prognostic factors, univariate Cox proportional hazards regression analyses were conducted for clinical, pathological, and imaging-derived parameters (SMI, SMD, and CXI).

Receiver operating characteristic (ROC) analysis was performed to evaluate the discriminative ability of CXI for predicting disease progression and death, yielding AUC values of 0.695 (95% CI 0.578–0.811, p = 0.002) for PFS and 0.686 (95% CI 0.566–0.807, p = 0.004) for OS. Although these values indicate discrimination above chance, neither reached the commonly accepted threshold of ≥ 0.70 for defining a clinically robust cut-off. In addition, examination of the ROC coordinate tables did not identify a stable threshold based on the Youden index, as sensitivity and specificity values changed gradually across the CXI range without a distinct peak. Therefore, to avoid overfitting and ensure a balanced group distribution, patients were dichotomized according to the median CXI value (44.9; IQR 27.1–81.7), a widely accepted approach in exploratory prognostic studies, particularly in the absence of a validated cut-off.

Variables that demonstrated statistical significance (p < 0.05) in the univariate analysis were subsequently included in multivariate Cox regression models. Hazard ratios (HRs) and 95% confidence intervals (CIs) were calculated, and a two-sided p-value < 0.05 was considered statistically significant. The administrative cutoff date for data collection was September 30, 2025.

Results

Patient characteristics

Eighty-four patients with HR +/HER2 − metastatic breast cancer treated with CDK4/6 inhibitors were included in this study. The median age was 56.5 years (IQR 43.0–66.7). When stratified by age group, 10 patients (11.9%) were aged < 40 years, 20 (23.8%) were aged 40–49 years, 31 (36.9%) were aged 50–64 years, 17 (20.2%) were aged 65–74 years, and 6 (7.1%) were aged ≥ 75 years. Overall, 60.7% of patients were postmenopausal. De novo metastatic disease was observed in 54.8% of patients, whereas 45.2% had recurrent disease. The majority of tumors (90.5%) were classified histologically as invasive invasive carcinoma of no special type (NST). Luminal A like and luminal B like subtypes were equally represented (50% each).

Palbociclib and ribociclib were used in 52.4% and 47.6% of patients, respectively; regarding endocrine therapy 66.7% received aromatase inhibitors and 33.3% received fulvestrant. Endocrine sensitivity was observed in 63.1% of patients, whereas 17.9% had primary endocrine resistance and 19.0% had secondary endocrine resistance. CDK4/6 inhibitors were administered in the first-line setting in 63.1% of cases.

Lung metastases were detected in 35.7% of patients, liver metastases in 29.8%, and CNS metastases in 7.1%. Bone-only disease was observed in 21.4%. The median number of metastatic sites was 2 (range 1–4). Detailed characteristics are summarized in Table 1.

Table 1.

Clinicopathological characteristics of the patients

Variable	N (%)
Age, median (IQR)	56.5 (43–66.7)
< 40 years	10 (11.9%)
40–49 years	20 (23.8%)
50–64 years	31 (36.9%)
65–74 years	17 (20.2%)
≥ 75 years	6 (7.1%)
BMI
< 25 kg/m2	31 (36.9%)
≥ 25 kg/m2	53 (63.1%)
Menopausal status
Premenopausal	33 (39.3%)
Postmenopausal	51 (60.7%)
Metastatic status
De novo	46 (54.8%)
Recurrent	38 (45.2%)
Histology
Invasive carcinoma of no special type (NST)	76 (90.5%)
Invasive lobular carcinoma (ILC)	6 (7.1%)
Other	2 (2.4%)
Grade
1–2	64 (76.2%)
3	20 (23.8%)
Luminal subtype
A like	42 (50.0%)
B like	42 (50.0%)
Metastatic burden
1 site	31 (36.9%)
≥ 2 sites	53 (63.1%)
Liver metastasis
Yes	25 (29.8%)
No	59 (70.2%)
Lung metastasis
Yes	30 (35.7%)
No	54 (64.3%)
CNS metastasis
Yes	6 (7.1%)
No	78 (92.9%)
Bone-only disease
Yes	18 (21.4%)
No	66 (78.6%)
Line of CDK4/6i
First-line	53 (63.1%)
≥ 2nd-line	31 (36.9%)
CDK4/6 inhibitor
Palbociclib	44 (52.4%)
Ribociclib	40 (47.6%)
Endocrine partner
AI	56 (66.7%)
Fulvestrant	28 (33.3%)
Endocrine sensitivity
Sensitive	53 (63.1%)
Primary resistance	15 (17.9%)
Secondary resistance	16 (19.0%)

Abbreviations: IQR Interquartile range, BMI Body mass index, NST No special type, ILC Invasive lobular carcinoma, CDK4/6i Cyclin-dependent kinase 4 and 6 inhibitor, CNS Central nervous system, AI Aromatase inhibitor

Body composition parameters

The median SMI was 32.1 cm²/m² (IQR 25.2–39.9). Sarcopenia, defined as SMI < 38.5 cm²/m², was present in 50 patients (64.3%). Myosteatosis, defined according to BMI-adjusted cutoffs (SMD < 33 HU for patients with BMI < 25 kg/m², and SMD < 41 HU for patients with BMI ≥ 25 kg/m²), was identified in 16 (19%) patients. Based on the median CXI value 44.9, IQR 27.1–81.7), patients were classified into low and high CXI groups.

Survival outcomes and prognostic factors

At a median follow-up of 42.0 months, 39 patients (46.4%) experienced disease progression and 34 patients (40.5%) died. The median PFS was 40.8 months (95% CI not reached), and the median OS was 49.6 months (95% CI not reached).

In the univariate Cox regression analysis (Table 2), luminal B-like subtype was significantly associated with shorter PFS compared with luminal A-like (HR 2.04, 95% CI 1.07–3.88, p = 0.031). Liver metastases were strongly associated with inferior PFS (HR 4.18, 95% CI 2.20–7.94, p < 0.001). Among body composition parameters, lower L3 SMD was associated with a higher risk of progression (HR 0.44, 95% CI 0.22–0.88, p = 0.021). Similarly, low L3 SMI was significantly associated with shorter PFS compared with high SMI (HR 0.42, 95% CI 0.19–0.88, p = 0.023). In contrast, a higher CXI (≥ 44.9) was associated with a significantly longer PFS (HR 0.33, 95% CI 0.17–0.65, p = 0.001). Kaplan–Meier curves demonstrated markedly longer PFS among patients with high CXI compared to those with low CXI (median PFS: not reached vs. 26.3 months; log-rank p = 0.001; Fig. 1A). Similarly, patients with high L3 SMI experienced significantly prolonged PFS compared with those with low SMI (median PFS: not reached vs. 29.9 months; log-rank p = 0.019; Fig. 1B).

Table 2.

Univariate and multivariate cox regression analysis for PFS

Variable	Univariate Cox Regression			Multivariate Cox Regression
	HR	95% CI	p-value	HR	95% CI	p-value
Age
< 65	Ref	_	_
≥ 65	0.69	0.32–1.50	0.352
Menopausal status
Premenopausal	Ref	_	_
Postmenopausal	1.11	0.58–2.10	0.743
Metastatic status
De novo	Ref	_	_
Recurrent	1.37	0.73–2.58	0.323
Histological
Grade 1–2	Ref	_	_
Grade 3	1.56	0.79–3.08	0.202
Luminal
Type A	Ref	_	_	Ref	_	_
Type B	2.04	1.07–3.88	0.031	1.52	0.77–3.00	0.218
CDK4/6i type
Palbociclib	Ref	_	_
Ribociclib	1.80	0.96–3.42	0.068
Endocrine therapy
AI	Ref	_	_
Fulvestrant	1.43	0.74–2.76	0.286
Endocrine sensitivity
Sensitive	Ref	_	_
Resistance	1.41	0.74- 2.67	0.262
Liver metastasis
Absent	Ref	_	_	Ref	_	_
Present	4.18	2.20–7.94	< 0.001	3.96	2.00–7.81	< 0.001
Lung metastasis
Absent	Ref	_	_
Present	1.31	0.69–2.47	0.411
Bone-only disease
No	Ref	_	_
Yes	0.46	0.18–1.17	0.104
Line of CDK4/6i
First	Ref	_	_
Later	1.48	0.79–2.78	0.224
BMI
Normal	Ref	_	_
Overweight/obese	0.65	0.35–1.23	0.185
Skeletal Muscle Density
Low	Ref	_	_	Ref	_	_
High	0.44	0.22–0.88	0.021	0.78	0.35–1.71	0.534
Skeletal Muscle Index
Low	Ref	_	_	Ref	_	_
High	0.42	0.19–0.88	0.023	0.72	0.32- 1.63	0.446
Cachexia Index
Low	Ref	_	_	Ref	_	_
High	0.33	0.17–0.65	0.001	0.34	0.16–0.71	0.004

Abbreviations: HR Hazard ratio, CI Confidence interval, CDK4/6i Cyclin-dependent kinase 4 and 6 inhibitor, AI Aromatase inhibitor, BMI Body mass index, Ref. Reference category

Fig. 1.

A Progression-free survival by cachexia index status. Median PFS was 26.3 months (95% CI, 17.1–35.5) in the low CXI group versus not reached in the high CXI group (log-rank p = 0.001). B Kaplan–Meier curves for progression-free survival according to Skeletal Muscle Index (SMI). Patients with high SMI (≥ 38.5) had significantly longer PFS compared with those with low SMI (< 38.5) (log-rank p = 0.019)

The multivariate model for PFS included variables achieving p < 0.05 in univariate analysis: cachexia index, skeletal muscle density, skeletal muscle index, liver metastasis, and luminal subtype (Table 2). High CXI remained independently associated with a reduced risk of disease progression (HR 0.34, 95% CI 0.16–0.71, p = 0.004). Conversely, the presence of liver metastases independently predicted shorter PFS (HR 3.96, 95% CI 2.00–7.81, p < 0.001). However, the prognostic impact of L3 SMD (HR 0.78, 95% CI 0.35–1.71, p = 0.534), L3 SMI (HR 0.72, 95% CI 0.32–1.63, p = 0.446), and luminal subtype (HR 1.52, 95% CI 0.77–3.00, p = 0.218) was not sustained in the multivariate Cox regression model.

In the univariate Cox regression analysis for OS (Table 3), liver metastases were the strongest adverse prognostic factor (HR 4.51, 95% CI 2.26–9.02, p < 0.001) and luminal B-like subtype was also associated with shorter OS (HR 2.35, 95% CI 1.16–4.76, p = 0.018). Patients treated with fulvestrant had significantly worse OS compared with those receiving aromatase inhibitors (HR 2.07, 95% CI 1.04–4.13, p = 0.039). Low SMD showed a trend toward reduced OS (HR 0.49, 95% CI 0.23–1.03, p = 0.059), whereas low L3 SMI was significantly associated with shorter OS (HR 0.37, 95% CI 0.16–0.87, p = 0.023). Importantly, a high CXI was strongly associated with improved survival (HR 0.35, 95% CI 0.17–0.74, p = 0.006). Kaplan–Meier analysis demonstrated significantly longer OS in patients with high CXI compared with those with low CXI (median OS: not reached vs. 33.3 months; log-rank p = 0.004; Fig. 2A). Similarly, patients with high L3 skeletal muscle index (≥ 38.5) experienced significantly improved OS compared with those with low SMI (median OS: not reached vs. 38.5 months; log-rank p = 0.018; Fig. 2B).

Table 3.

Univariate and multivariate cox regression analysis for OS

Variable	Univariate Cox Regression			Multivariate Cox Regression
	HR	95% CI	p-value	HR	95% CI	p-value
Age
< 65	Ref	_	_
≥ 65	1.12	0.54–2.35	0.757
Menopausal status
Premenopausal	Ref	_	_
Postmenopausal	1.45	0.72–2.94	0.302
Metastatic status
De novo	Ref	_	_
Recurrent	1.38	0.70–2.70	0.351
Histological
Grade 1–2	Ref	_	_
Grade 3	1.73	0.86–3.50	0.128
Luminal
Type A	Ref	_	_	Ref	_	_
Type B	2.35	1.16–4.76	0.018	1.93	0.94–3.96	0.074
CDK4/6i type
Palbociclib	Ref	_	_
Ribociclib	1.90	0.96–3.75	0.065
Endocrine therapy
AI	Ref	_	_	Ref	_	_
Fulvestrant	2.07	1.04–4.13	0.039	3.40	1.55–7.50	0.002
Endocrine sensitivity
Sensitive	Ref	_	_
Resistance	1.72	0.87- 3.40	0.116
Liver metastasis
Absent	Ref	_	_	Ref	_	_
Present	4.51	2.26–9.02	< 0.001	4.00	1.94–8.22	< 0.001
Lung metastasis
Absent	Ref	_	_
Present	1.17	0.59–2.34	0.653
Bone-only disease
No	Ref	_	_
Yes	0.43	0.15–1.22	0.113
Line of CDK4/6i
First	Ref	_	_
Later	1.74	0.89–3.41	0.108
BMI
Normal	Ref	_	_
Overweight/obese	0.76	0.39–1.51	0.437
Skeletal Muscle Density
Low	Ref	_	_
High	0.49	0.23–1.03	0.059
Skeletal Muscle Index
Low	Ref	_	_	Ref	_	_
High	0.37	0.16–0.87	0.023	0.60	0.25–1.45	0.256
Cachexia Index
Low	Ref	_	_	Ref	_	_
High	0.35	0.17–0.74	0.006	0.26	0.11–0.61	0.002

Abbreviations: HR Hazard ratio, CI Confidence interval, CDK4/6i Cyclin-dependent kinase 4 and 6 inhibitor, AI Aromatase inhibitor, BMI Body mass index, Ref. Reference category

Fig. 2.

A Kaplan–Meier curves for overall survival according to cachexia index status. Median OS was 33.3 months (95% CI, 26.2–40.4) in the low CXI group versus not reached in the high CXI group (log-rank p = 0.004). B Kaplan–Meier curves for overall survival according to Skeletal Muscle Index (SMI). Patients with high SMI (≥ 38.5) had significantly longer OS compared with those with low SMI (< 38.5) (log-rank p = 0.018)

The multivariate model for OS included the following variables significant on univariate analysis: cachexia index, skeletal muscle index, liver metastasis, endocrine therapy type (fulvestrant vs. aromatase inhibitor), and luminal subtype (Table 3). High CXI was independently associated with a 74% reduction in the risk of death (HR 0.26, 95% CI 0.11–0.61, p = 0.002). Liver metastases continued to predict markedly shorter OS (HR 4.00, 95% CI 1.94–8.22, p < 0.001). Treatment with fulvestrant, compared with aromatase inhibitors was associated with poorer outcomes (HR 3.40, 95% CI 1.55–7.50, p = 0.002). L3 SMI did not retain independent prognostic significance in the multivariate model (HR 0.60, 95% CI 0.25–1.45, p = 0.256), and the association between luminal B–like subtype and OS was not statistically significant after multivariate adjustment (HR 1.93, 95% CI 0.94–3.96, p = 0.074).

Treatment-related toxicity and dose modifications

Treatment-related toxicity and dose modifications according to CXI and SMI status are summarized in Table 4. Overall, grade ≥ 3 adverse events occurred in 27 patients (32.1%). When stratified by CXI status, no significant differences were observed between patients with low and high CXI in terms of grade ≥ 3 adverse events (33.3% vs. 31.0%, p = 0.815), dose reduction (19.0% vs. 14.3%, p = 0.558), or treatment interruption (31.0% vs. 31.0%, p = 1.000).

Table 4.

Treatment-related toxicity and dose modifications by CXI and SMI status

Variable	Low CXI n (%)	High CXI n (%)	p-value (CXI)	Low SMI n (%)	High SMI n (%)	p-value (SMI)
Grade ≥ 3 AE	14 (33.3%)	13 (31.0%)	0.815	23 (42.6%)	4 (13.3%)	0.006
Dose reduction	8 (19.0%)	6 (14.3%)	0.558	11 (20.4%)	3 (10.0%)	0.222
Treatment interruption	13 (31.0%)	13 (31.0%)	1.000	20 (37.0%)	6 (20.0%)	0.106

Abbreviations: CXI Cachexia index, SMI Skeletal muscle index, AE Adverse events

^*p-values calculated using Pearson Chi-Square test or Fisher's Exact Test as appropriate

In contrast, patients with low skeletal muscle index (SMI < 38.5 cm²/m²) experienced significantly higher rates of grade ≥ 3 adverse events compared with those with high SMI (42.6% vs. 13.3%, p = 0.006). However, no statistically significant associations were observed between SMI status and dose reduction (20.4% vs. 10.0%, p = 0.222) or treatment interruption (37.0% vs. 20.0%, p = 0.106).

Discussion

In this study, we demonstrated that CXI, a marker which integrates skeletal muscle mass, nutritional status, and systemic inflammation, is an independent prognostic factor for both PFS and OS in HR +/HER2 − metastatic breast cancer patients treated with CDK4/6 inhibitors. Patients with high CXI had a 66% reduced risk of disease progression (HR 0.34, 95% CI 0.16–0.71, p = 0.004) and a 74% reduced risk of death (HR 0.26, 95% CI 0.11–0.61, p = 0.002) compared to those with low CXI, independent of traditional clinicopathological factors such as luminal subtype and liver metastases. Our findings suggest that CXI, a readily available and cost-effective biomarker derived from routine imaging and laboratory parameters, may provide clinically meaningful risk stratification in this patient population for multidisciplinary approaches such as early nutritional support and resistance exercises.

Jafri et al. first introduced the CXI in patients with advanced non-small cell lung cancer and demonstrated that low CXI was associated with inferior survival outcomes [16]. Subsequently, prognostic value of CXI has been systematically validated across multiple cancer types. A recent meta-analysis including 22 studies and 5401 patients showed that low CXI consistently predicted inferior OS (HR 2.03, p < 0.001) and PFS (HR 1.86, p < 0.001) across gastrointestinal, hepatobiliary, renal and lung cancers, with remarkably low heterogeneity [24].

Despite this growing evidence, data on the prognostic relevance of CXI in breast cancer, especially in patients treated with CDK4/6 inhibitors, is limited. Recently, Baş et al. validated the prognostic significance of modified cachexia index (mCXI) calculated by using the urea-to-creatinine ratio (UCR) instead of SMI, in a larger cohort of 240 metastatic breast cancer patients treated with CDK4/6 inhibitors. In that study, low mCXI was independently associated with inferior PFS (HR 1.50, p = 0.02) and OS (HR 3.22, p < 0.001) [25]. Our study extends these findings by incorporating comprehensive CT-based body composition analysis, including direct assessment of sarcopenia and myosteatosis.

The prevalence of sarcopenia in our study (64.3%) was higher than that reported in the existing literature on metastatic breast cancer. A recent systematic review and meta-analysis reported pooled sarcopenia rates ranging from approximately 40% to 60%, depending on the definitions and imaging cutoffs applied [10]. The higher sarcopenia rate observed in our study may be partly explained by differences in patient characteristics including a substantial proportion of patients receiving second-line or later therapy (36.9%), variations in imaging methodology, and a higher metastatic burden. Our findings are consistent with the results of this meta-analysis regarding the heterogeneous impact of sarcopenia on clinical outcomes. While some studies included in the meta-analysis reported an association between sarcopenia and adverse survival outcomes, the pooled analyses did not demonstrate a statistically significant association between sarcopenia and either PFS or OS. In our cohort, sarcopenia was associated with inferior PFS and OS in univariate analyses, but it did not emerge as an independent prognostic factor after multivariate adjustment.

Yücel et al. evaluated CT-based body composition parameters in 52 patients with HR +/HER2 − metastatic breast cancer treated with CDK 4/6 inhibitors. They reported a sarcopenia prevalence of 50%. In that study, both low SMI and low visceral adipose tissue index (VAT-I) were associated with shorter PFS, with low SMI emerging as the only independent predictor in multivariate analysis. Notably, no association was observed between body composition parameters and grade 3–4 hematological toxicity [26].

In a recent prospective study, Imbimbo et al. investigated the relationship between body composition parameters and treatment-related toxicity in patients receiving CDK4/6 inhibitors [27]. Baseline skeletal muscle area was inversely correlated with the number of grade 3–4 adverse events; however, longitudinal changes in muscle and adipose tissue were not consistently associated with severe toxicity, dose reductions, or treatment discontinuation. In line with these findings, we observed that low SMI was associated with increased severe toxicity in our cohort, whereas CXI status was not associated with grade ≥ 3 adverse events, dose reductions, or treatment interruptions. These observations suggest that while skeletal muscle mass may influence treatment tolerability, CXI primarily reflects prognosis rather than treatment-related toxicity.

Beyond SMI, there is increasing evidence that myosteatosis, as reflected by SMD, may also be relevant to worse outcomes in cancer patients. Kim et al. demonstrated that low SMD was associated with inferior treatment response and survival in patients with advanced HR + breast cancer receiving CDK4/6 inhibitors combined with aromatase inhibitors [15]. In our cohort, low SMD was associated with shorter PFS (HR 0.44, 95% CI 0.22–0.88, p = 0.021) and showed a borderline association with OS (HR 0.49, 95% CI 0.23–1.03, p = 0.059) in univariate analyses; however, this effect did not persist in multivariate models. Taken together, these findings suggest that while isolated measures of muscle quantity or quality capture important aspects of host physiology, composite markers such as CXI may better reflect the complex interplay between muscle mass, nutritional status, and systemic inflammation that ultimately determines clinical outcomes.

An additional observation from our multivariate OS analysis was the association of fulvestrant use with worse survival (HR 3.40, 95% CI 1.55–7.50, p = 0.002). This finding should be interpreted with caution. In current clinical practice, fulvestrant in combination with a CDK4/6 inhibitor is more frequently used in the second-line or later setting, or in patients with primary or secondary endocrine resistance clinical contexts that are themselves associated with poorer outcomes independent of the endocrine backbone. In our cohort, endocrine resistance was present in 36.9% of patients and CDK4/6 inhibitor therapy was initiated beyond first-line in 36.9% of cases. Notably, neither treatment line nor endocrine resistance status reached statistical significance in univariate analysis and were therefore not included in the multivariate model per our pre-specified variable selection criteria. Consequently, the potential confounding effects of these variables could not be statistically adjusted for, and the observed association most likely reflects residual confounding by indication whereby fulvestrant is preferentially prescribed to patients with more advanced, treatment-refractory disease rather than a true pharmacological effect.

Our study has several limitations. First, it was conducted at a single center, which may limit the generalizability of our findings to broader patient populations. Therefore, multicenter validation studies are warranted before the CXI can be widely implemented in clinical practice. Second, the retrospective and cross-sectional nature of the study introduces the potential for unmeasured confounding factors. Additionally, the use of the median CXI value as the dichotomization threshold, rather than a ROC-derived cutoff, may limit clinical applicability, as the optimal threshold could differ across populations. Prospective studies are required to validate or refine this threshold. Furthermore, patient-reported outcomes, particularly quality of life, were not assessed. Finally, longitudinal changes in SMI, SMD, and CXI during treatment were not evaluated. Dynamic evaluation of these parameters over time may provide additional insights into disease progression, treatment response, and clinical outcomes, and should be explored in future prospective studies.

Conclusions

In conclusion, cancer cachexia is associated with adverse clinical outcomes, including increased treatment toxicity, impaired quality of life, and reduced survival [28]. Early identification of patients at high risk is therefore critical to enable timely preventive and supportive interventions.

In this study, we demonstrated that CXI serves as an independent prognostic biomarker for progression-free and overall survival in patients with HR +/HER2 − metastatic breast cancer treated with CDK4/6 inhibitors. Given its derivation from routinely available imaging and laboratory parameters, CXI represents a practical and cost-effective tool for clinical risk stratification. Further multicenter, prospective studies with larger sample sizes and longer follow-up are needed to validate the clinical utility of CXI.

Abbreviations

18F-FDG

Fluorine-18 fluorodeoxyglucose

AI

Aromatase inhibitor

BMI

Body mass index

CDK4/6

Cyclin-dependent kinase 4 and 6

CDK4/6i

CDK4/6 inhibitor

CI

Confidence interval

CNS

Central nervous system

CT

Computed tomography

ctDNA

Circulating tumor DNA

CXI

Cachexia index;

ECOG

Eastern Cooperative Oncology Group

ESMO

European Society for Medical Oncology

FOV

Field of view

HER2

Human epidermal growth factor receptor 2

HR

Hormone receptor

HR + 

Hormone receptor–positive

HU

Hounsfield units

IBM

International Business Machines

IDC

Invasive ductal carcinoma

ILC

Invasive lobular carcinoma

IQR

Interquartile range

kVp

Kilovolt peak

L3

Third lumbar vertebra

mA

Milliampere

MBC

Metastatic breast cancer

NLR

Neutrophil-to-lymphocyte ratio

NOS

Not otherwise specified

OS

Overall survival

PET/CT

Positron emission tomography/computed tomography

PFS

Progression-free survival

ROC

Receiver operating characteristic

SMA

Skeletal muscle area

SMD

Skeletal muscle density

SMI

Skeletal muscle index

SPSS

Statistical Package for the Social Sciences

UCR

Urea-to-creatinine ratio

VAT-I

Visceral adipose tissue index

Authors’ contributions

Conceptualization: GSYK, SI, Data collection and curation: GSYK, GGT, GY, VA, EKT, MBA, and SI, Formal analysis: GSYK, VA Methodology: GSYK, GGT, SI Writing – original draft: GSYK, Writing – review & editing: GSYK, GGT, VA, SI Supervision: SI, NK.

Funding

The authors declare that no funds, grants, or other support were received during the preparation of this manuscript.

Data availability

The data that support the findings of this study are not publicly available due to privacy reasons but are available from the corresponding author upon reasonable request.

Declarations

Ethics approval and consent to participate

This study was conducted in accordance with the principles of the Declaration of Helsinki. Ethical approval was obtained from the Gulhane Scientific Research Ethics Committee of Gulhane Training and Research Hospital, Ankara, Turkiye (Approval No: 2024–322; Date: 28 June 2024). Due to the retrospective nature of the study and the use of anonymized data, the requirement for patient informed consent was waived by the Ethics Committee.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.