Clinical impact of postoperative loss in psoas major muscle and nutrition index after radical cystectomy for patients with urothelial carcinoma of the bladder

Between January 2006 and October 2014, 105 UCB cases, without evidence of distant metastases, underwent curative RC. Sixteen cases were excluded because of insufficient radiographic and laboratory examination, leaving the data from 89 cases to be included in our retrospective analysis. The histopathological review was conducted by an experienced uropathologist (KS) to determine the T category (2010 AJCC TNM Staging system), tumor grade (2004 WHO classification) and presence of variant histology. Clinicopathological variables and laboratory data obtained at baseline and at all time points of measurement postoperatively were extracted from medical records for analysis of the time-course of change after surgery. Among the 89 cases, 37 (42%) and 14 (16%) received 2 or 3 cycles of platinum-based neoadjuvant chemotherapy and adjuvant chemotherapy, respectively.

Unenhanced CT images, performed for diagnostic or follow-up purposes, were used as inputs into the Volume Analyzer SYNAPSE VINCENT image analysis system (Fujifilm Medical, Tokyo, Japan) to quantify the body composition index. Measured variables were normalized to height (m²) and were expressed in cm³/m² or cm²/m² for between-subject comparisons. The following measures were obtained for analysis: total psoas major muscle volume (cm³/m²); partial volume of the psoas major muscle at the level of the third and fourth lumbar vertebra (L3-L4; cm³/m²); area of the psoas major muscle at L3 (cm²/m²); total area of the abdominal skeletal muscle (cm²/m²); and the volume of abdominal adipose tissue, commonly referred to as the volume of subcutaneous and visceral fat (cm³/m²). The visceral-to-subcutaneous adipose tissue volume ratio (VSR) was calculated to determine the distribution of abdominal adipose tissue. Representative images used for analyses of abdominal muscle and fat are shown in Fig. 1a-c.

Abdominal skeletal muscles include the psoas major, paraspinals (erector spinae and quadratus lumborum), and muscles of the abdominal wall (transversus abdominus, external and internal obliques, and rectus abdominus). The cross-sectional area of skeletal muscle at the level of L3 (cm²), which was identified using Hounsfield unit thresholds of −30 to +150, was normalized to height (m²) to yield the skeletal muscle index (SMI), expressed in cm²/m² (Fig. 1d) [16]. Image analysis was performed by two independent investigators (MM and YM) who were blinded to other parameters and outcomes at the time of measurement. Martin’s definition of the SMI [21] was adopted in our analysis as it is based on a large cohort and, therefore, would be applicable to various cancer patients with different BMI profiles, as follows: SMIs of <43 cm²/m² for males with a BMI <25 kg/m²; <53 cm²/m² for males with a BMI ≥25 kg/m²; and <41 cm²/m² for females.

The baseline blood data were obtained the day before RC. The PNI was calculated using the following formula: 10 × serum albumin (g/dl) + 0.005 × total lymphocyte count (per mm³) [18]. The CONUT score was determined on the basis of the serum albumin, peripheral lymphocyte count and the T-cholesterol, as previously described (Table 1) [22]. Both indices were reported as continuous variable for analysis. Cutoffs for the PNI and CONUT scores were set to the lower quartile and higher quartile values among the study group, respectively, based on the interquartile range (IQR).

Postoperative follow-up was performed at the following time points, according to our institutional protocol: 1 month, every 3 months for the first 2 years, every 6 months for the following 3 years, and annually thereafter for patients without evidence of recurrent disease [23]. Oncological evaluation included physical examination, routine blood tests, urine cytology, and CT scan of the chest, abdomen and pelvis. Body composition variables and nutritional index were retrospectively determined and their postoperative time-course change, from baseline, was calculated by absolute value and relative value analysis, with baseline data set to 100%.

Data were expressed by bar charts or box plots, and evaluated using Student’s t-test or the Mann–Whitney U-test as appropriate for the dataset. Disease-specific survival (DSS) and overall survival (OS) were estimated using the Kaplan–Meier method and compared using the log-rank test or log-rank test for trend analysis. Multivariate analysis was used to identify independent prognostic variables using a stepwise Cox proportional hazards regression model. Variables significantly affecting mortality (P < 0.05) in univariate analysis were included in the multivariate analysis. IBM SPSS Version 21 (SPSS Inc., Chicago, IL, USA) and PRISM software version 5.00 (San Diego, CA, USA) were used for statistical analyzes and plotting the data, respectively. A P value <0.05 was considered statistically significant.

The baseline clinicopathological variables, nutritional index, blood test results, and the administration of neoajuvant/adjuvant chemotherapy for the 89 cases included in our analysis are summarized in Table 2. The median follow-up period after RC for the survival analysis was 29 months (IQR, 10–60). Over the follow-up period, 35 patients (39%) died, 25 (28%) of UCB. The mean ± SD (cm²/m²) SMI before RC in men and women, respectively, was 52.1 ± 11.1 and 46.9 ± 7.6. Twenty two patients (25%) were diagnosed with sarcopenia in our study cohort. Baseline variables were compared between patients with and without sarcopenia (Table 2). Sarcopenia status was associated to a lower BMI and clinical T4 category, while there was a marginal association of sarcopenia status with a lower PNI, a lower hemoglobin, and a lower sodium.

The median (IQR) and mean ± SD of the abdominal muscle index, the abdominal adipose tissue index, and the nutritional indices at baseline are listed in Additional file 1: Table S1. Although there was no sex-specific difference in BMI, the visceral fat index, PNI and CONUT score, significant sex-specific differences were identified for all abdominal muscle indices with higher abdominal muscle indices in males, with the subcutaneous fat index and VSR being higher in females. Based on this result, relative values (expressed as a percentage of the baseline, which was set at 100%) were more appropriate to plot the data of male and female patients all together in the analysis of the postoperative change, from baseline, in body composition index than absolute values.

The change from baseline to 24 months after RC was plotted on line graphs (Fig. 2a-h), with absolute values shown in Additional file 2: Figure S1. A transient deterioration was identified for most variables at 1 and 3 months after RC, with the majority of scores recovering to baseline values over time. Although the SMI (all abdominal skeletal muscle volume at L3) exhibited little deterioration (Fig. 2d), the decrease in the psoas major muscle area at L3 remained significant at 3 months (Fig. 2c). A similar trend was seen in changes of two abdominal fat index and VSR (Fig. 2e-g).

Of 89 patients, 71 (80%) underwent ileal conduit, 1 (1%) underwent orthotopic ileal neobladder, and 17 (19%) underwent uretero-cutaneostomy. The former two requires intestinal anastomosis, while the latter does not. To test whether urinary diversion method affects time-course changes in skeletal muscle volume and nutrition status, we compared their changes between ileal conduit/neobladder and cutaneostomy (Fig. 3). Similar analyses were performed for the administration of adjuvant and neoadjuvant chemotherapy (Additional file 3: Figure S2 and Additional file 4: S3). However, none of urinary diversion method, adjuvant chemotherapy, and neoadjuvant chemotherapy affected significantly the changes in skeletal muscle volume and nutrition status.

Both nutritional indices showed a significant transient deterioration at 1 and 3 months after RC (Fig. 2h and i). We hypothesized that the extent of the partial volume loss of the psoas major muscle at L3 and the deterioration in nutritional indices after RC would be associated with a worse prognosis, as previously reported in a cohort of patients with colorectal cancer and lung cancer treated with chemotherapy [16, 19]. The postoperative time course of change in the area of the psoas major muscle area at L3 is shown in Fig. 4a for two representative patients. Median (IQR) points of maximal change during 1–3 months from baseline for the partial volume of psoas major muscle at L3 and for the PNI and CONUT score were −10.0% (−44.6 to −0.1), −5.0 (−20.3 to −0.3) and +2 (−3 to +4), respectively. The cutoff values for change after RC were determined based on median values as follows: −10% (partial psoas volume), −5 points (PNI) and 2 points (CONUT score). Patients experiencing a ≤ −10% loss in the volume of the psoas muscle after RC had a significantly shorter OS, whereas a change of ≤ − 5 on the PNI and ≥2 on the CONUT score was not associated with a worse survival outcome (Fig. 4b). In a subanalysis of patients without sarcopenia (n = 67), DSS and OS were significantly worse in patients experiencing a ≤ −5 change in the PNI, compared to those with > − 5 change in PNI (Fig. 4b). A ≤ −10% loss in the volume of the psoas muscle was associated with a shorter OS in all analyzes, including the entire cohort and subanalyzes for patients with and without sarcopenia (Fig. 4B). No prognostic effect of the CONUT score was identified. The Kaplan–Meier curves for DSS and OS are shown in Additional file 5: Figure S4.

We also evaluated the correlation between the number of days of hospitalization after RC and postoperative loss in psoas major muscle volume, decrease in the PNI and increase in the CONUT score (Fig. 4c-e). Both postoperative muscle loss and deterioration in nutritional indices were significantly related to a longer duration of hospitalization after RC.

To explore the prognostic impact of preoperative sarcopenia status, as well as of the postoperative volume loss of psoas muscle, we performed univariate analyzes followed by multivariate analysis (Table 3). Univariate analysis identified T4 category, N positive and sarcopenia as prognostic variables for DSS, with T4 category, N positive, sarcopenia, and postoperative volume loss of the psoas muscle identified as prognostic factors of OS. Multivariate analysis, controlling for possible confounding factors, identified positive N [hazard ratio (HR) 3.1, P = 0.02] as an independent prognostic factor of DSS, and sarcopenia status at baseline (HR 2.2, P = 0.03) and a 10% loss in the volume of the psoas muscle (HR 2.4, P = 0.02) as an independent prognostic factor of OS. Neither sarcopenia status at baseline nor a 10% loss in the volume of the psoas muscle were identified as independent prognostic factors for DSS in this study (P = 0.14 and 0.11, respectively).