Computed Tomography-Defined Sarcopenia in Outcomes of Patients with Unresectable Hepatocellular Carcinoma Undergoing Radioembolization: Assessment with Total Abdominal, Psoas, and Paraspinal Muscles

Introduction: Sarcopenia is an adverse prognostic factor in patients with liver cirrhosis and hepatocellular carcinoma (HCC). Image-based sarcopenia assessment allows a standardized method to assess abdominal skeletal muscle. However, which is an index muscle for sarcopenia remains unclear. Therefore, we investigated whether sarcopenia defined according to different muscle groups with computed tomography (CT) scans can predict the prognosis of HCC after radioembolization. Methods: In this retrospective study, we analyzed patients who underwent radioembolization for unresectable HCC between January 2010 and December 2019. Before treatment, the total abdominal muscle (TAM), psoas muscle (PM), and paraspinal muscle (PS) areas were evaluated using a single CT slice at the third lumbar vertebra. In previous studies, sarcopenia was determined using the TAM, PM, and PS after stratifying by sex. Finally, we investigated each muscle-defined sarcopenia to decide whether or not it can serve as a prognostic factor for overall survival (OS). Results: We included 92 patients (74 men and 18 women). TAM, PM, and PS areas were significantly higher in the men than in the women (all p < 0.05). The patients with sarcopenia defined using PM, but not TAM and PS, exhibited significantly poorer OS than those without sarcopenia (median 15.3 vs. 23.8 months, p = 0.034, 0.821, and 0.341, respectively). After adjustment for clinical variables, such as body mass index, liver function, alpha-fetoprotein level, clinical staging, treatment response, and posttreatment curative therapy, PM-defined sarcopenia (hazard ratio: 1.899, 95% confidence interval: 1.087–3.315) remained an independent predictor for the poor OS. Conclusion: CT-assessed sarcopenia defined using PM was an independent prognostic factor for the poorer prognosis of unresectable HCC after radioembolization.

© 2023 The Author(s). Published by S. Karger AG, Basel

Introduction

Hepatocellular carcinoma (HCC) is the sixth most commonly diagnosed cancer and the fourth leading cause of cancer deaths worldwide [1]. The Barcelona Clinic Liver Cancer (BCLC) staging system stratifies patients with HCC into very early, early, intermediate, advanced, and terminal stages, with the 5-year survival rates of 40–70%, 14–45%, 6–14%, and 10%, respectively [2]. Because of the limitation of treatment options, transarterial chemoembolization and systemic therapy have served as first-line treatment for intermediate to advanced HCC for more than a decade [3]. However, recent advancements in intra-arterial treatments, such as drug-eluting beads, radioembolization, and hepatic arterial infusion chemotherapy, have provided interventional radiologists with more tools to treat unresectable HCCs [4]. Radioembolization has demonstrated promising results in randomized controlled trials [5] and was recommended by recent practice guidelines for patients with multiple tumors or vascular invasion [6].

Patients with a large tumor burden or vascular invasion commonly exhibit sarcopenia and cachexia [7]. In contrast to the complicated and multifactorial criteria of cachexia, sarcopenia is defined as low muscle mass, decreased muscle strength, or poor performance [8]. Sarcopenia diagnosed based on low muscle mass can be easily accessed through computed tomography (CT), which is a standard diagnostic tool for HCC and routinely used for cancer staging. The total abdominal muscle (TAM) area of a single CT slice at the third lumbar vertebra (L3) strongly predicted fat-free mass [9]. Prado et al. [10] reported that TAM-defined sarcopenia, <52.4 cm2/m2 in men and <38.5 cm2/m2 in women after adjustment for body height (BH), independently predicted outcomes for patients with solid tumors of the respiratory and gastrointestinal tracts and was widely accepted by many studies.

Hamaguchi et al. [11] used BH-adjusted psoas muscle (PM) mass of <6.36 cm2/m2 for men and <3.92 cm2/m2 for women based on liver donors as criteria for healthy controls. A study demonstrated that sarcopenia defined using PM mass with cross-sectional imaging might be an independent prognostic factor for hepatic malignancy after intra-arterial therapy [12]. However, this study enrolled patients with various hepatic malignancies, including HCC; intrahepatic cholangiocarcinoma; and colorectal, neuroendocrine, and other liver metastases. Other studies conducted in Western countries have indicated that magnetic resonance imaging (MRI)-derived, paraspinal muscle (PS)-defined sarcopenia (<31.97 cm2 for men and <28.95 cm2 for women) might be associated with mortality and outcomes in patients with HCC receiving radioembolization [13, 14]. However, the definition of sarcopenia should be sex and ethnic specific. The analyses of TAM- and PS-defined sarcopenia are mainly based on the Western population and the survey of PM-defined sarcopenia consists of the Eastern people. Meanwhile, they used different cutoffs for diagnosing sarcopenia between men and women and various protocols for measuring muscle mass.

In summary, which muscle group can serve as a sentinel indicator for sarcopenia remains controversial [15, 16]. Furthermore, the difference in prognostic prediction among TAM, PM, and PS remains unclear. Whether the Western criteria can be transferred to the Eastern population and sarcopenia has a similar effect on the outcomes between male and female patients are also unknown. Therefore, we explored whether CT-based sarcopenia can be used as a prognostic factor for patients with unresectable HCC after radioembolization and determined an index for abdominal muscle mass to assess sarcopenia.

Materials and MethodsStudy Population

This study was based on a retrospective and observational cohort. We enrolled patients who underwent resin-based yttrium-90 radioembolization for intermediate to advanced HCCs between January 2010 and December 2019 at National Taiwan University Hospital (NTUH), a tertiary referral center in northern Taiwan. We analyzed abdominal CT images within 2 months before radioembolization. Those who had decompensated cirrhosis (Child-Pugh score >7), a high lung shunt fraction (>20%), an estimated tumor burden of >50% of the liver, or had no adequate CT images for analysis were excluded. Because of the retrospective nature of the study, the requirement of informed consent was waived. The study protocol followed the Declaration of Helsinki and was approved by the Institutional Review Board of NTUH (No. 201805070RIND).

Study Variables

In this study, individuals aged ≥70 years were termed elderly. Patients aged <70 years with a body mass index (BMI) of <20 kg/m2 and those aged ≥70 years with a BMI of <22 kg/m2 were categorized as reduced BMI according to the European Society of Clinical Nutrition and Metabolism consensus [17]. Patients who were positive for hepatitis B surface antigen or hepatitis C antibody were considered to have hepatitis B virus (HBV)-related HCC and hepatitis C virus (HCV)-related HCC, respectively. The liver function was evaluated using the albumin-bilirubin (ALBI) and Model for End-Stage Liver Disease (MELD) score as follows: log10 total bilirubin [in µmol/L] × 0.66) + (serum albumin [in g/L] × −0.085 and 9.57 × ln(creatinine) + 3.78 × ln(total bilirubin) + 11.2 × ln(INR) + 6.43 [18]. The ALBI grades were determined according to the ALBI scores (1 [−2.60 or lower], 2 [greater than −2.60 to −1.39, and 3 [greater than −1.39]); only objective parameters were collected for better evaluation of liver function than the Child-Pugh score [19]. The tumor burden was assessed using the maximal tumor size, tumor number, distribution, and volume. The demographic data of the study population are presented in Table 1.

Table 1.

Baseline characteristics of the study population (N = 92)

lic529676_t01.pnglic529676_t01_b.pngVariableValue (mean±SD, median [IQR], or number [%])Age, years65.6±10.9SexMale74 (80.4)Female18 (19.6)Body weight, kg64.5±10.6BH, cm164.4±7.5BMI, kg/m223.8±3.3HBV related60 (65.2)HCV related18 (19.6)Alanine aminotransferase, U/L49±44Albumin, g/dL4.3±0.5Total bilirubin, mg/dL0.95±0.55ALBI grade155 (59.8)>137 (40.2)MELD sore8±2Maximal tumor size, cm7.5±4.2Tumor size, cm≤530 (32.6)>562 (67.4)Tumor number129 (31.5)2–539 (42.4)>524 (26.1)Tumor distributionUnilobar47 (51.1)Bilobar45 (48.9)Total tumor volume, mL252 (405)Non-tumor liver volume, mL1,151±359AFP, ng/mL≤40052 (56.5)>40040 (43.5)BCLC stageB52 (56.5)C40 (43.5)ECOG performance status081 (88.0)111 (12.0)Administered activity (GBq)1.47±0.74Prior therapy64 (69.6)1-month objective response43 (46.7)4-month objective response43 (46.7)7-month objective response28 (30.4)Posttreatment curative therapy19 (20.7)Yttrium-90 Radioembolization and Follow-Up

In our hospital, the weekly multidisciplinary meeting, which included hepatologists, surgeons, medical/radiation oncologists, and intervention/nuclear medicine radiologists, reviewed the diagnosis of HCC and determined the indication for radioembolization, following the recommendations of the European Society for Medical Oncology and Taiwan Liver Cancer Association (TLCA) [6, 20]. According to the TLCA guideline, the diagnostic criteria of HCC include arterial phase hyperenhancement and washout in the portal venous phase under dynamic CT/MRI or tumor biopsy if the imaging pattern was nontypical [6]. Before radioembolization, mapping angiography was performed to explore possible extrahepatic blood supply to the tumor and nontarget organs. If necessary, vascular redistribution was achieved through coil embolization. Then, the intra-arterial administration of technetium 99m-labeled macroaggregated albumin and subsequent planar and single-photon emission CT was performed to evaluate lung shunt, detect extrahepatic perfusion, and calculate the fraction between the tumor and normal liver. After evaluation, SIR-spheres (Sirtex Medical, Sydney, Australia) were applied, and the radiation dose was determined using the body surface area or partition model. Finally, Bremsstrahlung subsequent planar and single-photon emission CT was performed to verify target embolization on the second day after radioembolization. Dynamic CT was routinely arranged 1 month after radioembolization to evaluate the efficacy of the technique. The patients then underwent serum biochemistry testing, alpha-fetoprotein (AFP) level measurement, and CT every 3 months. The treatment response at 1-month, 4-month, and 7-month follow-up CT was evaluated according to the modified Response Evaluation Criteria in Solid Tumors and stratified as complete response (CR), partial response (PR), stable disease (SD), and progressive disease (PD) [21]. Overall survival (OS) was calculated from the date of yttrium-90 administration to death or final follow-up (June 30, 2021). Progression-free survival (PFS) was determined from the treatment date to the date of disease progression, death, or final follow-up. Only the 4-month treatment response was included in survival analysis because the 1-month evaluation may be too early for radioembolization, and we usually shift further therapy at the sixth month.

Imaging Study and Processing

Abdominal CT was performed using a multidetector CT scanner (Brilliance iCT and Ingenuity CT; Philips Healthcare, Amsterdam, The Netherlands). Nonenhanced CT images at the L3 level were analyzed to determine TAM, PM, and PS areas and were processed on a compatible computer using open-source software (ImageJ version 1.51; National Institutes of Health, Bethesda, MD, USA). Muscle tissues were differentiated by setting Hounsfield units from unenhanced images based on specific attenuation values (−29, 150), and contours were obtained using a manual tracing method. A radiologist (C.-H.W.) with 10 years of experience in abdominal imaging processed the images based on the method described previously [22]. TAM, PM, and PS areas were obtained from CT images (Fig. 1), and TAM and PM indices were calculated as follows: TAM and PM area/(BH [m])2[10, 11]. Sarcopenia based on TAM (<52.4 cm2/m2 for men and <38.5 cm2/m2 for women after BH adjustment), PM (<6.36 cm2/m2 for men and <3.92 cm2/m2 for women after BH adjustment), and PS (<31.97 cm2 for men and <28.95 cm2 for women without BH adjustment) was determined by following the protocols proposed by Prado, Hamaguchi, and Guichet, respectively [10, 11, 13].

Fig. 1.

CT image reading demonstration of a 64-year-old man. Unenhanced CT (a) at the L3 level shows lipiodol retention at both hepatic lobes, which was analyzed to determine the (b) TAM, (c) PM, and (d) PS after image processing. TAM, PM, and PS areas were 105.85, 54.43, and 15.91 cm2, respectively. CT, computed tomography; L3, the third lumbar level; TAM, total abdominal muscle; PM, psoas muscle; PS, paraspinal muscle.

/WebMaterial/ShowPic/1502024Statistical Analysis

Data were analyzed using Excel 2013 (Microsoft, Redmond, WA, USA) and R 3.4.3. A p value <0.05 was considered statistically significant in all statistical analyses. The χ2 test was used to compare categorical variables, and the Student’s t test was used to compare continuous variables. If continuous variables do not conform to normal distribution, these values were described as median and interquartile range. The Wilcoxon-Mann-Whitney test was used to compare the difference between groups, and Fisher’s exact test was used for an estimated sample size of less than 5. The paired t test was used to compare muscle areas before and after radioembolization. The Kaplan-Meier method and the log-rank test were used to estimate OS and PFS. Finally, the Cox proportional hazards model was used to analyze survival outcomes in univariate and multivariate analyses.

ResultsPatient Characteristics

A total of 94 patients were enrolled. Two patients were excluded because they had no adequate CT images for analysis. Finally, we analyzed 92 patients whose nonenhanced abdominal CT images were successfully processed (Fig. 2). Of them, 60 (65.2%) had HBV-related HCC, 18 (19.6%) had HCV-related HCC, 2 (2.2%) had HBV/HCV coinfection, and 16 (17.4%) had no evidence of HBV or HCV infection. Moreover, 87 (94.6%) patients had Child-Pugh A and 5 (5.4%) had Child-Pugh B liver cirrhosis, 7 (9.5%) patients had extrahepatic metastasis, and 38 (41.3%) had vascular invasion. The mean MELD score was 8. Furthermore, 33 (35.9%) patients underwent operation, 60 (65.2%) underwent locoregional therapy, and 11 (12.0%) underwent systemic therapy before radioembolization. Among 33 patients who underwent operation, 26 and 7 received open hepatectomy and laparoscopic approach, respectively. Four patients underwent liver transplantation, 3 underwent liver resection, and 14 underwent radiofrequency ablation after radioembolization. There was a wide range of total tumor volumes (median: 252 mL, interquartile range: 405 mL) and the non-tumor liver volume was 1,151 ± 359 mL. The administrated activity was 1.47 ± 0.74 GBq with 45 and 47 patients using body surface area and partition models, respectively (Table 1).

Fig. 2.

Flowchart of patient enrollment.

/WebMaterial/ShowPic/1502023Abdominal Muscle Analysis and Sarcopenia Assessment

Compared with the men, the women had a significantly lower TAM area (p < 0.001), lower PM area (p = 0.020), lower PS area (p < 0.001), lower TAM index (p < 0.001), and lower PS index (p = 0.002). By contrast, no significant difference was observed in the PM index (mean: 6.29 ± 3.49 vs. 4.95 ± 3.19 cm2/m2, p = 0.140) between the men and women. Similarly, TAM- and PS-defined sarcopenia exhibited significant sex differences (p = 0.034 and 0.001) but not PM-defined sarcopenia (p = 0.293; Table 2). Furthermore, PM-defined sarcopenia was associated only with performance status but not with age, sex, BMI, albumin, tumor size, number, distribution, volume, non-tumor liver volume, administrated activity, and response (Table 3). In contrast, TSM- and PS-defined sarcopenia were associated with age and sex. The PS-defined sarcopenia was also associated with weight, BH, BMI, and albumin.

Table 2.

Areas and indices of the TAM, PM and PS and percentage of patients with sarcopenia among men and women

lic529676_t02.pnglic529676_t02_b.pngMeasurementMen (n = 74)Women (n = 18)p valueTAM area, cm2131.43±27.9088.64±18.34<0.001*PM area, cm217.56±9.7511.75±7.190.020*PS area, cm254.88±16.7136.12±13.23<0.001*TAM index, cm2/m247.08±9.1637.02±6.53<0.001*PM index, cm2/m26.29±3.494.95±3.190.140PS index, cm2/m219.62±5.5714.96±4.900.002*TAM-defined sarcopenia (%)59 (79.7)10 (55.6)0.034*PM-defined sarcopenia (%)39 (52.7)7 (38.9)0.293PS-defined sarcopenia (%)3 (4.1)5 (27.8)0.001*Table 3.

Comparison of clinical characteristics between sarcopenic and nonsarcopenic groups

lic529676_t03.pnglic529676_t03_b.pngVariableTAM-defined sarcopenia (n = 69)TSM-defined nonsarcopenia (n = 23)p valuePM-defined sarcopenia (n = 46)PM-defined nonsarcopenia (n = 46)p valuePS-defined sarcopenia (n = 8)PS-defined nonsarcopenia (n = 84)p valueaAge, year67.46±10.7261.70±10.640.028*66.60±10.7565.44±11.220.61473.50±8.0465.31±10.940.042*Sex (male)59 (85.5%)15 (65.2%)0.034*39 (84.8%)35 (76.1%)0.2933 (37.5%)71 (84.5%)0.007*Body weight, kg64.2±10.365.5±11.80.62163.5±9.965.4±11.30.39652.5±5.665.7±10.30.001*BH, cm165.0±6.9162.7±9.00.191164.6±7.7164.3±7.30.829152.4±7.3165.6±6.4<0.001*BMI, kg/m223.6±3.424.6±2.90.20423.5±3.224.2±3.30.29422.6±3.823.9±3.30.293TAM area, cm2116.14±26.56143.82±35.58<0.001*118.04±30.50128.09±31.580.12487.58±34.25126.44±29.000.001*PM area, cm215.19±9.2820.16±9.590.030*9.02±4.9623.84±6.92<0.001*15.87±10.8516.48±9.490.864PS area, cm249.20±15.8357.22±21.650.11254.68±17.8047.73±17.050.05919.44±10.7654.23±15.05<0.001*HBV related46 (66.7%)14 (60.9%)0.61334 (73.9%)26 (56.5%)0.0803 (37.5%)57 (67.9%)0.121HCV related12 (17.4%)6 (26.1%)0.3636 (13.0%)12 (26.1%)0.1154 (50.0%)14 (16.7%)0.044*Alanine aminotransferase, U/L48±4352±450.70047±3351±520.64161±3948±440.417Albumin, g/dL3.9±0.43.8±0.70.6073.8±0.53.9±0.50.5053.4±0.63.9±0.50.006*Total bilirubin, mg/dL0.91±0.411.09±0.810.1630.97±0.470.93±0.620.7760.89±0.500.96±0.550.725ALBI grade >128 (40.6%)9 (39.1%)0.90220 (43.5%)17 (37.0%)0.5246 (75.0%)31 (36.9%)0.057MELD score8±29±30.1958±28±30.6678±28±30.526Maximal tumor size, cm7.6±4.47.3±3.30.7307.8±4.67.3±3.70.5887.8±4.87.5±4.10.836Tumor size > 5 cm47 (68.1%)15 (65.2%)0.79731 (67.4%)31 (67.4%)0.9996 (75.0%)56 (66.7%)>0.999Tumor number >517 (24.6%)7 (30.4%)0.58316 (34.8%)8 (17.4%)0.0581 (12.5%)23 (27.4%)0.675Bilobar32 (46.4%)13 (56.5%)0.39918 (39.1%)27 (58.7%)0.0615 (62.5%)40 (47.6%)0.481Total tumor volume, mLb254 (411)249 (400)0.885300 (659)242 (293)0.417140 (191)261 (437)0.178Non-tumor liver volume, mL1,133±3721,205±3160.4051,191±3571,111±3600.2851,103±2401,156±3690.691AFP > 400, ng/mL30 (43.5%)10 (43.5%)>0.99924 (52.2%)16 (34.8%)0.0925 (62.5%)35 (41.7%)0.288BCLC stage C32 (46.4%)8 (34.8%)0.33123 (50.0%)17 (37.0%)0.2072 (25.0%)38 (45.2%)0.458ECOG performance status 110 (14.5%)1 (4.3%)0.1949 (19.6%)2 (4.3%)0.024*3 (37.5%)8 (9.5%)0.052Administered activity (GBq)1.42±0.671.63±0.900.2291.51±0.711.44±0.770.6391.09±0.471.51±0.080.123Prior therapy50 (72.5%)14 (60.9%)0.29532 (69.6%)32 (69.6%)0.9994 (50.0%)60 (71.4%)0.2401-month objective response32 (46.4%)11 (47.8%)0.90423 (50.0%)20 (43.5%)0.5314 (50.0%)39 (46.4%)>0.9994-month objective response29 (42.0%)14 (60.9%)0.11722 (47.8%)21 (45.7%)0.8344 (50.0%)39 (46.4%)>0.9997-month objective response21 (30.4%)7 (30.4%)>0.99913 (28.3%)15 (32.6%)0.6502 (25.0%)26 (31.0%)>0.999Posttreatment curative therapy15 (21.7%)4 (17.4%)0.65610 (21.7%)9 (19.6%)0.7970 (0.0%)19 (22.6%)0.198

During follow-up, significant increases were observed in PM index 1 month after radioembolization (6.14–7.17 cm2/m2, p = 0.040), but TSM and PS areas and indices were not. No significant difference in all imaging-based muscle measurements was noted at 4-month and 7-month follow-up images (online supplement 2; for all online suppl. material, see www.karger.com/doi/10.1159/000529676).

Treatment Response and Survival Analysis

During follow-up, 43, 21, and 17 patients had PR, SD, and PD on 1-month CT. 6, 37, 15, and 22 patients had PR, SD, and PD on 4-month CT. 7, 21, 6, and 22 patients had PR, SD, and PD on 7-month CT. The objective response rates (CR + PR) were 46.7% (43/92), 46.7% (43/92), and 30.4% (28/92) at the 1-month, 4-month, and 7-month follow-up CT. The disease control rates (CR + PR + SD) were 69.6% (64/92), 63.0% (58/92), and 37.0% (34/92) at the 1-month, 4-month, and 7-month follow-up CT.

The median OS rates of the entire cohort, men, and women were 16.2, 15.2, and 23.8 months, respectively. Reduced BMI, ALBI grade >1, MELD score, AFP level >400 ng/mL, BCLC stage C, 4-month objective response, and curative therapy after radioembolization were significant prognostic factors, but prior therapy before radioembolization was not. After applying optimal sex-specific cutoff points for TAM, PM, and PS, only the patients with PM-defined sarcopenia exhibited significantly poorer OS than did those without (median 15.3 vs. 23.8 months, p = 0.034; Fig. 3c), and PM-defined sarcopenia remained an independent predictor of poor OS (hazard ratio: 1.899, 95% confidence interval: 1.087–3.315, p = 0.024; Table 4), after adjustment for other clinical variables. TAM- and PS-defined sarcopenia were not significant prognostic factors in univariate and multivariate analyses (Fig. 3; Table 4).

Fig. 3.

Comparison of survival rates. OS was significantly different when using the PM criteria (c, p = 0.034) but did not significantly differ when using TAM and PS criteria (a, e, p = 0.821 and 0.341). PFS was not significantly different when using the TAM, PM, and PS criteria (b, d, and f, p = 0.679, 0.234, and 0.390). OS, overall survival; PM, psoas muscle; TAM, total abdominal muscle; PS, paraspinal muscle; PFS: progression-free survival.

/WebMaterial/ShowPic/1502022Table 4.

Univariate and multivariate OS analyses

lic529676_t04.pnglic529676_t04_b.pngUnivariate analysisMultivariate analysisHR95% CIp valueHR95% CIp valueElderly0.9320.536–1.6210.803Male1.9220.910–4.0600.130Reduced BMI2.0641.084–3.9310.028*Reduced BMI1.9820.995–3.9500.052HBV related1.4610.848–2.5160.172HCV related0.8040.417–1.5500.515ALBI grade >13.2621.931–5.509<0.001*ALBI grade >12.7111.565–4.695<0.001*MELD score1.1081.018–1.2070.017*MELD score1.1311.020–1.2540.019*Size >5 cm1.3630.792–2.3460.263Number >51.3490.790–2.3050.273Bilobar0.7740.466–1.2870.323AFP >400 ng/mL2.6031.551–4.371<0.001*AFP >400 ng/mL1.5250.834–2.7890.170BCLC stage C1.9781.184–3.3050.009*BCLC stage C0.8600.639–1.1580.320ECOG performance status 11.9460.954–3.9690.067Prior therapy0.6660.392–1.1320.1334-month objective response0.4310.257–0.724<0.001*4-month objective response0.4640.265–0.8120.007*Posttreatment curative thera

留言 (0)

沒有登入
gif