SEARCH

SEARCH BY CITATION

Keywords:

  • renal cell carcinoma;
  • body composition;
  • muscle density;
  • prognosis;
  • targeted therapy

Abstract

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. FUNDING SUPPORT
  8. CONFLICT OF INTEREST DISCLOSURES
  9. REFERENCES

BACKGROUND

Studies have shown that skeletal muscle and adipose tissue are linked to overall survival (OS) and progression-free survival (PFS). Because targeted therapies have improved the outcome in patients with metastatic renal cell carcinoma (mRCC), new prognostic parameters are required. The objective of the current study was to analyze whether body composition parameters play a prognostic role in patients with mRCC.

METHODS

Adipose tissue, skeletal muscle, and skeletal muscle density (SMD) were assessed with computed tomography imaging by measuring cross-sectional areas of the tissues and mean muscle Hounsfield units (HU). A high level of mean HU indicates a high SMD and high quality of muscle. OS and PFS were estimated using the Kaplan-Meier method and compared with the log-rank test. The multivariable Cox proportional hazards model was adjusted for Heng risk score and treatment.

RESULTS

In the 149 patients studied, the median OS was 21.4 months and was strongly associated with SMD; the median OS in patients with low SMD was approximately one-half that of patients with high SMD (14 months vs 29 months; P = .001). After adjustment for Heng risk score and treatment, high SMD was associated with longer OS (hazards ratio, 1.85; P = .004) and longer PFS (hazards ratio, 1.81; P = .002). Adding SMD will separate the intermediate-risk and favorable-risk groups into 3 groups, with different median OS periods ranging from 8 months (95% confidence interval [95% CI], 6 months-12 months) for an intermediate-risk Heng score/low SMD to 22 months (95% CI, 14 months-27 months) for an intermediate-risk Heng score/high SMD and a favorable-risk Heng score/low SMD to 35 months (95% CI, 24 months-43 months) for a favorable-risk Heng score/high SMD.

CONCLUSIONS

High muscle density appears to be independently associated with improved outcome and could be integrated into the prognostic scores thereby enhancing the management of patients with mRCC. Cancer 2013;119:3377–84. © 2013 American Cancer Society.


INTRODUCTION

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. FUNDING SUPPORT
  8. CONFLICT OF INTEREST DISCLOSURES
  9. REFERENCES

Tumors of the kidney represent approximately 2% to 3% of new cancer cases diagnosed each year and approximately one-third of patients will eventually develop metastatic disease. Better knowledge of prognostic and predictive factors could improve the management of metastatic renal cell cancer (mRCC) by identifying those patients at a higher risk of death or disease progression. The most commonly used prognostic scores are the Memorial Sloan-Kettering Cancer Center score and the recent Heng risk score for patients receiving targeted therapies.[1, 2]

Since the introduction of targeted therapies, survival has doubled. However, if new drugs have proven efficacy with regard to progression-free survival (PFS) in patients with mRCC, the discordance between the response to treatment and overall survival (OS) has been noted.[3] Therefore, predictive as well as prognostic markers need to be studied extensively.

Many prognostic scores incorporate clinical and biological factors, but to the best of our knowledge none integrate measurements of body composition such as quantity of fat or skeletal muscle. Patients with RCC who have a higher body mass index (BMI) have been reported to have a significantly better prognosis than those with a lower BMI.[4, 5] However, the favorable prognostic roles of overweight and obesity have been challenged.[6, 7] In addition, body composition parameters (ie, adipose tissue and skeletal muscle) have been shown to be linked with prognosis in patients with cancer. Previous studies have shown that low skeletal muscle mass (ie, sarcopenia) is linked to a shorter OS in obese patients with colorectal cancer and in obese patients with pancreatic cancer.[8, 9] High levels of muscle performance as have been reported in lung cancer,[10, 11] or high skeletal muscle density (SMD) as has been described in the elderly and in patients with stage III melanoma, are also prognostic of a longer OS.[12, 13] With regard to adipose tissue, the results are conflicting. High adipose tissue and especially visceral adipose tissue (VAT) have been reported as either related to better OS[14, 15] or to shorter survival.[16] The objective of the current study was to analyze whether skeletal muscle (either muscle mass or muscle function based on the SMD) and adipose tissue play a prognostic role in patients with mRCC who are treated with targeted therapy.

MATERIALS AND METHODS

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. FUNDING SUPPORT
  8. CONFLICT OF INTEREST DISCLOSURES
  9. REFERENCES

Patients

We selected all patients treated at the Institut Gustave-Roussy in Paris, France who were enrolled in 3 prospective multicenter trials: the TARGET trial (Treatment Approaches in Renal Cancer Global Evaluation Trial) (n = 83), which assessed the efficacy of a dose of sorafenib of 400 mg given twice daily versus placebo[17]; the RECORD trial (Renal Cell cancer treatment with Oral RAD001 given Daily) (n = 28), which assessed the efficacy of an oral dose of 10 mg/day of everolimus versus placebo[18]; and the sunitinib phase 2 trial (n = 38), which assessed the efficacy of sunitinib at a continuous daily dose of 37.5 mg.[19] Of note, the 83 patients from the TARGET trial were included in previous studies assessing the prognostic role of sarcopenia.[22]

For the original trials, patients provided informed consent, and these studies were approved by the research ethics board of all centers according to good clinical practice, the Declaration of Helsinki, and applicable regulations. Additional analysis of clinical data and the interpretation of body composition from computed tomography (CT) images were approved by the Institutional Review Board of the Institut Gustave-Roussy.

The inclusion criteria for these 3 studies were very similar: histologically proven clear cell mRCC and evidence of measurable disease, after at least 1 line of therapy. Full details of the main studies have been previously described.[17-19]

Anthropometry and Body Composition by Analysis of CT Imaging

Weight and height were measured prospectively at baseline as part of the protocol. Body mass index (BMI) was calculated (ie, BMI = weight in kg/height in m2).

Body composition features were evaluated using the same CT images obtained at baseline for tumor assessment. Baseline CT scans were performed within 28 days before the initiation of therapy and all the measurements were performed by 1 technician who was blinded to patient information, clinical treatment, and outcome. Directly measured variables in the current study were lumbar cross-sectional areas (cm2) of skeletal muscle, VAT, and subcutaneous adipose tissue (SAT) as described in previous studies.[20] Total adipose tissue (TAT) was calculated as TAT = VAT + SAT. The third lumbar vertebra (L3) was chosen as a landmark because L3 cross-sectional areas and whole-body measurements are linearly related.[21] CT images were analyzed using SliceOMatic software (version 4.3; TomoVision, Magog, Quebec, Canada). To evaluate the density of the skeletal muscle, we measured the mean radiation attenuation of skeletal muscle, which describes the input images read by the SliceOMatic software. The pixel values of these images displayed in shades of gray represent the physical properties of the scanned tissue expressed in a numerical form. The mean attenuation (expressed as the mean Hounsfield unit [HU]) has been extensively studied as a correlate of muscle density.[12] Muscle density assessed by this method reflects fatty muscle infiltration, with a lower mean HU indicating lower density and more fatty infiltration. This highly reproducible method correlates with muscle triglyceride contents on muscle biopsy.[12] As in previous studies, these values were normalized for height scale and are expressed as cm2/m2.[20, 22]

Statistical Analysis

Since the distribution of adipose tissue, skeletal muscle, and SMD differs greatly according to sex, for each parameter the population was dichotomized in 2 groups: patients with a value under the median value in patients of the same sex and patients with a value that was equal to or above the median value in patients of the same sex. Baseline values (at the time of treatment initiation in the trial) of body parameters were expressed as the median, interquartile range, and coefficient of variation corresponding to the standard deviation divided by the mean (noise-to-signal ratio). PFS was defined as the time from treatment to the first documentation of disease progression or death from any cause, whichever occurred first. Disease progression was defined as a ≥ 20% increase in SLD as per Response Evaluation Criteria In Solid Tumors (RECIST, version 1.0). OS was defined as the time from treatment to death or last contact. The PFS and OS were estimated using the Kaplan-Meier method with Rothman 95% confidence intervals (95% CI) and compared across the groups using the log-rank test.

The association between each body parameter with death and disease progression was evaluated using the Cox proportional hazards model. Univariable analysis assessed the association between OS and PFS and the following variables: sex (male vs female), Heng score risk groups (favorable-risk vs intermediate-risk vs poor-risk groups), active therapy (yes if sorafenib, sunitinib, or everolimus; no if placebo), type of therapy (vascular endothelial growth factor [VEGF] inhibitor-containing therapy vs other therapies), and BMI (< 18.5 kg/m2 vs 18.5-24.9 kg/m2 vs ≥ 25 kg/m2). The interaction between sex and all body parameters was systematically tested. Multivariable analyses were adjusted for variables with a P value < .20 in the univariable models using a backward stepwise strategy to eliminate nonsignificant variables at a P value < .05. All analyses were performed using the SAS statistical software (version 9.3; SAS Institute Inc, Cary, NC).

RESULTS

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. FUNDING SUPPORT
  8. CONFLICT OF INTEREST DISCLOSURES
  9. REFERENCES

Patients

After the exclusion of patients who did not meet the criteria for inclusion (n = 40), 142 patients and 119 patients had data available regarding tissue areas and SMD, respectively (Fig. 1). Among the 101 patients receiving active therapies, 83 received VEGF inhibitor therapy.

image

Figure 1. Flow chart of the study is shown. HU, Hounsfield unit.

Download figure to PowerPoint

The Heng risk score classified 61% of the patients as being of favorable risk, 36% as being of intermediate risk, and 3% as being of poor risk. Body composition parameters differed between men and women, reflecting the higher surface area of VAT and skeletal muscle in men (Table 1). Adipose tissue parameters were scattered, especially VAT, which presented high coefficients of variation in men (61%) and women (107%). Only 4 patients were considered to be malnourished (BMI of < 18.5 kg/m2).

Table 1. Baseline Characteristics: Age and Body Parameters
 MaleFemale
CharacteristicNo.Median (Q1-Q3)CVNo.Median (Q1-Q3)CV
  1. Abbreviations: BMI, body mass index; CV, coefficient of variation; HU, Hounsfield unit; VAT, visceral adipose tissue; VAT index, visceral adipose tissue index; SAT index, subcutaneous adipose tissue index; SAT, subcutaneous adipose tissue; TAT, total adipose tissue (TAT = VAT +SAT); Q1-Q3: 1st quartile -3rd quartile.

Age, y11360 (52–66)16%3658 (54–65)17%
Weight, kg11380 (72–88)17%3662 (52–69)23%
BMI, kg/m211326 (24–29)16%3623 (20–26)23%
VAT, cm2109153 (90–219)61%3344 (22–77)107%
VAT index, cm2/m210950 (28–72)62%3316 (9–29)113%
SAT, cm2109157 (114–209)47%33160 (95–183)57%
SAT index, cm2/m210953 (37–66)47%3360 (38–72)56%
TAT, cm2109326 (211–407)49%33201 (117–317)63%
TAT index, cm2/m2109106 (75–133)49%3378 (50–110)64%
VAT-to-TAT, %10948 (38–55)23%3323 (14–30)55%
Skeletal muscle, cm2109161 (147–173)13%3399 (91–110)21%
Skeletal muscle index, cm2/m210953 (48–58)14%3337 (35–42)25%
Mean muscle HU9038 (31–43)22%2936 (30–44)30%

Overall Survival

In the 149 patients included in the current study, the median OS was 21.4 months (95% CI, 18.4 months-23.9 months) and was strongly associated with SMD; the median OS in patients with high SMD (29 months [95% CI, 21 months-43 months]) was 2-fold longer than that of patients with low SMD (14 months [95% CI, 10 months-22 months]) (P = .001) (Figs. 2a and 2b). No association between OS and TAT or muscle area was found. OS was not found to be associated with BMI category (P = .99).

image

Figure 2. Kaplan-Meier estimates of (a and b) overall survival and (c and d) progression-free survival in the strata of skeletal muscle density (SMD; a and c) are shown. Black dashed line indicates greater than or equal to the median value in patients of the same sex; gray dashed line, less than the median value in patients of the same sex and (b and d) in the strata of SMD and Heng risk score. (1a) Favorable-risk Heng score and SMD greater than or equal to the median values are indicated by the black dashed line. (1b) Favorable-risk Heng score and SMD less than the median values are indicated by the gray solid line. (2a) Intermediaterisk Heng score and SMD greater than the median values are indicated by the black dashed line. (2b) Intermediate-risk Heng score and SMD less than the median values are indicated by the gray dotted line.

Download figure to PowerPoint

No interaction was found between the respective effects of sex and each body parameter on survival.

When adjusted for Heng risk score, SMD below the median value for patients of the same sex was associated with higher mortality (hazards ratio [HR], 1.9; 95% CI, 1.3-2.9) (Table 2), whereas OS was not found to be associated with either a low level of muscle area (HR, 1.2; 95% CI, 0.8-1.7) (P = .40) or a low TAT level (HR, 1.2; 95% CI, 0.8-1.7) (P = .38). Adding SMD to the Heng risk score was found to define 3 groups instead of 2 for the intermediate-risk and favorable-risk groups, with a different median OS ranging from 8 months (95% CI, 6 months-12 months) for an intermediate-risk Heng score/low SMD to 22 months (95% CI, 14 months-27 months) for an intermediate-risk Heng score/high SMD and favorable-riskHeng score/low SMD to 35 months (95% CI, 24 months-43 months) for a favorable-risk Heng score/high SMD.

Table 2. Crude Estimates and Adjusted Estimated Ratio of Death and of Death or Disease Progression
 OS HR of DeathPFS HR of Death or Disease Progression
 Crude EstimatesMultivariable EstimatesCrude EstimatesMultivariable Estimates
 HR (95% CI)PHR (95% CI)PHR (95% CI)PHR (95% CI)P
  1. Abbreviations: 95% CI, 95% confidence interval; HR, hazards ratio; OS, overall survival; PFS, progression-free survival; SMD, skeletal muscle density.

  2. a

    Median indicates the median value in patients of the same sex.

Heng scorePoor1<.0011<.0011<.0011<.001
Intermediate2.6 (1.8–3.8) 2.4 (1.6–3.9) 2.3 (1.6–3.3) 2.1 (1.4–3.3) 
Favorable7.5 (2.9–19.4) 24.9 (5.0–123.0) 4.7 (1.8–11.9) 7.5 (1.7–32.7) 
TreatmentPlacebo1.76  1<.0011<.001
Active0.9 (0.6–1.4)   0.4 (0.3–0.6) 0.5 (0.3–0.7) 
SMD≥Mediana1.0021.0021<.0011<.001
<Mediana2.0 (1.3–3.0) 1.9 (1.3–2.9) 1.9 (1.3–2.7) 2.0 (1.3–2.9) 

Progression-Free Survival

In the 149 patients included in the current study, the median PFS was 5.5 months (95% CI, 4.2 months-6.5 months) and was strongly associated with SMD; the median OS in patients with high SMD (8 months (95% CI, 6 months-11 months) was 2-fold longer than that of patients with low SMD (4 months; 95% CI, 4 months-6 months) (P < .001) (Figs. 2c and 2d). No association was found between PFS and TAT and muscle area. PFS was not associated with BMI category (P = .73)

Multivariable analyses were not stratified based on sex because there was no interaction noted between the respective effects of sex and each body parameter on PFS.

Active treatment was associated with better PFS (Table 2) but in treated patients there was no significant effect of the type of therapy noted (VEGF receptor [VEGFR] inhibitor vs mammalian target of rapamycin inhibitor: HR, 1.0; 95% CI, 0.6-1.7 [P = .97]).

When adjusted for Heng risk score and active treatment, PFS was found to be associated with SMD (P = .002); SMD below the median value for patients of the same sex was associated with a higher rate of death or disease progression (HR of death or disease progression, 2.0; 95% CI, 1.3-2.9) (Table 2), but the PFS was not associated with either a low level of muscle area (HR, 1.0; 95% CI, 0.7-1.4 [P = .86]) or with a low TAT level (HR, 0.9; 95% CI, 0.6-1.2 [P = .37]).

DISCUSSION

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. FUNDING SUPPORT
  8. CONFLICT OF INTEREST DISCLOSURES
  9. REFERENCES

The results of the current study of patients with mRCC who were treated with targeted therapies (VEGF inhibitors and mammalian target of rapamycin inhibitors), have demonstrated that muscle density as measured by CT was associated with an OS and PFS that appear to be reduced by approximately one-half (P = .001 and P < .001, respectively). By adding SMD obtained by the simple analysis of CT images, we were able to determine a new subcategory between favorable-risk and intermediate-risk groups based on Heng risk score. Instead of the 2 intermediate-risk and favorable-risk groups, we were able to establish 3 groups classified by an 8-month median OS (95% CI, 6 months-12 months) for patients with an intermediate-risk Heng score and low SMD to a 22-month median OS (95% CI, 14 months-27 months) for patients with an intermediate-risk Heng score and high SMD and a favorable-risk Heng score and low SMD, to a 35-month median OS (95% CI, 24 months-43 months) for patients with a favorable-risk Heng score and high SMD. We did not find any effect of BMI or VAT, SAT, or TAT on OS and PFS and, contrary to what was expected, we did not find any association between skeletal muscle area and OS or PFS.

Epidemiological studies have described a strong association between overweight or obesity and the incidence of mRCC. A high percentage of overweight or obese patients was found in the current study (68%) and in 2 other therapeutic trials (56% and 58%, respectively).[14, 16] The relations between RCC and overweight or obesity appear to be seem more complex than expected. The influence of BMI on OS and PFS has been found in patients before nephrectomy and removal of the tumor might affect outcome.[5, 15] With the exception of the study by Tang et al,[23] studies focusing on mRCC, including the current one, failed to demonstrate a prognostic role for BMI.[6] Overweight or obese patients appear to have a more favorable prognosis than those with a normal BMI among those with organ-confined RCC only, when surgery is planned and may be in a specific ethnic population.[5]

Adipose tissue, and especially VAT, is not simply a way to store energy, but should be considered as a true endocrine compartment with the secretion of various hormones that interact with tumor progression.[24] Two studies have found opposite results. In a study of 64 patients with mRCC who were treated with anti-VEGF, Ladoire et al demonstrated that high VAT was associated with a shorter OS and PFS (HR, 6.3) (P < .001).[16] Conversely, the study by Steffens et al reported that high VATs and SATs were both found to be independently associated with longer PFS and OS.[14] Two criticisms can be made in both studies: the cutoff points were not defined according to sex and both studies analyzed CT images measured at the umbilicus, which has never been correlated to the whole adipose and muscle tissue. Using measurements at L3 and adjusting the results to patient sex, we did not find any influence of VAT and SAT on either OS or PFS.

Muscle mass, as well as muscle functional status, have been shown to be linked with clinical outcomes. In patients with cancer, low skeletal muscle mass was shown to be linked to anticancer treatment toxicity[22, 25] and to OS in sarcopenic obese patients.[8, 9] Muscle function as assessed by a low level of peak oxygen consumption and a 6-minute walk test of < 400 meters (clinically significant cutoff) was found to be strongly associated with survival in patients with non-small cell lung cancer.[10, 11]

We did not find any correlation between skeletal muscle mass and OS. The discrepancy we observed in the current study between results for muscle mass and SMD could be similar to what has been previously reported in elderly patients for SMD measures and risk of hospitalization,[26] and in patients with stage III melanoma.[13] In this latter study, which was performed among 101 patients, SMD was found to be significantly associated with both disease-free survival (P = .04) and distant disease-free survival (P < .001). It is interesting to note that, similar to the results of the current study, no significant association was found between outcomes and muscle area. Muscle density has been shown to be closely related to muscle lipid content and muscle function,[12] and it is linked to inflammatory processes, whereas skeletal muscle mass may be associated with the imbalance between proteolysis and muscle anabolism. These 2 concepts are very similar but do not reflect the same mechanisms.

It is interesting to note that the current study enrolled only patients from clinical trials, and therefore the prognostic role of SMD should be validated in a real-world experience. However, 2 recent studies have demonstrated a similar relationship between SMD and patient outcome. Sabel et al[13] reported that the SMD was an important predictor of outcome in a cohort of patients with stage III melanoma. Martin et al also recently published interesting results.[27] They demonstrated that patients with cancers of the gastrointestinal and respiratory tracts, with involuntary weight loss, muscle depletion, and low SMD, shared a poor prognosis.[27].

Conclusions

There is increasing evidence of a close association between body composition and outcome in patients with mRCC. The main finding of the current study is that patients with a high quality of muscle (high muscle density) have an OS and PFS that are twice that of patients with a low quality of muscle (low muscle density). By integrating the SMD status in the Heng risk score, we were able to add a new prognostic group among the patients with intermediate-risk and favorable-risk Heng scores. There is a gradient in OS and PFS from the group with an intermediate-risk Heng score and low SMD, to the group with an intermediate-risk Heng score and high SMD or a favorable-risk Heng score and low SMD, to the group with a favorable-risk Heng score and high SMD, presenting the longest OS and PFS.

Prognostic factors for improved OS are lacking for patients with mRCC because the prognosis has been improved with targeted therapies. The results of the current study indicate that muscle density could be used as a prognostic factor and could improve the usual prognostic scores. However, further studies are needed to better define the threshold values for muscle density.

CONFLICT OF INTEREST DISCLOSURES

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. FUNDING SUPPORT
  8. CONFLICT OF INTEREST DISCLOSURES
  9. REFERENCES

Dr. Escudier has acted as a consultant and in an advisory role at Bayer Pharma, Pfizer, Novartis, GlaxoSmithKline, and Aveo and has received honoraria from Bayer, Roche, Pfizer, Genentech, Novartis, Aveo, and GlakoSmithKline. Dr. Lanoy has received travel support from GlaxoSmithKline for the 48th Annual Meeting of the American Society of Clinical Oncology, held June 1 to 5, 2012 in Chicago, Illinois. Dr. Albiges-Sauvin acts as a consultant and in an advisory role at and has received research funding from Novartis.

REFERENCES

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. FUNDING SUPPORT
  8. CONFLICT OF INTEREST DISCLOSURES
  9. REFERENCES