Article Text
Abstract
Background Sarcopenia is prevalent among older patients with cancer and is associated with poor outcomes.
Objective To explore the relationship between muscle mass, quality, and patient age with overall survival after surgery for advanced ovarian cancer.
Methods Patients with advanced stage (IIIC/IV) ovarian cancer who underwent primary cytoreductive surgery between January 2006 and July 2016 were included. Body composition measures were calculated from pre-operative CT imaging: skeletal muscle index (skeletal muscle index=skeletal muscle area normalized for height), skeletal muscle density, and skeletal muscle gauge (product of skeletal muscle index and skeletal muscle density). Each measure was transformed to a z-score and evaluated for association with risk of death using Cox proportional hazards models. Recursive partitioning was used to classify patients into homogeneous subgroups considering age and skeletal muscle gauge as predictors of overall survival.
Results The study included 429 patients (mean age 64.2 years). Increased age moderately correlated with decreased skeletal muscle gauge (r=−0.45). Decreasing skeletal muscle density and skeletal muscle gauge were significantly associated with increased risk of death; HR (95% CI) per 1-unit decrease in z-score of 1.24 (1.10 to 1.39) for skeletal muscle density and 1.27 (1.12 to 1.44) for skeletal muscle gauge. Associations were diluted after adjusting for age (1.13 (1.00 to 1.29) skeletal muscle density and 1.14 (0.99 to 1.30) skeletal muscle gauge). Recursive partitioning identified three subgroups: <60 years old, ≥60 years old with skeletal muscle gauge ≥937.3, and ≥60 years old with skeletal muscle gauge <937.3; median overall survival was 5.8, 3.3, and 2.3 years, respectively (p<0.001).
Conclusions Skeletal muscle gauge, a novel sarcopenia measure incorporating quantity and quality, was associated with poorer survival in patients with advanced ovarian cancer, particularly among patients older than 60. Expanding our knowledge of how sarcopenia relates to solid tumor outcomes among high-risk patients can modify our treatment approach.
- Ovarian Cancer
- Gynecologic Surgical Procedures
- Surgical Oncology
- Postoperative Care
- Cytoreduction surgical procedures
Data availability statement
Data are available upon reasonable request.
Statistics from Altmetric.com
- Ovarian Cancer
- Gynecologic Surgical Procedures
- Surgical Oncology
- Postoperative Care
- Cytoreduction surgical procedures
WHAT IS ALREADY KNOWN ON THIS TOPIC
Sarcopenia is prevalent among patients with advanced solid cancers and represents a potentially modifiable risk factor for poor outcomes.
WHAT THIS STUDY ADDS
This study used a novel measure of sarcopenia, skeletal muscle gauge, to evaluate both muscle quantity and quality and found that low skeletal muscle gauge is associated with poor outcomes in patients aged 60 or older with advanced ovarian cancer.
HOW THIS STUDY MIGHT AFFECT RESEARCH, PRACTICE OR POLICY
Continued research evaluating sarcopenia for pre-operative risk stratification in complex surgery and the influence of prehabilitation on this potentially modifiable risk factor is warranted in those at high risk for poor outcomes.
INTRODUCTION
Multiple factors are thought to contribute to the age-related increased morbidity and mortality seen among patients with cancer. These include increased prevalence of medical co-morbidities, decreased functional status, and decrease in muscle mass and function known as sarcopenia.1 2 The projected increase in older oncology patients in the next two decades will require novel, objective methods to assess patients’ fitness for treatment; sarcopenia is one such clinical factor that is easily and objectively measured.
Ovarian cancer is often diagnosed at an advanced stage in patients with an average age of 63 years; older age increases the risk of disease-specific death.3–7 Patients with advanced ovarian cancer require treatment with both chemotherapy and multiquadrant cytoreductive surgery.8–11 Significant morbidity is associated with tumor-directed treatment, especially among older adults.2 6 7 Understanding modifiable risk factors to decrease morbidity and mortality in older adults is an unmet clinical need within surgical oncology. Improved treatment stratification is needed for older women with ovarian cancer and objective measures such as sarcopenia have emerged as an easily accessible assessment of fitness for cancer treatments.
Sarcopenia is an objective measure that has been associated with worse outcomes in multiple cancer types, including ovarian cancer.5–7 12–21 Sarcopenia can be defined by several means; using CT, the most commonly defined measure is skeletal muscle index, or skeletal muscle area normalized for height (skeletal muscle index=skeletal muscle area/height). Skeletal muscle density, the average Hounsfield units (HU) of skeletal muscle on CT, has shown a stronger association with outcomes in women with advanced ovarian cancer and conveys the degree of fatty infiltration of muscle tissue.12 Additionally, Shachar et al evaluated the use of a novel measure combining muscle area and density, called skeletal muscle gauge (skeletal muscle gauge=skeletal muscle index × skeletal muscle density). This combined measure had the strongest association with overall survival and chemotherapy toxicity when evaluated in several cancer types.13 This novel approach may better reflect the complex association of sarcopenia and cancer allowing for improved pre-operative and oncologic risk stratification. Further, identification of sarcopenia may be useful as the target endpoint of interventions such as prehabilitation.22 23
Given the high rate of sarcopenia among patients with cancer and its association with decreased survival, continued investigation of this topic is needed.17 Important goals for the field to address include: (1) defining the best measure of sarcopenia among a diverse population, (2) determining the association between age, sarcopenia, and tolerance to treatment and survival, and (3) determining how to translate these findings into improvements in patient care, especially for older patients. As a first step toward addressing these research questions, our objective was to determine which CT-based measure of sarcopenia is most associated with an increased risk of death in advanced ovarian cancer. We hypothesized that the combined measure of both muscle quantity and quality—namely, skeletal muscle gauge, would be most associated with survival. Ultimately, we hope to use these measures to advance pre-operative risk stratification and improve outcomes for high-risk patients undergoing complex surgery.
METHODS
Patient Eligibility
We conducted a single institution, institutional review board approved, retrospective cohort study evaluating patients who received primary cytoreductive surgery for ovarian cancer between January 1, 2006 and July 31, 2016. Data were obtained from an electronic medical record review conducted as part of a prospectively maintained ovarian cancer surgical database, with additional medical record review for variables specific to this study. Inclusion criteria included stage IIIC or IV epithelial ovarian cancer (ovarian, fallopian tube, or primary peritoneal cancer, International Federation of Gynecology and Obstetrics (FIGO) 2009 and 2014 classifications), presence of a pre-operative CT scan with IV contrast of the abdomen and pelvis within 61 days prior to surgery, and active consent to the use of medical records for research. There was a change from FIGO 2009 to FIGO 2014 during the study time period, but given the focus on stage IIIC and IV patients, this change did not affect our study. A 61-day cut-off point was used to balance the need for both a clinically relevant and statistically sufficient population for analysis. Exclusion criteria included patients who underwent surgery for palliative intent, patients who received neoadjuvant chemotherapy, or if pre-operative CT images could not be analyzed due to poor quality. Clinicopathological characteristics collected and evaluated can be found in Table 1.24 25
Body Composition Measurements and Validation
Pre-operative CT images were analyzed for patients meeting study inclusion criteria. From axial images, the third lumbar vertebra was identified and analyzed for cross-sectional skeletal muscle area and mean skeletal muscle attenuation or density using one of two methods: Slice-o-Matic and a validated Mayo Clinic internal program.17 24
Slice-O-Matic software v4.3 (TomoVision) was used to manually evaluate the cross-sectional skeletal muscle area from identified axial images according to previously defined attenuation thresholds (HU −30 to +150).17 CT analysis transitioned to a Mayo internal program developed by Takahashi et al, BodyCompSlicer, during the completion of this study.24 This software, written in MatLab (Matlab 2015b, MathWorks, Natick, Massachusetts, USA), automatically segments the chosen L3 image into subcutaneous fat, skeletal muscle, and visceral fat compartments using three boundary lines (boundary 1: between external air and subcutaneous fat, boundary 2: between subcutaneous fat and abdominal wall/paraspinal muscles, boundary 3: between the abdominal wall/paraspinal muscles and visceral fat).24 The authors then evaluated and manually adjusted these automated boundary lines as appropriate for increased precision. Once complete, the software calculated the skeletal muscle area between boundary 2 and 3 using attenuation values of −30 to +150 HU and excluding the spine.24 The mean attenuation of included skeletal muscles was evaluated as skeletal muscle density by both methods. The software underwent internal validation as reported by Takahashi et al; additionally, we evaluated the agreement between the two quantitative methods for 30 CT scans in our study population. We found excellent agreement between the two methods with intra-class correlation coefficients of 0.973 (95% CI 0.944 to 0.987) for skeletal muscle area and 0.995 (95% CI 0.989 to 0.998) for skeletal muscle density.
The skeletal muscle index, skeletal muscle area normalized for patient height, was calculated as skeletal muscle area (cm2)/height (m2). Skeletal muscle gauge, as previously described by Williams et al, was calculated as the product of skeletal muscle index and skeletal muscle density, and is represented by arbitrary units.8 9 The term sarcopenia is used throughout the manuscript specifically in reference to CT-derived parameters of muscle quantity and quality and does not incorporate additional in-office measures of physical function and strength.
Study Outcomes and Statistical Analysis
The primary outcome of the study was overall survival. Secondary outcomes were progression-free survival, presence of a 30-day post-operative Accordion grade 3+ complication, and 90-day mortality. The Aletti surgical complexity score, which evaluates complexity and number of surgical procedures performed, was used to analyze surgical complexity.25 The Accordion grading system of surgical complications was used to classify and analyze complications of surgery.26 Data were summarized using standard descriptive statistics. Comparisons of clinicopathologic characteristics were evaluated using the two-sample t-test or F-test from a one-way analysis of variance for continuous variables and Χ2 test or Fisher’s exact test for categorical variables. For time-to-event analyses, follow-up was censored at date of last follow-up for any reason and follow-up of patients without a documented progression was censored at the date of last relevant clinical follow-up. Each of the four continuously scaled CT obtained body composition measurements were transformed to z-scores ((patient value – overall mean)/overall SD) to more directly compare results for each measure when each was evaluated in separate Cox proportional hazards models. First, each measure (in the original scale) was evaluated in a separate univariate Cox model using a smoothing spline to assess the functional form of the relationship with risk of death. Given each measure appeared to have a linear relationship with risk of death, each measure (as a z-score) was then evaluated as a linear term in a Cox model for its association with risk of death, both with and without adjustment for age. Lastly, a full multivariable Cox model was fit that included clinicopathologic factors and skeletal muscle gauge. Associations were summarized using the HR and corresponding 95% CI estimated by the models. As an exploratory analysis, a recursive partitioning technique, using the R package rpart, was used to classify patients into homogeneous groups based on evaluating age and skeletal muscle gauge as predictors of overall survival.
Analyses were performed using SAS version 9.4 statistical software (SAS Institute; Cary, North Carolina, USA) and version 1.4.1103–4 of RStudio. All calculated P values were two-sided and P values <0.05 were considered statistically significant.
RESULTS
Among 615 patients with stage IIIC or IV ovarian cancer who underwent primary cytoreductive surgery, 429 had a CT scan within 61 days prior to primary cytoreductive surgery that was of acceptable quality and were included in our study. Baseline patient characteristics of the 429 patients are summarized in Table 1. The mean age at surgery was 64.2 years; the majority had high grade serous histology. Surgical resection outcomes after primary surgery were optimal in 92.5% (55.2% RD0, 37.3% ≤1 cm) and suboptimal (>1 cm) in 7.5% of patients. The mean skeletal muscle index was 41.1 cm2/m2, and 38.0% were considered sarcopenic using an internationally accepted cut-off point of 39 cm2/m2 for skeletal muscle index.17 Clinical characteristics of those included (n=429) versus those not included (n=186) in this study are presented in online supplemental table 1).
Supplemental material
Among the 429 included patients, there were 293 documented deaths, and the median duration of follow-up was 6.1 years (IQR 3.9–8.0) for the remaining 136 patients. The median overall survival was 3.7 years. Decreasing skeletal muscle density and skeletal muscle gauge were both associated with an increased risk of death when evaluated as z-scores in univariate analyses (skeletal muscle density HR 1.24 per 1-unit decrease in skeletal muscle density z-score, skeletal muscle gauge HR 1.27 per 1-unit decrease in skeletal muscle density z-score (online supplemental figure 1). However, these associations were diluted when adjusted for age, with an adjusted HR of 1.13 (95% CI 1.00 to 1.29) for skeletal muscle density and HR 1.14 (95% CI 0.99 to 1.30) for skeletal muscle gauge (Table 2).
Supplemental material
Sarcopenia is known to correlate with increasing age, accordingly we observed that an increase in age was moderately correlated with a decrease in skeletal muscle gauge (Pearson correlation coefficient, r=−0.45; Figure 1A). Considering age and skeletal muscle gauge as predictors of overall survival, recursive partitioning identified three subgroups as depicted in Figure 1A (<60 years old, ≥60 years old with a skeletal muscle gauge ≥937.3, and ≥60 years old with a skeletal muscle gauge <937.3) with median overall survival of 5.8, 3.3, and 2.3 years, respectively, as depicted in Figure 1B. The overall survival was significantly different between each subgroup (p<0.001).
These three subgroups differed in age and proportion with American Society of Anesthesiologists (ASA) score of 3+, but otherwise had similar clinicopathologic characteristics (Table 3). However, the subgroups were significantly different in terms of intra-operative factors, including amount of residual disease and complexity of the surgical procedure (Table 3). The subgroup of women with the poorest overall survival (age ≥60 years old with a skeletal muscle gauge <937.3) were more than twice as likely than the other two subgroups (14.7% vs 6.2% and 6.0%) to have a suboptimal cytoreduction (>1 cm). In addition, the women in this subgroup were also more likely to have undergone a low complexity surgery (33.8% vs 13.8% and 8.6%).
Lastly, a full multivariable model was fit including clinicopathologic factors and skeletal muscle gauge (Table 4). Older age, non-serous histology, and suboptimal cytoreduction were significantly associated with an increased risk of death; however, the age-adjusted HR for skeletal muscle gauge was no longer statistically significant (HR per 1 per 1-unit decrease in skeletal muscle density z-score-, 1.02; 95% CI 0.88 to 1.18). The adjusted HR for skeletal muscle gauge was also no longer statistically significant in an additional multivariable model that included only pre-operative factors (age, BMI, ASA score, pre-operative albumin, and skeletal muscle gauge).
DISCUSSION
Summary of Main Results
In this study, we found that a combination of older age and lower skeletal muscle gauge were related to poor outcomes and an increased risk of death.13–16 Specifically, we found that a subgroup of older patients with lower skeletal muscle gauge had the shortest overall survival, and that skeletal muscle gauge, the value evaluating both muscle quantity and quality, had a stronger association with poorer overall survival than skeletal muscle index, the traditional image-based measure of skeletal muscle. We also found that this subgroup was less likely to undergo complex surgery and were more likely to have residual disease ≥1 cm, probably influencing the overall survival. These data further the understanding of sarcopenia and outcomes in patients with advanced solid tumors and may help us determine which patients are appropriate for complex surgery. It may also be used as a stratifier for prehabilitation efforts and a possible surrogate end point for prehabilitation trials.
Results in the Context of Published Literature
Surgeons are often faced with the difficult decision of determining a patient’s fitness for complex cancer treatment. Many factors are weighed when considering a potentially high-risk surgery in older patients with advanced cancer. Pre-treatment risk assessment aims to identify a group of patients with increased vulnerability to the stressors of cancer and its treatments.1 2 27–30 Both pre-operative risk-assessment algorithms and prehabilitation strategies are being developed to improve outcomes among those at highest risk of poor outcomes.1 2 27 28
Skeletal muscle area normalized for patient height (skeletal muscle index) has been traditionally used to define sarcopenia. However, we previously found that lower skeletal muscle density, and not skeletal muscle area, was associated with worse overall survival in women with advanced ovarian cancer.17 Clearly defining body composition variables relevant to a diverse oncologic patient population is still an area needing research. We evaluated the novel combination of skeletal muscle mass and density, skeletal muscle gauge, in a population with advanced ovarian cancer. Our work revealed an association between age, skeletal muscle gauge, and survival; the median overall survival for patients aged 60 or older with low skeletal muscle gauge was 1.5 years less than for the entire cohort and 1 year less than those 60 or older with higher skeletal muscle gauge. Patient age as well as tumor biology probably contribute to sarcopenia; the intersection of these three characteristics is not currently well understood. It is notable that while skeletal muscle gauge contributes to risk of death, the effect is attenuated when controlling for age. While age remains clinically relevant in operative decision-making algorithms, it is a non-modifiable factor. As such, we continue to aim our prehabilitation efforts on possibly modifiable pre-operative factors, such as sarcopenia. The combination of age and sarcopenia demonstrates the importance of body composition in older patients with advanced cancer.
Patients aged 60 or older with low skeletal muscle gauge were less likely to have a moderate or high complexity procedure and less likely to have cytoreduction to no gross residual disease. Given the high rate of complete resection in the group as a whole, this raises the question as to why these patients appear to have less aggressive surgery. We have considered two possible explanations: (1) practitioner bias regarding patient’s ability to tolerate complex surgery, (2) differences in tumor biology and aggressiveness in this group, resulting in tumors that are less amenable to surgical resection. Prior studies in other solid tumor types have shown that skeletal muscle gauge has high correlation with impaired physical functioning in older adults with cancer, including prolonged Timed Up and Go and impairments in instrumental activities of daily living.8 It is notable that 66% of patients who were 60 or older and had a low skeletal muscle gauge had an ASA score ≥3. This indicates that this group was indeed less fit, but the ASA score is an inherently subjective measure and is insufficient to identify the entire risk group.31 32 We previously reported differences in tumor spread pattern and resectability among specific molecular subtypes of epithelial ovarian cancer, notably that the mesenchymal subtype resulted in worse intra-peritoneal spread patterns and lower likelihood of complete resection.27 28 Ultimately, the interplay between tumor biology, surgical resectability, and pre-operative patient fitness for surgery is multifaceted, and further research is needed to understand these observations and whether they are modifiable.
Strengths and Weaknesses
Our study does have several strengths, including the size of the study cohort, detailed data repository, and availability of CT imaging for analysis. However, we do recognize several limitations, including its retrospective nature, the impact of practice changes over the course of this study, and the inability to validate the image-based body composition measurements with physical functioning. Our institution underwent several quality improvement initiatives focused on reducing treatment-related morbidity and mortality during this time period. These included the creation and continued modification of Enhanced Recovery after Surgery (ERAS) policies and selective triage of high-risk patients to neoadjuvant chemotherapy, both of which may influence results and limit overall generalizability. Additionally, we used a new imaging analysis modality created within our institution. Although this would not be readily available to other institutions, the validation performed shows that this methodology was sound. We used recursive partitioning in an exploratory manner to classify patients into homogeneous groups based on evaluating age and skeletal muscle gauge as predictors of overall survival; no validation of the identified cut-off points has yet been conducted. Further research is needed to identify cut-off points applicable to clinical decision-making.
Implications for Practice and Future Research
Older patients with cancer have been shown to receive suboptimal cancer-related treatment for multiple reasons, including perceived poor tolerability of aggressive treatment. Our research aims to better understand the complex relationships between age, sarcopenia, and solid tumor cancers, as well as the impact on patient outcomes and how to mitigate these effects. As we improve our understanding of this complex process, we aim to ultimately improve outcomes of patients at highest risk for morbidity and mortality with a more individualized care plan. This would allow for increased surgical efforts in those most fit for surgery, pre-operative optimization using prehabilitation, and relegating the use of neoadjuvant chemotherapy or chemotherapy alone in those at highest risk for death.
Next steps include evaluation of sarcopenia measures over the course of ovarian cancer and work focused on understanding the ability to mitigate body composition changes. Understanding the interplay between tumor biology and muscle biology/body composition is key to advancing this area of research and will foster a deeper understanding of the relationship between this concrete measure and the multiple factors influencing outcomes.
CONCLUSION
Sarcopenia is an age-related loss of muscle and function associated with worse outcomes in those with advanced solid cancers; sarcopenia is prevalent among patients with cancer. This study used a combined measure of sarcopenia incorporating both muscle quantity and quality obtained from clinically available CT scans and identified that those aged 60 and older with sarcopenia were at risk for poor outcomes. Continued work is needed to translate these findings into clinical practice, evaluate the influence of prehabilitation measures, and ultimately improve outcomes for this high-risk group.
Data availability statement
Data are available upon reasonable request.
Ethics statements
Patient consent for publication
Ethics approval
This study involves human participants and was approved by Mayo Clinic institutional review board ID: 13-002493. Participants gave informed consent to participate in the study before taking part.
References
Supplementary materials
Supplementary Data
This web only file has been produced by the BMJ Publishing Group from an electronic file supplied by the author(s) and has not been edited for content.
Footnotes
Presented at Sarcopenia in ovarian cancer: Sarcopenic patients have decreased survival Polen-De, C., Weaver, A., McGree, M., Jatoi, A., Cliby, W., Kumar, A. Western Association for Gynecologic Oncology June 2020 (virtual).
Contributors CP-D: conceptualization, data curation, investigation, writing – original draft; PF: data curation, investigation, writing – review and editing; ALW, MMc: formal analysis, methodology, writing – review and editing; MMo: methodology, resources, software, writing – review and editing; NT: resources, software, writing – review and editing; AJ, WC: conceptualization, methodology, supervision, writing – review and editing; NKL: conceptualization, writing – review and editing; AK: guarantor, conceptualization, investigation, sata curation, methodology, visualization, supervision, writing – review and editing.
Funding Funding provided to Dr Cliby by the Virgil S. Counsellor MD Professorship in Surgery, Mayo Clinic.
Competing interests MM has a restricted research grant Phillips Medical, unrelated to this project. Funding provided to WC by the Virgil S Counseller MD Professorship in Surgery, Mayo Clinic.
Provenance and peer review Not commissioned; externally peer reviewed.
Supplemental material This content has been supplied by the author(s). It has not been vetted by BMJ Publishing Group Limited (BMJ) and may not have been peer-reviewed. Any opinions or recommendations discussed are solely those of the author(s) and are not endorsed by BMJ. BMJ disclaims all liability and responsibility arising from any reliance placed on the content. Where the content includes any translated material, BMJ does not warrant the accuracy and reliability of the translations (including but not limited to local regulations, clinical guidelines, terminology, drug names and drug dosages), and is not responsible for any error and/or omissions arising from translation and adaptation or otherwise.
Linked Articles
- Letter
- Letter