Prognostic role of sarcopenia and body composition with long-term outcomes in obstructive colorectal cancer

Background The clinical signicance of sarcopenia in colorectal cancer obstruction has not yet been described. The present study aimed to determine the short and long-term oncologic impacts of sarcopenia in obstructive colorectal cancer. Methods


Abstract Background
The clinical signi cance of sarcopenia in colorectal cancer obstruction has not yet been described. The present study aimed to determine the short and long-term oncologic impacts of sarcopenia in obstructive colorectal cancer.

Methods
A total of 214 patients with obstructive colon cancer were included in this study, between January 2004 and December 2013. Initial staging computed tomography (CT) scans identi ed sarcopenia and visceral obesity by measuring the muscle and visceral fat areas at the third lumbar vertebra level. Both short-term postoperative and long-term oncologic outcomes were analyzed.

Conclusion
Sarcopenia is a clinical factor signi cantly associated with OS and DFS but not with short-term complications in obstructive colorectal cancer. In future, prospective studies should incorporate body composition data in patient risk assessments and oncologic prediction tools. Background Sarcopenia is de ned as a decrease in skeletal muscle volume and function [1] that has been reported to re ect patients' frailty. Major clinical guidelines have incorporated sarcopenia as a tool to assess cachexia in cancer patients [2]. Sarcopenia patients may possess unfavorable nutritional and immunological factors [3] and show lower compliance to consecutive anti-tumor treatments such as radiotherapy, surgery and chemotherapy [4].
Sarcopenia is a signi cant prognostic factor in colorectal cancer patients. In colorectal cancer surgery, sarcopenia has been reported to predict poorer postoperative short-term and oncologic long-term outcomes [5,6]. Sarcopenia patients undergoing resection of colorectal liver metastases have a shorter median survival and lower disease free survival (DFS) rates than non-sarcopenia patients [7]. Additionally, sarcopenia was negatively associated with overall survival (OS) in locally advanced rectal cancer patients who underwent neoadjuvant chemoradiation therapy and curative resection [8]. Despite this, the clinical importance of sarcopenia in obstructive colorectal cancer has not been fully described.
Obstructive colon cancer is a signi cant clinical event that affects the short-and long-term prognosis of patients. The rates of mortality and complications associated with emergency surgery for obstructive colorectal cancer are higher than those associated with elective surgeries [9, 10]. Regarding long-term oncologic prognosis, colon cancer obstruction has a negative impact on DFS and OS [11,12]. Thus, treatment guidelines recommend adjuvant chemotherapy for this high-risk group [13]. While there are several available studies on metastatic colon cancer, to our knowledge, none has focused on obstructive colon cancer. The present study aimed to determine the short and long-term oncologic impacts of sarcopenia in obstructive colorectal cancer and included a more de ned group of cancer patients.

De nitions
We have de ned obstructive colorectal cancer as pathological con rmation of adenocarcinoma originated from cecum to rectum with clinical symptom of obstruction (abdominal distention, pain, tenderness and no stool passage) and radiological nding of obstruction in computed tomography (CT) scan or failure of scope passing beyond cancer lesion. Also, we included only complete obstruction to this de nition, which clinically required immediate procedure.

Subjects
Between January 2004 and December 2013, a total of 214 consecutive patients with obstructive colorectal cancer were enrolled. The inclusion criteria were as follows: (1) Clinically (abdominal distention, pain, tenderness and no stool passage) and radiologically con rmed malignant large bowel obstruction; (2) the patients who required immediate procedure due to complete obstruction. The exclusion criteria were as follows: (1) palliative bypass surgery; (2) palliative stoma surgery; (3) palliative bypass surgery; (4) metastatic cancer; and (5) colon perforation. During initial staging computed tomography (CT) body composition evaluations were performed at the level of the transverse process of third lumbar vertebra as previously described [8]. The total body fat area, visceral fat area, subcutaneous fat area, and abdominal circumference were measured automatically (TeraRecon Aquarius Workstation, TeraRecon, Foster City, California, USA), and visceral fat area was used to identify visceral adiposity. The skeletal muscle areas (psoas, para-spinal, transverse abdominis, rectus abdominis, internal oblique, and external oblique muscles) were measured using a commercial system (Advantage Windows workstation, GE Healthcare, Milwaukee, Wisconsin, USA) entailing Houns eld unit thresholds between -29 to +150 [14]. To normalize the skeletal muscle area for patient height, skeletal muscle index (SMI) was calculated as the skeletal area (cm 2 ) divided by the square of the height (m 2 ). Sarcopenia was de ned using sexspeci c cutoff values for SMI (46.4 cm 2 / m 2 for men and 37.5 cm 2 / m 2 for women) [15]. The cutoff CT visceral fat area for the classi cation of visceral obesity has not yet been standardized. Previous studies have used 100 cm 2 and 130 cm 2 as cutoffs [16]. The present study de ned visceral obesity as an area of visceral fat of 130 cm 2 and above. Left colon cancer was de ned as cancer arising between the distal third of the transverse colon and the rectosigmoid junction of the colon.
Emergency surgery or self-expandable metallic stent (SEMS) placement was performed based on previously published data [17]. The surgeon determined the operative method for each case considering the patient's medical status and tumor location. A SEMS (HANARO stent, M. I. Tech Co., Ltd, Seoul, South Korea or Niti-S stent, Taewoong Medical, Co., Ltd, Gyeonggido, South Korea) was inserted through the working channel over the guidewire under uoroscopic guidance.
This study was approved by the St. Mary's Hospital Research Ethics Board (KC19RESI0152) and informed consent was obtained from all participants.

Statistical analysis
Chi-square and Fisher's exact tests were performed for association analysis. P-values <0.05 were considered statistically signi cant. Differences between groups were evaluated using Student's t-and x 2 tests for continuous and categorical variables, respectively. OS and DFS curves were analyzed using the Kaplan-Meier method and compared by log-rank test for univariate analysis. Multivariate analysis for survival was performed using the Cox proportional hazard model with the backward logistic regression method. Statistical analyses were performed performed using SPSS version 24.0 (IBM SPSS Statistics®, Armonk, NY, USA).

Results
A total of 214 patients (126 men and 90 women) were enrolled in this study. The detailed demographics and clinical treatment courses according to sarcopenia status are summarized in Table 1. Seventy-one patients had sarcopenia based on initial CT. Tumor-related factors (stage, location, carcinoembryonic antigen [CEA]), pathologic factors (stage, lymph node status, lymph-vascular invasion), and patientrelated factors (American Society of Anesthesiologist [ASA] physical status) did not differ signi cantly according to sarcopenia status. Age and sex have been reported to be correlated to sarcopenia status [18]. Here, sarcopenia was observed more frequently in male patients than in female patients (51 men, 70.8% vs 75 men, 52.1%, p = 0.008) The association between body composition factors and sarcopenia are shown in Table 2. The mean BMIs in the non-sarcopenia and sarcopenia groups did not differ signi cantly by sex (p = 0.486 and 0.687, respectively). The mean visceral fat area, abdominal circumference, and SMI in the non-sarcopenia group were signi cantly higher in male than in female (p < 0.001, p = 0.008 and p < 0.001, respectively). The subcutaneous fat area in the non-sarcopenia group was signi cantly higher in female than in male in the sarcopenia group (p <0.001). The mean visceral fat area, abdominal circumference and SMI in the Sarcopenia group were signi cantly higher in male than in female (p < 0.001, p < 0.001 and p < 0.001, respectively) ( Table 2).

Discussion
Sarcopenia was negatively associated with a long-term oncological prognosis. Sarcopenia patients showed signi cantly lower OS and DFS. However, there were no signi cant differences in short-term postoperative outcomes. Body composition factors other than sarcopenia did not signi cantly impact the patient's survival. There was an observed trend for a negative association between high visceral adiposity and long-term oncological prognosis, but this did not reach statistical signi cance in our patient population.
The results of our study suggest that sarcopenia is a negative prognostic factor for both OS and DFS in patients with obstructive colorectal cancer. To our knowledge, this is the rst report to describe the prognostic impact of sarcopenia speci cally in patients with obstructive colorectal cancer. Other studies reported sarcopenia to have a negative effect on OS in patients undergoing resection for locally advanced rectal cancer after neoadjuvant chemoradiation therapy [8] and as an independent predictor of worse OS and DFS in stage I to III colorectal cancer [19].
Early studies of sarcopenia were often based on the work by Prado et. al [20]. In their study, they included sarcopenia data based on solid tumors of the lung or gastrointestinal tract from patients referred to a regional medical oncology service in Canada. However, because there are signi cant differences in body composition between different ethnicities, more data are needed from the Asian populations [21].
Moreover, more data are needed to assess the optimal cutoff value for sarcopenia for each ethnicity [6]. The Asian cut-off value for sarcopenia should be different from the one used in Western countries.
Sarcopenia was associated with a signi cantly increased risk of developing major complications [22].
However, one study reported that sarcopenia was not a predictor of postoperative complications [8]. In the present study, no signi cant differences in minor (11.8% vs. 12.5%) or major postoperative complications (5.6% vs. 6.9%) were observed between non-sarcopenia and sarcopenia patients ( Table 3). All surgeries were performed at the three tertiary-referral hospitals where more than 100 colorectal cancer patients are treated annually by seven independent surgeons. These surgeons were quali ed through live demonstrations held by the Korean Laparoscopic Colorectal Surgery Study Group and each submitted a videotape of their laparoscopic rectal surgery, which was subsequently reviewed by a trial steering committee to assess the surgeon's oncological technique [23].
Visceral obesity was reported to be a signi cant prognostic factor in predicting DFS in patients with resectable colorectal cancer [24]. Similarly, viscerally obese patients with rectal cancer have poorer DFS [25]. In contrast, BMI measurements were not correlated with any survival outcomes [24,25]. Our data demonstrated that sarcopenia was negative associated with visceral obesity ( Table 2) but there were no other signi cant differences in predicting DFS and OS in patients with obstructive colorectal cancer. A lower BMI was correlated with sarcopenia but not with prognosis. SMI was the most meaningful prognostic value among body composition factors.
According to a recent meta-analysis study [6], most studies (11/22, 50%) used SMI when applying muscle mass criteria in CT scan-based assessments of skeletal muscle index (SMI). Their de ned cutoff values varied between 53.5 and 40.8 cm 2 / m 2 for men and 46.4 and 34.9 cm 2 / m 2 for women. To de ne a sarcopenia cutoff value for this study, we analyzed the association between body composition factors and sarcopenia in four different studies that included an Asian study population (46.4, 43.75, 43.2 and 40.8 cm 2 / m 2 for men and 37.5, 41.10, 35.3, 34.9 cm 2 /m 2 for women for each study, respectively) [15, 26-28]. Among these, the cutoff value proposed by Takagi was applied differently to men and women and the inclusion criteria selected in their study was similar to ours. Additionally, their selected cutoff value did not differ signi cantly from the cutoff value of other Asian references. Thus, we used the same cutoff value proposed by Takagi [15], which was ideal for our study in Asian patients.
There are some limitations to this study. First, this retrospective study involved three operators from a single institution. Therefore, a multi-institutional, prospective validation study will be necessary in future to con rm our ndings. Second, muscle function and strength such as walk test and grip test could not be measured. Third, there is still no consensus on CT-de ned cutoff values for a sarcopenia measurement. Despite these limitations, our analysis showed the oncologic signi cance of sarcopenia in obstructive colorectal cancer.
Obstructive colorectal cancer may occasionally require colonic stent insertion to improve the patient's condition and adjuvant chemotherapy treatment is also often administered to these patients. Obstructive colon cancer generally has poor prognosis and patients suffer with large clinical burdens, including sarcopenia. Results of this study suggest that in obstructive colorectal cancer sarcopenia, prevalence at initial diagnosis is negatively associated with both OS and DFS.
In conclusion, in order to overcome the large clinical burden in obstructive colon cancer patients, a careful consideration and analysis of the patient's body composition status should be performed, in particular, sarcopenia should be considered in the patients' risk assessments and oncologic prognosis strati cation.

Consent to participate
Informed consent from patients to be included in this study was omitted according to the policy of our IRB.

Consent for publication
Informed consent from patients to be included in this study was omitted according to the policy of our IRB.

Availability of data and material
The study data is not available