nueva página del texto (beta)
Inglés (pdf)
Artículo en XML
Referencias del artículo
Traducción automática
Enviar artículo por email
Citado por SciELO 
Similares en
SciELO 
Permalink
BACKGROUND
Body composition analysis by computed tomography (CT) has been studied in different medical outcomes. Adipose and muscle tissue distribution are strongly correlated to the patients' metabolic and performance status, respectively; both features could influence surgical outcomes. The visceral fat area (VFA) and skeletal muscle area (SMA) may be assessed by the quantification of each tissue in a transversal CT image at the third lumbar vertebra (L3) with excellent inter and intraobserver agreement. Both variables have been identified as prognostic factors for oncologic and surgical outcomes among colorectal cancer surgery patients. The use of image analysis to determine fat and muscle distribution is an advantage because CT is widely available among this population1,2.
Visceral obesity has been associated with longer operative times and hospital stays, as well as with increased morbidity3,4. Low skeletal muscle mass is also known to have a negative effect on post-operative outcomes5, mainly among patients with cancer, increasing their morbidity and mortality. Most data regarding SMA and adipose tissue analysis focus on the oncological population, but there is a lack of information regarding patients with benign entities. The importance of identifying the implications of SMA and VFA in surgical outcomes relies on the implementation of interventions that target SMA and VFA to improve these outcomes.
Sarcopenia is defined by the Sarcopenia Working Group as a syndrome characterized by the loss of skeletal muscle mass and strength6. Since we only analyzed muscle area and not function, we decided not to use the term sarcopenia; instead, the term SMA or skeletal muscle index (SMI) will be used accordingly. SMI is obtained by dividing the patients' SMA by their height.
The objective was to study the association of VFA and SMI with 30-day mortality in patients undergoing any colorectal surgery including primary anastomosis. A secondary aim was to analyze the association between these variables and anastomosis leak (AL).
METHODS
Study Population
This study is a retrospective analysis of all patients who underwent colorectal surgery with primary anastomosis from January 2007 to December 2018 at a tertiary referral hospital in Mexico City. We excluded 54 patients who had incomplete medical charts since the variables needed for the study could not be obtained from the electronic or physical files or because no pre-operative CT was available. Finally, only 548 patients were included in the study. All surgeries were performed by the two chief surgeons of our colorectal surgery department.
Data Collection
For each patient, the following variables were collected: (1) patient characteristics, including age, sex, preoperative albumin, diagnosis (cancer or benign), steroid use, current smoker, comorbidities calculated with Charlson comorbidity index (CCI) score, weight, height, body mass index (BMI), VFA, and SMA; (2) operative details, including the type of surgery (right hemicolectomy, left hemicolectomy, surgeries with rectal anastomosis, and others) blood loss, type of case (urgent or elective surgery), number of anastomoses, type of anastomosis (manual or stapled), approach (open or laparoscopic), conversion, protective stoma, and operative time; and (3) post-operative outcomes: AL, length of stay, 30-day readmission, and 30-day mortality. This study was approved by our institution's Research Ethics Committee with the reference number GAS 3184 19 20 1.
Body Composition Analysis
Pre-operative CTs were assessed with NIH ImageJ software, version 1.487. VFA and SMA were assessed at the level of L3. VFA was computed with a Hounsfield unit (HU) threshold between −30 and −190 HU, whereas SMA with a threshold between −29 and +150 HU. The L3-SMI was calculated by dividing SMA/height (cm2). Visceral obesity was defined as a VFA >100 cm2, and low SMI was diagnosed if patients had an SMI <52.4 cm2/m2 and <38.5 cm2/m2 in men and women, respectively8. Agreement between raters to classify patients as having or not having low SMI was assessed with Cohen's kappa. We also computed the intraclass correlation coefficient of the estimated SMA9.
Statistical Analysis
Statistical analysis was performed using Stata 14 (StataCorp, Texas, USA). Continuous variables were presented as means ± standard deviations and categorical variables as frequencies (n) and percentages. Variables were compared using the Chi-squared test or Fisher's exact test, and Student's t-test as appropriate. To study the association between body composition variables and the outcomes of interest (i.e., 30-day mortality and AL), we used logistic regression. Those variables with p < 0.05 were included in a multivariable analysis. We used the Box-Tidwell model to test the linearity of continuous variables and goodness-of-fit was assessed by the Hosmer-Lemeshow test.
RESULTS
General Characteristics of Patients
The study population consisted of 548 patients with a mean age of 57.4 + 17.4 years; 262 (47.7%) were male. The prevalence of low SMI was 52% (286 patients), and it was higher among men (60% vs. 45%, p < 0.001). The prevalence of visceral obesity was 53.7% (295 patients) and was also more common in men (69% vs. 40%, p < 0.001).
According to the BMI, among the overweight and obese patients, low SMI was found in 99 (48%) and 14 (21%), respectively; visceral obesity was found in 88 (36%) patients with normal BMI (Table 1). The prevalence of visceral obesity in conjunction with low SMI (i.e., low SMI and high VFA) was 28%; when obesity was defined by the BMI (i.e., >30) instead of VFA, the prevalence was 2.5%. The mean BMI was 25.2 + 5.1 kg/m2.
Table 1 Frequency of low SMI and visceral obesity by BMI (cm/m2) category
| Characteristics | BMI <18.5 (n = 31) | BMI 18.5-24.9 (n = 246) | BMI 25-29.9 (n = 205) | BMI>30 (n = 66) |
|---|---|---|---|---|
| Low SMIa, n (%) | 20 (64) | 153 (62) | 99 (48) | 14 (21) |
| Visceral obesityb, n(%) | 2 (6) | 88 (36) | 147 (72) | 58 (88) |
aDefined as an SMI at the level of L3 of < 52.4 cm2/m2 in men and < 38.5 cm2/m2 in women.
bDefined as visceral fat area >100 cm2.
BMI: body mass index; SMI: skeletal muscle index.
The most common surgery was right hemicolectomy in 174 patients (32.5%). Two hundred and eighteen (40%) cases were laparoscopic, and the conversion rate was 35%. Two hundred and seventy-six patients (50%) had colorectal cancer. The rest of the variables is described in Table 2.
Table 2 General population characteristics (n = 548)
| Patient characteristics | Total |
|---|---|
| Male, n (%) | 262 (48) |
| Age (years), mean + SD | 57.4 + 17.4 |
| Low SMIa, n (%) | 286 (52) |
| Visceral obesityb, n (%) | 295 (54) |
| BMI (kg/m2), mean + SD | 25.3+5.1 |
| Cancer diagnosis, n (%) | 276 (50) |
| Pre-operative albumin (g/dL), mean + SD | 3.8+0.7 |
| Current smoker, n (%) | 230 (42) |
| Steroid use, n (%) | 31 (6) |
| Operative details | |
| Type of surgery, n (%) | |
| Right hemicolectomy | 174 (32) |
| Sigmoidectomy | 128 ( 23) |
| Stoma closure | 75 (14) |
| Total colectomy | 25 (4) |
| Left hemicolectomy | 37 (7) |
| Low anterior resection | 42 (8) |
| Proctocolectomy + J pouch | 26 (5) |
| Othersc | 116 (21) |
| Estimated blood loss (mL), mean + SD | 374.5+371.5 |
| Emergency surgery, n (%) | 92 (17) |
| Surgical time (minutes), mean + SD | 247.0+94.7 |
| Manual anastomosis, n (%) | 59 (11) |
| Laparoscopic surgery, n (%) | 218 (40) |
| Conversion, n (%) | 71 (35) |
| Loop ileostomy, n (%) | 157 (29) |
| Post-operative outcomes | |
| 30-day readmission, n (%) | 38 (7) |
| 30-day mortality, n (%) | 23 (4) |
| AL, n (%) | 53 (9.7) |
| Clavien-Dindo, n (%) | |
| I | 76 (14.10) |
| II | 67 (12.43) |
| IIIA | 21 (3.90) |
| IIIB | 33 (6.12) |
| IVA | 42 (7.79) |
| IVB | 25 (4.64) |
| V | 23 (4) |
BMI: body mass index; SMI: skeletal muscle index;
SD: standard deviation; AL: anastomosis leak.
aDefined as an SMI at the level of L3 of <52.4 cm2/m2 in men and <38.5 cm2/m2 in women.
bDefined as visceral fat area >100 cm2.
cStoma closure and segmental resection.
Patients with low SMI were older (60.4 + 18.2 vs. 54.2 + 16.0 years; p < 0.001), had a lower BMI (23.9 + 3.8 vs. 26.8 + 5.8 kg/m2; p < 0.001), and had a higher 30-day mortality (6.6% vs. 1.5%; p = 0.003) than patients with a normal SMI. There were no differences in length of stay, 30-day readmission, or frequency of AL. Compared to non-obese patients, those with visceral obesity were more commonly men (62% vs. 32%; p < 0.001) and were older (60.2 + 15.3 vs. 54.1 + 19.1 years; p < 0.001). Table 3 compares patients according to their body composition. The 30-day mortality was similar in patients with and without visceral obesity in conjunction with low SMI (6.5% vs. 3.3%; p = 0.09). Cohen's kappa coefficient for the assessment of low SMI was 0.76 (p < 0.001), and the intraclass correlation coefficient for SMA was 0.96 (p < 0.001).
Table 3 General characteristics by body composition
| Variable | Normal SMI (n = 262) | Low SMIa (n = 286) | p-value | No visceral obesity (n = 253) | Visceral obesityb (n = 295) | p-value |
|---|---|---|---|---|---|---|
| Pre-operative characteristics | ||||||
| Male, n (%) | 105 (40) | 157 (55) | <0.001 | 80 (32) | 182 (62) | <0.001 |
| Age (years), mean + SD | 54.2 +16.0 | 60.4 +18.2 | <0.001 | 54.1 + 19.1 | 60.2 + 15.3 | <0.001 |
| Charlson comorbidity index >2, n (%) | 119 (45) | 156 (54) | 0.03 | 120 (47) | 155 (52) | 0.2 |
| Cancer diagnosis, n (%) | 134 (51) | 142 (50) | 0.7 | 132 (52) | 144 (49) | 0.2 |
| BMI (kg/m2), mean + SD | 26.8 + 5.8 | 23.9 + 3.8 | <0.001 | 22.8 + 3.8 | 27.4 + 5.1 | <0.001 |
| Albumin (g/dL), mean + SD | 3.8 + 0.8 | 3.8+0.7 | 0.4 | 3.7 + 0.8 | 3.9 + 0.7 | 0.07 |
| Surgical characteristics | ||||||
| Surgical time (min), mean + SD | 252.9+91.2 | 241.6 + 97.5 | 0.2 | 242.9 + 94.1 | 250.5 + 95.2 | 0.4 |
| Manual anastomosis, n (%) | 29 (11) | 30 (10) | 0.8 | 34 (13) | 25 (8) | 0.06 |
| Laparoscopic, n (%) | 119 (45) | 99 (35) | 0.009 | 92 (36) | 126 (43) | 0.1 |
| Blood loss (mL), mean + SD | 389.1.9 + 402.6 | 361.2 + 340.6 | 0.4 | 347.4 + 359.2 | 397.8 + 380.7 | 0.1 |
| Conversion, n (%) | 38 (34) | 33 (36) | 0.8 | 24 (28) | 47 (41) | 0.05 |
| Loop ileostomy, n (%) | 81 (31) | 76 (26) | 0.2 | 75 (29) | 82 (28) | 0.6 |
| Post-operative outcomes | ||||||
| AL, n (%) | 24 (9) | 29 (10) | 0.7 | 20 (8) | 33 (11) | 0.2 |
| Length of stay (days), mean + SD | 14.3 + 23.7 | 14. 7 + 24.8 | 0.8 | 13.9 + 23.1 | 15.0 + 25.3 | 0.6 |
| 30-day readmission, n (%) | 18 (7) | 20 (7) | 0.9 | 19 (8) | 19 (6) | 0.6 |
| 30-day mortality, n (%) | 4 (1.5) | 19 (6.6) | 0.003 | 10 (4) | 13 (4) | 0.8 |
aDefined as an SMI at the level of L3 of <52.4 cm2/m2 in men and <38.5 cm2/m2 in women.
bDefined as visceral fat area >100 cm2.
BMI: body mass index; SD: standard deviation; SMI: skeletal muscle index. AL: anastomosis leak.
Features Associated with Mortality and AL
Thirty-day mortality was 4%. On univariable analysis (Table 4), factors associated with mortality were low SMI (odds ratio [OR] 4.59, 95% confidence interval [CI] 1.54-13.67; p = 0.006), the L3-SMI (OR 0.92, 95% CI 0.87-0.97; p = 0.003), AL (OR 8.62, 95% CI 3.57-20.82; p < 0.001), pre-operative albumin (OR 0.34, 95% CI, 0.20-0.56; p < 0.001), estimated blood loss (OR 1.00, 95% CI 1.00-1.001; p = 0.02), age (OR 1.06, 95% CI 1.03-1.09; p < 0.001), CCI score >2 (OR 10.30, 95% CI 3.75-28.26; p < 0.001), steroid use (OR 5.33, 95% CI 1.83-15.48; p = 0.002), and other type of surgery (i.e., total colectomy, proctocolectomy, sigmoidectomy, and low anterior resection) (OR 2.78, 95% CI 0.95-8.07; p = 0.06). On multivariable analysis (Table 4), low SMI remained an independent risk factor for 30-day mortality as well as AL, age, steroid use, CCI >2, pre-operative serum albumin levels, estimated blood loss, and other types of surgery. The same variables were associated with 30-day mortality when including L3-SMI instead of low SMI in the multivariable analysis.
Table 4 Variables associated with 30-day mortality on univariable and multivariable analysis
| Variable | Univariable | Multivariablea | Multivariableb | |||
|---|---|---|---|---|---|---|
| OR (95% CI) | p-value | OR (95% CI) | p-value | OR (95% CI) | p-value | |
| Low SMIc | 4.59 (1.54-13.67) | 0.006 | 4.74 (1.22-18.36) | 0.02 | – | – |
| L3-SMI, cm2/m2 | 0.92 (0.87-0.97) | 0.003 | – | – | 0.93 (0.87-0.99) | 0.03 |
| Visceral obesityd | 1.12 (0.48-2.60) | 0.8 | – | – | – | – |
| VFA, cm2 | 1.00 (0.99-1.00) | 0.3 | – | – | – | – |
| Low L3-SMI and visceral obesity | 2.05 (0.88-4.79) | 0.09 | – | – | – | – |
| AL | 8.62 (3.57-20.82) | <0.001 | 5.08 (1.53 – 16.85) | 0.008 | 5.66 (1.70-18.83) | 0.005 |
| Albumin, g/dL | 0.34 (0.20-0.56) | 0.001 | 0.33 (0.16-0.68) | 0.002 | 0.36 (0.17-0.73) | 0.005 |
| Blood loss, mL | 1.00 (1.00-1.001) | 0.02 | 1.00 (0.99-1.00) | 0.3 | 1.00 (0.99-1.00) | 0.5 |
| Male sex | 1.20 (0.52-2.76) | 0.7 | – | – | – | – |
| Age, years | 1.06 (1.03-1.09) | <0.001 | 1.06 (1.01-1.09) | 0.007 | 1.06 (1.02-1.10) | 0.003 |
| Steroid use | 5.33 (1.83-15.48) | 0.002 | 4.87 (1.09-21.64) | 0.04 | 4.61 (1.09-19.51) | 0.04 |
| CCI >2 | 10.30 (3.75-28.26) | <0.001 | 4.64 (1.37-15.66) | 0.01 | 4.32 (1.30-14.33) | 0.02 |
| BMI, kg/m2 | 0.98 (0.90-1.08) | 0.8 | ||||
| Presence >1 anastomosis | 1.62 (0.36-7.28) | 0.5 | ||||
| Loop ileostomy | 1.09 (0.44-2.71) | 0.8 | ||||
| Type of surgery | ||||||
| Right hemicolectomy | Reference | 0.9 | Reference | 0.6 | Reference | 0.7 |
| Left hemicolectomy | 0.94 (0.11-8.28) | 1.65 (0.14-19.49) | 1.53 (0.11-19.71) | |||
| Surgeries with rectal anastomosise | 0.97 (0.27-3.41) | 0.9 | 1.71 (0.38-7.74) | 0.4 | 2.21 (0.49-9.80) | 0.3 |
| Others | 2.78 (0.95-8.07) | 0.06 | 8.89 (2.04-38.78) | 0.004 | 9.59 (2.20-41.73) | 0.003 |
aConsidering low SMI,
bconsidering L3-SMI.
cDefined as a SMI at the level of L3 of <52.4 cm2/m2 in men and <38.5 cm2/m2 in women.
dDefined as VFA >100 cm2.
eTotal colectomy, proctocolectomy, sigmoidectomy, and low anterior resection.
BMI: body mass index; CI: confidence interval; OR: odds ratio; SMI: skeletal muscle index. VFA: visceral fat area; AL: anastomosis leak;
CCI: Charlson comorbidity index.
AL developed in 53 (9.7%) patients; blood loss, CCI>2, and other types of surgery were variables associated with AL in univariable and multivariable analyses (Table 5).
Table 5 Variables associated with AL on univariable and multivariable analysis
| Variable | Univariable | Multivariable | ||
|---|---|---|---|---|
| OR (95% CI) | p-value | OR (95% CI) | p-value | |
| Low SMIa | 1.11 (0.63-1.97) | 0.7 | ||
| L3-SMI, cm2/m2 | 1.00 (0.97-1.03) | 0.7 | ||
| Visceral obesityb | 1.47 (0.82-2.62) | 0.2 | ||
| VFA, cm2 | 1.00 (0.99-1.00) | 0.1 | ||
| Low L3-SMI and visceral obesity | 1.37 (0.75-2.50) | 0.3 | ||
| Albumin, g/dL | 0.76 (0.53-1.10) | 0.1 | ||
| Blood loss, mL | 1.00 (1.00-1.001) | 0.006 | 1.00 (1.00-1.001) | 0.007 |
| Male sex | 1.36 (0.77-2.40) | 0.3 | ||
| Age, years | 1.01 (0.99-1.02) | 0.2 | ||
| Steroid use | 1.41 (0.47-4.21) | 0.5 | ||
| CCI >2 | 1.89 (1.05-3.40) | 0.03 | 2.07 (1.13-3.81) | 0.02 |
| BMI, kg/m2 | 1.03 (0.97-1.08) | 0.2 | ||
| Loop ileostomy | 1.32 (0.72-2.40) | 0.4 | ||
| Type of surgery | ||||
| Right hemicolectomy | Reference | Reference | ||
| Left hemicolectomy | 2.86 (0.90-9.10) | 0.07 | 3.36 (1.03-10.89) | 0.04 |
| Surgeries with rectal anastomosisc | 1.79 (0.77-4.18) | 0.2 | 1.96 (0.83-4.65) | 0.1 |
| Others | 3.12 (1.39-6.97) | 0.005 | 3.26 (1.44-7.37) | 0.004 |
aDefined as a SMI at the level of L3 of <52.4 cm2/m2 in men and <38.5 cm2/m2 in women.
bDefined as VFA > 100 cm2. cTotal colectomy, proctocolectomy, sigmoidectomy, and low anterior resection.
BMI: body mass index; CI: confidence interval; OR: odds ratio; SMI: skeletal muscle index. VFA: visceral fat area; AL: anastomosis leak;
CCI: Charlson comorbidity index.
A sub-analysis of geriatric and oncological patients can be found in Supplementary Tables 1 and 2.
DISCUSSION
Body composition assessment through radiological evaluation has become an important pre-operative tool to identify patients with increased risk of poorer surgical outcomes. In this study, we investigated the association between SMA, assessed with the SMI, and VFA with 30-day mortality among colorectal surgery patients.
In this work, we found low SMI to be independently associated with 30-day mortality but not with AL. On the other hand, neither visceral obesity nor visceral obesity in conjunction with low SMI was associated with any of these outcomes. To the best of our knowledge, this is the first study comparing the influence of both visceral obesity and SMI on post-operative outcomes of patients undergoing colorectal surgery for oncologic and benign disorders in a Mexican population.
The prevalence of low SMI in our population was 52%, and 54% of patients had visceral obesity. Despite the high prevalence of low SMI in oncological patients, no difference in low SMI prevalence was found between oncological and non-oncological patients in our study. This underlines the importance of assessing muscle mass and function in all surgical patients, and not only in elderly or oncological patients.
We found low SMI to be an independent predictor of 30-day mortality after colorectal surgery, as it has been previously described in patients with colorectal cancer10-13. Low SMI may imply a prolonged state of catabolism, impaired host immune function, and inadequate response to an inflammatory state14.
An association could not be found between AL and VFA or SMI, as described in other retrospective studies10,15-17. Even a prospective cohort study of 80 patients who underwent elective surgery for colon cancer found no relationship between SMI and any post-operative morbidity18.
In opposition to this, other authors10,19–23 have found low SMI to be a predictor of overall complications and length of stay. Herrod et al.21 evaluated psoas density as a surrogate for SMI and found that patients with smaller psoas muscles had an OR of 14.37 for AL and 6.33 for overall complications, though mortality was not affected. The morphologic change of the psoas has also been described as an independent risk factor for overall complications and AL in surgical colorectal cancer patients24. In addition, sarcopenic obesity in patients with gastrointestinal solid tumors has also been associated with8, longer length of stay, and higher readmission rates24. In our population, visceral obesity in conjunction with low SMI was not related to any adverse outcomes.
Other predictive factors for 30-day mortality besides skeletal muscle in multivariate analysis were AL, pre-operative albumin levels, age, CCI >2, "other" types of surgery, and steroid use. AL is a very threatening complication in colorectal surgery, whose association with 30-day has been previously described25.
Concerning pre-operative albumin, patients with lower albumin may have a poorer nutritional status that relates to negative outcomes26. Patients with greater estimated blood loss may have had more difficult operative courses, linked to longer operative times and more blood transfusions. All these elements may be related to adverse post-operative outcomes and, thus, more mortality. Moreover, blood loss has been related to a depleting effect in the immune system27,28. Steroid use may be associated with mortality because of impaired tissue healing and immune response29. Mortality is also expected to be more prevalent among older people, with more comorbidities as our study suggested. The increased 30-day mortality among patients who had other types of surgery (including stoma closure and segmental colonic resection) cannot be explained, as it is expected that surgeries with rectal anastomosis are more challenging25. However, since these surgeries were also related to AL, a high 30-day mortality is expected.
Visceral obesity has been related to adverse outcomes in colorectal surgery30-33, including a higher incidence of systemic complications presumably caused by longer operative times, a higher frequency of metabolic, cardiovascular comorbidities34, and more operative blood loss35. At least two meta-analyses by Yang et al.36 and Cakir et al.4 found that post-operative complications in laparoscopic procedures were more frequent in patients with visceral obesity. However, we could not corroborate these data. Visceral obesity was not associated with mortality or AL. Other authors have also found no increased morbidity among patients with visceral obesity37-39, and the reason behind this discrepancy may be that visceral obesity mainly seems to affect laparoscopic approaches, and only 40% of the cases in our cohort were laparoscopic. A meta-analysis on the association between visceral obesity and AL could not draw appropriate conclusions due to the heterogeneity between studies, but a tendency against an association between AL and visceral obesity was established40.
Our study has several limitations, the most important one being its retrospective nature. Although an association between low SMI and mortality was found, prospective studies are needed to confirm the association and explore whether preconditioning and muscle mass improvement may reduce mortality. Another limitation is that our study only evaluates SMI. Muscle function was not considered in this analysis, so its interaction with surgical outcomes could not be described. Finally, the number of outcomes in our study was limited, and therefore, the results of the multivariable analysis should be interpreted with caution.
In conclusion, SMI, but not visceral obesity, was independently associated with 30-day mortality. Neither low SMI nor visceral obesity was associated with AL. The role of preconditioning in elective surgeries to improve outcomes needs to be further explored.
Supplementary Data
Supplementary data are available at Revista de Investigación Clínica online (www.clinicalandtranslationalinvestigation.com). These data are provided by the corresponding author and published online for the benefit of the reader. The contents of supplementary data are the sole responsibility of the authors.
