Research ArticleClinical Studies

Dorsal Subcutaneous Fat Thickness as a Risk Factor for Pancreatic Fistula After Gastric Cancer Surgery

HIRONARI MIYAMOTO, MAMI YOSHII, HIROAKI TANAKA, SATOSHI NISHI, SOTA DEGUCHI, YUICHIRO MIKI, TATSURO TAMURA, TAKAHIRO TOYOKAWA, SHIGERU LEE and KIYOSHI MAEDA
Anticancer Research May 2023, 43 (5) 2171-2178; DOI: https://doi.org/10.21873/anticanres.16379

Abstract

Background/Aim: Recently, there have been many reports on the use of preoperative body composition analysis to predict postoperative complications in gastric cancer surgery, most of which used 3D image analysis software for the measurements. This study aimed to evaluate the risk of postoperative infectious complications (PICs), especially pancreatic fistulas, using a simple measurement method incorporating only preoperative computed tomography images. Patients and Methods: A total of 265 patients with gastric cancer underwent laparoscopic or robot-assisted gastrectomy with lymph node dissection at Osaka Metropolitan University Hospital between 2016 and 2020. To simplify the measurement method, we measured the length of each region of the subcutaneous fat area (SFA). Each area included a) umbilical depth, b) thickness of the longest ventral subcutaneous fat, c) thickness of the longest dorsal subcutaneous fat, and d) thickness of the median dorsal subcutaneous fat (MDSF) measurements. Results: PICs occurred in 27 of 265 cases, of which pancreatic fistula was present in 9. SFA for pancreatic fistulas showed high diagnostic accuracy (area under the curve=0.922). Among the subcutaneous fat lengths, the MDSF was the most useful, and the optimal cut-off value was 16 mm. MDSF and non-expert surgeons were found to be independent risk factors for pancreatic fistula. Conclusion: Since the possibility of developing pancreatic fistula is high in cases with MDSF ≥16 mm, careful surgical strategies, such as having a skilled physician, are necessary.

Key Words:

Recently, there have been many reports on the use of preoperative body composition analysis to predict postoperative complications of gastric cancer surgery. In particular, there are many reports that infectious complications, such as pancreatic fistula and anastomotic leakage, are increased in patients with a high visceral fat content (1-10). The reasons mentioned include the high amount of fat in the abdominal cavity, which restricts the operative field and obscures the anatomy, and the thick mesentery, which places tension on the anastomosis (5, 6). We did not find any reports discussing the relationship between the amount of subcutaneous fat and postoperative infectious complications (PICs), especially pancreatic fistulas. In addition, most previously reported methods of visceral fat measurement use 3D image analysis software and preoperative computed tomography (CT) images, which is a somewhat complicated process for some institutions. Therefore, this study examined the possibility of predicting postoperative pancreatic fistula by using a simpler method of measurement that utilized measuring the subcutaneous fat thickness only from preoperative CT images.

Patients and Methods

Ethical approval. This study was conducted in accordance with the declaration of Helsinki and the ethical approval was provided by the Osaka City University ethics committee (approval no. 4038). At the time of providing consent for their surgical procedure, all patients also provided consent for their medical records to be used for research purposes.

Patients. A total of 265 patients with gastric cancer underwent laparoscopic or robot-assisted gastrectomy with lymph node dissection at Osaka Metropolitan University Hospital between 2016 and 2020. According to Japanese guidelines, all patients underwent distal, proximal, or total gastrectomy, and lymph node dissection was selected as D1, 1+, or 2. A drain was placed near the anastomosis site in all patients.

Data collection. Perioperative elements were extracted from a database maintained by the hospital, and detailed patient information was obtained from their medical records. Visceral fat was measured in a single slice at the umbilical level using CT before surgery, and SYNAPSE VINCENT (FUJIFILM, Tokyo, Japan) software was used as the image analysis software to control the study. As a tissue-specific threshold, adipose tissue was automatically measured at −150 to −50 HU to collect the subcutaneous fat area (SFA) and visceral fat area (VFA). As a simpler measurement method for the present purpose, the length of each area of subcutaneous fat, at the umbilical level, was measured and collected from preoperative CT images. The definition of each measurement was as follows: a) umbilical depth (UD), measured as the distance from the epidermis to the peritoneum at the umbilicus; b) thickness of the longest ventral subcutaneous fat (LVSF), and c) thickness of the longest dorsal subcutaneous fat (LDSF), measured as the distances in the subcutaneous fat considered the longest when a line was drawn vertically from the muscle or bone to the epidermis on the ventral and dorsal aspects, respectively; and d) thickness of the median dorsal subcutaneous fat (MDSF), measured as the distance from the middle of the erector spinae muscles to the epidermis (Figure 1).

Figure 1.
Figure 1.

Coronal computed tomography image at the level of the umbilicus to show how the subcutaneous fat measurements were obtained. a) umbilical depth, b) thickness of longest ventral subcutaneous fat, c) thickness of longest dorsal subcutaneous fat, d) thickness of middle dorsal subcutaneous fat. Key: blue: subcutaneous fat, red: visceral fat, green: psoas major muscle.

Definition of PICs and pancreatic fistula. PICs were defined as pancreatic fistula, anastomotic leakage, intra-abdominal abscess, or surgical site infection (SSI) of grade 2 or higher on the Clavien-Dindo classification (11). Among them, pancreatic fistula was defined as a drainage amylase level of at least three times the serum amylase level for at least 3 postoperative days, regardless of drainage volume, and grade B or higher based on the criteria of the International Study Group of Pancreas Fistula (12).

Statistical analysis. All statistical analyses were performed using EZR (Saitama Medical Center, Jichi Medical University, Saitama, Japan) (13), which is a statistical software that extends the capabilities of R (The R Foundation for Statistical Computing, Vienna, Austria) and the R commander. All continuous variables were analysed using Student’s t-test. The association between each perioperative factor, including PICs and SFA, was analysed using the χ2 test. Univariate and multivariate analyses were performed using logistic regression models. A receiver operating characteristic (ROC) analysis was performed to confirm the predictive ability of each element in predicting pancreatic fistula by evaluating the area under the curve (AUC), and the appropriate cut-off values were determined. Statistical significance was set at a two-tailed probability p-value of 0.05.

Results

Factors associated with PICs. Of the 265 patients included in the study, 27 had PICs, with a frequency of 10.2%. The patient characteristics, based on the presence of PICs, are shown in Table I. No significant differences were found in sex, age, or body mass index (BMI) as an indicator of body composition; however, significant differences were found in SFA, LVSF, and MDSF. No significant differences were found in surgical factors, such as surgical procedures, degree of lymph node dissection, operative time, amount of blood loss, or expert surgeon (the holder of the qualification system for endoscopic surgical skill by the Japanese Society for Endoscopic Surgery). In the multivariate analysis with LVSF and MDSF as factors, none of them was an independent risk factor for the development of PICs.

View this table:
Table I.

Patient characteristics, operative details, and univariate and multivariate analysis of factors associated with postoperative infectious complications (PICs).

Relationship between pancreatic fistula and subcutaneous fat. Among the different PICs, we chose to focus on pancreatic fistula. A pancreatic fistula was found in 9 (3.4%) of the 265 cases (Table II). There was a significant difference in the development of pancreatic fistula in the SFA and also in whether the surgeon was an expert. The results of the ROC analysis showed an AUC of 0.922, 95% confidence interval (CI) of 0.879-0.965, sensitivity of 100%, and specificity of 84.0%, indicating a high diagnostic accuracy (Figure 2).

View this table:
Table II.

Patient characteristics, operative details, and univariate and multivariate analysis of factors associated with pancreatic fistula.

Figure 2.
Figure 2.

Receiver operating characteristic analysis for predicting pancreatic fistula by subcutaneous fat area (SFA). Pancreatic fistula by SFA revealed high diagnostic accuracy [area under the curve (AUC)=0.922, sensitivity 100, specificity 84.0]. The white arrow indicates the optimal cut-off point.

Relationship between visceral and subcutaneous fat and appropriate assessment methods to predict pancreatic fistula. A positive correlation was found between SFA and VFA (R=0.5, p<0.001) (Figure 3). In addition, as a simple measurement method for this study, each subcutaneous fat thickness (UD, LVSF, LDSF, and MDSF) measured on preoperative CT images showed significant differences between patients with and without pancreatic fistulas (Table II). When a ROC analysis was performed for each, MDSF showed the highest diagnostic accuracy (AUC 0.868, 95%CI=0.796-0.940, sensitivity 100%, specificity 68.8%), and the optimal cut-off value was 16 mm (Figure 4). All patients who developed pancreatic fistulas had an MDSF of ≥16 mm (Figure 5). In the multivariate analysis with MDSF and expert surgeon as factors, both MDSF and non-expert surgeons were independent risk factors.

Figure 3.
Figure 3.

There is a positive correlation (R=0.5, p<0.001) between subcutaneous fat area (SFA) and visceral fat area (VFA).

Figure 4.
Figure 4.

Receiver operating analysis for predicting pancreatic fistula by (a) umbilical depth, (b) thickness of the longest ventral subcutaneous fat, (c) thickness of the longest dorsal subcutaneous fat, and (d) thickness of the median dorsal subcutaneous fat. Pancreatic fistula by thickness of the median dorsal subcutaneous fat had the best diagnostic accuracy and revealed moderate diagnostic accuracy [area under the curve (AUC)=0.868, sensitivity 100, specificity 68.8]. The arrowhead indicates the optimal cut-off point.

Figure 5.
Figure 5.

All cases with actual pancreatic fistulas have a thickness of the median dorsal subcutaneous fat of ≥16 mm.

Effect of sex on the diagnostic accuracy of MDSF for pancreatic fistula. Results by sex are shown in Table III. In men, pancreatic fistulas were found in 5 (2.7%) of 185 eligible cases and in 5% of women. For each category, significant differences were observed in SFA, MDSF, and expert surgeon for both sexes. Similarly, ROC analysis for predicting the development of pancreatic fistula in MDSF by sex showed an AUC=0.891 (95%CI=0.797-0.985, sensitivity 100%, specificity 73.3%) with a cut-off value of 16 mm for men and an AUC=0.821 (95%CI=0.685-0.957, sensitivity 100%, specificity 63.2%) with a cut-off value of 17 mm for women (Figure 6).

View this table:
Table III.

Patient characteristics by sex.

Figure 6.
Figure 6.

Pancreatic fistula by thickness of the median dorsal subcutaneous fat in men reveals moderate diagnostic accuracy [area under the curve (AUC)=0.891, sensitivity 100, specificity 73.3]. The black arrowhead indicates the optimal cut-off point. Its value is 16. Pancreatic fistula by thickness of the median dorsal subcutaneous fat in women reveals moderate diagnostic accuracy (AUC=0.821, sensitivity 100, specificity 63.2). The white arrowhead indicates the optimal cut-off point. Its value is 17.

Discussion

In this study, MDSF and non-expert surgeons were identified as risk factors for postoperative pancreatic fistula in laparoscopic gastric cancer surgery. Moreover, SFA helped predict the development of pancreatic fistula, and MDSF was shown to be a simple index for the quantitative evaluation of SFA. Many reports have shown that infectious complications, such as pancreatic fistula and anastomotic leakage, increase in patients with high visceral fat mass as it is a risk factor for PICs (1-10). In particular, risk factors for pancreatic fistula include high BMI, high VFA, surgical procedure (total gastrectomy), degree of lymph node dissection (suprapancreatic marginal lymph node), anatomy of the pancreas, blunt trauma from compression of the pancreas, and burns from the continuous use of energy devices (1, 14, 15). A higher VFA tends to be associated with a higher surgical difficulty (higher open conversion rate) and higher complication rates (16), and in cases with higher VFA, blood loss and operative time were greater in non-expert surgeons than in expert surgeons (17). One of the reasons for this is that a large amount of fat in the abdominal cavity limits the operative field, obscuring the anatomy, and compromising the development of the surgical field. In particular, it is thought that the boundary between fat and pancreatic tissue becomes unclear, causing pancreatic injury, which leads to a pancreatic fistula. In this study, SFA for pancreatic fistulas showed very high diagnostic accuracy (AUC=0.922). These results indicate that SFA can predict the development of pancreatic fistulas. The reason for this may be that although body fat distribution varies among individuals, cases with high SFA tend to also have high VFA.

The measurement of SFA and VFA requires 3D analysis software, which may be somewhat complicated in some institutions. Based on the results of this study, we identified four locations that could be measured in a straight line with a single slice of CT as a simple measurement method. Among these, UD was considered affected by individual differences owing to its shape, and LVSF and LDSF were considered inappropriate because of their large measurement error since they measure the perpendicular line at the point we assume to be the longest and was not corroborated by the measurement software. As a result, MDSF was easy to measure, showed the highest statistical diagnostic accuracy, and appeared to be a very useful index. It is unclear why MDSF was not an independent risk factor for PICs as a whole (Table I) but only for pancreatic fistulas (Table II). However, MDSF may better reflect anatomical factors such as high peripancreatic fat content. Of the 265 cases in this study, we observed 90 with MDSF ≥16 mm, which included all cases of pancreatic fistula. In addition, since women tend to have more subcutaneous fat in general, it was expected that sex would affect the results. Therefore, when we examined the results separately for men and women, the results were similar to those without sex discrimination, and the cut-off values were almost the same. The results suggest that MDSF may be a useful indicator for both sexes.

This study appears to be the first to report that the MDSF length appears to be correlated with the incidence of pancreatic fistula formation, as we could not find any past literature on the subject. In addition to fatty tissue, other factors associated with pancreatic fistulae include non-expert surgeons. Clearly, high-quality surgical skills and careful surgical manipulation by the surgeon are essential in addition to patient factors. Especially for surgeons performing laparoscopic gastric cancer surgery for the first time, cases with MDSF ≥16 mm should be avoided.

One of the limitations of this study is that it was conducted with a small number of patients from a single institute. Another limitation is that since all data were from Japanese individuals, data on racial differences cannot be presented, and caution should be exercised when using defined cut-off values for other races with different body sizes. Another limitation is that the multivariate analysis was performed in a small number of cases.

In conclusion, in laparoscopic gastric cancer surgery, since MDSF and surgeon skills are strongly related to the development of pancreatic fistula, and the likelihood of developing a pancreatic fistula is higher in cases with an MDSF of ≥16 mm, careful surgical strategies, such as having a skilled surgeon perform the surgery, are necessary.

Acknowledgements

The Authors would like to thank Editage (www.editage.com) for English language editing.

Footnotes

  • Authors’ Contributions

    H.M. acquired, analysed, and interpreted the data, confirmed the authenticity of the data, and drafted the manuscript. M.Y. and H.T. made substantial contributions to the conception and design of the study, interpreted the data, confirmed the authenticity of the data, and revised the manuscript critically. S.N., S.D., Y.M., T.Ta., T.To., and S.L. acquired and analysed the data. K.M. contributed to the conception and design of the study and revised the manuscript critically. All Authors read and approved the final manuscript.

  • Conflicts of Interest

    There are no financial or other interests with regard to the submitted manuscript that might be construed as conflicts of interest.

  • Received February 7, 2023.
  • Revision received February 22, 2023.
  • Accepted February 23, 2023.

References

  1. ↵
    1. Tanaka K,
    2. Miyashiro I,
    3. Yano M,
    4. Kishi K,
    5. Motoori M,
    6. Seki Y,
    7. Noura S,
    8. Ohue M,
    9. Yamada T,
    10. Ohigashi H and
    11. Ishikawa O
    : Accumulation of excess visceral fat is a risk factor for pancreatic fistula formation after total gastrectomy. Ann Surg Oncol 16(6): 1520-1525, 2009. PMID: 19267237. DOI: 10.1245/s10434-009-0391-y
    OpenUrlCrossRefPubMed
    1. Tokunaga M,
    2. Hiki N,
    3. Fukunaga T,
    4. Ogura T,
    5. Miyata S and
    6. Yamaguchi T
    : Effect of individual fat areas on early surgical outcomes after open gastrectomy for gastric cancer. Br J Surg 96(5): 496-500, 2009. PMID: 19358176. DOI: 10.1002/bjs.6586
    OpenUrlCrossRefPubMed
    1. Kunisaki C,
    2. Makino H,
    3. Takagawa R,
    4. Sato K,
    5. Kawamata M,
    6. Kanazawa A,
    7. Yamamoto N,
    8. Nagano Y,
    9. Fujii S,
    10. Ono HA,
    11. Akiyama H and
    12. Shimada H
    : Predictive factors for surgical complications of laparoscopy-assisted distal gastrectomy for gastric cancer. Surg Endosc 23(9): 2085-2093, 2009. PMID: 19116746. DOI: 10.1007/s00464-008-0247-8
    OpenUrlCrossRefPubMed
    1. Yoshikawa K,
    2. Shimada M,
    3. Kurita N,
    4. Iwata T,
    5. Nishioka M,
    6. Morimoto S,
    7. Miyatani T,
    8. Komatsu M,
    9. Mikami C and
    10. Kashihara H
    : Visceral fat area is superior to body mass index as a predictive factor for risk with laparoscopy-assisted gastrectomy for gastric cancer. Surg Endosc 25(12): 3825-3830, 2011. PMID: 21688079. DOI: 10.1007/s00464-011-1798-7
    OpenUrlCrossRefPubMed
  5. ↵
    1. Sugisawa N,
    2. Tokunaga M,
    3. Tanizawa Y,
    4. Bando E,
    5. Kawamura T and
    6. Terashima M
    : Intra-abdominal infectious complications following gastrectomy in patients with excessive visceral fat. Gastric Cancer 15(2): 206-212, 2012. PMID: 21993853. DOI: 10.1007/s10120-011-0099-0
    OpenUrlCrossRefPubMed
  6. ↵
    1. Miyaki A,
    2. Imamura K,
    3. Kobayashi R,
    4. Takami M and
    5. Matsumoto J
    : Impact of visceral fat on laparoscopy-assisted distal gastrectomy. Surgeon 11(2): 76-81, 2013. PMID: 22840236. DOI: 10.1016/j.surge.2012.07.001
    OpenUrlCrossRefPubMed
    1. Takeuchi M,
    2. Ishii K,
    3. Seki H,
    4. Yasui N,
    5. Sakata M,
    6. Shimada A and
    7. Matsumoto H
    : Excessive visceral fat area as a risk factor for early postoperative complications of total gastrectomy for gastric cancer: a retrospective cohort study. BMC Surg 16(1): 54, 2016. PMID: 27494994. DOI: 10.1186/s12893-016-0168-8
    OpenUrlCrossRefPubMed
    1. Zhang Y,
    2. Wang JP,
    3. Wang XL,
    4. Tian H,
    5. Gao TT,
    6. Tang LM,
    7. Tian F,
    8. Wang JW,
    9. Zheng HJ,
    10. Zhang L,
    11. Gao XJ,
    12. Li GL and
    13. Wang XY
    : Computed tomography-quantified body composition predicts short-term outcomes after gastrectomy in gastric cancer. Curr Oncol 25(5): e411-e422, 2018. PMID: 30464692. DOI: 10.3747/co.25.4014
    OpenUrlCrossRefPubMed
    1. Wang SL,
    2. Ma LL,
    3. Chen XY,
    4. Zhou DL,
    5. Li B,
    6. Huang DD,
    7. Yu Z,
    8. Shen X and
    9. Zhuang CL
    : Impact of visceral fat on surgical complications and long-term survival of patients with gastric cancer after radical gastrectomy. Eur J Clin Nutr 72(3): 436-445, 2018. PMID: 29184135. DOI: 10.1038/s41430-017-0032-7
    OpenUrlCrossRefPubMed
  10. ↵
    1. Yang SJ,
    2. Li HR,
    3. Zhang WH,
    4. Liu K,
    5. Zhang DY,
    6. Sun LF,
    7. Chen XL,
    8. Zhao LY,
    9. Chen XZ,
    10. Yang K,
    11. Chen ZX,
    12. Zhou ZG and
    13. Hu JK
    : Visceral fat area (VFA) superior to BMI for predicting postoperative complications after radical gastrectomy: a prospective cohort study. J Gastrointest Surg 24(6): 1298-1306, 2020. PMID: 31161593. DOI: 10.1007/s11605-019-04259-0
    OpenUrlCrossRefPubMed
  11. ↵
    1. Clavien PA,
    2. Barkun J,
    3. de Oliveira ML,
    4. Vauthey JN,
    5. Dindo D,
    6. Schulick RD,
    7. de Santibañes E,
    8. Pekolj J,
    9. Slankamenac K,
    10. Bassi C,
    11. Graf R,
    12. Vonlanthen R,
    13. Padbury R,
    14. Cameron JL and
    15. Makuuchi M
    : The Clavien-Dindo classification of surgical complications: five-year experience. Ann Surg 250(2): 187-196, 2009. PMID: 19638912. DOI: 10.1097/SLA.0b013e3181b13ca2
    OpenUrlCrossRefPubMed
  12. ↵
    1. Bassi C,
    2. Dervenis C,
    3. Butturini G,
    4. Fingerhut A,
    5. Yeo C,
    6. Izbicki J,
    7. Neoptolemos J,
    8. Sarr M,
    9. Traverso W,
    10. Buchler M and International Study Group on Pancreatic Fistula Definition
    : Postoperative pancreatic fistula: an international study group (ISGPF) definition. Surgery 138(1): 8-13, 2005. PMID: 16003309. DOI: 10.1016/j.surg.2005.05.001
    OpenUrlCrossRefPubMed
  13. ↵
    1. Kanda Y
    : Investigation of the freely available easy-to-use software ‘EZR’ for medical statistics. Bone Marrow Transplant 48(3): 452-458, 2013. PMID: 23208313. DOI: 10.1038/bmt.2012.244
    OpenUrlCrossRefPubMed
  14. ↵
    1. Tanaka T,
    2. Takayama T,
    3. Matsumoto S,
    4. Wakatsuki K,
    5. Enomoto K,
    6. Migita H and
    7. Nakajima Y
    : Evaluation of postoperative pancreatic fistula after gastrectomy. The Japanese Journal of Gastroenterological Surgery 44(6): 657-664, 2022. DOI: 10.5833/jjgs.44.657
    OpenUrlCrossRef
  15. ↵
    1. Washio M,
    2. Yamashita K,
    3. Niihara M,
    4. Hosoda K and
    5. Hiki N
    : Postoperative pancreatic fistula after gastrectomy for gastric cancer. Ann Gastroenterol Surg 4(6): 618-627, 2020. PMID: 33319151. DOI: 10.1002/ags3.12398
    OpenUrlCrossRefPubMed
  16. ↵
    1. Fu G,
    2. Niu Z,
    3. Zhou Y,
    4. Zhou X,
    5. Wang H and
    6. Su Z
    : [Influence of visceral fat area on laparoscopic radical gastrectomy in patients with gastric carcinoma]. Zhonghua Wei Chang Wai Ke Za Zhi 18(8): 804-807, 2015. PMID: 26303691.
    OpenUrlPubMed
  17. ↵
    1. Shin HJ,
    2. Son SY,
    3. Cui LH,
    4. Byun C,
    5. Hur H,
    6. Lee JH,
    7. Kim YC,
    8. Han SU and
    9. Cho YK
    : Is there any role of visceral fat area for predicting difficulty of laparoscopic gastrectomy for gastric cancer? J Gastric Cancer 15(3): 151-158, 2015. PMID: 26468412. DOI: 10.5230/jgc.2015.15.3.151
    OpenUrlCrossRefPubMed
Back to top

In this issue

Print
Download PDF
Article Alerts
Email Article
Citation Tools
Reprints and Permissions
Dorsal Subcutaneous Fat Thickness as a Risk Factor for Pancreatic Fistula After Gastric Cancer Surgery
HIRONARI MIYAMOTO, MAMI YOSHII, HIROAKI TANAKA, SATOSHI NISHI, SOTA DEGUCHI, YUICHIRO MIKI, TATSURO TAMURA, TAKAHIRO TOYOKAWA, SHIGERU LEE, KIYOSHI MAEDA
Anticancer Research May 2023, 43 (5) 2171-2178; DOI: 10.21873/anticanres.16379
Twitter logo Facebook logo Mendeley logo

Jump to section

Keywords