Temporary Stomas after Rectal Cancer Resection; Predilection of Being Permanent and Predictors of Complications?

Abstract

Stomas are essential for colorectal surgery and are widely used not only for selected cases for bowel obstructions but also in rectal cancer operations to divert stool away from low rectal anastomosis. On the other hand, complications with stomas/ stomas reversal are not uncommon. In this study, we aimed at studying the frequency and the predictors of temporary stomas being permanent, and the contributing factors of surgical stoma/stoma closure related complications. In our cohort, only about 40% of the patient closed their initially planned temporary stomas. The occurrence of intestinal leak, wound sepsis, or any type of morbidity with 30 days of operation were significant predictors of permanent stomas. In addition, alarmingly although Hartmann’s procedure was uncommon in our practice, only 9% of those who underwent Hartmann’s have had it reversed. Moreover, the only factor that significantly increased stoma related complications was having an end colostomy. There was a tendency toward late closure of stomas with median 8.2 months, however early closure did not correlate to complications. In conclusion, further studies are needed to delineate the low rate of stoma closure. Patients who develop postoperative complications, even wound sepsis, would be at a higher risk of living with permanent stomas. Hartmann’s procedures are commonly associated with stoma problems, and reluctance to reverse the stomas.

Keywords:
Rectal cancer; Proctectomy; Hartmann’s procedure; Stoma

Introduction

In rectal cancer surgery, creation of a diverting stoma potentially minimizes the fatal consequences of anastomotic leakage, albeit it may not substantially decrease its incidence.¹ In addition, stoma creation in patients undergoing a low/ultralow anterior resection after neoadjuvant chemo-radiation may be better for the patient’s well-being as in this group of patients the predicted poor bowel function in the immediate postoperative period could be avoided.²

A group of patients who underwent sphincter-saving rectal surgeries with a temporary stoma may consequently end with a permanent stoma status, including non-reversal or re-creationofanotherstoma after the initial stoma closure surgery.² The incidence of this scenario is non-consistent, it was up to 17% in one series³ and up to 23% in another one.⁴

Furthermore, the timing of temporary stoma reversal is not agreed upon. However, some authors suggested that it is safe to close a temporary ileostomy as early as 8 to 13 days after proctectomy and anastomosis for rectal cancer in selected patients without clinical or radiological signs of anastomotic leakage.⁵

There is lack of studies assessing the rate, predictors, and timing of stoma reversal and complications associated with temporary stomas in the Middle East.

Patients and Methods

This is a retrospective study, where the institutional registry at oncology center Mansoura University (OCMU) is thoroughly revised for Rectal cancer cases that attended the hospital from 2006 till August 2020.

All procedures performed in the study involving human participants were in accordance with the ethical standards according to 1964 Helsinki declaration and its later amendments. The study has been approved by Mansoura Faculty of Medicine Institutional Research Board (MFM-IRB) with approval code (R/22.06.1748).

Patients with rectal adenocarcinoma where stoma was constructed were included. While patients were excluded if they have colon cancer, no definite pathologic diagnosis, inadequate registered data, only non-resectional surgery was done (including ostomy) or underwent abdominoperineal resection.

The primary outcome was the rate of reversal of stomas and factors predicting failure of reversal. While the secondary outcomes were complications related to stoma construction and stoma reversal and factors predicting these complications.

The data were analyzed, and statistical values were obtained using SPSS version 22 (Inc, Chicago, IL). Continuous variables are presented as mean when symmetrical or median and range when asymmetrical. Categorical variables are presented as proportions. Univariate analysis was done using Chi-Square test, Fisher’s exact test (if cell count less than 5), Mann-Whitney test, and student t-test. Multivariate analysis was done using binary logistic regression. P value <0.05 was considered significant.

Results

Clinico-epidemiologic and Pathologic Criteria of Patients (►Table 1)

Table 1
Basic clinic-epidemiologic criteria, pathology, operative details, and outcomes of the studied patients

Variable	Value
Age mean +/- SD	48.5 +/- 13.2 years
Sex
Male	61 (43%)
Female	81 (57%)
Rectal site
Low	52 (36.6%)
Mid	52 (36.6%)
High	25 (17.6%)
Pathology
Adenocarcinoma, NOS	103 (72.5%)
Mucinous carcinoma	35 (24.6%)
Signet ring carcinoma	4 (2.8%)
Metastasis
No	122 (85.9%)
Yes	13 (9.2%)
Preoperative therapy
No	47 (33.1%)
Yes	90 (63.4%)
Intestinal Obstruction
No	135 (94.4%)
Yes	8 (5.6%)
Operation
Anterior resection	98 (69%)
Intersphincteric resection	32 (22.5%)
Hartmann’s procedure	11 (7.7%)
Proctocolectomy	1 (0.7%)
Approach
Open	98 (69%)
Laparoscopic	32 (22.5%)
Transanal total mesorectal excision	12 (8.5%)
Anastomosis type*
End to end	115 (87.8%)
Other (side to end, end to side, coloplasty, pouch)	12 (9.2%)
Anastomosis*
Hand-sewn	71 (54.1%)
Stapler	52 (39.7%)
Stoma type
Ileostomy	114 (80.3%)
Colostomy	27 (19%)
Ileocolostomy	1 (0.7%)
Permanent Stoma
Yes	84 (59.2%)
No	58 (40.8%)
Time to stoma closure median (range)	8.5 (1-24) months
Pathologic stage
0	10 (7%)
I	20 (14.1%)
II	45 (31.7%)
III	57 (40.1%)
IV	9 (6.3%)
30-day Morbidity
No	95 (66.9%)
Yes	46 (32.4%)
Wound sepsis**
No	111 (78.2%)
Yes	30 (21.1%)
Intestinal fistula*
No	101 (77.1%)
Yes	14 (10.7%)
Tumour recurrence
No	101 (71.1%)
Yes	33 (23.2%)
Stoma Complications
No	132 (93%)
Retraction	2 (1.4%)
Obstruction	1 (0.7%)
Parastomal hernia	2 (1.4%)
Peristomal collection	1 (0.7%)
Peristomal excoriation	1 (0.7%)
Fistula upon stoma closure	3 (2.1%)

• *Percentage calculated in the 131 patients who underwent primary anastomosis.
• **Wound sepsis defined as infection superficial or deep in the wound.

Out of 370 operated rectal cancer patients at the enrollment time, 215 had sphincter preserving surgery for adenocarcinoma. Of them, 142 patients had stomas constructed and were included in the study (►Figure 1). Mean age 48.5 +/-SD 13.2. Females represent 57%. About a third of the cases (36.6%) were low lying rectal cancer (<5cm from verge). The majority were adenocarcinoma not otherwise specified (NOS) 72.5%, followed by mucinous carcinoma (24.6%). About 2/3 of the patients (63.4%) received neoadjuvant therapy. Only 8 (5.6%) received primary resection in an emergency context secondary to bowel obstruction. 2/3 of the patients (98, 69%) underwent anterior resection (high, low, or ultra-low). Followed by intersphincteric resection (32,22.5%), then Hartmann’s procedure (11, 7.7%). Again 2/3 of patients were operated by open surgery (69%). The anastomosis was configured as end to end in 115/131 who underwent resection anastomosis and was hand-sewn in more than half of them 71/131.

Fig. 1
Flowchart showing included and excluded patients in this study.

13 (9.2%) patients were metastatic, as following: 3 had their metastasis vanished after neoadjuvant therapy, 4 underwent concomitant hepatectomy, 1 was operated in context of bowel obstruction, 2 on palliative basis, 2 had concomitant non-regional node dissection (portocaval & inguinal) and 1 was planned for staged pneumonectomy.

Most of the patients were pathologic stage III (40.1%), followed by stage II (31.7%).

Recurrence occurred in 33 patient (23.2%).

Stoma Construction

All patientshad stomas, 2 of them the stoma was constructed as an emergency before the primary resection surgery, another 2 patients the stoma was exteriorized after surgery (secondary stoma) for leak management, while the rest 138 patients (97.2%) had primary stoma constructedat the time of resection of the tumor. In 80.3% the stoma was ileal, while in the rest it was colonic, with one double barrel ileocolostomy.

Stoma Closure

Only 58 patients (40.8%) of patients are stoma free at the end of follow up, while the rest lived with a permanent stoma either their stoma closure has never been attempted (79 patients, 55.6%) or reversed the stoma and recreated another one (5 patients, 3.5%).

Predictors of Failure of Stoma Closure (►Table 2)

Table 2
Univariate and multivariate analysis of factors predicting permanent stoma.

Variable	Persistent stoma	Stoma reversed	P-value	HR (95% CI)	Significance
Age mean +/- SD	49.3 +/- 14.5	47.4 +/- 11	0.403
Sex			0.86
Male	37	24
Female	47	34
Tumour site			0.37
Low	34	18
Mid	27	25
High	14	13
Pathology			0.16
Adenocarcinoma, NOS	62	41
Mucinous carcinoma	18	17
Signet ring carcinoma	4	0
Preoperative therapy			0.36
No	25	22
Yes	56	34
Obstruction (Emergency)			0.47
No	78	56
Yes	6	2
Operation			0.47
Open	60	38
Minimally invasive	24	20
Type of operation			0.09
Anterior resection	57	41
Intersphincteric	17	15
Hartmann’s procedure	10	1
Proctocolectomy	0	1
Primary anastomosis			.028	8.2 (0.98-68)	0.052
No	10	1
Yes	74	57
Anastomosis configuration			0.76
End to end	66	49
Other	6	6
Anastomosis method			0.36
Hand-sewn	42	29
Stapler	26	26
Stoma type			0.39
Ileostomy	65	49
Colostomy	18	9
Pathologic stage			.023		0.18
0	6	4
I	5	15
II	28	17
III	38	19
IV	6	3
30-day morbidity			.017	2.5 (1.1-5.8)	.025
No	49	46
Yes	34	12
Wound sepsis			.002	6.5 (2.1-20)	.001
No	57	53
Yes	26	5
Pelvic abscess			0.76
No	71	47
Yes	8	4
Intestinal fistula			.021	5.2 (1.3-33.6)	.04
No	59	54
Yes	12	2
Anastomotic stenosis			0.35
No	68	51
Yes	16	7
Tumour Recurrence			0.54
No	57	44
Yes	21	12
Time to oral median (range)	2 (1-17)	2 (0-6)	0.71
Hospital stay median (range)	9 (3-50)	7 (3-45)	.005	0.94 (0.9-1)	0.052

The only predictors of failed stoma closure were Hartmann’s procedure (.028), higher pathologic stage (p-value = .023), 30-day morbidity (p-value = .017), occurrence of postoperative wound sepsis (p-value = .002), overt intestinal fistula (.021), andprolonged hospital stay (.005). Otherwise, neither age, sex, tumor site, stoma site, pelvic abscess, recurrence, anastomosis configuration nor the use of staplers affect the probability of stoma closure.

Running binary logistic regression with primary anastomosis, pathologic stage and either 30-day morbidity, hospital stay, intestinal leak, or wound sepsis, those who maintained significant risk of permanent stoma were morbidity (2.5, p = .025), intestinal leak (5.2 fold, p = .04), and wound sepsis (6.5 fold, p = .001), while Hartmann’s procedure was borderline significant (8.2 fold, p = .052) probably a type II statistical error as for only 11 patients underwent Hartmann’s procedure.

Stoma Related Complications

Only 10 major stomal complications were reported: 2 retraction, 2 parastomal hernias, 1 obstruction, 1 peristomal collection, 1 skin excoriation, and 3 anastomotic leaks with stoma closure. 1 patient died after reversal of stoma from pulmonary embolism.

Predictors of Stoma Related Complications (►Table 3)

Table 3
Analysis of factors contributing to surgical stoma related complications

Variable	Stoma complications		Significance
	No	Yes
Age mean +/-SD	48.2 +/- 13.2	52.4 +/- 13.5	0.34
Sex			0.19
Male	59	2
Female	73	8
Preoperative therapy			0.09
No	41	6
Yes	86	4
Approach			1
Open	91	7
MIS	41	3
Stoma type			0.4
Ileostomy	107	7
Colostomy	24	3
Stoma configuration			.031
Loop	124	7
Terminal	8	3
Obstruction (Emergency)			0.45
No	125	9
Yes	7	1
Timing of stoma			0.054
construction
Before resection	1	1
Primary stoma	129	9
Secondary stoma	2	0
Timing of stoma closure			1
Early	8	0
Late	45	3

The only predictor of stoma complications was stoma configuration where end stomas were a risk factor (p-value = .023). However, stoma type (ileostomy vs. colostomy), age, sex, surgical approach (open vs. minimally invasive) did not affect the complications rate.

Discussion

The use of stomas is an integral part of rectal surgery. The rate of primary diverting stomas varied from 36%⁶ to 54.2%⁷ of low anterior resection patients. In our institution 66% of the patients who underwent sphincter preserving surgery for rectal adenocarcinoma have had a stoma constructed either primary (about 97% of stoma patients) or secondary after a major leak or even in a separate initial surgery to relieve obstruction.

The percentage of patients who closed the stomas in this study was quite low, less than half of the patients, with the majority of the rest never opted to stoma closure. Socioeconomic issues, improper counselling, or failure to refer patients back from their GPs/medical oncologists to reverse stomas may be a factor. Kuryba et al. has mentioned in explanation of his findings that patients from more deprived backgrounds may be less likely to access follow-up services after their initial surgery or that they may be more willing to live with a stoma to avoid further surgery.⁸

In Holmgren et al. multicenter study, partial mesorectal excision (PME) correlated with a stoma-free outcome. While non-reversal was considerably more prevalent among patients with leakage and Stage IV disease. Interestingly stage III patients at first had a decreased reversal rate, which increased after the initial year of surgery.⁹ In contrast, in this series, the site of tumor(high, mid, low) and a ssuch PME or TME resection was not a predictor of stoma free. Moreover, although late stages were significantly associated with more permanent stomas in univariate analysis, this was not significant in multivariate analysis. However, major intestinal leak, wound sepsis and overall, 30-day morbidity were the only significant predictors of failure of stoma closure.

In a study of NHS patients, within 1.5 years from anterior resection, 72.5% of the patients had undergone an ileostomy reversal. The reversal rate was lower in the following circumstances: older patients, male gender, higher American Society of Anesthesiologists (ASA) grade, more advanced cancer, socioeconomic deprivation, comorbidity, and open surgical procedure.⁸ In our data, neither age, sex, stage, or surgical approach (open VS. laparoscopic) significantly predicts stoma reversal.

Whilst 85% of patients with primary anastomosis (PA) have had their stoma reversed, only 58% of patients with an Hartmann’s procedure (HP) had a stoma reversal in one series.¹⁰ In this study, we have a quite small number of HP (11 patients), this can be explained by the tendency to manage patients with malignant bowel obstruction secondary to cancer by proximal diversion (loop colostomy) followed by neoadjuvant therapy and then resection on an elective basis. Albeit that, the relation between stoma closure and primary anastomosis was clearer where 57/131 (43.5%) of the patients with PA successfully reversed their stomas, while only 1/11 (9.1%) of HP patients did.

In one study, the postoperative stoma-related complications developed in 17.8% of patients. Of them stoma outlet obstructions occurred in 7.0%. Thick subcutaneous fat at the stoma-marking site (vertical distance ≥ 20 mm) was the only predictor of stoma obstruction.¹¹ In our series non-medical complications related to stoma creation/closure occurred in 7% only, with stoma obstruction affecting only one patient.

As regard stoma closure complications, 9% of the patients who underwent reversal of a stoma in one series experienced major complications requiring a return to theatre, need for intensive care or mortality.⁹ In our study, only 3 patients out of 63 who attempted stoma closure (4.8%) developed major complications post stoma closure, 1 succumbed of pulmonary embolism and another 2 developed intestinal leak necessitating the creation of another stoma.

It is known that a diverting loop colostomy is associated with ostomy prolapse and parastomal hernia, while a diverting loop ileostomy is in particularly associated with the risk for high-output ostomy.¹² In a metanalysis, age ≥65 years, body mass index ≥30kg/m2, diabetes mellitus, hypertension, renal comorbidity, regular diuretic use, ileal pouch-anal anastomosis procedure and length of stay after index admission were associated with dehydration readmission, while a diagnosis of colorectal cancer was less likely to result in readmission.¹³ In our cohort terminal colostomies were the only statistically significant predictor for stoma related complications, however; none of our patients were admitted with dehydration following ileostomy construction. This is important to highlight, although this may mean that they are just managed on a primary care level, it may truly reflect what we notice in practice that readmission for stoma related dehydration, which is common in western patients, is not that feared in our patients probably due to younger age (mean 48), being all treated for colorectal cancer and that only 1 patient had ileal pouch-anal anastomosis. In addition, the stoma type (ileostomy VS. colostomy), age, sex, surgical approach (open VS. minimally invasive) did not affect the stoma complications rate.

In one study stoma closure before 109 days of construction increased risk of complications, which were mainly prolonged ileus.¹⁴ A meta-analysis suggested that early closure (<14 days) of a defunctioning loop ileostomy is effective and safe in carefully selected patients without increasing overall postoperative complications,¹⁵ while in a later systematic review the results were slightly inclined toward early closure of loop ileostomy.¹⁶ In this series, there was tendency to late closure (>3 months) where 82.8% patients closed late. The median time to closure was 8.5 months, and we could not correlate stoma complication to timing of closure, however; paralytic ileus was not recognized.

This study was limited by being retrospective, as such some data as socioeconomic status and BMI, ASA of patients are not adequately assessed. In addition, complications related to stoma that was not escalated to our service by peripheral health care facilities, this would mainly include minor complications, were probably missed.

Conclusion

More than half of rectal cancer patients with intended temporary stomas will end up with a permanent one. 30-day morbidity, wound sepsis and intestinal leak and may be Hartmann’s procedure are predictors of permanency. End colostomies are more liable to stoma related complications. Finally, the optimum timing for stoma closure could not be identified as timing of closure does affect outcomes.

References

• ¹
  Tan WS, Tang CL, Shi L, Eu KW. Meta-analysis of defunctioning stomas in low anterior resection for rectal cancer. Br J Surg 2009; 96(05):462–472
• ²
  Huh JW. Stoma Creation After Surgery for Rectal Cancer: Temporary or Permanent? Ann Coloproctol 2015;31(03):82–82
• ³
  Kim YA, Lee GJ, Park SW, Lee WS, Baek JH. Multivariate Analysis of Risk Factors Associated With the Nonreversal Ileostomy Following Sphincter-Preserving Surgery for Rectal Cancer. Ann Colo-proctol 2015;31(03):98–102
• ⁴
  Lim SW, Kim HJ, Kim CH, Huh JW, Kim YJ, Kim HR. Risk factors for permanent stoma after low anterior resection for rectal cancer. Langenbecks Arch Surg 2013;398(02):259–264
• ⁵
  Danielsen AK, Park J, Jansen JE, et al. Early closure of a temporary ileostomy in patients with rectal cancer. Ann Surg 2017;265(02): 284–290
• ⁶
  Shiomi A, Ito M, Saito N, etal. The indications for a diverting stoma in low anterior resection for rectal cancer: a prospective multi-centre study of 222 patients from Japanese cancer centers. Colorectal Dis 2011;13(12):1384–1389
• ⁷
  Shimizu H, Yamaguchi S, Ishii T, et al. Who needs diverting ileostomy following laparoscopic low anterior resection in rectal cancer patients? Analysis of 417 patients in a single institute. Surg Endosc 2020;34(02):839–846
• ⁸
  Kuryba AJ, Scott NA, Hill J, van der Meulen JH, Walker K. Determinants of stoma reversal in rectal cancer patients who had an anterior resection between 2009 and 2012 in the English National Health Service. Colorectal Dis 2016;18(06):O199–O205
• ⁹
  Holmgren K, Kverneng Hultberg D, Haapamäki MM, Matthiessen P, Rutegård J, Rutegård M. High stoma prevalence and stoma reversal complications following anterior resection for rectal cancer: a population-based multicentre study. Colorectal Dis 2017;19(12):1067–1075
• ¹⁰
  Alizai PH, Schulze-Hagen M, Klink CD, et al. Primary anastomosis with a defunctioning stoma versus Hartmann’s procedure for perforated diverticulitis–a comparison of stoma reversal rates. Int J Colorectal Dis 2013;28(12):1681–1688
• ¹¹
  Tamura K, Matsuda K, Yokoyama S, et al. Defunctioning loop ileostomy for rectal anastomoses: predictors of stoma outlet obstruction. Int J Colorectal Dis 2019;34(06):1141–1145
• ¹²
  Bertelsen CA, Andreasen AH, Jørgensen T, Harling HDanish Colorectal Cancer Group. Anastomotic leakage after anterior resection for rectal cancer: risk factors. Colorectal Dis 2010;12 (01):37–43
• ¹³
  Liu C, Bhat S, Sharma P, Yuan L, O’Grady G, Bissett I. Risk factors for readmission with dehydration after ileostomy formation: A systematic review and meta-analysis. Colorectal Dis 2021;23(05): 1071–1082
• ¹⁴
  Yin T-C, Tsai HL, Yang PF, et al. Early closure of defunctioning stoma increases complications related to stoma closure after concurrent chemoradiotherapy and low anterior resection in patients with rectal cancer. World J Surg Oncol 2017; 15(01):80
• ¹⁵
  Menahem B, Lubrano J, Vallois A, Alves A. Early closure of defunctioning loop ileostomy: is it beneficial for the patient? A meta-analysis. World J Surg 2018;42(10):3171–3178
• ¹⁶
  Aljorfi AA, Alkhamis AH. A systematic review of early versus late closure of loop ileostomy. Surg Res Pract 2020;2020:9876527

Publication Dates

• Publication in this collection
  23 Oct 2023
• Date of issue
  2023

History

• Received
  07 Apr 2023
• Accepted
  21 June 2023