Tech Coloproctol DOI 10.1007/s10151-013-1085-9
ORIGINAL ARTICLE
Outcome of colonic fistula surgery in the modern surgical era M. S. Tam • M. Abbass • A. T. Tsay M. A. Abbas
•
Received: 18 July 2013 / Accepted: 14 October 2013 Ó Springer-Verlag Italia 2013
Abstract Background Various conditions lead to the development of colonic fistulas. Contemporary surgical data is scarce and it is unclear whether advances in surgical care have impacted outcome. The aim of the present study was to review the short- and long-term outcome of patients treated surgically for colonic fistula over an 8-year period at a tertiary institution. Methods A retrospective review was performed, focusing on the type of operative interventions, short- and long-term complications, length of hospital stay, readmission rate, mortality rate, and fistula recurrence. Results Forty-five patients were treated for colonic fistula. The most common etiology was diverticulitis (74 %). Fistula type was colovesical (58 %), colocutaneous (18 %) and colovaginal (15 %). Laparoscopic resection was performed in 42 % of cases. An intraoperative complication occurred in 4 %. A primary anastomosis was performed in 96 % of patients and 10 (23 %) had a temporary stoma. Median length of hospital stay was 6 days. Postoperative complications were common (47 %) and wound infection was noted in 20 % of patients. The readmission rate was This manuscript was a poster presentation at the 2013 American Society of colon and Rectal Surgeons Annual meeting in Phoenix, Arizona April 27th–May 1st. M. S. Tam M. Abbass A. T. Tsay M. A. Abbas Department of Surgery, Kaiser Permanente, 4760 Sunset Boulevard, 3rd Floor, Los Angeles, CA 90027, USA e-mail: [email protected] M. A. Abbas (&) Digestive Disease Institute, Cleveland Clinic Abu Dhabi, Al Maqam Tower, Sowwah Square, Al Maryah Island, P.O. BOX 112412, Abu Dhabi, UAE e-mail: [email protected]
29 % and the 90-day mortality was 4 %. All patients healed their fistula with no recurrences noted during a median follow-up of 37 months. Conclusions Surgical intervention healed the majority of patients with colonic fistula. However postoperative complications were common and readmission occurred in onethird of the cases. Laparoscopic excision was feasible in nearly half of the patients. Keywords Colonic fistula Colovesical Colovaginal Colocutaneous Surgical outcome
Introduction Colonic fistula is an uncommon condition but its impact on the patient’s quality of life can be significant [1]. Sigmoid diverticulitis is the most common etiology of colonic fistulas. Several types have been reported including colovesical, colovaginal, colocutaneous, or a combination of various types [1–9]. Surgical resection is the established gold standard for definitive eradication of colonic fistulas [1–4, 6–21]. While numerous studies have been published on this topic, it is unclear whether the morbidity and outcome of surgical intervention have changed with the recent advances in surgical care and the increasing use of minimally invasive techniques. The aim of this study was to review the short and long-term outcome of patients operated for colonic fistula.
Materials and methods The study was approved by the Institutional Review Board of Kaiser Permanente Southern California. A retrospective
123
Tech Coloproctol
review was conducted of all patients who underwent operative intervention for colonic fistula over an 8-year period (October 2004–December 2012). All procedures were performed by one colorectal surgeon (MAA) at Kaiser Permanente Los Angeles Medical Center, a regional tertiary institution in Southern California that serves a population of approximately 3.4 million patients. All patients were prepared with polyethylene glycol-electrolyte solution the day prior to their operation and received perioperative intravenous antibiotics within 1 h of incision (ampicillin/sulbactam, piperacillin/tazobactam, or a combination of cefazolin and metronidazole). Patients with chronic pelvic abscess received postoperative antibiotics for 10–14 days. Deep venous thrombosis prophylaxis was given preoperatively and continued postoperatively with subcutaneous heparin. The inpatient and outpatient electronic healthcare records were reviewed. Data collected included patient-related preoperative factors such as demographics, body mass index (BMI) in kg/m2, American Society of Anesthesiologists (ASA) classification, prior surgical history, malnutrition (albumin\3.3 gm/dL), anemia (hematocrit level: male\42, female \37), smoking, diabetes, etiology and type of the fistula. Intraoperative data included operation type, ureteral catheter usage, stoma, operative time, blood loss and need for transfusion, and intraoperative complications. Postoperative outcome measures included length of hospital stay, intensive care unit admission, overall complications, need for postoperative transfusion, need for readmission, healing of fistula, and 90-day mortality.
Results Forty-five patients were surgically treated for colonic fistula during the study period. Table 1 summarizes the patients’ characteristics. There were 25 females (56 %) and the median age was 59 years (range 25–82 years). The median BMI was 29 kg/m2 (range 18.5–47.1 kg/m2). A prior abdominopelvic operation was performed in 64 % of the patients. However only 3 patients (7 %) had a prior operation for the fistula. We defined the fistulas of these 3 patients as persistent fistula as they had never healed following the prior repair. The remaining 57 % had abdominopelvic operations unrelated to the treatment of the fistula (such as hysterectomy or bowel resection). A prior hysterectomy was noted in all patients with colovaginal fistula. In general, for the minority of patients who had undergone a prior attempt at repair, a minimum of 6 months elapsed prior to considering additional surgical intervention. Nineteen patients (42 %) presented with a chronic abdominopelvic abscess. The most common etiology of the
123
Table 1 Characteristics of patients with colonic fistula N = 45 Median age, years (mean, range)
59 (59, 25–82)
Gender M/F
20/25 (44 %/56 %)
BMI (kg/m2) (mean, range)
29 (29, 18.5–47.1)
ASA I
1 (2 %)
II
22 (49 %)
III
22 (49 %)
Prior abdominopelvic operation
29 (64 %)
Prior operation for fistula
3 (7 %)
Anemia
35 (78 %)
Chronic abdominopelvic abscess Malnutrition
19 (42 %) 17 (38 %)
Diabetes
14 (31 %)
Smoking
7 (16 %)
Etiology Diverticulitis
33 (74 %)
Postoperative
5 (11 %)
Post radiation
4 (9 %)
Malignant
2 (4 %)
Inflammatory bowel disease
1 (2 %)
Fistula type Colovesical
26 (58 %)
Colocutaneous
8 (18 %)
Colovaginal
7 (15 %)
Colovaginal and colovesical
4 (9 %)
BMI body mass index, ASA American Society of Anesthesiologists
fistula was chronic diverticulitis (74 %). Fistula type was colovesical (58 %), colocutaneous (18 %), colovaginal (15 %), and combined colovaginal/colovesical (9 %). The presence of fistula was confirmed preoperatively in all patients. Several diagnostic modalities were used including computed tomography, gastrografin enema and contrast vaginogram, tandem vaginoscopy with colonoscopy, cystoscopy, and cystogram. Table 2 highlights the operative findings in the 45 patients. The most common operations were anterior resection (49 %) and sigmoidectomy (38 %). Laparoscopic surgery was performed in 19 patients (42 %). The majority of patients (96 %) had a primary anastomosis. Ureteral catheters were placed in 58 % of the patients. Indication for ureteral catheter placement included postoperative or radiationrelated fistula, prior pelvic surgery to repair the fistula, diverticulitis with inflammation involving the retroperitoneum, or the presence of hydronephrosis. In patients with a colovesical fistula, the bladder was decompressed with a urethral catheter for 10 days postoperatively. The urinary catheter was removed after cystographic confirmation of healing. The fistulous opening in the bladder was not
Tech Coloproctol Table 2 Operative findings of patients with colonic fistula
Table 3 Postoperative outcome of patients with colonic fistula
N = 45 Operation type Anterior resection Sigmoidectomy Ileocolic resection Pelvic exenteration Hemicolectomy Revision of colostomy Laparoscopic resection Ureteral catheters Splenic flexure mobilization Stoma Temporary ileostomy Permanent colostomy Median operative time, minutes (mean, range) Median blood loss, milliliters (mean, range) Intraoperative transfusion Intraoperative complications Ureteral laceration Bladder laceration
22 (49 %) 17 (38 %) 2 (4 %) 2 (4 %) 1 (2 %) 1 (2 %) 19 (42 %) 26 (58 %) 15 (33 %) 12 (27 %) 10 (23 %) 2 (4 %) 263 (292, 63–725) 300 (468, 50–3,600) 9 (20 %) 2 (4 %) 1 (2 %) 1 (2 %)
routinely closed. Formal bladder repair by the colorectal surgeon was performed for large full thickness defects where the bladder lumen could be visualized. The splenic flexure was mobilized in one-third of the patients. A temporary ileostomy was established in 10 patients (23 %) and all were subsequently reversed. Fecal diversion was performed selectively. Indications included prior failed repair, radiated patients, low anastomosis (i.e. proctectomy), or any intraoperative concerns about the integrity of the anastomosis. The median operative time was 263 min (range 63–725 min) with a median blood loss of 300 ml (range 50–3,600 ml). Two patients (4 %) suffered an intraoperative complication involving laceration of the bladder and the left ureter and both were recognized and repaired intraoperatively. The median length of stay was 6 days (range 2–65 days) and 18 % of the patients required intensive care. The overall complication rate was 47 % (Table 3). Wound infection was the most common complication (20 %). No anastomotic leak was noted but 1 patient (2 %) developed a postoperative abdominopelvic abscess. The readmission rate was 29 %. A small bowel obstruction was the most common reason for readmission (9 %). Two patients (4 %) required an operation at readmission. All patients healed their fistula and during a median follow-up of 37 months (range 8–102 months), no recurrence was noted. Ninety-day mortality was 4 % with 2 patients succumbing to acute renal failure with cardiopulmonary failure.
N = 45 Median length of stay, days (mean, range)
6 (9, 2–65)
Intensive care hospitalization
8 (18 %)
Overall complicationsa
21 (47 %)
Wound infection/dehiscence
12 (27 %)
Sepsis, urinary tract infections
4 (9 %)
Small bowel obstruction
4 (9 %)
Deep venous thrombosis, thrombophlebitis
3 (7 %)
Postoperative bleeding
2 (4 %)
Abdominopelvic abscess
1 (2 %)
Hydronephrosis
1(2 %)
Vocal cord paralysis Trocar site hernia
1 (2 %) 1 (2 %)
Postoperative transfusion
7 (16 %)
Readmission
13 (29 %)
Small bowel obstruction
4 (9 %)
Abdominal pain
2 (4 %)
Rectal bleeding
2 (4 %)
Hydronephrosis
1 (2 %)
Abdominopelvic abscess
1 (2 %)
Wound dehiscence
1 (2 %)
Deep venous thrombosis
1 (2 %)
Urinary tract infection
1 (2 %)
Trocar site hernia Reoperation during readmission Hydronephrosis Small bowel obstruction from trocar site hernia Recurrence of fistula 90-day mortality a
1 (2 %) 2 (4 %) 1 (2 %) 1 (2 %) 0 2 (4 %)
Some patients with more than one complication
Discussion Colonic fistulas are relatively uncommon and are most often the result of complicated diverticulitis. Surgical intervention remains the treatment modality of choice in patients with fistulas refractory to medical therapy. Resection and/or fecal diversion are the most commonly performed procedures for this condition. This study aimed at determining the outcome of patients treated surgically for this condition in recent years. Current surgical care includes standardization of perioperative antimicrobial regimen administration, routine use of chemical prophylaxis for deep venous thromboembolism, and increased use of the laparoscopic technique. The majority of fistulas were related to chronic diverticulitis and a significant number of patients in this study presented with a chronic abdominopelvic abscess. Diverticulitis is the most common cause of colonic fistulas with published rates ranging from 62 to 75 % of all fistulas [2–8]. Seventy-four percent of the
123
123
Albert Schweitzer Hospital, The Netherlands
Kaiser Permanente, California
Smeenk et al. [13]
Current study
b
Range Mixed colovaginal and colovesical fistulas
Urologishe Klinik und Poklinik, Germany
a
Mount Sinai, New York
Leicht et al. [11]
The Leister General Hospital, UK
Garcea et al. [4]
Lynn et al. [2]
Darent Valley Hospital, UK
Solkar et al. [21]
Barnet and Chase Farm NHS Trust, UK
Virginia Commonwealth University, Virginia
Najjar et al. [5]
Holoryd et al. [3]
Hospital Cantonal De Fribourg, Switzerland
Menenakos et al. [17]
Dandery Hospital, Sweden
Glasgow Royal Infirmary, UK
Walker et al. [7]
Johannes Guterberg University, Germany
Aberdeen Royal Hospital, UK
Driver et al. [8]
Melchior et al. [10]
Cleveland Clinic, Ohio
Woods et al. [1]
Hjern et al. [14]
Institution
Author
Table 4 Summary of recent studies on colonic fistulas
2013
2012
2012
2012
2012
2009
2007
2006
2005
2004
2003
2002
1997
1988
Year
45
40
54
72
37
49
60
72
50
12
18
19
67
84
N
58
88
100
100
86
100
100
100
100
100
83
100
100
65
Colovesical (%)
Fistula type
15
12
–
–
14
–
–
–
–
–
17
–
–
28
Colovaginal (%)
18
–
–
–
–
–
–
–
–
–
–
–
–
–
Colocutaneous (%)
48 48
9b
11
–
22
8
28
42
35
–
27
84
10
27
–
–
–
–
–
–
–
–
–
–
–
–
7
Coloenteric (%)
Morbidity (%)
4%
8%
0%
10 %
17 %
0%
0%
7%
0
1%
0
1%
10 %
3.5 %
Mortality
0
–
0%
11 %
5%
0%
–
–
2%
0%
5.5 %
0%
14 %
0%
Recurrence
40
–
62
31
23
68
12
–
–
2–108a
26
35
–
–
Mean follow-up, months
Tech Coloproctol
Tech Coloproctol
fistulas in our series were attributable to diverticulitis. The remaining fistulas were postoperative, radiation-induced, malignant, or related to Crohn’s disease. The bladder was involved in over half of the patients. Colovesical fistula is the most common type of colonic fistula and has been reported in 50–100 % of published series on this condition [1–21]. Anterior resection was the most frequently performed operation in our study and nearly half of the patients underwent a laparoscopic resection. Postoperative complications were common and often infectious in nature but the mortality rate was low. Postoperative complications led to a readmission in about one-third of the patients in this study. Several recent studies have reported morbidity rates of 8–48 % and mortality rates of 0–18 % [3, 10, 11, 13]. In our study a primary anastomosis was performed in 96 % of the patients and no anastomotic leak was noted. In one of the largest series published to date on colonic fistula, the Cleveland Clinic reported a 5 % leak rate in 84 patients undergoing resection and primary anastomosis [1]. However resolution of the fistula was noted in all of the patients with no recurrences, a finding similar to our results. There were some limitations to this study. Because the incidence of colonic fistula is low in the general population, our sample size of 45 patients over an 8-year period was relatively small. Furthermore, this study was conducted at a tertiary institution with a large population of complex and/or elderly patients with significant co-morbidities which may have skewed the morbidity and mortality findings. Lastly, the study was retrospective in nature which inherently may introduce bias in terms of patient selection and follow-up. Despite the acknowledged shortcomings, this study provided valuable insight into the short- and long-term outcome of the surgical management of colonic fistulas in the modern surgical era. There is a paucity of data from the United States regarding the surgical outcome of patients treated for colonic fistulas with the majority of existing literature originating from Europe (Table 4) [1–5, 7, 8, 10– 14, 17, 21]. Furthermore, most of the available data in the literature describes the outcome of patients with colovesical fistulas. The largest colovesical fistula series to date comes from Lynn et al. [2] from Mount Sinai Hospital in New York. Seventy-two patients underwent operative repair for colovesical fistula with a recurrence rate of 11 %. No difference in recurrence rate was noted in that study between patients who underwent a single stage procedure versus a multi-stage repair. Our study included all fistula types including colovesical, colocutaneous, and colovaginal. During a mean follow-up of 40 months, no recurrence was noted. The reported recurrence rate in the literature ranges from 0 to 14 % [1–5, 7, 8, 10–14, 17, 21]. Recently, laparoscopic colectomy has gained popularity for elective resection but has not been universally accepted
for complicated diverticular disease including colonic fistulas. A laparoscopic resection was performed in nearly half of our patients with good success and no conversion. The lack of conversion in this study maybe attributed to both the advanced laparoscopic skills of the lead surgeon and to proper selection of patients. Hence a selection bias (i.e. choosing the cases deemed feasible with laparoscopy) could explain the lack of conversion. Menenakos et al. [17] studied 18 patients who underwent laparoscopic resection for colonic fistulas and found only 1 recurrence (5.5 %). Many subsequent studies have supported the laparoscopic approach for diverticular fistula based on low rates of postoperative morbidity and low conversion rates, though both of these outcome measures are higher in patients with diverticular fistulas than in patients with non-complicated diverticulitis [18–20].
Conclusions This retrospective study from a tertiary colorectal surgery unit demonstrated that colonic fistula can be successfully eradicated with surgical intervention in the majority of patients. The most common fistula type was colovesical and was often secondary to chronic diverticulitis. The 90-day mortality rate was low following surgical intervention. However postoperative complications occurred in approximately half of the patients with one-third requiring readmission. Despite recent advances in surgical care such as prophylactic use of intravenous antibiotics to prevent wound infection, subcutaneous heparin for venous thrombosis prophylaxis, and an increasing use of laparoscopic surgery, the morbidity associated with operative intervention for colonic fistula remains significant. Conflict of interest
None.
References 1. Woods RJ, Lavery IC, Fazio VW, Jahgelman DG, Weakley FL (1988) Internal fistulas in diverticular disease. Dis Colon Rectum 31:591–596 2. Lynn ET, Ranasinghe NE, Dallas KB, Divino CM (2012) Management and outcomes of colovesical fistula repair. Am Surg 78:514–518 3. Holroyd DJ, Banerjee S, Beavan M, Prentice R, Vijay V, Warren SJ (2012) Colovaginal and colovesical fistulae: the diagnostic paradigm. Tech Coloproctol 16:119–126 4. Garcea G, Majid I, Sutton CD, Pattenden CJ, Thomas WM (2006) Diagnosis and management of colovesical fistulae; six-year experience of 90 consecutive cases. Colorectal Dis 8:347–352 5. Najjar SF, Jamal MK, Savas JF, Miller TA (2004) The spectrum of colovesical fistula and diagnostic paradigm. Am J Surg 188:617–621 6. Pollard SG, Macfarlane R, Greatorex R, Everett WG, Hartfall WG (1987) Colovesical fistula. Ann R Coll Surg Engl 69:163–165
123
Tech Coloproctol 7. Walker KG, Anderson JH, Iskander N, McKee RF, Finlay IG (2002) Colonic resection for colovesical fistula: 5-year follow-up. Colorectal Dis 4:270–274 8. Driver CP, Anderson DN, Findlay K, Keenan RA, Davidson AI (1997) Vesico-colic fistulae in the Grampain region: presentation, assessment, management and outcome. J R Coll Surg Edinb 42:182–185 9. Wassef R, Morgan S, Tasse´ D, Bernard D (1983) Fistulas in diverticular disease of the colon: study of 29 cases. Can J Surg 26:546–549 10. Melchior S, Cudovic D, Jones J, Thomas C, Gillitzer R, Thu¨roff J (2009) Diagnosis and surgical management of colovesical fistulas due to sigmoid diverticulitis. J Urol 182:978–982 11. Leicht W, Thomas C, Thu¨roff J, Roos F (2012) Colovesical fistula caused by diverticulitis of the sigmoid colon: diagnosis and treatment. Urology 51:971–974 (German) 12. King RM, Beart RW, Mcllrath DC (1982) Colovesical and rectovesical fistulas. Arch Surg 117:680–683 13. Smeenk RM, Plaisier PW, van der Hoeven JA, Hesp WL (2012) Outcome of surgery for colovesical and colovaginal fistulas of diverticular origin in 40 patients. J Gastrointest Surg 16:1559– 1565
123
14. Hjern F, Goldberg SM, Johansson C, Parker SC, Mellgren A (2007) Management of diverticular fistulae to the female genital tract. Colorectal Dis 9:438–442 15. Bahadursingh AM, Longo WE (2003) Colovaginal fistulas. Etiology and management. J Reprod Med 48:489–495 16. Greenstein AJ, Sachar DB, Tzakis A (1984) Course of enterovesical fistulas in Crohn’s disease. Am J Surg 17:788–792 17. Menenakos E, Hahnloser D, Nassiopoulos K, Chanson C, Sinclair V, Petropoulos P (2003) Laparoscopic surgery for fistulas that complicate diverticular disease. Langenbecks Arch Surg 388:189–193 18. Bartus CM, Lipof T, Sarwar CM et al (2005) Colovesical fistula: not a contraindication to elective laparoscopic colectomy. Dis Colon Rectum 48:233–236 19. Zapletal C, Woeste G, Bechstein WO, Wullstein C (2007) Laparoscopic sigmoid resections for diverticulitis complicated by abscesses or fistulas. Int J Colorectal Dis 22:1515–1521 20. Royds J, O’Riordan JM, Eguare E, O’Riordan D, Neary PC (2012) Laparoscopic surgery for complicated diverticular disease: a single-center experience. Colorectal Dis 14:1248–1254 21. Solkar MH, Forshaw MJ, Sankararajah D, Stweart M, Parker MC (2005) Colovesical fistula—is a surgical approach always justified? Colorectal Dis 7:467–471
ORIGINAL ARTICLE
Outcome of colonic fistula surgery in the modern surgical era M. S. Tam • M. Abbass • A. T. Tsay M. A. Abbas
•
Received: 18 July 2013 / Accepted: 14 October 2013 Ó Springer-Verlag Italia 2013
Abstract Background Various conditions lead to the development of colonic fistulas. Contemporary surgical data is scarce and it is unclear whether advances in surgical care have impacted outcome. The aim of the present study was to review the short- and long-term outcome of patients treated surgically for colonic fistula over an 8-year period at a tertiary institution. Methods A retrospective review was performed, focusing on the type of operative interventions, short- and long-term complications, length of hospital stay, readmission rate, mortality rate, and fistula recurrence. Results Forty-five patients were treated for colonic fistula. The most common etiology was diverticulitis (74 %). Fistula type was colovesical (58 %), colocutaneous (18 %) and colovaginal (15 %). Laparoscopic resection was performed in 42 % of cases. An intraoperative complication occurred in 4 %. A primary anastomosis was performed in 96 % of patients and 10 (23 %) had a temporary stoma. Median length of hospital stay was 6 days. Postoperative complications were common (47 %) and wound infection was noted in 20 % of patients. The readmission rate was This manuscript was a poster presentation at the 2013 American Society of colon and Rectal Surgeons Annual meeting in Phoenix, Arizona April 27th–May 1st. M. S. Tam M. Abbass A. T. Tsay M. A. Abbas Department of Surgery, Kaiser Permanente, 4760 Sunset Boulevard, 3rd Floor, Los Angeles, CA 90027, USA e-mail: [email protected] M. A. Abbas (&) Digestive Disease Institute, Cleveland Clinic Abu Dhabi, Al Maqam Tower, Sowwah Square, Al Maryah Island, P.O. BOX 112412, Abu Dhabi, UAE e-mail: [email protected]
29 % and the 90-day mortality was 4 %. All patients healed their fistula with no recurrences noted during a median follow-up of 37 months. Conclusions Surgical intervention healed the majority of patients with colonic fistula. However postoperative complications were common and readmission occurred in onethird of the cases. Laparoscopic excision was feasible in nearly half of the patients. Keywords Colonic fistula Colovesical Colovaginal Colocutaneous Surgical outcome
Introduction Colonic fistula is an uncommon condition but its impact on the patient’s quality of life can be significant [1]. Sigmoid diverticulitis is the most common etiology of colonic fistulas. Several types have been reported including colovesical, colovaginal, colocutaneous, or a combination of various types [1–9]. Surgical resection is the established gold standard for definitive eradication of colonic fistulas [1–4, 6–21]. While numerous studies have been published on this topic, it is unclear whether the morbidity and outcome of surgical intervention have changed with the recent advances in surgical care and the increasing use of minimally invasive techniques. The aim of this study was to review the short and long-term outcome of patients operated for colonic fistula.
Materials and methods The study was approved by the Institutional Review Board of Kaiser Permanente Southern California. A retrospective
123
Tech Coloproctol
review was conducted of all patients who underwent operative intervention for colonic fistula over an 8-year period (October 2004–December 2012). All procedures were performed by one colorectal surgeon (MAA) at Kaiser Permanente Los Angeles Medical Center, a regional tertiary institution in Southern California that serves a population of approximately 3.4 million patients. All patients were prepared with polyethylene glycol-electrolyte solution the day prior to their operation and received perioperative intravenous antibiotics within 1 h of incision (ampicillin/sulbactam, piperacillin/tazobactam, or a combination of cefazolin and metronidazole). Patients with chronic pelvic abscess received postoperative antibiotics for 10–14 days. Deep venous thrombosis prophylaxis was given preoperatively and continued postoperatively with subcutaneous heparin. The inpatient and outpatient electronic healthcare records were reviewed. Data collected included patient-related preoperative factors such as demographics, body mass index (BMI) in kg/m2, American Society of Anesthesiologists (ASA) classification, prior surgical history, malnutrition (albumin\3.3 gm/dL), anemia (hematocrit level: male\42, female \37), smoking, diabetes, etiology and type of the fistula. Intraoperative data included operation type, ureteral catheter usage, stoma, operative time, blood loss and need for transfusion, and intraoperative complications. Postoperative outcome measures included length of hospital stay, intensive care unit admission, overall complications, need for postoperative transfusion, need for readmission, healing of fistula, and 90-day mortality.
Results Forty-five patients were surgically treated for colonic fistula during the study period. Table 1 summarizes the patients’ characteristics. There were 25 females (56 %) and the median age was 59 years (range 25–82 years). The median BMI was 29 kg/m2 (range 18.5–47.1 kg/m2). A prior abdominopelvic operation was performed in 64 % of the patients. However only 3 patients (7 %) had a prior operation for the fistula. We defined the fistulas of these 3 patients as persistent fistula as they had never healed following the prior repair. The remaining 57 % had abdominopelvic operations unrelated to the treatment of the fistula (such as hysterectomy or bowel resection). A prior hysterectomy was noted in all patients with colovaginal fistula. In general, for the minority of patients who had undergone a prior attempt at repair, a minimum of 6 months elapsed prior to considering additional surgical intervention. Nineteen patients (42 %) presented with a chronic abdominopelvic abscess. The most common etiology of the
123
Table 1 Characteristics of patients with colonic fistula N = 45 Median age, years (mean, range)
59 (59, 25–82)
Gender M/F
20/25 (44 %/56 %)
BMI (kg/m2) (mean, range)
29 (29, 18.5–47.1)
ASA I
1 (2 %)
II
22 (49 %)
III
22 (49 %)
Prior abdominopelvic operation
29 (64 %)
Prior operation for fistula
3 (7 %)
Anemia
35 (78 %)
Chronic abdominopelvic abscess Malnutrition
19 (42 %) 17 (38 %)
Diabetes
14 (31 %)
Smoking
7 (16 %)
Etiology Diverticulitis
33 (74 %)
Postoperative
5 (11 %)
Post radiation
4 (9 %)
Malignant
2 (4 %)
Inflammatory bowel disease
1 (2 %)
Fistula type Colovesical
26 (58 %)
Colocutaneous
8 (18 %)
Colovaginal
7 (15 %)
Colovaginal and colovesical
4 (9 %)
BMI body mass index, ASA American Society of Anesthesiologists
fistula was chronic diverticulitis (74 %). Fistula type was colovesical (58 %), colocutaneous (18 %), colovaginal (15 %), and combined colovaginal/colovesical (9 %). The presence of fistula was confirmed preoperatively in all patients. Several diagnostic modalities were used including computed tomography, gastrografin enema and contrast vaginogram, tandem vaginoscopy with colonoscopy, cystoscopy, and cystogram. Table 2 highlights the operative findings in the 45 patients. The most common operations were anterior resection (49 %) and sigmoidectomy (38 %). Laparoscopic surgery was performed in 19 patients (42 %). The majority of patients (96 %) had a primary anastomosis. Ureteral catheters were placed in 58 % of the patients. Indication for ureteral catheter placement included postoperative or radiationrelated fistula, prior pelvic surgery to repair the fistula, diverticulitis with inflammation involving the retroperitoneum, or the presence of hydronephrosis. In patients with a colovesical fistula, the bladder was decompressed with a urethral catheter for 10 days postoperatively. The urinary catheter was removed after cystographic confirmation of healing. The fistulous opening in the bladder was not
Tech Coloproctol Table 2 Operative findings of patients with colonic fistula
Table 3 Postoperative outcome of patients with colonic fistula
N = 45 Operation type Anterior resection Sigmoidectomy Ileocolic resection Pelvic exenteration Hemicolectomy Revision of colostomy Laparoscopic resection Ureteral catheters Splenic flexure mobilization Stoma Temporary ileostomy Permanent colostomy Median operative time, minutes (mean, range) Median blood loss, milliliters (mean, range) Intraoperative transfusion Intraoperative complications Ureteral laceration Bladder laceration
22 (49 %) 17 (38 %) 2 (4 %) 2 (4 %) 1 (2 %) 1 (2 %) 19 (42 %) 26 (58 %) 15 (33 %) 12 (27 %) 10 (23 %) 2 (4 %) 263 (292, 63–725) 300 (468, 50–3,600) 9 (20 %) 2 (4 %) 1 (2 %) 1 (2 %)
routinely closed. Formal bladder repair by the colorectal surgeon was performed for large full thickness defects where the bladder lumen could be visualized. The splenic flexure was mobilized in one-third of the patients. A temporary ileostomy was established in 10 patients (23 %) and all were subsequently reversed. Fecal diversion was performed selectively. Indications included prior failed repair, radiated patients, low anastomosis (i.e. proctectomy), or any intraoperative concerns about the integrity of the anastomosis. The median operative time was 263 min (range 63–725 min) with a median blood loss of 300 ml (range 50–3,600 ml). Two patients (4 %) suffered an intraoperative complication involving laceration of the bladder and the left ureter and both were recognized and repaired intraoperatively. The median length of stay was 6 days (range 2–65 days) and 18 % of the patients required intensive care. The overall complication rate was 47 % (Table 3). Wound infection was the most common complication (20 %). No anastomotic leak was noted but 1 patient (2 %) developed a postoperative abdominopelvic abscess. The readmission rate was 29 %. A small bowel obstruction was the most common reason for readmission (9 %). Two patients (4 %) required an operation at readmission. All patients healed their fistula and during a median follow-up of 37 months (range 8–102 months), no recurrence was noted. Ninety-day mortality was 4 % with 2 patients succumbing to acute renal failure with cardiopulmonary failure.
N = 45 Median length of stay, days (mean, range)
6 (9, 2–65)
Intensive care hospitalization
8 (18 %)
Overall complicationsa
21 (47 %)
Wound infection/dehiscence
12 (27 %)
Sepsis, urinary tract infections
4 (9 %)
Small bowel obstruction
4 (9 %)
Deep venous thrombosis, thrombophlebitis
3 (7 %)
Postoperative bleeding
2 (4 %)
Abdominopelvic abscess
1 (2 %)
Hydronephrosis
1(2 %)
Vocal cord paralysis Trocar site hernia
1 (2 %) 1 (2 %)
Postoperative transfusion
7 (16 %)
Readmission
13 (29 %)
Small bowel obstruction
4 (9 %)
Abdominal pain
2 (4 %)
Rectal bleeding
2 (4 %)
Hydronephrosis
1 (2 %)
Abdominopelvic abscess
1 (2 %)
Wound dehiscence
1 (2 %)
Deep venous thrombosis
1 (2 %)
Urinary tract infection
1 (2 %)
Trocar site hernia Reoperation during readmission Hydronephrosis Small bowel obstruction from trocar site hernia Recurrence of fistula 90-day mortality a
1 (2 %) 2 (4 %) 1 (2 %) 1 (2 %) 0 2 (4 %)
Some patients with more than one complication
Discussion Colonic fistulas are relatively uncommon and are most often the result of complicated diverticulitis. Surgical intervention remains the treatment modality of choice in patients with fistulas refractory to medical therapy. Resection and/or fecal diversion are the most commonly performed procedures for this condition. This study aimed at determining the outcome of patients treated surgically for this condition in recent years. Current surgical care includes standardization of perioperative antimicrobial regimen administration, routine use of chemical prophylaxis for deep venous thromboembolism, and increased use of the laparoscopic technique. The majority of fistulas were related to chronic diverticulitis and a significant number of patients in this study presented with a chronic abdominopelvic abscess. Diverticulitis is the most common cause of colonic fistulas with published rates ranging from 62 to 75 % of all fistulas [2–8]. Seventy-four percent of the
123
123
Albert Schweitzer Hospital, The Netherlands
Kaiser Permanente, California
Smeenk et al. [13]
Current study
b
Range Mixed colovaginal and colovesical fistulas
Urologishe Klinik und Poklinik, Germany
a
Mount Sinai, New York
Leicht et al. [11]
The Leister General Hospital, UK
Garcea et al. [4]
Lynn et al. [2]
Darent Valley Hospital, UK
Solkar et al. [21]
Barnet and Chase Farm NHS Trust, UK
Virginia Commonwealth University, Virginia
Najjar et al. [5]
Holoryd et al. [3]
Hospital Cantonal De Fribourg, Switzerland
Menenakos et al. [17]
Dandery Hospital, Sweden
Glasgow Royal Infirmary, UK
Walker et al. [7]
Johannes Guterberg University, Germany
Aberdeen Royal Hospital, UK
Driver et al. [8]
Melchior et al. [10]
Cleveland Clinic, Ohio
Woods et al. [1]
Hjern et al. [14]
Institution
Author
Table 4 Summary of recent studies on colonic fistulas
2013
2012
2012
2012
2012
2009
2007
2006
2005
2004
2003
2002
1997
1988
Year
45
40
54
72
37
49
60
72
50
12
18
19
67
84
N
58
88
100
100
86
100
100
100
100
100
83
100
100
65
Colovesical (%)
Fistula type
15
12
–
–
14
–
–
–
–
–
17
–
–
28
Colovaginal (%)
18
–
–
–
–
–
–
–
–
–
–
–
–
–
Colocutaneous (%)
48 48
9b
11
–
22
8
28
42
35
–
27
84
10
27
–
–
–
–
–
–
–
–
–
–
–
–
7
Coloenteric (%)
Morbidity (%)
4%
8%
0%
10 %
17 %
0%
0%
7%
0
1%
0
1%
10 %
3.5 %
Mortality
0
–
0%
11 %
5%
0%
–
–
2%
0%
5.5 %
0%
14 %
0%
Recurrence
40
–
62
31
23
68
12
–
–
2–108a
26
35
–
–
Mean follow-up, months
Tech Coloproctol
Tech Coloproctol
fistulas in our series were attributable to diverticulitis. The remaining fistulas were postoperative, radiation-induced, malignant, or related to Crohn’s disease. The bladder was involved in over half of the patients. Colovesical fistula is the most common type of colonic fistula and has been reported in 50–100 % of published series on this condition [1–21]. Anterior resection was the most frequently performed operation in our study and nearly half of the patients underwent a laparoscopic resection. Postoperative complications were common and often infectious in nature but the mortality rate was low. Postoperative complications led to a readmission in about one-third of the patients in this study. Several recent studies have reported morbidity rates of 8–48 % and mortality rates of 0–18 % [3, 10, 11, 13]. In our study a primary anastomosis was performed in 96 % of the patients and no anastomotic leak was noted. In one of the largest series published to date on colonic fistula, the Cleveland Clinic reported a 5 % leak rate in 84 patients undergoing resection and primary anastomosis [1]. However resolution of the fistula was noted in all of the patients with no recurrences, a finding similar to our results. There were some limitations to this study. Because the incidence of colonic fistula is low in the general population, our sample size of 45 patients over an 8-year period was relatively small. Furthermore, this study was conducted at a tertiary institution with a large population of complex and/or elderly patients with significant co-morbidities which may have skewed the morbidity and mortality findings. Lastly, the study was retrospective in nature which inherently may introduce bias in terms of patient selection and follow-up. Despite the acknowledged shortcomings, this study provided valuable insight into the short- and long-term outcome of the surgical management of colonic fistulas in the modern surgical era. There is a paucity of data from the United States regarding the surgical outcome of patients treated for colonic fistulas with the majority of existing literature originating from Europe (Table 4) [1–5, 7, 8, 10– 14, 17, 21]. Furthermore, most of the available data in the literature describes the outcome of patients with colovesical fistulas. The largest colovesical fistula series to date comes from Lynn et al. [2] from Mount Sinai Hospital in New York. Seventy-two patients underwent operative repair for colovesical fistula with a recurrence rate of 11 %. No difference in recurrence rate was noted in that study between patients who underwent a single stage procedure versus a multi-stage repair. Our study included all fistula types including colovesical, colocutaneous, and colovaginal. During a mean follow-up of 40 months, no recurrence was noted. The reported recurrence rate in the literature ranges from 0 to 14 % [1–5, 7, 8, 10–14, 17, 21]. Recently, laparoscopic colectomy has gained popularity for elective resection but has not been universally accepted
for complicated diverticular disease including colonic fistulas. A laparoscopic resection was performed in nearly half of our patients with good success and no conversion. The lack of conversion in this study maybe attributed to both the advanced laparoscopic skills of the lead surgeon and to proper selection of patients. Hence a selection bias (i.e. choosing the cases deemed feasible with laparoscopy) could explain the lack of conversion. Menenakos et al. [17] studied 18 patients who underwent laparoscopic resection for colonic fistulas and found only 1 recurrence (5.5 %). Many subsequent studies have supported the laparoscopic approach for diverticular fistula based on low rates of postoperative morbidity and low conversion rates, though both of these outcome measures are higher in patients with diverticular fistulas than in patients with non-complicated diverticulitis [18–20].
Conclusions This retrospective study from a tertiary colorectal surgery unit demonstrated that colonic fistula can be successfully eradicated with surgical intervention in the majority of patients. The most common fistula type was colovesical and was often secondary to chronic diverticulitis. The 90-day mortality rate was low following surgical intervention. However postoperative complications occurred in approximately half of the patients with one-third requiring readmission. Despite recent advances in surgical care such as prophylactic use of intravenous antibiotics to prevent wound infection, subcutaneous heparin for venous thrombosis prophylaxis, and an increasing use of laparoscopic surgery, the morbidity associated with operative intervention for colonic fistula remains significant. Conflict of interest
None.
References 1. Woods RJ, Lavery IC, Fazio VW, Jahgelman DG, Weakley FL (1988) Internal fistulas in diverticular disease. Dis Colon Rectum 31:591–596 2. Lynn ET, Ranasinghe NE, Dallas KB, Divino CM (2012) Management and outcomes of colovesical fistula repair. Am Surg 78:514–518 3. Holroyd DJ, Banerjee S, Beavan M, Prentice R, Vijay V, Warren SJ (2012) Colovaginal and colovesical fistulae: the diagnostic paradigm. Tech Coloproctol 16:119–126 4. Garcea G, Majid I, Sutton CD, Pattenden CJ, Thomas WM (2006) Diagnosis and management of colovesical fistulae; six-year experience of 90 consecutive cases. Colorectal Dis 8:347–352 5. Najjar SF, Jamal MK, Savas JF, Miller TA (2004) The spectrum of colovesical fistula and diagnostic paradigm. Am J Surg 188:617–621 6. Pollard SG, Macfarlane R, Greatorex R, Everett WG, Hartfall WG (1987) Colovesical fistula. Ann R Coll Surg Engl 69:163–165
123
Tech Coloproctol 7. Walker KG, Anderson JH, Iskander N, McKee RF, Finlay IG (2002) Colonic resection for colovesical fistula: 5-year follow-up. Colorectal Dis 4:270–274 8. Driver CP, Anderson DN, Findlay K, Keenan RA, Davidson AI (1997) Vesico-colic fistulae in the Grampain region: presentation, assessment, management and outcome. J R Coll Surg Edinb 42:182–185 9. Wassef R, Morgan S, Tasse´ D, Bernard D (1983) Fistulas in diverticular disease of the colon: study of 29 cases. Can J Surg 26:546–549 10. Melchior S, Cudovic D, Jones J, Thomas C, Gillitzer R, Thu¨roff J (2009) Diagnosis and surgical management of colovesical fistulas due to sigmoid diverticulitis. J Urol 182:978–982 11. Leicht W, Thomas C, Thu¨roff J, Roos F (2012) Colovesical fistula caused by diverticulitis of the sigmoid colon: diagnosis and treatment. Urology 51:971–974 (German) 12. King RM, Beart RW, Mcllrath DC (1982) Colovesical and rectovesical fistulas. Arch Surg 117:680–683 13. Smeenk RM, Plaisier PW, van der Hoeven JA, Hesp WL (2012) Outcome of surgery for colovesical and colovaginal fistulas of diverticular origin in 40 patients. J Gastrointest Surg 16:1559– 1565
123
14. Hjern F, Goldberg SM, Johansson C, Parker SC, Mellgren A (2007) Management of diverticular fistulae to the female genital tract. Colorectal Dis 9:438–442 15. Bahadursingh AM, Longo WE (2003) Colovaginal fistulas. Etiology and management. J Reprod Med 48:489–495 16. Greenstein AJ, Sachar DB, Tzakis A (1984) Course of enterovesical fistulas in Crohn’s disease. Am J Surg 17:788–792 17. Menenakos E, Hahnloser D, Nassiopoulos K, Chanson C, Sinclair V, Petropoulos P (2003) Laparoscopic surgery for fistulas that complicate diverticular disease. Langenbecks Arch Surg 388:189–193 18. Bartus CM, Lipof T, Sarwar CM et al (2005) Colovesical fistula: not a contraindication to elective laparoscopic colectomy. Dis Colon Rectum 48:233–236 19. Zapletal C, Woeste G, Bechstein WO, Wullstein C (2007) Laparoscopic sigmoid resections for diverticulitis complicated by abscesses or fistulas. Int J Colorectal Dis 22:1515–1521 20. Royds J, O’Riordan JM, Eguare E, O’Riordan D, Neary PC (2012) Laparoscopic surgery for complicated diverticular disease: a single-center experience. Colorectal Dis 14:1248–1254 21. Solkar MH, Forshaw MJ, Sankararajah D, Stweart M, Parker MC (2005) Colovesical fistula—is a surgical approach always justified? Colorectal Dis 7:467–471
