Colosplenic fistula diagnosis and management: a case series and review of literature (2024)

Highlights

• The largest review of rare colosplenic fistula (CsF) cases, including 30 patients internationally.

• Summary of CsF causes, clinical features, diagnosis, and management - Surgical intervention was the most successful treatment for symptom resolution.

• Colonoscopy caused no complications in the evaluation of CsF.

Eligibility criteria

For the review, we included studies reporting patients diagnosed with CsF, regardless of etiology or management.

Data sources and searches

We applied a search strategy developed in collaboration with an experienced librarian to find potentially eligible studies in Ovid MEDLINE, Ovid EMBASE, Scopus, Web of Science, and Wiley Cochrane Library from each database’s inception until June 2022. A review of the literature was performed using the search terms ‘colosplenic fistula’, ‘splenocolic fistula’, and ‘splenocolonic fistula’. No language restrictions were imposed. Conference abstracts were included, and literature reviews were excluded. The reference lists of selected studies were searched to identify additional publications (Supplementary Search Strategy, Supplemental Digital Content 3, http://links.lww.com/JS9/B808).

The Cleveland Clinic Foundation – IRB 08-670 was used to conduct a retrospective review of the available CsF cases at our institution. Potential patients were identified from the electronic medical records using natural language processing and searching the operative reports of all surgeries related to using the search terms ‘colosplenic fistula’, ‘splenocolic fistula’, and ‘splenocolonic fistula’.

Study selection

The search strategy results were uploaded to an online software program (Covidence Systematic Review Software; Veritas Health Innovation)³⁰. Next, the abstracts and full texts were screened. The reviewers (R.S. and F.A.C.) worked independently and in duplicate. Before initiating the study selection process, a pilot study with five articles was conducted to clarify the eligibility criteria. For abstract screening, articles included by at least one reviewer were considered for a full-text review. In the full-text review phase, only articles included by both reviewers were deemed eligible for this systematic review and in the case of disagreements, these were resolved by consensus with a third reviewer (E.P.L.). Full-text review agreement was substantial (Cohen’s kappa=0.7). Articles written in languages other than English or Spanish were translated using Google Translate (GT).

Data collection

Data from the included studies were extracted independently and in duplicate in a standardized manner by two reviewers (R.S. and F.A.C.). A third review checked all the included studies for accuracy (E.P.L.). The following information was extracted: general characteristics (first author’s last name, publication date, and country), patient characteristics (age, sex, and comorbidities), CsF data collected included cause, symptoms at presentation, diagnosis approach, management approach, pathology findings, intraoperative complications, postoperative complications (Clavien–Dindo and 30-day mortality), and prognosis.

Methodological quality in individual studies

We used a tool proposed by Murad et al.³¹ to evaluate the methodological quality of the case reports. This tool is composed of four domains assessed by eight questions: (1) selection, (2) ascertainment, (3) causality, and (4) reporting. Case reports were assessed by two reviewers (E.P.L. and O.H.D.), with disagreements resolved by consensus with a third reviewer (T.U.). When only two or fewer questions were not satisfied, the risk of bias was considered low. Studies with three unsatisfied questions were assessed as unclear risk. Case reports with four or more questions not satisfied were considered to have a high risk of bias (Supplementary Table A.2, Supplemental Digital Content 4, http://links.lww.com/JS9/B809).

Statistical analysis

A narrative synthesis of the included studies was provided. We used descriptive statistics (e.g. percentages) to report our findings and summarized them in figures and tables.

Open data

To support reproducible research, we made all our files freely available on an online platform. Documents used in the screening phases (references screened in abstract screening and full-text screening) and extra information (tables and figures) are available at https://figshare.com/s/4130858d368e899525ec.

Table 1

Summary of colosplenic fistula cases at our institution.

Table 2

Summary of colosplenic fistula review.

Clinical features

Most patients with CsF were middle-aged males. The presenting symptoms vary depending on the underlying etiology. Nevertheless, several similar clinical features were observed in this small patient pool. About three-quarters of the patients reported pain or discomfort in the left upper quadrant of the abdomen, which may suggest location of pathology. However, a palpable left flank mass was rarely present and seems to not be an indicator of CsF. Notably, approximately a quarter of the cases reviewed reported LGIB, which is thought to be from the direct evacuation of the splenic pulp into the colon via the fistula^10,13,26. Anemia was also repeatedly seen, which could be due LGIB or chronic disease from an underlying malignancy. Another common feature was fever, leukocytosis, and elevated CRP to suggest an inflammatory process or infection. Unfortunately, CsF can present with hemodynamic instability and septic shock requiring urgent intervention.

Etiology and pathology findings

CsFs were first described in a 1985 case report as a rare complication of Crohn’s disease¹. Subsequent case reports have identified CsFs in other contexts such as colonic lymphoma and colorectal cancer. Our review of available cases identifies colonic lymphoma as one of the most common causes of CsF. Similarly, in our institution, three of the four CsF cases were also diagnosed with colonic lymphoma. Interestingly, the gastrointestinal (GI) tract is the most common extranodal site of lymphoma, although the colon is rarely the site of involvement^32–34. Although less than 3% of colonic lymphomas perforate³⁵, our analysis suggest colonic lymphoma perforations at the splenic flexure may lead to the formation CsF. Naschitz et al.²⁵ hypothesized that colonic lymphomas perforate and form fistulas due to the massive tumorous involvement of tissues, extensive coagulative necrosis, and absence of desmoplastic reaction. This hypothesis is supported by several cases in our review, which found histological evidence of colonic lymphoma cells along the malignant fistulous tract and inside the spleen. Evidence from our review suggests that in the rare case of left colonic lymphoma, clinicians should be wary of CsFs as a complication.

CsFs have also been repeatedly observed as a complication of colon cancer of the splenic flexure. The cases reviewed in this study reported fistula formation mostly because of malignancy-driven perforation at the splenic flexure, rather than from direct invasion into the spleen. Colon cancer perforation is either due to tumor necrosis at the cancer site, or to distended colon blowout from a distal malignant obstruction³⁶. Gervaise et al.²² postulated that after perforation of the colon at the splenic flexure, the local inflammatory process and adhesion formation trigger the development of a fistulous tract with the spleen. Although splenic flexure tumors only account for 3–5% of all colon cancer cases^37,38, CsFs may be a complication of tumors at this location.

Crohn’s is another cause of CsF. Crohn’s disease causes chronic transmural inflammation throughout the gastrointestinal tract. Up to 50% of patients with Crohn’s disease develop fistulas when this inflammation causes a transmural ulcer that penetrates from the affected bowel to adjacent structures³⁹. The majority of intra-abdominal fistulae are entero-enteric⁴⁰. However, our review revealed that patients with Crohn’s disease can form a fistula from their diseased colon to the adjacent spleen. Many of these cases have reported deep ulcerations in the colon adjacent to the fistula, supporting Crohn’s colitis as the etiology^7,8.

Several cases of CsF have unique etiologies but a similar proposed mechanism of fistula formation. Two case reports suggested that a massive spleen can cause pressure necrosis in the adjacent colon and result in a fistula^4,13. Others have reported infectious causes of fistula development. For example, drainage from splenic abscess^4,5 or echinococcal hydatid cyst²⁶. Similarly, our institution reported a case of endocarditis resulting in a splenic abscess and ultimately a CsF.

Diagnostic modalities

Several imaging modalities and endoscopic tools have been used to evaluate and diagnose CsF. In our review, chest radiography and abdominal plain radiography rarely offered diagnostic value. Fistulograms were effective under special circ*mstances in which the spleen was intubated with a percutaneous drain. In such cases, the contrast injected through the tube was shown to flow into the colon. Contrast enemas were also successful in showing contrast flow through the fistula and did not require percutaneous access. Enemas can be used as a confirmatory adjunct after noncontrast CT or splenic ultrasound raised suspicion of the fistula. Abdominal or splenic ultrasound (US) is helpful in identifying an enlarged spleen, splenic lesions, or air in the spleen. However, they have limited utility for the diagnosis of CsF. CT imaging either effectively diagnosed CsF or was highly suggestive based on the identification of a splenic abscess. CT with intravenous contrast helped identify splenic lesions and colonic thickening, whereas enteral contrast was useful in identifying the fistula tract as several cases showed enteral contrast entering the spleen. MRI was not performed in any case. A systematic review by Panes et al.⁴¹ suggested that CT, US, and MRI have high accuracy (sensitivity and specificity >80%) in identifying intra-abdominal fistulas, but the selection between MRI and US is dependent on local expertise and availability. Colonoscopy was helpful in identifying the underlying etiology, but seldom identified a CsF. Snell² suggested that colonoscopy had the potential to cause splenic rupture via insufflation through a CsF. However, several cases in this review performed colonoscopies without incident and therefore appears to be safe in the setting of CsF^{11–13,21,25,26}.

Management

As most patients presented with fever and leukocytosis, they were empirically treated with intravenous (IV) fluid and intravenous broad-spectrum antibiotics. Depending on the clinical severity of the presenting symptoms or etiology, patients were either managed nonoperatively or with upfront surgery. Nonoperative approaches consist of addressing the splenic collection/abscess, which most typically involves placing an image-guided percutaneous drain into the spleen. However, this approach remained inadequate in addressing the fistula, as only one of the five cases did not require surgical treatment for resolution of symptoms. Reasons for failure included continued fever, worsening abdominal pain, continued septicemia, and uncontrolled feculent drainage from the drain. Therefore, percutaneous drains can help with diagnosis and temporizing measures, but have lower curative potential. Similarly, medical therapy without invasive procedure or surgery was helpful in stabilizing patients, but half eventually required surgical intervention.

Surgical management of CsFs usually includes open splenectomy and segmental resection of the involved colon. Undergoing both in conjunction seems to adequately address the fistula from both ends of the tract. One case attempted just an inferior resection of the spleen at the site of the fistula, but this patient died from complications of the surgery. Although only one case of this partial splenectomy was found, it appears that a full splenectomy is safer and effective in treatment of CsF. Additionally, due to the high incidence of neoplasia and Crohn’s as cause of CsF, it is recommended that a segmental resection be performed. Although, if a neoplastic process is present, the extent of colon and spleen resection should be evaluated and aligned with patient’s wishes and oncological team. For example, there was a case in which the etiology of CsF was advanced and surgical intervention would cause more harm than benefit. The patient was transitioned to comfort care and died shortly after from their malignancy²⁰.

Prognosis

Overall, patients presenting with symptoms resolved after intervention to address CsF. Relatively few complications were reported in the nonoperative group and in those that received surgical intervention. Almost all postoperative recoveries were uneventful and showed resolution of CsF symptoms on follow-up. As expected, resolution of CsF occurred with surgical resection of the affected organs. Of the 30 cases reviewed, we identified two patients with intraoperative complications and five patients with postoperative complications. Four of these cases had relatively insignificant Clavien–Dindo I complications, including respiratory insufficiency and atrial fibrillation. However, one patient at our institution had Clavien–Dindo IV postoperative complications that required reoperation and intensive care management for missed gastrostomy. In the three mortalities, two of the cases had advanced underlying malignancy with progression prompting palliative care. Only one case of mortality could be possible related to the surgical treatment. This was a surgery in a patient with splenic flexure colon cancer complicated by perforation, CsF, and splenic abscess²². In this case, there was an attempt at partial splenectomy of the inferior portion of the spleen.

Strengths and limitations

The main limitation of our review was the paucity of data regarding CsFs, which may reflect the true rarity of this complication or simply the lack of publication on the topic. In addition, some studies lacked granular data on the outcomes of interest. We were unable to statistically compare or analyze the outcomes. Finally, for the abovementioned reasons, we remain limited in our ability to draw strong recommendations, as we could only generate an overview of what is happening with patients with CsF. Therefore, this piece can serve as an educational tool.

Despite these limitations, the strengths of this study were notable. To our knowledge, this is the first global perspective systematic review describing the spectrum related to the care of patients with CsFs. Second, our inclusion of conference abstracts and the lack of language restrictions enabled us to include a satisfactory number of patients. Lastly, we followed a systematic methodology by applying prespecified and detailed data tabulation and extraction and standardized evaluation of evidence quality and publication bias. All steps were rigorously performed by multiple researchers.

SUPPLEMENTARY MATERIAL:

Oscar Hernandez Dominguez and Eddy P. Lincango are claiming first authorship.

This abstract has been accepted for the selected as a poster presentation to be presented at the ASCRS 2023 Annual Scientific Meeting, June 3-6, 2023 at the Seattle, WA.

Sponsorships or competing interests that may be relevant to content are disclosed at the end of this article.

Supplemental Digital Content is available for this article. Direct URL citations are provided in the HTML and PDF versions of this article on the journal’s website, www.lww.com/international-journal-of-surgery.

Published online 6 February 2024

1. Wechter DG, Willson RA. A rare splenic abscess complication of Crohn’s colitis. Dig Dis Sci1985;30:802–806. [PubMed] [Google Scholar]

2. Snell DB. Colon cancer with colosplenic fistula. Gastroenterol Hepatol Open Access2017;7:6–7. [Google Scholar]

3. Hashemzadeh S, Pouryousef K, Oliaei-Motlagh M, et al.. Colosplenic fistula as a complication of splenic abscess. ANZ J Surg2016;86:623. [Google Scholar]

4. Debat Zoguereh D, Dubau A, Badiaga S. Calcified splenic abscess, colonic fistula and ascites in a chronic carrier of Salmonella typhi. Proc EBC Congr1998;22:1102–1105. [Google Scholar]

5. Cowie MR, Hoffbrand BI, Grant DS. Liencolonic fistula following splenic abscess. J R Soc Med1992;85:636–637. [PMC free article] [PubMed] [Google Scholar]

6. Winter MW, Lee S. Colosplenopleural fistula: an unusual colonic fistula in a 44-year-old male with Crohn’s disease. Radiol Case Rep2014;9:1028. [PMC free article] [PubMed] [Google Scholar]

7. Pappalardo E, Ricci A, Dray X, et al.. Splenic abscess secondary to a colosplenic fistula in Crohn’s disease. Acta Chir Belg2007;107:323–324. [PubMed] [Google Scholar]

8. Rowell DL, Longstreth GF. Colosplenic fistula and splenic abscess complicating crohn’s colitis. J Clin Gastroenterol1995;21:74–75. [PubMed] [Google Scholar]

9. McCrystal S, Hatzifotis M. Splenocolonic fistula following non-operative management of splenic rupture. Trauma2013;15:86–90. [Google Scholar]

10. Kesieme EB, Dongo AE, Osime CO, et al.. Gastric duplication: a rare cause of massive lower gastrointestinal haemorrhage, chest wall mass, and enterocutaneous fistula. Case Rep Gastrointest Med2012;2012:1–3. [Google Scholar]

11. Goldsmith P, Pine JK, Smith AM. An unusual case of rectal bleeding colosplenic fistula complicating pancreatitis. J ECT2011;27:175–177. [Google Scholar]

12. Benizri EI, Rahili A, Bernard JL, et al.. Primary cyst of the spleen presenting as a splenocolic fistula. Clin Res Hepatol Gastroenterol2011;35:511–512. [PubMed] [Google Scholar]

13. Means JR, Villella ER, Stahlfeld KR. Splenocolic fistula in a patient with polycythemia vera. Am J Surg2003;185:173–174. [PubMed] [Google Scholar]

14. Awotar GK, Luo F, Zhao Z, et al.. Splenic abscess owing to cancer at the splenic flexure: a case report and comprehensive review. Medicine (Baltimore)2016;95:e4941. [PMC free article] [PubMed] [Google Scholar]

15. Alrahbi R, Al Ayoubi F, Bessayah A, et al.. Colosplenic fistula: case report. ANZ J Surg2018;88:E636. [Google Scholar]

16. Ashouri M, Kuhi Z, Ataie-Ashtiani Z, et al.. Colosplenic fistula presentation in the context of undiagnosed colon cancer: case report and review of literature. Int J Surg Case Rep2022;92:106828. [PMC free article] [PubMed] [Google Scholar]

17. Yokoyama Y, Kashyap S, Ewing E, et al.. Gastrosplenocolic fistula secondary to non-Hodgkin B-cell lymphoma. J Surg Case Rep2020;2020:1–3. [Google Scholar]

18. Radulescu A, Arrese D, Bach J. Colosplenic contained perforation secondary to colonic lymphoma. Int J Crit Illn Inj Sci2015;5:213. [PMC free article] [PubMed] [Google Scholar]

19. Chun ES, Demos TC, Gaynor ER. Colosplenic fistula in a patient treated with interleukin-2 for malignant melanoma. J Comput Assist Tomogr1997;21:674–676. [PubMed] [Google Scholar]

20. Goldberg JB, Moses RA, Holubar SD. Colosplenic fistula: a highly unusual colonic fistula. J Gastrointest Surg2012;16:2338–2340. [PubMed] [Google Scholar]

21. Al-Zahir AA, Meshikhes AWN. Colonic lymphoma presenting acutely with perforated colo-splenic fistula. Int J Surg Case Rep2012;3:368–371. [PMC free article] [PubMed] [Google Scholar]

22. Gervaise A, De Saint Roman C, Sockeel P, et al.. Splenic abscess secondary to a colosplenic fistula as the presenting manifestation of colon cancer. J Radiol2010;91:1259–1262. [PubMed] [Google Scholar]

23. Nishiwaki M. A case of mucinous cystadenocarcinoma of the pancreas with splenic abscess and fistula between the splenic abscess and the colon. Japanese Soc Gastroenterol Surg2005;38:190–195. [Google Scholar]

24. Paramelle PJ, Ferretti G, Desroches E, et al.. Cancer of the left colonic angle with colonic-splenic fistula, thrombosis of the splenic vein. Proc EBC Congr2001;82:511–513. [Google Scholar]

25. Naschitz JE, Yeshurun D, Horovitz IL, et al.. Spontaneous colosplenic fistula complicating immunoblastic lymphoma. Dis Colon Rectum1986;29:521–523. [PubMed] [Google Scholar]

26. Teke Z, Yagci AB, Atalay AO, et al.. Splenic hydatid cyst perforating into the colon manifesting as acute massive lower gastrointestinal bleeding: an unusual presentation of disseminated abdominal echinococcosis. Singapore Med J2008;49:13–16. [Google Scholar]

27. Hernandez Dominguez O, Lincango E, Spivak R, et al.. S101 Colosplenic fistula in inflammatory bowel disease: a scoping review. Am J Gastroenterol2022;117:S26–S26. [Google Scholar]

28. Page MJ, McKenzie JE, Bossuyt PM, et al.. The PRISMA 2020 statement: an updated guideline for reporting systematic reviews. Int J Surg2021;88:105906. [PubMed] [Google Scholar]

29. Shea BJ, Reeves BC, Wells G, et al.. AMSTAR 2: a critical appraisal tool for systematic reviews that include randomised or non-randomised studies of healthcare interventions, or both. BMJ2017;358:4008. [Google Scholar]

30. Veritas Health Innovation. Covidence systematic review software.

31. Murad MH, Sultan S, Haffar S, et al.. Methodological quality and synthesis of case series and case reports. Evid Based Med2018;23:60–63. [Google Scholar]

32. Lightner AL, Shannon E, Gibbons MM, et al.. Primary Gastrointestinal Non-Hodgkin’s Lymphoma of the Small and Large Intestines: a Systematic Review. doi: 10.1007/s11605-015-3052-4

34. Romaguera J, Hagemeister FB. Lymphoma of the colon. Curr Opin Gastroenterol2005;21:80–84. [PubMed] [Google Scholar]

35. Vaidya R, Habermann TM, Donohue JH, et al.. Bowel perforation in intestinal lymphoma: incidence and clinical features. Ann Oncol2013;24:2439. [PMC free article] [PubMed] [Google Scholar]

36. Otani K, Kawai K, Hata Keisuke, et al.. Colon cancer with perforation 2019;49:15–20. [Google Scholar]

37. Shaikh IA, Suttie SA, Urquhart M, et al.. Does the outcome of colonic flexure cancers differ from the other colonic sites?. doi: 10.1007/s00384-011-1292-7

38. Bhangu A, Kiran RP, Slesser A, et al.. Survival after resection of colorectal cancer based on anatomical segment of involvement. Oncol2013;20:4161–4168. [Google Scholar]

39. Bell SJ, Williams AB, Wiesel P, et al.. The clinical course of fistulating Crohn’s disease. doi: 10.1046/j.0269-2813.2003.01561.x

40. Benson AA, Aviran E, Yaari S, et al.. Clinical and radiologic characteristics of intra-abdominal fistulizing Crohn’s disease. Res Artic Dig2020;101:667–673. [Google Scholar]

41. Panés J, Bouzas R, Chaparro M, et al.. Systematic review: the use of ultrasonography, computed tomography and magnetic resonance imaging for the diagnosis, assessment of activity and abdominal complications of Crohn’s disease. Aliment Pharmacol Ther2011;34:125–145. [PubMed] [Google Scholar]