FormalPara Key Summary Points

There is a great need for more epidemiology work to determine the true burden of CCF in the global population.

Primary healing rates associated with local surgical and intersphincteric ligation procedures for CCF were 57.1–100%.

Recurrence rates associated with local surgical and intersphincteric ligation procedures for CCF were 4.9–60.7%.

Failure of local surgical and intersphincteric ligation procedures for CCF occurred in 2.8–18.0% of patients.

There are substantial knowledge gaps and a great need for future research to help improve our understanding of how best to support and treat patients with CCF.

Introduction

Anal fistulas are abnormal passages connecting the anal canal to the skin near the anus [1]. Approximately 90% of anal fistulas are idiopathic [2]. Parks’ cryptoglandular theory on the pathogenesis of anal fistulas hypothesizes that infected anal glands and surrounding anal abscesses eventually progress to fistulas [3]. About 40–70% of patients with an anal abscess have a concomitant anal fistula; and even months or years after abscess drainage, 30% of patients will be diagnosed with an anal fistula [1].

Cryptoglandular fistulas (CF) are generally classified by the anatomic location of the primary tract relative to the anal sphincter muscles: intersphincteric, transsphincteric, suprasphincteric, and extrasphincteric fistulas [2, 4]. CF with significant involvement of the external sphincter or multiple tracts are classified as “complex” [5]. Patients with complex CF (CCF) often experience compromised quality of life due to painful defecation, constant discharge, reduced social functioning, and/or recurrence [6]. Eradicating the anal fistula(s) and preventing recurrence while maintaining fecal continence are the goals of managing anal fistulas [1]. The surgical management of anal fistulas is a trade-off between the extent of operative sphincter division and postoperative functional loss. For example, higher rates of fecal incontinence (FI) and longer healing times are often associated with effective, but sphincter-dividing, options such as fistulotomy and fistulectomy [1, 5]. Conversely, less invasive sphincter-sparing interventions, such as loose seton and fibrin glue, have varying success rates and patients often face multiple operations [4].

This systematic literature review (SLR) aimed to: (1) identify the global incidence/prevalence of CF; and (2) evaluate and summarize evidence published within the past 5 years on treatment outcomes of local surgical (fistulotomy, lay open fistulotomy, fistulectomy, modified Parks’ technique, and advancement flap) and intersphincteric ligation procedures (ligation of the intersphincteric fistula tract [LIFT], BioLIFT (LIFT with a bioprosthetic graft), and transanal opening of intersphincteric space [TROPIS]) for CCF.

Materials and Methods

The SLR was conducted according to the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) [7]. The protocol was registered with the International Prospective Register of Systematic Reviews (PROSPERO; registration number CRD42020177732). This article is based on previously conducted studies and does not contain any new studies with human participants or animals performed by any of the authors.

Search Strategy and Eligibility

Eligibility was based on study Population, Intervention, Comparison, Outcomes, Time, and Study design (PICOTS) (Supplementary Material Table S1). The electronic search was conducted on March 25, 2020 in PubMed and Embase (Supplementary Material Tables S2 and S3) using human studies published in English up to 10 years prior to the search date. An additional manual search of key publications and references was conducted to identify any studies missed by the electronic search. Only studies reporting incidence/prevalence of CF or outcomes for CCF for local surgical or intersphincteric ligation procedures of interest were included in this manuscript.

Titles and abstracts of identified studies were independently screened by two reviewers to determine whether they met the PICOTS criteria. If so, the full-text articles were independently assessed by each reviewer to determine eligibility for data abstraction. Discrepancies in either phase were resolved by consensus. If consensus could not be achieved, a third senior reviewer made the determination. Out-of-scope studies and those for which the full text was unavailable, and the abstract did not include sufficient information were excluded with a documented rationale. Data from eligible studies were independently abstracted by two reviewers using a standardized data abstraction form. Both reviewers jointly examined abstraction spreadsheets to synthesize the data into one master spreadsheet. Data were extracted for multiple variables, including study type, design, population, outcomes, and limitations.

Studies included in this report met the following criteria: (1) reported on CF; (2) used an observational study design; (3) measured incidence/prevalence (for any CF) or clinical outcomes of interest (healing/failure/recurrence rates, pain, FI) (for CCF only); and (4) were original research (Supplementary Material Table S1). Case series were designated as cohort studies if they met all of the following pre-specified criteria: more than 10 patients per fistula type, patients sampled on the basis of exposure (not outcome), outcome assessed over a pre-specified follow-up period or mean/median follow-up reported, and information available to calculate the absolute/relative risk. In addition, sampling had to be labeled as “consecutive” or text had to indicate that all eligible patients were included to avoid selection bias. CCF also had to be reported separately from other types of fistulas.

Study populations were classified as surgery-naïve and/or surgery-experienced. Patients receiving drainage and incision or prior seton were classified as surgery naïve as these procedures are often performed as preparatory procedures for the current surgery.

Risk of Bias Assessment

Two independent reviewers assessed risk of bias in each article using the Risk of Bias in Non-randomized Studies of Interventions (ROBINS-I) tool for observational studies [8]. Any disagreements were settled by consensus, with a third reviewer making the determination if consensus could not be reached.

Results

Literature Search

The electronic search returned 514 articles; an additional 68 were identified from a manual search of other sources; 121 duplicates were deleted. Of the 461 records screened on the basis of titles and abstracts, we included 316 in the full-text assessment. Of those, 149 were excluded on the basis of inclusion/exclusion criteria. CF were assessed in 148 studies (PRISMA flow diagram, Fig. 1). Of these, two studies reported the incidence/prevalence of CF [9, 10]. Owing to the large volume of studies identified (n = 43) that reported outcomes of local surgical treatments and intersphincteric ligation procedures, the current synthesis is limited to those published on CCF between January 1, 2015 and March 25, 2020 (n = 18). Details of the studies are included in Table 1. Results from combined surgical procedures (e.g., mucosal anal flap [MAF] combined with injection of platelet-rich plasma) were out-of-scope; however, fistulotomy and primary sphincteroplasty (FIPS) was included because the two techniques are essentially two steps of a single procedure. Criteria for what comprised a complex fistula were determined by each respective study author and are included in Table 1.

Fig. 1
figure 1

PRISMA flow diagram

Full size image
Table 1 Characteristics of included studies of local surgical procedures (n = 10) and intersphincteric ligation (n = 8) included in the SLR
Full size table

Risk of Bias Assessment

Of the 18 included outcomes studies, 17 were cohort studies or case series that met the review definition for cohort studies. Of these, seven studies were prospective [11,12,13,14,15,16,17] and 10 were retrospective [18,19,20,21,22,23,24,25,26,27]. One study was a retrospective cross-sectional study [28]. Two papers [20, 21] were judged as having a serious risk of bias and 16 as having a moderate risk of bias (Table 1).

Epidemiology of CF

Two studies from the UK estimated incidence or prevalence of CF; a population-based study using The Health Improvement Network (THIN) UK primary care database estimated the prevalence of CF in patients without Crohn’s disease at 1.35/10,000 patients in 2017 in the UK, down from 1.83/10,000 patients in 2014. The standardized prevalence of CF in the EU in 2017 was 1.39 (1.26–1.52) per 10,000. The authors suggest that the declining fistula prevalence could be an artifact of a decline in active patients in the database [9]. Another study examined the incidence of anal fistula among patients with a hospital admission for anal abscess in the Hospital Episode Statistics database, an administrative data set with almost complete capture of all hospital episodes in England since its inception in 1987. The authors reported that 52.6% (95% CI 51.6% to 53.0%) of patients without inflammatory bowel disease progressed from anorectal abscess to fistula over 12 months [10]. No studies were identified estimating the incidence or prevalence of CCF specifically.

Clinical Outcomes of Selected Surgical Procedures for CCF

The studies identified in this review described clinical outcomes of healing, recurrence, FI, and pain following local surgical procedures and intersphincteric ligation procedures for CCF. Local surgical procedures included MAF (n = 6), fistulectomy (n = 3), FIPS, (n = 1), and modified Parks’ technique (n = 1). No studies were identified performing lay open fistulotomy. Intersphincteric ligation procedures included LIFT or BioLIFT (n = 7) and TROPIS (n = 1). Outcomes are summarized in Tables 2 and 3.

Table 2 Selected outcomes of local surgical procedures for complex cryptoglandular fistulas
Full size table
Table 3 Selected outcomes of intersphincteric ligation procedures for complex cryptoglandular fistulas
Full size table

Fistula Healing (Surgical Success)

Fourteen of 18 studies (seven local surgical and seven intersphincteric ligation studies) reported on fistula healing, or “success” of the intervention. Definitions of these outcomes varied (see Tables 2 and 3); however, many authors defined healing as closure of the opening, healing of the wound, and absence of purulent discharge. Primary healing rates, or the healing rate after the initial intervention of interest without follow-up intervention, in the three studies of MAF reporting this outcome ranged from 65.0% at 1 year in a mix of 31 surgery-experienced and surgery-naïve patients to 86.9% after 6 months in 61 surgery-experienced patients [11, 20, 23]. In another study of 121 patients with a median duration of 74 months (range 8–148 months), patients underwent up to two additional MAFs, until a 100% healing rate was reached [22].

One study reported that 100% of 173 patients healed after fistulectomy, and most patients (168/173) healed within 3–4 weeks. Healing rates were reported for FIPS in one study [21], where 93.2% of 103 patients healed after a mean follow-up of 55.9 months. The authors of these two studies did not report whether the patients were surgically naïve or experienced. One study that assessed healing after modified Parks’ technique reported initial healing in 93.8% of 32 surgery-naïve or surgery-experienced patients with subsequent healing at 100% after a median of 12 months (range 4–24 months) [13].

For LIFT/BioLIFT procedures, primary healing occurred in a range of 57.1% after a median of 12 weeks in a mix of 28 surgery-experienced or surgery-naïve patients to 88% after a mean of 14.6 months (standard error 1.7 months) in a mix of 75 surgery-experienced or surgery-naïve patients [15, 19, 24,25,26]. Healing after initial TROPIS surgery was reached in 84.6% of 61 patients after a median of 9 months (range 6–21 months) in one available study. Including patients who underwent a second TROPIS procedure, the overall healing rate increased to 90.4% [16]. The authors did not report whether these patients were surgery-experienced or surgery-naïve.

Fistula Recurrence/Failure (Surgical Failure)

Seventeen papers reported on fistula recurrence or treatment failure and demonstrated a wide range of findings. The authors defined these outcomes in various ways (see Tables 2 and 3); some authors equated intervention “failure” with “recurrence,” and some reported results separately for these outcomes. Many authors defined recurrence as the clinical occurrence of the fistula, an abscess, or purulent discharge after recovery of the surgical wound within various time periods. In studies of the MAF procedure, recurrence occurred in 4.9–44.4% of patients [11, 12, 18, 20, 22, 23]. Boenicke et al. reported recurrence in three surgery-experienced patients (3/61, 4.9%), one each taking place at 9, 13, and 15 months [11]. Emile et al. reported recurrence in a mix of four surgery-experienced or surgery-naïve patients (4/9, 44.4%) within 1 year of their procedure [18]. Three studies [11, 18, 23] also reported separate failure or disruption of flap rates that ranged from 16% (5/31; three of which occurred within the first week) to 55.5% (5/9 within 1 year of the procedure) of patients [18].

Recurrence rates after fistulectomy ranged from 8.1% (at 1-year follow-up among 175 patients whose surgery experience was not reported) to 60.7% (after a median follow-up period of 26 months [range 2–118 months] among 28 surgery-experienced or surgery-naïve patients) [12, 14, 28]. One study of the modified Parks’ technique reported recurrence after a median of 12 months (range 4–24 months) follow-up in 6.3% of 32 surgery-experienced or surgery-naïve patients who had horseshoe fistula with supralevator extension [13]. The single study of FIPS did not report recurrence rates for patients with CCF [21].

Among studies reporting intersphincteric ligation procedures, recurrence occurred in a range of 7.5% (3 of 40 surgery-experienced and surgery-naïve patients after a mean of 14.2 months follow-up) [17] to 42.9% (12 of 29 surgery-experienced or surgery-naïve patients relapsed within 12 weeks) [15] of patients receiving LIFT/BioLIFT [15, 17, 24, 25, 27]. Treatment failure was experienced in 2.8% (2 of 71 surgery-experienced or surgery-naïve patients after 12 months) [25] to 16.1% (10 of 62 surgery-naïve patients after a median of 24.5 months; range 12–51 months) [26] of patients [19, 25, 26]. The single study of TROPIS did not report recurrence or failure rates [16].

Fecal Incontinence

Tables 2 and 3 indicate the scales and definitions of FI used in each paper. Two of the four studies reporting on FI following MAF procedures did so using the Wexner score, which ranges from 0 (perfect functionality) to 20 (complete incontinence) [29]. Boenicke et al. reported Wexner scores of 0.46 ± 0.97 points at patients’ 6-month follow-up [11]. Lee et al. reported a Wexner score of 0 in 77.8% of patients, 14.8% had a score of 1–5, and 7.4% had a score of 11–13 [20]. Podetta et al. reported FI using the Miller scoring system with range 0–18, with higher numbers representing more frequent incontinence-related symptoms [22, 30]. Of 32 patients who received a second mucosal flap, two patients reported incontinence symptoms. Of patients who received two additional MAF procedures, one patient reported rare gas incontinence after the first MAF and no change in incontinence scores with the subsequent procedures.

One study reported Wexner-identified incontinence in 18.4% of patients with CCF following the FIPS procedure [21]. In El-Said et al., postoperative new-onset minor FI (according to the Wexner score) was reported in 3% of patients following modified Parks’ technique. None of the studies of fistulectomy reported FI by surgery type and for patients with CCF [13].

Seven of eight studies of intersphincteric ligation procedures reported on the outcome of FI. In studies of LIFT and BioLIFT, five studies used Wexner scoring and three used patient-reported scales, including the Fecal Incontinence Quality of Life (FIQL) and the Fecal Incontinence Severity Index (FISI). FI was reported in 0% of patients in two studies [15, 26] using Wexner scoring, and in one study using a self-reported scale [19] (scale name not reported). Ye et al. reported no incontinence in patients postoperatively by Wexner score and FISI [27]. Schulze et al. reported increased incontinence in 1.3% of patients following LIFT [24]. Sun et al. reported improvement in Wexner scores for flatus incontinence after LIFT in 5.7% of patients, and significant improvements in lifestyle, coping, and depression domains of the FIQL [25].

In the one study that reported on TROPIS [16], the authors used the Vaizey incontinence score with a range of 0 (perfect continence) to 24 (complete incontinence) and reported mean scores of less than 1 with no significant change in scores pre- and postoperatively.

Pain

Few studies reported pain as a clinical outcome. The majority used a mix of clinician-reported and patient-reported scales and measured postoperative pain versus perianal pain specifically. One of the two studies of MAF procedures reported that 8.1% (5/61) of patients had experienced postoperative pain that was self-limiting and responsive to analgesics 30 days post-procedure [11]. The scale used in this study was not reported. The second study of MAF used the Numeric Rating Scale (NRS) (1 = no pain; 10 = worst pain imaginable). The mean score did not increase significantly postoperatively (mean score preoperatively 1.4 ± 0.6 vs 3 months postoperatively 1.2 ± 0.5) [23]. One study [21] reported on pain following FIPS and noted that no patients developed postoperative intractable pain after a mean of 55.9 months (range 12–143 months). Using the Short Form-36 Health Survey, version 2, where each item is scored on a range of 0–100 and higher scores indicate more favorable health states, El-Said et al. reported a preoperative mean pain score of 37.5 ± 9.3 and 6-month postoperative mean score of 65.1 ± 7.2 following modified Parks’ technique [13].

Only one study reported on perianal pain associated with intersphincteric ligation procedures. Perianal pain was experienced in 3% of patients following LIFT [24].

Discussion

To the best of our knowledge, this is the first SLR reporting the incidence/prevalence of CF and outcomes associated with local surgical and intersphincteric ligation procedures for CCF. CCF are more difficult to treat than simple CF, resulting in higher failure rates and functional disability [31]. Treatments that heal CCF and reduce recurrence could provide hope for patients experiencing the substantial physical and social impacts of this condition [6]. Limited real-world evidence exists on CCF, and there is a need to critically evaluate and assess existing epidemiological data. We therefore summarized outcomes of fistula healing, recurrence/failure, FI, and pain following specific CCF interventions.

The important finding from this SLR is that in studies of local surgical or intersphincteric ligation procedures for CCF, primary healing rates were 57.1–100.0%, indicating a moderate degree of initial success in these patients. Another critical finding was that recurrence and treatment failure were reported in 4.9–60.7% and 2.8–18.0% of patients, respectively. The imprecision of these estimates suggest a need for standardization in the reporting of outcomes to better assess the risks and benefits of individual treatment approaches. The limited number of publications that report on postoperative FI and postoperative pain suggests these outcomes are rare after the procedures highlighted in this study. Many of these studies were indicative of no long-term incontinence and no long-term postoperative pain.

Notable strengths of the current review include compliance with established guidelines for SLRs inclusion of a pre-specified protocol and search criteria. Selection, data extraction, and adjudication of risk of bias were done by two independent reviewers. The protocol was registered with PROSPERO to promote transparency and allow for future replication or updates and was conducted by two independent reviewers.

A major limitation of the current review is that it does not include studies published after March 25, 2020 to the present. New studies have been published in the past 3 years (e.g., for TROPIS, select newer publications include Li et al. 2022 [32], Huang et al. 2021 [33], Garg et al. 2021 [34], and Jayne et al. 2021 [35]) and it is unknown how this additional body of literature would impact any conclusions in the current manuscript. Future updated reviews in this area should be explored. Other limitations include at least a moderate risk of bias in all the included studies. Several studies were limited by single-center design with small sample sizes and short follow-up durations. Although this review was designed to capture a wide range of literature, it was limited to English-language studies from the past 5 years and hence might not be representative of the full body of published literature. Additionally, identified publications reported various study designs, follow-up durations, and definitions of key outcomes, making comparisons between studies infeasible.

Conclusion

Despite limitations, this SLR provides a unique critical summary of available data and highlights evidence gaps that can be addressed with further research. Specifically, there is a need for global observational studies on the incidence/prevalence of CF and CCF. Furthermore, the available literature lacks consistent approaches for assessing outcomes which could be used to facilitate comparison of treatment approaches. This SLR provides a comprehensive and critical summary of published epidemiology of CF and healing, recurrence/failure, incontinence, and pain outcomes from local surgical and intersphincteric ligation procedures for CCF. Success rates vary by surgery type, and differences in treatment indication, population, duration, or other aspects of study design prevent direct comparison. However, reported overall healing rates indicate the potential for relief from the substantial burden of CCF with low-to-modest rates of recurrence and rare reports of long-term postoperative incontinence or long-term postoperative pain.