Clinical Outcomes of Medical Management Options for Chronic Anal Fissures in a Long-Term Follow-up: Systematic Review and Meta-Analysis

Anal fissures are linear ulcers extending from the pectinate line to the anal margin affecting the anoderm and covering the lower half of the internal anal sphincter [1]. They are usually diagnosed in young adults, although it can be seen at any age with a similar incidence in both sexes [2‒4]. Fissures persisting for longer than 4 weeks, or recurrent fissures, are generally defined as chronic anal fissures (CAF) [5]. The diagnosis is entirely clinical, with the presence of visible sphincter fibers at the fissure base, anal papillae, sentinel piles, and indurated margins; however, some clinicians define CAF with the presence of signs or symptoms [6]. The pathogenesis remains controversial [7]; two factors are considered to be the most important: local ischemia and hypertonicity of the internal anal sphincter [8‒10]. Currently, there is not a standard treatment for CAF. Historically, surgical treatment has been the main approach, with healing rates up to 90%; however, high rates of incontinence ranging from 0% to 27% have been described, with persistent incontinence of up to 9% [11‒15]. Other drawbacks of surgery include bleeding, pain, and infections [16, 17].

In recent years, topical and oral treatments have been an important therapy option. Glyceryl trinitrate, calcium channel blockers, and botulinum toxin being the most studied drugs in various systematic reviews. However, varying response rates have raised questions about efficacy [18‒20]. One of the main causes of the variability of response to treatment is the combination of oral and topical treatments in the same analysis, as well as the different follow-up periods ranging from 4 weeks to 2 years. It has been reported that these treatment options have adverse effects, leading to the discontinuation and therapeutic failure, which according to several studies have been reported in 33% at 9-month follow-up and up to 67% at 29-month follow-up [21, 22].

Due to clinical heterogenicity of previous meta-analyses, an objective evaluation of the true cure rates of topical treatment is complex, considering that CAF typically has a cyclical history of intermittent healing and recurrence. About 35% will eventually heal at least temporarily without medical intervention; it is important to know the response to treatment after an appropriate follow-up [18]. The aim of this manuscript was to evaluate the effectiveness of local treatments with calcium antagonists, nitroglycerin, and botulinum toxin in the healing of CAF in adult patients, as well as to identify the rate of recurrence. A systematic review was performed comparing topical treatment options.

The systemic review and meta-analysis were developed in accordance with the guidelines and recommendations of The PRISMA Statement for Reporting Systematic Reviews and Meta-Analyses of Studies That Evaluate Health Care Interventions: Explanation and Elaboration[23]. This study was carried out according to a predefined protocol, and it was registered in PROSPERO, an international database of prospectively registered systematic reviews (Registration number: CRD42020120386) [24]. MEDLINE, EMBASE, Cochrane Library, and Web of Science bibliographic databases were searched for randomized controlled trials in adults that compared the effectiveness of different medical treatment modalities in patients with CAF.

Reviewer team used a combination of medical descriptors (MeSH or DeCS, depending on the database) and the following textual terms: “adult,” “middle-aged,” “aged,” “fissure in anus,” “anal fissure” “calcium channel blockers,” botulinum toxins,” “nitroglycerin,” “diltiazem,” “nifedipine,” “wound healing,” “non-recurrence.” The search in Pubmed was done as follows: (“Adult”(Mesh) AND “Fissure in Anus”(Mesh)) AND ((“calcium channel blockers”(Mesh) OR “botulinum toxins”(Mesh) OR “nitroglycerin”(Mesh) OR “diltiazem”(Mesh) OR “nifedipine”(Mesh))) AND (“Wound Healing”(Mesh) OR “non-recurrence”)). In addition, we queried databases of the Virtual Health Library for Latin America and the Caribbean: BIREME and LILACS using the terms “treatment” and “Chronic anal fissure.” The literature review was performed between September 2017 and August 2020. We contacted authors to clarify incomplete data but we did not receive any answer.

Two reviewers (J.P.E. and M.C.E.) independently screened and selected the articles for inclusion. In all cases, the research was restricted to articles published in English, Spanish, or Portuguese. Publications, such as brief reports, letters to the editor, or editorials were not included. To be included, publications had to have a clinical definition of CAF with at least one of the following signs or symptoms: visible sphincter fibers at the base of the fissure, anal papillae, sentinel piles, and indurated margins. The symptoms had to be chronic for at least 4 weeks and the treatment option had to be medical interventions, studies with surgical treatment, oral treatments, or combinations of these were excluded. All of this with the aim of find evidence regarding the best local treatment option for patients with a diagnosis of CAF.

The data of the studies were extracted in an Excel spreadsheet. General study data, clinical characteristics of participants, chronicity of the fissures, diagnostic methods (clinical, imaging, and laboratories), and different treatment strategies were included. Three outcomes were evaluated: cure rate, recurrence of clinical symptoms, and adverse events. The information was obtained according to what was defined by each author.

The assessment of methodological quality was do it using the Cochrane Collaboration tool according to these domains: random sequence generation, allocation concealment, blinding of participants and staff, blinding of assessors, incomplete outcome data, and finally selective report.

Pooled hazard ratio estimates were calculated for diltiazem, nifedipine, and botulinum toxin. The risk ratio was defined as the simple ratio between the number of cases over the total population exposed once, divided by the number of cases in the nonexposed population. For papers that not reporting the number of cases, we multiplied the percentage of cases by the total population. No statistical corrections of incidence or risk ratios were considered. Random- and fixed-effects models were used to calculate the combined outcome of the binary data. Publication bias was evaluated by Begg’s and Egger’s test, and graphically by funnel plots. To evaluate heterogeneity, I², and χ² were used. The software package Revman 5.3^® was used in the statistical analysis.

A total of 189 records were reviewed for inclusion, after applying the selection criteria, 17 articles were selected [26‒41], 12 evaluated the use of calcium channel blockers [31‒43], and 5 evaluated botulinum toxin [26‒30] (Fig. 1). The general characteristics of the selected articles, including chronicity of symptoms, patient ages, treatment modalities, outcomes, and type of adverse events, are presented in Table 1. The methodological evaluation showed serious limitations, for instance, the inadequate description of allocation concealment and blinding of patients and assessors (Fig. 2).

Diltiazem and nifedipine were used as therapeutic strategies [31‒38]; diltiazem at 2% compared to glyceryl trinitrate ointment or with glyceryl stearate was evaluated in 8 studies [31‒38] and nifedipine was used in concentrations of 0.3% [39] and 0.5% [40, 41] in 3 studies.

Of the 5 studies that used botulinum toxin as an intervention, two studies used glyceryl trinitrate ointment (0.2%) as a comparator [27, 28]; the other 3 studies used placebo injection and isosorbide dinitrate ointment [26], isosorbide dinitrate in spray [29], and finally, a botulinum toxin type A injection and glyceryl trinitrate [27, 30].

The results of the evaluation of diltiazem showed a superior effect when compared to glycerin (RR = 1.16 [95% CI = 1.05–1.30]; I² = 18%; NNT = 12 [95% CI = 6–132]) and nifedipine compared to lidocaine (RR: 4.53 [95% CI = 2.99–6.86]; I² = 28%; NNT = 2 [95% CI = 2–3]) (Fig. 3. Healing outcomes. Comparison 3.1, 3.2).

Botulinum toxin showed no superiority when compared to glycerin (RR = 0.81 [95% CI = 0.02–29.36]; I² = 93%; NNT = 250 [95% CI = 6-∞]) or to isosorbide dinitrate (RR = 1.45 [95% CI = 0.32–6.54]; I² = 85%; NNT = 250 [95% CI = 6-∞]). There is high heterogeneity between the included studies (Fig. 3. Comparison 3.3.1, 3.3.2).

When recurrence was evaluated there was a tendency to decrease recurrences in subjects treated with diltiazem (RR = 0.92 [95% CI = 0.38–2.24]; I² = 36%; NNT = 237); and also an improvement in recurrence was observed with the use of nifedipine compared with lidocaine (RR = 0.18 [95% CI = 0.08–0.44]; NNT = 237) (Fig. 4. Comparison 4.1, 4.2).

Recurrence was lower in the group treated with botulinum toxin; however, no significant differences were found (botulinum toxin vs. isosorbide trinitrate (RR = 0.53 [95% CI = 0.15–1.85]; I² = 0%; NNT = 237) (Fig. 4. Comparison 4.3).

Adverse events were less likely to occur when treated with diltiazem compared to glyceryl trinitrate (RR: 0.13 [95% CI = 0.04–0.042]; I² = 87%; NNT = 237) No differences were found in patients treated with nifedipine when compared with lidocaine (RR: 0.80 [95% CI = 0.30–2.16]) (Fig. 5. Comparison 5.1, 5.2).

The presentation of adverse events was lower in patients treated with botulinum toxin than in patients who received glycerol trinitrate (botulinum toxin vs. other RR = 0.88 [95% CI = 0.47–1.67]) (Fig. 5. Comparison 5.3).

No publication bias was observed (Begg’s test p = 0.181; Egger’s test p = 0.6). There is the presence of atypical data, but the majority is within the band that represents the confidence interval for the model.

The treatment of anal fissures remains challenging. In the modern world, it has moved from surgical to medical modalities because of the risk of incontinence and infection associated with surgery and other complications [32]. The available common treatment options, such as nitroglycerin ointment relax the sphincter muscle; however, side effects such as severe headaches limit its administration [42]. Therefore, calcium medications like calcium channel blockers and topical diltiazem were introduced [43, 44]. Nonetheless, some controversy remains about the evidence in favor or against the available treatment alternatives; a study of long-term outcomes and follow-up could resolve this evaluating effectiveness and adverse events, aim of this review, and meta-analysis.

This study included 17 articles after an exhaustive search in the main medical databases; these had varied sample sizes according to the comparison group. Remarkably, the number of calcium channel blockers studies was higher (n = 12) than botulinum toxin studies (n = 5). Despite that, calcium channel blockers showed greater efficacy in healing CAF, showing a better outcome when using nifedipine, than when using other treatments (glyceryl trinitrate or stearate ointment, stool softeners, lidocaine ointment 1.5%, and hydrocortisone acetate ointment 1%). Botulinum toxin compared with other topical treatments showed no superiority in the cure of CAF, as demonstrated by other studies [11, 27, 45]. The results of this study indicate that the use of botulinum toxin as a primary agent in the treatment of anal fissures is not recommended, instead, it could offer real benefits as a second-line treatment. Nevertheless, as proposed by Festen et al. [26], it can be considered as a strategy for patients presenting adverse events to topical treatments or unable to comply with the application regime.

In terms of heterogenicity, this study evaluated the variation in results according to the frequency of adverse effects, revealing no significant differences in the pooled analysis; however, the reports showed significant heterogenicity in the RRs. It is worth mentioning that one of the main causes of therapeutic failure is the halting of the medication due to suspension because of the presentation of adverse events [46, 47], but we evidenciate that the form of reporting is not clear and many studies do not describe this within the objective of the study. The research included the largest databases of indexed articles, most of which yielded results meeting the requirements of the study. Based on the results of the review, it is fair to suggest that the included studies reasonably reflect common therapeutic alternatives for dealing with CAF. When we contacted the authors for missing data, we received no response; therefore, the RRs were calculated as the simple proportion of numbers of all cases of general correspondents by groups. Furthermore, the relevant figures can be found in all included studies.

Given the variations in the results, the use of combined RRs from fixed-effects models can be called into question; however, in view of the total number of studies included in each analytical subgroup, differences in the included items and treatments provided, random-effects models presented in our study should also be used with caution. It is also important to point out that the studies had methodological limitations that can artificially modify the effect of the interventions [28, 32]. Another point to consider in future studies is the control in the diets of the patients within the medical treatment [48].

This meta-analysis documented that the use of calcium channel blockers to heal CAF is more effective compared with other treatments in a long-term follow-up. Further studies are required to understand the underlying factors related to the healing process and better clinical prognosis of CAF, including large local or regional multicenter randomized clinical trials.

An ethics statement is not applicable because this study is based exclusively on published literature. In our institution before 2021, approval for meta-analysis was not necessary and this was built throughout 2019–2020, however, the manuscript protocol was approved by the clinical studies and clinical epidemiology subdirection of the institution. Patient consent was not required as this study was based on publicly available data as it is a systematic review.

The authors declare that they have no conflicts of interest that may affect the results of this study.

The authors did not participate in any open calls or received any financial support from their institution or any other entity during the development of the study. Therefore, the authors did not receive any research funding, nor do they currently receive any research funding, so there are no conflicting interests.

Fernando Sierra: methodological reviewer and thematic expert. José Antonio De la Hoz: reviewer of the manuscript and helps in the writing and interpretation of the data. Juan Pablo Espinosa: literature review and search. José Moreno Montoya: analysis of collected data. Mariana Vásquez Roldan and Erika D. Perez Riveros: drafting of the manuscript, collection and systematization of information, submission of article.

We confirm that all the information present in the manuscript is stored in the redcap repository managed by the institution and when the journal or any interested party requests it, it will be delivered at its discretion. Further inquiries can be directed to the corresponding author.