Pilonidal sinus disease: a 25-year experience and long-term results of different surgical techniques

The incidence of pilonidal disease is increasing. The choice of surgical approach differs between surgeons and countries. With better understanding of the etiology of the disease, there is a shift toward more successful concepts of treatment. In many cases, management can be challenging owing to the number of previous failed operations. The aim of this retrospective single-center cohort study was to compare recurrence rates and postoperative wound complications between five treatment arms. A total of 299 patients who underwent surgery for pilonidal disease between November 1994 and May 2019 were included. Primary endpoint was time to recurrence, secondary endpoint was wound care complication rate. Median follow-up was 85.8 months in 286 patients. An overall recurrence rate of 16.1% was observed at 24 months, 21.4% at 60 months, and 47.4% at 303 months; 24 months postoperatively, there was a range from 10.5% for excision with primary midline closure to 30.0% for the Bascom I procedure. Recurrence in excision with primary midline closure was 71.8% 268 months postoperatively. No statistically significant differences were observed between the five groups (p = 0.54). The highest prevalence of wound complications (46.3%) was in excision with midline closure. Cox regression showed that previous pilonidal operations are an independent prognostic factor for developing recurrence (p = 0.006). Multivariate logistic regression revealed that previous pilonidal operations have a significant predictive value for developing postoperative wound complications (odds ratio = 4.04, 95% confidence interval [1.61–10.18]; p = 0.003). In order to improve surgical outcomes, emphasis should be given to adoption of techniques with high success rates.


Introduction
The preferences of surgical approach in the therapy of pilonidal sinus disease differ slightly between surgeons and countries. A recent meta-analysis showed that recurrence rates depend on surgical approach and follow-up time [1]. The impact of geographical factors such as genetic mechanisms responsible for hair growth and hair thickness, healthcare settings, and socioeconomic factors on recurrence rates are revealed in the retrospective study, suggesting that some approaches have extraordinarily good outcomes in specific countries; however, flap techniques remain superior [2]. Postoperative complications 240 25 Years of Pilonidal Sinus Treatment K original article and recurrence significantly impair the quality of life of patients, especially when symptoms are severe. In decades past, excision with primary midline closure or lay open were considered the gold standard; however, these methods are associated with high complication and recurrence rates. In many cases, management can be challenging owing to the number of previous failed operations. In addition, when nonhealing or extended perianal wounds after excisional surgery occur, technical modifications of the subsequent flap procedure need to be done. According to the recently published German guidelines, excision with midline closure should not be performed at all [3].
With better understanding of the etiology of the disease, there is a shift toward more successful treatment concepts. Bascom showed that early epidermal changes are amplified by further deep tissue disruption from moisture, anerobic conditions, hair, and bacteria [4]. Remarkably, in the analysis of hair samples from pilonidal sinus cavities, short hair fragments with rootless sharp cut ends were found [5]. Furthermore, testing of hair from three different body regions suggests that occipital hair is regularly present in pilonidal sinus nests [6]. To improve outcomes, the origin of the disease must be understood. Differential diagnoses such as anal fistula can be excluded by magnetic resonance imaging [7], and more successful methods should be applied. These methods should include strategies to reshape and flatten the gluteal cleft and relocate the incision off the midline. Surprisingly, certain limitations and acceptance of newer techniques appear to impact the choice of the method.
We were interested in differences between surgical groups in regard to recurrence rates and wound care complications. We hypothesized that risk factors such as previous pilonidal operations, abscess incisions, duration of disease, smoking status, and body mass index (BMI) are associated with higher recurrence rates as well as with a higher incidence of wound care complications.

Materials and methods
This study was designed as a retrospective singlecenter cohort study and conducted at the University Clinic of General Surgery of the Medical University of Vienna. All patients gave informed consent. The Ethics Committee of the Medical University of Vienna approved the study (EC reference: 2019/2020).

Eligibility
Adult patients who underwent surgery for both primary and recurrent pilonidal sinus disease from November 1994 to May 2019 were included in the study. Exclusion criteria were patients under the age of 18 years, patients with emergency surgery for pilonidal abscess, simultaneous hidradenitis suppu-rativa, sinus pilonidalis squamous cell carcinoma, and fistulizing rectal cancer. Surgical techniques with less than 10 cases and repetitive procedures in case of recurrent disease were eliminated from further analysis. In the few patients undergoing the same procedure twice, only the first procedure was included in order to maintain statistical independence of observations within treatment procedures and thus validity of pvalues.
We reviewed the routinely recorded data of the Vienna General Hospital AKIM and collected BMI, patient age, sex, smoking status, duration of disease, number of previous pilonidal operations, abscess incisions, and postoperative wound care complications. Five treatment arms were compared: excision with primary midline closure, excision with marsupialization, Karydakis flap, Dufourmentel flap, and Bascom I procedure. Three groups of different techniques, i.e., split thickness skin graft (n = 2), excision and lay open (n = 6), and excision with negative pressure wound therapy (n = 1), were eliminated from analyses due to a small sample size. Eight repetitive surgeries in patients with recurrence were also eliminated: excision and primary midline closure (n = 1), excision and marsupialization (n = 7).

Outcomes
The first endpoint of the study is the time to recurrence and the evaluation of risk factors for time to recurrence in five investigated subgroups. A recurrence was defined as new disease occurring after the wound was completely healed. The secondary outcome was postoperative wound care complication rate among the five analyzed treatment arms. Complications were divided into three groups: mild, severe, and none. Examples of mild complications were surgical site infection (SSI), local superficial dehiscence, and seroma. Severe complications were dehiscence or abscess with drain placement, or revisional surgery requiring a hospital stay. For statistical analyses, only the occurrence of complications in the five treatment arms was of interest.

Follow-up
Disease recurrence was assessed either by a visit at the outpatient ward or by telephone interview. Data on postoperative wound care complications were collected from our internal patient database. To minimize recall bias and to avoid any confusion in terms of interpretation and clinical manifestation of recurrence, especially in patients with lower health literacy, we conducted the calls using a standardized questionnaire.

Statistical analysis
Categorical data were described using absolute and relative frequencies and group differences were tested by chi-square test or Fisher's exact test in case of sparse data. For continuous data, the median, minimum, and maximum were calculated due to nonnormally distributed data. Group differences for continuous data were calculated by Kruskal-Wallis test. Kaplan-Meier curves were estimated to visualize time to recurrence and log-rank test was performed to assess differences between groups. Cox proportional hazard regression was applied to model prognostic factors for time to recurrence. All p-values are twosided and p ≤ 0.05 is considered statistically significant. All calculations were performed using SAS (Version 9.4, SAS Institute Inc. ©, Cary, NC, USA). This manuscript has been designed in accordance with the STROBE guidelines [8].  time, a flap technique, in particular the Dufourmentel flap, was adopted. Later, between 2008 and 2011, this was the most frequently performed procedure. The first off-midline procedure, namely the Karydakis flap, was carried out in 2008 and has been on the rise since 2012. A minimally invasive Bascom I procedure did not find broader use, except for in 2016. A graphical display of yearly spectra of performed techniques is shown in Fig. 3.

Results
The analyzed patient population demonstrated a representation of mostly younger males, whilst the proportion of patients younger than 30 years varied between 55.4 and 72.1% among the cohorts. There was no significant difference in median BMI between the five treatment arms, which ranged from 24.  Table 1.

Recurrence-free survival
Kaplan-Meier estimator displays the recurrence-free survival for the five surgical techniques as shown in Fig. 4. From the 299 patients, 286 were included in the analysis (95.7%) and 13 lacked follow-up data. We see widely varying follow-up times among different procedures resulting from the year when the technique was adopted for the first time. No statistically significant differences were observed (p = 0.54). Median follow-up was 85.8 months in 286 patients and an overall recurrence rate of 16.1% was observed at 24 months, 21.4% at 60 months, and 47.4% at 303 months; 24 months postoperatively, there was a range from 10.5% for excision with primary midline closure to 30.0% for the Bascom I procedure. Unsurprisingly, the recurrence rate in excision with primary midline closure was as high as 24.7% at 60 months and 71.8% at last follow-up 268 months postoperatively. Patients undergoing excision and marsupialization showed the fourth lowest recurrence rate at 24 months (13.7%) and the lowest recurrence rate at 60 months ( Table 2. From 286 patients with follow-up data, 76 patients' reports lacked demographic data, leaving 210 patients to be included in a multivariate Cox proportional hazards regression model. It showed that previous pilonidal surgeries are an independent prognostic factor for developing recurrence (p = 0.006). Non-smokers were more likely to experience the event (p = 0.002). None of the other factors demonstrated a significant predictive value for developing recurrence (Table 3).  Table 4.

Wound care complications
From 284 patients with wound care complication data, 70 patients' reports lacked demographic data, leaving 214 patients to be included in multivariate logistic regression. This revealed that previous failed pilonidal operations in the patient history have significant predictive value for developing postoperative wound care complications (odds ratio = 4.04, 95% confidence interval [1.61-10.18]; p = 0.003) as shown in Table 3.

Discussion
Our retrospective study compared five different surgical approaches in the therapy of pilonidal disease in terms of recurrence rates and wound care complications. Recurrence rates vary between different methods depending on follow-up time. We found out that previous failed pilonidal operations are an independent prognostic factor for recurrence and development of wound complications.
We identified the following limitations in our study: not all techniques were included in the study, as certain procedures were used only on rare occasions. Although the group size of the most recent technique is still small, we included this group for explorative purposes and to give the most complete overview of all applied techniques. Information on disease recurrence was retrieved in some patients through recall during telephone interview, which may be another source of bias. Patients may not be able to recall accurately, and certain details may be omitted. To minimize this bias, we created a standardized scheme for conducting calls.
Finally, it is important to consider that an essential component of high-quality studies, which enables comparison of results of different surgical groups, is a universally adopted and well-recognized classification of the disease [9].
To date, many studies have reported outstanding results using off-midline methods for both primary and recurrent pilonidal disease [10][11][12][13][14]. Importantly, two decades ago, the deep, moist, and bacteriafriendly gluteal cleft with anerobic conditions was  [4]. Therefore, flattening of the natal cleft and bringing the incision off midline play a crucial role in the management of the disease. Although radical excision was earlier considered as a goal of treatment [15], it now can no longer be advocated. This paradigm shift represents a pivotal change for the choice of treatment strategy. Opponents of flap techniques often highlight their own subjective positive experience with excisional procedures. The reasons for the resistance to change could be comfort with the status quo and temporary incompetence caused by a shallow learning curve when learning without an experienced mentor. In our study, the group of excision with primary midline closure demonstrated the second highest recurrence rate 60 months postoperatively and the highest wound complication rate, at 24.7% and 46.3%, respectively. When comparing 24-month follow-up and the last follow-up at 268 months, the recurrence rates vary between 10.5 and 71.8%. Excisional methods do not address the deep gluteal cleft and when they fail, patients end up with a distorted gluteal cleft and nonhealing wounds, which is likely to have a profoundly negative impact on further treatment. A retrospective study on 124 patients with recurrent pilonidal disease comparing two treatment arms, namely excisional method and Karydakis flap, showed a 43% 1-year recurrence rate in the first group and 3% in the second group (p < 0.0001) [9]. Consequently, patients with recurrent pilonidal disease would have the most to gain from off-midline procedures.  [11,19,20]. One of the key principles of construction of advancement flaps is complete wound lateralization. If the final suture line or its caudal portion ends up in the midline, some additional skin needs to be excised [12,21,22]. Although, the minimally invasive procedures are associated with higher recurrence rates, their benefits are well known. From a cost standpoint, they are 246 [23]. A German retrospective study on 153 patients reported a recurrence rate of 28% at 24 months postoperatively and median time interval to recurrence of 2.8 months [24]. Their study also revealed a significant predictive value of disease duration > 6 months and BMI > 25.0 kg/m 2 for development of recurrence [23]. Similar results were reported in the study by Bascom, with mean time to complete healing of 3 weeks [25]. The last investigated group, excision with marsupialization showed a recurrence rate as high as 13.7% at 24 months, 18.1% at 60 months, and an overall wound care complication rate of 23.9%. Two of the major concerns in this method are prolonged healing and significant postoperative pain.

Conclusion
To achieve better results, we need to deliver the right treatment at the right time. The emphasis should be given to adoption of techniques with high success rates. In order to improve surgical outcomes, it is imperative to respect the key principles of their construction. Well-designed prospective studies focusing on both treatment strategies-minimally invasive techniques as well as flap procedures-are necessary.
Author Contribution All authors contributed to the study conception and design. All authors read and approved the final manuscript.