METHOD
Identify relevant studies and study selection A systematic literature search was conducted on the 23th of December 2019, in the PubMed, Cochrane and EMBASE databases to identify articles comparing outcome data from different surgical techniques for SBC removal (syntax provided in Appendix 1). No restriction regarding publication data and language were applied. This study is reported according to the Preferred Reporting Items for Systematic Review and Meta-analysis (PRISMA) statement (14).
Study selectionTwo authors (S.M., R.M.) independently screened the retrieved articles on title and abstract using pre-defined inclusion and exclusion criteria (Fig. 1). The selected articles were read in full-text by the aforementioned two authors. Only studies comparing different surgical techniques in one cohort of patients were selected. The reference lists of the selected articles were reviewed for a cross-reference check to select relevant studies that were not identified in the initial search. All authors were involved in the discussion leading to final article inclusion. Disagreement between authors was solved by discussion.
Data extractionData from all included studies was independently extracted by the authors. The predefined included data contained: year of publication, number of included patients (total and patients with SBC specifically), occurrence of bilateral anomalies, pre-operative SBC infection, gender, age at surgery, pre-operative imaging with: computed tomography, magnetic resonance imaging or ultrasound, operation technique, operating time, incision type and length, follow up duration, recurrence and complication rates, and scar satisfaction.
Quality assessment Four authors (S.M., H.B., E.v.d.V. and M.v.d.A.) critically appraised selected articles regarding directness of evidence (DoE) and risk of bias (RoB) (Table 1). We assessed the DoE using three criteria: 1) domain: studies comparing surgical techniques for SBC removal 2) determinant: clear description of the selected surgical technique and 3) surgical outcome: report on recurrence and complication rates. To perform RoB assessment on the selected studies, authors applied an appraisal tool derived from the Cochrane risk of bias Tool (15) . Each criterion was rated satisfactory (●), partly satisfactory (○), or unsatisfactory (-) (explanatory legend of Table 1). Overall DoE was rated as high (H), moderate (M) or low (L). Only studies with a high directness of evidence were selected for final inclusion. No studies were excluded based on RoB, adhering to the Grading of Recommendations Assessment, Development and Evaluation (GRADE) system (16).
Data synthesis Pooling of data was considered in case of homogeneity between studies (if I2 was <50%)(17).
RESULTS An overview of the study selection process and reasons for exclusion is presented in figure 1. (flowchart). Four articles had a high DoE and were included for final selection. This resulted in the inclusion of the treatment of 140 cysts. No patients with bilateral cysts were included. These four studies (6,9,10,12) contained two randomized controlled trails (RCTs) and two prospective trials. For an overview of the included surgical techniques see Figure 2 and the description in Appendix 2. Two authors (9,10) were contacted to provide additional information regarding the applied incision type, however, no additional information was provided. The inclusion dates of the patient cohorts of Chen et al. 2012 (10) and Chen et al. 2014 (9) did not overlap and therefore, both studies were included in the current review. Meta-analysis was not performed in this review due to heterogeneity regarding: baseline characteristics, study design and applied surgical techniques.
Data extraction: Table 2 shows the data extraction of four included studies that directly compared outcomes between conventional surgery and modern removal techniques in patients presenting with unilateral SBCs. All patients from these studies underwent pre-operative imaging using CT-scan or ultrasound scanning and pre-operative fine needle cytology to confirm the diagnosis (data not shown). Chen et al. (10) compared SBC removal results between conventional, curvilinear, cervical incisions along a natural skin crease (3-4 cm below the lower border of the mandible) to the endoscopic RAHI technique. Adult patients were randomly assigned between both techniques (Table 2). None of the included patients suffered from a pre-operative SBC infection. No recurrence occurred during a follow up of at least six months. There was no significant difference in operating time between both techniques; however, there was a significantly (p = < .001) higher scar satisfaction rate in the RAHI group. Scar satisfaction was measured six months postoperatively using a visual analog scale ranging from 0-10. Chen et al. (9) compared SBC removal using a curvilinear cervical incision along a natural skin crease (4-5 cm below the lower border of the mandible) to an endoscopic approach of the lateral neck using two randomly assigned patient groups. Twenty adult patients were assigned to the conventional cervical incision, whereas 21 patients were assigned to the endoscopic lateral neck approach. None of the included patients suffered from a pre-operative SBC infection. No recurrence occurred during a follow up of at least six months. Although no significant difference in operating time was reported between both groups, incision length and scar satisfaction did significantly (p < .05) differ in favor of the endoscopic technique. Scar satisfaction was measured six months postoperatively using a visual analog scale ranging from 0-10. Ahn et al . (12) compared SBC removal outcomes between a conventional approach (by making a curvilinear incision directly over the anomaly) and an open RAHI approach in a prospective case control study. Thirteen adult patients were operated by the open RAHI approach while 17 adult patients underwent a (conventional) cervical incision. Ahn et al. reported a pre-operative SBC infection rate of 30.8% in the patients who were operated using the open RAHI technique. No recurrence occurred during a follow up of three months. Of the patients who underwent conventional surgery, 11.8% suffered from a postoperative hematoma or seroma, compared to 7.7% of the patients who underwent open RAHI surgery (non-significant difference ). Only patients of the open RAHI group suffered from postoperative neurological damage that spontaneously resolved (23.1%). The retro-auricular approach entailed significantly longer operating time (p = .019), however, resulted in significantly higher scar satisfaction (p= < .001). Scar satisfaction was measured three months postoperatively using a visual analog scale ranging from 0-10. Iaremenko et al. (6) compared SBC removal outcomes between a conventional approach (by making a skin incision 2.0-2.5 centimeter below the lower border of the mandible) and an endoscopic occipital approach using a controlled study design. The latter technique is comparable to the endoscopic RAHI technique of Chen et al. (10) from a surgical perspective. Twenty-two adult patients were operated by the occipital endoscopic approach, while 22 adult patients underwent a (conventional) cervical incision. No recurrence occurred during a follow up of six months. Of the conventional group, 4.5% developed a hematoma and 4.5% developed temporary neurological damage. In the occipital approach group 27.3% reported temporary pain and difficulty at sideward arm raise. The endoscopic approach resulted in a significantly higher scar satisfaction (p = .05), but took significantly longer operating time (p = .05). Scar satisfaction was measured six months postoperatively using the criteria ‘emotional component’ of the ‘Attitude to health’ questionnaire. Since no recurrence was reported in any of the included studies, no data regarding revision surgery were retrieved.
DISCUSSION Summary of findings: In this systematic literature review, we compared the clinical outcome (complication and recurrence rates and scar satisfaction) of SBC removal between conventional surgery and less invasive removal techniques (endoscopic surgery or open/endoscopic RAHI). Only four studies (6,9,10,12) were identified that compared the conventional technique with newer techniques within one patient cohort. All of these included studies are of low quality due to: short follow up periods, small patient groups and a study design prone to bias due to both selection criteria (e.g. no inclusion of pre-operatively infected cysts) and lack of blinding. Since evidence is scarce, it remains difficult to provide evidence-based surgical treatment advice. Results demonstrate that surgical treatment of SBC results in a complication rate ranging between 0 to 27.3%. The most reported complications were temporary earlobe hypoesthesia in patients who underwent endoscopic or (endoscopic) RAHI surgery (most likely due to perioperative greater auricular nerve manipulation), temporary pain and difficulty of sideward arm raise (most likely resulting from spinal accessory nerve manipulation). No permanent neurological damage was reported. Surgical treatment provides a definitive treatment with no reported recurrence using either one of the techniques. Studies that compared both techniques within the same adult patient cohort demonstrated that both the (endoscopic) RAHI approach as well as other endoscopic techniques resulted in higher scar satisfaction. Therefore, available evidence demonstrates that application of less invasive SBC removal techniques to treat uninfected second branchial cleft cysts results in relatively higher, temporary complication rates, however, with a significantly higher scar satisfaction. An interesting result, since the operating area is in a prominently visible location in a patient population containing young adults.
Two included studies (9,10) excluded patients presenting with fistulas and sinuses, pre-operatively infected SBCs and patients who underwent prior neck surgery or radiotherapy. Only Ahn et al. reported on open RAHI treatment of patients with pre-operatively infected SBCs. Although 30.8% of these patients suffered from a pre-operative infection, no relatively higher complication rate was reported for this population compared to the cervical incision group. Iaremenko et al. did not report whether any pre-operatively infected SBC were included in their study cohort.
Comparison with other studies and techniques: This is the first literature review reporting on the clinical outcome of SBC removal comparing different surgical techniques. Cohort studies(9,11,13) investigating either open RAHI or endoscopic RAHI found similar results: absence of recurrence in combination with low complication rates, with an average follow up of (at least) 6, 14.5 and 42 months respectively. The only reported complications in open RAHI surgery were temporary hypoesthesia of the earlobe and hypertrophic scars.
In previous studies, conventional second branchial arch anomaly removal techniques have been studied intensively. Appendix 3 shows an overview of these studies that were identified through the same literature search. This Appendix also includes patients presenting with fistulas and sinuses (mostly children). Precise data regarding the distribution of (included) cysts, sinuses and fistulas, side of the anomalies, description of the used surgical technique or duration of follow up is lacking in most studies (Appendix 3). Only retrospective studies were identified with a complication rate ranging between 0 and 32% and a recurrence rate ranging between 0 and 4.9%. These recurrence rates were reported when follow up lasted till four or even ten years (18,19). Therefore, the follow-up of the included studies in this review (ranging between 3-24 months) could be too short to identify recurrence following surgery. Long-term recurrence rates are of major importance because recurrence of disease will cause high morbidity and can make revision surgery complex. Furthermore, this short follow up could also affect the reported scar satisfaction since three to six months after surgery the final scar result may not be visible yet.
Multiple authors (20–22) refer to Deane et al. (23) regarding recurrence percentages: Deane et al described the results of a retrospective study (performed between 1954 and 1975) including 274 patients with branchial cleft remnants below the mandible. The surgical technique is described as a local excision of the lesion (plus tract when necessary). The reported recurrence rate was 2.7% in patients without a prior infection, 14% in patients with a pre-operative infection and 21.2% in patients who underwent revision surgery. The average follow up duration was 12.4 years (1-22 years). No distinction between second, third and fourth branchial anomalies was made.
Quality of evidence and potential biases The overall quality of the included studies was low (IIb -IV regarding the Oxford Centre for Evidence-Based Medicine guidelines ): only two studies used a RCT to compare the clinical outcome between surgical techniques. In these RCTs, selection bias could not be ruled out since lack of blinding. The quality of evidence regarding SBAA removal was mostly affected by small patient cohorts resulting in Type II error (i.e., failing to reject a false null hypothesis), short follow up, unclear inclusion criteria and selective reporting.