Abstract
Objective:
Orocutaneous fistula (OCF) after reconstruction for oral cavity resection can lead to prolonged hospitalization and adjuvant treatment delay. Few studies have examined factors leading to OCF after oral cavity resection. Primary objective: evaluate overall incidence and factors associated with OCF after oral cavity reconstruction.
Data Sources:
Scopus 1960—database was searched for terms: “orocutaneous fistula,” “oro cutaneous fistula,” “oral cutaneous fistula,” “orocervical fistula,” “oral cavity salivary fistula.”
Review Methods:
English language studies with >5 patients undergoing reconstruction after oral cavity cancer resection were included. About 1057 records initially screened; 214 full texts assessed; 78 full-texts included. PRISMA guidelines were followed, and MINORS criteria used to assess risk of bias. Data were pooled using random-effects model. Primary outcome was OCF incidence. Meta-analysis to determine the effect of preoperative radiation on OCF conducted on 12 eligible studies. Pre-collection hypothesis was that prior radiation therapy is associated with increased OCF incidence. Post-collection analyses: free versus pedicled flaps; mandible-sparing versus segmental mandibulectomy.
Results:
Seventy-eight studies were included in meta-analysis of overall OCF incidence. Pooled effect size showed overall incidence of OCF to be 7.71% (95% CI, 6.28%-9.13%) among 5400 patients. Meta-analysis of preoperative radiation therapy on OCF showed a pooled odds ratio of 1.68 (95% CI, 0.93-3.06). OCF incidence was similar between patients undergoing free versus pedicled reconstruction, or segmental mandibulectomy versus mandible-sparing resection.
Conclusion:
Orocutaneous fistula after oral cavity resection has significant incidence and clinical impact. Risk of OCF persists despite advances in reconstructive options; there is a trend toward higher risk after prior radiation.
Introduction
Oral cavity cancer is the most common non-cutaneous site of head and neck cancer, with approximately 32 000 new cases in the United States annually. 1 The approach to oral cavity cancer has long been upfront surgery followed by appropriate adjuvant therapy, and survival has improved over the last 50 years as surgical and adjuvant therapies have evolved. 2 Oral cavity tumor extirpation can have enormous impact on patient speech, swallowing, airway, and esthetics; reconstructive techniques have advanced over the last several decades, but challenges remain. En bloc resection of locally advanced tumors commonly leads to intraoperative communication between the oral cavity and the neck, with the potential for persistent orocutaneous or orocervical fistula after surgery. Fistulas following head and neck reconstructions often lead to delays in adjuvant treatment, which can worsen overall oncologic outcomes. 3 Worse yet, fistulas may lead to exposure of the great vessels in the neck and carotid blowout.
While incidence of and factors leading to pharyngocutaneous fistula after total laryngectomy have been extensively reported in the literature, 4 the incidence of orocutaneous fistula (OCF) after oral cavity resection and reconstruction is not as well studied. Few studies have analyzed factors that predispose patients to OCF, and surgeons must largely rely on data from other anatomic subsites to counsel patients regarding perioperative risk of OCF.
Here, we performed a systematic review of the literature regarding orocutaneous fistula after oral cavity cancer resection and reconstruction, and we performed a meta-analysis with the goals of identifying the overall incidence OCF and patient or surgical factors that may be predispose to OCF.
Methods
Information Sources, Search Strategy, and Study Selection
A comprehensive search strategy was designed to identify studies investigating or reporting data on fistula after oral cavity resection and reconstruction. Scopus database from 1960—was searched in July 2020. Key words searched were: “orocutaneous fistula,” “oro cutaneous fistula,” “oral cutaneous fistula,” orocervical fistula,” and “oral cavity salivary fistula.” Study references were also searched to include relevant studies. Preferred reporting items for systematic reviews and meta analyses (PRISMA) guidelines were used throughout. 5
Study inclusion criteria were English (or available in English translation) studies on adult human patients. Both prospective and retrospective studies were eligible. Studies reporting resection and reconstruction for non-oral cavity sites were included if discrete data were available on oral cavity patients within the larger group. Studies were required to include at least 75% of patients undergoing resection for oral cavity cancer, in order to avoid confounding factors such as reconstruction for trauma or osteoradionecrosis. Studies were excluded for having 5 or fewer patients, to avoid case studies. Abstracts without full text articles available were not included. Abstracts were assessed by 1 author (P.T.) for full text review, and full text review as well as reference review was conducted by 2 authors (P.T. and T.G.).
Data Extraction, Synthesis, and Statistical Analysis
Risk of bias was assessed for each full text by 2 authors (P.T. and T.G.) using the methodological index for non-randomized studies (MINORS) criteria. 6 Within each study, available data on indication for surgery, patient history of preoperative radiation therapy, type of reconstruction (regional vs free flap), and surgery requiring segmental mandibulectomy were collected. The primary outcome variable was the presence of orocutaneous fistula during available follow-up. Orocutaneous fistula was defined as communication between the oral cavity and cervical skin in the perioperative period after an attempt at reconstructive closure to prevent this communication. Secondary outcome, when applicable, was the necessity of return to operating room for management of fistula.
A random effects meta-analysis was performed to evaluate the proportion of patients with orocutaneous fistula for: all patients; patients with free versus regional flap reconstruction; and patients with segmental mandibulectomy versus not. These analyses were performed using R 4.0.2 and the meta package (R foundation for Statistical Computing, Vienna, Austria). For studies that included detailed data on preoperative radiation therapy and OCF, odds ratios for OCF as a binary outcome were calculated as described by Tierney et al. 7 Odds ratios were then used in a generic inverse variance, random effects meta-analysis. This analysis was performed using RevMan 5.1 (Nordic Cochrane Centre, Copenhagen, Denmark). I2 statistic was used to evaluate statistical heterogeneity.
Hypothesis
Our hypothesis prior to data collection was that preoperative radiation therapy would be associated with increased incidence of OCF. Following data collection, we added hypotheses: segmental mandibulectomy would be associated with OCF; free flaps would not have an advantage over regional flaps in OCF outcomes; incidence of OCF would decrease over time.
Results
Systematic Review
Initial literature review identified 1189 candidate abstracts. Scan for duplicates identified 131 duplicates, leading to 1057 unique abstracts. About 1057 abstracts were screened, and 843 were excluded for failure to meet inclusion criteria. The remaining 214 full text articles were assessed for eligibility, with 78 studies included for systematic review and meta-analysis (Figure 1).

Flowchart diagram of search and study selection.
Three studies were prospective trials of techniques to reduce orocutaneous fistula: 1 trial of salivary gland botulinum toxin injection 8 ; 1 trial of 2-octyl cyanoacrylate 9 ; and 1 single-arm prospective trial of using a skin island pectoralis major myocutaneous flap for oral cavity reconstruction in a rural Indian hospital. 10 The remaining 75 studies were retrospective reviews.11-85 Study characteristics are included in Table 1.
Study Characteristics Included in Pooled Meta-Analysis of Fistula Rates.
Abbreviations: DP, delopectoral; FAMM, facial artery myomucosal; Marg mand, marginal mandibulectomy; OC, oral cavity; PMMC, pectoralis major myocutaneous; SCCa, squamous cell carcinoma; Seg mand, segmental mandibulectomy; STSG, split thickness skin graft.
The trial of salivary gland botulinum toxin injection took place preoperatively in a non-blinded, single-arm study of 43 patients undergoing oral cavity resection and free flap reconstruction; 2 patients (4.6%) experienced orocutaneous fistula, without a control or comparison group. 8 The trial of 2-octyl cyanoacrylate took place intraoperatively in a non-blinded, single-arm study of 46 patients undergoing oral cavity resection and free flap reconstruction, and study patients were compared to historical matched controls; 8 (17.4%) patients in the experimental group developed orocutaneous fistula compared to 7 (15.2%) of matched controls, and the authors concluded that no improvement in fistula rate was achieved with this intervention. 9
Risk of bias as assessed using the MINORS criteria showed a low risk of bias for most studies, with consecutive series of disease site or reconstructive technique being common (Supplemental Table S1).
Meta-Analysis
Meta-analysis of 5400 patients who underwent oral cavity resection and reconstruction showed a pooled risk of 7.71% (95% confidence interval [CI], 6.28%-9.13%) for orocutaneous fistula (Figure 2). The statistical heterogeneity was high at I2 = 70%. Meta-analysis of 1331 patients from 28 evaluable studies8,9,20,23,28,30,33-36,38-40,42,47,48,50,53-55,57,58,61,64,74,85 showed a pooled risk of OCF of 8.55% (95% CI, 5.71%-11.4%, I2 = 71.8%) among patients undergoing free flap reconstruction (Figure 3A), compared to 8.02% (95% CI, 5.27%-10.8%, I2 = 72.7%) among 1449 patients undergoing regional flap reconstruction from 34 evaluable studies10,13,14,16,19,21,24-26,32,41,44,45,51,56,59,60,62,63,65-67,69,71,73,75,76,78,80,82-84 (Figure 3B). Pooled risk of fistula was 6.74% (95% CI, 4.50%-8.97%, I2 = 38.6) among 1133 patients undergoing segmental mandibulectomy in 21 studies,13,16,20,22,23,28,33,35,39,40,42,47,48,50,51,53,54,57,69,75,77 compared to 5.48% (95% CI, 2.41%-8.45%, I2 = 54.6%) among 556 patients without segmental mandibulectomy in 24 studies.10,12,30,31,34,38,41,43-45,52,54-56,59,65,67,68,71,73,74,78,80,82

Pooled meta-analysis of rates of orocutaneous fistula among 78 studies including 5400 patients undergoing oral cavity resection and reconstruction.

Pooled meta-analysis of rate of orocutaneous fistula among patients undergoing free flap (A) or regional flap (B) reconstruction after oral cavity resection.
A total of 591 patients from 12 studies9,17,18,20,23,33,41,47,49,63,65,70 were included in a generic inverse variance, random effects meta-analysis of preoperative radiation therapy, and risk of OCF. This analysis showed a pooled odds ratio of 1.68 (95% CI, 0.93-3.06, P = .09), a trend toward preoperative radiation therapy association with orocutaneous fistula (Figure 4).

Meta-analysis of odds ratios for orocutaneous fistula formation after prior radiation or not. Overall odds ratio is 1.68 (95% CI, 0.93-3.06).
The pooled risk of OCF was investigated over time by grouping studies into 4 temporal groups: 1972 to 1990 (n = 11 studies, 690 patients), 1991 to 2000 (n = 15 studies, 662 patients), 2001 to 2010 (n = 17 studies, 802 patients), and 2011 to 2020 (n = 35 studies, 3246 patients). Pooled risk was highest among the earliest group, but not significantly higher than later groups (Figure 5).

Pooled meta-analysis of rates of orocutaneous fistula over time. Error bars represent standard error.
Thirty-three studies included information on fistula management. Among 258 evaluable patients with orocutaneous fistula, 108 (41.9%) required operative management for closure, with the remaining 150 managed conservatively without need for return to the operating room.
Discussion
Orocutaneous fistula after oral cavity resection and reconstruction is a surgical complication that ranges in severity from mildly problematic to acutely life-threatening. While not all OCFs may have significant long-term effects on patients, certainly all of them are felt as adverse events by both patient and surgeon. Here, we demonstrate a pooled incidence of OCF of 7.7% over the last 5 decades. OCF, therefore, impacts a significant number of patients, and will certainly be encountered by surgeons caring for patients with oral cavity cancer.
Unexpectedly, the presence of preoperative radiation therapy did not reach statistical significance for association with OCF, though it did trend toward significance. This stands in contrast to the well-established risk of pharyngocutaneous fistula among patients undergoing total laryngectomy after radiation therapy. One explanation may be that oral cavity radiation is often given only in the adjuvant setting after prior oral cavity resection and neck dissection; therefore, previously radiated patients may not have had extensive neck surgery at the time of their salvage resections, leading to a lower incidence of OCF in some series. Most studies included in this meta-analysis did not provide patient-specific data to be able to determine whether post-radiated patients also underwent neck dissection or exploration, but this would be an interesting prospective study question for future investigations.
Still, because of radiation therapy’s known effects on wound healing and tissue fibrosis, these authors will continue to counsel patients preoperatively of an increased risk for wound complications in the post-radiated setting.
Our meta-analysis demonstrated a similar incidence of OCF whether free or regional flaps were used for reconstruction after oral cavity resection; this result is not surprising. The decisions for oral cavity reconstruction are complex, and depend heavily on the defect, the patient, and the surgeon. As nearly all the included studies were retrospective, a selection decision, and therefore a selection bias, has already occurred before patients are analyzed. Such decisions occur in a “black box” in retrospective studies. Just as the patients in this meta-analysis are a heterogeneous population, the decisions for optimal reconstruction are heterogeneous. The similar incidence between free and regional flaps likely reflects thoughtful decisions regarding patient and surgeon factors, and highly reliable nature of both free and regional flap reconstructions.
While free or regional flaps were found to be equally effective at preventing orocutaneous fistula in this study, we did not investigate other important outcomes such as oral intake, trach dependence, and donor site morbidity. All of these should be considered when making decisions about individual patient reconstructions.
One of our preoperative hypotheses was that patients undergoing segmental mandibulectomy would have an increased incidence of OCF, because of the increased complexity of reconstruction. However, even though many patients with segmental mandibulectomy underwent soft-tissue-only regional reconstructions, the OCF rate remained low. This finding is likely due to the fact that the current study is focused on short-term perioperative OCF; patients undergoing non-bony reconstructions after segmental mandibulectomy are susceptible to long-term complications of plate fracture and exposure, 86 but those complications are not captured in the current study.
There was a trend toward decreasing rates of fistula over time, which may correlate with the introduction of increasingly advanced regional and free flap techniques; however, the OCF rate even in the most recent 10 years of publications remains significant. Risk of OCF should still be discussed with patients preoperatively, and reconstructive decisions should account for its possibility.
Among evaluable patients with fistula, 42% required operative intervention to resolve the fistula. While patient-specific data on adjuvant therapy are lacking in this meta-analysis, it is reasonable to expect that an unexpected return to the operating room may delay the initiation of adjuvant treatment, as has been demonstrated in prior studies. 3 This underscores the impact of orocutaneous fistula on patients.
Limitations
Nearly all of the included studies are retrospective, and come with limitations. Often, an included study described a specific technique for closure of oral cavity defects; investigators are unlikely to report series of non-successful reconstructions or reconstructive techniques. Such patients are likely under-represented in the literature, and therefore in this systematic review and meta-analysis.
Importantly, this study made an attempt to exclude patients undergoing reconstruction for osteoradionecrosis (ORN) by including studies with no more than 25% of patient with non-cancer resections. Patients with ORN often have fistulas or active infections at the time of reconstruction that might make post-operative wound complications more likely. While overall this study may include some patients with ORN, extending findings from this study to inform expectations for patients with ORN may be misleading.
Additionally, factors not analyzed here may predispose to fistula formation, such as flap failures, preoperative nutrition status, and hardware use. The inconsistency of reporting these factors in most studies analyzed here makes such post-hoc analyses difficult, though the factors themselves may be important.
Likewise, the current study represents an “all-comers” approach to orocutaneous fistula; in reality, there is likely a spectrum of risk for fistula after oral cavity resection that is not captured in a meta-analysis of many different types of oral cavity resections and reconstructions. Additionally, patient-level data are not available on all studies here, and may include a small number of patients whose resections did not place them at risk of fistula formation.
Conclusions
Orocutaneous fistula occurs in a small but significant percentage of patients undergoing oral cavity resection and reconstruction. There is a trend toward increased risk of OCF among patients with prior radiation therapy.
Supplemental Material
sj-pdf-1-aor-10.1177_00034894211047463 – Supplemental material for Orocutaneous Fistula After Oral Cavity Resection and Reconstruction: Systematic Review and Meta-Analysis
Supplemental material, sj-pdf-1-aor-10.1177_00034894211047463 for Orocutaneous Fistula After Oral Cavity Resection and Reconstruction: Systematic Review and Meta-Analysis by Patrick Tassone, Tabitha Galloway, Laura Dooley and Robert Zitsch in Annals of Otology, Rhinology & Laryngology
Footnotes
Declaration of Conflicting Interests
The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding
The author(s) received no financial support for the research, authorship, and/or publication of this article.
Supplemental Material
Supplemental material for this article is available online.
References
Supplementary Material
Please find the following supplemental material available below.
For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.
For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.
