Exploring the use and efficacy of 5-fluorouracil in the management of odontogenic keratocysts: a systematic review and meta-analysis.

Introduction
Odontogenic keratocyst (OKC) accounts for about 10% of all odontogenic cysts and has long been a topic of clinical interest due to its unique characteristics and behavior [1]. OKCs are locally aggressive cystic lesions which commonly affects the posterior mandible and may present as a painless swelling, bone expansion, or an incidental radiographic finding [1, 2].
In 2005, the World Health Organization (WHO) reclassified OKC as a tumor under the term keratocystic odontogenic tumor (KCOT) [2, 3], based on its aggressive clinical course, high recurrence rate after simple enucleation [4], and distinct histopathological features [5]. These include high proliferative activity and expression of tumor-related markers such as PTCH1 mutations [6, 7], similar to those seen in basal cell carcinomas. However, in the 2017 WHO/IARC classification, the OKC was reclassified back as a cystic lesion [3].
Treatment of OKC remains controversial [8]. The lesion’s thin, friable lining and the presence of daughter cysts beyond its visible margins contribute significantly to its recurrence [6]. Conservative methods such as enucleation and marsupialization are associated with high recurrence rates which could reach 60% in case of simple enucleation [6]. Aggressive procedures, on the other hand, reduce recurrence but increase morbidity. To overcome this balance between recurrence and morbidity, various chemical and physical adjuvants have been used after cyst removal, including Carnoy’s solution, modified Carnoy’s solution (MCS), cryotherapy, and topical 5-Fluorouracil (5-FU) [9–11]. Despite its effectiveness, Carnoy’s solution is no longer widely used due to the potential carcinogenic effects of chloroform, which led to its ban by the United States Food and Drug Administration (FDA) [12, 13].
5-FU is an antimetabolite that inhibits thymidylate synthase and interferes with DNA and RNA synthesis, leading to selective cytotoxicity in rapidly proliferating epithelial cells [7, 14, 15]. Although originally developed as an anticancer agent [14], its topical use has been successfully extended to locally aggressive but non-malignant lesions. OKCs epithelium shows high proliferative activity and overexpression of markers such as Ki-67 and p53 [14]. Immunohistochemical studies have also demonstrated expression of both thymidylate synthase and thymidine phosphorylase, especially in inflamed areas, enzymes known to influence tissue responsiveness to 5-FU [14]. Inflammation-induced upregulation of thymidine phosphorylase enhances the local conversion of 5-FU to its active metabolite, fluorodeoxyuridine monophosphate, thereby promoting selective destruction of residual cystic epithelium [7, 14].
Given these properties, 5-FU has emerged as a potential topical agent capable of achieving local control with minimal tissue damage. This systematic review and meta-analysis aims to summarize current evidence regarding the effectiveness and safety of 5-FU as an adjunctive treatment for OKC, comparing it with conventional adjuvant methods.

Materials and methods

Eligibility criteria
This systematic review and meta-analysis followed the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) 2020 guidelines [16]. The study protocol was prospectively registered in PROSPERO (CRD42024578698).
Studies were included if they met the following criteria: (1) involved patients diagnosed with OKC (2) evaluated 5-FU as an adjunctive treatment, either alone or in combination with surgery, and (3) reported outcomes such as recurrence, nerve injury, or postoperative complications.
Case reports, cohort studies, and randomized controlled trials were eligible. Reviews, editorials, letters, animal studies, in vitro research and studies without extractable clinical data were excluded. There were no language or publication date restrictions.

Information sources and search strategy
A comprehensive literature search was conducted in PubMed/Medline, Embase, Scopus, and Google Scholar, covering all publications up to August 1 st, 2024. The search strategy followed PRISMA 2020 recommendations and used combinations of the following terms: “fluorouracil” or “5-fluorouracil” and “odontogenic cyst” or “odontogenic tumor” or “keratocyst,” targeting their presence in titles or abstracts.
The complete search strings for each database are presented in Appendix A in Supplementary Material. The search also included gray literature and reference lists of relevant studies.
Two reviewers (R.M. and S.K.) independently screened titles and abstracts to identify eligible studies. Full-text articles were then retrieved and assessed according to the inclusion and exclusion criteria.

Data extraction
Data extracted from the selected studies included:

Study design, year of publication and data collection period.

Treatment received by the control group, if found.

Inclusion and exclusion criteria.

Patients’ demographics (age, gender).

Lesion characteristics (location, size, locularity, association with a tooth, history of previous recurrences).

Treatment protocols (application of 5-FU, surgical methods).

Follow-up details (duration, recurrence rates, adverse outcomes).

All extracted data were organized into a database for comprehensive assessment and comparison.

Risk of bias
Risk of bias was independently assessed by using two validated tools depending on the study design. For cohort studies, the Newcastle–Ottawa Scale (NOS) was applied, which evaluates three domains: [17]

Selection (up to 4 stars): A star was given for representativeness of cohorts, a non-exposed cohort drawn from the same population, ascertainment of exposure, and confirmation that the outcome was absent at baseline.

Comparability (up to 2 stars): A star was assigned if the study controlled for the most important factor and an additional star if it controlled for any additional factor.

Outcome (up to 3 stars): A star was awarded for independent blind assessment of the outcome, sufficient length of follow-up (at least 24 months), and completeness of follow-up (≥ 90%).

For randomized controlled trials, the Cochrane Risk of Bias 2 (ROB-2) tool was used [18], evaluating five domains: (1) randomization process, (2) deviations from intended interventions, (3) missing outcome data, (4) measurement of the outcome, and (5) selection of the reported result. Each domain was rated as “low risk,” “some concerns,” or “high risk,” and the overall risk was determined according to the Cochrane guidelines. The ROBVIS online tool was utilized to generate traffic light plots.

Data synthesis
Recurrence rates and nerve damage frequencies were pooled and reported with their corresponding 95% confidence intervals (CIs). Heterogeneity across the included studies was assessed using Cochran’s chi-square test and the I² statistic, with values above 75% indicating substantial heterogeneity. The inverse variance weighting method was used for pooling the data. Since the I² values were below 75% for all analyses, a fixed-effect model was applied to calculate the pooled risk difference. In cases where at least one study reported zero events, a continuity correction of 0.5 was added to all studies. All statistical analyses were conducted using SPSS software (IBM SPSS Statistics for Windows, Version 29, IBM Corp., Armonk, NY, USA, 2023). A p-value of < 0.05 was considered statistically significant for all statistical analyses.

Results

Study selection
The initial search identified 249 records (PubMed: 37, Scopus: 118, Embase: 53, Google Scholar: 31). No additional records were identified through gray literature sources. After removing 92 duplicates, 157 titles and abstracts were screened, and 32 full-text articles were assessed for eligibility.
Fourteen studies met the inclusion criteria and were included in the final analysis, including 4 randomized controlled trials, 5 retrospective cohort studies, and 5 case reports. The study selection process is illustrated in the PRISMA flow diagram (Fig. 1). A detailed list of excluded studies and reasons for exclusion is provided in Appendix B in Supplementary Material.

Study characteristics
The 14 included studies comprised 282 OKC lesions, 156 of them treated with enucleation and peripheral ostectomy followed by 5-FU, 95 with MCS, 7 with segmental resection and 24 underwent enucleation and peripheral ostectomy without chemical adjuvant therapy.
Among the comparative studies, four evaluated the treatment efficacy of 5-FU versus MCS [7, 14, 19, 20], one compared 5-FU, MCS, and segmental resection [21], and another one compared 5-FU with enucleation and peripheral ostectomy [22]. The remaining studies did not include a control group.
Study characteristics and treatment comparisons are summarized in Tables 1 and 2.

5-FU VS MCS
The meta-analysis included five studies comparing 86 lesions treated with 5-FU and 95 lesions treated with MCS. All cases in the 5-FU group underwent enucleation and peripheral ostectomy, followed by the application of 5% 5-FU cream on gauze, which was left in the surgical cavity for 24 h.
Among the five studies, three excluded patients with recurrent lesions (Mishra N, et al. 2023 [19], Wanve S, et al. 2023 [20], and Ledderhof N, et al. 2017 [14]). Lone P, et al. 2020 [21] did not specify whether the lesions were recurrent, and only one study reported previous recurrence data. Caminiti et al. [7] noted that 9 of the 34 lesions in the 5-FU group and 4 of the 36 lesions in the MCS group were recurrent (p = 0.13). No significant differences were found between groups in gender (p = 0.092), lesion location (p = 0.095), recurrence history (p = 0.149), or locularity (p = 0.985) (Table 3).

No recurrences occurred in any 5-FU treated cases, while 23 recurrences (24.21%) were reported in the MCS group (Fig. 2). The pooled effect significantly favored 5-FU (p < 0.001, effect size = −0.199, 95% CI: −0.283 to −0.114). Moderate heterogeneity was observed (I² = 57.3%), suggesting consistent findings across studies.

Postoperative paresthesia occurred in 18.82% cases treated with 5-FU (n = 16), with 56.25% showing complete recovery (n = 9). In contrast, 37.89% of MCS-treated cases (n = 36) experienced paresthesia, and 41.67% developed permanent sensory deficits (Fig. 3). The pooled effect again favored 5-FU compared to MCS (p = 0.012, effect size = −0.158, 95% CI: −0.281 to −0.035), with moderate heterogeneity (I² = 37.4%, p = 0.172).

For permanent paresthesia, Ledderhof N, et al. [14] reported a trend favoring 5-FU, but the pooled effect was not statistically significant (effect size = −0.061, 95% CI: −0.140 to 0.018, p = 0.129). (Fig. 4)

Although funnel plots were used to evaluate potential publication bias, the small number of included studies limited their interpretability. Nevertheless, the plots appeared symmetric, suggesting no major publication bias. The direction and magnitude of the pooled estimates remained consistent, supporting the robustness of results despite the limited dataset.
Finally, during the follow-up period, no systemic or other local side effects of 5-FU were reported.

5-FU VS segmental resection
A single study by Lone P, et al. [21] compared 5-FU treatment with segmental mandibulectomy. Eleven patients received 5-FU, and seven underwent segmental mandibulectomy. No recurrences were reported in either group (p > 0.99). However, only one temporary paresthesia (9.09%) occurred in the 5-FU group during 2–4 years of follow-up, while all patients treated with segmental mandibulectomy (100%) experienced permanent paresthesia during 4–9 years of follow-up (p < 0.001). Functional and cosmetic deformities occurred exclusively in the segmental mandibulectomy group (100%).

5-FU VS enucleation and peripheral ostectomy
Eid M, et al. [22] compared outcomes between enucleation and peripheral ostectomy alone and the same procedure followed by 5-FU application. Twenty-four were treated with 5-FU, and 24 underwent surgery without adjuvant therapy. Both groups followed for 24 months.
Clinical evaluations were performed at 1 week, 1 month, and every 3 months thereafter, with CT scans were taken at 24-hour, 12-month, and 24-months.
All lesions were located in the mandible (Tables 1 and 2). No recurrences were observed in either group, and all cases demonstrated uneventful healing. Bone density was significantly higher in the 5-FU group at both 12 and 24 months (p = 0.006 and p < 0.001, respectively). Temporary paresthesia occurred in two 5-FU-treated patients, compared with three cases (including one permanent) in the control group, with no significant difference between groups in term of recurrence and paresthesia (p > 0.99).

Uncontrolled studies
Eight uncontrolled studies reported 46 lesions treated with 5-FU in 43 patients [23–30].
The mean patient age was 27.77 years ± 6.55, with 15 females (34.88%) and 28 males (65.12%). One patient (2.33%) had Gorlin syndrome [27]. Twenty-two lesions (47.82%) were symptomatic, 35 (76.09%) were located in the mandible, and 41.3% were multilocular. Four lesions (8.7%) had a history of recurrence.
Of the 46 lesions, 43 underwent enucleation and peripheral ostectomy before 5-FU application, while three were treated in two stages: decompression followed by enucleation and peripheral ostectomy with 5-FU.
Regarding the method of 5-FU application:15 lesions (32.61%) were treated with topical 5% 5-FU cream applied to all exposed bony cavity surfaces, including any soft tissue with bony dehiscence or fenestrations, and followed by absorbable gelatin sponge (Gelfoam) soaked with 5-FU and left place postoperatively [24].

One lesion (2.17%) treated with 5% 5-FU solution for 5 min, followed by irrigation [25].

30 lesions (65.22%) treated with 5% 5-FU cream on gauze for 24 h [23, 26–30]. In one study of Barua and colleagues, weekly iodoform gauze was applied for 6 months [26].

No recurrences were observed (0%). Paresthesia was reported in six cases (13.04%), with only one permanent case (2.17%). The mean follow-up duration was 27.16 months ± 18.1. No systemic or serious local complications were reported.

Risk of bias
All four RCTs included in the review demonstrated an overall unclear risk of bias [19–22], as summarized in Fig. 5. Among the cohort studies, Essa E, et al. 2023 [23] and Casino A, et al. 2023 [24], each scored six stars on the NOS, while Caminiti M, et al. 2021 [7] and Ledderhof N, et al. 2017 [14] each scored eight stars. Detailed NOS assessments are shown in Fig. 5.

Discussion
Managing OKCs has long been debated [11], primarily concerning the balance between reducing recurrence and minimizing surgical morbidity [31]. Recurrence may result from incomplete removal, residual satellite cysts, or epithelial remnants left behind during surgery [6]. Additional contributing factors include lesion size, location, infection, tooth involvement, and syndromic association [8, 32, 33]. However, these factors are inconsistently reported, and many studies exclude syndromic patients from their analyses [13]. Such variations in study populations complicated cross-study comparisons and make accurate evaluation of recurrence rates challenging.
This systematic review provides a comprehensive assessment of topical 5-FU as an adjunctive treatment for OKC, comparing it with established adjuvant options. Various treatment modalities have been investigated over the years [8–10, 12], each with variable outcomes and adverse effects. The main objectives of this review were to assess the recurrence rates and adverse events associated with 5-FU therapy. By compiling data from 282 OKCs, this represents the largest review to date on the use of 5-FU in OKC management.
Most included studies (n = 12) used a consistent protocol, applying 5% 5-FU cream or solution on gauze left in surgical cavity for 24 h [7, 14, 19–23, 26–30]. Two studies applied slightly different techniques: one applied the cream directly to bony surfaces with a cotton swab, followed by the placement of a Gelfoam sponge coated with 5-FU that was not removed after application [21]; another used a 5% 5-FU solution applied for five minutes followed by irrigation [22]. While clinical outcomes were similar across techniques, the precise mechanism by which 5-FU acts on residual cystic linings and satellite cysts remains unclear.
Ledderhof et al. [14] identified biomarkers related to 5-FU sensitivity, reporting reduced thymidylate synthase expression, an enzyme involved in DNA damage, indicating increased susceptibility of OKC epithelium to 5-FU. Moreover, elevated levels of thymidine phosphorylase and dihydropyrimidine dehydrogenase were detected in inflamed OKC linings, potentially enhancing local activation and efficacy of 5-FU.
Current evidence suggest that 5-FU is an effective and safe adjunctive treatment for OKC, with no reported recurrences across all included studies. Compared to segmental resection, 5-FU treatment results in markedly reduced morbidity [21]. It also appears to be a favorable alternative to MCS, offering lower paresthesia rates and fewer complications. Although 5-FU requires a longer application period than MCS [14. 24], it is simple to handle, widely available, and easy to store and apply in either cream or solution form [7, 25].
When compared with peripheral ostectomy alone, 5-FU demonstrated an additional benefit: although recurrence rates were comparable, bone density was significantly higher in the 5-FU group, suggesting enhanced bone regeneration [22].
It’s important to recognize that OKC can recur years after treatment. Most recurrences occur within the first 60 months [34, 35], emphasizing the need for long-term follow-up of at least five years [8]. In the reviewed studies, follow-up durations ranged from 6 to 64 months, with most below five years. Consequently, the absence of recurrence in 5-FU–treated cases should be interpreted cautiously, and extended monitoring is essential to confirm its long-term effectiveness.
Unlike systemic 5-FU therapy, which is associated with well-known systemic toxicities [36], topical application has primarily been linked to mild, localized reactions such as irritation, sloughing, and ulceration, typically observed in dermatologic use [37]. Systemically, 5-FU is contraindicated in individuals with dihydropyrimidine dehydrogenase (DPD) gene polymorphisms due to potential sever toxicity [38]. However, the degree of systemic absorption following topical use in bone cavities, and its potential effects on the oral mucosa, remain unclear [39]. Therefore, while not contraindicated in patients with DPD polymorphisms, caution is recommended until more data on systemic absorption becomes available.
Despite being the most comprehensive review to date on 5-FU in OKC management, several limitations should be acknowledged. None of the included studies directly compared 5-FU with Carnoy’s solution, historically considered the gold standard adjuvant. Only five studies qualified for meta-analysis, including four randomized controlled trials. Additionally, several studies lacked details such as locularity, cyst size, association with teeth, and prior recurrence history.
In summary, OKC remains a challenging benign lesion requiring treatment strategies that minimize recurrence while reducing patient morbidity. As summarized in the evidence table, 5-FU emerges as a highly promising alternative adjuvant therapy, offering excellent local control, minimal side effects, and straightforward clinical use. However, given the small sample sizes and potential bias in some studies, these results should be interpreted with caution. Future large-scale, well-designed randomized trials with long-term follow-up are necessary to confirm these findings and establish 5-FU as a standard adjunctive therapy for OKC.

Supplementary Information