Antimetabolitos como complemento de la dacriocistorrinostomía para la obstrucción del conducto nasolagrimal
Resumen
Antecedentes
La obstrucción del conducto nasolagrimal (OCNL) es una afección que provoca un lagrimeo excesivo (epífora) o una infección del saco nasolagrimal (dacriocistitis). La etiología de la OCNL adquirida es multifactorial y no se comprende del todo. La dacriocistorrinostomía (DCR) es la corrección quirúrgica de la OCNL, cuyo objetivo es establecer una nueva vía de drenaje entre el saco lagrimal y la nariz. El éxito de la DCR es variable; la causa más común de fracaso es la fibrosis y la estenosis del ostium quirúrgico. Se ha demostrado que los antimetabolitos como la mitomicina‐C (MMC) y el 5‐fluorouracilo (5‐FU) son seguros y efectivos para reducir la fibrosis y mejorar los resultados clínicos en otros contextos de cirugía oftálmica (por ejemplo, en la cirugía de glaucoma y de córnea). Se ha estudiado la aplicación de antimetabolitos en el momento de la DCR, pero aún no se conoce la utilidad de estos tratamientos.
Objetivos
Objetivo primario: Determinar si el tratamiento adyuvante con antimetabolitos mejora el éxito funcional en el contexto de la DCR en comparación con la DCR sola.
Objetivos secundarios: Determinar si el éxito anatómico de la DCR aumenta con el uso de antimetabolitos, y si el ostium quirúrgico es mayor en los participantes tratados con antimetabolitos.
Métodos de búsqueda
Se realizaron búsquedas en el Registro Cochrane de Ensayos Controlados (Cochrane Register for Controlled Trials, CENTRAL) (que contiene el Registro de Ensayos del Grupo Cochrane de Trastornos de los Ojos y la Visión, Cochrane Eye and Vision Trials Register) (2019, número 9); Ovid MEDLINE; Embase.com; PubMed; LILACS (Latin American and Caribbean Health Sciences Literature Database); ClinicalTrials.gov; y en la International Clinical Trials Registry Platform (ICTRP) de la Organización Mundial de la Salud (OMS). No se aplicó ninguna restricción de fecha o idioma en las búsquedas electrónicas. Se buscó por última vez en las bases de datos electrónicas el 6 de septiembre 2019.
Criterios de selección
Solo se incluyeron ensayos controlados aleatorizados. Los estudios elegibles fueron los que compararon la administración de antimetabolitos en cualquier dosis y concentración versus placebo u otro tratamiento activo en participantes con OCNL sometidos a DCR primaria y a una nueva cirugía. Solo se incluyeron estudios que habían reclutado a adultos a partir de los 18 años de edad. También se incluyeron estudios que utilizaron la intubación con silicona como parte del procedimiento de DCR.
Obtención y análisis de los datos
Se utilizaron los procedimientos metodológicos estándar previstos por Cochrane. Dos autores de la revisión examinaron de forma independiente los resultados de la búsqueda, evaluaron el riesgo de sesgo y extrajeron los datos de los estudios incluidos mediante un formulario electrónico de recopilación de datos.
Resultados principales
Se incluyeron 31 estudios en la revisión, de los cuales 23 (1309 participantes) proporcionaron datos relacionados con los resultados primarios y secundarios. Muchos de los 23 estudios evaluaron el éxito funcional, mientras que otros también evaluaron los resultados secundarios del éxito anatómico o el tamaño del ostium, o ambos.
Características de los estudios
Las características de los participantes variaron a través de los estudios, y la edad de los participantes osciló entre los 30 y los 70 años. Los participantes fueron predominantemente mujeres. Estos datos demográficos corresponden a los que se ven más frecuentemente afectados por la obstrucción del conducto nasolagrimal. Casi todos los estudios utilizaron MMC como antimetabolito, y solo uno utilizó 5‐FU. La mayoría de los ensayos se evaluó como en riesgo de sesgo poco claro para la mayoría de los dominios. No se informó con frecuencia de conflictos de intereses, aunque los antimetabolitos utilizados son medicamentos genéricos, y no era probable que se realizaran estudios por interés financiero.
Hallazgos
Veinte estudios proporcionaron datos sobre el resultado primario del éxito funcional, de los cuales 7 (356 participantes) proporcionaron datos a los 6 meses y 14 (909 participantes) proporcionaron datos más allá de los 6 meses. A los seis meses, los resultados no mostraron evidencia del efecto de los antimetabolitos en el éxito funcional (riesgos relativos [RR] 1,12; intervalo de confianza [IC] del 95%: 0,98 a 1,29; evidencia de certeza baja). Más allá de los seis meses, los resultados favorecieron al grupo de antimetabolitos (RR 1,15; IC del 95%: 1,07 a 1,25; evidencia de certeza moderada).
Catorce estudios presentaron datos sobre el resultado secundario del éxito anatómico, de los cuales 4 (306 participantes) presentaron datos a los 6 meses y 12 (831 participantes) proporcionaron datos más allá de los 6 meses. Los resultados a los seis meses no mostraron evidencia del efecto de los antimetabolitos en el éxito anatómico (RR 1,02; IC del 95%: 0,95 a 1,11; evidencia de certeza baja). Después de seis meses, los participantes del grupo de antimetabolitos tuvieron más probabilidades de lograr el éxito anatómico que los que recibieron DCR solamente (RR 1,09; IC del 95%: 1,04 a 1,15; evidencia de certeza moderada).
A los seis meses y más allá de los seis meses de seguimiento, dos estudios informaron del cambio medio en el tamaño del ostium. No se realizó un metanálisis para los diversos períodos de seguimiento debido a la heterogeneidad clínica, metodológica y estadística. Sin embargo, las estimaciones puntuales de estos estudios a los seis meses favorecieron sistemáticamente a los participantes del grupo de antimetabolitos (evidencia de certeza baja). Más allá de los seis meses, mientras que las estimaciones puntuales de un estudio favorecieron a los participantes del grupo de antimetabolitos, las estimaciones de otro estudio no mostraron evidencia de una diferencia entre los dos grupos. La certeza de la evidencia en ambos puntos temporales fue baja.
Eventos adversos
Los eventos adversos fueron poco frecuentes. Un estudio informó que un participante del grupo de MMC experimentó un retraso en la cicatrización de la herida. Otros estudios no informaron de ningún evento adverso significativo relacionado con la aplicación de antimetabolitos.
Conclusiones de los autores
Hay evidencia de certeza moderada de que la aplicación de antimetabolitos en el momento de la DCR aumenta el éxito funcional y anatómico de la DCR cuando se realiza un seguimiento de los pacientes durante más de seis meses después de la cirugía, pero no hay evidencia de una diferencia a los seis meses, evidencia de certeza baja. Hay evidencia de certeza baja de que la combinación de antimetabolitos con DCR aumenta el tamaño del ostium lagrimal a los seis meses. Sin embargo, después de seis meses, la evidencia siguen siendo incierta. Los efectos adversos de la aplicación de antimetabolitos fueron mínimos.
PICO
Resumen en términos sencillos
Antimetabolitos como complemento de la dacriocistorrinostomía para la obstrucción del conducto nasolagrimal
¿Cuál es el objetivo de la revisión?
La dacriocistorrinostomía (DCR) es un tipo de cirugía que crea una nueva vía de drenaje lagrimal entre el párpado y la nariz para aliviar los síntomas lagrimales (éxito funcional), mejorar la apertura del conducto lagrimal a la irrigación (éxito anatómico) y aumentar el tamaño de la abertura en la nariz (tamaño del ostium). El objetivo era evaluar si los medicamentos contra la formación de cicatrices (antimetabolitos) pueden aumentar el éxito funcional, el éxito anatómico y el tamaño del ostium logrados con la DCR.
Resultados clave
Se encontró que los antimetabolitos pueden mejorar el éxito funcional y anatómico (en relación con la DCR solamente) en un período de seguimiento de más de seis meses. Los antimetabolitos también pueden mejorar el tamaño del ostium a los seis meses.
¿Qué se estudió en la revisión?
El sistema lagrimal del ojo produce lágrimas, que nutren la superficie del ojo y lo mantienen húmedo. Después de atravesar la superficie ocular, las lágrimas drenan hacia la nariz a través del aparato de drenaje lagrimal. La obstrucción del conducto nasolagrimal (OCNL) es el bloqueo de este canal, el cual puede causar un lagrimeo excesivo. La OCNL no suele provocar dolor y puede afectar uno o ambos ojos. La OCNL también puede dar lugar a una infección en el ojo. La OCNL se trata de forma quirúrgica con un procedimiento conocido como dacriocistorrinostomía (DCR), que establece una nueva vía al crear una entre el saco lagrimal y la nariz. Los antimetabolitos se han utilizado para mejorar las tasas de éxito de este procedimiento. Se deseaba saber si la DCR en combinación con antimetabolitos puede mejorar los resultados del éxito funcional, el éxito anatómico y el tamaño del ostium en comparación con la DCR sola. Para responder esta pregunta, se recopilaron y analizaron todos los ensayos controlados aleatorizados relevantes.
¿Cuáles son los principales resultados de la revisión?
Se identificaron 31 estudios pertinentes para su inclusión, la mayoría de los cuales se originaron en la zona meridional y oriental de Asia e incluyeron predominantemente a mujeres. Estos estudios compararon a participantes sometidos a DCR con metabolitos versus participantes sometidos a DCR sola. Veintitrés de estos estudios (1309 participantes) proporcionaron datos sobre los resultados de interés.
La DCR con antimetabolitos puede mejorar el éxito funcional y anatómico cuando el seguimiento de los pacientes se realiza durante más de seis meses después de la cirugía; la certeza de esta evidencia fue moderada. No hubo diferencias en el éxito funcional y anatómico a los seis meses entre los participantes sometidos a DCR con antimetabolitos en comparación con los participantes sometidos a DCR sola; la certeza de la evidencia es baja.
A los seis meses, los participantes sometidos a DCR con antimetabolitos pueden haber presentado un aumento del tamaño del ostium en comparación con los que recibieron DCR sola. Sin embargo, después de seis meses, no hay evidencia de una diferencia entre los participantes sometidos a DCR con antimetabolitos en comparación con los participantes sometidos a DCR sola. La certeza de la evidencia fue baja debido a la variabilidad considerable entre los estudios que evaluaron este resultado. Los efectos adversos de los antimetabolitos fueron mínimos.
¿Cuán actualizada está esta revisión?
Se revisaron los estudios publicados hasta el 6 de septiembre 2019.
Authors' conclusions
Summary of findings
| Mitomycin C dacryocystorhinostomy compared to dacryocystorhinostomy alone for nasolacrimal duct obstruction | ||||||
| Patient or population: nasolacrimal duct obstruction | ||||||
| Outcomes | Anticipated absolute effects* (95% CI) | Relative effect | № of participants | Certainty of the evidence | Comments | |
| Risk with DCR alone | Risk with MMC DCR | |||||
| Functional success, defined as the relief of epiphora Follow‐up: 6 months | Study population | RR 1.12 | 356 | ⊕⊕⊝⊝ | ||
| 81 per 100 | 90 per 100 | |||||
| Functional success, defined as the relief of epiphora Follow‐up: > 6 months | Study population | RR 1.15 | 909 | ⊕⊕⊕⊝ | ||
| 73 per 100 | 84 per 100 | |||||
| Anatomic success, defined as patency to lacrimal irrigation Follow‐up: 6 months | Study population | RR 1.02 | 306 | ⊕⊕⊝⊝ | ||
| 87 per 100 | 89 per 100 | |||||
| Anatomic success, defined as patency to lacrimal irrigation Follow‐up: > 6 months | Study population | RR 1.09 | 831 | ⊕⊕⊕⊝ | ||
| 82 per 100 | 89 per 100 | |||||
| Ostium size on nasal endoscopy Follow‐up: 6 months | The mean ostium size on nasal endoscopy ranged from 7 to 10 mm2. | Point estimates from two studies that reported mean change in ostium size at six months follow‐up. Both studies consistently show that participants treated with MMC are more likely to have larger ostium size in (mean difference (MD) 16.27, 95% CI 11.39 to 21.15; 1 study, 15 participants) and (MD 3.70, 95% CI 2.09 to 5.31; 1 study, 50 participants). | 65 | ⊕⊕⊝⊝ | As fewer than 10 studies assessed this outcome, publication bias could not be quantitatively assessed, however there may still be some but not very serious publication bias. We did not downgrade the certainty of evidence. | |
| Ostium size on nasal endoscopy at Follow‐up: > 6 months | The mean ostium size on nasal endoscopy ranged from 2 to 13 mm2. | Beyond 6 months, one study found no evidence a difference in ostium size beyond six months follow up (MD 1.40, 95% CI 0.57 to 2.23; 1 study, 50 participants), and another found that participants who were treated with MMC may experience larger ostium size (MD 8.20, 95% CI 6.14 to 10.26; 1 study 50 participants) | 100 | ⊕⊕⊝⊝ | As fewer than 10 studies assessed this outcome, publication bias could not be quantitatively assessed, however there may still be some but not very serious publication bias. We did not downgrade the certainty of evidence. | |
| *The risk in the intervention group (and its 95% confidence interval) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI). | ||||||
| GRADE Working Group grades of evidence | ||||||
| 1Downgraded (‐1) due to risk of bias. | ||||||
Background
Description of the condition
The lacrimal system of the eye includes specialized glands that naturally produce tears. The tears nourish the ocular surface and keep the eye moist. After passing along the ocular surface, tears drain into the nose. The conduit for tears between the eye and the nose is known as the lacrimal drainage apparatus. This system includes a series of four key anatomic features: the puncta (opening on the surface of each eyelid), the canaliculi (small channels that connect the puncta with the sac), the nasolacrimal sac (where tears collect), and the nasolacrimal duct (the passage from the sac that leads into the nose). Disruption of any part of the lacrimal drainage apparatus can lead to an overflow of tears. Nasolacrimal duct obstruction (NLDO) refers to a blockage of the nasolacrimal duct.
NLDO is an important ophthalmic problem. One study found an annual incidence rate of 20.24 people with NLDO per 100,000 (Woog 2007). The demographics of NLDO include a higher incidence among older people and women. In Woog 2007, the male‐to‐female ratio was about 1:3 and the mean age 60 years. It is not known if the etiology of NLDO differs by race or socioeconomic status. NLDO may be partially due to anatomic changes in the diameter of the bony lacrimal canal (Janssen 2001), which occurs with aging. These bony changes appear to affect women more than men (because women have a smaller diameter lacrimal duct at baseline) and tend to progress with time.
NLDO is usually painless unless there is an associated infection. The condition can affect one or both eyes. People with NLDO commonly present with epiphora (watery eyes), which significantly impacts their quality of life (Shin 2015). The condition can also lead to dacryocystitis (infection of the lacrimal sac), which raises the risk of secondary infections such as endophthalmitis (infection inside the eye) after cataract surgery.
NLDO is diagnosed by assessing the patency of the lacrimal drainage system with lacrimal irrigation. Typically, a tube, known as a cannula, is placed into the puncta and canaliculi and saline is irrigated. Complete reflux from the other punctum of the same eye is diagnostic of NLDO.
NLDO can be divided into congenital and acquired. Congenital NLDO is primarily treated with probing, followed by balloon catheter dilation if probing fails (Casady 2006). Congenital NLDO that has not responded to probing or balloon catheter dilation may necessitate dacryocystorhinostomy (DCR). Acquired NLDO is primarily treated surgically with DCR.
The aim of DCR is to establish a new drainage pathway by creating a connection between the lacrimal sac and the nasal mucosa. This connection requires removal of maxillary and lacrimal bone that separates those tissues. DCR may be performed via either the traditional external approach (EX‐DCR), in which a surgical incision is made through the skin of the eyelid, or the endonasal approach (EN‐DCR), in which there is no skin incision and the osteotomy is made through a nasal mucosal incision site. An endoscope is typically used to visualize the operative site for the internal approach. The success of the DCR procedure ranges from 70% to 95% (Huang 2014). While successful DCR surgery results in improved quality of life for patients, unsuccessful DCR has a negative impact on patient health (Spielmann 2009). Adjuvant methods, such as silicone stents and antimetabolites, have been used to try to improve success rates. The authors of one systematic review have summarized the effects of these various interventions in EN‐DCR (Marcet 2014). The effectiveness of interventions for congenital NLDO is discussed in another Cochrane Review (Petris 2017).
Description of the intervention
Antimetabolites are adjunctive agents that alter the wound‐healing process by inhibiting postoperative fibrosis. Two common antimetabolites used in ocular surgical procedures are mitomycin C (MMC) and 5‐fluorouracil (5‐FU). MMC is a toxic natural product of certain bacteria that causes the cross‐linking of DNA. It is typically delivered to the eye in a 0.02% to 0.04% concentration. Antimetabolites may be applied topically or injected directly into the tissues. 5‐FU blocks DNA synthesis through its action as a thymidylate synthase inhibitor of collagen gene expression, which could play a role in altering scar formation (Wendling 2003). These actions prevent normal wound‐healing responses by inhibiting cellular proliferation and fibrosis.
Intraoperative MMC has proven useful for trabeculectomy in cases at high risk of bleb failure in glaucoma surgery. Its use is associated with a significantly lower intraocular pressure after five years' follow‐up in people who underwent glaucoma filtration surgery (Bindlish 2002; Wilkins 2005). Intraoperative MMC has also been shown to be more efficacious in reducing the rate of bleb failure from scarring compared with 5‐FU given postoperatively (Skuta 1992). However, 5‐FU has found a role in cases of bleb failure due to its antifibrotic effect in bleb needling (Kapasi 2009). A randomized controlled trial comparing conjunctival autograft with MMC to prevent recurrence after pterygium surgery demonstrated that the two methods were equivalent and reduced recurrence compared with bare sclera excision (Chen 1995).
The use of antimetabolites in eye surgery should be undertaken with caution as serious complications have been reported with their use (Rubinfeld 1992). Because of previous reports of vision‐threatening complications, the minimum amount of topical antimetabolite should be used (Rubinfeld 1992). Antimetabolites have been found to be useful in nasal applications, for example in the use of reduction of fibrosis in choanal atresia surgery (Prasad 2002). In DCR surgery, antimetabolites are applied intraoperatively to the surgical ostium to prevent postoperative closure of the opening. The concentration and length of application of the agents may vary.
How the intervention might work
In certain ophthalmology procedures (i.e. glaucoma filtration and pterygium surgeries), the development of scar tissue is associated with failure of the procedure. By reducing the development of fibrosis, MMC is thought to increase the success rates of these procedures. One of the key causes of failure with DCR is a blocked ostium due to membranous scarring (Hull 2013). MMC may reduce the scarring that often causes the drainage pathway created from DCR to decrease in size, a factor that presumably leads to DCR failure (Chan 2013).
Why it is important to do this review
A Cochrane Review showed that antimetabolites reduce surgical failures in glaucoma surgery, especially in high‐risk patients (Wilkins 2005). Antimetabolites reduce surgical failure in glaucoma surgery by preventing fibrosis that results in bleb failure. It is unclear if antimetabolites would also have the same biological mechanism and clinical benefit in participants undergoing DCR. While one randomized controlled trial showed a possible benefit to using antimetabolites as an adjunct to DCR, other studies have combined the use of antimetabolites with other interventions, such as silicone stents (Dogan 2013b; Mudhol 2013b), making it difficult to infer direct conclusions about the effects of MMC and 5‐FU. The comparative effectiveness and safety of antimetabolites in dacryocystorhinostomy for nasolacrimal duct obstruction is therefore unclear.
Objectives
Primary objective: To determine if adjuvant treatment with antimetabolites improves functional success in the setting of DCR compared to DCR alone.
Secondary objectives: To determine if anatomic success of DCR is increased with the use of antimetabolites, and if the surgical ostium is larger in participants treated with antimetabolites.
Methods
Criteria for considering studies for this review
Types of studies
We included only randomized controlled trials (RCTs). Eligible RCTs were those that compared the administration of antimetabolites versus placebo or other active treatments in participants undergoing DCR.
Types of participants
We included studies in which participants underwent primary DCR and reoperation for NLDO indication. We only included studies of adults 18 years or older.
Types of interventions
We included studies in which the use of antimetabolites (MMC or 5‐FU) at any concentration and dose was compared with placebo or another active treatment as an adjunct to either EN‐DCR or EX‐DCR. We also included studies that used silicone intubation.
Types of outcome measures
Primary outcomes
-
Functional success, defined as the relief of epiphora at six months postoperatively.
Secondary outcomes
-
Anatomic success, defined as patency to lacrimal irrigation at six months postoperatively.
-
Ostium size on nasal endoscopy at six months postoperatively.
Adverse events
We compared adverse events related to treatments, such as hemorrhage, infection, and scarring.
In addition to the primary time point of six months, we evaluated outcomes reported at follow‐up times greater than six months when data were available.
Search methods for identification of studies
Electronic searches
The Cochrane Eyes and Vision Information Specialist searched the following electronic databases for RCTs. There were no restrictions on language or year of publication. We last searched the electronic databases on 6 September 2019.
-
Cochrane Central Register of Controlled Trials (CENTRAL; 2019, Issue 9) (which contains the Cochrane Eyes and Vision Trials Register) in the Cochrane Library (searched 6 September 2019) (Appendix 1).
-
MEDLINE Ovid (1946 to 6 September 2019) (Appendix 2).
-
Embase.com (1947 to 6 September 2019) (Appendix 3).
-
PubMed (1948 to 6 September 2019) (Appendix 4).
-
LILACS (Latin American and Caribbean Health Sciences Literature database) (1982 to 6 September 2019) (Appendix 5).
-
US National Institutes of Health Ongoing Trials Register ClinicalTrials.gov (www.clinicaltrials.gov; searched 6 September 2019) (Appendix 6).
-
World Health Organization (WHO) International Clinical Trials Registry Platform (ICTRP) (www.who.int/ictrp; searched 6 September 2019) (Appendix 7).
Searching other resources
We searched the reference lists of included studies to identify additional studies. We used the Web of Science database to search for reports that have cited the studies in this review. We did not handsearch journals or conference proceedings for the specific purposes of this review.
Data collection and analysis
Selection of studies
Two review authors (PP and MM) independently reviewed the titles and abstracts identified by the electronic searches according to the Criteria for considering studies for this review, classifying each record as 'definitely relevant', 'possibly relevant', or 'definitely not relevant'. Any disagreements were resolved through discussion. We retrieved the full‐text reports for records classified as 'definitely relevant' or 'possibly relevant', and two review authors independently assessed each of these as 'include' or 'unsure'. We contacted the study investigators for those reports classified as 'unsure' for further information to determine eligibility as required. Any disagreements were resolved through discussion. We reported studies excluded after full‐text review and the reasons for their exclusion in the Characteristics of excluded studies table. We classified as 'ongoing' any included studies that met the eligibility criteria but have not yet been completed or for which the study results were not available.
Data extraction and management
Two review authors independently extracted and recorded study methods, participant characteristics, and outcome data using forms developed by Cochrane Eyes and Vision. One review author entered data into Review Manager 5 (Review Manager 2014), and a second review author verified all values. Any discrepancies were resolved through discussion.
Assessment of risk of bias in included studies
Two review authors (PP and MM) independently assessed the included studies for risk of potential bias according to the guidelines in Chapter 8 of the Cochrane Handbook for Systematic Reviews of Interventions (Higgins 2019). We evaluated each study for potential bias based on the following criteria: sequence generation and allocation concealment (selection bias), masking of participants and study personnel (performance bias), masking of outcome assessors (detection bias), incomplete outcome data (attrition bias), selective outcome reporting (reporting bias), and other sources of bias. We reported the judgement for each study for each criterion as 'low risk of bias', 'high risk of bias', or ‘unclear' (information is insufficient to assess risk of bias). Any discrepancies were resolved through discussion. We contacted the study investigators for clarification as required after reviewing the study report. When the study investigators did not respond within two weeks, we based our 'Risk of bias' assessment on the available information. One review author entered data into the Characteristics of included studies table, and a second review author verified the data entry.
Measures of treatment effect
For dichotomous outcomes, we calculated risk ratios (RR) with 95% confidence intervals (CIs). Dichotomous outcomes for this review included functional success and anatomic success. We also considered the proportion of participants that had an adverse event as a dichotomous outcome. For continuous outcomes, we considered the normality of distributions and calculated mean differences (MDs) with 95% CIs when the measurements were considered normally distributed. We calculated standardized mean differences (SMDs) when continuous outcomes were measured using different scales. Continuous outcomes for this review included ostium size.
Unit of analysis issues
The unit of analysis was the participant (one eye per person). If two eyes were included per participant and received the same treatment, when possible we considered the unit of analysis to be the participant by calculating average values, or selecting one eye for analysis, per the guidelines in Chapter 16 of the Cochrane Handbook for Systematic Reviews of Interventions (Higgins 2011). When both eyes of the same participant were included, and one eye was assigned to one treatment group and the other eye was assigned to the second treatment group (i.e. paired‐eye design), we referred to Chapter 16 of the Cochrane Handbook for Systematic Reviews of Interventions for guidelines regarding considerations of correlation between the two eyes of one person (Higgins 2011).
Dealing with missing data
We contacted the study investigators for incomplete or unclear information regarding study details, outcome data, and standard deviations for means. When the investigators did not respond within two weeks, we used the available information as reported in the study. We did not impute any data.
Assessment of heterogeneity
We assessed clinical and methodological heterogeneity by examining potential variations in participant characteristics, interventions compared (EN‐DCR and EX‐DCR), and design features. We used the I2 statistic (%) to determine the proportion of variation due to statistical heterogeneity, considering a value above 50% as indicative of substantial statistical heterogeneity. We also examined the probabilities from Chi2 tests that suggested heterogeneity and the degree of overlap in CIs of effect estimates from the included studies. We considered poor overlap as indicating the presence of heterogeneity.
Assessment of reporting biases
We assessed selective outcome reporting by comparing the outcomes reported versus the outcomes listed in the study protocols or design articles, when these were available. We planned to assess small study‐effects using funnel plots for each meta‐analysis that included 10 or more trials and to examine the funnel plots for asymmetry. An asymmetric funnel plot may imply possible selection or publication bias, poor reporting of small trials, true heterogeneity, or chance.
Data synthesis
We performed a meta‐analysis when studies were clinically and methodologically comparable. We combined the outcomes from included studies in meta‐analysis using a random‐effects model, unless fewer than three studies were included, in which case we used a fixed‐effect model. When we found substantial statistical heterogeneity (I2 greater than 50%) and the direction of treatment effects was inconsistent across studies, we did not combine results in a meta‐analysis but instead presented a narrative summary.
Subgroup analysis and investigation of heterogeneity
We had planned subgroup analyses by agent used (MMC and 5‐FU) and by primary DCR and reoperation after failure. However, studies in these individual groups were insufficient to pursue a meaningful subgroup analysis. We had not planned subgroup analyses based on type of approach for DCR, but we decided post hoc to conduct subgroup analysis by stratifying data according to the approach used to visualize the operative site, either via the internal approach (EN‐DCR) or the external approach (EX‐DCR).
Sensitivity analysis
We had planned to performed sensitivity analyses to determine the impact of excluding studies at high risk of bias for incomplete outcome data and selective outcome reporting, but did not do this because many of the included studies had unclear risk of bias. We had also planned to perform sensitivity analyses by excluding studies funded by industry and those that were unpublished at the time of this review, but did not do this because no studies with these characteristics were included in the review.
Summary of findings
We summarized the main findings (see summary of findings Table for the main comparison table), including the strengths and limitations of evidence for all outcomes assessed in this review. We provided a summary of the effectiveness of the interventions and a general interpretation of the evidence in the context of other evidence, and implications for practice and future research. We used a 'Summary of findings' table according to the methods described in Chapters 11 and 12 of the Cochrane Handbook for Systematic Reviews of Interventions (Schünemann 2011a; Schünemann 2011b). Two review authors independently graded the overall certainty of the evidence for each outcome using the GRADE classification (www.gradeworkinggroup.org).
Results
Description of studies
Results of the search
The electronic search yielded 764 records (Figure 1). After removal of duplicates, we screened the remaining 520 records and excluded a further 448 records based on title and abstract review. We obtained the full‐text reports of 72 records for further investigation. We included 31 reports from 31 studies (see Characteristics of included studies table) and excluded 36 reports after full‐text screening (see Characteristics of excluded studies). We identified five ongoing studies that potentially meet the inclusion criteria, which we will assess when data become available (see Characteristics of ongoing studies).
Study flow diagram.
In an additional top‐up search conducted on 6 September 2019 that yielded 928 records, we screened 230 titles and abstracts after removal of duplicates, of which 223 records were excluded. We excluded one report after full‐text review as well as four ongoing studies that were duplicates of ongoing studies identified in the 17 August 2019 search. We listed the remaining two records as studies awaiting classification.
Overall, we included 31 studies (31 reports), excluded 37 studies (37 reports), classified 5 studies (5 reports) as ongoing studies, and identified 2 records (1 full‐text and 1 ongoing study) from the top‐up search, which we assessed as awaiting classification (Figure 1).
Included studies
See Characteristics of included studies.
Types of studies
We included 31 studies in the systematic review (Ahmad 2002; Alañón 2006; Ari 2009; Bakri 2003; Cai 2003; Chavan 2018; Costa 2007; Dogan 2013a; Eshraghy 2012; Ghosh 2006; Gonzalvo 2000; Kao 1997; Kim 2002; Liao 2000; Mukhtar 2014; Ozkiris 2012; Park 2000; Penttilä 2011; Prasannaraj 2012; Qadir 2014; Qiu 2000; Ragab 2012; Roozitalab 2004; Shaikh 2015; Tirakunwichcha 2011; Wadhera 2013; Xie 2015; Yalaz 1999; Yan 2002; Yildirim 2007; You 2001). Most studies recruited participants in Asia: six in India, five in Turkey, five in China, two in Taiwan, two in South Korea, two in Iran, one in Thailand, one in Saudi Arabia, and one in Pakistan. Outside of Asia, two studies recruited participants in Spain, one in Finland, one in England, one in Egypt, and one in Brazil.
Of the 31 included studies, eight were not included in the meta‐analysis because they either did not report review specific primary or secondary outcome data or had a follow‐up duration of less than 6 months (Alañón 2006; Costa 2007; Qiu 2000; Shaikh 2015; Xie 2015; Yalaz 1999; Chavan 2018; Mukhtar 2014). We included 23 studies in the meta‐analyses of various outcomes (Ahmad 2002; Ari 2009; Bakri 2003; Cai 2003; Dogan 2013a; Eshraghy 2012; Ghosh 2006; Gonzalvo 2000; Kao 1997; Kim 2002; Liao 2000; Ozkiris 2012; Park 2000; Penttilä 2011; Prasannaraj 2012; Qadir 2014; Ragab 2012; Roozitalab 2004; Tirakunwichcha 2011; Wadhera 2013; Yan 2002; Yildirim 2007; You 2001).
The study with the earliest enrollment of participants from meta‐analysis began in 1994 (Kao 1997), and only two studies were published prior to 2000 (Kao 1997; Yalaz 1999). Study follow‐up time varied significantly, but all had at least 6 months of follow‐up, with the maximum follow‐up being 24 months (Dogan 2013a). None of the included studies declared any sources of funding or financial interests.
Type of participants
The 31 studies enrolled a total of 2299 participants (range from 15 to 200 participants per study). The youngest mean age was 30 years, in You 2001, and the oldest mean age was 70 years, in Penttilä 2011. Study participants were generally younger than expected in previous demographic studies of NLDO (Woog 2007). Among the 15 studies that reported information on gender (Ari 2009; Bakri 2003; Cai 2003; Eshraghy 2012; Gonzalvo 2000; Mukhtar 2014; Ozkiris 2012; Park 2000; Penttilä 2011; Qadir 2014; Roozitalab 2004; Shaikh 2015; Tirakunwichcha 2011; Wadhera 2013; You 2001), participants were predominantly female, except in three studies (Eshraghy 2012; Ozkiris 2012; Wadhera 2013). The diagnosis of NLDO varied among studies, with some studies including participants with primary acquired nasolacrimal duct obstruction and others those diagnosed with recurrent nasolacrimal duct obstruction. All studies excluded individuals with congenital NLDO.
Type of interventions
Of the 31 included studies, 11 compared EX‐DCR in combination with MMC to EX‐DCR alone (Ahmad 2002; Ari 2009; Ghosh 2006; Gonzalvo 2000; Kao 1997; Liao 2000; Mukhtar 2014; Qadir 2014; Roozitalab 2004; Shaikh 2015; Yildirim 2007). Ten studies compared treatment with EN‐DCR in combination with MMC to EN‐DCR alone (Chavan 2018; Kim 2002; Ozkiris 2012; Park 2000; Penttilä 2011; Prasannaraj 2012; Ragab 2012; Tirakunwichcha 2011; Wadhera 2013; Xie 2015). Five studies comparing treatment with DCR in combination with MMC, Cai 2003; Eshraghy 2012; Qiu 2000; Yan 2002, or 5‐FU, Costa 2007, did not specify what approach (EN‐DCR or EX‐DCR) was used. One study each compared treatment with EX‐DCR with different doses of MMC, You 2001, or treatment with EX‐DCR with different doses of MMC and 5‐FU, Yalaz 1999. The remaining studies compared endonasal and endocanalicular dacryocystorhinostomy with diode laser (TLA‐ELA DCR) in combination with MMC to TLA‐ELA DCR alone (Alañón 2006); or endonasal laser dacryocystorhinostomy (ELDCR) in combination with MMC to ELDCR alone (Bakri 2003); or endocanalicular dacryocystorhinostomy (ECL‐DCR) in combination with MMC to ECL‐DCR alone (Dogan 2013a).
Of the 23 studies included in the meta‐analyses, a subgroup of nine studies compared treatment with EN‐DCR in combination with MMC to EN‐DCR alone (Dogan 2013a; Kim 2002; Ozkiris 2012; Park 2000; Penttilä 2011; Prasannaraj 2012; Ragab 2012; Tirakunwichcha 2011; Wadhera 2013); one study utilized a laser in the EN‐DCR (Dogan 2013a). Another subgroup of 13 studies compared EX‐DCR in combination with MMC to EX‐DCR alone (Ahmad 2002; Ari 2009; Cai 2003; Eshraghy 2012; Ghosh 2006; Gonzalvo 2000; Kao 1997; Liao 2000; Qadir 2014; Roozitalab 2004; Yan 2002; Yildirim 2007; You 2001). One study compared endoscopic laser DCR with 5‐FU to endoscopic laser DCR alone (Bakri 2003).
Type of outcomes
Although 31 studies were included in the review, four studies did not provide analyzable outcomes data (Alañón 2006; Qiu 2000; Xie 2015; Yalaz 1999). A further four studies assessed outcomes at less than six months follow‐up (Chavan 2018; Costa 2007; Mukhtar 2014; Shaikh 2015). Twenty‐three of the 31 RCTs provided analyzable data on either primary or secondary outcomes, or both. At 6 months and beyond, 20 RCTs provided data on functional success of DCR, and 14 had data on anatomic success. Three studies reported on ostium size. Proportions of participants experiencing complications were variably reported among the included studies.
Excluded studies
Of the 81 full‐text articles assessed for eligibility, we excluded 39 with reasons: 20 were not RCTs; 14 did not evaluate the intervention of interest; four were duplicates; and one was conducted in a different patient population (see Characteristics of excluded studies). Four were duplicates of studies already identified in previous search and classified as ongoing studies (Figure 1).
Ongoing studies and studies awaiting classification
We identified five ongoing studies and two records from the top‐up search that we assessed as awaiting classification (see Characteristics of ongoing studies and Characteristics of studies awaiting classification).
Risk of bias in included studies
The risk of bias in the included trials is summarized in Figure 2 and Figure 3.
Risk of bias summary: review authors' judgements about each risk of bias item for each included study.
Risk of bias graph: review authors' judgements about each risk of bias item presented as percentages across all included studies.
Allocation
Random sequence generation
Eight studies reported using a computer‐based random generator to generate the random allocation sequence, a method that we considered to be at low risk of bias (Ari 2009; Kim 2002; Mukhtar 2014; Ozkiris 2012; Penttilä 2011; Prasannaraj 2012; Ragab 2012; Tirakunwichcha 2011). We rated one study as having high risk of bias because the randomization was based on order of visitation (Cai 2003). The remaining 22 studies did not report the method of generating the allocation sequence and were assessed as at unclear risk of bias.
Allocation concealment
We assessed five studies that described the method used to conceal the treatment allocation sequence as at low risk of bias (Kim 2002; Penttilä 2011; Prasannaraj 2012; Ragab 2012; Tirakunwichcha 2011). One study used alternate allocation by order of visitation, therefore we determined that treatment allocation was not concealed de facto and assessed this study as at high risk of bias (Cai 2003). We assessed the remaining 25 studies as at unclear risk of bias.
Blinding
Five studies reported masking of participants (Ari 2009; Bakri 2003; Ozkiris 2012; Ragab 2012; Tirakunwichcha 2011), while five other studies reported masking of outcome assessors (Cai 2003; Gonzalvo 2000; Kao 1997; Ozkiris 2012; Roozitalab 2004); we assessed all of these studies as at low risk of bias. We assessed one study as at high risk of bias because participants and study personnel were not blinded (Ahmad 2002). We judged the remaining studies to be at unclear risk of bias due to lack of reporting of blinding of participants, study personnel, and outcome assessors.
Incomplete outcome data
We assessed 25 studies as at low risk of bias for incomplete outcome data because there were no missing data for the outcomes of our review (Ahmad 2002; Ari 2009; Cai 2003; Costa 2007; Dogan 2013a; Ghosh 2006; Gonzalvo 2000; Kao 1997; Kim 2002; Liao 2000; Mukhtar 2014; Ozkiris 2012; Park 2000; Penttilä 2011; Prasannaraj 2012; Qadir 2014; Qiu 2000; Ragab 2012; Roozitalab 2004; Shaikh 2015; Tirakunwichcha 2011; Wadhera 2013; Yalaz 1999; Yan 2002; Yildirim 2007; You 2001). We assessed two studies as at high risk of attrition bias because either they conducted analyses on as‐treated basis (Bakri 2003), or there were missing data that were not balanced across intervention arms, and reasons for missing data were not provided (Chavan 2018). We assessed the remaining three RCTs as at unclear risk of bias.
Selective reporting
We considered the risk of reporting bias as high in one study because syringing was performed, but there was no reporting of anatomic patency as a result (Ghosh 2006). The remaining studies had no study registration or published protocol available for comparison to ascertain selective outcome reporting and were therefore judged as at unclear risk of reporting bias.
Other potential sources of bias
We assessed six studies as free from other sources of bias (Ahmad 2002; Ari 2009; Bakri 2003; Kim 2002; Prasannaraj 2012; Ragab 2012). Information was insufficient to judge whether the remaining 25 studies were at low or high risk of other potential sources bias, therefore we assessed these studies as at unclear risk of bias.
Effects of interventions
Functional success
Twenty studies reported data on functional success. Meta‐analysis of 7 studies (356 participants) suggests that antimetabolite had no evidence of benefit at 6 months (risk ratio (RR) 1.12, 95% confidence interval (CI) 0.98 to 1.29). There was moderate statistical heterogeneity (I2 = 44%) (Figure 4; Analysis 1.1). The certainty of the evidence was low, downgrading for risk of bias and imprecision. However, beyond six months, antimetabolite probably improves functional success as demonstrated in a meta‐analysis of 14 studies (909 participants) (RR 1.15, 95% CI 1.07 to 1.25). There was moderate statistical heterogeneity (I2 = 34%) (Figure 4; Analysis 1.1). Visual inspection of funnel plots for functional success outcomes at six months and beyond revealed no obvious funnel plot asymmetry (Figure 5). We assessed the certainty of the evidence as moderate, downgrading one level for risk of bias. The test for subgroup differences indicated no evidence of subgroup effect at six months (P = 0.72). However, the test for subgroup differences suggest evidence of a difference in subgroup effect (P=0.05) (Figure 6; Analysis 2.1, Figure 7; Analysis 2.2) suggesting that beyond six months, DCR approaches (EN‐DCR versus EX‐DCR) significantly modifies the effect of MMC DCR in comparison to DCR alone. The treatment effect beyond six months favors EX‐DCR over EN‐DCR.
Forest plot of comparison: 1 Mitomycin C dacryocystorhinostomy versus dacryocystorhinostomy alone, outcome: 1.1 Functional success, defined as the relief of epiphora.
Funnel plot of comparison: 1 Mitomycin C dacryocystorhinostomy versus dacryocystorhinostomy alone, outcome: 1.1 Functional success, defined as the relief of epiphora.
Forest plot of comparison: 2 Subgroup: Mitomycin C dacryocystorhinostomy versus dacryocystorhinostomy alone, outcome: 2.1 Functional success, defined as the relief of epiphora at 6 months.
Forest plot of comparison: 2 Subgroup: Mitomycin C dacryocystorhinostomy versus dacryocystorhinostomy alone, outcome: 2.2 Functional success, defined as the relief of epiphora at > 6 months.
Anatomic success
Fourteen studies reported data on anatomic success. Meta‐analysis of 4 RCTs (306 participants) indicated that antimetabolites had little or no effect on anatomic success at 6 months (RR 1.02, 95% CI 0.95 to 1.11) (Figure 8; Analysis 1.2). There were no concerns regarding statistical heterogeneity across the included studies (I2 = 0%). We assessed the certainty of the evidence as low, downgrading for risk of bias and imprecision. The beneficial effect was greater beyond 6 months of follow‐up, as observed in pooled analysis of 12 RCTs (831 participants) (RR 1.09, 95% CI 1.04 to 1.15), with low statistical heterogeneity (I2 = 0%). Visual inspection of funnel plots for anatomic success revealed no obvious funnel plot asymmetry, with the exception of anatomic success at six months, where a small study‐effect appeared to be present but was not serious enough to warrant a downgrade of the certainty of the evidence (Figure 9; Analysis 1.2). We rated the certainty of the evidence as moderate, downgrading one level for risk of bias. The test for subgroup differences indicated that there is no statistically significant subgroup effect at six months (P = 0.98) or beyond six months (P = 0.27) (Figure 10; Analysis 2.3, Figure 11; Analysis 2.4), suggesting that DCR approaches (EN‐DCR versus EX‐DCR) do not modify the effect of MMC DCR in comparison to DCR alone at both time points.
Forest plot of comparison: 1 Mitomycin C dacryocystorhinostomy versus dacryocystorhinostomy alone, outcome: 1.2 Anatomic success, defined as patency to lacrimal irrigation.
Funnel plot of comparison: 1 Mitomycin C dacryocystorhinostomy versus dacryocystorhinostomy alone, outcome: 1.2 Anatomic success, defined as patency to lacrimal irrigation.
Forest plot of comparison: 2 Subgroup: Mitomycin C dacryocystorhinostomy versus dacryocystorhinostomy alone, outcome: 2.3 Anatomic success, defined as patency to lacrimal irrigation at 6 months.
Forest plot of comparison: 2 Subgroup: Mitomycin C dacryocystorhinostomy versus dacryocystorhinostomy alone, outcome: 2.4 Anatomic success, defined as patency to lacrimal irrigation at > 6 months.
Ostium size
Two studies reported mean change in ostium size at six months follow‐up. At 6 months, Kao 1997 reported data on 15 participants which demonstrated significantly larger ostium size in participants treated with MMC (mean difference (MD) 16.27, 95% CI 11.39 to 21.15). The 50 participants in Tirakunwichcha 2011 similarly demonstrated significantly increased ostium size in participants treated with MMC (MD 3.70, 95% CI 2.09 to 5.31). However, we observed considerable heterogeneity (I2 = 96%) and therefore did not perform a meta‐analysis, but instead presented point estimates in a forest plot (Figure 12; Analysis 1.3). We graded the certainty of the evidence as low, downgrading one level each for risk of bias and inconsistency.
Forest plot of comparison: 1 Mitomycin C dacryocystorhinostomy versus dacryocystorhinostomy alone, outcome: 1.3 Ostium size on nasal endoscopy at 6 months postoperatively.
Beyond 6 months, two studies reported data on ostium size. Among the 50 participants in Tirakunwichcha 2011, those treated with MMC had no evidence of a difference in ostium size at follow up (MD 1.40, 95% CI 0.57 to 2.23). Among the 50 participants in You 2001, the ostium size of those treated with MMC was measured during the final follow up period between 23 and 42 months. In this study, the ostium size in participants who were treated with 0.2 mg/mL MMC (n = 16) vs. 0.5 mg/mL MMC (n =16) vs. external DCR alone (n = 18) were compared. The mean ostium size at the final follow‐up visit was 22.2 ± 5.0 mm2 in the group treated with 0.2 mg/mL MMC, 20.6 ± 5.0 mm2 in 0.5 mg/mL MMC group, and 13.2 ± 2.7 mm2 in the control group. Overall, investigators observed that those treated with MMC were likely to experience larger ostium size (MD 8.20, 95% CI 6.14 to 10.26) compared to those treated with DCR alone. Similarly, we did not conduct meta‐analysis for data reported beyond six months due to considerable heterogeneity (I2 = 97%), instead presenting point estimates in a forest plot (Figure 12; Analysis 1.3). We graded the certainty of the evidence as low, downgrading one level each for risk of bias and inconsistency.
Data were insufficient at both six months and beyond six months to perform subgroup analyses between EN‐DCR and EX‐DCR; if data by DCR approach become available in future updates of this review, we will include these subgroup analyses.
Adverse events
Adverse events were rare. One participant in the antimetabolites group experienced delayed wound healing due to what was thought to be wound disruption related to the accidental application of an MMC‐soaked sponge on the skin. The other studies reported no significant adverse events related to the application of antimetabolites.
Discussion
Summary of main results
We identified 31 studies that compared the adjuvant treatment of antimetabolites in the setting of DCR to DCR alone. After reviewing the available evidence we summarized our findings in summary of findings Table for the main comparison for the main comparison section. We evaluated 20 studies comparing treatment with antimetabolites in combination with DCR to DCR alone on functional success. Data from seven studies indicated that participants with NLDO randomized to antimetabolites showed no evidence of effect on functional success at six months post‐DCR. The certainty of the evidence was low, with moderate statistical heterogeneity. Fourteen studies assessed functional success beyond six months, results suggests that participants randomized to antimetabolites were 1.15 times more likely to experience improvement in functional success beyond six months post‐DCR. The certainty of the evidence was moderate with moderate statistical heterogeneity.
Fourteen included studies examined anatomic success. Data from four studies indicated that participants with NLDO randomized to antimetabolites showed no evidence of a difference in anatomic success at six months. The certainty of the evidence was low. Beyond six months, participants randomized to antimetabolites were likely to experience a small increase in anatomic success compared to the control group. The certainty of the evidence was moderate. However, the effect size was generally small, and as the majority of studies that contributed data to this outcome lacked trial registration, selective outcome reporting cannot be ruled out.
Additionally, in three studies examined ostium size a six months and beyond, point estimates consistently indicated that participants randomized to antimetabolites were more likely to experience improvement in mean ostium size six month post intervention. However, beyond six months, one study found no evidence of effect antimetabolites on ostium size and another observed a difference in favor of participants receiving antimetabolites. There was considerable statistical heterogeneity that rendered meta‐analysis inappropriate for both time points. The certainty of the evidence was low.
Overall completeness and applicability of evidence
We included only RCTs in this review. Our search strategy was comprehensive. We believe that we identified a high proportion if not all published studies on antimetabolite intervention in combination with DCR for the treatment of NLDO. Specific racial or ethnic groups may be underrepresented, since most randomized participants were from South and East Asia, so our conclusions may not translate to other populations. Treatment prior to DCR in the studies were varied, with participants undergoing a revision DCR in some cases. Additionally, the approach used for interventions was not the same (EN‐DCR versus EX‐DCR approach); however, we found no significant differences between the EN‐DCR and EX‐DCR subgroups on functional success at six months and anatomic success at both time points evaluated. Furthermore, none of the included studies reported any sources of funding or financial interests, and any undeclared financial interest or support from industry is likely to impact the level of certainty of the evidence (Guyatt 2011) .
Quality of the evidence
The certainty of the evidence was moderate for the functional and anatomic success outcomes of DCR participants who were followed beyond six months. We considered the certainty of the evidence for functional and anatomic success outcomes at six months and ostium size at six months and beyond as low. Most studies did not report how the random sequence was generated or the method of concealing treatment allocation. We assessed most trials as at unclear risk of detection bias because outcome assessors were not masked. None of the trials were registered or were CONSORT compliant. Most studies were at low risk of attrition bias. Additionally, considerable statistical heterogeneity among studies that examined ostium size precluded meta‐analysis.
Potential biases in the review process
We worked with an Information Specialist to conduct broad electronic searches of multiple databases including trial registries. Although visual inspection of funnel plots revealed no obvious funnel plot asymmetry, with the exception of anatomic success at six months (Figure 5; Figure 9), publication bias for studies that demonstrated an effect of antimetabolites could not be ruled out, as visual inspection of funnel plots alone may not be a reliable way to rule out publication bias (Terrin 2005). Two review authors independently completed all steps outlined in the methods section of this review in order to reduce bias during study selection, 'Risk of bias' assessment, and data extraction.
Agreements and disagreements with other studies or reviews
Our review is generally in agreement with Cheng 2013, the only other published review on this topic that we found, in which the authors observed that intraoperative combination of MMC and EN‐DCR is safe and could improve success rate after primary and revision EN‐DCR as well as reduce the closure rate of the ostium size after EN‐DCR (Cheng 2013). Cheng and colleagues reviewed 11 randomized and non‐randomized studies conducted mostly in Asia, which included 574 eyes and defined success as patency of the nasolacrimal canal and improvement of symptoms. They found higher success rates in favor of the MMC group compared with control group (RR 1.12, 95% CI 1.04 to 1.20; P = 0.004) (Cheng 2013). However, after excluding the two non‐randomized trials from their analysis, they observed little or no difference in success rates between the two groups (Cheng 2013). When analyzing a subgroup of primary and revision EN‐DCR, and EN‐DCR without silicone intubation, they observed higher success rates in favor of the MMC group compared with the control group, but no difference in the subgroup with silicone intubation (Cheng 2013). Similar to our review, the authors of Cheng 2013 also observed bigger ostium size at osteotomy site at 3 months (weighted mean difference (WMD) 7.65, 95% CI 0.33 to 14.98; P = 0.041) and 6 months (WMD 9.28, 95% CI 2.45 to 16.11; P = 0.008), but little or no difference at 12 months after surgery (WMD 11.63, 95% CI 21.04 to 24.29; P = 0.072) (Cheng 2013).
Study flow diagram.
Risk of bias summary: review authors' judgements about each risk of bias item for each included study.
Risk of bias graph: review authors' judgements about each risk of bias item presented as percentages across all included studies.
Forest plot of comparison: 1 Mitomycin C dacryocystorhinostomy versus dacryocystorhinostomy alone, outcome: 1.1 Functional success, defined as the relief of epiphora.
Funnel plot of comparison: 1 Mitomycin C dacryocystorhinostomy versus dacryocystorhinostomy alone, outcome: 1.1 Functional success, defined as the relief of epiphora.
Forest plot of comparison: 2 Subgroup: Mitomycin C dacryocystorhinostomy versus dacryocystorhinostomy alone, outcome: 2.1 Functional success, defined as the relief of epiphora at 6 months.
Forest plot of comparison: 2 Subgroup: Mitomycin C dacryocystorhinostomy versus dacryocystorhinostomy alone, outcome: 2.2 Functional success, defined as the relief of epiphora at > 6 months.
Forest plot of comparison: 1 Mitomycin C dacryocystorhinostomy versus dacryocystorhinostomy alone, outcome: 1.2 Anatomic success, defined as patency to lacrimal irrigation.
Funnel plot of comparison: 1 Mitomycin C dacryocystorhinostomy versus dacryocystorhinostomy alone, outcome: 1.2 Anatomic success, defined as patency to lacrimal irrigation.
Forest plot of comparison: 2 Subgroup: Mitomycin C dacryocystorhinostomy versus dacryocystorhinostomy alone, outcome: 2.3 Anatomic success, defined as patency to lacrimal irrigation at 6 months.
Forest plot of comparison: 2 Subgroup: Mitomycin C dacryocystorhinostomy versus dacryocystorhinostomy alone, outcome: 2.4 Anatomic success, defined as patency to lacrimal irrigation at > 6 months.
Forest plot of comparison: 1 Mitomycin C dacryocystorhinostomy versus dacryocystorhinostomy alone, outcome: 1.3 Ostium size on nasal endoscopy at 6 months postoperatively.
Comparison 1 Mitomycin C dacryocystorhinostomy versus dacryocystorhinostomy alone, Outcome 1 Functional success, defined as the relief of epiphora.
Comparison 1 Mitomycin C dacryocystorhinostomy versus dacryocystorhinostomy alone, Outcome 2 Anatomic success, defined as patency to lacrimal irrigation.
Comparison 1 Mitomycin C dacryocystorhinostomy versus dacryocystorhinostomy alone, Outcome 3 Ostium size on nasal endoscopy.
Comparison 2 Subgroup: Mitomycin C dacryocystorhinostomy versus dacryocystorhinostomy alone, Outcome 1 Functional success, defined as the relief of epiphora at 6 months.
Comparison 2 Subgroup: Mitomycin C dacryocystorhinostomy versus dacryocystorhinostomy alone, Outcome 2 Functional success, defined as the relief of epiphora at > 6 months.
Comparison 2 Subgroup: Mitomycin C dacryocystorhinostomy versus dacryocystorhinostomy alone, Outcome 3 Anatomic success, defined as patency to lacrimal irrigation at 6 months.
Comparison 2 Subgroup: Mitomycin C dacryocystorhinostomy versus dacryocystorhinostomy alone, Outcome 4 Anatomic success, defined as patency to lacrimal irrigation at > 6 months.
| Mitomycin C dacryocystorhinostomy compared to dacryocystorhinostomy alone for nasolacrimal duct obstruction | ||||||
| Patient or population: nasolacrimal duct obstruction | ||||||
| Outcomes | Anticipated absolute effects* (95% CI) | Relative effect | № of participants | Certainty of the evidence | Comments | |
| Risk with DCR alone | Risk with MMC DCR | |||||
| Functional success, defined as the relief of epiphora Follow‐up: 6 months | Study population | RR 1.12 | 356 | ⊕⊕⊝⊝ | ||
| 81 per 100 | 90 per 100 | |||||
| Functional success, defined as the relief of epiphora Follow‐up: > 6 months | Study population | RR 1.15 | 909 | ⊕⊕⊕⊝ | ||
| 73 per 100 | 84 per 100 | |||||
| Anatomic success, defined as patency to lacrimal irrigation Follow‐up: 6 months | Study population | RR 1.02 | 306 | ⊕⊕⊝⊝ | ||
| 87 per 100 | 89 per 100 | |||||
| Anatomic success, defined as patency to lacrimal irrigation Follow‐up: > 6 months | Study population | RR 1.09 | 831 | ⊕⊕⊕⊝ | ||
| 82 per 100 | 89 per 100 | |||||
| Ostium size on nasal endoscopy Follow‐up: 6 months | The mean ostium size on nasal endoscopy ranged from 7 to 10 mm2. | Point estimates from two studies that reported mean change in ostium size at six months follow‐up. Both studies consistently show that participants treated with MMC are more likely to have larger ostium size in (mean difference (MD) 16.27, 95% CI 11.39 to 21.15; 1 study, 15 participants) and (MD 3.70, 95% CI 2.09 to 5.31; 1 study, 50 participants). | 65 | ⊕⊕⊝⊝ | As fewer than 10 studies assessed this outcome, publication bias could not be quantitatively assessed, however there may still be some but not very serious publication bias. We did not downgrade the certainty of evidence. | |
| Ostium size on nasal endoscopy at Follow‐up: > 6 months | The mean ostium size on nasal endoscopy ranged from 2 to 13 mm2. | Beyond 6 months, one study found no evidence a difference in ostium size beyond six months follow up (MD 1.40, 95% CI 0.57 to 2.23; 1 study, 50 participants), and another found that participants who were treated with MMC may experience larger ostium size (MD 8.20, 95% CI 6.14 to 10.26; 1 study 50 participants) | 100 | ⊕⊕⊝⊝ | As fewer than 10 studies assessed this outcome, publication bias could not be quantitatively assessed, however there may still be some but not very serious publication bias. We did not downgrade the certainty of evidence. | |
| *The risk in the intervention group (and its 95% confidence interval) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI). | ||||||
| GRADE Working Group grades of evidence | ||||||
| 1Downgraded (‐1) due to risk of bias. | ||||||
| Outcome or subgroup title | No. of studies | No. of participants | Statistical method | Effect size |
| 1 Functional success, defined as the relief of epiphora Show forest plot | 20 | Risk Ratio (IV, Random, 95% CI) | Subtotals only | |
| 1.1 Follow‐up: 6 months | 7 | 356 | Risk Ratio (IV, Random, 95% CI) | 1.12 [0.98, 1.29] |
| 1.2 Follow‐up: > 6 months | 14 | 909 | Risk Ratio (IV, Random, 95% CI) | 1.15 [1.07, 1.25] |
| 2 Anatomic success, defined as patency to lacrimal irrigation Show forest plot | 14 | Risk Ratio (IV, Random, 95% CI) | Subtotals only | |
| 2.1 Follow‐up: 6 months | 4 | 306 | Risk Ratio (IV, Random, 95% CI) | 1.02 [0.95, 1.11] |
| 2.2 Follow‐up: > 6 months | 12 | 831 | Risk Ratio (IV, Random, 95% CI) | 1.09 [1.04, 1.15] |
| 3 Ostium size on nasal endoscopy Show forest plot | 3 | Mean Difference (IV, Fixed, 95% CI) | Subtotals only | |
| 3.1 Follow‐up: 6 months | 2 | 65 | Mean Difference (IV, Fixed, 95% CI) | 4.93 [3.40, 6.46] |
| 3.2 Follow‐up: > 6 months | 2 | 100 | Mean Difference (IV, Fixed, 95% CI) | 2.35 [1.58, 3.12] |
| Outcome or subgroup title | No. of studies | No. of participants | Statistical method | Effect size |
| 1 Functional success, defined as the relief of epiphora at 6 months Show forest plot | 7 | 356 | Risk Ratio (IV, Random, 95% CI) | 1.12 [0.98, 1.29] |
| 1.1 EN‐DCR | 3 | 144 | Risk Ratio (IV, Random, 95% CI) | 1.17 [0.88, 1.56] |
| 1.2 EX‐DCR | 4 | 212 | Risk Ratio (IV, Random, 95% CI) | 1.10 [0.94, 1.29] |
| 2 Functional success, defined as the relief of epiphora at > 6 months Show forest plot | 14 | 909 | Risk Ratio (IV, Random, 95% CI) | 1.15 [1.07, 1.25] |
| 2.1 EN‐DCR | 6 | 436 | Risk Ratio (IV, Random, 95% CI) | 1.07 [0.98, 1.18] |
| 2.2 EX‐DCR | 8 | 473 | Risk Ratio (IV, Random, 95% CI) | 1.22 [1.11, 1.34] |
| 3 Anatomic success, defined as patency to lacrimal irrigation at 6 months Show forest plot | 4 | 306 | Risk Ratio (IV, Random, 95% CI) | 1.02 [0.95, 1.11] |
| 3.1 EN‐DCR | 1 | 76 | Risk Ratio (IV, Random, 95% CI) | 1.03 [0.84, 1.27] |
| 3.2 EX‐DCR | 3 | 230 | Risk Ratio (IV, Random, 95% CI) | 1.03 [0.93, 1.15] |
| 4 Anatomic success, defined as patency to lacrimal irrigation at > 6 months Show forest plot | 12 | 831 | Risk Ratio (IV, Random, 95% CI) | 1.09 [1.04, 1.15] |
| 4.1 EN‐DCR | 5 | 379 | Risk Ratio (IV, Random, 95% CI) | 1.06 [0.99, 1.14] |
| 4.2 EX‐DCR | 7 | 452 | Risk Ratio (IV, Random, 95% CI) | 1.12 [1.05, 1.20] |
