Retinopexia neumática versus cerclaje escleral para la reparación del desprendimiento de retina regmatógeno simple
Resumen
Antecedentes
Un desprendimiento de retina regmatógeno (DRR) es una separación de la retina neurosensorial del epitelio pigmentario de la retina causada por una rotura de espesor total asociada a una tracción vítrea. Aunque la retinopexia neumática (RN), el cerclaje escleral (CE) y la vitrectomía son intervenciones quirúrgicas bien recibidas para los ojos con DRR, su efectividad relativa todavía es controvertida.
Objetivos
Evaluar la efectividad y la seguridad de la RN versus la CE o la RN versus un tratamiento combinado de CE y vitrectomía para las personas con DRR y resumir los datos sobre las medidas económicas y la calidad de vida.
Métodos de búsqueda
Se realizaron búsquedas en CENTRAL, que contiene el Registro de ensayos del Grupo Cochrane de Salud ocular y de la visión (Cochrane Eyes and Vision Trials Register, número 3, 2021); Ovid MEDLINE; Ovid Embase; y otras cuatro bases de datos el 11 de marzo de 2021. En las búsquedas electrónicas de ensayos, no se aplicaron restricciones de fecha ni idioma.
Criterios de selección
Se incluyeron todos los ensayos controlados aleatorizados o cuasialeatorizados que compararon la efectividad de la RN versus la CE (con o sin vitrectomía) para los ojos con DRR.
Obtención y análisis de los datos
Después de comprobar la elegibilidad, dos autores de la revisión de forma independiente extrajeron las características, los métodos y los desenlaces de los estudios. Se siguieron las normas para las revisiones sistemáticas como establece Cochrane.
Resultados principales
En esta actualización, se identificó e incluyó un nuevo ensayo controlado aleatorizado. Junto con dos ensayos de la versión de 2015 de la revisión, se incluyeron tres ensayos (276 ojos de 274 participantes) que comparaban la efectividad de la RN versus la CE. Ninguno comparó la RN versus un tratamiento combinado de CE y vitrectomía.
De los tres ensayos, uno era un estudio pequeño (publicado en 1996) con 20 participantes (20 ojos) inscritos en Irlanda y seguidos durante una media de 16 meses; el segundo (publicado en 1989) incluía 196 participantes (198 ojos) en los EE.UU. seguidos durante al menos seis meses, y el tercero (publicado en 2021) se realizó en Italia e inscribió a 58 participantes (58 ojos) con un seguimiento de 12 meses. En general, la mala calidad de la publicación de información dio lugar a riesgos de sesgo poco claros o altos.
Se encontraron evidencias de certeza baja de que la RN podría lograr la reinserción de la retina con una frecuencia ligeramente menor que la CE (razón de riesgos [RR] 0,91; intervalo de confianza [IC] del 95%: 0,81 a 1,02; I2 = 0%; tres estudios, 276 ojos). Los ojos sometidos a RN también podrían mostrar un mayor riesgo de desprendimiento de retina recurrente (evidencia de certeza baja), pero las estimaciones del RR fueron muy imprecisas (RR 1,70; IC del 95%: 0,97 a 2,98; I2 = 0%; tres estudios, 276 ojos).
Los tres estudios describieron la agudeza visual (AV) final tras las dos intervenciones. Sin embargo, los resultados se proporcionaron utilizando métricas diferentes y no pudieron combinarse. Un estudio (196 participantes) informó sobre la proporción de ojos con una AV final de 20/40 o mayor y favoreció la RN (RR 1,31; IC del 95%: 1,04 a 1,65; evidencia de certeza baja), mientras que en el estudio 2021, ambos grupos mostraron una mejoría en la AV final y no hubo evidencia de una diferencia entre ambos (diferencia de medias [DM] ‐0,03; IC del 95%: ‐0,25 a 0,19; evidencia de certeza baja).
Ningún estudio aportó datos sobre la calidad de vida o las medidas económicas.
Los desenlaces de seguridad posoperatorios favorecieron en general a la RN versus la CE (evidencia de certeza baja/muy baja); sin embargo, hubo una incertidumbre considerable en cuanto al riesgo de cualquier evento adverso ocular operatorio (RR 0,55; IC del 95%: 0,28 a 1,11; 276 ojos), glaucoma (RR 0.31, IC del 95%: 0,01 a 7,46; 198 ojos), arrugas maculares (RR 0,65, IC del 95%: 0,20 a 2,11; 256 ojos), vitreorretinopatía proliferativa (RR 0,94, IC del 95%: 0,30 a 2,96; 276 ojos) y diplopía persistente (RR 0,24, IC del 95%: 0,03 a 2,09; 256 ojos). Los ojos sometidos a RN experimentaron menos aparición posoperatoria de cataratas (RR 0,40; IC del 95%: 0,21 a 0,75; 153 ojos), desprendimiento de coroides (RR 0,17; IC del 95%: 0,05 a 0,57; 198 ojos) y desplazamiento miope (RR 0,03; IC del 95%: 0,01 a 0,10; 256 ojos).
Conclusiones de los autores
La presente actualización confirma las conclusiones de la revisión anterior. La RN podría dar lugar a menores tasas de reinserción y mayores tasas de recurrencia que la CE, pero conlleva una menor carga de complicaciones postoperatorias. Los efectos de estas dos intervenciones sobre otros desenlaces funcionales y la calidad de vida siguen siendo inciertos. La evidencia disponible sigue siendo insuficiente y de baja calidad.
PICO
Resumen en términos sencillos
Intervenciones quirúrgicas para los desprendimientos de retina regmatógenos: alternativas a la vitrectomía
Pregunta de la revisión
Se ha actualizado esta revisión con el mismo objetivo de conocer cuál de las dos técnicas quirúrgicas, la hebilla escleral (CE) o la retinopexia neumática (RN), es mejor para el tratamiento de ciertos tipos de desprendimiento de retina regmatógeno (DRR).
Antecedentes
El desprendimiento de retina es la separación de la retina, el tejido sensible a la luz en la parte posterior del ojo, de su capa subyacente adherida a la superficie posterior interna del ojo. El DRR ocurre cuando se produce la separación debido a roturas o desgarros retinianos, generalmente debido a un tirón (tracción) del humor vítreo, la sustancia que llena el centro del ojo.
Se utilizan tres intervenciones quirúrgicas para reparar el/los desgarro/s retiniana/s en el DRR: RN, CE y vitrectomía. En la RN, se inyecta una burbuja de gas en la cavidad vítrea del centro del ojo para proporcionar un sellado mecánico (taponamiento) a los desgarros de retina hasta que éstas puedan sellarse con calor (láser) o frío (crioterapia). En la CE, se aplica una presión local a los desgarros de la retina mediante la sutura de material en la parte exterior del ojo (esclerótica) para hundirla (doblarla) hacia dentro. En la vitrectomía, se extrae el vítreo para aliviar la tracción del vítreo sobre la retina y se puede utilizar un gas, o aceite de silicona, para facilitar la cicatrización.
Características de los estudios
Se encontraron tres ensayos aleatorizados (en los que las personas fueron colocadas al azar en uno de dos o más grupos de tratamiento) que incluyeron a 274 participantes (276 ojos) de Irlanda, Estados Unidos e Italia. Todos los ensayos evaluaron si la RN o la CE era un mejor tratamiento para la DRR. El estudio realizado en Estados Unidos (1989) contó con 196 participantes con un seguimiento de entre seis meses y dos años. El estudio de Irlanda (1996) contó con 20 participantes con cinco a 27 meses de seguimiento. El estudio realizado en Italia (2021) incluyó a 58 participantes con 12 meses de seguimiento. La evidencia está actualizada hasta el 10 de marzo de 2021.
Fuentes de financiación de los estudios
Los estudios fueron financiados por las instituciones de los autores o por recursos desconocidos.
Resultados clave
Los resultados de los tres estudios sugieren que el CE podría funcionar mejor o tan bien como el RN en términos de tasas de reinserción y para reducir el riesgo de recurrencia del desprendimiento. Se produjeron pocos episodios adversos oculares (efectos secundarios relacionados con los ojos) durante cualquiera de las dos intervenciones, y las diferencias en algunos episodios adversos que se produjeron después de las cirugías fueron muy inciertas. Más ojos del grupo CE experimentaron cataratas y un cambio de refracción hacia la miopía (cambio a miopía que podría ser un signo de aparición de cataratas) que los ojos del grupo RN.
Calidad de la evidencia
La calidad de la evidencia fue en su mayoría baja debido a la mala información sobre cómo se realizaron los estudios. Cada estudio informó de la agudeza visual (claridad o nitidez de la visión) de forma diferente. Ninguno de los estudios analizó desenlaces importantes como la calidad de vida o los costes relacionados con los propios tratamientos.
Authors' conclusions
Summary of findings
| PR versus scleral buckle for repairing simple rhegmatogenous retinal detachments | ||||||
| Population: people with simple rhegmatogenous retinal detachments Settings: ophthalmology centers Intervention: PR Comparison: SB | ||||||
| Outcomes | Illustrative comparative risks* (95% CI) | Relative effect | No of eyes | Certainty of the evidence | Comments | |
|---|---|---|---|---|---|---|
| Assumed risk | Corresponding risk | |||||
| SB | PR | |||||
| Reattachment of the retina at 6–12 months | 843 per 1000 | 767 per 1000 | RR 0.91 (0.81 to 1.02) | 276 (3 studies) | ⊕⊕⊝⊝ | RR < 1 favors SB |
| Recurrence of retinal detachment at 6–12 months | 126 per 1000 | 214 per 1000 | RR 1.70 (0.97 to 2.98) | 276 (3 studies) | ⊕⊕⊝⊝ | RR > 1 favors SB |
| Mean change in BCVA at 12 months from baseline | — | — | MD −0.03 (95% CI −0.25 to 0.19) | 54 (1 study) | ⊕⊕⊝⊝ | Lower values mean better vision |
| Proportion of eyes with final BCVA 20/40 or better at 6 months | 526 per 1000 | 689 per 1000 (547 to 868) | RR 1.31 (1.04 to 1.65) | 198 (1 study) | ⊕⊕⊝⊝ | RR > 1 favors PR |
| Any operative ocular adverse event at 6–24 months | 142 per 1000 | 78 per 1000 (40 to 158) | RR 0.55 (0.28 to 1.11) | 276 (3 studies) | ⊕⊕⊝⊝ | RR < 1 favors PR
|
| Quality of life | Not measured in any study. | |||||
| BCVA: best‐corrected visual acuity; CI: confidence interval; MD: mean difference; PR: pneumatic retinopexy; RR: risk ratio; SB: scleral buckle. *The basis for the assumed risk is the SB group risk across studies. The corresponding risk (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). | ||||||
| The Grading of Recommendations Assessment, Development and Evaluation (GRADE) Working Group grades of evidence: | ||||||
| aDowngraded one level for risk of bias: many study details from both included studies were not reported resulting in unclear risk of bias assessments for most domains. | ||||||
Background
Description of the condition
Retinal detachment is the separation of the sensory retina from the underlying retinal pigment epithelium (Sodhi 2008). Retinal detachments are classified according to the cause of this separation; the four categories are rhegmatogenous (RRD), tractional, combined tractional/rhegmatogenous, and exudative (serous) (Sodhi 2008).
RRDs are caused by a full thickness retinal break which originates from a vitreoretinal traction and allows fluid from the vitreous cavity to enter the subretinal space (Ghazi 2002). RRDs usually require surgical management, the urgency of which depends on whether the macula is still spared (i.e. macula‐on detachments) or if it has already been involved by the detachment (i.e. macula‐off detachments). In tractional retinal detachments, a vitreoretinal membrane generates tractional strain without a full‐thickness tear in the retina, whereas in exudative retinal detachments, serous fluid accumulates underneath the sensory retina. Common causes of tractional detachments include diabetes and of exudative retinal detachment inflammatory conditions such as posterior uveitis. In the latter cases, therapy involves treatment of the underlying ocular condition or surgery, or both. In this review, we considered the surgical interventions for repairing certain types of RRD.
Worldwide, the reported incidence rate of RRD varies dramatically in different countries. It was reported to be 14 per 100,000 people per year in Sweden (Algvere 1999); 12.6 per 100,000 people per year in Minnesota, US (Rowe 1999); and 7.98 per 100,000 people per year in Beijing (Li 2003). One Danish reportshowed an increase in age and sex‐standardized RRD incidence rate of more than 50% during 2000 to 2016 and suggested the increase might be related to the increased frequency of cataract surgery (Nielsen 2020). RRD occurs most commonly in people aged 40 to 70 years with pre‐existing or concurrent posterior vitreous detachments that lead to retinal tears. The highest incidence rate is in people aged 60 to 70 years (Mitry 2010). While some studies report success rates from 90% to 95% for surgical reattachment of the retina, as many as 40% of these patients have final visual acuities of 20/50 or lower. While the occurrence of RRD in the general population is low, there are several factors that increase the risk of experiencing RRD, including lattice degeneration (lifetime risk of RRD close to 1%; Byer 1989), extreme myopia (15 to 200 times higher life‐time risk for RRD in pathologic myopia; Moisseiev 2017), cataract surgery (overall risk ranges from 0.26% to 4%; Olson 2017), and ocular trauma or infection (Sodhi 2008).
Vitreoretinal traction and underlying weakness in the peripheral retina combine to cause the retinal breaks responsible for RRD. However, not all retinal breaks will in turn cause a retinal detachment (Byer 1998). In fact, there are numerous adhesive forces that can counteract the deleterious effects of retinal breaks and maintain the stability of the vitreous‐retina border. For instance, the movement of ions and fluid by retinal pigment epithelium cells, choroid‐subretinal oncotic pressure differentials, intraocular pressure‐associated hydrostatic forces, and subretinal adhesive‐like mucopolysaccharides all can work to offset a retinal break, thus preventing progression to RRD (Ghazi 2002). When these adhesive forces are not sufficient to compensate for vitreoretinal traction, fluid can enter the subretinal space and RRD can occur (Sodhi 2008).
People with RRD often have a history of flashing lights, vitreous floaters, or both, caused by an acute posterior vitreous detachment. After a variable time, the person may notice a peripheral visual field defect, which may progress to involve central vision (Gariano 2004). However, some people do not experience these premonitory symptoms. In these people, the first sign of RRD can be a black shadow that may or may not affect visual acuity. Involvement of the macula in an RRD, which is a common cause of decreased vision in a retinal detachment, is an important prognostic marker; people without macular involvement will have better visual outcomes.
In the clinic, examination of people with RRD may reveal pigmented cells, also known as tobacco dust, in the vitreous and the anterior chamber. Other clinical findings include transparent subretinal fluid and an opaque, furrowed‐appearing retina that may ripple with the patient's eye movements (Ross 2000).
Description of the intervention
Three separate surgical interventions are used in current clinical practice to repair retinal break(s) in RRD: pneumatic retinopexy (PR), scleral buckle (SB), and vitrectomy.
PR may be performed as an outpatient clinical procedure. In this procedure, a gas bubble is injected into the vitreous cavity to provide tamponade for the detached retina, followed by cryotherapy or laser that are applied to the area of the retinal tear. A variation to the standard technique includes the initial drainage of the subretinal fluid via a posterior sclerotomy and the consequent intravitreal infusion of balanced salt solution (BSS) to flatten the detached retina. Eyes with RRD meeting the following criteria are ideal cases for surgery: single retinal tear to one clock hour or less in size, tear located in the superior half of the retina, and no associated peripheral retinal degeneration. However, eyes not meeting these criteria (e.g. larger breaks, limited lattice degeneration) also may be candidates for PR.
The SB procedure involves localizing the position of all retinal breaks, treating all retinal breaks with a cryoprobe, and supporting them with an SB. The buckle can be positioned radially, in a segmental manner, or it can encircle the entire eye. A combination of SB and PR is also possible.
Vitrectomy involves operating inside the eye and removing the vitreous to relieve vitreoretinal traction. The retina is reattached by various techniques depending on the location and extent of the detachment. At the conclusion of the vitrectomy, a gas bubble is usually injected into the eye to provide tamponade for the retina to heal (reattach). SB surgery can be combined with vitrectomy when the retinal detachment is complex.
How the intervention might work
While of dubious benefit for people with asymptomatic RRD, surgical intervention is the clear course of action for those who experience symptoms; if symptomatic RRDs are not treated, the affected eyes will be at risk for involvement of the entire retina and further vision loss.
In PR, retinal breaks are tamponaded by the intravitreal gas bubble, closed, and sealed by the chorioretinal adhesion induced by cryotherapy. The SB indents the eye wall, brings the detached retina closer to the eye wall, and relieves vitreoretinal traction. In vitrectomy, vitreous is removed and all the vitreoretinal traction on any of the breaks excised. The patient's retina is flattened intraoperatively using a gas bubble.
Why it is important to do this review
Despite the continuous improvements in surgical techniques and instrumentation witnessed in the last decades, the overall success rate of RRD repair still lies around 85% (Sultan 2020). Even though some recent developments have largely expanded the role of vitrectomy, there is no "one fits all" treatment and a tailored approach for each patient is mandatory to obtain the maximum benefit from the chosen procedure.
While various studies have proposed different paradigms for management of retinal detachment, few sufficiently powered randomized controlled trials (RCTs) have established any one therapy as clearly superior (Sodhi 2008). In this updated review, we systematically examined the existing evidence on the effectiveness of PR versus SB as two major surgical treatments for RRD.
Objectives
To assess the effectiveness and safety of PR versus SB or PR versus a combination treatment of SB and vitrectomy for people with RRD, and to summarize any data on economic measures and quality of life.
Methods
Criteria for considering studies for this review
Types of studies
We included RCTs and quasi‐RCTs with at least six months of follow‐up as specified in our published protocol (Ramchand 2010), and previous review (Hatef 2015).
Types of participants
We included studies that enrolled people with RRD for surgical treatment. Because RRD most commonly occurs in people aged 40 to 70 years with pre‐existing or concurrent posterior vitreous detachments leading to retinal tears, we planned to consider studies with participants in this age range. However, we did not exclude studies with participants outside this age range or studies that did not provide information on ages of participants.
Types of interventions
We included studies that compared PR with SB. We found no studies that compared PR with a combination of SB and vitrectomy surgery. As future studies are reported, we plan to include studies that have compared PR with a combination of SB and vitrectomy surgery.
Types of outcome measures
The previous review planned to examine outcomes at 12 months but also allowed for the inclusion of eligible studies with at least six months of follow‐up that reported early adverse events (Hatef 2015). Therefore, we included RCTs with at least six months of follow‐up and reported outcomes at other time points as displayed by the included studies.
Primary outcomes
Anatomical outcomes at postoperative six to 12 months.
-
Proportion of participants with successful reattachment of the retina after their initial surgery.
-
Proportion of participants with recurrence of retinal detachment.
Functional outcomes at the longest follow‐up time of the study (newly added outcome, see Differences between protocol and review).
-
Mean change in best‐corrected visual acuity (BCVA) from baseline.
-
Proportion of eyes with visual acuity 20/40 or better.
Adverse effects at postoperative six to 12 months.
-
Number of surgical complications such as progression of cataract in phakic eyes, postoperative choroidal detachments, proliferative vitreoretinopathy, diplopia, macular pucker, increase in intraocular pressure, and others reported in the included studies.
Secondary outcomes
Quality‐of‐life measures at postoperative six to 12 months.
-
We planned to include data on quality‐of‐life measures, however no included study reported these data.
Economic data at postoperative six to 12 months.
-
Included studies did not report economic data. When new studies are added, we plan to summarize any available data on economic measures.
Search methods for identification of studies
Electronic searches
The Cochrane Eyes and Vision Information Specialist searched the following databases for RCTs and controlled clinical trials. There were no restrictions to language or year of publication. The date of the search was 11 March 2021.
-
Cochrane Central Register of Controlled Trials (CENTRAL; 2021, Issue 3) (which contains the Cochrane Eyes and Vision Trials Register) in the Cochrane Library (searched 11 March 2021) (Appendix 1).
-
MEDLINE Ovid (1946 to 11 March 2021) (Appendix 2).
-
Embase Ovid (1980 to 11 March 2021) (Appendix 3).
-
LILACS (Latin American and Caribbean Health Science Information database; 1982 to 11 March 2021) (Appendix 4).
-
ISRCTN registry (www.isrctn.com/editAdvancedSearch; searched 11 March 2021) (Appendix 5).
-
US National Institutes of Health Ongoing Trials Register ClinicalTrials.gov (www.clinicaltrials.gov; searched 11 March 2021) (Appendix 6).
-
World Health Organization (WHO) International Clinical Trials Registry Platform (ICTRP) (www.who.int/ictrp; searched 11 March 2021) (Appendix 7).
Searching other resources
We handsearched the reference lists of the included trials to identify other possible trials. We sought to obtain information about any ongoing studies by contacting the relevant trial investigators.
Data collection and analysis
Selection of studies
Two review authors (RK and GV) independently selected the studies for inclusion using a two‐stage process. In the first stage, we screened the titles and abstracts of all the records identified by electronic searches and handsearching. Each review author classified each record as follows: 1. definitely relevant, 2. possibly relevant, or 3. definitely not relevant. In the second stage, we retrieved the full‐text report for all records classified by at least one review author as 1. definitely relevant or 2. possibly relevant. We then assessed each full‐text report and classified as: a. include, b. awaiting classification, or c. exclude. For studies classified as b., we requested additional information from study investigators. The same two review authors compared their individual classifications, and then resolved any differences by discussion or requested review by a third review author (SR). We documented all studies classified as c. exclude. We retrieved and reviewed all pertinent references from each included study to provide the most complete published information possible about study design, methods, and findings.
Data extraction and management
Two review authors (RK and SL) independently extracted data from studies included in the review using data extraction forms developed by the Cochrane Eyes and Vision. We resolved any discrepancies through discussion and by consulting a third review author when necessary. One review author (RK) entered data into Review Manager software (RevMan Web 2021), and a second review author (GV) verified the entries. Categories of information to be extracted for each study included methods (e.g. study design, number of participants, and setting), intervention details, outcomes (definitions and endpoints), and results for each outcome (sample size, missing data, summary data for each intervention). We contacted study authors whenever we needed additional information or clarification.
Assessment of risk of bias in included studies
We assessed the risk of bias as recommended in Chapter 8 of the Cochrane Handbook for Systematic Reviews of Interventions (Higgins 2019). Two review authors (RK and SL) independently assessed the risk of bias. This assessment required a description and a judgment on questions about selection bias, performance bias, detection bias, attrition bias, and reporting bias. We assessed each study as being at low, high, or unclear risk of bias. We judged studies as being at unclear risk of bias whenever lack of information or our uncertainty over the potential for bias rendered another classification impossible. Specific questions for assessing risk of bias focused on adequate sequence generation, allocation concealment before randomization, masking (blinding), adequate handling of incomplete outcome data, absence of selective outcome reporting, and absence of other potential sources of bias. Whenever the information available in the published trial reports was inadequate to assess risk of bias, we contacted the study investigators for clarification. We classified the trial based on the available information if they did not respond within two weeks. We resolved discrepancies through discussion and by consulting a third review author (GV) when necessary.
Measures of treatment effect
Data analysis followed guidelines set in Chapter 9 of the Cochrane Handbook for Systematic Reviews of Interventions (McKenzie 2021). We presented dichotomous data as risk ratios (RR) with 95% confidence intervals (CI). We reported outcomes based on the follow‐up times reported for each study. Planned dichotomous outcome measures included the proportion of participants who had successful reattachment of the retina after their initial surgery, the proportion of participants with recurrence of retinal detachment after surgical reattachment, the proportion of participants with an adverse event, and the proportion of participants with improvement (yes or no) in quality‐of‐life measures (e.g. relief of symptoms). We planned to calculate the mean difference (MD) and 95% CI for continuous outcome measures. Planned continuous outcome measures included mean change in BCVA), mean change in quality‐of‐life scores, and economic (cost) outcomes.
Unit of analysis issues
The unit of analysis was the individual (one eye of each participant included), with the exception of two participants for whom both eyes were included.
Dealing with missing data
We conducted analysis including studies with missing data in accordance with the guidelines in Chapter 16 of the Cochrane Handbook for Systematic Reviews of Interventions (Higgins 2011). We conducted the primary analysis based on the data as reported, although there was a mismatch between time points of reported outcomes and prespecified time points of interest. Whenever information was missing from the published trial reports or unclear, we contacted the primary trial investigators. We used the data available in the trial reports and described limitations of this method when applicable if they did not respond within two weeks.
Assessment of heterogeneity
We assessed heterogeneity by examining study characteristics and forest plots of the results. We used the I2 statistic to assess the impact of statistical heterogeneity and considered an I2 value of 50% or higher as indicating substantial statistical heterogeneity (Deeks 2021).
Assessment of reporting biases
We included only three studies; therefore, we could not examine funnel plots. When future studies are added to yield at least 10 studies in meta‐analyses, we plan to examine funnel plots for each outcome to assess for publication bias. We assessed for selective outcome reporting as part of the risk of bias assessment for each included study.
Data synthesis
We planned to analyze data using a random‐effects model unless there were fewer than three trials. When meta‐analysis was not appropriate due to substantial clinical or methodological heterogeneity, we reported results for each study individually and did not pool data across trials.
Subgroup analysis and investigation of heterogeneity
We were unable to perform subgroup analysis due to there being insufficient data available. When future studies are added and subgroup data become available, we plan to perform subgroup analysis based on the size and location of retinal detachment, that is whether the detachment involves the macula.
Sensitivity analysis
As we included only three studies in meta‐analyses, there was no need for a sensitivity analysis. When future studies are added, we plan to conduct a sensitivity analysis to determine the impact of exclusion of studies with lower methodological quality, unpublished studies, and industry‐funded studies. We would classify studies as of lower methodological quality based on the research design, such as studies that did not document how randomization was performed.
Summary of findings and assessment of the certainty of the evidence
We prepared summary of findings Table 1 for the following primary and secondary outcomes, which includes relative and absolute risks based on the risks across intervention groups in the included studies.
Anatomical outcomes at postoperative six to 12 months.
-
Proportion of participants with successful reattachment of the retina after their initial surgery.
-
Proportion of participants with recurrence of retinal detachment.
Functional outcomes at the longest follow‐up time of the study (newly added outcome, see Differences between protocol and review).
-
Mean change in BCVA from baseline.
-
Proportion of eyes with visual acuity 20/40 or better.
Adverse effects at postoperative six to 12 months.
-
Number of surgical complications such as progression of cataract in phakic eyes, postoperative choroidal detachments, proliferative vitreoretinopathy, diplopia, macular pucker, increase in intraocular pressure, and others reported in the included studies.
Quality‐of‐life measures at postoperative six to 12 months.
-
We planned to include data on quality‐of‐life measures, however no included study reported these data.
Two review authors (RK and GV) independently graded the overall certainty of the evidence for each outcome using the GRADE classification (GRADEpro GDT).
Results
Description of studies
Results of the search
In the previous version of the review (Hatef 2015), authors screened 1806 studies, excluded eight studies (see Excluded studies section for the reasons for exclusion), and included two eligible trials (Mulvihill 1996; Tornambe 1989); there was one study awaiting classification (Betran‐Loustaunau 1997).
While updating the review in March 2021, we identified 489 new titles and abstracts (Figure 1). After removing duplicate records, we screened 397 titles and abstracts, excluded 391 records, and assessed six full‐text reports for eligibility. We excluded five full‐text reports and recorded reasons for exclusion (Gauthier 2017; Hillier 2019; Kartasasmita 2016; Martinez‐Mujica 2018; Paulus 2017), and included one new study (Morescalchi 2021).
Study flow diagram.
In total, the review included three trials (Morescalchi 2021; Mulvihill 1996; Tornambe 1989) for evidence synthesis and excluded 13 studies (see Excluded studies section). As we were unable to extract any new information about Betran‐Loustaunau 1997, it remains under 'awaiting classification'.
Included studies
We included three RCTs (276 eyes of 274 participants) in this review (Morescalchi 2021; Mulvihill 1996; Tornambe 1989; see Characteristics of included studies table).
Settings and participants
The characteristics of participants recruited in the two previously included studies were similar (Mulvihill 1996; Tornambe 1989). Both studies included participants who were generally good candidates for PR (e.g. single retinal tear to one clock hour or less in size, tear located in the superior half of the retina, absence of proliferative vitreoretinopathy, absence of uncontrolled glaucoma). There was no reason to expect differences among participants with simple RRD from different countries or settings. Tornambe 1989 enrolled participants from a teaching hospital in the US; Mulvihill 1996 was conducted in Ireland, but did not report how participants were recruited.
While still respecting the above‐mentioned criteria for being eligible for a PR procedure, the participants included by Morescalchi 2021 were characterized by the presence of a severe superior bullous retinal detachment (i.e. a retinal detachment that partially or totally masked the vision of the posterior pole). Participants were enrolled at the ophthalmic center of the Spedali Civili di Brescia hospital, Italy.
Interventions
All studies compared PR with SB surgery. In Mulvihill 1996 and Tornambe 1989, surgeons could inject either sulfur hexafluoride (SF6) or perfluoropropane gas tamponades and use either cryopexy or laser to seal retinal breaks when performing PR. Both studies also employed similar procedures for SB surgery. Procedures generally included draining of the subretinal fluid, cryopexy to seal retinal breaks, gas injections, and suturing of the SB.
The PR technique utilized by Morescalchi 2021 was slightly different from the one performed by the other two studies. Morescalchi 2021 performed transscleral subretinal fluid drainage and flattening of the detached retina via intravitreal BSS infusion prior to the injection of the gas tamponade. By doing so, the authors might have had higher chances of obtaining better outcomes than those resulting from the other two studies. SF6, which was injected in only 6/29 participants in the SB group, was the only gas utilized in this study.
Outcomes
All three studies assessed reattachment of the retina, visual acuity, recurrence of retinal detachment, and adverse events. Morescalchi 2021 and Tornambe 1989 further assessed changes in refractive error. Tornambe 1989 reported time to resolution of symptoms and length of hospital stay, and Morescalchi 2021 assessed the duration of the surgeries themselves and the persistence of subretinal fluid. No study assessed quality of life or economic measures.
Mulvihill 1996 had a minimum of five months' follow‐up for all participants (average follow‐up of 16 months), Tornambe 1989 reported outcomes at a primary endpoint of six months' follow‐up (198/200 eyes, 99%) and secondary endpoint of 24 months' follow‐up (169/200 eyes, 85%), whereas Morescalchi 2021 had a follow‐up period of 12 months while also reporting the persistence of subretinal fluid at an intermediate time point (six months).
Excluded studies
Overall, we excluded 13 potentially relevant studies after review of the full‐text report (Avitabile 2004; Barr 1995; Figueroa 2000; Gauthier 2017; Hillier 2019; Hsu 2014; Kartasasmita 2016; Maia 2007; Martinez‐Mujica 2018; Massin 1971; Paulus 2017; Topbas 2013; Veckeneer 2001; see Excluded studies). Our reasons for exclusion included lack of randomization and ineligible interventions of interest.
Risk of bias in included studies
The certainty of evidence of the previously included studies was unclear as no study reported key methodological details (Mulvihill 1996; Tornambe 1989). Most of the domains had unclear risk of bias, but both studies had a low risk of attrition bias (Figure 2). The newly included study had a low risk of selective reporting while displaying high risks of performance, detection, and attrition bias.
Risk of bias summary: review authors' judgments about each risk of bias item for each included study.
Allocation
We assessed the risk of selection bias as unclear in both previously included studies, as neither reported adequate methods of random sequence generation or allocation concealment (Mulvihill 1996; Tornambe 1989). Although Mulvihill 1996 reported using closed envelopes to assign groups, it was unclear whether the envelopes were sequentially numbered or opaque to ensure treatment assignments were concealed.
The risk was unclear also for Morescalchi 2021 who used a 1:1 randomization without further specifying any details about the randomization process. There was no method for allocation concealment reported.
Blinding
Although it was not possible to mask the surgeons performing the procedures, it is reasonable that outcome assessors could have been masked. We judged both previously included studies to have unclear risk of performance bias, as they did not report masking of participants (Mulvihill 1996; Tornambe 1989). We also judged both studies to have unclear risk of detection bias, as they did not report masking of surgeons who assessed the outcomes. Tornambe 1989 masked visual acuity examiners to treatment groups.
In Morescalchi 2021, the protocol stated the study was "open label" and that there was no masking (NCT04139746 on ClinicalTrials.gov). Thus, we assessed the study at high risk of performance bias and detection bias.
Incomplete outcome data
Both the previously included studies (Mulvihill 1996; Tornambe 1989) displayed a low risk of attrition bias, as they had missing follow‐up data for less than 1% of participants.
Results reported by Morescalchi 2021 uniquely based on the participants achieving anatomical success by a single procedure. Two of 29 (6.9%) participants were excluded from the final analysis in both the PR and the SB groups, thus the risk of attrition bias was high.
Selective reporting
We planned to compare outcomes set in study protocols with reported outcomes to assess for selective outcome reporting. We judged both previously included studies at unclear risk of reporting bias, as we were unable to procure the protocol for either study. Morescalchi 2021 reported results for all outcome measures previously stated on ClinicalTrials.gov. Therefore, the reporting bias risk for the study was low.
Other potential sources of bias
Mulvihill 1996 and Tornambe 1989 did not report about sources of funding or conflicts of interest. Morescalchi 2021 was at low risk for other sources of bias, as they stated that the research was supported by the University of Brescia.
Effects of interventions
Reattachment of the retina
We defined successful reattachment of the retina as reattachment after the initial surgery. We combined reattachment results reported by Morescalchi 2021 and Mulvihill 1996 at different time points (12 months and mean 16 months) in a post‐hoc manner to approximate the surgical success at the postoperative six to 12 months, as most instances of reattachment would have been clinically observed by this time. Tornambe 1989 reported reattachment rates at six months.
Most participants achieved reattachment in both groups (raw sum: 109/142 [76.8%] with PR; 113/134 [84.3%] with SB). Slightly fewer participants in the PR group achieved retinal reattachment compared with participants in the SB group (RR 0.91, 95% CI 0.81 to 1.02; I2 = 0%; 3 studies, 276 eyes; Figure 3). The certainty of the evidence was low because of risk of bias (downgraded one level) and imprecision (downgraded one level) (summary of findings Table 1).
Recurrence of retinal detachment
Morescalchi 2021 and Mulvihill 1996 reported this outcome.
There was recurrence of retinal detachment in 29/142 eyes (20.4%) with PR and 16/134 (11.9%) with SB. Recurrence may have been more common in the PR group than in the SB group (RR 1.70, 95% CI 0.97 to 2.98; I2 = 0%; 3 studies, 276 eyes; Figure 4). The certainty of this evidence was low because of risk of bias (downgraded one level) and imprecision (downgraded one level).
Tornambe 1989 additionally reported that one eye in the PR group detached at seven months after the initial surgery and one eye in the SB group detached at 11 months after initial surgery.
Best‐corrected visual acuity
Neither of the two previously included studies reported mean BCVA at six to 12 months (Mulvihill 1996; Tornambe 1989). At 12 months, Morescalchi 2021 reported an MD in final BCVA of −0.3 (95% CI −0.25 to 0.19; low‐certainty evidence).
Mulvihill 1996 and Morescalchi 2021 did not report the proportion of eyes with final BCVA of 20/40 or better. At six months, Tornambe 1989 reported that 71/103 (69%) eyes in the PR group and 50/95 (53%) eyes in the SB group had BCVA of 20/40 or better (RR 1.31, 95% CI 1.04 to 1.65; low‐certainty evidence). For participants who were followed up to postoperative 24 months, Tornambe 1989 also reported large proportions of participants achieving BCVA of 20/40 or better in both groups (81/92 [88%] eyes in the PR group versus 57/77 [74%] eyes in the SB group; RR 1.19, 95% CI 1.02 to 1.38; low‐certainty evidence; Analysis 1.3).
Adverse events
Mulvihill 1996 reported 2/10 (20%) participants in the PR group developed proliferative vitreoretinopathy (one eye became phthisical) and 1/10 (10%) participants in the SB group developed subretinal hemorrhage at the drainage site. The authors did not report other adverse events. Tornambe 1989 presented operative and postoperative adverse events up to six months of follow‐up; whereas Morescalchi 2021 followed participants up to 12 months and divided the adverse events into intraoperative, early postoperative, and late postoperative categories without providing exact definitions for the cut‐off periods.
In total, there were (intra‐)operative adverse events in 11/142 (7.7%) eyes with PR and 19/134 (14.2%) eyes with SB. Fewer participants may have experienced operative complications with PR, but estimates were imprecise and included no difference (RR 0.55, 95% CI 0.28 to 1.11; I2 = 0%; 3 studies, 276 eyes; Figure 5). The certainty of this evidence was low because of risk of bias (downgraded one level) and imprecision (downgraded one level). Intra‐operative adverse events included vitreous incarceration in paracentesis site, anterior hyaloidal gas injection, anterior lens capsule touch, choroidal detachment, vitreous hemorrhage, subretinal hemorrhage, hyphema, retinal perforation, subretinal gas, and scleral punctures.
Adverse events that are specific to SB, as they are known to be associated with a greater degree of bulb manipulation, include: myopic shift (RR 0.03, 95% CI 0.01 to 0.10; I2 = 0%; 2 studies, 256 eyes), choroidal detachment (RR 0.17, 95% CI 0.05 to 0.57; 1 study, 198 eyes), and diplopia (RR 0.24, 95% CI 0.03 to 2.09; I2 = 0%; 2 studies, 256 eyes).
As classical complications of retinal detachment surgeries, there was no evidence of a difference in postoperative development of proliferative vitreoretinopathy and macular pucker between groups (proliferative vitreoretinopathy: RR 0.94, 95% CI 0.30 to 2.96; I2 = 45%; 3 studies, 276 eyes; macular pucker: RR 0.65, 95% CI 0.20 to 2.11; I2 = 0%; 2 studies, 256 eyes).
Of interest, participants undergoing PR developed a cataract less frequently than those in the SB group (RR 0.40, 95% CI 0.21 to 0.75; I2 not estimable; 2 studies, 155 eyes; Figure 5).
Morescalchi 2021 reported no evidence of a difference for the presence of persistent subretinal fluid found via optical coherence tomography between groups (RR 2.00, 95% CI 0.19 to 20.86; 1 study, 58 eyes).
Quality of life
None of the included studies reported quality of life.
Economic outcomes
None of the included studies reported economic data.
Discussion
Summary of main results
In this updated review, we identified one recently published study (Morescalchi 2021), and combined it with two previously identified studies comparing the effectiveness and safety outcomes of PR with SB for eyes with RRD. We found that about 77% of participants achieved retinal reattachment with PR by six months, but these may still be slightly fewer compared with SB (about 84%). Moreover, eyes treated with PR may have been more likely to have recurrence of retinal detachment by six months' follow‐up. This evidence was low certainty and the CIs for both these outcomes could not rule out no difference between the procedures.
The most recent study reported mean change in visual acuity and estimates were very imprecise (Morescalchi 2021). In Tornambe 1989, there was low‐certainty evidence that more participants achieved good vision with PR. None of the three studies reported outcomes regarding quality‐of‐life or economic measures.
There were operative adverse events in 1/13 eyes receiving PR compared with 1/7 eyes receiving SB; however, this evidence was low certainty, and the CIs did not rule out no difference between the procedures. For most adverse events there was uncertainty in the effect between groups, although eyes in the SB group were more susceptible to choroidal detachment, myopic shift, and cataract development than eyes in the PR group.
Overall completeness and applicability of evidence
When evaluating the results of the three included studies, we were aware that the first two studies were conducted much earlier (in the 1980s and 1990s) than the latest one (in 2021) and that the latter included a study population with severe bullous retinal detachment being treated with a classic SB approach on one eye and a revised PR procedure on the other.
Quality of the evidence
The certainty of the evidence was low for all outcomes due to inadequate reporting of several bias domains or lack of masking of outcome assessors, which is possible at least for functional outcomes. We judged each of these studies at unclear risk of bias for most of the bias domains assessed, whereas the study from Morescalchi 2021 displayed high risk of performance, detection, and attrition bias and a low risk of reporting bias.
Potential biases in the review process
We followed standard Cochrane systematic review methodology to minimize potential biases in the review process. We used no language or date restrictions in the electronic search for trials. We also searched clinical trial registries for ongoing trials.
Agreements and disagreements with other studies or reviews
We found no other systematic reviews comparing PR with SB. Other studies and reviews provide indirect or non‐comparative evidence on PR.
One narrative review summarizing the pathogenesis, diagnosis, and management of RRD concluded the success rate of overall RRD surgery to be around 85% in most large modern series (Sultan 2020). Particularly, primary anatomical success rates for SB ranged from 53% to 83% and for PR from 41% to 81%.
Hillier 2019 conducted an RCT comparing PR with vitrectomy and found that PR offered better visual acuity outcomes than the newer, more technically advanced technique. A post hoc analysis of Hillier 2019 demonstrated how a disruption of the outer retinal layers (i.e. ellipsoid zone and external limiting membrane) was more frequent in vitrectomized eyes than in those undergoing PR (Muni 2021). These data, together with the encouraging results of retrospective studies (Schmidt 2019; Yannuzzi 2021), suggest there still is a place for different surgical approaches to RRD.
The outcomes recorded in Yannuzzi 2021 are of particular interest since they present the real‐world use of PR in 9659 eyes of 9553 patients in the US. They found a single‐operation success rate of 68.5% and a final visual acuity of 0.24 logMAR in eyes with success and 0.43 logMAR in eyes with recurrence. Moreover, women had a better success rate than males and current smokers did worse than non‐smokers.
Sultan 2020 concluded that "despite the optimum method to repair detached retinas to allow maximal visual recovery (…) is gradually becoming more defined, (…) surgeon experience and preference will still remain major factors affecting technique choice." Although different clinical findings might determine the indication for different surgical approaches, it is still unclear which management is the most effective for the treatment of uncomplicated RRDs. Modern vitreoretinal surgery increasingly adopts primary pars plana vitrectomy and lately, depending on the geographic area, fewer SB and PR procedures are being performed (Bucher 2020; Reeves 2018; Williams 2014).
Study flow diagram.
Risk of bias summary: review authors' judgments about each risk of bias item for each included study.
Comparison 1: Pneumatic retinopexy versus scleral buckle, Outcome 1: Reattachment of the retina at 6–12 months' follow‐up
Comparison 1: Pneumatic retinopexy versus scleral buckle, Outcome 2: Recurrence of retinal detachment at 6–12 months' follow‐up
Comparison 1: Pneumatic retinopexy versus scleral buckle, Outcome 3: Proportion of participants with best corrected visual acuity of 20/40 or better
Comparison 1: Pneumatic retinopexy versus scleral buckle, Outcome 4: Adverse events at 6–24 months' follow‐up
| PR versus scleral buckle for repairing simple rhegmatogenous retinal detachments | ||||||
| Population: people with simple rhegmatogenous retinal detachments Settings: ophthalmology centers Intervention: PR Comparison: SB | ||||||
| Outcomes | Illustrative comparative risks* (95% CI) | Relative effect | No of eyes | Certainty of the evidence | Comments | |
|---|---|---|---|---|---|---|
| Assumed risk | Corresponding risk | |||||
| SB | PR | |||||
| Reattachment of the retina at 6–12 months | 843 per 1000 | 767 per 1000 | RR 0.91 (0.81 to 1.02) | 276 (3 studies) | ⊕⊕⊝⊝ | RR < 1 favors SB |
| Recurrence of retinal detachment at 6–12 months | 126 per 1000 | 214 per 1000 | RR 1.70 (0.97 to 2.98) | 276 (3 studies) | ⊕⊕⊝⊝ | RR > 1 favors SB |
| Mean change in BCVA at 12 months from baseline | — | — | MD −0.03 (95% CI −0.25 to 0.19) | 54 (1 study) | ⊕⊕⊝⊝ | Lower values mean better vision |
| Proportion of eyes with final BCVA 20/40 or better at 6 months | 526 per 1000 | 689 per 1000 (547 to 868) | RR 1.31 (1.04 to 1.65) | 198 (1 study) | ⊕⊕⊝⊝ | RR > 1 favors PR |
| Any operative ocular adverse event at 6–24 months | 142 per 1000 | 78 per 1000 (40 to 158) | RR 0.55 (0.28 to 1.11) | 276 (3 studies) | ⊕⊕⊝⊝ | RR < 1 favors PR
|
| Quality of life | Not measured in any study. | |||||
| BCVA: best‐corrected visual acuity; CI: confidence interval; MD: mean difference; PR: pneumatic retinopexy; RR: risk ratio; SB: scleral buckle. *The basis for the assumed risk is the SB group risk across studies. The corresponding risk (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). | ||||||
| The Grading of Recommendations Assessment, Development and Evaluation (GRADE) Working Group grades of evidence: | ||||||
| aDowngraded one level for risk of bias: many study details from both included studies were not reported resulting in unclear risk of bias assessments for most domains. | ||||||
| Outcome or subgroup title | No. of studies | No. of participants | Statistical method | Effect size |
| 1.1 Reattachment of the retina at 6–12 months' follow‐up Show forest plot | 3 | 276 | Risk Ratio (M‐H, Fixed, 95% CI) | 0.91 [0.81, 1.02] |
| 1.2 Recurrence of retinal detachment at 6–12 months' follow‐up Show forest plot | 3 | 276 | Risk Ratio (M‐H, Fixed, 95% CI) | 1.70 [0.97, 2.98] |
| 1.3 Proportion of participants with best corrected visual acuity of 20/40 or better Show forest plot | 1 | Risk Ratio (M‐H, Fixed, 95% CI) | Subtotals only | |
| 1.3.1 At 6 months | 1 | 198 | Risk Ratio (M‐H, Fixed, 95% CI) | 1.31 [1.04, 1.65] |
| 1.3.2 At 24 months | 1 | 169 | Risk Ratio (M‐H, Fixed, 95% CI) | 1.19 [1.02, 1.38] |
| 1.4 Adverse events at 6–24 months' follow‐up Show forest plot | 3 | Risk Ratio (M‐H, Fixed, 95% CI) | Subtotals only | |
| 1.4.1 Any operative ocular adverse event | 3 | 276 | Risk Ratio (M‐H, Fixed, 95% CI) | 0.55 [0.28, 1.11] |
| 1.4.2 Proliferative vitreoretinopathy | 3 | 276 | Risk Ratio (M‐H, Fixed, 95% CI) | 0.94 [0.30, 2.96] |
| 1.4.3 Cataract | 2 | 155 | Risk Ratio (M‐H, Fixed, 95% CI) | 0.40 [0.21, 0.75] |
| 1.4.4 Glaucoma | 1 | 198 | Risk Ratio (M‐H, Fixed, 95% CI) | 0.31 [0.01, 7.46] |
| 1.4.5 Macular pucker | 2 | 256 | Risk Ratio (M‐H, Fixed, 95% CI) | 0.65 [0.20, 2.11] |
| 1.4.6 Choroidal detachment | 1 | 198 | Risk Ratio (M‐H, Fixed, 95% CI) | 0.17 [0.05, 0.57] |
| 1.4.7 Myopic shift | 2 | 256 | Risk Ratio (M‐H, Fixed, 95% CI) | 0.03 [0.01, 0.10] |
| 1.4.8 Persistent diplopia | 2 | 256 | Risk Ratio (M‐H, Fixed, 95% CI) | 0.24 [0.03, 2.09] |
| 1.4.9 Persistent subretinal fluid | 1 | 58 | Risk Ratio (M‐H, Fixed, 95% CI) | 2.00 [0.19, 20.86] |
