Diseño del borde óptico de la lente intraocular para la prevención de la opacificación de la cápsula posterior tras la cirugía de cataratas
Resumen
Antecedentes
La opacificación de la cápsula posterior (OCP) es una opacidad de la parte posterior de la cápsula del cristalino, una estructura transparente parecida a la piel, que rodea el cristalino en el ojo humano. La OCP es la complicación posoperatoria más frecuente después de la cirugía de cataratas moderna con implante de una lente intraocular (LIO) de cámara posterior. Los principales síntomas de la OCP son la disminución de la agudeza visual, la visión "nublada" y borrosa y la reducción de la sensibilidad al contraste. La OCP se trata con un láser de neodimio:YAG (Nd:YAG) para crear una pequeña abertura en la cápsula opaca y recuperar un eje visual central claro. Esta capsulotomía puede provocar otras complicaciones oculares, como el aumento de la presión intraocular o la inflamación de la retina central (edema macular). Este procedimiento también supone una importante carga financiera para los sistemas sanitarios de todo el mundo. En las últimas décadas se han producido avances en la selección de los materiales de las LIO y en la optimización de los diseños de las LIO para ayudar a prevenir la formación de la OCP tras la cirugía de cataratas. Entre ellos se encuentran los cambios en las estructuras laterales que sujetan la lente en el centro del saco capsular de la lente, denominados hápticos de la LIO, y los diseños del borde óptico de la LIO.
Objetivos
Comparar los efectos de diferentes diseños de bordes ópticos de las LIO en la OCP después de la cirugía de cataratas.
Métodos de búsqueda
Se realizaron búsquedas en CENTRAL, Ovid MEDLINE, Ovid Embase, la Latin American and Caribbean Health Sciences Literature Database (LILACS), el registro ISRCTN, ClinicalTrials.gov y la Plataforma de registros internacionales de ensayos clínicos (ICTRP) de la Organización Mundial de la Salud (OMS) hasta el 17 de noviembre de 2020.
Criterios de selección
Se incluyeron los ensayos controlados aleatorizados (ECA) que compararon diferentes tipos de diseño de borde óptico de las LIO. El desenlace principal predeterminado de esta revisión fue la proporción de ojos con capsulotomía con Nd:YAG al año después de la cirugía. Los desenlaces secundarios incluyeron la puntuación de la OCP, la agudeza visual a distancia mejor corregida (AVDMC) y la puntuación de la calidad de vida al año. Debido a la disponibilidad de datos importantes a largo plazo, también se presentaron los datos al seguimiento a más largo plazo, lo cual es un cambio post hoc en el protocolo.
Obtención y análisis de los datos
Se utilizaron los métodos estándar previstos por Cochrane y el método GRADE para evaluar la certeza de la evidencia.
Resultados principales
Se incluyeron diez estudios (1065 personas, 1834 ojos) que compararon LIO de bordes afilados y redondeados. Ocho de estos estudios fueron estudios intrapersonales en los que un ojo recibió una LIO de borde afilado y el otro una LIO de borde redondeado. Los materiales de las LIO fueron acrílicos (dos estudios), silicona (cuatro estudios), polimetilmetacrilato (PMMA, tres estudios) y diferentes materiales (un estudio). Los estudios se realizaron en Austria, Alemania, India, Japón, Suecia y el Reino Unido. Cinco estudios presentaron un alto riesgo de sesgo en al menos un dominio. Se consideró que dos estudios tuvieron un bajo riesgo de sesgo en todos los dominios.
Hubo pocos casos de capsulotomía con Nd:YAG al año (desenlace principal): 1/371 en el grupo con bordes afilados y 4/371 en el grupo con bordes redondeados. La estimación del efecto estuvo a favor de las LIO de bordes afilados, pero los intervalos de confianza fueron muy amplios y compatibles con una mayor o menor probabilidad de capsulotomía con Nd:YAG en las lentes de bordes afilados, en comparación con las de bordes redondeados (odds ratio [OR] de Peto 0,30; IC del 95%: 0,05 a 1,74; I2 = 0%; seis estudios, 742 ojos). Lo anterior corresponde a siete casos menos de capsulotomía con Nd:YAG por cada 1000 LIO de borde afilado insertada en comparación con las LIO de borde redondeado (IC del 95%: 9 menos a 7 más). Se consideró que la evidencia fue de certeza baja, ya que la calificación se disminuyó por imprecisión y riesgo de sesgo.
Se observó una reducción similar en el riesgo de capsulotomía con Nd:YAG en las LIO de borde afilado en comparación con las de borde redondeado a los dos, tres y cinco años, pero como el número de eventos de capsulotomía con Nd:YAG aumentó con un seguimiento más prolongado, este efecto se midió con mayor precisión en un seguimiento más largo: dos años, razón de riesgos (RR) 0,35 (0,16 a 0,80); 703 ojos (seis estudios); 89 casos menos por 1000; tres años, RR 0,21 (0,11 a 0,41); 538 ojos (seis estudios); 170 casos menos por 1000; cinco años, RR 0,21 (0,10 a 0,45); 306 ojos (cuatro estudios); 331 casos menos por 1000. Los datos a los nueve y 12 años solo estuvieron disponibles en un estudio.
Todos los estudios proporcionaron una puntuación de OCP. Cuatro estudios informaron la puntuación AQUA (Automated Quantification of After‐Cataract), cuatro estudios informaron la puntuación EPCO (Evaluation of PCO) y dos estudios informaron sobre otro método de cuantificación de la OCP. No fue posible agrupar estos datos debido a la forma en la que se proporcionaron, pero todos los estudios proporcionaron de forma sistemática una puntuación media de OCP más baja y estadísticamente significativa (del orden de 0,5 a 3 unidades) con las LIO de borde afilado en comparación con las LIO de borde redondeado. Esta evidencia se consideró de certeza moderada, disminuida debido al riesgo de sesgo.
La puntuación de agudeza visual logMAR fue menor (mejor) en los ojos que recibieron una LIO de bordes afilados, pero la diferencia fue pequeña y probablemente no sea clínicamente importante al año (diferencia de medias [DM] ‐0,06 logMAR; IC del 95%: ‐0,12 a 0; dos estudios, 153 ojos; evidencia de certeza baja). Se observaron efectos similares en períodos de seguimiento más prolongados, pero los datos estadísticamente no significativos se notificaron en menor medida: dos años, DM ‐0,01 logMAR (‐0,05 a 0,02); dos estudios, 311 ojos; tres años, DM ‐0,09 logMAR (‐0,22 a 0,03); dos estudios, 117 ojos; los datos a los cinco años solo estuvieron disponibles en un estudio.
Ninguno de los estudios informó sobre la calidad de vida. La evidencia de certeza muy baja sobre los eventos adversos no indicó diferencias importantes entre los grupos.
Conclusiones de los autores
Esta revisión aporta evidencia de que es probable que las LIO de bordes afilados se asocien con una menor formación de OCP que las LIO de bordes redondeados, con menos capsulotomías con Nd:YAG. La certeza sobre los efectos en la agudeza visual es menor. No se ha evaluado el impacto de estas lentes en la calidad de vida y solo existen datos comparativos de certeza muy baja sobre los eventos adversos.
PICO
Resumen en términos sencillos
¿El diseño de las lentes de reemplazo utilizadas en la cirugía de cataratas ayudan a detener la opacidad de la nueva lente una vez dentro del ojo?
Mensajes clave
En la cirugía de cataratas, las lentes de reemplazo que tienen un diseño de borde afilado podrían ser menos propensas a desarrollar opacidad en su superficie que las lentes con un borde redondeado. Las personas que reciben lentes nuevas con bordes afilados podrían tener una visión más clara, y podrían tener menos probabilidades de necesitar cirugía láser, que las personas que reciben lentes con bordes redondeados.
¿Cómo se tratan las cataratas?
El cristalino es un pequeño disco transparente dentro del ojo que enfoca los rayos de luz para crear imágenes claras de los objetos observados. Las cataratas aparecen cuando se desarrollan manchas nubladas en el cristalino del ojo. A medida que las manchas nubladas aumentan con el tiempo, la vista se vuelve brumosa y borrosa.
La cirugía es la única manera de mejorar la vista si se tiene cataratas. En la cirugía de cataratas se hace un pequeño corte en el ojo, se retira el cristalino viejo que ha perdido su transparencia y se coloca una nueva lente de plástico en su lugar.
Después de una operación de cataratas
Un problema común después de la operación de cataratas es que la parte posterior del nuevo cristalino desarrolla manchas opacas en la superficie, lo que provoca nuevamente una visión borrosa o nublada. El problema se trata utilizando un láser para crear una pequeña abertura en la parte posterior del cristalino que permite el paso de la luz. Sin embargo, este procedimiento podría causar otros problemas como el aumento de la presión dentro del ojo o podría dañar otras partes del ojo.
¿Por qué se ha elaborado esta revisión Cochrane?
Se han diseñado diferentes estilos de lentes de plástico para intentar detener o reducir la opacidad una vez que están insertadas en el ojo. Los cambios incluyen hacer que los bordes de la lente sean redondeados o afilados. Se quería averiguar si los bordes de la lente con una forma diferente harían que fuera menos probable que se nublara.
¿Qué se hizo?
Se buscaron estudios que probaran diferentes diseños de lentes de plástico utilizados en la cirugía de cataratas. Se quería saber cómo afectaban los bordes del cristalino con formas diferentes la necesidad de tratamiento ocular con láser un año después de la cirugía de cataratas.
Fecha de búsqueda: se incluyó la evidencia publicada hasta el 17 de noviembre de 2020.
Qué se encontró
Se encontraron diez estudios con 1065 personas con cataratas relacionadas con la edad que se sometieron a una operación para reemplazar el cristalino en 1834 ojos. Los estudios se realizaron en Austria, Alemania, India, Japón, Suecia y el Reino Unido. Los estudios duraron al menos un año. Algunos estudios realizaron un seguimiento de los participantes hasta 12 años después de la operación.
Los estudios compararon lentes de reemplazo con bordes redondeados frente a lentes de reemplazo con bordes afilados.
Una empresa farmacéutica financió un estudio y otra financió parcialmente otro estudio.
¿Cuáles son los resultados principales de esta revisión?
En los diez estudios, los ojos a los que se les insertó una lente de borde afilado mostraron menos opacidad en el cristalino, en comparación con los ojos a los que se les administró una lente de borde redondeado.
‐ Al cabo de dos, tres y cinco años, los ojos a los que se les ha colocado una lente de borde afilado podrían tener menos probabilidades de necesitar cirugía láser que aquellos a los que se les ha colocado una lente de borde redondeado (evidencia de seis estudios en los que participaron 742 personas).
‐ Al cabo de un año y de tres años, la visión podría ser más clara en los ojos a los que se les han colocado lentes con bordes afilados en comparación con los ojos a los que se les han colocado lentes con bordes redondeados (evidencia de dos estudios en los que participaron 260 personas).
‐ Al año de la operación, no estuvo claro si el diseño de la lente (de bordes afilados o redondeados) afectó al número de personas que necesitó someterse a una cirugía con láser para tratar los parches nublados de la nueva lente. Esto se podría deber a que no muchas personas de ninguno de los dos grupos necesitaron cirugía láser al cabo de un año.
No está claro cómo los dos diseños de lentes afectaron al número de efectos no deseados que presentaron las personas, ya que éstos no se informaron de manera consistente. No se encontró evidencia de diferencias importantes entre los dos tipos de lentes en el número de complicaciones informadas.
Ninguno de los estudios midió el bienestar de las personas.
Confianza en los resultados
Se tiene una confianza moderada en que las lentes con bordes afilados muestran una puntuación media de opacidad inferior a la de las lentes con bordes redondeados. La evidencia adicional podría aumentar la confianza en este resultado. Se tiene menos confianza en que la visión sea más clara y que la necesidad de cirugía láser sea menor en los ojos que reciben lentes con bordes afilados. Es probable que nuevos estudios de investigación aumenten la confianza en estos resultados.
Authors' conclusions
Summary of findings
| Sharp‐edged compared to round‐edged IOLS for the prevention of posterior capsule opacification after cataract surgery | ||||||
| Patient or population: people undergoing cataract surgery | ||||||
| Outcomes | Anticipated absolute effects* (95% CI) | Relative effect | № of participants | Certainty of the evidence | Comments | |
|---|---|---|---|---|---|---|
| Risk with round‐edged IOL | Risk with sharp‐edged IOL | |||||
| Nd:YAG capsulotomy at one year | Study population | OR 0.30 | 742 | ⊕⊕⊝⊝ | At 2 years: RR 0.35 (0.16 to 0.80); 703 eyes (6 studies); 89 fewer cases per 1000. At 3 years: RR 0.21 (0.11 to 0.41); 538 eyes (6 studies); 170 fewer cases per 1000. At 5 years: RR 0.21 (0.10 to 0.45); 306 eyes (4 studies); 331 fewer cases per 1000. Data at 9 years and 12 years were only available from one study. | |
| 10 per 1000 | 3 per 1000 | |||||
| Posterior capsule opacification score at one year As measured by AQUA score, Evaluation of PCO (EPCO) score and other methods of quantifying PCO | All studies consistently reported a lower average PCO score (of the order of 0.5 to 3 units) with sharp‐edged IOLs compared with round‐edged IOLs | (10 RCTs) | ⊕⊕⊕⊝ | It was not possible to pool these data due to the way they were reported | ||
| BCDVA at one year, measured using logMAR score (range ‐0.3 to 1.3, low scores = better vision, a score of 0 is equivalent to 6/6 or 20/20 vision) | The mean logMAR BCDVA at one year ranged from 0 to 0.13 | MD 0.06 logMAR units lower (better) | ‐ | 153 | ⊕⊕⊝⊝ | At 2 years: MD ‐0.01 logMAR (‐0.05 to 0.02; 2 studies, 311 eyes. At 3 years MD ‐0.09 logMAR (‐0.22 to 0.03; 2 studies, 117 eyes. Data at 5 years were only available from one study. |
| Quality of life | None of the studies reported this outcome | |||||
| Adverse effects | Adverse effects were reported inconsistently. There was no evidence of any important differences in Intraocular complications, entopic phenomena, IOL tilt/decentration or anterior capsule fibrosis between groups. | ⊕⊝⊝⊝ VERY LOWa,c | ‐ | |||
| *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 | ||||||
| aDowngraded for risk of bias (‐1): only 2 studies at low risk of bias in all domains | ||||||
Background
Extracapsular cataract extraction (ECCE) with posterior chamber intraocular lens (IOL) implantation is the preferred surgical technique for the treatment of opacification of the crystalline lens (cataract) in high‐income countries, and most surgeons in low‐ and middle‐income countries are now being trained in this method (Ang 2014; Riaz 2013). In this technique, the lens contents are removed, leaving the posterior lens capsule intact. The lens capsule forms an anatomical barrier between the anterior and posterior segments of the eye. This is thought to reduce the risk of posterior segment complications following surgery (such as swelling of the central area of the retina (cystoid macular oedema) and retinal detachment) compared to intracapsular cataract extraction (ICCE), in which the whole lens with intact capsule is removed from the eye. Whereas phacoemulsification with implantation of a posterior chamber IOL is standard for cataract treatment in high‐income countries, this technique represents only 10% of all cataract surgeries in low‐ and middle‐income countries (LMICs) (Khanna 2011). The main reason for such a low percentage may be the high costs of phacoemulsification equipment (Findl 2010; Riaz 2006). However, leaving the capsule intact may result in long‐term complications, mainly that of posterior capsule opacification (PCO) (Apple 1992; Schaumberg 1998). People with PCO experience decreased visual acuity, impaired contrast sensitivity and glare.
Description of the condition
PCO is the most common postoperative complication after cataract surgery, and PCO proportions of 11% to 43% have been described within the first year after surgery, depending on surgical technique, IOL design and material, and concomitant ocular diseases (Findl 2010; Riaz 2013; Schaumberg 1998; Wormstone 2002). Clinically, PCO (or after‐cataract) is subdivided into two types: regeneratory and fibrotic PCO. Regeneratory PCO is much more common and is the main cause of a decrease in visual function after cataract surgery. Regeneratory PCO is a result of migration of lens epithelial cells along the posterior capsule, behind the IOL. These cells proliferate to form layers of lens material and Elschnig pearls, leading to opacification of the posterior capsule. In fibrotic PCO, lens epithelial cells of the anterior capsule undergo transformation to myofibroblasts, causing fibrosis and contraction of the capsule bag. This can lead to decentration of the IOL and hinder visualisation of the peripheral retina. If present on the posterior capsule, this may cause vision to deteriorate. Diagnosis is made by slit lamp examination by an ophthalmologist. PCO is usually treated by cutting an opening into the posterior capsule with a neodymium:YAG (Nd:YAG) laser, or rarely by surgical opening (capsulotomy). However, laser treatment may damage the IOL optic, transiently increase intraocular pressure (IOP), induce cystoid macular oedema, increase the incidence of retinal detachment and often does not improve visualisation of the peripheral retina (Javitt 1992). Additionally, Nd:YAG laser treatment is associated with high costs for the healthcare system. In the USA, Nd:YAG capsulotomy is the second most commonly performed ophthalmic surgical procedure after cataract surgery. The annual cost of Nd:YAG capsulotomy was estimated to be USD 250 million in the USA alone (Apple 1992). The lack of availability of laser treatment is one of the major reasons why ECCE with IOL implantation is not extensively used in low‐ and middle‐income countries.
Description of the intervention
To prevent PCO in eyes after cataract surgery with implantation of an IOL, industry and clinicians make efforts to prevent PCO by modifying IOL loops (or haptics) and optic design. The IOL optic edge may be round or sharp. A sharp edge means that the posterior surface and the lateral edge of the IOL meet at a right angle. The firm contact with the posterior capsule created by the sharp optic edge seems to be important in the inhibition of cell migration (Kohnen 2008; Nishi 2007). The sharpness of the optic edge may vary according to the type of material used to manufacture the lens (Nanavety 2008; Nanavety 2019).
How the intervention might work
IOL optic edges with a square edge reduce the migration of remnant lens epithelial cells from the equatorial zone of the capsular bag. Lens epithelial cells proliferating from the periphery of the capsule stop when they reach the sharply bent edge of the capsule in contact with the sharp edge of the IOL. As a result, this optic edge design may reduce progression of PCO towards the centre of the capsule and, therefore, towards the visual axis. Various manufacturers are offering IOLs with sharp optic edges in order to reduce postoperative PCO proportion (Nishi 2007).
The introduction of sharp optic edges appears to reduce the incidence of PCO (Findl 2010). It is unclear whether differences in the style of IOL haptics play a role in PCO inhibition.
Beside this inhibiting effect on PCO formation, the square edge of IOL optics has been shown to be a potential source for the occurrence of entopic phenomena such as positive and negative dysphotopsia. Several ray‐tracing models and in vivo studies have investigated sharp IOL optic edges as cause for these optic side effects in pseudophakic eyes (Holladay 2012; Masket 1993; Masket 2020; Schwiegerling 2016). Sharp and truncated edges may induce dysphotopsia, especially in combination with material of a high index of refraction (Radmall 2015). Therefore, some manufacturers have enhanced their IOL designs ‐ for example, by rounding the anterior optic edge.
Why it is important to do this review
Three Cochrane Reviews evaluating different surgical interventions for age‐related cataract have already been published (Ang 2014; de Silva 2014; Riaz 2013). There is an extensive body of published data on interventions to reduce the incidence of PCO. Several factors seem to influence the interaction of the IOL with the capsule bag and therefore play an important role in the prevention of PCO progression (Findl 2010). These include surgical technique (Hollick 1999; Khalifa 1992), IOL material and biocompatibility (Auffarth 2004; Findl 2010), and different IOL optic edge designs (Findl 2010; Hollick 1999; Nishi 1998; Nishi 2004). Additionally, Walker 2008 has investigated intraocular pharmaceutical agents for their influence on lens epithelial cell growth and PCO formation. In order to draw conclusions regarding the different IOL optic edge designs' influence on preventing PCO, we have undertaken a systematic review of randomised controlled trials (RCTs). This will be important to inform cataract surgeons and patients as to the best choice of IOL during cataract surgery.
Objectives
To compare the effects of different IOL optic edge designs on PCO after cataract surgery.
Methods
Criteria for considering studies for this review
Types of studies
This review included only randomised controlled trials (RCTs) with a follow‐up period of at least 12 months.
Types of participants
Participants in the trials were people with age‐related cataract who underwent cataract surgery (all extracapsular surgical techniques, including femtosecond laser‐assisted cataract surgery) with implantation of an IOL into the capsule bag. We did not exclude people with ocular comorbidities such as age‐related macular degeneration (AMD), diabetes, glaucoma or uveitis, and other sight‐threatening diseases, because low PCO rates and therefore better postoperative visual outcomes might also be an advantage for these individuals.
Types of interventions
We compared sharp (or square) IOL optic edges to round‐edge IOLs, including IOLs with special optical properties, like toric, multifocal and accommodating IOLs.
Types of outcome measures
We considered the following outcome measures.
Primary outcomes
-
Proportion of eyes receiving Nd:YAG laser capsulotomy at 12 months postoperatively. Because PCO formation is very uncommon in the first months after surgery, a minimum of 12 months follow‐up time was defined.
Nd:YAG capsulotomy may depend on a clinician's subjective evaluation, whereas PCO scores are objective measurement techniques. However, PCO scores do not indicate whether PCO is located on the central or peripheral posterior capsule and they may not reflect individuals' subjective visual impairment (Li 2013). Therefore, we selected the capsulotomy proportion as the primary outcome.
Secondary outcomes
-
Mean intensity of PCO, as assessed with scoring techniques or quantification by image analysis at 12 months and 24 months postoperatively.
-
Mean best‐corrected distance visual acuity (BCDVA) (logMAR). The minimum length of follow‐up to be included for meta‐analysis was 12 months. Data from earlier follow‐up were not included.
-
Mean quality of life score as measured by any validated questionnaire.
Assessment of PCO intensity or severity is complicated by the fact that there is no commonly accepted scoring method for PCO (Findl 2003). There are several well‐known subjective scoring systems, such as the EPCO (Evaluation of PCO) computer‐aided scoring system, as well as 'semi‐objective' or objective systems, such as POCO (posterior capsule opacification) software and AQUA (Automated Quantification of After‐Cataract) systems, both based on automated analysis of retro‐illumination PCO images. In addition, there is a frequently used objective PCO assessment method based on analysis of Scheimpflug images.
Analysis both of PCO scores and visual acuity data may be additionally complicated by the fact that some people may have an Nd:YAG capsulotomy. This problem is sometimes handled by estimating the missing PCO scores after a capsulotomy (Buehl 2005; Findl 2005). We documented how the problem of Nd:YAG was handled in each study.
Adverse effects
-
Intraoperative surgical complications (e.g. capsule rupture)
-
Entopic phenomena (e.g. negative dysphotopsia)
-
Clinically relevant IOL tilt and decentration
-
Distinctive anterior capsule fibrosis and capsule phimosis
-
Any other adverse effects as reported by studies
We collected data at 12, 24, 36 months and longer time periods where available.
Search methods for identification of studies
Electronic searches
The Cochrane Eyes and Vision Information Specialist searched the following electronic databases for randomised controlled trials and controlled clinical trials. There were no restrictions on language or year of publication. The electronic databases were last searched on 17 November 2020.
-
Cochrane Central Register of Controlled Trials (CENTRAL; 2019; Issue 11) (which contains the Cochrane Eyes and Vision Trials Register) in the Cochrane Library (searched 17 November 2020) (Appendix 1).
-
MEDLINE Ovid (1946 to 17 November 2020) (Appendix 2).
-
Embase Ovid (1980 to 17 November 2020) (Appendix 3).
-
Latin American and Caribbean Health Sciences Literature Database (LILACS) (1982 to 17 November 2020) (Appendix 4).
-
ISRCTN registry (www.isrctn.com/editAdvancedSearch; searched 17 November 2020) (Appendix 5).
-
US National Institutes of Health Ongoing Trials Register ClinicalTrials.gov (www.clinicaltrials.gov; searched 17 November 2020) (Appendix 6).
-
World Health Organization (WHO) International Clinical Trials Registry Platform (ICTRP) (www.who.int/ictrp; searched 17 November 2020) (Appendix 7).
Searching other resources
We manually searched reference lists of relevant trial reports. We did not manually search any conference abstracts for this review.
Data collection and analysis
Selection of studies
Two review authors (SM, AH) reviewed the titles and abstracts resulting from the searches. The same authors obtained full‐text copies of possibly and definitely relevant trials and assessed them according to the criteria set out above. There were no restrictions concerning publication date. Both authors worked independently from each other, and resolved discrepancies through discussion.
Data extraction and management
Three review authors (SM, AH, JE) independently extracted data using a pre‐piloted customised data extraction template in Covidence. The review authors compared these and resolved discrepancies through discussion. Data were imported directly from Covidence into Review Manager 5 (Review Manager 2014). When data were missing, we tried to contact the study authors by email and asked them to provide the missing information within a period of two months. One review author (SM) checked data entered in Review Manager 5 for accuracy.
Assessment of risk of bias in included studies
We assessed risk of bias according to the methods set out in Chapter 8 of the Cochrane Handbook for Systematic Reviews of Interventions (Higgins 2017). We used five components to determine risk of bias: selection bias (sequence generation, allocation concealment), performance bias, detection bias and attrition bias, and selective outcome reporting. We graded each component as low risk of bias, high risk of bias or unclear risk of bias. Masking (blinding) of study participants and personnel may reduce the risk of performance bias, and masking of outcome assessors may reduce risk of detection bias. Therefore, both measures, if applicable, were of special interest in our assessment of risk of bias in the studies included for this review. Two review authors (SM, AH) independently assessed the risk of bias and resolved disagreements through discussion, the use of additional referees, or both. The review authors were not masked to any trial details during the assessment. If we were unable to make a decision about the classification of a study due to lack of information, we tried to contact the study authors.
Measures of treatment effect
We presented dichotomous outcomes, such as the proportion of people needing capsulotomy, as risk ratios, and continuous outcomes, such as BCDVA and PCO scores, as the mean difference. For the analysis of ND:YAG capsulotomy at one year, there were only five events so we used the Peto odds ratio, following guidance in Chapter 10 of the Cochrane Handbook for Systematic Reviews of Interventions (Sterne 2011).
Unit of analysis issues
In our protocol, we anticipated a variety of unit of analysis issues (Differences between protocol and review). For the review, the majority of studies are within‐person studies whereby one eye receives one type of lens and the other eye receives the other. This is a paired design and, ideally, the included studies should have reported a paired analysis. For some outcomes (such as PCO score) in some studies, this was done. However, in general, the paired design was ignored by the included studies. We have used the data as reported. This is a conservative approach ‐ confidence intervals will be wider than if a paired analysis had been done.
Dealing with missing data
Most missing data were due to loss of follow‐up of participants. To assess whether the data were 'missing at random', whenever possible, we tried to contact the original investigators to identify reasons for loss of follow‐up and request incomplete data, if available.
Assessment of heterogeneity
We examined heterogeneity between trial results using a Chi2 test and the I2 statistic. Due to the fact that it has low power in the situation of a meta‐analysis when studies have small sample size or are few in number, a P value of 0.10 was used to determine statistical significance. According to Chapter 9 of the Cochrane Handbook for Systematic Reviews of Interventions, concerning the I2 statistic, values of 30% to 60% may represent moderate heterogeneity, 50% to 90% may represent substantial heterogeneity, and values between 75% and 100% may indicate considerable heterogeneity (Deeks 2017).
Assessment of reporting biases
The main reporting biases that we considered were publication bias and outcome reporting bias. For publication bias, there were not enough trials contributing data to undertake our plan to do a funnel plot to assess whether small trials had different effects. To assess outcome reporting bias we were guided by the ORBIT classification (Kirkham 2010).
Data synthesis
We summarised data from studies collecting comparable outcome measures with similar follow‐up times. We used a random‐effects model, unless there were fewer than three trials in a comparison or the data were sparse, when we used a fixed‐effect model.
Subgroup analysis and investigation of heterogeneity
We planned three subgroup analyses: firstly, to investigate whether effects varied by IOL optic material; secondly, to see whether effects varied by surgical technique; and thirdly, to see the effect of ocular co‐morbidities. There were only data available to assess the first of these planned subgroup analyses.
Sensitivity analysis
We investigated the effect of excluding studies at risk of bias. We repeated the analyses removing studies judged to be at high risk of selection bias, performance bias or detection bias.
Summary of findings and assessment of the certainty of the evidence
We created a 'Summary of findings' table including the following outcomes at one year: Nd:YAG capsulotomy, PCO score, BCDVA, quality of life and adverse effects. Two authors (SM, JE) working together assessed the certainty of the evidence for each outcome using the GRADE approach (GRADEpro GDT).
Results
Description of studies
See Characteristics of included studies and Characteristics of excluded studies.
Results of the search
The electronic searches yielded a total of 990 records (Figure 1). After 257 duplicate records were removed, the Cochrane Information Specialist (CIS) screened the remaining 733 records and removed 452 records which were not relevant to the scope of the review. We screened the remaining 281 records and identified 62 references that potentially met the inclusion criteria and obtained full‐text copies of these for further investigation. After assessing the full‐text references, we included 15 reports of 10 studies (Characteristics of included studies), and excluded 47 reports of 47 studies (Characteristics of excluded studies).
Included studies
We included 10 studies (1065 people, 1834 eyes) that compared sharp‐ and round‐edged IOLs (Table 1). The studies were conducted in Austria, Germany, India, Japan, Sweden and the UK.
| Study | Country | Number of people | Number of eyes | Type of study | Type of lens material | |
|---|---|---|---|---|---|---|
| 1 | Austria | 53 | 106 | Within‐person | Acrylic | |
| 2 | Austria | 52 | 104 | Within‐person | Silicone | |
| 3 | Austria | 32 | 64 | Within‐person | PMMA | |
| 4 | India | 100 | 200 | Within‐person | PMMA | |
| 5 | Japan | 75 | 150 | Within‐person | Acrylic | |
| 6 | Germany | 288 | 576 | Within‐person | Silicone | |
| 7 | Austria | 51 | 102 | Within‐person | Silicone | |
| 8 | India, UK | 118 | 236 | Within‐person | PMMA | |
| 9 | Sweden | 116 | 116 | Parallel group | Silicone | |
| 10 | Sweden | 180 | 180 | Parallel group | Acrylic, PMMA, silicone | |
| Total | 1065 | 1834 |
PMMA: polymethyl methacrylate
Types of study
Eight of the studies were within‐person studies, whereby eyes within one person were allocated to sharp‐ or round‐edged lenses. Two studies were parallel group studies whereby people were randomly allocated to treatment and one eye per person included in the study (Sundelin 2005; Wejde 2003).
Types of participants
People with bilateral age‐related cataract. In most of the studies, participants with ocular sight‐threatening diseases (glaucoma, diabetic retinopathy, age‐related macular degeneration or corneal pathologies) or other circumstances that would make a postoperative visual acuity of 20/40 (decimal equivalent 0.5) or better unlikely, were excluded.
Types of interventions
Uncomplicated cataract surgery in both eyes with implantation of an IOL with sharp‐edged optic or round‐edged optic. Different studies used different IOLs. Four studies used silicone IOLs (Buehl 2007; Mester 2004; Sacu 2005; Sundelin 2005), three studies used polymethyl methacrylate (PMMA) (Findl 2005; Haripriya 2017; Shah 2007), two studies used acrylic (Buehl 2005; Hayashi 2005), and one study used all three types (Wejde 2003).
Types of outcome measures
Nd:YAG capsulotomy rates; intensity of PCO (PCO scores ‐ POCO, EPCO, AQUA Score ‐ fraction of PCO); anterior capsule opacification (ACO); visual acuity (BCDVA); contrast sensitivity; glare sensitivity; IOL centration; and capsulorhexis‐IOL overlap.
Excluded studies
Altogether, we excluded 47 studies (please see Characteristics of excluded studies for further information):
-
19 studies, because they didn't have the correct study design (e.g. were not RCTs, or they were reviews) (Abela‐Formanek 2002; Auffarth 2003; Beltrame 2002; Casprini 2002; Cumming 1994; Hancox 2007; Iashinskas 2005; Kruger 2000; Li 2008; Miyata 2007; Morgan‐Warren 2013; Nishi 2007; Nixon 2004; Oner 2000; Sacu 2004b; Sacu 2005; Spyridaki 2010; Vyas 2007; Zemaitiene 2004);
-
20 studies, because IOLs of the same optic edge design were being compared (Alio 2016; Alio 2018; Cardona 2018; Gangwani 2011; Gundersen 2017; Hirnschall 2013; Hirnschall 2018; Iwase 2011; Kahraman 2015; Kugelberg 2008; Leydolt 2013; Leydolt 2017; Miháltz 2018; Prosdocimo 2003; Schrecker 2018; Voronin 2017; Yang 2017; Yu 2016; Zemaitiene 2007; Zemaitiene 2011);
-
three studies, because of the wrong outcomes (e.g. intraocular inflammatory response) (Sacu 2004b; Sacu 2004c; Wejde 2004);
-
two studies, because no copy of the paper was available after trying to contact the author (Krajcova 2007; Zhang 2002);
-
two studies because of wrong indication (IOL design comparison: three‐piece versus one‐piece IOL) (Leydolt 2007; Nejima 2006);
-
one study, because the last follow‐up was only six months postoperatively (Choi 2018).
Risk of bias in included studies
See Figure 2.
Risk of bias summary: review authors' judgements about each risk of bias item for each included study
Allocation
Only two studies used and reported methods of generating and concealing the allocation schedule sufficient for a low risk of bias judgment (Findl 2005; Hayashi 2005). Three studies described a procedure for generating a random allocation but did not describe its concealment (Haripriya 2017; Shah 2007; Sundelin 2005). The remaining studies did not report this aspect in sufficient detail.
Blinding
Two studies reported procedures to mask participants, personnel and outcome assessment (Findl 2005; Hayashi 2005). Three studies were not masked (Mester 2004; Sundelin 2005; Wejde 2003). The remaining studies provided insufficient information.
Incomplete outcome data
In cases of incomplete outcome data reporting, we tried to contact the authors of the studies to gain additional information. We graded all studies with incomplete outcome data as 'unclear risk'.
Selective reporting
We did not identify any obvious cases of selective reporting, but in the absence of access to protocols, it is likely that we could have missed instances of selective reporting. All studies reported Nd:YAG capsulotomy and PCO score. None of the studies reported quality of life but it is likely that this outcome was not measured. Adverse effects were reported inconsistently. Three studies reported visual acuity as a continuous measure and are included in the analyses (Findl 2005; Shah 2007; Sundelin 2005), two studies reported visual acuity as a dichotomous variable (Haripriya 2017; Wejde 2003), two studies reported only on lack of statistical significance (ORBIT classification A, high risk of bias) (Buehl 2007; Mester 2004) and three studies reported on lack of statistical significance and reported the data in a form that could not be included in the review (ORBIT classification C, low risk of bias) (Buehl 2005; Hayashi 2005; Sacu 2005).
Other potential sources of bias
We detected no additional potential sources of bias among the studies included in the review.
Effects of interventions
See summary of findings Table 1.
Nd:YAG capsulotomy
Six studies reported Nd:YAG capsulotomy proportions one year postoperatively (Analysis 1.1). There were few events: 1/371 in sharp‐edged and 4/371 in round‐edged IOL groups. The effect estimate was in favour of sharp‐edged IOLs but the confidence intervals were very wide and compatible with higher or lower chance of Nd:YAG capsulotomy in sharp‐edged compared with round‐edged lenses (Peto odds ratio (OR) 0.30, 95% CI 0.05 to 1.74; I2 = 0%; 6 studies, 742 eyes). This corresponds to seven fewer cases of Nd:YAG capsulotomy per 1000 sharp‐edged IOLs inserted compared with round‐edged IOLs (95% CI 9 fewer to 7 more). We judged this as low‐certainty evidence, downgrading for imprecision and risk of bias.
Similar effects were seen at other time points (Analysis 1.2; Analysis 1.3; Analysis 1.4; Analysis 1.5; Analysis 1.6), and are summarised in the following table. As the number of events in the round‐edged comparator arm increased over time, the number of cases of Nd:YAG capsulotomy potentially prevented increased over time with 89 fewer cases per 1000 at two years, 161 fewer cases per 1000 at three years, and 331 fewer cases per 1000 at five years. Data at 9 years and 12 years were only available from one study.
| Time point | Eyes (studies) | Risk ratio (95% CI) | Fewer cases of Nd:YAG capsulotomy in sharp‐edged IOLs per 1000 (95% CI) |
|---|---|---|---|
| 2 years | 703 (6) | 0.35 (0.16 to 0.80) | 89 (27 to 115) |
| 3 years | 538 (6) | 0.21, (0.11 to 0.41) | 170 (127 to 191) |
| 5 years | 306 (4) | 0.21 (0.10 to 0.45) | 331 (230 to 377) |
| 9 years | 66 (1) | 0.06 (0.01 to 0.42) | 484 (299 to 510) |
| 12 years | 179 (1) | 0.74 (0.49 to 1.13) | 113 (56 more to 221 fewer) |
Subgroup analyses
There was no evidence that these effects varied by type of IOL lens material (test for interaction at one year, P = 0.77; two years, P = 0.80; three years, P = 0.53; five years, P = 0.92), but the number of studies in each group was small and the analyses were underpowered to detect differences (Analysis 1.1; Analysis 1.2; Analysis 1.3; Analysis 1.4).
Sensitivity analyses
We investigated the effect of excluding studies at risk of bias. We repeated the analyses removing studies judged to be at high risk of selection bias, performance bias or detection bias (Table 2). The estimates of effect for the outcome Nd:YAG capsulotomy were reasonably robust and did not change much when excluding studies at high risk of bias.
| Including all studies | Excluding studies at high risk of selection, performance or detection bias | |||||
|---|---|---|---|---|---|---|
| Analysis | Number of studies | Effect estimate | 95% CI | Number of studies | Effect estimate | 95% CI |
| Analysis 1.1 Nd:YAG capsulotomy at one year | 6 | Peto OR 0.30 | 0.05 to 1.74 | 4 | Peto OR 0.14 | 0.00 to 6.82 |
| Analysis 1.2 Nd:YAG capsulotomy at two years | 6 | RR 0.35 | 0.16 to 0.80 | 4 | RR 0.37 | 0.12 to 1.18 |
| Analysis 1.3 Nd:YAG capsulotomy at three years | 6 | RR 0.21 | 0.11 to 0.41 | 3 | RR 0.19 | 0.07 to 0.52 |
| Analysis 1.4 Nd:YAG capsulotomy at five years | 4 | RR 0.21 | 0.10 to 0.45 | 2 | RR 0.30 | 0.13 to 0.72 |
| Analysis 1.5 Nd:YAG capsulotomy at nine years | Not applicable only one study | ‐ | ‐ | ‐ | ‐ | ‐ |
| Analysis 1.6 Nd:YAG capsulotomy at 12 years | Not applicable only one study | ‐ | ‐ | ‐ | ‐ | ‐ |
| Analysis 1.8 BCDVA at one year | 2 | MD ‐0.06 logMAR | ‐0.12 to 0 | 1 | MD ‐0.06 logMAR | ‐0.14 to 0.02 |
| Analysis 1.8 BCDVA at two years | 2 | MD ‐0.01 logMAR | ‐0.05 to 0.02 | 1 | MD ‐0.01 logMAR | ‐0.04 to 0.03 |
| Analysis 1.8 BCDVA at three years | 2 | MD ‐0.09 logMAR | ‐0.22 to 0.03 | 1 | MD 0.03 logMAR | ‐0.19 to 0.25 |
| Analysis 1.8 BCDVA at five years | Not applicable only one study | ‐ | ‐ | ‐ | ‐ | ‐ |
CI: confidence interval; OR: odds ratio; RR: risk ratio; MD: mean difference
PCO score
Ten studies reported a PCO score (Analysis 1.7). Four studies reported the AQUA score, four studies reported the EPCO score and two studies reported another method of quantifying PCO. It was not possible to pool these data due to the way they were reported, but all studies consistently reported a statistically significant lower average PCO score (of the order of 0.5 to 3 units) with sharp‐edged IOLs compared with round‐edged IOLs. We judged this to be moderate‐certainty evidence, downgrading for risk of bias.
Subgroup and sensitivity analyses could not be performed.
Best‐corrected distance visual acuity (BCDVA)
Three studies reported visual acuity as a continuous measure and are included in the analyses (Findl 2005; Shah 2007; Sundelin 2005), two studies reported visual acuity as a dichotomous variable (Haripriya 2017; Wejde 2003), three studies reported data in a form that could not be included in the review (Buehl 2005; Hayashi 2005; Sacu 2005) and two studies reported only on lack of statistical significance (Buehl 2007; Mester 2004).
Several studies showed better BCDVA in eyes with IOLs with sharp optic edges but the difference was small: one year postoperatively MD ‐0.06 logMAR (95% CI ‐0.12 to 0; 2 studies, 153 eyes; Analysis 1.8); two years postoperatively MD ‐0.01 logMAR (95% CI ‐0.05 to 0.02; 2 studies, 311 eyes; Analysis 1.9); three years postoperatively MD ‐0.09 logMAR (95% CI ‐0.22 to 0.03; 2 studies, 117 eyes; Analysis 1.10), and five years postoperatively MD ‐0.08 logMAR (95% CI ‐0.22 to 0.06; 1 study, 34 eyes; Analysis 1.11).
Subgroup analyses
Subgroup analyses could not be performed due to the low number of studies in each IOL lens material group.
Sensitivity analyses
We investigated the effect of excluding studies at risk of bias. We repeated the analyses removing studies judged to be at high risk of selection bias, performance bias or detection bias (Table 2). The estimates of effect for BCDVA were not affected by excluding high risk of bias studies at one and two years but the effect was reduced (changed) at three years, however, the number of studies was small.
Quality of life
None of the studies reported quality of life data or other subjective outcomes.
Adverse events
Table 3 summarises the available information on adverse events. The included studies reported adverse effects inconsistently. There was little evidence of any important differences between groups.
| Study | Intraoperative surgical complications | Entopic phenomena (e.g. negative dysphotopsia) | Clinically relevant IOL tilt and decentration | Distinctive anterior capsule fibrosis and capsule phimosis |
|---|---|---|---|---|
| Not reported | "Concerning subjective visual symptoms, there were no differences between the groups. There were 1 case of day glare in the AR40 group and 2 cases in the AR40e group and 2 cases of halos and/or photophobia in each group." | "There was no significant difference between the 2 groups at any time for decentration of the IOL or capsulorhexis–IOL overlap (1 case of buttonholing in each group)." | "There was no difference between groups in cellular ongrowth on the IOL from the anterior capsule. There were 34 cases in both groups after 1 month, 6 cases in both groups after 1 year, and 1 case in both groups after 2 years; however, there was no case in either group 3 years after surgery." | |
| "There were no surgical complications that would have led to patient exclusion." | "There was also no statistically significant difference in subjective visual symptoms described by the patients (one case of day glare in the SI40 group and two cases in the Clariflex group; three cases of halos and/or photophobia in each group)." | "Two patients were excluded after surgery, one patient because of endophthalmitis and one patient because of pronounced fibrosis‐induced decentration of an SI40 IOL that required resurgery." | Not reported | |
| "There were no surgical complications that would have led to patient exclusion." | Not reported | "We found no significant difference between the 2 groups at any time regarding decentration of the IOL (1 case in the sharp‐edge group), rhexis/IOL overlap (1 case of buttonholing in the round‐edge group), ..." | "We found no significant difference between the 2 groups at any time regarding... cases of cellular ongrowth onto the anterior IOL surface from the rhexis margin (chi‐square tests, all Ps % 0.05)." | |
| "There were no surgical complications, such as posterior capsular rupture, leading to subject disqualification." | Not reported | Not reported | Not reported | |
| "Furthermore, 1 surgeon performed all of the operations, and there were no surgical complications" | Not reported | Not reported | Not reported | |
| "Of the 288 randomized patients, 41 (14%) had to be excluded from | Not reported | "In all eyes, slitlamp examination revealed correct IOL positioning in the bag with no significant tilt or decentration." | "There was no difference between the 3 IOLs in anterior chamber cells, flare, or chronic inflammatory cells on the IOL surface by slitlamp examination." | |
| "There were no surgical complications that would have led to patient exclusion." | Not reported | Not reported | "One week postoperatively, full circumferential overlap of the peripheral optic by the anterior capsule leaf as a prerequisite for capsular bend formation was verified in all cases. In cases in which part of the optic circumference was exposed at the 5‐year follow‐up examination (one eye in the sharp edge and four eyes in the round edge optic group, P = 0.13), this had resulted from secondary buttonholing generally occurring during months 1 and 6." | |
| "No serious surgical complications were seen." | Not reported | 6/107 in round‐edge group had decentered IOL compared with 4/107 in the sharp‐edge group. | 1 case of iris atrophy, 1 case of posterior synechiae, 41 cases capsulorhexis contact with IOL surface in round‐edge group. 51 cases of capsulorhexis contact with IOL surface in square‐edge IOL group. | |
| "Surgical complications" 1 case in each group given as a reason for incomplete follow‐up. | Not reported | "At the 2‐month visit it was noticed that in some patients the optic was partly outside the capsulorhexis rim. This was due either to an asymmetric or overly large capsulorhexis, or to the IOL being slightly decentred in the capsular bag. In many patients the pupil could not be dilated sufficiently to inspect the entire optic/ capsulorhexis relationship at the 2‐month visit. A total of 113 patients attended the 2‐month visit; in 55 patients the entire capsulorhexis was on the optic, in 17 patients some part of the capsulorhexis was outside the optic edge, and in 41 patients the optic/capsulorhexis relationship could not be assessed" | Not reported | |
| Not reported | Not reported | Not reported | Not reported |
Intraoperative complications
-
Two studies did not report intraoperative complications (Buehl 2005; Wejde 2003).
-
Six studies reported that there were either no intraoperative complications (Hayashi 2005; Shah 2007), or no intraoperative complications leading to exclusion (Buehl 2007; Findl 2005; Haripriya 2017; Sacu 2005).
-
Two studies reported intraoperative complications. In Mester 2004, 10/288 people were excluded due to severe complications during surgery but it was not clear to which groups these people were allocated. In Sundelin 2005, one case in each group had "surgical complications" given as a reason for incomplete follow‐up.
Entopic phenomena ‐ negative dysphotopsia
Two studies reported cases of positive dysphotopsia (Buehl 2005; Buehl 2007). However, the number of reported cases was low and no difference could be found between different IOL optic edge designs.
Most studies did not report either the presence or absence of entopic phenomena. Two studies reported glare, halos and photophobia but these occurred rarely and in both groups (Buehl 2005; Buehl 2007).
Clinically relevant IOL tilt and decentration
-
Four studies did not report on IOL tilt or decentration (Haripriya 2017; Hayashi 2005; Sacu 2005; Wejde 2003).
-
One study reported that there was correct IOL position in all eyes with no important tilt or decentration (Mester 2004).
-
In the remaining five studies, IOL tilt or decentration was reported but in most cases this occurred rarely and evenly between groups.
Distinctive anterior capsule fibrosis and capsule phimosis
-
Five studies did not report on this (Buehl 2007; Haripriya 2017; Hayashi 2005; Sundelin 2005; Wejde 2003).
-
The remaining studies reported no important differences between groups.
Discussion
Summary of main results
We included 10 studies (1834 eyes) that compared sharp‐ and round‐edged IOLs. Eight of these studies were within‐person studies: one eye received a sharp‐edged IOL and the fellow eye a round‐edged IOL. The IOL materials were acrylic (two studies), silicone (four studies), PMMA (three studies) and different materials (one study). The studies were conducted in Austria, Germany, India, Japan, Sweden and the UK. We judged five studies to be at high risk of bias in at least one domain. We judged only two studies to be at low risk of bias in all domains.
Low‐certainty evidence suggested a reduction in the chance of Nd:YAG capsulotomy at one year with sharp‐edged IOLs compared to round‐edged IOLs: there were 7 fewer cases of Nd:YAG capsulotomy per 1000 (95% confidence interval (CI) 9 fewer to 7 more) (Peto odds ratio (OR) 0.30, 95% CI 0.05 to 1.74; 6 studies, 742 eyes). A similar relative effect was seen at two, three and five years. As the number of Nd:YAG capsulotomy events increased with longer follow‐up, the number of cases of Nd:YAG capsulotomy prevented increased over time. There was no evidence that these effects varied by type of IOL lens material but the number of studies in each group was small and the analyses underpowered to detect any differences in effect.
All of the studies reported a PCO score. Four studies reported the AQUA score, four studies reported the EPCO score and two studies reported another method of quantifying PCO. It was not possible to pool these data due to the way they were reported, but all studies consistently reported a lower average PCO score (of the order of 0.5 to 3 units) with sharp‐edged IOLs compared with round‐edged IOLs. We judged this to be moderate‐certainty evidence, downgrading for risk of bias.
Low‐certainty evidence suggested the logMAR visual acuity score was lower (better) in eyes that received a sharp‐edged IOL but the difference was small at one year and probably clinically unimportant. There was a greater difference observed at longer follow‐up times but statistically non‐significant results were less likely to be reported fully and incorporated in the analyses.
Very low‐certainty evidence on adverse events did not suggest any important differences between the groups. None of the studies reported quality of life.
Overall completeness and applicability of evidence
Participants in the included studies were likely to be representative of a standard cataract population. Although we did not exclude participants with ocular comorbidities in this review, most of the included studies excluded people with ocular comordibities. We believe that people with age‐related macular degeneration (AMD), glaucoma, diabetes and other sight‐threatening diseases should also be part of these study populations, as they are part of cataract populations in daily clinical practice. We assume that most authors excluded ophthalmic comorbidities to reduce noise, especially for visual acuity data, and make results comparable.
All of the study authors describe modern standard cataract surgery with implantation of an IOL. Nevertheless, many of the IOLs mentioned (including most PMMA IOLs) in the included studies are no longer being used and are unavailable on the market. Some of the included studies were published more than 15 years ago. In the meantime, scientists and manufacturers have produced advances in IOL technology. As a result, many new foldable IOLs with better biocompatibility, lower PCO rates and possibly with better visual acuity are on the market today.
We found that, independent of the IOL material and design, only very few Nd:YAG capsulotomies were reported one year after cataract surgery. We decided to include longer term follow‐ups at two, three, five, nine and 12 years, in addition to our initially defined primary time point. At these later time points, significant differences concerning PCO formation between different IOL designs could be highlighted. The effect of higher PCO proportion with longer postoperative follow‐ups corresponds well with clinical experience, because ‐ with modern IOL designs ‐ only very few laser treatments need to be performed within the first year of cataract surgery.
Only two studies reported cases of dysphotopsia. In both studies, the number of participants affected was low and equal for both IOLs with sharp and round optic edge design. Nevertheless, according to the literature, the number of people experiencing optical side effects associated with IOL material and design is high, and dysphotopsia in pseudophakic eyes is a leading cause of low postoperative patient satisfaction (Kinard 2013). Therefore, the studies included in this review – which focused on objective outcomes such as PCO and capsulotomy rates – may not sufficiently capture subjective outcomes.
Quality of the evidence
The studies were poorly reported in general, and several studies were at risk of bias in one or more domain. In general, we downgraded for risk of bias. Some of the outcomes were imprecisely measured, particularly at shorter follow‐up times with fewer events. The results were largely consistent. We did not have enough studies to reliably assess publication bias.
Potential biases in the review process
Our prespecified primary outcome was the proportion of eyes with Nd:YAG capsulotomy one year after surgery. However, the number of events at this time point was low, and so we made a post hoc decision to include and present data for longer time periods. We think this is justified but it should be noted as a post hoc decision.
The majority of studies included in this review were within‐person studies whereby one eye received one type of lens and the other eye received the other type. This is a paired design and, as we specified in our methods, ideally the included studies should have reported a paired analysis. However, a paired analysis was not done consistently for all outcomes. We have used the data as reported. This is a conservative approach ‐ confidence intervals are wider than if a paired analysis had been done.
Agreements and disagreements with other studies or reviews
Other reviews and meta‐analyses have shown similar findings to the current review; namely, that IOLs with sharp optic edges are associated with less PCO, lower laser capsulotomy rates and better visual acuity compared with IOLs with round optic edges (Cheng 2007). A recent network meta‐analysis sponsored by Alcon compared sharp‐edged lens Acrysof with other lenses and concluded that this sharp‐edged lens had a lower risk of Nd:YAG capsulotomy compared to all the other round‐edged IOLs (Thom 2019). This network meta‐analysis included more studies than the current review. This is because we restricted our search to studies that formally compared optic edge design.
Risk of bias summary: review authors' judgements about each risk of bias item for each included study
Comparison 1: Sharp versus round optic design, Outcome 1: Nd:YAG capsulotomy at one year
Comparison 1: Sharp versus round optic design, Outcome 2: Nd:YAG capsulotomy at two years
Comparison 1: Sharp versus round optic design, Outcome 3: Nd:YAG capsulotomy at three years
Comparison 1: Sharp versus round optic design, Outcome 4: Nd:YAG capsulotomy at five years
Comparison 1: Sharp versus round optic design, Outcome 5: Nd:YAG capsulotomy at 9 years
Comparison 1: Sharp versus round optic design, Outcome 6: Nd:YAG capsulotomy at 12 years
| PCO score | ||
| Study | Method of measuring PCO and method of dealing with people who had Nd:YAG capsulotomy | Mean difference between sharp and round edge (95% CI or P value) |
|---|---|---|
| Buehl 2005 | AQUA Score 0=clear capsule to 10=exceptionally severe PCO | Mean difference (P value) 1 year: ‐1.18 (P < 0.05) 2 years: ‐1.38 (P < 0.05) 3 years: ‐1.36 (P < 0.05) including YAG cases 3 years: ‐1.05 (P < 0.05) excluding YAG cases |
| Buehl 2007 | AQUA Score 0=clear capsule to 10=exceptionally severe PCO | Mean difference (P value) Excluding YAG cases 1 year: ‐0.83 (P < 0.05) 2 years: ‐0.86 (P < 0.05) 3 years: ‐0.88 (P < 0.05) Including YAG cases 2 years: ‐1.07 (P < 0.05) 3 years: ‐1.4 (P < 0.05) |
| Findl 2005 | AQUA Score 0=clear capsule to 10=exceptionally severe PCO | Mean within‐person differences (95% CI) 1 year: ‐1.29 (‐0.51 to ‐2.07) 3 years: ‐2.31 (‐1.62 to ‐3.00) 5 years: ‐2.83 (‐1.66 to ‐4.00) |
| Haripriya 2017 | EPCO Score Density of opacification (0 to 4) times fractional area. Missing data due to YAG imputed (how?) | Mean difference (P value) 1 year: ‐0.37 2 years: ‐0.056 (P < 0.05) 3 years: ‐0.362 (P < 0.05) 4 years: ‐0.432 (P < 0.05) 5 years: ‐0.673 (P < 0.05) 9 years: ‐0.999 (P < 0.05) |
| Hayashi 2005 | PCO value calculated by subtracting the scattering light density of the anterior IOL surface area from that of the posterior capsule area and averaged over 4 meridians. | "The PCO value in the sharp‐edge group was significantly less than that in the rounded‐edge group throughout the follow‐up period". |
| Mester 2004 | EPCO Score | Median within‐person difference (range) 1 year: ‐0.001 (‐0.855 to 0.328) 3 years: ‐0.069 (‐0.874 to 0.442) |
| Sacu 2005 | AQUA score | Mean difference (P value) 5 years: ‐1.2 (P = 0.001). |
| Shah 2007 | PCO score based on the percentage of the area within the capsulorhexis showing opacity using the POCO system | Median PCO 1 year:
"The square‐edged IOL eyes had significantly less PCO than the round‐edged IOL eyes (p<0.001, Wilcoxon signed rank test)." 2 years:
"The square‐edged IOL eyes had significantly less PCO than the round‐edged IOL eyes (p=0.006, Wilcoxon signed rank test)." |
| Sundelin 2005 | EPCO | "There was a statistically significant difference between the EPCO values for the two different IOLs at 2 (p=0.00014) and 3 years (p=0.02), but not at 1 year." |
| Wejde 2003 | EPCO | 2 years:
5 years:
12 years (Rönbeck 2014)
(note: HSM PMMA and Silicone IOLs were round‐edged, AcrySof group was sharp‐edged). |
Comparison 1: Sharp versus round optic design, Outcome 7: PCO score
Comparison 1: Sharp versus round optic design, Outcome 8: BCDVA (logMAR) at one year
Comparison 1: Sharp versus round optic design, Outcome 9: BCDVA (logMAR) at two years
Comparison 1: Sharp versus round optic design, Outcome 10: BCDVA (logMAR) at three years
Comparison 1: Sharp versus round optic design, Outcome 11: BCDVA (logMAR) at five years
| Sharp‐edged compared to round‐edged IOLS for the prevention of posterior capsule opacification after cataract surgery | ||||||
| Patient or population: people undergoing cataract surgery | ||||||
| Outcomes | Anticipated absolute effects* (95% CI) | Relative effect | № of participants | Certainty of the evidence | Comments | |
|---|---|---|---|---|---|---|
| Risk with round‐edged IOL | Risk with sharp‐edged IOL | |||||
| Nd:YAG capsulotomy at one year | Study population | OR 0.30 | 742 | ⊕⊕⊝⊝ | At 2 years: RR 0.35 (0.16 to 0.80); 703 eyes (6 studies); 89 fewer cases per 1000. At 3 years: RR 0.21 (0.11 to 0.41); 538 eyes (6 studies); 170 fewer cases per 1000. At 5 years: RR 0.21 (0.10 to 0.45); 306 eyes (4 studies); 331 fewer cases per 1000. Data at 9 years and 12 years were only available from one study. | |
| 10 per 1000 | 3 per 1000 | |||||
| Posterior capsule opacification score at one year As measured by AQUA score, Evaluation of PCO (EPCO) score and other methods of quantifying PCO | All studies consistently reported a lower average PCO score (of the order of 0.5 to 3 units) with sharp‐edged IOLs compared with round‐edged IOLs | (10 RCTs) | ⊕⊕⊕⊝ | It was not possible to pool these data due to the way they were reported | ||
| BCDVA at one year, measured using logMAR score (range ‐0.3 to 1.3, low scores = better vision, a score of 0 is equivalent to 6/6 or 20/20 vision) | The mean logMAR BCDVA at one year ranged from 0 to 0.13 | MD 0.06 logMAR units lower (better) | ‐ | 153 | ⊕⊕⊝⊝ | At 2 years: MD ‐0.01 logMAR (‐0.05 to 0.02; 2 studies, 311 eyes. At 3 years MD ‐0.09 logMAR (‐0.22 to 0.03; 2 studies, 117 eyes. Data at 5 years were only available from one study. |
| Quality of life | None of the studies reported this outcome | |||||
| Adverse effects | Adverse effects were reported inconsistently. There was no evidence of any important differences in Intraocular complications, entopic phenomena, IOL tilt/decentration or anterior capsule fibrosis between groups. | ⊕⊝⊝⊝ VERY LOWa,c | ‐ | |||
| *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 | ||||||
| aDowngraded for risk of bias (‐1): only 2 studies at low risk of bias in all domains | ||||||
| Study | Country | Number of people | Number of eyes | Type of study | Type of lens material | |
|---|---|---|---|---|---|---|
| 1 | Austria | 53 | 106 | Within‐person | Acrylic | |
| 2 | Austria | 52 | 104 | Within‐person | Silicone | |
| 3 | Austria | 32 | 64 | Within‐person | PMMA | |
| 4 | India | 100 | 200 | Within‐person | PMMA | |
| 5 | Japan | 75 | 150 | Within‐person | Acrylic | |
| 6 | Germany | 288 | 576 | Within‐person | Silicone | |
| 7 | Austria | 51 | 102 | Within‐person | Silicone | |
| 8 | India, UK | 118 | 236 | Within‐person | PMMA | |
| 9 | Sweden | 116 | 116 | Parallel group | Silicone | |
| 10 | Sweden | 180 | 180 | Parallel group | Acrylic, PMMA, silicone | |
| Total | 1065 | 1834 | ||||
| PMMA: polymethyl methacrylate | ||||||
| Including all studies | Excluding studies at high risk of selection, performance or detection bias | |||||
|---|---|---|---|---|---|---|
| Analysis | Number of studies | Effect estimate | 95% CI | Number of studies | Effect estimate | 95% CI |
| Analysis 1.1 Nd:YAG capsulotomy at one year | 6 | Peto OR 0.30 | 0.05 to 1.74 | 4 | Peto OR 0.14 | 0.00 to 6.82 |
| Analysis 1.2 Nd:YAG capsulotomy at two years | 6 | RR 0.35 | 0.16 to 0.80 | 4 | RR 0.37 | 0.12 to 1.18 |
| Analysis 1.3 Nd:YAG capsulotomy at three years | 6 | RR 0.21 | 0.11 to 0.41 | 3 | RR 0.19 | 0.07 to 0.52 |
| Analysis 1.4 Nd:YAG capsulotomy at five years | 4 | RR 0.21 | 0.10 to 0.45 | 2 | RR 0.30 | 0.13 to 0.72 |
| Analysis 1.5 Nd:YAG capsulotomy at nine years | Not applicable only one study | ‐ | ‐ | ‐ | ‐ | ‐ |
| Analysis 1.6 Nd:YAG capsulotomy at 12 years | Not applicable only one study | ‐ | ‐ | ‐ | ‐ | ‐ |
| Analysis 1.8 BCDVA at one year | 2 | MD ‐0.06 logMAR | ‐0.12 to 0 | 1 | MD ‐0.06 logMAR | ‐0.14 to 0.02 |
| Analysis 1.8 BCDVA at two years | 2 | MD ‐0.01 logMAR | ‐0.05 to 0.02 | 1 | MD ‐0.01 logMAR | ‐0.04 to 0.03 |
| Analysis 1.8 BCDVA at three years | 2 | MD ‐0.09 logMAR | ‐0.22 to 0.03 | 1 | MD 0.03 logMAR | ‐0.19 to 0.25 |
| Analysis 1.8 BCDVA at five years | Not applicable only one study | ‐ | ‐ | ‐ | ‐ | ‐ |
| CI: confidence interval; OR: odds ratio; RR: risk ratio; MD: mean difference | ||||||
| Study | Intraoperative surgical complications | Entopic phenomena (e.g. negative dysphotopsia) | Clinically relevant IOL tilt and decentration | Distinctive anterior capsule fibrosis and capsule phimosis |
|---|---|---|---|---|
| Not reported | "Concerning subjective visual symptoms, there were no differences between the groups. There were 1 case of day glare in the AR40 group and 2 cases in the AR40e group and 2 cases of halos and/or photophobia in each group." | "There was no significant difference between the 2 groups at any time for decentration of the IOL or capsulorhexis–IOL overlap (1 case of buttonholing in each group)." | "There was no difference between groups in cellular ongrowth on the IOL from the anterior capsule. There were 34 cases in both groups after 1 month, 6 cases in both groups after 1 year, and 1 case in both groups after 2 years; however, there was no case in either group 3 years after surgery." | |
| "There were no surgical complications that would have led to patient exclusion." | "There was also no statistically significant difference in subjective visual symptoms described by the patients (one case of day glare in the SI40 group and two cases in the Clariflex group; three cases of halos and/or photophobia in each group)." | "Two patients were excluded after surgery, one patient because of endophthalmitis and one patient because of pronounced fibrosis‐induced decentration of an SI40 IOL that required resurgery." | Not reported | |
| "There were no surgical complications that would have led to patient exclusion." | Not reported | "We found no significant difference between the 2 groups at any time regarding decentration of the IOL (1 case in the sharp‐edge group), rhexis/IOL overlap (1 case of buttonholing in the round‐edge group), ..." | "We found no significant difference between the 2 groups at any time regarding... cases of cellular ongrowth onto the anterior IOL surface from the rhexis margin (chi‐square tests, all Ps % 0.05)." | |
| "There were no surgical complications, such as posterior capsular rupture, leading to subject disqualification." | Not reported | Not reported | Not reported | |
| "Furthermore, 1 surgeon performed all of the operations, and there were no surgical complications" | Not reported | Not reported | Not reported | |
| "Of the 288 randomized patients, 41 (14%) had to be excluded from | Not reported | "In all eyes, slitlamp examination revealed correct IOL positioning in the bag with no significant tilt or decentration." | "There was no difference between the 3 IOLs in anterior chamber cells, flare, or chronic inflammatory cells on the IOL surface by slitlamp examination." | |
| "There were no surgical complications that would have led to patient exclusion." | Not reported | Not reported | "One week postoperatively, full circumferential overlap of the peripheral optic by the anterior capsule leaf as a prerequisite for capsular bend formation was verified in all cases. In cases in which part of the optic circumference was exposed at the 5‐year follow‐up examination (one eye in the sharp edge and four eyes in the round edge optic group, P = 0.13), this had resulted from secondary buttonholing generally occurring during months 1 and 6." | |
| "No serious surgical complications were seen." | Not reported | 6/107 in round‐edge group had decentered IOL compared with 4/107 in the sharp‐edge group. | 1 case of iris atrophy, 1 case of posterior synechiae, 41 cases capsulorhexis contact with IOL surface in round‐edge group. 51 cases of capsulorhexis contact with IOL surface in square‐edge IOL group. | |
| "Surgical complications" 1 case in each group given as a reason for incomplete follow‐up. | Not reported | "At the 2‐month visit it was noticed that in some patients the optic was partly outside the capsulorhexis rim. This was due either to an asymmetric or overly large capsulorhexis, or to the IOL being slightly decentred in the capsular bag. In many patients the pupil could not be dilated sufficiently to inspect the entire optic/ capsulorhexis relationship at the 2‐month visit. A total of 113 patients attended the 2‐month visit; in 55 patients the entire capsulorhexis was on the optic, in 17 patients some part of the capsulorhexis was outside the optic edge, and in 41 patients the optic/capsulorhexis relationship could not be assessed" | Not reported | |
| Not reported | Not reported | Not reported | Not reported |
| Outcome or subgroup title | No. of studies | No. of participants | Statistical method | Effect size |
| 1.1 Nd:YAG capsulotomy at one year Show forest plot | 6 | 742 | Peto Odds Ratio (Peto, Fixed, 95% CI) | 0.30 [0.05, 1.74] |
| 1.1.1 Acrylic | 1 | 92 | Peto Odds Ratio (Peto, Fixed, 95% CI) | 0.14 [0.00, 6.82] |
| 1.1.2 PMMA | 3 | 366 | Peto Odds Ratio (Peto, Fixed, 95% CI) | 0.14 [0.00, 6.82] |
| 1.1.3 Silicone | 2 | 284 | Peto Odds Ratio (Peto, Fixed, 95% CI) | 0.51 [0.05, 4.96] |
| 1.2 Nd:YAG capsulotomy at two years Show forest plot | 6 | 703 | Risk Ratio (M‐H, Random, 95% CI) | 0.35 [0.16, 0.80] |
| 1.2.1 Acrylic | 2 | 204 | Risk Ratio (M‐H, Random, 95% CI) | 0.13 [0.02, 0.72] |
| 1.2.2 PMMA | 2 | 298 | Risk Ratio (M‐H, Random, 95% CI) | 0.34 [0.02, 4.80] |
| 1.2.3 Silicone | 1 | 68 | Risk Ratio (M‐H, Random, 95% CI) | 0.25 [0.03, 2.12] |
| 1.2.4 Mixed materials | 1 | 133 | Risk Ratio (M‐H, Random, 95% CI) | 0.38 [0.12, 1.19] |
| 1.3 Nd:YAG capsulotomy at three years Show forest plot | 6 | 538 | Risk Ratio (M‐H, Random, 95% CI) | 0.21 [0.11, 0.41] |
| 1.3.1 Acrylic | 1 | 62 | Risk Ratio (M‐H, Random, 95% CI) | 0.08 [0.01, 0.60] |
| 1.3.2 PMMA | 2 | 116 | Risk Ratio (M‐H, Random, 95% CI) | 0.15 [0.02, 1.13] |
| 1.3.3 Silicone | 3 | 360 | Risk Ratio (M‐H, Random, 95% CI) | 0.27 [0.12, 0.62] |
| 1.4 Nd:YAG capsulotomy at five years Show forest plot | 4 | 306 | Risk Ratio (M‐H, Random, 95% CI) | 0.21 [0.10, 0.45] |
| 1.4.1 PMMA | 2 | 106 | Risk Ratio (M‐H, Random, 95% CI) | 0.13 [0.01, 2.20] |
| 1.4.2 Silicone | 1 | 48 | Risk Ratio (M‐H, Random, 95% CI) | 0.11 [0.01, 1.96] |
| 1.4.3 Mixed material | 1 | 152 | Risk Ratio (M‐H, Random, 95% CI) | 0.19 [0.07, 0.50] |
| 1.5 Nd:YAG capsulotomy at 9 years Show forest plot | 1 | Risk Ratio (M‐H, Fixed, 95% CI) | Totals not selected | |
| 1.6 Nd:YAG capsulotomy at 12 years Show forest plot | 1 | Risk Ratio (M‐H, Fixed, 95% CI) | Totals not selected | |
| 1.7 PCO score Show forest plot | 10 | Other data | No numeric data | |
| 1.8 BCDVA (logMAR) at one year Show forest plot | 2 | 153 | Mean Difference (IV, Fixed, 95% CI) | ‐0.06 [‐0.12, 0.00] |
| 1.8.1 PMMA | 1 | 46 | Mean Difference (IV, Fixed, 95% CI) | ‐0.06 [‐0.14, 0.02] |
| 1.8.2 Silicone | 1 | 107 | Mean Difference (IV, Fixed, 95% CI) | ‐0.06 [‐0.16, 0.04] |
| 1.9 BCDVA (logMAR) at two years Show forest plot | 2 | 311 | Mean Difference (IV, Fixed, 95% CI) | ‐0.01 [‐0.05, 0.02] |
| 1.9.1 PMMA | 1 | 214 | Mean Difference (IV, Fixed, 95% CI) | ‐0.01 [‐0.04, 0.03] |
| 1.9.2 Silicone | 1 | 97 | Mean Difference (IV, Fixed, 95% CI) | ‐0.11 [‐0.25, 0.03] |
| 1.10 BCDVA (logMAR) at three years Show forest plot | 2 | 117 | Mean Difference (IV, Fixed, 95% CI) | ‐0.09 [‐0.22, 0.03] |
| 1.10.1 PMMA | 1 | 34 | Mean Difference (IV, Fixed, 95% CI) | 0.03 [‐0.19, 0.25] |
| 1.10.2 Silicone | 1 | 83 | Mean Difference (IV, Fixed, 95% CI) | ‐0.15 [‐0.30, 0.00] |
| 1.11 BCDVA (logMAR) at five years Show forest plot | 1 | Mean Difference (IV, Fixed, 95% CI) | Totals not selected | |
