Print version ISSN 0042-9686
Bull World Health Organ vol.78 n.3 Genebra Mar. 2000
Randomized controlled trial of anterior-chamber intraocular lenses in Nepal: long-term follow-up
Essai contrôlé randomisé des implants intraoculaires de chambre antérieure au Népal : suivi à long terme
Ensayo aleatorizado controlado de la implantación de lentes intraoculares de cámara anterior en Nepal: seguimiento a largo plazo
J.R. EvansI; A. HennigII; D. PradhanI; A. FosterIII; R. LagnadoIII; A. PoulsonIII; G.J. JohnsonIII; R.P.L. WormaldI
IGlaxo Department of Ophthalmic Epidemiology, Institute of Ophthalmology (University College London), London, England; and Moorfields Eye Hospital, City Road, London EC1V 2PD, England
IILahan Eye Hospital, Nepal
IIIDepartment of Preventive Ophthalmology, Institute of Ophthalmology (UCL), Bath Street, London, England
Most of the estimated 20 million people who are blind with cataracts live in rural areas of developing countries, where expert surgical resources are scarce. We have studied the use of multiflex open-loop anterior-chamber intraocular lenses (ACIOL) in high-volume low-cost surgery. Between 1992 and 1995, a total of 2000 people attending Lahan Eye Hospital, Nepal, with bilateral cataracts reducing vision to < 6/36 were randomly allocated to receive intracapsular extraction (ICCE) with aphakic spectacles, or ICCE with an ACIOL. We re-examined the cohort (1305/2000, 65%) between November 1996 and April 1997 and report the findings in this article. There were 13 new cases of poor visual outcome (best corrected vision <6/60) arising after one year: 9 in the ACIOL group and 4 in the control group; odds ratio 2.1 (95% confidence interval, 0.599.55). The causes of poor outcome were as follows: ACIOL group retinal detachment (4 cases), cystoid macular oedema (2), epiretinal membrane (1), age-related macular degeneration (1), and late endophthalmitis (1); control group retinal detachment (2 cases), late endophthalmitis (1), and primary open-angle glaucoma with age-related macular degeneration (1). In rural areas of developing countries, well-manufactured multiflex open-loop ACIOLs can be implanted safely by experienced ophthalmologists after routine ICCE, avoiding the disadvantages of aphakic spectacle correction.
Keywords: lenses, intraocular; cataract extraction; treatment outcome; vision, low; randomized controlled trials; Nepal.
On estime à 20 millions le nombre de personnes atteintes de cécité bilatérale due à la cataracte, qui pour la plupart vivent dans des zones rurales de pays en développement où les ressources chirurgicales sont rares. On connaît encore mal le rôle des implants intraoculaires de chambre antérieure (IIO-CA) dans les situations où un important volume dinterventions chirurgicales à coût modique est nécessaire. La présente étude a en conséquence été réalisée pour évaluer la sécurité des IIO-CA à anse ouverte en Multiflex. Nous avions déjà rapporté nos observations sur une cohorte un an après lintervention; nous rapportons maintenant les résultats du suivi ultérieur de cette cohorte.
Au total, 2000 personnes venues consulter au Lahan Eye Hospital, dans le sud du Népal, pour une cataracte bilatérale ayant réduit leur vision à < 6/36 ont été réparties par tirage au sort dans un groupe traité par extraction intracapsulaire (EIC) et correction de laphakie par des lunettes (groupe témoin), ou dans un groupe traité par EIC puis pose dun IIO-CA dans le premier il opéré. Ces personnes ont été revues six semaines et un an après lintervention (résultats déjà rapportés). Le recrutement pour la présente étude a eu lieu entre début février 1992 et fin février 1995. Nous avons réexaminé la cohorte (1305/2000, 65 %) entre novembre 1996 et avril 1997.
Treize nouveaux cas de résultats médiocres sur le plan visuel (meilleure vision corrigée < 6/60) sont survenus dans un délai supérieur à un an après lintervention (9 dans le groupe IIO-CA et 4 dans le groupe témoin) (odds ratio 2,1 ; intervalle de confiance à 95% : 0,59-9,55 ; p = 0,2). Les causes de ces mauvais résultats étaient, dans le groupe IIO-CA : décollement de la rétine (4 cas), dème maculaire cystoïde (2 cas), membrane épirétinienne (1 cas), dégénérescence maculaire liée à lâge (1 cas) et endophtalmie tardive (1 cas), et dans le groupe témoin : décollement de la rétine (2 cas), endophtalmie tardive (1 cas) et glaucome primitif à angle ouvert avec dégénérescence maculaire liée à lâge (1 cas). Aucun cas de décompensation cornéenne na été observé. Davantage de patients présentaient une pression intraoculaire > 30 mmHg dans le groupe IIOAC (5 yeux contre 1 ; p = 0,2); dans ce groupe, on comptait cinq yeux atteints de glaucome secondaire contre 2 dans le groupe témoin.
Cette étude montre que, dans les zones rurales des pays en développement, des IIO-CA à anse ouverte en Multiplex de fabrication correcte peuvent être posés en toute sécurité par des ophtalmologistes expérimentés après extraction intracapsulaire de routine, ce qui évite linconvénient dune correction de laphakie par des lunettes. Bien que lemploi des IIO-CA soit associé à un taux plus élevé duvéite et de glaucome secondaire que lutilisation de verres correcteurs, la proportion de personnes ayant un résultat médiocre sur le plan visuel après intervention chirurgicale est similaire.
Se estima que hay unos 20 millones de personas con ceguera bilateral por cataratas, la mayoría de las cuales viven en zonas rurales de países en desarrollo, donde escasean los recursos quirúrgicos especializados. Actualmente no se conoce con exactitud la utilidad de la implantación de lentes intraoculares de cámara anterior (LIOCA) en las situaciones en que hay que practicar un elevado número de intervenciones quirúrgicas a bajo costo. El presente estudio se emprendió con objeto de evaluar la seguridad de las LIOCA multiflex de asa abierta. Hemos informado anteriormente de los resultados observados en una cohorte al cabo de un año de la intervención quirúrgica. Ahora informamos aquí de los resultados del posterior seguimiento de esa cohorte.
Un total de 2000 personas que acudieron al hospital oftalmológico Lahan, en el sur de Nepal, con cataratas bilaterales que reducían su visión a <6/36 fueron asignados al azar para someterse bien a una extracción intracapsular (EIC) seguida de uso de gafas para afaquia, o bien a EIC con LIOCA en el primer ojo operado. Dichas personas fueron sometidas a seguimiento a las seis semanas y al cabo de un año de la intervención quirúrgica (resultados publicados anteriormente). El reclutamiento para el presente trabajo tuvo lugar entre principios de febrero de 1992 y finales de febrero de 1995. Reexaminamos la cohorte (1305/ 2000, 65%) entre noviembre de 1996 y abril de 1997.
Al cabo de un año detectamos 13 nuevos casos de visión precaria (mejor visión corregida < 6/60): 9 en el grupo sometido a implantación de LIOCA, y 4 en el grupo de control): OR = 2,1 (intervalo de confianza del 95%: 0,59-9,55, P = 0,2). Las causas de visión precaria en esos 13 ojos fueron las siguientes: grupo con LIOCA: desprendimiento de retina (4 casos), edema macular cistoide (2), membrana epirretiniana (1), degeneración macular relacionada con la edad (1), y endoftalmitis tardía (1); y en el grupo de control: desprendimiento de retina (2 casos), endoftalmitis tardía (1), y glaucoma primario de ángulo abierto con degeneración macular relacionada con la edad (1). No se observaron casos de descompensación corneal. Se detectaron más ojos con presión intraocular > 30 mmHg en el grupo con LIOCA (5 frente a 1, P = 0,2); en este grupo, 5 ojos presentaban signos de glaucoma secundario, frente a 2 en el grupo de control.
Los resultados de este estudio respaldan la idea de que, en las zonas rurales de los países en desarrollo, oftalmólogos experimentados pueden implantar sin riesgos LIOCA multiflex de asa abierta tras la EIC de rutina, evitando así los inconvenientes asociados a las gafas de corrección de la afaquia. Aunque la incidencia de uveítis y de glaucoma secundario es mayor con la LIOCA que con las citadas gafas, la proporción de personas con visión precaria tras la cirugía es similar en los dos casos.
Globally, cataract is the most frequent cause of blindness, affecting more than 20 million people (1). Most of these people live in rural areas of developing countries, where there are limited ophthalmologically trained personnel or services.
There are two ways to remove the cataractous lens. It can be extracted in its entirety (intracapsular cataract extraction, ICCE) or the lens contents can be removed leaving the posterior lens capsule in place (extracapsular cataract extraction, ECCE). ICCE is a quicker and simpler technique, but is associated with more complications related to disturbance of the vitreous face. It is still commonly performed in rural Africa and Asia, whereas ECCE is the preferred technique in industrialized countries. The disadvantage of ECCE is that, in a proportion of cases, the lens capsule becomes opaque, requiring further intervention in order to maintain good vision.
It is generally agreed that inserting an intraocular lens into the eye during surgery is a better method of correcting refraction than using spectacles. In situations where surgeons are routinely performing ICCE, it is not possible to use a posterior-chamber intraocular lens (PCIOL). The alternative is an anterior-chamber lens (ACIOL), but due to complications resulting from the use of these types of lenses in the 1970s and early 1980s, surgeons are reluctant to use them. Modern multiflex open-loop ACIOL implants are much improved and are now used by surgeons in industrialized countries when a PCIOL cannot be inserted after ECCE.
We conducted a randomized controlled trial among 2000 patients attending Lahan Eye Hospital, southern Nepal and, one year after surgery, reported results that demonstrated that open-loop ACIOLs can be used safely as a primary implant in developing countries (2). However, it took several years for some of the complications of the old closed-loop ACIOLs to become evident. We therefore re-examined our cohort 1.55 years after surgery in order to determine the incidence and causes of new cases of poor visual outcome arising one year after the surgery. This article presents the main findings of public health interest.
Participants. The study participants were aged 4064 years, lived within selected accessible districts, and attended Lahan Eye Hospital between February 1992 and March 1995 with bilateral cataracts that reduced their vision to < 6/36 in both eyes. The following exclusion criteria were applied: known preexisting ocular disease, hypertension, or diabetes. All the participants (n = 2000) gave their free and informed consent to take part in the study, which was approved by the Medical Research Council of Nepal.
Interventions. The interventions consisted of ICCE with ACIOL versus ICCE with aphakic spectacles (control group). Surgery was performed by one of two experienced ophthalmologists (A.H., D.P.) under local anaesthetic using a 4.5 × loupe. A standard 19.0 or 19.5 dioptre single-piece four-point fixation CILCO Kelman Multiflex III lens was used in the ACIOL group. Control patients were given +11D spectacles.
Follow-up. Follow-up was performed at discharge, 6 weeks, and 1 year after surgery (results reported previously, see (2)). The whole cohort was contacted between November 1996 and April 1997 and asked to return for a further examination. Those not returning for follow-up were visited by an ophthalmologist at home where possible. For anyone not seen at follow-up, we obtained data from any hospital visits more than 1 year after surgery.
Outcome measures. The primary outcome measure was poor vision defined as visual acuity <6/60 in the operated eye. A modified Snellen E chart with 4 optotypes on each line was used to measure visual acuity. Vision was taken as the last line on which at least three optotypes were read correctly. Functional vision (vision used in everyday life, with or without prescribed spectacles), and best corrected vision (vision after refraction, or with a pinhole when refraction could not be performed) were recorded. All patients seen at the hospital received a slit lamp examination and full examination of the retina. Glaucoma was defined as end organ damage (i.e., cupping of the optic disc characteristic of glaucoma) plus an intraocular pressure > 22 mmHg. Glaucoma was termed secondary when the optic discs had been recorded as normal both at discharge and at the 6-week follow-up examination. Secondary glaucoma was also diagnosed in one case where the IOL had been removed, vision was evaluated as no light perception, and the intraocular pressure was 52 mmHg even though the disc could not be seen because of a pupillary membrane; and in a further case with vision evaluated as no light perception and phthisis at the 6-week follow-up, but a pressure of 58 mmHg and corneal oedema at 1 year.
The sample size was estimated for results at one year only (2). For the long-term follow-up reported here, the aim was to re-examine as many of the cohort as possible.
Statistical analysis. All analyses were performed on the basis of intent to treat. Results are reported for the first operated eye only (see Assignment below). The incidence of poor visual outcome (best corrected vision <6/60) after one year was compared between the two groups, taking as the denominator everyone seen at one year after surgery who had good vision (best corrected vision > 6/60). The results at one year showed a higher rate of functional blindness (vision with own correction <3/60) in the control group and an increased incidence of glaucoma in the ACIOL group was hypothesized, so we also analysed these outcomes. Data from hospital visits represent a self-selected sample of the cohort, and therefore are not included in the analyses. All analyses were carried out using Epi Info version 6 (Centers for Disease Control and Prevention, Atlanta, GA, USA) and SAS 6.11 (SAS Institute Inc, Cary, NC, USA).
The unit of randomization was the patient. The first-operated eye was entered into the trial. Treatment allocation was computer-generated and notified in serially numbered sealed opaque envelopes in the operating theatre. No person involved in the study in Nepal was told how the randomization list was prepared and no copy of the list was kept in Nepal. ACIOL patients requesting a second eye operation received an ECCE procedure with PCIOL; the second eyes of control group patients underwent ECCE procedure with spectacles.
As there were obvious physical differences between the treatment groups the trial was not masked.
Fig. 1 shows the trial profile. A total of 39 people died before the 1-year follow-up (12 in the ACIOL group, 27 controls), and a further 87 died between the 1-year follow-up and the examination reported here (36 ACIOL, 51 controls), leaving 1874 in the cohort alive (954 ACIOL, 920 controls) of whom 1305 were examined (677 (71%) in ACIOL, and 628 (68%) controls). The median follow-up time was 3.1 years (range, 1.85.1 years). Most (97%) of the examinations were carried out at the hospital and the remainder at home. Table 1 shows the distribution of follow-up by year of recruitment. Follow-up was similar between the two groups, except for the last year of the study where more of the ACIOL group than the control group were followed up (ACIOL 74%; control 64%; c² test = 9.2, P = 0.002). Of the 569 individuals not seen at the current examination, information on hospital visits after 1-year follow-up was available for 51 (31 ACIOL, 20 controls). The median time since surgery for these visits was 2.3 years (range, 1.34.7 years).
Outcome in individuals who had good vision 1 year after surgery
At 1-year follow-up there were 1782 participants whose best-corrected vision was > 6/60 (ACIOL 893, controls 889). Of these, 1228 (69%) were examined at the 1.55-year follow-up (ACIOL 632 (71%); controls 596 (67%)). There were 13 new cases of poor visual outcome (best-corrected vision <6/60) in the cohort after the 1-year follow-up, of which 9/632 (1.4%) were in the ACIOL group and 4/596 (0.7%) in the control group: odds ratio (OR) = 2.1; 95% confidence interval (CI) = 0.599.55, P = 0.2. The causes of poor visual outcome in these 13 eyes were as follows: ACIOL group retinal detachment (4 cases), cystoid macular oedema (2), epiretinal membrane (1), age-related macular degeneration (1), and late endophthalmitis (1); and in the control groupretinal detachment (2 cases), late endophthalmitis (1), and primary open-angle glaucoma with macular degeneration (1).
Hospital visit data were available for a further 45 individuals (29 ACIOL, 16 controls). All had vision > 6/60 with the exception of one person in the control group who had developed a retinal detachment 17 months after surgery.
Of 141 people seen more than 4.5 years after surgery, 3 had visual loss that occurred after one year (2.1%; 95% CI = 0.55 6.6%): 2/76 eyes in the ACIOL group and 1/65 in the control group. The causes of visual loss were retinal detachment (1 ACIOL, 1 control) and age-related macular degeneration (1 ACIOL).
Outcome in individuals who had poor vision 1 year after surgery
At the 1-year follow-up, 44 individuals were identified as having best-corrected vision <6/60 (24 ACIOL, 20 controls). Of these, 3 died and 24 were re-examined; 7 of the latter had improved to a corrected vision > 6/60 at the current examination (6 ACIOL, 1 controls). Three of the 17 not seen had visited the hospital more than 1 year after surgery (1 ACIOL, 2 controls) and all of them had vision > 6/60 in the trial eye.
The causes of poor visual outcome at 1 year among individuals whose vision subsequently improved were as follows: in the ACIOL group uveitis (6) and cystoid macula oedema (1); and in the control group uveitis (1), optic atrophy (1), and high myopia (1).
Table 2 shows the number of cases of glaucoma and high intraocular pressure at the clinical examination 1.55 years after surgery. Five eyes in the ACIOL group and 2 in the control group had secondary glaucoma (OR = 2.32, exact 95% CI = 0.3824, Fishers exact P = 0.45). Although mean intraocular pressure was similar in the two groups, 11 eyes in the ACIOL group had intraocular pressure > 22 mmHg, compared with 7 eyes in the control group (OR = 1.46; 95% CI = 0.524.2, P = 0.43); 5 versus 1 had intraocular pressure > 30 mmHg (OR = 4.66, exact 95% CI = 0.52221, Fishers exact P = 0.22).
Table 3 shows the proportion of eyes with a functional acuity <3/60, by year of follow-up. Those in the aphakic glasses group were 3.5 times more likely to be functionally blind than those in the ACIOL group (6.4% versus 1.9%). Of the 628 individuals questioned in the control group, 18 (2.9%) no longer had a pair of aphakic glasses and 22 (3.5%) owned but did not wear glasses. A total of 21% of the control group (126) had received a replacement pair of spectacles since the operation.
The results of our study show that 1.55 years after surgery, the number of new cases of poor visual outcome among individuals receiving an ACIOL is small (1.4%, 9 cases). There were fewer cases of new visual loss in the control group after 1 year (0.7%, 4 cases). However, in view of the small numbers of events, there is no evidence to suggest an increase in poor visual outcome in the lens group compared with the aphakic spectacles group: OR = 2.1; 95% CI = 0.599.55.
Previously we have reviewed the evidence for use of ACIOLs (2). Only one other trial has evaluated ACIOLs (3). Previous studies have been case series that were too small to give good estimates of the incidence of sight-threatening complications. To our knowledge, the present trial is the largest and longest follow-up of ACIOLs that has been carried out. We did not find any indication that lens-related problems increased in frequency over time: 76 of the ACIOL group were seen more than 4.5 years after surgery, and only two had had new visual loss since the 1-year follow-up (1 due to retinal detachment, 1 due to age-related macular degeneration).
Owing to the longer period between surgery and 1.55-year follow-up, we did not obtain such a good response rate as at the 1-year follow-up, when we were able to examine 91% of the cohort. At the 1.55-year follow-up we examined 68% of the original cohort randomized to receive an ACIOL and 63% of the control group. In order to investigate whether any bias had arisen as a result of the lower response rate, we repeated the analyses for the four districts close to the hospital, where we were able to obtain a response rate of 84%. For this group, 3/257 ACIOL and 2/234 controls had new visual loss after 1 year (OR = 1.37; 95% CI = 0.1616.53, Fishers exact P = 1.0).
An unexpected finding was that more of the control group died than the intervention group (48 versus 78; c² test = 7.75, P = 0.005). We do not have an explanation for this: our randomization schedule was carefully concealed and baseline analyses showed that the two groups were comparable with respect to age, sex, occupation, and preoperative visual acuity (2). We enquired about visual status in people who had died and were able to obtain information on 44/48 of the ACIOL group and 74/78 of the control group. There were no reports of blindness in either of these groups; 1 individual in the ACIOL group and 2 individuals in the control group were reported to have had some problems with their vision. We think it unlikely that the difference in mortality can be attributed to differences in postoperative visual acuity and it may be a chance finding. We have only examined visual outcomes of cataract surgery, however, and cannot exclude an effect of intraocular lens implantation on other outcomes; for example, social and mental functioning.
Our study provides detailed information on the outcome for 1002 eyes that received an ACIOL. One year after surgery, 2.6% (24 cases) had a poor visual outcome, i.e. had best corrected acuity < 6/60 (2). The causes of visual loss were uveitis/ secondary glaucoma (13 cases), endophthalmitis (4 cases all operated on the same day), cystoid macular oedema (2 cases) and corneal ulcer. A total of 15 of these 24 cases were reexamined at the 1.55 year follow-up, and 7 were found to have improved to > 6/60. The acuity of a further 9 individuals dropped to < 6/60 after 1 year: 4 from retinal detachment, 2 from cystoid macular oedema, 1 from late endophthalmitis (due to suture removal and unrelated to the intraocular lens), 1 from epiretinal membrane and 1 from age-related macular degeneration.
There was some evidence that raised intraocular pressure and secondary glaucoma occurred more frequently in the ACIOL group than the control group. However, none of the differences reached conventional statistical significance. A total of 11/674 ACIOL eyes had an intraocular pressure > 22 mmHg; 2 had primary and 5 had secondary glaucoma. There was no case of corneal decompensation in the ACIOL group.
In summary, the major problem in the ACIOL eyes was uveitis, which occurred only in the first year, and caused a poor visual outcome in 13 eyes, of which 6 improved to > 6/60 after 1 year.
This study was undertaken to determine whether ACIOLs are a safe alternative to aphakic spectacles in situations were ICCE procedures are performed. We believe that the results demonstrate that in the hands of experienced surgeons the open-loop ACIOL lens is safe. Furthermore, the study has demonstrated that functional blindness occurs more commonly in people receiving aphakic spectacles. This is due to lost and broken spectacles, despite the policy of the study hospital of making replacement spectacles readily available and affordable to all patients.
In conclusion, we believe that an ACIOL is a preferred alternative to aphakic spectacles for patients who are undergoing an ICCE procedure, providing the surgeon has received training in the technique.
We are grateful to all the patients who took part in the trial and without whom the study could not have been conducted. We acknowledge the contribution of the following individuals: Bindeshwar Mahato, who coordinated the study in Lahan; Kedar Timilsina, Janardan Khatiwada, Ganesh Prasad Updahaya, Mahendra Mahato and Ramchandra Pasman, who assisted in the recruitment of patients and execution of the trial including data collection; Kristina Hennig, who managed the data entry and databases in Kathmandu; and Richard Gregson, who provided clinical support to the surgeons. At various times over the course of the study, members of the Steering Committee included Noel Rice (Chair), Dominique Négrel (WHO), Bjorn Thylefors (WHO), Christian Garms (CBM), Dr Pokhrel and Professor Richard Hayes, who also advised on the statistical analysis of the study. ALCON (UK) generously donated the lenses. The study was funded by Christoffel Blindenmission/ Christian Blind Mission International.
The study was executed by Albrecht Hennig and Damodar Pradhan, who carried out all the operations. Ruby Lagnado and Arabella Poulson conducted the clinical examinations. They were assisted in the design and execution of the study by (in alphabetical order) Jennifer Evans, Allen Foster, Gordon Johnson and Richard Wormald. The data were analysed by Jennifer Evans, Albrecht Hennig and Allen Foster.
1. Thylefors B et al. Global data on blindness. Bulletin of the World Health Organization, 1995, 73: 115121. [ Links ]
2. Hennig A et al. Randomised controlled trial of anterior-chamber intraocular lenses. Lancet, 1997, 349: 11291133. [ Links ]
3. Snellingen T et al. The South Asian Cataract Management Study. Complications, vision outcomes and corneal endothelium cell loss in a randomised multicenter clinical trial comparing cataract extraction with and without anterior chamber intraocular lens. Ophthalmology, 2000, 107: 231240. [ Links ]