Dear Editor

We thank Dr Rabiah for his helpful comments. Unfortunately Dr Rabiah's article [1] appeared in print after we have already submitted our paper for publication.

We designed this study principally to look at the influence of cataract surgery during the first month of life on the risk of subsequent glaucoma, rather than to compare the risk of glaucoma with surgery carried out before and after the first year of life [2]. Our study was motivated by a recent report from Watts et al. [3] which suggested that surgery within the first 2 weeks of life is paradoxically associated with lower risk of glaucoma. This report contradicted a previous study by Magnusson et al. [4].

For bilateral cataracts, eyes operated before 1 month of age were associated with a significantly higher risk of glaucoma by 5 years after operation and there was no significant difference in 5-year visual outcome between patients operated before and shortly after one month old. However, this observation should not be regarded as an absolute risk-free threshold as we still had a few cases of glaucoma in patients who had operations carried out between 2nd and 5th month of life.

For unilateral cataracts and unilateral PHPVs, the small number of cases within these two subgroups has prevented us from confirming any possible risk associated with early surgery. We included PHPV in this study, as Johnson et al. [5] had previously demonstrated that patients with PHPV had similar level of risk of glaucoma as patients with congenital cataract.

As stated in our paper, apart from 6 patients lost to follow up, all patients had at least 5-year post-operative follow-up. Kaplan-Meier survival analysis enabled us to perform risk analysis from a cohort of subjects with non-uniform length of follow-up.

Many of Dr. Rabiah's criticisms are related to the inherent limitations of retrospective case notes studies. We were keen to avoid introducing bias by 'artificially' recreating standard diagnostic criteria for glaucoma from hindsight. We would emphasise that in all cases, the diagnosis of glaucoma was made by consultant paediatric ophthalmologists at our unit (including 2 of the co-authors DT, IRE) and all our post- operative patients were monitored regularly with tonometry, refraction, cornea measurements and fundoscopy.

Given the limited number of cases and missing data for some key clinical variables (e.g. HCD measurement, as it was not standard practice at our unit before 1996 to record horizontal corneal diameter pre- operatively in all cases), we were unable to conduct a rigorous multi- factorial analysis.

A multi-centre randomized controlled trial may be the way forward. However it will require extensive planning before the implementation of such a costly and time-consuming undertaking. The question remains as to how best to manage a newborn with bilateral visually significant cataracts? We believe that the results of our study serve to inform evidence-based decision-making.

References

(1) Frequency and predictors of glaucoma after pediatric cataract surgery. Rabiah PK. Am J Ophthalmol. 2004 Jan;137(1):30-7.

(2) Mills MD, Robb RM. Glaucoma following childhood cataract surgery. J Pediatr Ophthalmol Strabismus 1994;31:355-60

(3) Watts P, Abdolell M, Levin AV. Complications in infants undergoing surgery for congenital cataract in the first 12 weeks of life: is early surgery better? J AAPOS 2003;7:81-5

(4) Magnusson G, Abrahamsson M, Sjostrand J. Glaucoma following congenital cataract surgery: an 18-year longitudinal follow-up. Acta Ophthalmol Scand

(5) Johnson CP, Keech RV. Prevalence of glaucoma after surgery for PHPV and infantile cataracts. J Pediatr Ophthalmol Strabismus 1996;33:14-17.

Dear Editor

A recent report stated it might be prudent to delay paediatric cataract surgery, at least for bilateral cases, until 4 weeks of age to lower the risk of later glaucoma,[1] and this suggestion was fully supported in an accompanying editorial [2]. Such a suggestion that could have an impact on the practice patterns of ophthalmologists and consequently the care of children with cataract should have strong data to back it. For the following reasons, the data presented in the report do not seem to rise to that level.

First, as postcataract surgery glaucoma is often a delayed complication, it is critical to know how long study patients were followed; and aside from what can be inferred from the presented statements and survival analyses, such data are not provided. Second, study patients with glaucoma often presented with the isolated finding of elevated intraocular pressure, but no mention is made of whether all study patients were screened for glaucoma using intraocular pressure measurement (or other means) at any time postoperatively, let alone for the duration of follow-up. This is important given the challenges of detecting glaucoma in young children, which the authors acknowledge. Ultimately, glaucoma was diagnosed if a study clinician elected to treat for glaucoma; but specific, reliable, and mutually agreed upon parameters for making the diagnosis were not defined. Third, from a study design perspective, the decision to include a subgroup of patients (those with persistent hyperplastic primary vitreous) in the study solely on the basis of the glaucoma frequency within that subgroup (as determined from a preliminary analysis) is problematic. Fourth, while other risk factors aside from a young age at surgery have been associated with the development of aphakic glaucoma, such as microcornea and secondary membrane surgery, no mention is made of whether any patients in the study had such features. It might be assumed that patients with microcornea were excluded from the study since patients with “other ocular malformations” were excluded; if this is the case, however, it would seem unusual that the authors would elect to include patients with persistent hyperplastic primary vitreous (who generally have microphthalmos, not to mention other abnormalities). In any case, no multivariable statistical technique was employed that would take into account other important potential risks for glaucoma aside from simply an early age at cataract surgery; without such a multivariable analysis, the validity of the authors’ conclusions would rest on the presumption that other potential risks for glaucoma aside from age at surgery were equally distributed among all age groups studied. Fifth, the study purports to compare the glaucoma frequencies among those patients operated on in the first month of life to those operated on later during the first year of life, yet relatively few patients were operated on after 3 months (and especially after 5 months) of age.

A review of presently available reports tends to implicate early cataract surgery in the pathogenesis of aphakic glaucoma, but exactly what “early” is has not been absolutely established. In their discussion, the authors cite prior studies indicating a higher risk of glaucoma when surgery is performed in the first month of life; these studies, as well as others indicating a high risk of glaucoma extending up to a later age at surgery or indicating no relationship to age at surgery at all, have been briefly reviewed elsewhere [3]. In each of these prior studies, at least one shortcoming from among the following limit the certainty of any conclusion relating age at surgery for isolated cataract to later glaucoma: limited follow-up, small sample size, inclusion of patients with ocular anomalies aside from cataract that might independently predispose to glaucoma, inclusion of patients operated on with various techniques, and lack of information regarding whether all study patients were actually screened for glaucoma.

A recent study on a large group of cataract patients who lacked other ocular anomalies aside from microcornea, who were operated on with modern and standard technique, who were all screened for glaucoma, and who all had a minimum of 5 years’ postcataract surgery follow-up found that glaucoma was frequent when the cataract surgery was performed during the first 9 months of life (37% of 272 patients’ eyes), without seeming matter if the surgery was performed earlier or later within this timeframe [3]. (Of further note, the frequency of glaucoma was 24% of 25 patients’ eyes in a subgroup undergoing surgery in the first month of life [previously unpublished data, Peter K. Rabiah, 2004].) Multivariable analysis confirmed young age at surgery as a predictor of later glaucoma. Like all studies on the subject, this study is not without its flaws, but along with other available reports, it calls into question the usefulness of delaying surgery until 4 weeks of age. Admittedly, this study may not have been available to the report’s authors during their manuscript preparation, though importantly was available to the writer of the editorial.

The balance of available data does not clearly support a recommendation to delay cataract surgery until 4 weeks, or even 6 weeks, of age if the expectation is a reduced risk of later glaucoma; in fact, a far longer delay might very well be required, in which case the problem of amblyopia would then weigh more heavily. Further study is required to learn the optimal time to perform cataract surgery on an infant, with an important goal being to achieve the most favorable balance between the risks of later glaucoma and amblyopia.

Sincerely,

Peter K. Rabiah, M.D.

The author has no relevant financial interest in the subject matter of this letter or the article/editorial on which he comments.

References

(1) Vishwanath M, Cheong-Leen R, Taylor D, Russell-Eggitt I, Rahi J. Is early surgery for congenital cataract a risk factor for glaucoma? Br J Ophthalmol 2004;88:905-910.

(2) Lambert SR. Treatment of congenital cataract. Br J Ophthalmol 2004;88:854-855. 3 Rabiah PK. Frequency and predictors of glaucoma after pediatric cataract surgery. Am J Ophthalmol 2004;137:30-37.