Outcome of the Treatment of Retinopathy of Prematurity in Qatar

El Shafei M.M., Rodriguez V.R., Martinez F.E.
Ophthalmology Section, Department of Surgery, Hamad Medical Corporation
Doha, Qatar

Abstract:

Introduction:

Retinopathy of prematurity (ROP), first described in 1942(1), is a multifactorial proliferative retinopathy which occurs almost exclusively in premature infants. Many risk factors and associated conditions have been suggested and reported, the most significant being low birth weight, lower gestational age and the need for supplemental oxygen.

The incidence of ROP has been closely related to the improvement in neonatal care. For about the last 15 years, a fetus with a gestational age of 25 weeks and/or a birth weight of 700 grams has more than a 50 % probability of survival(2). Hence the population at risk for developing ROP has increased. It is known that in newborns of less than 1700 grams birth weight the incidence of ROP reaches 50.9 % and 4.7 % of them develop end-stage disease(3). If there were no treatment for this disease the consequences would be devastating. In a survey carried out in schools for the blind as many as 39 % of cases of blindness or severe visual impairment have been found to be due to ROP(4). Nevertheless, with the increased awareness and improved life support capabilities at Neonatal Intensive Care Units (NICU), along with treatment for threshold ROP disease, this figure has been reduced(5).

As a result of newborn immaturity, large areas of the retina are avascular, presumably producing an ischemic mediated release of angiogenic factors causing a neovascular response(6). Destruction of these non-vascularized zones could be a rationale for removing this triggering condition.

Attempts at treating ROP started in 1968. After the International Classification of ROP (IC ROP) improved the understanding of the nature of the disease, and the utilization of screening eye examination for ROP detection, a multicenter trial of cryotherapy for ROP was carried out, demonstrating the efficacy of cryoablation in reducing the risk of an unfavorable outcome in newborns with “threshold” ROP(7). Later, laser photocoagulation proved to be as effective as cryotherapy(8), even becoming an important alternative when zone I treatment was needed.

In 1997 the Ophthalmology Section, Surgery Department in Hamad Medical Corporation, Doha, Qatar, began a screening program for premature infants as well as treatment of threshold ROP. As soon as the first cases of threshold ROP were detected, cryotherapy was performed. In 1999 a laser indirect ophthalmoscope was received and was used immediately to apply this technique to these infants. The results of the treated patients are reported and analyzed below.

Patients and Methods:

The charts were reviewed retrospectively of 46 infants with a birth weight (BW) of 1500 grams or less and a gestational age (GA) of 32 weeks or less admitted to Hamad Hospital Neonatal Intensive Care Unit (NICU) between 1st January 1999 and 31st December 2002. These were screened, diagnosed and staged for ROP based on the criteria and recommendations of the ICROP Study(9) (Table 1) and were treated with laser photocoagulation for threshold ROP. Of the 46, three were excluded because of insufficient follow-up data (6.52 %); one died three days after treatment as a result of systemic complications unrelated to the treatment, and two left the country soon after treatment. Finally 43 infants were analyzed (Graph 1).
 

Infants eligible for treatment were those with threshold disease as Cryo-ROP Study criteria(10) (Table 1). A detailed parental informed consent was required before treatment. Diode laser photocoagulation was used within 48 hours of diagnosis of the condition. The use of a lid speculum and pupillary dilatation was always necessary. For dilatation purposes, we used an infant mydriatic drops mixture prepared in our pharmacy (phenylephrine 1%, tropicamide 0.45 % and cyclopentolate 0.45 %) applied twice with a 15 minute interval. Treatment was applied always by two of the authors (MME and VRR) only to the whole avascular retinal zone anterior to the ridge and up to the ora serrata.
 

A diode laser unit with indirect ophthalmoscope delivery system (IRIS Oculight Slx LIO 500, IRIS Medical Instruments; Mountain View, CA; USA) and an aspheric 20 diopters Volk lens, with scleral indentation (when needed) were used. Laser spots were placed one-half “burn width” apart, with a faint white endpoint laser mark. Laser power range was 200 to 500 mW; duration of single applications was 200 msec.; Burn numbers ranged from 801-1216, with an average of 978.

Re-treatment using cryotherapy was necessary when no regression or non-treated areas with residual activity were discovered during the early follow-up. Several applications, with retinal whitening as an endpoint, and a freeze time of about 2-3 seconds was the chosen technique. A 2 mm diameter neonatal cryoprobe (DORC. Ophthalmic Cryounit, Spembly Medical Ltd. Hampshire, UK) was used. The mean number of cryo applications was seven (range 4-9). No cases needed conjunctival incision for application of the treatment.

Treatment was carried out always in the operating theatre at Hamad Hospital under general anesthesia consisting of anticholinergic premedication to stop oculocardiac reflex and a narcotic-relaxation technique to maintain the cardio-respiratory stability with monitoring by a neonatologist and anesthetist. The treatment was halted temporarily if that was requested by the medical support team due to the child’s condition becoming medically unstable. Post surgery treatment consisted of combined antibiotic and steroid eye drops three times a day for three days.

Treatment was followed up after 24 hours and then every week until total regression of the disease was achieved. Re-treatment was considered if no signs of regression were detected two weeks after the initial treatment. If full regression was achieved, follow-up visits were scheduled at three, six and twelve months of age. At the age of one year patients were re-examined again under general anesthesia and refraction was performed (data not analyzed).

Only two categories of treatment outcome were considered: favorable or unfavorable (Table 1). Baseline independent variables such as birth weight (BW), gestational age (GA) at birth, post-conceptional age (PCA), presence of systemic complications (not related to the surgical event) and also the clinical characteristics of the disease (zones involved and extension of neovascularization) were analyzed. For statistical analysis of this retrospective study, descriptive statistics (%, mean, SD) were used for baseline clinical-demographic variables. To evaluate the possible influences of baseline variables and disease characteristic on treatment outcome, a correlation test (r) was performed. For outcome categories analysis, unpaired T Student’ test for continuous variables and chi-square test for non-continuous were applied. Level of significance was assumed for an a value of 0. 05.

Results:

Forty three infants (86 eyes) with a diagnosis of bilateral symmetrical threshold ROP were treated in HMC-NICU during the period. There were no statistical differences regarding gender or nationality on treated patients (p> 0.5). Mean GA was 26.2 weeks, and mean BW was 811 grams. Mean PCA at ROP Threshold diagnosis and treatment was 35 weeks (Table 2). Mean interval time between diagnosis and treatment was 1.8 days. Laser treatment was applied initially to all patients. Ten infants (23.25%) needed re-treatment by cryo application (Graphic 1).

“Favorable” outcome with regression of ROP was achieved in 40 patients (93.02%), 33 of them (76.64 %) after the initial treatment. Only three infants (6.97%) did not show regression, even after re-treatment (Graphic 1). Regression of plus disease occurred within one week in all infants. Mean time span between treatment and ROP regression or other characteristics of a favorable outcome was 16.3 days (range 13-18).

Cryotherapy was the selected method for re-treatment of ten patients, seven of whom (70%) showed a favorable outcome. Regression mean time was shorter than in the primary treatment but the difference was not significant. Most common causes for re-treatment were the presence of “skipped” previously untreated areas (4 out of 10) and/or insufficient regression after first applications (6 out of 10).

No local (ocular) complications occurred during laser treatment. In the case of cryotherapy, only mild chemosis and periorbital swelling after the applications were noticed, without any sequels. No systemic complications occurred during the procedures or related to them.

“Unfavorable” outcomes occurred in three patients after re-treatment due to residual neovascular activity. Two developed tractional retinal detachment (one stage 4B and the other stage 5 with vitreous hemorrhage). The third patient regressed from acute ROP but developed a gross retinal fold that reached the macula.

Comparative analysis of favorable and unfavorable outcomes is characterized by the very small number of failed cases. Nevertheless, important differences were noticed in variables such as gestational age (GA) (p=0.01), presence of infant’s systemic complications (p=0.008) and presence of ROP disease in zone I (p= 0.0003) (Table 3). Regarding the last two variables, 11 of 40 infants had systemic complications in the favorable group (and all three in the unfavorable group); and only two of 40 had ROP in Zone I in the favorable group (and two of the three in the unfavorable group). Mean birth weight (BW), post-conceptional age (PCA) at threshold’s disease diagnosis, and ROP extension failed to show significant differences between the outcome groups (p>0.05).

Table 3:  Treatment Outcome and Baseline Clinical Data

BW = Birth Weight  PCA = Post Conceptional Age  GA = Gestational Age (+) = Mean     (-) = Statistically Significant (X2 Test) (*) - Statistically Significant (T' Student Test)

However, when baseline variables were re-analyzed, particularly stratified and correlated with the presence of an “unfavorable” outcome (Table 4), some were found to be statistically significantly related with this event. The presence of BW lower than 800 grams, GA less than 27 weeks, presence of systemic complications and ROP on zone I were statistically correlated with failed outcome (p< 0.05). The strongly correlated variables were the presence of systemic complications (r=0.82) and the GA (r=0.73).

Table 4 : Perinatal Risk Factors Correlated with Unfavorable Outcome                (Only Statistically Significant Results r ; p < 0.05)

Discussion:

Our series showed no differences when comparing it with the Cryo ROP Study regarding BW and GA (p>0.05)(11) but our range of favorable outcomes was greater than the Cryo ROP Study (93.02 % vs. 74.4 %). This remarkable difference could basically be the result of some particularities of our series. Despite having no differences in BW (811 g. vs. 800 g) or GA (26.2 vs. 26.3 weeks) with the Cryo ROP series, there were twice the number of newborn infants with threshold ROP in zone I (9.30 % vs. 4.12%). Assuming that this ROP characteristic was statistically correlated with treatment failure, it seems that the large success ratio was definitely influenced by the treatment used. All our patients were treated with laser photocoagulation and it is well known that the presence of threshold ROP in zone I reveals particular conditions regarding outcome when treatment is focused on(12-15) because of the easier accessibility of posterior areas of the retina with the laser. Even more, this difference in outcome becomes much less significant when this series is compared with other studies in which laser photocoagulation is used with similar structural design and outcome evaluation. For example, in the McNamara(16) and Fleming(17) series, there were no differences with our patients in BW or GA, and the success ratio was similar. Also, in the Hunter study there were similar results(18).

Laser photocoagulation has already become the preferred treatment for threshold ROP because of its many advantages and effectiveness even in severe cases(19-21).

A laser indirect ophthalmoscope became available in our department in 1999 since when it has become the preferred treatment. The learning curve was short for experienced surgeons and the complications are less than for cryotheraphy. Despite the potential of immediate complications as inadvertent macular burns, posterior segment hemorrhages at all levels and thermal injuries to the cornea, iris and lens, local (ocular) complications are fewer and less significant. Conjunctival hematomas, scarring or lacerations, all reported in Cryo-ROP studies, are much less frequent with laser treatment(21). Induced myopia after treatment is more common with cryotherapy than with laser treatment(22,23) and retino-choroidal scarring caused by cryotherapy is also reduced with the potential advantage of lower incidence of rhegmatogenous retinal detachment.

Nevertheless laser treatment is not totally harmless; excessive treatment cause thermal damage to the inner scleral tissues and ciliary nerves and vessels and the incidence of cataract after treatment appears to be greater when compared with cryotherapy(24). Fortunately, in this series there were none of these complications with laser treatment.

Cryotherapy has some advantages in its application because it requires a less clear optical status (including corneal or lens opacities, and vitreous or pre-retinal hemorrhages, which are frequent in severe cases) and it can still be applied even when the pupil is poorly dilated. Since most of our patients with an insufficient first treatment procedure needed retinal ablation mostly in the far periphery, an easily accessible area with the cryoprobe, cryotherapy was the preferred method for retreatment.

In our study all the patients in the unfavorable outcome group required re-treatment as a result of an insufficient first therapeutic approach. It is known that light treatment with laser may not affect the spindle cell endothelial capillary precursors of the inner retinal layers, which, in the immature retina, may affect the course of ROP as a potentially triggering factor for neovascularisation(25).

All the treatments were conducted under general anesthesia in the presence of a neonatologist and an anesthetist using anticholinergic premedication and a narcotic relaxation technique to maintain cardiorespiratory stability. At the same time the immobility of the infant during the procedure was guaranteed. We believe that the management of those immature infants under these care conditions is ideal. In comparison with other reports(26), we did not have any systemic treatment-related complications.

The influence of some baseline general and ocular conditions in the final outcome was variable. Nevertheless, the presence of severe systemic complications in newborn infants with ROP, whether or not it was related with the treatment procedure, had a significant influence. Episodes of hypoxemia, long term oxygen supplementation, hypocarbia, sepsis, blood transfusion or ventricular hemorrhages, are, among others, well known risk factors for developing severe ROP(27,28). In this study we restricted the analysis to the presence or absence of severe, life-threatening systemic complications. About thirty per cent of the infants were presented with severe systemic complications unrelated to treatment including all three of the unfavorable outcome group. Considering that ROP is a multifactorial disease, many neonatal risk factors other than the most important ones (gestational age and birth weight) may play a significant role, although they are not fully understood. They could also have an influence on the outcome of treatment. However, caution should be taken when interpreting or assuming conclusions based on the characteristics our unfavorable outcome group because it contained only three patients. Despite some limitations in this series, such as the relatively short follow-up, very encouraging results have been achieved since the start of ROP screening and treatment Hamad Medical Corporation. The World Health Organization “Vision 2020 Program” has identified ROP as an important cause of blindness in both high and middle income countries(29). Consequently, the long term visual results (beyond the scope of this study) and the impact on the incidence of child blindness surely will be very important in the Qatari society and health plans.

References:

29. Steinkuller PG, Du L Gilbert C, et al: Childhood blindness. J AAPOS 1999; 3: 26-32.