Intravitreal chemotherapy-induced cataract in retinoblastoma: challenges and outcome

Abstract

A toddler with bilateral retinoblastoma (Rb) received intravenous chemotherapy for advanced tumours in both eyes. The right eye required enucleation due to a poor response, but the left eye was salvaged by additional treatment with intravitreal chemotherapy for tumour vitreous seeds. Though the vitreous seeds regressed, a gradually progressive cataract developed likely due to drug toxicity. On ensuring continued regression of tumour within the eye, cataract extraction was performed. Surgery was uneventful and the child recovered good functional vision in the left eye. Salvaging the eyeball followed by vision restoration through cataract surgery in the only remaining eye of the child necessitated a multidisciplinary approach involving the ocular oncologist, the paediatric medical oncologist and the paediatric ophthalmologist. Ensuring tumour quiescence within the eye clinically before cataract surgery was the most crucial element of management. Additionally, close monitoring of the progress of the cataract to decide when to intervene was also necessary to keep the risk of amblyopia at bay. The patient recovered good functional vision in the left eye after the surgery and remained free of any tumour recurrence or systemic metastasis at 10 months of follow-up.

Background

Cataract in cases of bilateral retinoblastoma (Rb) is a difficult entity to manage. Intraocular surgery in these eyes carries a risk of extraocular tumour spread that could be potentially lethal. In view of this risk, it is believed to be safe to ensure clinical regression of the intraocular tumour as best as possible before proceeding with cataract surgery.¹ Tumour seeding of the vitreous humour within the eye can often be a difficult problem to treat as tumoricidal concentration of chemotherapeutic agents is difficult to achieve in the vitreous cavity.² Intravitreal chemotherapy (IVitC) helps in treating tumour vitreous seeds by delivering high pharmacological concentrations of chemotherapeutic drugs directly into the vitreous.^{2 3} IVitC has progressively gained acceptance in the management of vitreous seeding in Rb at more and more specialist Rb treatment centres worldwide over the past decade³ and has now become an integral part of Rb treatment protocols. Development of a cataract is one of its attendant risks and the management of such cataracts throws up a special challenge. Here, the authors present the challenges encountered in the management of and the outcome in a case of IVitC-induced cataract.

Case presentation

A toddler was brought for an ophthalmic consultation by the parents on sustaining trauma from a fall causing redness in the right eye that had been diagnosed elsewhere as a congenital cataract with traumatic iridocyclitis. On examination performed under anaesthesia, bilateral Rb was diagnosed, with a group E tumour (with neovascular glaucoma) in the right eye and a group D tumour in the left eye respectively as per the International Intraocular Retinoblastoma Classification grouping system (figure 1A, B). Genetic testing for heritable disease was not available. However, clinical fundus screening of both parents and an older sibling did not reveal any evidence of tumour involvement in the other family members.

Figure 1.

The patient presented with bilateral retinoblastoma, International Intraocular Retinoblastoma Classification group E in the right eye (A) and group D in the left eye (B). All the three described morphologies of vitreous seeds (dust, sphere and cloud) are clearly demonstrated in the left eye (B). Utrasonography of the right eye shows a large intraocular mass lesion with clumps of intralesional calcified deposits on the B scan (C) and the respective corresponding medium and high reflective spikes on the A scan (D).

Investigations

At presentation, ocular ultrasound scans showed intraocular soft tissue mass lesions with medium internal reflectivity and multiple highly reflective clumped intralesional calcific foci with the right eye more extensively affected (figure 1C, D) than the left eye. These features were highly suggestive of Rb. MRI scans showed the disease to be intraocular, with the optic nerves, the integrity of the ocular coats and the brain being essentially within normal limits. Systemic work-up including cerebrospinal fluid cytology and bone marrow aspiration biopsy did not show any evidence of metastatic disease.

Differential diagnosis

It is not uncommon for Rb to be misdiagnosed. In this case, the referral diagnosis was that of congenital cataract with traumatic iridocyclitis due to the presence of a white pupillary reflex (leucocoria) and congestion of the eye caused post trauma. The other common differentials that were considered clinically in this case were Coats retinopathy and retinal dysplasia. Clinical differentiation was made on the basis of bilateral involvement of the eyes (Coats retinopathy and retinal dysplasia are more commonly unilateral), gender and age (Coats retinopathy is more common in - boys and tends to present at an older age of around 5–6 years), absence of any associated ocular malformation like microphthalmia (commonly associated with retinal dysplasia) or general syndromic features (retinal dysplasia associated with Norrie’s disease and incontinentia pigmenti). The findings on ocular ultrasonography were most in favour of Rb (figure 1C, D), and this coupled with the clinical appearance of the lesion (figure 1A, B) ruled out these differentials, thus substantiating the final diagnosis of Rb in both eyes.

Treatment

The child was treated with six cycles of triple-agent (vincristine, etoposide and carboplatin) neoadjuvant intravenous chemotherapy (IVC) and one injection of periocular carboplatin (15 mg) followed by three injections of intravitreal melphalan (25 mcg each) for vitreous seeds in the left eye. The right eye received additional transpupillary focal laser thermotherapy with the 810 nm diode laser. Both eyes showed good tumour and vitreous seed regression. Tumour recurrence was noted in the right eye 4 months after completion of the first six cycles of neoadjuvant IVC. The parents insisted on an attempt at eye salvage and further IVC was started. However, after four additional cycles of IVC, it was apparent that the tumour was not responding well. The right eye was enucleated and histopathology showed minimal choroidal invasion with postchemotherapeutic tumour changes. The left eye remained stable with regressed disease (figure 2A, B) until it started developing a cataract 11 months after the last intravitreal melphalan injection (figure 2C, D). The cataract progressed over the next 16 months and vision dropped from 20/160 to 20/380 over this time period. Pharmacotherapy with topical homatropine drops was performed for the left eye during the early stages of the development of cataract to counter potential development of amblyopia in the only seeing eye. Occlusion therapy was not applicable since the child had already undergone enucleation of the contralateral eye. The child then underwent clear corneal lens aspiration with primary posterior capsulotomy and anterior vitrectomy with a posterior chamber intraocular lens (IOL) implant. We used a single piece foldable monofocal hydrophobic acrylic intraocular lens (Alcon–Acrysof SA60AT) for in-the-bag implantation. The power/degree of the IOL was +26 diopters (D) and was calculated using the Sanders-Retzlaff-Kraff II (SRK II) formula. Serial clinical examinations under anaesthesia had documented regression of the primary tumour and the vitreous seeds for 19 months prior to cataract surgery.

Figure 2.

Sequential pictures showing the regression of vitreous seeds on Intravitreal chemotherapy with melphalan (A, B) and the subsequent development of central plaque-like posterior subcapsular and cortical cataract (C, D).

Outcome and follow-up

Postoperative recovery after cataract surgery was uneventful, and at 10 months of follow-up, the patient was doing well (figure 3A, B) with regressed calcified tumour remnants within the eye (figure 3C, D) and without systemic metastatic disease. Retinal pigment epithelial (RPE) changes typical of melphalan toxicity and mild optic pallor were evident post IVitC treatment (figure 3C, D). However, fundus fluorescein angiography or electroretinograms were not performed for retinal functional assessment post drug toxicity. The best-corrected visual acuity in the left eye was 20/60 at the last follow-up, and the child was comfortable with all routine activities of daily life. The child’s visual acuity was measured using Teller Acuity Cards (TAC) assessment. The set used consisted of the complete set of 17 TAC II (Stereo Optical, Chicago, Illinois, USA) as per the recommendations of the standard reference and instruction manual. The appropriate cards for the child’s age were selected and the test was administered by an experienced paediatric optometry faculty while viewing the child’s face through a 4 mm peephole in each card during testing. Under constant ambient illumination (luminance 130–150 cd/m²), the test was conducted with cards placed 55 cm away from the child’s eye, with optical correction, when necessary. The examiner judged the child’s reaction by presenting a start card without knowing whether the stripes were on the right side or on the left side. The examiner then rotated the card 180° and presented it for a second time. If the child looked in the opposite direction, the examiner checked the location of the stripes. If the child’s line of sight was the same as the striped side, the examiner chose a card with a spatial frequency one octave higher. The child’s visual acuity was determined by the card with the thinnest stripes for which the child responded.

Figure 3.

Follow-up at 10 months postoperatively shows a well-placed intraocular lens in the bag with anterior capsular fibrosis and a clear visual axis (A, B). Fundus pictures show regressed calcified tumour and vitreous seeds remnants with patchy retinal pigment atrophy and mild disc pallor resulting from melphalan toxicity (C, D).

As is already known, there is a postoperative myopic shift with time in paediatric patients undergoing IOL implantation. This shift is greatest in the younger age group and persists until 7–8 years of age. To address this myopic shift, the IOL power was selected to provide an age-appropriate postoperative target refraction as recommended by Enyedi et al.⁴ After cataract surgery, the refraction of the left eye was plano for distance at both 1-month and 3-month follow-up visits, with a near add of +3 D. At the last follow-up 10 months post surgery, the refraction in the left eye was −0.50 D cylinder at 90° for distance with a near add of +3 D. Considering the relatively stable postoperative refraction over the follow-up, the authors currently are not aiming for an IOL exchange in the future.

Discussion

Cataract surgery outcomes in children treated for Rb have been largely reported for radiation-induced cataracts⁵ and cataracts consequent to direct or other biochemical effects of the tumour and local tumour therapy.¹ With a revival of focus on IVitC for vitreous disease in Rb,³ an associated challenge of dealing with IVitC-induced cataracts has also surfaced. Treating vitreous seeds in Rb is undoubtedly challenging, but the introduction of safe injection techniques,³ excellent disease regression (83%–100%)^{3 6} and globe salvage rates (87%–100%),^{3 6} and an extremely low risk of extraocular tumour spread (proportion of 0.003–0.007)⁷ have led to IVitC being widely accepted at specialised Rb treating centres across the globe. In low-resource settings (as in the authors’ case), it is an invaluable addition to the armamentarium of an ocular oncologist managing Rb. However, it is not without its own set of adverse effects on the eye.

Multiple toxicities are reported with IVitC. They have been considered broadly under two categories: ‘significant’ (including vision loss, retinal detachment, treatment-limiting or sight-threatening vitreous haemorrhage, cataract, iris atrophy, chorioretinal atrophy, posterior synechiae and phthisis bulbi) and ‘minor’ (transient intraocular pressure elevation, transient conjunctivitis/keratitis, transient vitreous opacity, RPE alterations and injection site inflammation/haemorrhage).^{8 9} Among these, development of cataract is a challenging situation, especially in children who are in the midst of their therapeutic regimen, as it can preclude reliable assessment of the tumour ophthalmoscopically. To the best of the authors’ knowledge, there is no consensus in the literature regarding specific guidelines on the timing of surgical intervention in eyes with retinoblastoma that develop cataracts. Periods of preoperative tumour quiescence ranging from a median of 26.0 months to 34.6 months have been reported.^{1 10}

Literature on cataract surgery specifically for IVitC-induced cataracts is very sparse. In a reported series of 29 eyes with treated Rb undergoing cataract surgery by Kaliki et al,¹ only one eye had IVitC-induced cataract. More recently, Koç et al¹¹ have reported a series of five eyes of five patients with Rb who developed cataracts secondary to IVitC and subsequently underwent clear corneal lensectomy and anterior vitrectomy with or without an IOL implant. These patients also received intravitreal melphalan or topotecan intraoperatively as prophylaxis against possible tumour dissemination and subsequently received repeat monthly injections as required. The age at surgery varied from 5 years to 10 years, and all five eyes had received intra-arterial chemotherapy for treatment of their primary disease. Treatment-free period before surgery ranged from 3 months to 10 months. The range of time from the last injection to cataract detection was 2–10 months, and the time from the last injection to cataract surgery varied from 3 months to 20 months. Follow-up varied from 12 months to 16 months. None on the five eyes underwent enucleation. In comparison, the authors’ reported patient was in early childhood years at surgery and had received IVC for the treatment of the primary disease. The cataract developed 11 months after the last IVitC injection. Cataract surgery was performed 27 months after receiving the last IVitC injection. Follow-up was for 10 months.

Cataract surgery in eyes with treated Rb is preferably planned after adequate resolution of active tumour documented either clinically or ultrasonographically.¹ However, dealing with IVitC-induced cataracts may sometimes entail taking a call on performing surgery on an eye that harbours either potentially (or actually) active tumour, especially when the development of cataract interrupts an ongoing therapeutic regimen. Herein lies the challenge of appropriate tumour control in an eye prior to an intraocular surgical procedure that carries a potential of extraocular tumour spread. In Kaliki et al’s series,¹ post cataract surgery, one eye was found to have underlying active tumour and two eyes had tumour recurrence, one of which progressed to extraocular disease. This last patient had been monitored by serial ultrasound scans over 11 months to document tumour quiescence before being taken up for cataract surgery. This underscores the importance of ensuring tumour quiescence prior to cataract surgery and also demonstrates the challenges faced in monitoring tumour activity by surrogate measures like ultrasound scans in the presence of opaque media due to cataract formation that precludes direct visualisation of the tumour.

Darkly pigmented eyes (as in Asian Indians) have also been postulated to manifest more severe inflammation to intravitreal melphalan as compared with less pigmented eyes.¹² This may be a possible stimulus for development of cataract as well as a cause of increased toxicity to the RPE. Visual recovery may depend significantly on this latter factor even after the cataract has been appropriately addressed.

To conclude, no consensus guidelines exist on the period of quiescence to be ensured before performing cataract surgery in eyes with IVitC-induced cataracts. A bigger challenge is to assess the activity of vitreous seeds or tumour in eyes with opaque media due to the presence of a cataract. Visual outcomes are dependent not only on the cataract but also on the underlying toxicity that can be caused to the RPE and the optic disc by the injected drug. Given the sparse information available in the literature currently, it appears that management of cataracts secondary to IVitC for vitreous seeds in retinoblastoma needs to be tailor-made on a case-by-case basis.

Learning points.

• Intravitreal chemotherapy-induced cataracts should ideally be operated on after ensuring complete quiescence of activity of the vitreous seeds as well as any intraocular tumour. In the absence of any consensus guidelines in the literature, decisions may often need to be tailor-made for the individual patient.

• Monitoring activity of seeds in the presence of opaque media precluding direct visualisation is a major challenge. Serial ophthalmic ultrasound scans may be used to monitor activity. However, as evident from the literature, this may not always be a completely reliable method of assessment of seed activity.

• Visual outcomes are dependent not only on the cataract but also on post-treatment changes consequent to drug toxicity on the retinal pigment epithelial and the optic disc. This needs to be clearly communicated to the patient’s caregivers.

• Darkly pigmented eyes may have a relatively more severe inflammatory response to intravitreal melphalan and may have a higher risk of developing drug-induced toxicity, including cataract formation.

Ethics statements

Patient consent for publication

Consent obtained from parent(s)/guardian(s).

Ethics approval

This report adhered to all the tenets of the Declaration of Helsinki.