Intravenous Chemotherapy for Retinoblastoma in the Era of Intravitreal Chemotherapy: A Systematic Review

Abstract

Purpose

The published data on ocular survival following intravenous chemotherapy of retinoblastoma (RB) seems to be skewed by evolving practice patterns induced by use of intravitreal chemotherapy (iVitc). We aimed to explore potential role of iVitc for vitreous seeding for patients treated with intravenous chemotherapy (IVC).

Methods

A literature search was performed to identify cases of RB treated with primary IVC prior to advent of iVitc by various search engines (PubMed, Medline, and Google) from 1992 to 2018. Studies were excluded if number of cases were less than 40 or lacked data related to type of recurrence and its treatment. Rates and patterns of recurrence and its management were categorized.

Results

Out of 15 studies identified, only 10 studies (797 eyes) met the inclusion criteria. The mean age at presentation was 15.3 months (range 0–192.8 months). Unilateral cases represented 25% of the cohort. The ocular survival rate with primary IVC was 63% (500/797 eyes). Of the 297 eyes (37%) that failed IVC therapy, additional 99 eyes could be salvaged with EBRT (599/797 eyes, 75%). Remaining 198 eyes were enucleated (198/797 eyes 25%). K-M survival analysis could not be done due lack of sufficient data. Recurrences that occurred (mean 12.2 months) after completion of primary IVC included relapse of retinal tumor (143 eyes [48%]), vitreous seeding (73 eyes [25%]), subretinal seeding (49 eyes [16%]), or any combination (103 eyes [35%]). Out of 73 eyes with vitreous seeding, additional 66 eyes (90%) would have been salvaged with iVitc, potentially improving ocular survival rates to 71% (500 + 66/797).

Conclusions

Evolving practice patterns of RB treatment have unfavorably skewed published ocular survival rates following IVC. With incorporation of iVitc, the ocular survival rates with IVC can be potentially improved to be non-inferior to those achieved with intra-arterial chemotherapy.

Introduction

In the last two decades, intravenous chemotherapy (IVC), aimed at reducing tumor bulk before focal therapies, has been the mainstay of therapy for retinoblastoma (RB) [1, 2]. Various centers had reported excellent outcomes in salvaging nearly 100% of group A, B, and C eyes when combined with adjuvant laser therapy, cryotherapy, and episcleral brachytherapy [1, 3, 4, 5]. Advanced intraocular RB comprising of group D and group E eyes, however, carried a modest prognosis for globe salvage (24–59%) [6].

Because of concerns for systemic toxicity like neutropenia, infections, ototoxicity, and second cancers later in life [7] and to achieve higher globe salvage rates, since 2008 there has been a trend towards intra-arterial chemotherapy (IAC) [8]. Indeed, reported outcomes for advanced intraocular RB (group D–E eyes) are more favorable with IAC compared to IVC [9]. The globe salvage rates have ranged from 36 to 100% for group D eyes [10, 11, 12, 13] and 17–87% for group E eyes [10, 12, 14]. In a systematic review of published studies, the overall globe salvage rate was 74% with IAC as a first-line treatment modality [15].

Tumor seeding is one of the major predictive factors for failure of both IVC [16] and IAC. [17]. Intravitreal chemotherapy (iVitc), a targeted approach that delivers the highest concentration of drug in the vitreous cavity without systemic toxicity, has emerged since 2012 as an effective adjuvant treatment for vitreous seeding [18]. By definition, group C, D, and E eyes are associated vitreous seeding, and in such eyes, complete regression of vitreous seeds can be achieved in 82–100% eyes (iVitc, melphalan) [13, 18].

The role of concomitant and subsequent local therapies is an important contributor to ocular survival, particularly, the use of iVitc [17]. Because of evolving practice patterns of RB management, studies with IVC prior to 2012, have reported results without adjunctive use of iVitc, whereas recent studies with IAC report ocular survival rates with the use of iVitc [19]. As iVitc became widely acceptable during the era of IAC [18], the contributory role of iVitc in reported high globe survival rates with IAC cannot be ignored [17, 20].

As most of the patients treated during IVC era did not receive iVitc [2], we hypothesize that some of IVC failures could have been avoided with the use of iVitc. We performed a systematic review to determine ocular survival, recurrence-free survival, and event-free survival [21] and recurrence patterns following IVC for RB prior to the advent of iVitc (1992–2018). The explicit purpose was to explore potential benefit of iVitc in patients treated with IVC.

Methods

Search Method and Selection Criteria

A systematic literature search was done using search engine (PubMed, Medline, and Google) to identify all the studies related to primary IVC for treatment of RB from year 1992 to 2018 in English language (Fig. 1). Search key words included: retinoblastoma, systemic chemotherapy, intravenous chemotherapy, chemoreduction, recurrence of retinoblastoma, globe salvage rates, and failure rates. Additional information about the cases was identified from the reference lists in these publications. Studies were excluded if the number of cases in the series was less than 40 or the details related to type of recurrence or its treatment was not mentioned. As the study duration spanned over 3 decades, various authors have tried to classify RB using either Reese-Ellsworth classification or the current widely accepted international classification of intraocular retinoblastoma (ICRB).

Fig. 1.

Search and inclusion criteria for systematic review of intravenous chemotherapy for retinoblastoma.

Outcome Measures

Ocular survival, recurrence rate, and recurrence pattern.

Treatment success was defined as control of the primary tumor with IVC along with focal treatment (laser therapy, transpupillary thermotherapy, cryotherapy, and/or episcleral brachytherapy).

Definitions and Terminology

Primary treatment failure/recurrence was defined as any eye with either vitreous and/or subretinal seeding and/or recurrence of primary tumor after completion of IVC. Secondary treatment failure/recurrence was defined as eyes failing EBRT treatment and subsequently subjected to enucleation. Eyes enucleated before or during the course of IVC treatment (primary) were excluded from this study. Secondary enucleation was defined as eyes subjected to enucleation without undergoing prior EBRT whereas tertiary enucleation included cases wherein EBRT failed and hence eyes needed to be enucleated.

Statistical Analysis

All continuous variables were denoted using mean and range, whereas categorical variables were denoted using frequency and percent. K-M survival analysis could not be done due lack of sufficient data.

Results

Out of 15 major studies (1,128 eyes), 10 studies [6, 16, 22, 23, 24, 25, 26, 27, 28, 29] (797 eyes) met the inclusion criteria which were included for the detailed analysis and 5 studies were excluded [30, 31, 32, 33, 34] (Table 1).

Table 1.

Primary intravenous chemotherapy of retinoblastoma: recurrence rates and patterns (included studies)

Author [Ref.] (year)	Country	Total eyes, n	Success rate, n (%)	Recurrences, n	Type of recurrence (eyes)				Secondary treatment (eyes)			Mean follow-up, months (mean time to enucleation)
					vitreous seeding, n	subretinal seeding, n	relapse/new tumor	tumor+ seeding	plaque	EBRT (% success)	enucleation
Bartuma [28] (2001–2011)	Sweden	46	30 (65%)	16	5	5		11	4 (50%)	14	60 (14)
Manjandavida [6] (2000–2010)	India	101	53 (54%)	48	12		21		33 (73%)	24	50 (NA)
Gündüz [22] (1998–2003)	Turkey	95	53 (56%)	42	11	4	34		2	26 (77%)	22	25.7 (12.5)
Kunkele [27] (1995–2004)	Germany	56	42 (75%)	14			21		13 (77%)	4	101.6 (NA)
Shields [16] (1994–1999)	USA	158	110 (70%)	48	21	40			15	38 (29%)	37	29 (14)
Lumbroso [25] (1998–2002)	France	115	86 (75%)	29	20	5	69		1	13 (46%)	23	51 (NA)
Schiavetti [23] (1992–1998)	Italy	58	29 (50%)	29			29	3	10 (70%)	22	53 (7)
Fabian [29] (2002–2014)	UK	64	40 (63%)	24	4		20	12	5 (0%)	24	64 (12)
Berry [26] (2000–2009)	USA	55	26 (47%)	29		29		1	24 (79%)	10	54.4 (4.8)
Chung [24] (2000–2006)	South Korea	49	31 (63%)	18		6	12			18	28.9 (NA)
Summary	10 studies	797 eyes	500/797 (63%)	297/797 (37%)	73	49	143	103	45	166 (60%)	198/297 (66%)

All eyes with bilateral presentation were included as systemic chemotherapy was initiated only in bilateral cases. The mean age at presentation was 15.3 months (range 0–192.8 months). All the eyes classified according to ICRB were analyzed separately, from the old Reese-Ellsworth classification group. The groupwise distribution of all the studies according to ICRB was: group A 06%; group B 23%; group C 11%; group D 51%; group E 9% (Fig. 2). Out of 10 studies, in 8 studies the children underwent 6 cycles of standard chemotherapy (vincristine, etoposide, and carboplatin) for 2 days, whereas in 2 studies (France and Italy) only etoposide and carboplatin were given for 6 cycles. All the eyes underwent focal treatment with laser, cryotherapy, and/or brachytherapy after the tumor was consolidated with IVC. The mean follow-up for all the studies was 51.7 months (range 25.7–101.6).

Fig. 2.

Treatment response and ocular survival following intravenous chemotherapy.

The primary treatment success or ocular survival rate was 63% (500/797 eyes). In 297 eyes (37%), the primary IVC treatment failed to control the tumor and hence was termed as treatment failure. These cases were subsequently subjected to either EBRT or secondary enucleation.

Recurrence of the primary tumor was noted in form of relapse of retinal tumor (143 eyes; 48%), vitreous seeds (73 eyes; 25%), subretinal seeds (49 eyes; 16%), or combination of tumor with either type of seeding (103 eyes; 35%) (Fig. 3). Out of 166 eyes, 99 eyes (58%) were salvaged after EBRT, whereas 67 (40%) eyes underwent secondary enucleation in view of massive tumor recurrence, unresponsive to EBRT treatment or complications like neovascular glaucoma, painful blind eye, and phthisis bulbi.

Fig. 3.

Recurrence patterns following intravenous chemotherapy.

In 44 eyes of treatment failure, the data was lacking related to other mode of treatment administered like laser, cryotherapy, brachytherapy, or additional chemotherapy prior to treatment like EBRT or enucleation and hence were excluded from further analysis.

The total number of eyes enucleated were 198 (66%) out of 297 treatment failed eyes. The overall globe survival rate which included eyes successfully treated with EBRT (99 eyes) was 599 eyes (75%).

Seventy-three eyes of vitreous seeding were subjected to either EBRT or enucleation. 66 of these eyes (90% efficacy of iVitc) [13] could have been salvaged with iVitc. The presumed primary ocular survival rate would have increased from 63 to 71% (566 eyes) without use of EBRT and from 75 to 83% (665 eyes) including those that were treated with EBRT.

Discussion

Analysis of recurrence patterns following any form of therapy can guide towards development of new therapies. Regarding chemotherapy of RB, the published data depicts outcomes that do not lend to easy and fair evidence-based comparisons. In addition to adaptation of uniformity in reporting well-defined outcome measures [21], details of recurrence patterns and their management should be included so that formal meta-analysis can be performed, particularly applicable to rare disease such as RB with a small number of patients in each published series [35].

As iVitc became acceptable during the era of IAC (2008 onwards) [18], the majority of the patients treated with IVC during the study time period (1990–2018) did not receive iVitc (Fig. 4) [2]. The published data of ocular survival following chemotherapy of RB seems to be skewed by evolving practice patterns induced by use of iVitc [19]. We aimed at estimating treatment failures due to vitreous seeding that could have been salvaged with iVitc, thereby generating “missing data” in an effort to synthesize a balanced historic data wherein ocular survival rates following IVC can be fairly compared with IAC.

Fig. 4.

Evolving practice patterns of chemotherapy of retinoblastoma. The timeline represents best approximation. IVC, intravenous chemotherapy; IAC, intraarterial chemotherapy; iVitC, intravitreal chemotherapy.

The major cause of failure was primary tumor recurrence (48%). Nevertheless, vitreous seeding, amenable to iVitc was observed in 73 eyes (25%). Other patterns of recurrence included sub retinal seeding (16%) or a combination of both (35%). Given that iVitc is highly effective in controlling vitreous seeding, we assumed 90% response rate [13, 18]. Incorporation of iVitc can potentially improve ocular survival rates to 71% (500 + 66/797) from 63% (without use of iVitc), which seems to be comparable to results reported with IAC [15]. In fact, a few recent studies have shown improved ocular survival (70–80%) with use of concurrent or subsequent administration of iVitc for refractory or recurrent vitreous seeding [13, 36, 37, 38, 39, 40]. Similar enhancing effect on ocular survival has also been reported in patients treated with IAC with (5-year ocular survival of 85–100%, group D) [9] or without use of iVitc (64% at 2 years, group C, D, and E) [41], again a reflection of evolving practice patterns. iVitc has shown some promising results for the treatment of subretinal tumor and subretinal seed recurrence thereby potentiating the role of adjunctive iVitc and hence improving overall survival. We have not included the cases of tumor and subretinal seed recurrence in our analysis as these are initial results which needs further validation [42, 43].

We are aware of the limits of the present systematic analysis, as all the included studies were retrospective in nature. Despite our best efforts, not all studies/data can be included for the analysis that by itself could have induced selection bias. K-M survival analysis could not be done due lack of sufficient data. Among many variables, the case mix of ICRB classification in the IVC studies may not be directly comparable to those with IAC studies. Secondly, variations in classification systems used in each study can upgrade or downgrade eyes with group D and E eyes making it impossible to compare outcomes between studies [44]. We do recommend a single uniform classification system for advanced intraocular RB to avoid discrepancies when comparing results between clinical series. Our large cohort of 797 eye represented the spectrum of ICRB with the majority in the group C–E eyes (71%). Even so, critical review of available data can be revealing, particularly when evolving practice patterns seem to have skewed ocular survival rates against use of IVC.

Now that the Children's Oncology Group (COG) adjuvant therapy trial has revealed acceptable systemic safety profile of IVC in children with RB [45], the argument against use of IVC based on systemic toxicity profile seems to be mitigated. Another advantage of IVC is lack of any known ocular toxicity, whereas IAC carries significant risk of vascular ischemia, retinal detachment, and vitreous hemorrhage ranging from 13 to 23% [46, 47].

The published data of ocular survival rates following chemotherapy of RB is skewed by evolving practice patterns (use of iVitc since 2012) that implies enhanced efficacy of IAC and diminished efficacy of IVC. Based upon analysis of recurrence patterns following IVC, we project that enhanced globe survival rates in the era of iVitc can be achieved with IVC that are comparable to IAC. Based upon efficacy and safety profile of IVC, there is sufficient evidence to justify a non-inferiority randomized clinical trial comparing IVC with IAC.

Conflicts of Interest Statement

The authors have no conflicts of interest to declare.

Funding Sources

This work was supported by an unrestricted departmental grant from Research to Prevent Blindness Challenge Grant, to the Cole Eye Institute, Cleveland Clinic.

Author Contributions

V.R. and A.S. conceived and designed the study, performed data interpretation, and drafted the manuscript. V.R. collected the data. H.S. and R.C.B. critically revised the manuscript. V.R. and A.S. performed data analysis. All authors critically reviewed and approved the final version of the paper.