Journal of Pediatric Ophthalmology and Strabismus

Original Article 

Risk Factors for Tumor Recurrence Following Primary Intravenous Chemotherapy (Chemoreduction) for Retinoblastoma in 869 Eyes of 551 Patients

Lauren A. Dalvin, MD; Zeynep Bas, MD; Sameeksha Tadepalli, MD; Raksha Rao, MD; Sarangdev Vaidya, BA; Richard Pacheco, BA; Carol L. Shields, MD

Abstract

Purpose:

To identify risk factors for retinoblastoma recurrence following chemoreduction.

Methods:

This was a retrospective review of patients with retinoblastoma treated from 1994 to 2019 using chemoreduction with analysis for recurrence using Kaplan–Meier, Cox regression, and logistic regression.

Results:

There were 869 eyes of 551 patients with retinoblastoma treated with chemoreduction. Follow-up in 556 eyes revealed main solid tumor recurrence (n = 355, 64%), subretinal seed recurrence (n = 244, 44%), vitreous seed recurrence (n = 162, 29%), and/or new tumor (n = 118, 21%) requiring management with focal therapy (transpupillary thermotherapy, cryotherapy) (n = 294, 53%), intra-arterial chemotherapy (n = 125, 22%), intravitreal chemotherapy (n = 36, 6%), plaque radiotherapy (n = 120, 22%), external beam radiotherapy (n = 57, 10%), and/or enucleation (n = 49, 9%). Of all recurrences, 62% were detected by 1 year, 86% by 2 years, 94% by 3 years, 98% by 5 years, 99% by 10 years, and 100% by 15 years. Risk factors for recurrence on multivariate analysis included younger patient age at presentation (odds ratio [OR] = 1.02 [1.00 to 1.04] per 1 month decrease, P = .02), greater International Classification of Retinoblastoma group (OR = 1.24 [1.05 to 1.47] per 1 more advanced group, P = .01), shorter tumor distance to optic disc (OR = 1.11 [1.01 to 1.21] per 1 mm decrease, P = .03), and presence of subretinal seeds (OR = 1.66 [1.09 to 2.53], P = .02).

Conclusions:

Retinoblastoma recurrence after chemoreduction is usually detected within the first 3 years following treatment. Younger patients with more advanced, posteriorly located tumors and subretinal seeds at presentation are at increased risk, but recurrence can often be managed with globe-sparing therapy.

[J Pediatr Ophthalmol Strabismus. 2020;57(4):224–234.]

Abstract

Purpose:

To identify risk factors for retinoblastoma recurrence following chemoreduction.

Methods:

This was a retrospective review of patients with retinoblastoma treated from 1994 to 2019 using chemoreduction with analysis for recurrence using Kaplan–Meier, Cox regression, and logistic regression.

Results:

There were 869 eyes of 551 patients with retinoblastoma treated with chemoreduction. Follow-up in 556 eyes revealed main solid tumor recurrence (n = 355, 64%), subretinal seed recurrence (n = 244, 44%), vitreous seed recurrence (n = 162, 29%), and/or new tumor (n = 118, 21%) requiring management with focal therapy (transpupillary thermotherapy, cryotherapy) (n = 294, 53%), intra-arterial chemotherapy (n = 125, 22%), intravitreal chemotherapy (n = 36, 6%), plaque radiotherapy (n = 120, 22%), external beam radiotherapy (n = 57, 10%), and/or enucleation (n = 49, 9%). Of all recurrences, 62% were detected by 1 year, 86% by 2 years, 94% by 3 years, 98% by 5 years, 99% by 10 years, and 100% by 15 years. Risk factors for recurrence on multivariate analysis included younger patient age at presentation (odds ratio [OR] = 1.02 [1.00 to 1.04] per 1 month decrease, P = .02), greater International Classification of Retinoblastoma group (OR = 1.24 [1.05 to 1.47] per 1 more advanced group, P = .01), shorter tumor distance to optic disc (OR = 1.11 [1.01 to 1.21] per 1 mm decrease, P = .03), and presence of subretinal seeds (OR = 1.66 [1.09 to 2.53], P = .02).

Conclusions:

Retinoblastoma recurrence after chemoreduction is usually detected within the first 3 years following treatment. Younger patients with more advanced, posteriorly located tumors and subretinal seeds at presentation are at increased risk, but recurrence can often be managed with globe-sparing therapy.

[J Pediatr Ophthalmol Strabismus. 2020;57(4):224–234.]

Introduction

The introduction of intravenous chemotherapy (chemoreduction) for the treatment of retinoblastoma in the early 1990s revolutionized retinoblastoma management.1–3 Chemoreduction largely supplanted the use of external beam radiotherapy (EBRT) for globe salvage, with protection against pineoblastoma, avoidance of radiation-related toxicity to the eye and orbital growth retardation, and lower risk for second cancers.4–10 In combination with focal therapies, including cryotherapy and transpupillary thermotherapy (TTT), chemoreduction has been proven efficacious for tumor control.1–3,11–13 In a 2006 study by Shields et al14 on 249 consecutive eyes with retinoblastoma managed with chemoreduction and with a mean follow-up of 6.2 years, globe salvage varied by International Classification of Retinoblastoma (ICRB) group at 100% in group A, 93% in group B, 90% in group C, and 47% in group D, whereas group E eyes were generally managed with primary enucleation. We recently reviewed the 20-year long-term outcomes of chemoreduction for retinoblastoma in 964 eyes of 554 patients and found 2-year Kaplan–Meier estimated tumor control with avoidance of enucleation or EBRT at 96% in group A, 91% in group B, 91% in group C, 71% in group D, and 32% in group E with minimal change at 20 years of follow-up.11

In 2002, Shields et al15 investigated factors predictive of solid tumor, subretinal seed, and vitreous seed recurrence following chemoreduction for retinoblastoma in 158 eyes. The Kaplan–Meier estimated risk for recurrence peaked at 51% at 3 years for solid tumor, 62% at 3 years for subretinal seed, and 50% at 5 years for vitreous seed. The only factor predictive of solid tumor or vitreous seed recurrence was presence of subretinal seeds on initial examination, whereas largest tumor basal diameter greater than 15 mm and patient age of 12 months or younger at diagnosis contributed to the risk for subretinal seed recurrence. In 2004, Shields et al16 investigated risk factors for tumor recurrence in 457 eyes with retinoblastoma managed with chemoreduction. Eyes were managed by ICRB group with chemoreduction alone or in combination (consolidation) with TTT and/or cryotherapy. Only 11 group D eyes and no group E eyes were included. With a 7-year follow-up analysis, risk factors for tumor recurrence by multivariate analysis included non-White race, greater tumor thickness, macular tumor location, and lack of consolidation therapy. Given the long-term 20-year follow-up of nearly twice as many eyes, and greater number of group D and new inclusion of group E eyes, updated evaluation of risk factors for tumor recurrence is required.

We evaluate risk factors for local tumor recurrence of main solid tumor, subretinal seed, vitreous seed, and new tumor formation in eyes with retinoblastoma managed with primary chemoreduction over 20 years of follow-up. We included a large number of advanced group D and E eyes, as well as those managed in recent years with the availability of intra-arterial chemotherapy (IAC) for improved globe salvage, and we compared whether predictors of recurrence are similar to those previously reported in cohorts predominantly consisting of groups A, B, and C eyes in the era before IAC.

Patients and Methods

Additional analysis was performed on a cohort of patients with retinoblastoma treated with primary chemoreduction on the Ocular Oncology Service, Wills Eye Hospital, Thomas Jefferson University, in conjunction with the Department of Pediatric Oncology, Children's Hospital of Philadelphia, from January 1, 1994, through June 1, 2019.16 Inclusion criteria were intraocular retinoblastoma with no evidence of optic nerve; choroidal, scleral, or orbital involvement; management with primary chemoreduction; and minimum of 3 months of follow-up after chemo-reduction. Exclusion criteria were primary treatment with other modalities, such as primary enucleation, or follow-up after chemoreduction shorter than 3 months. This study complied with the Health Insurance Portability and Accountability Act (HIPAA), Institutional Review Board approval was obtained from Wills Eye Hospital, and informed consent was obtained for all treatments.

Patients underwent a complete eye examination under anesthesia by a senior ocular oncologist (CLS) with detailed retinal drawing, RetCam photography (Clarity Medical Systems), ocular ultrasonography, intravenous fluorescein angiography as needed, and, in recent years, optical coherence tomography (OCT). Drawing and imaging were repeated at each follow-up examination. Chemoreduction consisted of a planned six-cycle regimen of vincristine, etopo-side, and carboplatin (VEC) at the standard dose for groups A, B, and C or an intensified dose for groups D and E as previously described.1,11 Records were reviewed for patient demographics, clinical tumor features, treatment parameters, and outcomes.

Demographics included age, race, sex, involved eye, laterality, and germline mutation status. Clinical features at presentation included ICRB classification, number of tumors per eye, largest tumor basal diameter and thickness, tumor distance to foveola and optic disc, presence of vitreous or subretinal seeds, subretinal fluid, anterior chamber seeds, iris neovascularization, and nystagmus. Treatment features included chemotherapy regimen, focal consolidation with TTT and/or cryotherapy, and requirement for additional IAC, intravitreal chemotherapy, plaque radiotherapy, or EBRT. Additional therapy was required for recurrence and in some cases with no recurrence for incomplete tumor regression after chemoreduction therapy. Ocular outcomes at date last seen included duration of follow-up and tumor regression type as previously defined (type 0 = no remnant, type 1 = calcified remnant, type 2 = noncalcified remnant, type 3 = partially calcified remnant, or type 4 = flat scar).17

Patients were assessed for recurrence (main solid tumor, subretinal seed, vitreous seed) or new tumor formation. Recurrence was defined as recurrent tumor 3 months or more after initial tumor regression from chemoreduction, whereas activity prior to 3 months of quiescence was considered tumor progression. New tumor formation could occur any time after date first seen. To purely address the question of recurrence, risk factors were assessed for main solid tumor, subretinal seed, and vitreous seed recurrence alone. However, to ultimately address all aspects of tumor activity that may require secondary treatment, additional analysis was undertaken to examine the combined risk of tumor recurrence and/or new tumor formation. Additional features included type of recurrence (main solid tumor, subretinal seed, vitreous seed, or new tumor), time from date first seen to recurrence or new tumor, and treatment for recurrence or new tumor. Treatments for recurrence or new tumor included TTT, cryotherapy, IAC, intravitreal chemotherapy, plaque radiotherapy, EBRT, or enucleation. Reason for and time to enucleation were recorded. Systemic outcomes included metastasis, pineoblastoma, second cancers, and death.

Statistical analysis was performed using SPSS Statistics Software version 22 (IBM Corporation). Demographics, clinical features, treatment features, and outcomes were compared for eyes with recurrence (main solid tumor, subretinal seed, vitreous seed, or new tumor) versus those without recurrence. Categorical variables were compared using the Fisher's exact test or chi-square test, and continuous variables were compared using the Student's t test. Comparison of time to recurrence by type (main tumor vs subretinal seed vs vitreous seed) was performed with analysis of variance. Cox regression analysis was used to calculate the hazard ratio (95% CI) by ICRB group for recurrence or new tumor at 1, 2, 3, 5, 10, 15, and 20 years. Patients were censored at the time of last follow-up, time of recurrence or new tumor, or time of enucleation. Logistic regression analysis was used to determine risk factors for recurrence or new tumor formation by univariate and multivariate analysis. A P value less than .05 was considered statistically significant.

Results

There were 869 eyes of 551 patients with retinoblastoma managed with primary chemoreduction included in this series, with 95 eyes excluded due to fewer than 3 months of available follow-up by our service. Of the 869 included eyes, 556 eyes of 408 patients developed some degree of recurrence (main solid tumor, subretinal seed, vitreous seed, or new tumor) following chemoreduction versus 313 eyes of 262 patients with no recurrence or new tumor.

Patient demographics are listed in Table 1. A comparison (recurrence [main solid tumor, subretinal seed, vitreous seed, new tumor] vs no recurrence) revealed that recurrence was found with younger mean patient age at presentation (10 vs 13 months, P = .02). There was no difference in patient race, sex, involved eye, unilateral or bilateral tumor involvement, or hereditary pattern of somatic, germline, or 13q deletion syndrome.

Patient Demographics

Table 1:

Patient Demographics

Clinical features at date first seen are described in Table 2. A comparison (recurrence [main solid tumor, subretinal seed, vitreous seed, new tumor] vs no recurrence) revealed recurrence with more advanced disease, including fewer group A (4% vs 9%, P = .01) and group B (20% vs 27%, P = .02) eyes, more group D (30% vs 20%, P = .002) and unclassified (14% vs 8%, P = .004) eyes, a greater number of tumors per eye (mean 2 vs 2 tumors, median 2 vs 1 tumor, P = .01), a greater mean largest tumor basal diameter (12.1 vs 10.1 mm, P < .001) and thickness (6.3 vs 5.3 mm, P < .001), shorter mean tumor distance to foveola (1.9 vs 3.2 mm, P < .001) and optic disc (1.5 vs 2.9 mm, P < .001), and lower frequency of no subretinal seeds (60% vs 71%, P = .003) or subretinal fluid (52% vs 63%, P = .002). The recurrence group had lower frequency of anterior chamber seeds (<1% vs 2%, P = .01). When a view was present, there was no difference between groups in frequency of vitreous seeds. There was no difference in frequency of iris neovascularization or nystagmus.

Clinical Features at Diagnosis

Table 2:

Clinical Features at Diagnosis

Treatment features are described in Table 3. A comparison (recurrence [main solid tumor, subretinal seed, vitreous seed, new tumor] versus no recurrence) revealed recurrence with more frequent requirement for additional globe-sparing treatment, including cryotherapy (60% vs 35%, P < .001), IAC (26% vs 3%, P < .001), intravitreal chemotherapy (11% vs 1%, P < .001), plaque radiotherapy (30% vs 2%, P < .001), or EBRT (17% vs 7%, P < .001). There was no difference in the frequency of chemoreduction started outside of Wills Eye Hospital, regimen used for chemoreduction, number of chemotherapy cycles, or requirement for TTT.

Treatment Features and Outcomes

Table 3:

Treatment Features and Outcomes

Clinical outcomes are described in Table 3. Of the 556 eyes with recurrence, there were 355 (64%) with main tumor recurrence, 244 (44%) with sub-retinal seed recurrence, 162 (29%) with vitreous seed recurrence, and 118 (21%) with new tumor. Mean time to recurrence of any type was longer than mean time to new tumor formation (15 months to any recurrence [main solid tumor, subretinal seed, vitreous seed] versus 10 months to new tumor, P = .03). By recurrence type (main solid tumor, subretinal seed, vitreous seed), mean time was shortest for subretinal seed recurrence (16 vs 12 vs 18 months, P = .002). Recurrence or new tumor formation was managed with focal therapy (cryotherapy/TTT) (53%), IAC (22%), intravitreal chemotherapy (6%), plaque radiotherapy (22%), EBRT (10%), or enucleation (9%). A comparison (recurrence [main solid tumor, subretinal seed, vitreous seed, new tumor] vs no recurrence) revealed recurrence with more frequent regression type 1 (17% vs 9%, P = .003) and less frequent regression types 2 (6% vs 10%, P = .02) and 4 (9% vs 14%, P = .04). The recurrence group had more frequent requirement for enucleation (21% vs 14%, P = .01). There was no difference in duration of follow-up or systemic outcomes of metastasis, pineoblastoma, second cancers, or death, which were rare in both groups.

Risk for recurrence or new tumor formation by Cox regression analysis is described in Table 4. Of all recurrences (main solid tumor, subretinal seed, vitreous seed), 62%, 86%, 94%, 98%, 99%, and 100% were detected by 1, 2, 3, 5, 10, and 15 years, respectively. The risk for tumor recurrence increased by ICRB group with hazard ratio (compared to group A) at 1.4 for group B, 1.4 for group C, 2.3 for group D, and 2.0 for group E (overall P < .001) at 5 years with minimal change through 20 years of follow-up. Of all tumor activity including recurrences and new tumor formation, 62%, 86%, 94%, 98%, 99%, and 100% were detected by 1, 2, 3, 5, 10, 15, and 20 years, respectively. The risk for tumor recurrence/new tumor formation increased by ICRB group with hazard ratio (compared to group A) at 1.2 for group B, 1.1 for group C, 1.8 for group D, and 1.6 for group E (overall P < .001) at 2 years with minimal change through 20 years of follow-up.

Outcomes by Cox Regression Analysis for Recurrence and New Tumora

Table 4:

Outcomes by Cox Regression Analysis for Recurrence and New Tumor

Logistic regression analysis for predictors of tumor recurrence/new tumor formation is described in Table 5. By univariate analysis, predictors of recurrence only (main solid tumor, subretinal seed, vitreous seed) included younger patient age at diagnosis (odds ratio [OR] = 1.02 per 1 month decrease in age, P = .003), ICRB group (OR = 1.27 per 1 more advanced group, P < .001), number of tumors (OR = 1.12 per 1 additional tumor, P = .01), larger tumor basal diameter (OR = 1.05 per 1 mm larger, P < .001) and thickness (OR = 1.08 per 1 mm larger, P < .001), closer tumor distance to foveola (OR = 1.13 per 1 mm closer, P < .001) and optic disc (OR = 1.18 per 1 mm closer, P < .001), presentation with subretinal seeds (OR = 2.27, P < .001), subretinal fluid (OR = 1.68, P = .001), or anterior chamber seeds (OR = 5.22, P = .04), and regression type 1 (OR = 1.53, P = .04). Regression types 2 (OR = 0.57, P = .03) and 4 (OR = 0.59, P = .02) were less likely to be associated with recurrence. On multivariate analysis, factors predictive of recurrence alone (main solid tumor, subretinal seed, vitreous seed) included younger patient age at diagnosis (OR = 1.02, P = .02), greater ICRB group (OR = 1.24, P = .01), shorter tumor distance to optic disc (OR = 1.11, P = .03), and greater presence of subretinal seeds (OR = 1.66, P = .02).

Predictors of Recurrence and New Tumor by Logistic Regression Analysis

Table 5:

Predictors of Recurrence and New Tumor by Logistic Regression Analysis

By univariate analysis, predictors of any tumor activity, encompassing tumor recurrence (main solid tumor, subretinal seed, vitreous seed) and new tumor formation, included younger patient age at diagnosis (OR = 1.02 per 1 month decrease in age, P < .001), germline mutation or 13q deletion syndrome (OR = 1.39, P = .02), ICRB group (OR = 1.21 per 1 more advanced group, P = .001), number of tumors (OR = 1.13 per 1 additional tumor, P = .02), larger tumor basal diameter (OR = 1.05 per 1 mm larger, P < .001) and thickness (OR = 1.07 per 1 mm larger, P = .001), closer tumor distance to foveola (OR = 1.11 per 1 mm closer, P < .001) and optic disc (OR = 1.17 per 1 mm closer, P < .001), presentation with subretinal seeds (OR = 2.07, P < .001), subretinal fluid (OR = 1.87, P < .001), or anterior chamber seeds (OR = 6.44, P = .02), and regression type 1 (OR = 1.86, P = .01). Regression types 2 (OR = 0.52, P = .01) and 4 (OR = 0.59, P = .02) were less likely to be associated with recurrence or new tumor. On multivariate analysis, factors predictive of recurrence or new tumor included younger patient age at diagnosis (OR = 1.03, P = .003), germline mutation or 13q deletion syndrome (OR = 1.47, P = .02), greater ICRB group (OR = 1.20, P = .03), and shorter tumor distance to optic disc (OR = 1.12, P = .01).

Discussion

Intravenous chemotherapy (chemoreduction) remains an important component of retinoblastoma management for many patients, especially those with bilateral and germline disease.1 Benefits of intravenous chemotherapy, in addition to control of the intraocular tumor, include prevention of distant metastasis, prevention of pineoblastoma, and reduction in second cancers.6–8 In a recent series of 994 eyes with retinoblastoma managed with primary chemoreduction, Shields et al11 found 2-year tumor control (without enucleation or EBRT) by ICRB group (A vs B vs C vs D vs E) at 96% versus 91% vs 91% vs 71% vs 32%, respectively (P < .001) with minimal change at long-term follow-up of 20 years. Thus, chemoreduction remains an effective tool for retinoblastoma treatment. Nevertheless, retinoblastoma management is highly complex and commonly requires multimodality therapy, including IAC, intravitreal/intracameral chemotherapy, and plaque radiotherapy, for globe salvage. In this study, we specifically reviewed risk factors for local tumor recurrence and new tumor formation to provide guidance for individualized patient management.

Retinoblastoma recurrence and new tumor formation are common when using chemotherapy-based globe salvage treatments, making follow-up and possible adjuvant treatments critical aspects of care for these patients.18 In a study of 457 eyes, Shields et al16 found local tumor recurrence at 45% for tumors managed with intravenous chemotherapy alone and 18% for tumors managed with chemoreduction plus additional consolidation with cryotherapy and/or TTT. Risk factors for recurrence included non-White race, greater tumor thickness, macular tumor location, and lack of consolidation therapy.16 A prior study of 158 eyes found 3-year Kaplan–Meier estimated tumor recurrence after chemoreduction at 51%.15 For retinal tumor and vitreous seed recurrence, the only factor predictive of recurrence was presence of subretinal seeds on initial examination. Of those eyes with subretinal seeds at initial presentation, factors predictive of recurrence included largest tumor basal diameter of greater than 15 mm and patient age 12 months or younger at the time of diagnosis. Recurrences were typically detected within the first 3 years of follow-up.15 Regarding new tumor formation, a study of 162 eyes managed with chemoreduction found the Kaplan–Meier estimated frequency of new tumor development was 23% at 1 year with minimal increase to 24% by 5 years.19 Risk factors for new tumor formation included younger patient age at presentation and family history of retinoblastoma.19

In this study of 869 eyes, we found recurrence/new tumor in 556 eyes (64%), of which there were 355 (64%) with main tumor recurrence, 244 (44%) with subretinal seed recurrence, 162 (29%) with vitreous seed recurrence, and 118 (21%) with new tumor. Of the recurrences and new tumor formation, 94% were detected within the first 3 years of follow-up, similar to prior literature.15 The risk for recurrence (main solid tumor, subretinal seed, vitreous seed) increased by ICRB group and was highest for group D tumors with a hazard ratio of 2.3 compared with group A tumors at 2 years, with minimal change thereafter. Similarly, group D tumors were at highest combined risk for any recurrence or new tumor formation at 2 years with a hazard ratio of 1.8. On multivariate analysis, risk factors for recurrence (main solid tumor, subretinal seed, vitreous seed) included younger patient age at diagnosis, more advanced ICRB group, closer tumor distance to optic disc, and presentation with subretinal seeds. Risk factors for any tumor activity, including recurrence or new tumor formation, included younger patient age, germline mutation or 13q deletion syndrome, more advanced ICRB group, and closer tumor distance to optic disc. It is logical that germline mutation may contribute substantially to the risk of new tumor formation but not to the risk of tumor recurrence.

Given the long time period during which this study population was treated, recurrence and new tumor formation were managed with a variety of treatments that evolved over the years, including focal therapy (cryotherapy/TTT), IAC, intravit-real chemotherapy, plaque radiotherapy, EBRT, and enucleation. We currently use more IAC and intravitreal chemotherapy and avoid EBRT due to risk for radiation toxicity, orbital growth retardation, and second cancers.4–10 Moreover, newer modalities of IAC, intravitreal chemotherapy, and intracameral chemotherapy have significantly improved globe salvage rates, especially for advanced group D and E eyes.14,20–23 In the 20-year long-term follow-up of 994 eyes managed with primary chemoreduction, Shields et al11 found that subsequent IAC or plaque radio-therapy was able to achieve globe salvage in an additional 5% versus 26% versus 28% versus 27% versus 19% (P < .001) of eyes by ICRB group (A vs B vs C vs D vs E). Thus, these additional treatment modalities are vitally important for globe salvage, especially for more advanced tumors. Patients with risk factors for recurrence or new tumor formation, including young age, germline mutation, more advanced ICRB group, posterior tumor location, and subretinal seeding, may benefit from adjuvant or primary treatment with these newer, more robust modalities to provide the best chance for globe salvage.

Study limitations include a retrospective design and lack of a uniform treatment algorithm aside from the planned six cycles of intravenous chemo-therapy. However, treatment uniformity was not practical, because therapy for retinoblastoma must be tailored to each individual patient, and treatment modalities have evolved substantially over the long study period. We also acknowledge that our results could be somewhat skewed due to the minimum follow-up criteria of 3 months, but our mean follow-up in this cohort was 81 months. Given that most recurrences were detected within the first 3 years and approximately 30% of patients had a follow-up period shorter than 3 years, recurrences could have been missed in these patients, resulting in an underreporting of the cumulative recurrence rates. Although some lack of follow-up is due to inclusion of patients treated in more recent years, national and international patients sometimes return home when stability is achieved, further contributing to loss of follow-up at the tertiary referral center level. Finally, we recognize that the criteria for primary chemo-reduction therapy have evolved. Given the superior globe salvage rates and lower systemic chemotherapy exposure with use of IAC,23,24 chemoreduction is now typically reserved for patients with germline mutation and bilateral disease. This is demonstrated by the fact that 75% of patients in this study had bilateral disease, likely reflecting a shift away from primary chemoreduction for unilateral disease over time. This introduces some selection bias among patients treated in more recent years. Study strengths are the large number of patients included and long-term follow-up data.

We have reviewed risk factors for recurrence and new tumor formation following primary chemoreduction for retinoblastoma over 20 years of follow-up. We found that most recurrences or new tumors were detected within the first 3 years of treatment. Younger patients with more advanced, posteriorly located tumor and subretinal seeds at presentation were at increased risk for recurrence, and germline mutation contributed to the risk for new tumor formation. Newer, targeted treatments were successfully employed to treat many recurrences without enucleation. These modalities, including IAC, intravitreal/intracameral chemotherapy, and plaque radiotherapy, should be considered as adjuvant or possibly primary therapy for at-risk patients.

References

  1. Shields CL, Lally SE, Leahey AM, et al. Targeted retinoblastoma management: when to use intravenous, intra-arterial, periocular, and intravitreal chemotherapy. Curr Opin Ophthalmol. 2014;25(5):374–385. doi:10.1097/ICU.0000000000000091 [CrossRef]
  2. Mendoza PR, Grossniklaus HE. Therapeutic options for retinoblastoma. Cancer Contr. 2016;23(2):99–109. doi:10.1177/107327481602300203 [CrossRef]
  3. Grossniklaus HE. Retinoblastoma: fifty years of progress. The LXXI Edward Jackson Memorial Lecture. Am J Ophthalmol. 2014;158(5):875–891. doi:10.1016/j.ajo.2014.07.025 [CrossRef]
  4. Kleinerman RA, Tucker MA, Tarone RE, et al. Risk of new cancers after radiotherapy in long-term survivors of retinoblastoma: an extended follow-up. J Clin Oncol. 2005;23(10):2272–2279. doi:10.1200/JCO.2005.05.054 [CrossRef]
  5. Wong JR, Morton LM, Tucker MA, et al. Risk of subsequent malignant neoplasms in long-term hereditary retinoblastoma survivors after chemotherapy and radiotherapy. J Clin Oncol. 2014;32(29):3284–3290. doi:10.1200/JCO.2013.54.7844 [CrossRef]
  6. Shields CL, Meadows AT, Shields JA, Carvalho C, Smith AF. Chemoreduction for retinoblastoma may prevent intracranial neuroblastic malignancy (trilateral retinoblastoma). Arch Ophthalmol. 2001;119(9):1269–1272. doi:10.1001/archopht.119.9.1269 [CrossRef]
  7. Ramasubramanian A, Kytasty C, Meadows AT, Shields JA, Leahey A, Shields CL. Incidence of pineal gland cyst and pineoblastoma in children with retinoblastoma during the chemoreduction era. Am J Ophthalmol. 2013;156(4):825–829. doi:10.1016/j.ajo.2013.05.023 [CrossRef]
  8. Turaka K, Shields CL, Meadows AT, Leahey A. Second malignant neoplasms following chemoreduction with carboplatin, etoposide, and vincristine in 245 patients with intraocular retinoblastoma. Pediatr Blood Cancer. 2012;59(1):121–125. doi:10.1002/pbc.23278 [CrossRef]
  9. Chantada GL, Fandiño AC, Schvartzman E, Raslawski E, Schaiquevich P, Manzitti J. Impact of chemoreduction for conservative therapy for retinoblastoma in Argentina. Pediatr Blood Cancer. 2014;61(5):821–826. doi:10.1002/pbc.24857 [CrossRef]
  10. Kim JW, Abramson DH, Dunkel IJ. Current management strategies for intraocular retinoblastoma. Drugs. 2007;67(15):2173–2185. doi:10.2165/00003495-200767150-00005 [CrossRef]
  11. Shields CL, Bas Z, Tadepalli S, et al. Long-term (20-year) real-world outcomes of intravenous chemotherapy (chemoreduction) for retinoblastoma in 964 eyes of 554 patients at a single centre [published online ahead of print February 12, 2020.]. Br J Ophthalmol. doi:10.1136/bjophthalmol-2019-315572 [CrossRef]
  12. Singh U, Katoch D, Kaur S, Dogra MR, Bansal D, Kapoor R. Retinoblastoma: a sixteen-year review of the presentation, treatment, and outcome from a tertiary care institute in northern India. Ocul Oncol Pathol. 2017;4(1):23–32. doi:10.1159/000477408 [CrossRef]
  13. Chawla B, Jain A, Seth R, et al. Clinical outcome and regression patterns of retinoblastoma treated with systemic chemoreduction and focal therapy: a prospective study. Indian J Ophthalmol. 2016;64(7):524–529. doi:10.4103/0301-4738.190143 [CrossRef]
  14. Shields CL, Mashayekhi A, Au AK, et al. The International Classification of Retinoblastoma predicts chemoreduction success. Ophthalmology. 2006;113(12):2276–2280. doi:10.1016/j.ophtha.2006.06.018 [CrossRef]
  15. Shields CL, Honavar SG, Shields JA, Demirci H, Meadows AT, Naduvilath TJ. Factors predictive of recurrence of retinal tumors, vitreous seeds, and subretinal seeds following chemoreduction for retinoblastoma. Arch Ophthalmol. 2002;120(4):460–464. doi:10.1001/archopht.120.4.460 [CrossRef]
  16. Shields CL, Mashayekhi A, Cater J, Shelil A, Meadows AT, Shields JA. Chemoreduction for retinoblastoma: analysis of tumor control and risks for recurrence in 457 tumors. Trans Am Ophthalmol Soc. 2004;102:35–44;discussion 44–45.
  17. Shields CL, Palamar M, Sharma P, et al. Retinoblastoma regression patterns following chemoreduction and adjuvant therapy in 557 tumors. Arch Ophthalmol. 2009;127(3):282–290. doi:10.1001/archophthalmol.2008.626 [CrossRef]
  18. Berry JL, Kogachi K, Murphree AL, Jubran R, Kim JW. A review of recurrent retinoblastoma: Children's Hospital Los Angeles classification and treatment guidelines. Int Ophthalmol Clin. 2019;59(2):65–75. doi:10.1097/IIO.0000000000000269 [CrossRef]
  19. Shields CL, Shelil A, Cater J, Meadows AT, Shields JA. Development of new retinoblastomas after 6 cycles of chemoreduction for retinoblastoma in 162 eyes of 106 consecutive patients. Arch Ophthalmol. 2003;121(11):1571–1576. doi:10.1001/archopht.121.11.1571 [CrossRef]
  20. Gobin YP, Dunkel IJ, Marr BP, Brodie SE, Abramson DH. Intra-arterial chemotherapy for the management of retinoblastoma: four-year experience. Arch Ophthalmol. 2011;129(6):732–737. doi:10.1001/archophthalmol.2011.5 [CrossRef]
  21. Shields CL, Manjandavida FP, Lally SE, et al. Intra-arterial chemotherapy for retinoblastoma in 70 eyes: outcomes based on the international classification of retinoblastoma. Ophthalmology. 2014;121(7):1453–1460. doi:10.1016/j.ophtha.2014.01.026 [CrossRef]
  22. Shields CL, Jorge R, Say EA, et al. Unilateral retinoblastoma managed with intravenous chemotherapy versus intra-arterial chemotherapy. Outcomes based on the International Classification of Retinoblastoma. Asia Pac J Ophthalmol (Phila). 2016;5(2):97–103. doi:10.1097/APO.0000000000000172 [CrossRef]
  23. Dalvin LA, Kumari M, Essuman VA, et al. Primary intra-arterial chemotherapy for retinoblastoma in the intravitreal chemotherapy era: five years of experience. Ocul Oncol Pathol. 2019;5(2):139–146. doi:10.1159/000491580 [CrossRef]
  24. Abramson DH, Fabius AW, Francis JH, et al. Ophthalmic artery chemosurgery for eyes with advanced retinoblastoma. Ophthalmic Genet. 2017;38(1):16–21. doi:10.1080/13816810.2016.1244695 [CrossRef]

Patient Demographics

Patient DemographicsRecurrencea (556 Eyes of 408 Patients) (%)No Recurrence (313 Eyes of 262 Patients) (%)P869 Eyes of 551b Patients (%)
Age at presentation (months), mean (median, range)10 (7, 0 to 80)13 (9, 0 to 84).02c11 (8, 0 to 84)
Race
  White257 (63)169 (65).74355 (64)
  African American44 (11)26 (10).8056 (10)
  Asian42 (10)26 (10).9058 (11)
  Hispanic52 (13)29 (11).5563 (11)
  Central/South Asian7 (2)6 (2).789 (2)
  Middle Eastern6 (1)6 (2).5510 (2)
Sex
  Male211 (52)134 (51).94283 (51)
  Female197 (48)128 (49)268 (49)
Affected eye
  Right292 (53)156 (50).48448 (52)
  Left264 (47)157 (50)421 (48)
Laterality
  Unilateral retinoblastoma89 (22)51 (19).50140 (25)
  Bilateral retinoblastoma319 (78)211 (81)411 (75)
Hereditary pattern
  Somatic189 (46)135 (52).21278 (50)
  Germline199 (49)118 (45).38250 (45)
  13q deletion syndrome20 (5)9 (3).4423 (4)

Clinical Features at Diagnosis

Clinical FeaturesRecurrencea (556 Eyes of 408 Patients) (%)No Recurrence (313 Eyes of 262 Patients) (%)PbTotal (869 Eyes of 551 Patients) (%)
ICRB Classification
  A25 (4)29 (9).0154 (6)
  B113 (20)85 (27).02198 (23)
  C73 (13)54 (17).11127 (15)
  D166 (30)63 (20).002229 (26)
  E99 (18)58 (19).85157 (18)
  Unknownc80 (14)24 (8).004104 (12)
No. of tumors per eye, mean (median, range)2 (2, 1 to 10)2 (1, 1 to 10).012 (2, 1 to 10)
Largest tumor diameter (mm), mean (median, range)12.1 (12.0, 0.3 to 24.0)10.1 (9.0, 0.4 to 24.0)< .00111.4 (11.0, 0.3 to 24.0)
Largest tumor thickness (mm), mean (median, range)d6.3 (5.6, 0.0 to 23.0)5.3 (4.8, 0.0 to 19.7)< .0015.9 (5.0, 0.0 to 23.0)
Distance to foveola (mm), mean (median, range)1.9 (0.0, 0.0 to 15.0)3.2 (1.0, 0.0 to 15.0)< .0012.4 (0.0, 0.0 to 15.0)
Distance to optic disc (mm), mean (median, range)1.5 (0.0, 0.0 to 16.0)2.9 (2.0, 0.0 to 16.0)< .0012.0 (0.0, 0.0 to 16.0)
Vitreous seeds
  None364 (65)210 (67).65574 (66)
  1 quadrant61 (11)30 (10).5791 (10)
  2 quadrants33 (6)12 (4).2045 (5)
  3 quadrants15 (3)6 (2).5021 (2)
  4 quadrants49 (9)24 (8).6173 (8)
  No view34 (6)31 (10).0465 (7)
Subretinal seeds
  None335 (60)221 (71).003556 (64)
  1 quadrant57 (10)13 (4).00270 (8)
  2 quadrants70 (13)18 (6)< .00188 (10)
  3 quadrants17 (3)10 (3).9927 (3)
  4 quadrants44 (8)19 (6).3463 (7)
  No view33 (6)32 (10).0265 (7)
Subretinal fluid
  None287 (52)196 (63).002483 (56)
  ≤ 1 quadrant54 (10)28 (9).7282 (9)
  2 quadrants55 (10)17 (5).0372 (8)
  3 quadrants28 (5)15 (5).9943 (5)
  4 quadrants98 (18)26 (8)< .001124 (14)
  No view34 (6)31 (10).0465 (7)
Anterior segment findings
  Anterior chamber seeds2 (<1)7 (2).019 (1)
  Iris neovascularization21 (4)16 (5).3837 (4)
  Nystagmus38 (7)16 (5).3854 (6)

Treatment Features and Outcomes

Treatment FeaturesRecurrencea (556 Eyes of 408 Patients) (%)No Recurrence (313 Eyes of 262 Patients) (%)PbTotal (869 Eyes of 551 Patients) (%)
Chemotherapy started elsewhere208 (37)111 (35).61319 (37)
Chemotherapy regimen
  Vincristine, etoposide, carboplatin479 (86)269 (86).99748 (86)
  Otherc77 (14)44 (14)121 (14)
No. of chemotherapy cycles, mean (median, range)d6 (6, 1 to 24)6 (6, 1 to 6).116 (6, 1 to 24)
Focal consolidation required
  TTT249 (45)119 (38).05368 (42)
    Mean no. (median, range)3 (3, 1 to 8)3 (3, 1 to 6).063 (3, 1 to 8)
  Cryotherapy332 (60)109 (35)< .001441 (51)
    Mean no. (median, range)3 (3, 1 to 17)2 (2, 1 to 11)< .0013 (3, 1 to 17)
Additional therapy required
  IAC143 (26)8 (3)< .001151 (17)
  Intravitreal chemotherapy61 (11)4 (1)< .00165 (7)
  Plaque162 (30)5 (2)< .001167 (19)
  EBRT96 (17)21 (7)< .001117 (13)

Ocular Outcomes

Follow-up (months), mean (median, range)84 (63, 3 to 282)75 (57, 3 to 299).0681 (60, 3 to 299)
Tumor regression type
  02 (1)1 (< 1).993 (< 1)
  193 (17)29 (9).003122 (14)
  233 (6)32 (10).0265 (7)
  3365 (66)185 (59).06550 (63)
  451 (9)43 (14).0494 (11)
  Regression type not available12 (2)23 (7)< .00135 (4)
Main tumor recurrence355 (64)0 (0)NA355 (41)
  Mean time to recurrence (median, range)16 (11, 3 to 165)NA16 (11, 3 to 165)
Subretinal seed recurrence244 (44)0 (0)NA244 (28)
  Mean time to recurrence (median, range)12 (8, 3 to 133)NA12 (8, 3 to 133)
Vitreous seed recurrence162 (29)0 (0)NA162 (19)
  Mean time to recurrence (median, range)18 (14, 3 to 150)NA18 (14, 3 to 150)
New tumor formation118 (21)0 (0)NA118 (14)
  Mean time to new tumor (median, range)10 (4, 1 to 187)NA10 (4, 1 to 187)
Treatment for recurrence or new tumor
  Focal (TTT or cryotherapy)294 (53)294 (34)
  IAC125 (22)125 (14)
  Intravitreal chemotherapy36 (6)NANA36 (4)
  Plaque120 (22)120 (14)
  EBRT57 (10)57 (7)
  Enucleation49 (9)49 (6)
Enucleation118 (21)44 (14).01162 (19)
Reason for enucleationn = 118n = 44N = 162
  Tumor progression84 (71)29 (66).57113 (70)
  Blind eye with NVG/RD/VH/phthisis34 (29)15 (34)49 (30)
  Time to enucleation, mean (median, range)25 (20, 3 to 187)9 (6, 1 to 63)< .00121 (15, 1 to 187)

Systemic Outcomesen = 408n = 262N = 551

Metastasis9 (2)6 (2).9911 (2)
Pineoblastoma5 (1)5 (2).528 (1)
Second cancerf5 (1)5 (2).529 (2)
Death5 (1)2 (1).717 (1)

Outcomes by Cox Regression Analysis for Recurrence and New Tumora

TimeICRB GroupCumulative Recurrenceb (%)HRc [95% CI]PdPeCumulative Recurrence and New Tumor Formation (%)HRc [95% CI]PdPe
1 yearA30NANANA35NANANA
B331.1 [0.6 to 1.9].70.54381.1 [0.6 to 1.7].84.54
C270.9 [0.5 to 1.5].60.17310.8 [0.5 to 1.4].47.18
D471.7 [1.0 to 2.9].04.60521.6 [1.0 to 2.6].06.56
E431.6 [0.9 to 2.8].08.21461.5 [0.9 to 2.4].15.20
All38NA< .001.00142NA.001.001

2 yearsA33NANANA41NANANA
B441.3 [0.9 to 2.2].31.82491.2 [0.7 to 1.9].52.67
C451.3 [0.7 to 2.2].38.74491.1 [0.7 to 1.8].67.62
D622.1 [1.3 to 2.5].003.32661.8 [1.2 to 2.9].01.36
E531.8 [1.1 to 3.0].02.20571.6 [1.0 to 2.5].06.19
All51NA< .001.0255NA< .001.01

3 yearsA35NANANA44NANANA
B481.3 [0.8 to 2.2].24.90531.2 [0.7 to 1.8].52.63
C501.3 [0.8 to 2.2].28.90531.1 [0.7 to 1.8].68.63
D642.1 [1.3 to 3.4].002.28691.8 [1.2 to 2.7].01.37
E581.9 [1.2 to 3.1].01.15621.6 [1.0 to 2.5].04.17
All54NA< .001.0259NA< .001.01

5 yearsA35NANANA44NANANA
B501.4 [0.8 to 2.3].20.99551.2 [0.8 to 1.9].44.59
C501.4 [0.8 to 2.3].24.93541.1 [0.7 to 1.8].62.65
D682.3 [1.4 to 3.6].001.20721.9 [1.2 to 2.9].004.30
E592.0 [1.2 to 3.2].01.13631.6 [1.1 to 2.6].03.13
All56NA< .001.0161NA< .001.004

10 yearsA35NANANA46NANANA
B501.4 [0.9 to 2.3].18.94551.2 [0.7 to 1.8].52.64
C531.4 [0.9 to 2.4].17.88561.1 [0.7 to 1.8].59.71
D692.3 [1.4 to 3.7].001.16721.8 [1.2 to 2.7].01.32
E592.0 [1.2 to 3.2].01.11631.6 [1.0 to 2.5].04.15
All57NA< .001.0161NA< .001.01

15 yearsA35NANANA46NANANA
B511.4 [0.9 to 2.3].16.84561.2 [0.8 to 1.8].45.60
C541.4 [0.9 to 2.4].14.85581.2 [0.8 to 1.9].47.80
D692.3 [1.4 to 3.7].001.16731.8 [1.2 to 2.8].01.31
E592.0 [1.2 to 3.3].01.11631.6 [1.0 to 2.5].03.14
All57NA< .001.0262NA< .001.01

20 yearsA35NANANA46NANANA
B511.4 [0.9 to 2.3].16.84571.2 [0.8 to 1.9].39.73
C541.4 [0.9 to 2.4].14.85581.2 [0.8 to 1.9].45.80
D692.3 [1.4 to 3.7].001.16731.8 [1.2 to 2.8].004.32
E592.0 [1.2 to 3.3].01.11631.6 [1.0 to 2.5].03.14
All57NA< .001.0262NA< .001.01

Predictors of Recurrence and New Tumor by Logistic Regression Analysis

Predictive Factors of RecurrenceaOR [95% CI]PbPredictive Factors of Recurrence and New Tumor FormationOR [95% CI]Pb
Univariate analysis
Age at diagnosis (per 1 month decrease)1.02 [1.01 to 1.03].003Age at diagnosis (per 1 month decrease)1.02 [1.01 to 1.04]< .001
Non-White race1.16 [0.87 to 1.55].31Non-White race1.07 [0.80 to 1.43].64
Male sex1.16 [0.89 to 1.52].28Male sex1.07 [0.81 to 1.41].64
Unilateral disease1.01 [0.70 to 1.45].98Unilateral disease1.02 [0.70 to 1.49].91
Germline mutation or 13q deletion1.22 [0.93 to 1.60].16Germline mutation or 13q deletion1.39 [1.05 to 1.83].02
ICRB classification1.27 [1.13 to 1.43]< .001ICRB classification1.21 [1.08 to 1.36].001
No. of tumors1.12 [1.02 to 1.24].01No. of tumors1.13 [1.02 to 1.24].02
Tumor largest basal diameter (per 1 mm increase)1.05 [1.03 to 1.08]< .001Tumor largest basal diameter (per 1 mm increase)1.05 [1.03 to 1.08]< .001
Tumor thickness (per 1 mm increase)1.08 [1.04 to 1.12]< .001Tumor thickness (per 1 mm increase)1.07 [1.03 to 1.11].001
Distance to foveola (per 1 mm decrease)1.13 [1.08 to 1.18]< .001Distance to foveola (per 1 mm decrease)1.11 [1.07 to 1.16]< .001
Distance to optic disc (per 1 mm decrease)1.18 [1.12 to 1.24]< .001Distance to optic disc (per 1 mm decrease)1.17 [1.11 to 1.23]< .001
Vitreous seeds at presentation1.33 [0.97 to 0.83].08Vitreous seeds at presentation1.27 [0.91 to 1.76].16
Subretinal seeds at presentation2.27 [1.64 to 3.15]< .001Subretinal seeds at presentation2.07 [1.48 to 2.89]< .001
Subretinal fluid at presentation1.68 [1.25 to 2.25].001Subretinal fluid at presentation1.87 [1.37 to 2.54]< .001
Anterior chamber seeds at presentation5.22 [1.08 to 25.33].04Anterior chamber seeds at presentation6.44 [1.33 to 31.21].02
Tumor regression typeTumor regression type
  00.32 [0.03 to 3.59].36  01.07 [0.10 to 11.81].96
  11.53 [1.01 to 2.31].04  11.86 [1.19 to 2.89].01
  20.57 [0.34 to 0.94].03  20.52 [0.31 to 0.87].01
  31.25 [0.93 to 1.67].14  31.16 [0.8 to 1.56].34
  40.59 [0.38 to 0.90].02  40.59 [0.39 to 0.92].02
Multivariate analysis
Age at diagnosis (per 1 month decrease)1.02 [1.00 to 1.04].02Age at diagnosis (per 1 month decrease)1.03 [1.01 to 1.05].003
ICRB classification1.24 [1.05 to 1.47].01Germline mutation or 13q deletion1.47 [1.06 to 2.04].02
Distance to optic disc (per 1 mm decrease)1.11 [1.01 to 1.21].03ICRB classification1.20 [1.02 to 1.43].03
Subretinal seeds at presentation1.66 [1.09 to 2.53].02Distance to optic disc (per 1 mm decrease)1.12 [1.03 to 1.23].01
Authors

From the Ocular Oncology Service, Wills Eye Hospital, Thomas Jefferson University, Philadelphia, Pennsylvania (LAD, ZB, ST, RR, SV, RP, CLS); and the Department of Ophthalmology, Mayo Clinic, Rochester, Minnesota (LAD).

Supported in part by the Eye Tumor Research Foundation, Philadelphia, Pennsylvania (CLS).

The authors have no financial or proprietary interest in the materials presented herein.

Correspondence: Carol L. Shields, MD, Ocular Oncology Service, Suite 1440, Wills Eye Hospital, 840 Walnut Street, Philadelphia, PA 19107. Email: carolshields@gmail.com

Received: February 23, 2020
Accepted: March 30, 2020

10.3928/01913913-20200417-01

Sign up to receive

Journal E-contents