Vitreoretinal lymphoma (VRL) is a rare intraocular malignancy that can affect the vitreous, retina, subretinal pigment epithelium (RPE), and, rarely, the optic nerve.1 Originally described as reticulum cell sarcoma in 1968, the association of VRL with central nervous system (CNS) lymphoma was established in the 1980s.2,3 VRL can present as isolated primary VRL or preceding, concurrent, or subsequent to CNS lymphoma.4,5 Most cases of VRL are classified as diffuse large B-cell lymphomas (DLBCL).1 Of patients with primary CNS lymphoma, 15% to 25% will develop VRL, whereas 56% to 90% of patients presenting with VRL will ultimately develop CNS lymphoma.1,6 Due to the association with CNS lymphoma, life prognosis for VRL is consistently poor, even with treatment.4,7
Historically, median survival for VRL has been 1 to 2 years.7–9 More recent studies of VRL report better overall survival (22–58 months), but outcomes remain relatively poor and most studies do not support the conclusion that survival is improved by systemic chemotherapy or prophylactic brain radiotherapy.10–14 Due to the rarity of the disease and the absence of a national registry, few studies have examined predictors of survival in patients with VRL. In one study, survival was shortest in patients who had concurrent VRL and CNS lymphoma with survival of 18 months compared with 31 months in the entire cohort.13 Herein, we examine a consecutive series of VRL patients seen at a single tertiary-care ocular oncology practice to identify clinical features predictive of time to CNS lymphoma and time to death.
The medical records were reviewed to identify all patients diagnosed with VRL at the Ocular Oncology Service, (withheld for blinded review) from January 1, 1984 to July 30, 2018. Patients with VRL were included regardless of CNS lymphoma status at date of presentation. Institutional Review Board approval was obtained from (withheld for blinded review), and this study was in compliance with the Health Insurance Portability and Accountability Act. Informed consent was obtained.
All patients underwent a complete eye examination by an ocular oncologist, including anterior segment evaluation and funduscopic examination using indirect ophthalmoscopy. Color fundus photography, B scan ultrasonography, fluorescein angiography, and optical coherence tomography were performed as available and as needed. Patients were managed in conjunction with a medical neuro-oncology team.
Clinical, photographic, and cytopathology records were retrospectively reviewed for past, concurrent, or subsequent history of CNS or systemic lymphoma, previous vitrectomy, previous ocular or systemic treatment, patient demographics (age, sex, race), and presenting symptoms. Clinical features included laterality (unilateral, bilateral), presenting Snellen visual acuity, anterior chamber cell or keratic precipitates, vitreous cellularity, retinal or subretinal infiltration, retinal detachment, sub-RPE infiltration, optic nerve abnormalities (edema, hemorrhage, atrophy), and lymphoma subtype. Treatment modalities included observation after previous diagnostic vitrectomy, systemic chemotherapy, intravitreal chemotherapy, and external beam radiotherapy (EBRT). Outcomes included follow-up duration, initial response to therapy (regression, no regression), recurrence (defined as disease relapse >3 months after discontinuing treatment), final tumor control at date last seen (no regression, partial regression, or complete regression defined as absence of vitreous cell and resolution of any previous intraretinal or sub-RPE infiltration), association with CNS or systemic lymphoma, final Snellen visual acuity, and death. Visual acuity outcomes and response to treatment were reported for patients with follow-up of at least 3 months. Obituaries were searched and primary care or medical oncology offices were called to determine date of death for all patients who were no longer actively following up with the Ocular Oncology Service at (withheld for blinded review). Time from ocular symptoms to development of CNS lymphoma and death was recorded.
Statistical analysis was performed using SPSS Statistics Software Version 22 (IBM, Armonk, NY). Mean times to CNS lymphoma and death were calculated per unique patient using Kaplan-Meier analysis, with LogRank test used to compare mean times by risk factor (age, male sex, CNS lymphoma status, systemic lymphoma status, bilateral ocular involvement, retinal or subretinal infiltration, sub-RPE involvement, and treatment modality). When risk factors were significant, multivariate analysis was performed to adjust for potential confounders. Patients who presented with CNS lymphoma before VRL were censored at time 0 when considering mean time to CNS lymphoma. Patients without follow-up were censored at time 0 when considering time to CNS lymphoma, and were censored at time 0 when considering time to death unless an obituary was found or date of death was confirmed by another healthcare provider. Cox regression analysis was used to calculate the odds of developing CNS lymphoma or death by risk factor. Odds ratios (ORs) are reported as OR (95% confidence interval). Competing risk analysis was performed using R statistical software. Statistical significance was defined as P < 0.05.
There were 168 eyes of 95 patients diagnosed with VRL at the Ocular Oncology Service, (withheld for blinded review) from January 1, 1984 to July 30, 2018. Demographics are listed in Table 1. Mean patient age at presentation was 67 years (median 66, range 28–91 years). Sex was evenly distributed with 52 (55%) males and 43 (45%) females, with most patients of white race (n = 82, 86%). Ocular involvement was unilateral in 22 (23%) and bilateral in 73 (77%) patients.
Clinical features are listed in Table 1. The most common presenting features were blurred vision in 95 (57%), floaters in 52 (31%), and asymptomatic presentation in 21 (13%) eyes. Visual acuity at presentation ranged from 20/20 to no light perception, but nearly half of eyes presented with good visual acuity of 20/40 or better (n = 80, 48%). Anterior chamber cell was found in 37 (22%) and keratic precipitates in 35 (21%) eyes. Vitreous cellular infiltration was present in 131 (78%) eyes, retinal or subretinal infiltration in 39 eyes (23%), and sub-RPE infiltration in 45 eyes (27%). Optic nerve involvement was found with optic disc swelling in 6 (4%), hemorrhage in 2 (1%), and atrophy in 7 (4%) eyes. Subtype of lymphoma was most commonly DLBCL (n = 90, 54%) or unspecified on cytology reports (n = 76, 45%).
Management details are listed in Table 2. There were 125 (74%) eyes of 70 patients with follow-up and 43 (26%) eyes of 25 patients with no follow-up that were excluded when considering treatment modalities, VRL control, and visual acuity outcomes. Of the 125 eyes of 70 patients with at least 3 months follow-up, there were 58 (35%) eyes initially managed with observation after diagnostic vitrectomy, of which 34 (59%) required subsequent additional treatment. Initial treatment for 101 eyes of 51 patients included systemic chemotherapy in 8 (8%), intravitreal chemotherapy in 17 (17%), concurrent systemic chemotherapy and intravitreal chemotherapy in 5 (5%), and EBRT in 71 (70%) eyes. Overall, of 101 eyes treated with chemotherapy or radiotherapy, systemic chemotherapy was used in 61 (60%), intravitreal chemotherapy in 39 (39%), and EBRT in 83 (82%) eyes.
Outcomes regarding VRL control and visual acuity for 125 eyes of 70 patients with at least 3 months’ follow-up are listed in Table 3. Of the 101 eyes initially treated with interventions (chemotherapy or radiotherapy), 94 (93%) eyes showed initial regression, but subsequent recurrence was detected in 22 (23%) eyes. At final mean follow-up of 25 months (median 12, range 3–229 months) for 125 (74%) eyes of 70 available patients, complete regression of VRL was achieved in 101 (81%) eyes. Visual acuity at date last seen for 125 eyes of 70 patients with follow-up ranged from 20/20 to no light perception, with nearly half of eyes maintaining visual acuity of 20/40 or better (n = 61, 49%).
Outcomes regarding CNS lymphoma and death for all 95 study patients are listed in Table 3. CNS lymphoma was found in 50 (53%) patients, diagnosed before (n = 36, 38%) or after (n = 14, 15%) VRL. Systemic lymphoma was found in 18 (19%) patients, diagnosed before (n = 15, 16%) or after (n = 3, 3%) VRL; of these, both CNS and systemic lymphoma were found in 9 (9%) patients. There were 54 (57%) patients who died, at mean age of 66 years (median 69, range 39–94 years) from CNS lymphoma (n = 20, 37%) or unknown cause (n = 34, 63%).
Kaplan-Meier estimated times to CNS lymphoma with ORs by clinical factor are listed in Table 4. Mean time to CNS lymphoma for all patients was 56 months (median 35 months). Mean time to CNS lymphoma in patients with isolated VRL did not differ for those patients treated with systemic chemotherapy compared with those who received no systemic treatment [53 vs 53 months, OR 0.6 (0.3–1.4), P = 0.24]. Time to CNS lymphoma did not differ based on patient age, sex, bilateral ocular involvement, presence of retinal infiltration, presence of sub-RPE infiltration, presence or timing of systemic lymphoma, or primary method of VRL treatment (P > 0.05).
Kaplan-Meier estimated times to death with odds ratios by clinical factor are listed in Table 5. Mean survival time after onset of ocular symptoms was 66 months (median 49 months) for all patients and did not differ when comparing CNS lymphoma diagnosed before VRL versus after VRL versus no CNS lymphoma at all (67 vs 60 vs 64 months, P > 0.05). The presence of sub-RPE infiltration was associated with shorter mean time to death [46 vs 76 months, OR 1.9 (1.0–3.5), P = 0.04], and remained a significant risk factor [OR 2.4 (1.1–5.4), P = 0.04] on multivariate analysis after adjustment for potential confounders, including patient age, bilateral ocular involvement, CNS lymphoma, systemic lymphoma, and treatment modality (systemic chemotherapy, intravitreal methotrexate, intravitreal melphalan, and EBRT) (Fig. 1). Older patient age, when analyzed as a continuous variable, was associated with increased risk of death [OR 1.0 (1.0–1.6), P = 0.02], but this association disappeared on multivariate analysis after adjustment for potential confounders, including patient age, bilateral ocular involvement, CNS lymphoma, systemic lymphoma, and treatment modality (P = 0.10). Mean survival did not differ based on patient sex, bilateral ocular involvement, presence of retinal infiltration, presence or timing of CNS or systemic lymphoma, or primary method of VRL treatment (P > 0.05). Given that cause of death was unknown in many cases, competing risk analysis was performed to assess the impact of sub-RPE infiltration on lymphoma-related death. When all deceased patients without known lymphoma-related death were considered to have death from causes other than lymphoma, patient survival did not differ by presence of sub-RPE infiltration (P = 0.12).
VRL is a rare, histologically aggressive, intraocular lymphoid malignancy that can present with or without CNS lymphoma.4,5 Although improved survival has been reported in recent years (up to 9 years in one study),15 the impact of current treatments on survival is questionable, and there is still no widely accepted standard of care for these patients.10–13,15 With recent advances in medical oncology, more effective therapy could be on the horizon for patients with VRL, especially those with CNS involvement.16 To most effectively target appropriate candidates for early systemic therapy, ophthalmologists and oncologists should be able to identify those patients at highest risk for poor survival.
In this study, we investigated clinical risk factors associated with time to CNS lymphoma and time to death in patients presenting to a single Ocular Oncology center with VRL. We did not identify any factors associated with time to CNS lymphoma. Specifically, systemic chemotherapy for isolated ocular disease did not prevent CNS lymphoma or prolong the time to CNS lymphoma. Regarding time to death, presence of sub-RPE infiltration was associated with shorter survival time by approximately 2.5 years and remained a significant risk factor for death on multivariate analysis. However, this association disappeared on competing risk analysis for cases of confirmed lymphoma-related death. Older age was also associated with shorter survival time on univariate analysis, but this association disappeared on multivariate analysis. There was no association between survival and sex, bilateral ocular involvement, retinal or subretinal infiltration, presence of CNS or systemic lymphoma, or treatment modality for VRL.
The literature on VRL is limited given the rarity of this disease. Ahmed et al reviewed the Surveillance, Epidemiology, and End Results (SEER) database of 396 patients with ophthalmic DLBCL and found the 5-year overall survival for patients with VRL DLBCL at 41.4% compared with those with ocular adnexal/uveal DLBCL at 59.1%. Furthermore, median survival for VRL DLBCL was 38 months compared with ocular adnexal/uveal DLBCL at 96 months.17 Concurrent presentation of VRL with CNS lymphoma has also been associated with shorter median survival than patients presenting with primary VRL or VRL after CNS lymphoma, with median survival of 18 months compared with 31 months for the entire cohort.13 The presence of any CNS lymphoma throughout the course of disease has been associated with decreased 5-year cumulative survival of 35% versus 68% with no CNS lymphoma.14 In our study, however, we did not find any association between survival and the presence or timing of CNS lymphoma relative to VRL.
Based on the published retrospective datasets, systemic chemotherapy has questionable survival benefit in VRL patients. In one series of 72 patients diagnosed with VRL at Mayo Clinic, 69 of which were DLBCL (33 primary VRL, 18 secondary VRL, and 18 concurrent VRL and CNS or systemic lymphoma), systemic chemotherapy with high-dose methotrexate administered concurrently with local intravitreal chemotherapy increased time to DLBCL relapse or progression in any location and increased time to CNS lymphoma relapse but not overall time to death.15 A retrospective review from the International Primary Central Nervous System Lymphoma Collaborative Group on 83 patients from 16 centers with isolated VRL revealed no difference in survival with local ocular therapy versus therapy with systemic chemotherapy and/or whole brain radiotherapy.10 A similar study from a collaborative European group evaluating 78 patients from 17 centers with isolated VRL revealed development of CNS lymphoma in 10 of 31 (32%) patients treated with local ocular radiotherapy and/or local intravitreal chemotherapy only, 9 of 21 (43%) patients treated with systemic chemotherapy, whole brain radiotherapy, and/or peripheral blood stem cell transplantation, and 9 of 23 (39%) patients treated with a combination of local and systemic therapy, with no difference in 5-year cumulative survival between the three treatment groups.14 Similarly, our study did not show improved survival in patients with VRL only or VRL with CNS lymphoma using systemic chemotherapy.
Although methotrexate is currently the mainstay of therapy for systemic treatment of VRL and related CNS lymphoma, relapse rates are high.1,18 Newer therapies could potentially be more effective, preventing CNS disease and improving survival. A recent study demonstrated durable remission of an aggressive MYC-rearranged subtype of CNS lymphoma with dose-adjusted EPOCH-R (etoposide, prednisone, vincristine, cyclophosphamide, doxorubicin, and rituximab) with 48-month progression-free survival in 71% and 48-month overall survival in 77%.19 However, novel, targeted therapies could also improve outcomes.16 VRL and associated CNS lymphoma are unique in their presentation in an immune-privileged site, making this a distinct molecular disease with high prevalence of MYD88 and CD79B mutations.16,20 Studies have suggested that aberrant activation of the B-cell receptor (BCR) signaling pathway could play an important role in VRL and CNS lymphoma pathogenesis.16 Drugs targeting BCR signaling via inhibition of bruton tyrosine kinase (ibrutinib) or phosphatidylinositol 3-kinase (PI3K) isoforms (idelalisib, duvelisib) could have a role in VRL treatment.16,21,22 Programmed death ligand-1 and 2 have also been implicated in primary CNS lymphoma, indicating a possible therapeutic role for checkpoint inhibitors.16,23,24 Patients at high risk for poor survival could be considered for treatment with novel systemic therapies.
In this study, we found that VRL that initially presented with sub-RPE infiltration could be associated with inferior survival, indicating a subpopulation of patients who might benefit substantially from prophylactic systemic treatment. To our knowledge, this association has not been previously reported and confirmatory studies are required. Potentially, in a cohort with predominantly DLBCL, sub-RPE infiltration could indicate a more aggressive subtype. For example, double and triple hit DLBCL with aberrations in MYC, BCL2, and/or BCL6 have increased risk of CNS lymphoma and worse outcomes with standard systemic therapy.25 It could be that patients with sub-RPE infiltration are more likely to harbor this type of disease and require an intensified chemotherapeutic regimen.
Limitations of our study include its retrospective nature, small number of patients given the rarity of the disease, and the inherent real world limitation of nonstandardized treatment regimens, especially in a disease that has no clearly established standard of care. In particular, systemic chemotherapy was not used in a standardized fashion across all patients, with systemic chemotherapy sometimes being administered prophylactically and sometimes for concurrent CNS disease, various drugs and dosages used depending on the choice of the medical oncologist. Therefore, we cannot draw any definitive conclusions regarding survival benefit or lack thereof with systemic chemotherapy. Additionally, patients diagnosed with CNS lymphoma before VRL were not examined in our office at the time of CNS lymphoma diagnosis. Therefore, it is possible that some patients had asymptomatic VRL earlier than recognized in this report. There were also a significant proportion of patients lost to follow-up (26%). Although obituaries were reviewed for date of death, cause of death could not be ascertained in most cases of lost follow-up, and the association between sub-RPE infiltration and inferior survival was not significant on competing risk analysis when all unconfirmed causes of death were presumed to be unrelated to lymphoma. Nevertheless, we suspect that many patients who died had likely developed CNS lymphoma, as isolated VRL should not be associated with death. Moreover, there could have been additional patients who developed CNS lymphoma after being lost to follow-up in our office, which might account for the lack of association between CNS disease and survival in this study. Strengths of our study include the large number of patients from a single center and careful, standardized documentation of ophthalmic findings with manual review of medical and imaging records.
In summary, in this large single-center study, we found that sub-RPE infiltration in VRL could be associated with shorter survival. Although all VRL patients require co-management with an ocular oncologist and a medical neuro-oncologist, and a sizable majority could benefit from the availability of more effective systemic therapies, high-risk subsets that should be preferentially targeted for more aggressive therapy have not previously been identified. In this regard, patients presenting with primary VRL and sub-RPE infiltration might warrant systemic therapy, especially with novel agents targeting BCR signaling that have shown efficacy in primary CNS lymphoma. Additional studies, especially with larger patient numbers and confirmed causes of death, are warranted to confirm the association of sub-RPE infiltration with decreased survival and investigate whether specific molecular aberrations are associated with this clinical presentation.
1. Chan CC, Rubenstein JL, Coupland SE, et al. Primary vitreoretinal lymphoma
: a report from an International Primary Central Nervous System Lymphoma
Collaborative Group symposium. Oncologist
2. Vogel MH, Font RL, Zimmerman LE, Levine RA. Reticulum cell sarcoma of the retina and uvea. Report of six cases and review of the literature. Am J Ophthalmol
3. Char DH, Margolis L, Newman AB. Ocular reticulum cell sarcoma. Am J Ophthalmol
4. Reichstein D. Primary vitreoretinal lymphoma
: an update on pathogenesis, diagnosis and treatment. Curr Opin Ophthalmol
5. Salomão DR, Pulido JS, Johnston PB, et al. Vitreoretinal
presentation of secondary large B-cell lymphoma
in patients with systemic lymphoma
. JAMA Ophthalmol
6. Pe’er J, Hochberg FH, Foster CS. Clinical review: treatment of vitreoretinal lymphoma
. Ocul Immunol Inflamm
7. Pulido JS, Johnston PB, Nowakowski GS, et al. The diagnosis and treatment of primary vitreoretinal lymphoma
: a review. Int J Retina Vitreous
8. Barr CC, Green WR, Payne JW, et al. Intraocular reticulum-cell sarcoma: clinico-pathologic study of four cases and review of the literature. Surv Ophthalmol
9. Kaplan HJ, Meredith TA, Aaberg TM, Keller RH. Reclassification of intraocular reticulum cell sarcoma (histiocytic lymphoma
). Immunologic characterization of vitreous cells. Arch Ophthalmol
10. Grimm SA, Pulido JS, Jahnke K, et al. Primary intraocular lymphoma
: an International Primary Central Nervous System Lymphoma
Collaborative Group Report. Ann Oncol
11. Cassoux N, Merle-Beral H, Leblond V, et al. Ocular and central nervous system lymphoma
: clinical features and diagnosis. Ocul Immunol Inflamm
12. Jahnke K, Korfel A, Komm J, et al. Intraocular lymphoma
2000–2005: results of a retrospective multicentre trial. Graefes Arch Clin Exp Ophthalmol
13. Cho BJ, Kim DY, Park UC, et al. Clinical features and treatment outcomes of vitreoretinal lymphoma
according to its association with CNS lymphoma
. Ocul Immunol Inflamm
14. Riemens A, Bromberg J, Touitou V, et al. Treatment strategies in primary vitreoretinal lymphoma
: a 17-center European collaborative study. JAMA Ophthalmol
15. Castellino A, Pulido JS, Johnston PB, et al. Role of systemic high-dose methotrexate and combined approaches in the management of vitreoretinal lymphoma
: a single center experience 1990-2018. Am J Hematol
16. Grommes C, Nayak L, Tun HW, Batchelor TT. Introduction of novel agents in the treatment of primary CNS lymphoma
. Neuro Oncol
17. Ahmed AH, Foster CS, Shields CL. Association of disease location and treatment with survival
in diffuse large B-cell lymphoma
of the eye
and ocular adnexal region. JAMA Ophthalmol
18. Chan CC, Sen HN. Current concepts in diagnosing and managing primary vitreoretinal
. Discov Med
19. Dunleavy K, Fanale MA, Abramson JS, et al. Dose-adjusted EPOCH-R (etoposide, prednisone, vincristine, cyclophosphamide, doxorubicin, and rituximab) in untreated aggressive diffuse large B-cell lymphoma
with MYC rearrangement: a prospective, multicentre, single-arm phase 2 study. Lancet Haematol
20. Pulido JS, Salomao DR, Frederick LA, Viswanatha DS. MyD-88 L265P mutations are present in some cases of vitreoretinal lymphoma
21. Wilson WH, Young RM, Schmitz R, et al. Targeting B cell receptor signaling with ibrutinib in diffuse large B cell lymphoma
. Nat Med
22. Lee JH, Jeong H, Choi JW, et al. Clinicopathologic significance of MYD88 L265P mutation in diffuse large B-cell lymphoma
: a meta-analysis. Sci Rep
23. Dalvin LA, Shields CL, Orloff M, et al. Checkpoint inhibitor immune therapy: systemic indications and ophthalmic side effects. Retina
24. Chapuy B, Roemer MG, Stewart C, et al. Targetable genetic features of primary testicular and primary central nervous system
25. Rosenthal A, Younes A. High grade B-cell lymphoma
with rearrangements of MYC and BCL2 and/or BCL6: double hit and triple hit lymphomas and double expressing lymphoma
. Blood Rev