Vertical Hyperreflective Lesions on Optical Coherence Tomography in Vitreoretinal Lymphoma | Ophthalmic Imaging | JAMA Ophthalmology | JAMA Network
[Skip to Navigation]
Sign In
Figure 1.  Example Image of Vertical Hyperreflective Lesion (VHRL)
Example Image of Vertical Hyperreflective Lesion (VHRL)

A, Infrared image, with green line showing scan line. B, Enlarged optical coherence tomography (OCT) image showing VHRL (arrowhead) with midreflectivity spanning between the retinal nerve fiber layer/ganglion cell layer and the retinal pigment epithelium (RPE). The RPE shows mild irregularity. Enlarged fundus photograph (C) and autofluorescence image (D) of the region of VHRL. The green line represents the scan line. Some atrophic lesions are visible in both modalities temporally, but the VHRL itself is not visible.

Figure 2.  Follow-up Infrared and Optical Coherence Tomography Images of Patient 2
Follow-up Infrared and Optical Coherence Tomography Images of Patient 2

A, Taken at baseline examination, the patient had a history of primary vitreoretinal lymphoma (PVRL) and had numerous local and systemic treatments for recurrent disease. The image shows vertical hyperreflective lesions (VHRLs) and minimal retinal pigment epithelium (RPE) irregularity. B, Four days after the baseline before intravitreal methotrexate and rituximab injection showing persistent VHRLs and a marked increase in sub-RPE deposits. C, Four weeks after panel B showing a resolution of sub-RPE deposits and a mild remnant of VHRL in the outer retina. D, Two months after panel C after receiving radiation therapy to the eyes with all the intraretinal and sub-RPE infiltrates resolved and only focal inner segment/outer segment layer loss.

Figure 3.  Follow-up Infrared and Optical Coherence Tomography Images of Patient 1
Follow-up Infrared and Optical Coherence Tomography Images of Patient 1

A, At presentation with a history of secondary vitreoretinal lymphoma following years after being treated for systemic lymphoma. Images show a vertical hyperreflective lesion (VHRL) with minimal retinal pigment epithelium (RPE) irregularity. B, Four weeks after intravitreal rituximab with persistent VHRL and still minimal RPE irregularity. C, Four weeks after repeated intravitreal rituximab injection with an increased number of VHRLs, temporally frank funduscopically visible retinal infiltration (arrowhead), and a slight increase in sub-RPE deposits. D, Two months after systemic high-dose methotrexate and rituximab chemotherapy with a decrease in VHRLs and the funduscopically visible retinal infiltrate with only remnants of both visible but with a marked increase in sub-RPE deposits.

Table.  Patient Demographics
Patient Demographics
1.
Levasseur  SD, Wittenberg  LA, White  VA.  Vitreoretinal lymphoma: a 20-year review of incidence, clinical and cytologic features, treatment, and outcomes.  JAMA Ophthalmol. 2013;131(1):50-55. doi:10.1001/jamaophthalmol.2013.569PubMedGoogle ScholarCrossref
2.
Grimm  SA, Pulido  JS, Jahnke  K,  et al.  Primary intraocular lymphoma: an International Primary Central Nervous System Lymphoma Collaborative Group Report.  Ann Oncol. 2007;18(11):1851-1855. doi:10.1093/annonc/mdm340PubMedGoogle ScholarCrossref
3.
Fend  F, Süsskind  D, Deuter  C, Coupland  SE.  Maligne Lymphome des Auges.  Pathologe. 2017;38(6):515-520. doi:10.1007/s00292-017-0378-6PubMedGoogle ScholarCrossref
4.
Forooghian  F, Merkur  AB, White  VA, Shen  D, Chan  C-C.  High-definition optical coherence tomography features of primary vitreoretinal lymphoma.  Ophthalmic Surg Lasers Imaging. 2011;42 Online(0):e97-e99. doi:10.3928/15428877-20110922-02PubMedGoogle Scholar
5.
Vasconcelos-Santos  DV, De Puy E Souza  GH, de Faria  BB,  et al.  Subretinal pigment epithelial infiltrates in primary vitreoretinal lymphoma.  J Ophthalmic Inflamm Infect. 2011;1(4):171-171. doi:10.1007/s12348-011-0034-xPubMedGoogle ScholarCrossref
6.
Saito  T, Ohguro  N, Iwahashi  C, Hashida  N.  Optical coherence tomography manifestations of primary vitreoretinal lymphoma.  Graefes Arch Clin Exp Ophthalmol. 2016;254(12):2319-2326. doi:10.1007/s00417-016-3395-xPubMedGoogle ScholarCrossref
7.
Chan  C-C, Rubenstein  JL, Coupland  SE,  et al.  Primary vitreoretinal lymphoma: a report from an International Primary Central Nervous System Lymphoma Collaborative Group symposium.  Oncologist. 2011;16(11):1589-1599. doi:10.1634/theoncologist.2011-0210PubMedGoogle ScholarCrossref
8.
Coupland  SE, Chan  CC, Smith  J.  Pathophysiology of retinal lymphoma.  Ocul Immunol Inflamm. 2009;17(4):227-237. doi:10.1080/09273940903168696PubMedGoogle ScholarCrossref
9.
Fend  F, Ferreri  AJM, Coupland  SE.  How we diagnose and treat vitreoretinal lymphoma.  Br J Haematol. 2016;173(5):680-692. doi:10.1111/bjh.14025PubMedGoogle ScholarCrossref
10.
Davis  JL.  Intraocular lymphoma: a clinical perspective.  Eye (Lond). 2013;27(2):153-162. doi:10.1038/eye.2012.250PubMedGoogle ScholarCrossref
11.
Barry  RJ, Tasiopoulou  A, Murray  PI,  et al.  Characteristic optical coherence tomography findings in patients with primary vitreoretinal lymphoma: a novel aid to early diagnosis.  Br J Ophthalmol. 2018;102(10):1362-1366. doi:10.1136/bjophthalmol-2017-311612PubMedGoogle ScholarCrossref
12.
Marsiglia  M, Gallego-Pinazo  R, Cunha de Souza  E,  et al.  Expanded clinical spectrum of multiple evanescent white dot syndrome with multimodal imaging.  Retina. 2016;36(1):64-74. doi:10.1097/IAE.0000000000000685PubMedGoogle ScholarCrossref
Original Investigation
November 29, 2018

Vertical Hyperreflective Lesions on Optical Coherence Tomography in Vitreoretinal Lymphoma

Author Affiliations
  • 1Feinberg School of Medicine, Department of Ophthalmology, Northwestern University, Chicago, Illinois
  • 2Department of Ophthalmology and Optometry, Medical University of Vienna, Vienna, Austria
JAMA Ophthalmol. 2019;137(2):194-198. doi:10.1001/jamaophthalmol.2018.5835
Key Points

Question  Are there optical coherence tomography (OCT) features that might suggest the diagnosis of vitreoretinal lymphoma?

Findings  This case series study found that microscopic lesions were visible as hyperreflective columns between the retinal nerve fiber layer and the retinal pigment epithelium on OCT in 5 of 7 patients (7 of 12 eyes) with vitreoretinal lymphoma. Vertical hyperreflective lesions were often seen on subretinal pigment epithelial deposits and in many cases preceded the appearance of such lesions.

Meaning  Vertical hyperreflective lesions on OCT may suggest the diagnosis of vitreoretinal lymphoma and provide clues as to the pathogenesis of this disease.

Abstract

Importance  Vitreoretinal lymphoma is a diagnostic challenge and the pathophysiology is still unclear.

Objective  To describe an imaging finding seen on optical coherence tomography (OCT) of patients with vitreoretinal lymphoma.

Design, Setting, and Participants  This case series study was a retrospective medical record review of patients who received a diagnosis of vitreoretinal lymphoma at the Department of Ophthalmology at Northwestern University between July 2014 and January 2016.

Main Outcomes and Measures  Optical coherence tomography findings in vitreoretinal lymphoma.

Results  We identified 7 patients (4 women [57.1%]; mean [range] age, 62.4 [45-75] years; 12 eyes) with intraocular lymphoma involving the retina (5 patients [71.4%] with primary vitreoretinal or central nervous system lymphoma with ocular involvement, 1 patient [14.3%] with testicular lymphoma with secondary central nervous system lymphoma and vitreoretinal lymphoma, and 1 patient [14.3%] with secondary vitreoretinal lymphoma). We identified vertical hyperreflective lesions that showed moderate or high reflectivity and affected all layers of the neuroretina in 5 patients (7 of 12 eyes [58.3%]). These often preceded the development of subretinal pigment epithelial deposits and were often localized around second-order and third-order retinal vessels. In most cases, they resolved with minimal or no scarring after the initiation of chemotherapy.

Conclusions and Relevance  Vertical hyperreflective lesions are a common physical finding on OCT in eyes with vitreoretinal lymphoma.

Introduction

Vitreoretinal lymphomas (VRL) are rare diseases. They comprise about 1% to 3% of intraocular tumors and the incidence varies between 0.02 to 0.05 of 100 000 people in the North American population.1

In almost 90% of cases, VRLs are diffuse large B cell lymphomas (DLBCLs) and are considered to be a subtype of the primary DLBCLs of the central nervous system (CNS). Sixteen percent to 34% of patients with VRL have CNS involvement at the time of diagnosis and 35% to 90% will develop CNS lymphoma during the course of their disease. Conversely, 15% to 25% of patients with CNS lymphoma develop VRL.2,3

Optical coherence tomography (OCT) is an important noninvasive diagnostic tool in diagnosing and managing retinal diseases. In VRL, most of the morphologic features that are seen on OCT are nonspecific.4-6 The aim of this study was to describe an OCT feature in patients with confirmed VRL.

Methods

We conducted a retrospective medical record review of patients who received a diagnosis of VRL at the Department of Ophthalmology at Northwestern University. The protocol was approved by the institutional ethics committee, who waived patient consent, and adhered to the Declaration of Helsinki. Patients who had a diagnosis of intraocular lymphoma who were seen by 2 senior attending physicians (D.A.G. and L.M.J.) between July 2014 and January 2018 were identified using a search of electronic medical records. Patient records and all imaging modalities that were carried out at the initial visit and subsequent visits were collected and reviewed for completeness and consistency. The exclusion criteria were an uncertain diagnosis and inadequate OCT image quality. Optical coherence tomography examinations were conducted with the Spectralis OCT (Heidelberg Engineering) with a raster scan covering the whole macular region (30° × 25°), and in 4 of 7 cases (57.1%) additional scans were centered on the temporal retina or the temporal vascular arcades. The standard scanning pattern consisted of 61 scans (with one exception of 121 scans) and an Automatic Real-time Tracking setting between 9 and 14. Color photos, fundus autofluorescence, and infrared reflectance images were analyzed using Adobe Photoshop (version 13.0.1 CS6 Extended; Adobe Systems Inc) using the levels adjustments to correct brightness, contrast, and gamma.

Results

We identified 12 eyes of 7 patients (4 women [57.1%] and 3 men [42.9%]) with VRL. The mean (range) age at initial presentation was 62.4 (45-75) years. The Table shows the baseline characteristics of the patients. Five patients (71.4%) had primary VRL and 2 patients (28.6%) had secondary VRL. One of the patients with secondary VRL received a diagnosis of systemic DLBCL 3 years before developing intraocular lymphoma that involved the retina and vitreous. The other patient received a diagnosis of testicular lymphoma with brain and intraocular involvement that was confirmed by chorioretinal biopsy results. In both of these cases, the ocular lymphoma involved the vitreous, neuroretina, and the subretinal pigment epithelial space (sub-RPE) but stayed above the Bruch membrane and as such was classified as secondary VRL. All of the remaining patients with primary VRL received a diagnosis of DLBCL based on either vitreous biopsy results (2 patients [40%]) or brain biopsy results (3 patients [60%]).

In 7 eyes (58.3%) of 5 patients, we identified vertical hyper-reflective columns that varied in width but extended from the inner retina (ganglion cell layer [GCL] or retinal nerve fiber layer [RNFL]) to the outermost part of the neuroretina and the retinal pigment epithelium (RPE). We describe these findings as vertical hyperreflective lesions (VHRLs) (Figure 1). The intensity of VHRLs showed great variability, ranging from a hyperreflectivity similar to the RPE to a much more subtle midreflectivity similar to the GCL, but they were always more hyperreflective than the surrounding retinal tissue. Vertical hyperreflective lesions were most commonly located along the major vessel arcades or temporal to the fovea and were only rarely visible on OCT scans centered on the fovea. We found that VHRLs were often but not always adjacent to retinal vessels (Figure 1, Figure 2, and Figure 3; eFigure in the Supplement). None of the VHRLs were detectable on color fundus photography or on the infrared images even after magnification and appropriate brightness, contrast, and gamma adjustments (Figure 1). On fundus autofluorescence images that were of gradable quality, VHRLs colocalized with sub-RPE deposits were usually hyperautofluorescent, but VHRLs without sub-RPE deposits were undetectable. The eFigure in the Supplement describes a collection of representative VHRLs in the cohort.

On follow-up examination, in some cases VHRLs resolved as quickly as 2 weeks after intravitreal or systemic immunotherapy or chemotherapy. In other patients, VHRLs persisted for up to 4 to 8 weeks. After resolution of the lesions there were no apparent signs of atrophy in the inner retina. In 2 patients (28.6%) in whom VHRLs were located on sub-RPE deposits we observed small RPE defects. After the resolution of the lesions, a small patch of RPE and inner segment/outer segment layer atrophy developed (eFigure in the Supplement; Figure 2). In the 2 patients for whom there were follow-up examination results before any therapy we observed a spontaneous resolution of VHRLs and an appearance of new lesions elsewhere. In several cases, sub-RPE deposits colocalized with VHRLs and in some instances sub-RPE deposits appeared subsequent to the development of VHRLs at the same location (Figure 2 and 3). These sub-RPE deposits had medium reflectivity, were either solitary or confluent “drusen like,” and were clearly located above the Bruch membrane in all cases. Choroidal lesions (under the Bruch membrane) were not observed in any of the patients.

Discussion

There remains uncertainty about the pathophysiology of VRL. Our understanding of VRLs is increasing, but because of the rarity of the disease and the difficulty in obtaining substantial tissue specimens, most of our knowledge is derived from studies done on primary CNS lymphomas. To our knowledge it is still not completely clear where the abnormal lymphocytes originate and how and why they target the specific tissues where they finally manifest.7-9 It is still debated whether lymphoma cells infiltrate the retina from the retinal vessels, from the optic nerve, or from the choroid through the Bruch membrane and the RPE.10

Vertical hyperreflective lesions are located between the inner layers of the retina (GCL and RNFL) and the RPE. They are not stationary but seem to be constantly evolving and resolving; while they disappeared in one place, new ones appear elsewhere. We have also observed—as shown in Figure 2 and Figure 3—the appearance of sub-RPE infiltrates in the same locations where VHRLs were seen previously, although we could not associate every sub-RPE infiltrate with prior VHRLs. We also found that most of the VHRLs were around retinal vessels and some appeared to connect retinal vessels with sub-RPE deposits.

Our hypothesis is that VHRLs represent early microinfiltrates that are not visible on fundus examination, infrared imaging, and fundus autofluorescence. They may originate from second-order and third-order retinal vessels and capillaries. They infiltrate the retina and subsequently the sub-RPE space. As such, they follow the fluid flow that is seen in diseases with blood-retina barrier breakdown in which fluid originating from the retinal blood vessels penetrates through the retina and is pumped to the choroid by the RPE. However, because we could not identify VHRLs before the appearance of every sub-RPE deposits we cannot rule out that sub-RPE deposits might occur without VHRLs. A histologic examination of retinal tissue would be needed to confirm this hypothesis.

Vitreoretinal lymphoma has various findings on OCT. Saito et al6 described “focal round lesions in the neural retinal layer.” On their article’s figures, some of these lesions are larger, funduscopically visible retinal infiltrates while others seem to be similar microinfiltrates to the VHRL that we describe, although they do not describe the characteristic features of VHRL. In their series “focal round lesions” were present in 5 of 26 eyes. In their methods, they describe that horizontal and vertical OCTs were done at the macular level and through lesions identified on fundus photographs. We found that VHRLs are much more common (we found them in 7 of 12 eyes [58.3%]), are usually outside of the macular area, and are not visible on fundus photography, so the imaging done by Saito et al6 may have missed some lesions. In our cohort, the 2 patients in whom VHRLs were not found only had macular scans as baseline and during follow-up, so we may have missed the lesions as well in these cases.

In a recent publication, Barry et al11 describe OCT changes seen in primary VRL in a cohort of 22 patients. They note hyperreflective infiltrates in the inner layers of the retina in 18.8% of eyes, discrete nodules of hyperreflective foci in the subretinal space in 21.9%, and confluent bands of hyperreflective foci in the subretinal space in 31.3%, and suggest that the latter was highly suggestive of vitreoretinal lymphoma. In our series, we only saw 1 patient with a similar confluent band of hyperreflectivity, but we found VHRLs in 7 affected eyes (58.3%). We also observed that some of the VHRLs were oblique to the cut of the OCT scan and on dense scans lesions appearing in the inner retina were traceable to the RPE on subsequent B-scan slices. We suggest that some of the inner or outer retinal infiltrates seen by Barry et al11 may have been VHRLs that were only partially imaged. Furthermore, some of the figures in their article show completely imaged VHRLs, but they are unrecognized as such. While other diseases, such as multiple evanescent white dot syndrome, may have hyperreflective material extending toward the outer nuclear layer, in these diseases the material has not been found to extend to the GCL/RNFL.12

The high variability of changes seen on OCT in VRL has made OCT a less reliable surrogate than in other retinal diseases as similar morphologic findings can be seen in various types of posterior uveitis, age-related macular degeneration, or diabetic retinopathy. Optical coherence tomography features that are highly suggestive of VRLs, such as the one proposed by Barry et al,11 can help in the diagnostic process.

Limitations

A limitation of our study is its retrospective nature and the limited number of imaging studies that were performed. In 3 patients (42.9%), only OCT scans centered on the fovea were performed but no peripheral scans. Of these 6 eyes we found VHRLs in only 1 eye (16.7%), but we cannot rule out that VHRLs might have been seen in the other eyes if a detailed imaging had been performed. We found that fundus autofluorescence was often not gradable because of bad image quality in the setting of vitreous infiltration. On the images that were gradable, VHRLs alone were not detectable and only VHRLs on top of sub-RPE deposits were, but the still image quality might have influenced our assessment. Furthermore, none of the patients underwent OCT angiography examinations. Optical coherence tomography angiography may have provided further evidence of the connections between VHRLs and retinal vessels.

Conclusions

We believe that VHRL is a sign that may suggest the diagnosis of VRL and provide clues as to its pathogenesis. Performing OCT scans outside of the central macular area in patients in whom VRL is suspected might reveal the presence of VHRLs. Performing dense scan patterns helps in the detection as it enables the tracing of obliquely oriented VHRLs. If further confirmation is noted, the finding of VHRLs on OCT might justify adding VRL to the differential diagnosis in patients with otherwise undiagnosed retinal diseases.

Back to top
Article Information

Accepted for Publication: October 9, 2018.

Corresponding Author: Gábor Gy. Deák, MD, Department of Ophthalmology and Optometry, Medical University of Vienna, Währinger Gürtel 18-20, 1090, Vienna, Austria (gabor.deak@meduniwien.ac.at).

Published Online: November 29, 2018. doi:10.1001/jamaophthalmol.2018.5835

Author Contributions: Dr Deák had full access to the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.

Concept and design: Goldstein, Jampol.

Acquisition, analysis, or interpretation of data: All authors.

Drafting of the manuscript: Deak, Jampol.

Critical revision of the manuscript for important intellectual content: Goldstein, Zhou, Fawzi, Jampol.

Supervision: Goldstein, Fawzi, Jampol.

Conflict of Interest Disclosures: Dr Goldstein reports personal fees from AbbVie Pharmaceutical, Clearside Biomedical, and Santen Pharmaceutical; grant support from Clearside Biomedical; and membership on advisory panels at Allergan and Bausch & Lomb. No other disclosures are reported.

References
1.
Levasseur  SD, Wittenberg  LA, White  VA.  Vitreoretinal lymphoma: a 20-year review of incidence, clinical and cytologic features, treatment, and outcomes.  JAMA Ophthalmol. 2013;131(1):50-55. doi:10.1001/jamaophthalmol.2013.569PubMedGoogle ScholarCrossref
2.
Grimm  SA, Pulido  JS, Jahnke  K,  et al.  Primary intraocular lymphoma: an International Primary Central Nervous System Lymphoma Collaborative Group Report.  Ann Oncol. 2007;18(11):1851-1855. doi:10.1093/annonc/mdm340PubMedGoogle ScholarCrossref
3.
Fend  F, Süsskind  D, Deuter  C, Coupland  SE.  Maligne Lymphome des Auges.  Pathologe. 2017;38(6):515-520. doi:10.1007/s00292-017-0378-6PubMedGoogle ScholarCrossref
4.
Forooghian  F, Merkur  AB, White  VA, Shen  D, Chan  C-C.  High-definition optical coherence tomography features of primary vitreoretinal lymphoma.  Ophthalmic Surg Lasers Imaging. 2011;42 Online(0):e97-e99. doi:10.3928/15428877-20110922-02PubMedGoogle Scholar
5.
Vasconcelos-Santos  DV, De Puy E Souza  GH, de Faria  BB,  et al.  Subretinal pigment epithelial infiltrates in primary vitreoretinal lymphoma.  J Ophthalmic Inflamm Infect. 2011;1(4):171-171. doi:10.1007/s12348-011-0034-xPubMedGoogle ScholarCrossref
6.
Saito  T, Ohguro  N, Iwahashi  C, Hashida  N.  Optical coherence tomography manifestations of primary vitreoretinal lymphoma.  Graefes Arch Clin Exp Ophthalmol. 2016;254(12):2319-2326. doi:10.1007/s00417-016-3395-xPubMedGoogle ScholarCrossref
7.
Chan  C-C, Rubenstein  JL, Coupland  SE,  et al.  Primary vitreoretinal lymphoma: a report from an International Primary Central Nervous System Lymphoma Collaborative Group symposium.  Oncologist. 2011;16(11):1589-1599. doi:10.1634/theoncologist.2011-0210PubMedGoogle ScholarCrossref
8.
Coupland  SE, Chan  CC, Smith  J.  Pathophysiology of retinal lymphoma.  Ocul Immunol Inflamm. 2009;17(4):227-237. doi:10.1080/09273940903168696PubMedGoogle ScholarCrossref
9.
Fend  F, Ferreri  AJM, Coupland  SE.  How we diagnose and treat vitreoretinal lymphoma.  Br J Haematol. 2016;173(5):680-692. doi:10.1111/bjh.14025PubMedGoogle ScholarCrossref
10.
Davis  JL.  Intraocular lymphoma: a clinical perspective.  Eye (Lond). 2013;27(2):153-162. doi:10.1038/eye.2012.250PubMedGoogle ScholarCrossref
11.
Barry  RJ, Tasiopoulou  A, Murray  PI,  et al.  Characteristic optical coherence tomography findings in patients with primary vitreoretinal lymphoma: a novel aid to early diagnosis.  Br J Ophthalmol. 2018;102(10):1362-1366. doi:10.1136/bjophthalmol-2017-311612PubMedGoogle ScholarCrossref
12.
Marsiglia  M, Gallego-Pinazo  R, Cunha de Souza  E,  et al.  Expanded clinical spectrum of multiple evanescent white dot syndrome with multimodal imaging.  Retina. 2016;36(1):64-74. doi:10.1097/IAE.0000000000000685PubMedGoogle ScholarCrossref
×