A term first coined by Gass in 1990,1 bilateral diffuse uveal melanocytic proliferation (BDUMP) refers to a peculiar and rare paraneoplastic ocular disorder. This condition characteristically leads to melanocytic propagation within the choroid, along with retinal pigment epithelial (RPE) hypertrophy and degeneration.1,2 Often, rapid mortality is expected from underlying malignancy within 1 to 2 years.1 Herein, we explore with multimodal imaging the potentially unique features of our case, including bacillary layer detachment (BLD),3–5 gravitating retinal detachment,6 and multifocal choroidal hyperpermeability, to enhance our understanding of the pathophysiology of this paraneoplastic disorder.
A 77-year-old White man presented for tertiary retinal evaluation. His history was significant for neovascular age-related macular degeneration (AMD), right worse than left eye, treated with intravitreal anti-vascular endothelial growth factor for 3 years. Phacoemulsification cataract surgery with intraocular lens placement was performed in each eye 2 to 3 years previously. He had a history of hypertension, hypercholesterolemia, non-insulin-dependent diabetes, and a remote smoking history. Current medications included oral carvedilol, clopidogrel, aspirin, amlodipine, losartan, sitagliptin, simvastatin, and doxazosin with conventional dosing.
Snellen acuity was 20/25 in the right eye and 20/150 in the left. Anterior segment examinations revealed arcus senilis, with centered and clear posterior chamber intraocular lenses, as well as clear vitreous bilaterally. Dilated examination and ultra-widefield fundus photography (Optos, Dunfermline, United Kingdom) showed thickening of the left macula, most prominent centrally, and juxtapapillary outer retinal cystic changes and shallow retinal detachment of the right eye in addition to bilateral areas of yellow subretinal deposits and pigmentary alterations including dark choroidal plaques (Figure 1A). The left eye had more extensive intraretinal fluid and detachment involving the fovea, particularly in areas overlying these plaques. Ultra-widefield fundus autofluorescence (FAF) (Optos, Dunfermline, United Kingdom) showed a prominent alternating hyperauto-fluorescent and hypoautofluorescent pattern of the left greater than the right perimacula, as well as a retinal pigment epithelial tract extending inferiorly on the left, consistent with an antecedent dependent detachment (Figure 1B). Spectral-domain optical coherence tomography (SD-OCT) with Heidelberg HRA+OCT (Heidelberg Engineering, Heidelberg, Germany) revealed outer retinal deposits and cystic thickening bilaterally, as well as a fibrinous, BLD involving the left fovea (Figure 1C). There was prominent choroidal thickening and highly reflective placoid pigmented choroidal foci with shadowing of the outer choroid, but intact choriocapillaris, bilaterally. Ultra-widefield fluorescein angiography (Optos, Dunfermline, United Kingdom) showed alternating hyperfluorescence and hypofluorescence (Figure 2A) that reciprocated the pattern demonstrated by FAF, consistent with atrophy or window defects of the RPE and blockage from pigment clumping, respectively. Swept-source OCT angiography (SS-OCTA) (PLEX Elite 9000; Zeiss, Oberkochen, Germany) demonstrated a lack of any flow signal that would be consistent with retinal or choroidal neovascularization bilaterally (Figure 2B). Ultra-widefield indocyanine green angiography (ICGA) (Optos, Dunfermline, United Kingdom) revealed scattered areas of hyperfluorescence, punctate hypofluorescence, and late staining consistent with multifocal areas of choroidal hyperpermeability and pigmentary clumping (“spaghetti and parmesan”), bilaterally (Figure 2C). He then revealed a 5-year history of localized, persistent, high-grade urothelial cancer involving bladder and ureteral junction, status post multiple resections, but without any systemic chemotherapy or radiation. A diagnosis of bilateral diffuse uveal melanocytic proliferation with BLD was made.
Multimodal imaging. Multifocal areas of pigmentation and drusen-like deposits (arrow) are seen with ultra-widefield fundus photography, as well as areas of macular fluid including at least one nasally (arrowhead, A). Corresponding fundus autofluorescence shows alternating hyperautofluorescence and hypoautofluorescence extending around the macula, and in the left eye, a hyper-autofluorescent track extending inferiorly (arrowheads) consistent with previous fluid (B). Near-infrared reflectance shows hyperreflectance of subretinal drusen-like material visible on cross section through B-scan with spectral-domain optical coherence tomography (arrow), as well as intraretinal fluid nasally (arrowhead, compare with A) in the right eye, and extensive fluid and fibrin with foveal-involving bacillary layer detachment (*) in the left eye (C, right; D, left). Shadowing of the outer choroid from highly reflective placoid pigmented choroidal foci is observed (double-arrow) among a generally intact choriocapillaris.
Ultra-widefield recirculation-phase fluorescein angiography demonstrates alternating hyperfluorescence and hypofluorescence that reciprocates the fundus autofluorescence (A, compare with 1B), consistent with retinal pigment epithelial (RPE) degeneration and RPE hypertrophy, respectively. Ultra-widefield early phase indocyanine green angiography (ICGA) shows variability in choroidal vessel hyperfluorescence and small spots of hypofluorescence corresponding with hyperpigmentation (B). Late-phase ICGA shows multifocal choroidal hyperpermeability with over-lying punctuated spots of hypofluorescence also consistent with this hyper-pigmentation (C). Magnified view of hyperpermeable choroidal vessels from left eye in B (right panel) shows large choroidal vessel hyperfluorescence (arrows, “spaghetti”) and discontinuity that can be attributed to punctate pattern of hypofluorescence from overlying RPE hyperplasia and proliferating choroidal melanocytes (arrowheads, “parmesan,” D). Note there are no areas suspicious for neovascularization with the dye tests. Swept-source optical coherence tomography angiography shows the bacillary layer detachment with fluid and fibrin accumulation (*), but absence of flow signal in the retina suggestive of any obvious neovascularization, and overall intact choriocapillaris adjacent to the bottom pink segmentation line (E).
BDUMP is a rare form of a paraneoplastic disease expressed as a proliferative rather than primarily atrophic or degenerative phenomenon. The diagnosis of BDUMP is presumptive but quite definitive based on the typical clinical and imaging features of the disease: exudative retinal detachment, recent cataract progression, melanocytic infiltration and thickening of the choroid with pigmented choroidal plaques, and alternating autofluorescence that reciprocates fluorescence consistent with respective RPE degeneration and hyperplasia. Our case of BDUMP demonstrates several features that both expand the clinical spectrum of the disease, as well as overlap with other posterior segment conditions, particularly with its exudative findings.
Our case demonstrated the presence of a BLD. A BLD was originally thought to represent a histologic artifact.4 It has now been associated with several diseases in the macula. The first was described in toxoplasmosis, referring specifically to the splitting of the photoreceptors at the inner segment myoid as demonstrated using pseudo-correlation with the histologic artifact, but precise localization is still debated.3–5 It has a characteristic appearance of exudation and fibrin within the outer retina. To our knowledge, this presentation is the first illustration with description of a BLD with BDUMP, though it appears to have been illustrated without comment previously.2 It is possible that bladder cancer (and factors mediating its paraneoplastic process) may specifically render an increased likelihood of BLD in the setting of BDUMP, now that two cases have exhibited it when no others to our knowledge have before. Uroplakins are unique and essential proteins for urothelial tissues, including permeability barriers.7 It is conceivable, then, that the paraneoplastic response from sensitization to these constituents with bladder cancer may mediate this specific phenotype with BDUMP. One other case of bladder cancer with BDUMP has utilized OCT, but with insufficient scans to determine the presence of a BLD.8
It appears there are exceedingly few cases of bladder cancer with BDUMP, and are estimated to occur in only 3% of BDUMP presentations.9 Most of these other cases conferred survival greater than 4 years,8,10,11 longer than typically found with BDUMP.1 It is possible that a sensitization response to uroplakin proteins may occur earlier than proteins of other malignancies, contributing to earlier manifestations of BDUMP and therefore apparently longer survival. Although it may seem that primary bladder cancer could be detected and treated earlier than other malignancies, it is a cancer known to have high morbidity and mortality.12 Alternatively, these surviving patients with BDUMP and bladder cancer may have a more robust immune system. An additionally novel finding here is the absence of neovascularization beneath or within the BLD here as seen with SS-OCTA, supported by the indifference to 3 years of intravitreal anti-vascular endothelial growth factor treatment.
It has been hypothesized that the configuration of the BLD can be attributed to the swiftness of fluid and fibrin accumulation in exudative disease,5 particularly with acute conditions like acute idiopathic maculopathy (AIM).3 Despite the chronicity of the disease in our case with 3 years of anti-angiogenic therapy, and lack of any obvious neovascularization or inflammation, this patient still exhibited a BLD. This discrepancy in time-line suggests that the phenotype could converge from varying underlying pathophysiologic mechanisms. Although rapid inner choroidal ischemia and RPE dysfunction is thought to lead to BLD in inflammatory conditions like AIM,3 here the choroidal hyperpermeability and RPE degeneration likely results in a comparable exudation with BLD, even though the underlying causes and rapidity are inherently different. Therefore, BLD may not exclusively form secondary to rapid exudation. The cystic changes here could be equally exudative as they are degenerative, particularly in this setting of prolonged anti-vasogenic treatment. The BLD, may, however, depend on the volume of exudation given the lack of its presence among the relatively mild cystic changes of the contralateral retina. Reports of BLD in a variety of other posterior segment diseases including Vogt-Koyanagi-Harada syndrome, metastatic diseases, central serous chorioretinopathy (CSCR), and posterior scleritis highlight the variability of presentation.5 We believe that BDUMP should be added to the differential diagnosis of BLD. Although the chronicity of the disease differs, it is possible that the RPE degeneration allows for the particular accumulation of fluid and fibrin in the bacillary layer, just as the RPE dysfunction and ischemia may allow for the BLD in AIM, for example.3 We further hypothesize that the fibrin may play an integral part of the BLD formation, given previous histologic demonstration of choroidal invasion with proliferating epithelioid cells in BDUMP.1 It is thought that fibrin plays a role in outer retinal conditions ranging from blowout tears with avulsion from choroidal neovascularization to bullous variants of CSCR.13 Through “sticky” tangential forces imposed by the fibrinous change, it is possible that with enough exudation, it may be forced into the plane of the bacillary layer.
The patterns observed in this case by ICGA appeared consistent with choroidal hyperpermeability (reminiscent of the “pachychoroid phenotype”)13 and punctuated RPE hyperplasia, as well as proliferating choroidal melanocytes. Debate remains on whether proliferating choroidal melanocytes are entirely benign, somewhat malignant, or both, but recent findings and discussion suggest they can exhibit a spectrum of these histologic features.10 A previous report of BDUMP with bladder cancer proposes a “string of sausages” sign with ICGA, as hyperfluorescent choroidal vessels with alternating hypofluorescence from overlying RPE hypertrophy and lipofuscin accumulation.2 However, our case shows that the alternating fluorescence with ICGA also occurs outside the large choroidal vessels. It may more accurately show speckled hypofluorescence from both scattered RPE hyperplasia and infiltrating melanocytes1 overlying hyperfluorescent choroidal hyperpermeability, more consistent with an overall “spaghetti and parmesan” appearance. The ICGA findings are striking and may be helpful in differentiating BDUMP from other conditions. The late staining of the choroid with ICGA not only highlights the permeable, inflamed choroidal vessels, but may correspond with intrastromal fibrin deposition.
The abundance of fibrin in our case is particularly interesting. It has been established previously that fibrin is implicated with choroidal hyperpermeability.14 We believe that the late multifocal staining in the choroid with ICGA can be attributed to the molecular affinity of the ICG molecule upon conjugation with the fibrin. This explanation for the abundance of fibrin may then account for both the BLD described above and staining pattern with ICGA found in this case.
The gravitational tract inferiorly in the left eye could suggest an overlapping mechanism with chronic CSCR (known to frequently exhibit gravitational tracks),6,15 which can further complicate the overlap of presentations with this disease and delay diagnosis or recognition of malignancy. Gravitational tracks can also be nonspecific as they are observed in conditions like choroidal hemangioma in addition to choroidal neovascularization with AMD.16 To our knowledge, like the BLD, this feature has also not been described previously with BDUMP. It is possible that this gravitating track could have been related to fluid and fibrin within the BLD; further investigation is warranted to determine if the BLD can lead to these gravitating tracks. The presence of drusen (or similar-appearing deposits) implies concomitant exudative disease, and in this case, consideration of neovascular AMD, that can further obscure the correct diagnosis.
BDUMP is classically thought to lead to mortality within 1 to 2 years, although survival is reported as long as 9 years.1 Our case illustrates that despite 5 years of known cancer, and 3 years of eye disease presumably consistent with exudation in the setting of BDUMP, the diagnosis cannot be excluded upon duration alone. Moreover, an association of BDUMP with bladder cancer may confer increased survival relative to other malignancies, though reported numbers remain small. In addition to further characterization of imaging features, additional investigation is needed for augmenting the prognosis and treatment options for patients with BDUMP and their underlying malignancy.
- Gass JDM, Gieser RG, Wilkinson CP, Beahm DE, Pautler SE. Bilateral diffuse uveal melanocytic proliferation in patients with occult carcinoma. Arch Ophthalmol. 1990;108(4):527–533. doi:10.1001/archopht.1990.01070060075053 [CrossRef] PMID:2322154
- O'Day R, Michalova K, Campbell WG. Bilateral diffuse uveal melanocytic proliferation associated with bladder cancer: a novel imaging finding. Ophthalmic Surg Lasers Imaging Retina. 2019;50(8):525–528. doi:10.3928/23258160-20190806-10 [CrossRef] PMID:31415702
- Fernández-Avellaneda P, Breazzano MP, Fragiotta S, et al. Bacillary layer detachment overlying reduced choriocapillaris flow in acute idiopathic maculopathy. Retin Cases Brief Rep. 2019;Publish Ahead of Print. doi:10.1097/ICB.0000000000000943 [CrossRef] PMID:31764886
- Mehta N, Chong J, Tsui E, et al. Presumed foveal bacillary layer detachment in a patient with toxoplasmosis chorioretinitis and pachychoroid disease. Retin Cases Brief Rep. 2018;Publish Ahead of Print. doi:10.1097/ICB.0000000000000817 [CrossRef] PMID:30142112
- Cicinelli MV, Giuffré C, Marchese A, et al. The bacillary detachment in posterior segment ocular diseases. Ophthalmol Retina. 2020;4(4):454–456. doi:10.1016/j.oret.2019.12.003 [CrossRef] PMID:31937472
- Mrejen S, Balaratnasingam C, Kaden TR, et al. Long-term visual outcomes and causes of vision loss in chronic central serous chorioretinopathy. Ophthalmology. 2019;126(4):576–588. doi:10.1016/j.ophtha.2018.12.048 [CrossRef] PMID:30659849
- Babu Munipalli S, Yenugu S. Uroplakin expression in the male reproductive tract of rat. Gen Comp Endocrinol. 2019;281:153–163. doi:10.1016/j.ygcen.2019.06.003 [CrossRef] PMID:31181195
- Alrashidi S, Aziz AA, Krema H. Bilateral diffuse uveal melanocytic proliferation: a management dilemma. BMJ Case Rep. 2014;2014(may22 1):2014. doi:10.1136/bcr-2014-204387 [CrossRef] PMID:24855079
- Klemp K, Kiilgaard JF, Heegaard S, Nørgaard T, Andersen MK, Prause JU. Bilateral diffuse uveal melanocytic proliferation: case report and literature review. Acta Ophthalmol. 2017;95(5):439–445. doi:10.1111/aos.13481 [CrossRef] PMID:28636126
- Rahimy E, Coffee RE, McCannel TA. Bilateral diffuse uveal melanocytic proliferation as a precursor to multiple systemic malignancies. Semin Ophthalmol. 2015;30(3):206–209. doi:10.3109/08820538.2013.835835 [CrossRef] PMID:24124896
- De Salvo G, Prakash P, Rennie CA, Lotery AJ. Long-term survival in a case of bilateral diffuse uveal melanocytic proliferation. Eye (Lond). 2011;25(10):1385–1386. doi:10.1038/eye.2011.180 [CrossRef] PMID:21799522
- Martinez Rodriguez RH, Buisan Rueda O, Ibarz L. Bladder cancer: present and future. Med Clin (Barc). 2017;149(10):449–455. doi:10.1016/j.medcli.2017.06.009 [CrossRef] PMID:28736063
- Balaratnasingam C, Freund KB, Tan AM, et al. Bullous variant of central serous chorioretinopathy: expansion of phenotypic features using multimethod imaging. Ophthalmology. 2016;123(7):1541–1552. doi:10.1016/j.ophtha.2016.03.017 [CrossRef] PMID:27084564
- Guyer DR, Yannuzzi LA, Slakter JS, Sorenson JA, Ho A, Orlock D. Digital indocyanine green videoangiography of central serous chorioretinopathy. Arch Ophthalmol. 1994;112(8):1057–1062. doi:10.1001/archopht.1994.01090200063023 [CrossRef] PMID:8053819
- Yannuzzi LA, Shakin JL, Fisher YL, Altomonte MA. Peripheral retinal detachments and retinal pigment epithelial atrophic tracts secondary to central serous pigment epitheliopathy. Ophthalmology. 1984;91(12):1554–1572. doi:10.1016/S0161-6420(84)34117-3 [CrossRef] PMID:6084221
- Eandi CM, Sugin S, Spaide RF. Descending atrophic tracts associated with choroidal hemangioma. Retina. 2005;25(2):216–218. doi:10.1097/00006982-200502000-00018 [CrossRef] PMID:15689816