Professional Documents
Culture Documents
2008 Shields Faf Melanoma
2008 Shields Faf Melanoma
Ocular Oncology Service, Wills ABSTRACT when making a judgement on malignant poten-
Eye Institute, Thomas Jefferson Aim: To describe the autofluorescence features of tial.6 7 9
University, Philadelphia, PA, USA There is little information on the role of
choroidal melanoma.
Correspondence to: Design: Non-comparative case series. autofluorescence for intraocular tumours.12–18 In
Dr C L Shields, Ocular Oncology Participants: 51 consecutive patients. this report, we evaluate the autofluorescence
Service, Suite 1440, Wills Eye Methods: Standard fundus photography and autofluor- features of choroidal melanoma and related retinal
Institute, 840 Walnut Street,
Philadelphia, PA 19107, USA; escence photography (580 nm excitation, 695 nm barrier and retinal pigment epithelial (RPE) changes.
carol.shields@shieldsoncology. filter) were performed on all patients. Clinical features
com were correlated with autofluorescence features. METHODS
Main outcome measure: Autofluorescence features of The clinical records, fundus photographs and
CLS has had full access to all
the data in the study and takes choroidal melanoma and overlying retinal pigment optical coherence tomography (OCT) and fundus
responsibility for the integrity of epithelium (RPE). autofluorescence images of 51 consecutive patients
the data and the accuracy of the Results: The mean patient age was 59 years. The with choroidal melanoma were reviewed. The
data analysis. choroidal melanoma was a mean of 3.6 mm from the fundus autofluorescence was performed with
Accepted 11 January 2008 optic disc and 2.6 mm from the foveola. The mean standard filters2 (580 nm excitation, 695 nm bar-
tumour basal dimension was 11 mm and the mean rier filter) to avoid imaging the autofluorescence of
tumour thickness was 4 mm. The choroidal melanoma the lens, using a Zeiss camera (Carl Zeiss Meditec
showed intrinsic hypoautofluorescence (39%), isoauto- Inc, Jena, Germany) and Ophthalmic Imaging
fluorescence (6%) and hyperautofluorescence (55%). Systems (OIS) (Sacramento, California, USA) soft-
Slightly increased hyperautofluorescence of the mela- ware. The autofluorescence features of choroidal
noma was found in pigmented tumours (versus non- melanoma were evaluated on the basis of the
pigmented), those with greater thickness and basal appearance of the choroidal tumour relative to
dimensions, and those with overlying disrupted RPE. surrounding normal choroid. Related autofluores-
Related RPE hyperplasia and atrophy showed hypoauto- cence features of the retina and RPE were also
fluorescence, drusen, RPE detachment and subretinal fluid studied. The features were graded according to
showed slight hyperautofluorescence, and orange pig- autofluorescence (hypoautofluorescent, isoauto-
ment displayed the brightest hyperautofluorescence. fluorescent or hyperautofluorescent relative to
Conclusions: Choroidal melanoma generally shows slight the surrounding normal choroid/RPE), degree of
intrinsic hyperautofluorescence and the brightness autofluorescence (trace (1+), moderate (2+), or
increases with pigmented tumours, larger tumours, and marked (3+)) and granular pattern of autofluores-
those associated with disrupted RPE. Overlying orange cence (non-granular, fine granular, coarse granu-
pigment shows remarkably bright hyperautofluorescence. lar). These features were then correlated with the
clinical features.
Table 1 Clinical features of 51 consecutive eyes with Table 3 General summary of autofluorescence of choroidal melanoma
choroidal melanoma in 51 consecutive eyes relative to tumour pigmentation, location and size
Clinical feature Autofluorescence
Table 4 Granularity of autofluorescence of choroidal melanoma in 51 Table 6 General summary of autofluorescence of choroidal melanoma
consecutive eyes relative to tumour pigmentation in 51 consecutive eyes relative to retinal and retinal pigment epithelial
Autofluorescence granularity (RPE) features
Granular Granular Autofluorescence
Tumour pigmentation Non-granular fine coarse Hypo Iso Hyper
Choroidal melanoma overall (n = 51) 4 (8) 7 (14) 40 (78) RPE hyperplasia (n = 6) +
Choroidal melanoma pigmented 2 (5) 3 (8) 32 (86) RPE detachment (n = 1) ++
(n = 37) RPE fibrous metaplasia (n = 3) +
Choroidal melanoma non-pigmented 2 (14) 4 (29) 8 (57) RPE atrophy (n = 8) ++
(n = 14)
Orange pigment (n = 39) ++
Values are number (%). Drusen (n = 9) +
Subretinal fluid (n = 46) +
Cystoid macular oedema (n = 12) +
3+, marked; 2+, moderate; 1+, trace.
Table 5 Autofluorescence of choroidal melanoma in 51 consecutive eyes relative to retinal and retinal pigment epithelial (RPE) features
Autofluorescence
Hypo Hyper
Feature
Clinical feature 3+ 2+ 1+ Iso 1+ 2+ 3+ not present
Table 7 Autofluorescence of choroidal melanoma in 51 consecutive eyes relative to status of overlying retinal pigment epithelium (RPE)
Autofluorescence
Hypo Hyper
Clinical feature 3+ 2+ 1+ Iso 1+ 2+ 3+
Melanoma overall
(n = 51)
Overlying RPE 1 (11) 1 (11) 3 (33) 1 (11) 1 (11) 1 (11) 1 (11)
intact (n = 9)
Overlying RPE not 0 (0) 6 (14) 9 (21) 2 (5) 13 (31) 8 (19) 4 (10)
intact* (n = 42)
Pigmented melanoma
(n = 37)
Overlying RPE 1 (25) 0 (0) 0 (0) 1 (25) 0 (0) 1 (25) 1 (25)
intact (n = 4)
Overlying RPE not 0 (0) 5 (15) 6 (18) 2 (6) 10 (30) 6 (18) 4 (12)
intact* (n = 33)
Non-pigmented
melanoma (n = 14)
Overlying RPE 0 (0) 1 (20) 3 (60) 0 (0) 1 (20) 0 (0) 0 (0)
intact (n = 5)
Overlying RPE not 0 (0) 1 (11) 3 (33) 0 (0) 3 (33) 2 (22) 0 (0)
intact* (n = 9)
Values are number (%).
*RPE not intact includes RPE hyperplasia, detachment, fibrous metaplasia, atrophy and orange pigment.
3+, marked; 2+, moderate; 1+, trace.
Table 8 General summary of autofluorescence of choroidal melanoma In the 46 patients with active subretinal fluid, the fluid rim
in 51 consecutive eyes relative to status of overlying retinal pigment appeared with slightly more hyperautofluorescence than the
epithelium (RPE) fluid centre. In the five patients with a break in the Bruch’s
Autofluorescence membrane, the edge of the rupture showed slightly brighter
Clinical feature Hypo Iso Hyper
autofluorescence than the centre.
tool might be useful for evaluation and differentiation of uveal features. Indeed, melanoma had relatively mild autofluorescent
melanoma and nevus. Very little was published on this subject properties with the current standard technique of 580 nm
until 2007 when Lavinsky and associates13 commented on five excitation and 695 nm barrier filter on a digital fundus camera.
eyes with choroidal melanoma in vivo, in which the autofluor- We found that melanomas displayed hypoautofluorescence
escence findings were mostly related to the RPE alterations, and (39%), isoautofluorescence, (6%) and hyperautofluorescence
little information was provided on the tumour itself. Gunduz and (55%), and the hyperautofluorescence increased with larger
associates14 recognised a direct correlation of orange pigment and tumours, pigmented tumours and those with disrupted overlying
hyperpigmentation with increased autofluorescence in 23 patients RPE. The autofluorescence property of melanoma was of a
with choroidal nevus or melanoma. Bakri and coworkers15 showed granular quality in 92% of cases and non-granular in 8%. Fundus
autofluorescence emitted by lipofuscin in two eyes with choroidal autofluorescence of choroidal melanoma was studied by Lavinsky
melanoma. Shields et al17 commented on the brilliant hyperauto- and coworkers13 using scanning laser ophthalmoscopy with
fluorescence of orange pigment (lipofuscin) overlying small slightly different wavelength parameters (488 nm excitation and
choroidal melanoma, even when the lipofuscin was relatively 500 nm barrier filter), but there were no observable findings
inconspicuous on colour photography. within the tumour. Our results show perhaps more intrinsic
In this analysis, we evaluated the autofluorescence of both autofluorescence of choroidal lesions than those of Lavinsky et al,
the melanoma and the related retinal/RPE alterations. as the excitation wavelength in our model was longer. Farina and
Autofluorescence features were correlated with the clinical coworkers8 evaluated patients with pigmented cutaneous lesions
using selected wavelengths between 420 and 1040 nm and found conduct of this study, in the collection, analysis or interpretation of the data, or in the
that the ideal wavelength was 940 nm for discriminating preparation, review or approval of the manuscript.
melanoma from nevus and 578 nm for discerning tumour Competing interests: None declared.
margins. Perhaps future studies on the role of fundus autofluor-
escence in investigating choroidal melanocytic lesions should REFERENCES
evaluate near-infrared wavelengths (780–820 nm), which pene- 1. Lohmann W, Paul E. Native fluorescence of unstained cryosection of the skin with
melanomas and nevi. Naturwissenschaften 1988;76:424–6.
trate deeper into the choroid than shorter wavelengths and can 2. Lohmann W, Paul E. In situ detection of melanoma by fluorescence measurements.
excite melanin-associated fluorophores.23 25 Naturewissenschaften 1988;75:201–2.
Our group recently evaluated 64 eyes with choroidal nevus 3. Lohmann W, Nilles M, Boedeker RH. In situ differentiation between nevi and
using fundus autofluorescence.18 Overall, choroidal nevi showed malignant melanomas by fluorescence measurements. Naturwissenschaften
1991;78:456–7.
darker autofluorescence than melanoma, as nevi were hypoau- 4. Sterenborg HJCM, Motamedi M, Wagner RF, et al. In vivo fluorescence
tofluorescent in 56%, isoautofluorescent in 19% and hyperauto- spectroscopy and imaging of human skin tumours. Lasers Med Sci 1994;9:191–201.
fluorescent in 25% of cases. Similar to melanoma, choroidal 5. Sterenborg HCM, Thomsen S, Jacques SL, et al. In vivo autofluorescence of an
unpigmented melanoma in mice: correlation of spectroscopic properties to
nevus showed slightly brighter autofluorescence with increasing microscopic nature. Melanoma Res 1995;5:211–16.
size, increasing intrinsic pigmentation and disruption of over- 6. Tomatis S, Bartoli C, Bono A, et al. Spectrophotometric imaging of cutaneous
lying RPE. Despite these observations, the information obtained pigmented lesions: discriminant analysis, optical properties and histological
with standard autofluorescence (580 nm excitation and 695 nm characteristics. J Photochem Photobiol B 1998;42:32–9.
7. Bono A, Tomatis S, Bartoli C, et al. The ABCD system of melanoma detection: a
barrier filter) was subtle. Similar to melanoma, the brightest spectrophotometric analysis of the Asymmetry, Border, Color, and Dimension. Cancer
hyperautofluorescence was found with overlying orange pig- 1999 1;85:72–7.
ment, RPE detachment, subretinal fluid and drusen, in decreas- 8. Farina B, Bartoli C, Bono A, et al. Multispectral imaging approach in the diagnosis of
cutaneous melanoma: potentiality and limits. Phys Med Biol 2000;45:1243–54.
ing order. With current settings, it is doubtful whether intrinsic 9. Chwirot BW, Chwirot S, Sypniewska N, et al. Fluorescence in situ detection of
tumour autofluorescence can differentiate choroidal nevus from human cutaneous melanoma: study of diagnostic parameters of the method. J Invest
melanoma, but information from the overlying RPE could be Dermatol 2001;117:1449–51.
10. Marchesini R, Bono A, Bartoli C, et al. Optical imaging and automated melanoma
important in the diagnostic differentiation, as bright hyper-
detection: questions and answers. Melanoma Res 2002;12:279–86.
autofluorescence of orange pigment and subretinal fluid is more 11. Huang Z, Zeng H, Hamzavi I, et al. Cutaneous melanin exhibiting fluorescence
often found with melanoma. In contrast, areas of dark RPE emission under near-infrared light excitation. J Biomed Opt 2006;11:34010.
hypoautofluorescence generally signify atrophy or hyperplasia, 12. Lohmann W, Wiegand W, Stolwijk TR, et al. Endogenous fluorescence of ocular
malignant melanomas. Ophthalmologica 1995;209:7–10.
features often associated with nevi. 13. Lavinsky D, Belfort RN, Navajas EV, et al. Fundus autofluorescence of choroidal
The origin of autofluorescence in skin melanoma is not nevus and melanoma. Br J Ophthalmol 2007;91:1299–302.
completely understood. It is speculated that it results from 14. Gunduz K, Pulido J, Bakri SJ, et al. Fundus autofluorescence in choroidal melanocytic
lesions. Retina 2007;27:681–7.
endogenous fluorophores, including collagen, elastin, desmosine 15. Bakri K, Pulido J, Bakri SJ. Autofluorescence of choroidal melanomas. Retinal Cases
and keratin.1–3 5 Within the eye, several tissues are known to be & Brief Reports 2007;1:117–19.
autofluorescent, including the cornea, lens, RPE, sclera and 16. Shields CL, Pirondini C, Bianciotto C, et al. Autofluorescence of congenital
melanocytes within the choroid.5 hypertrophy of the retinal pigment epithelium. Retina 2007;27:1097–100.
17. Shields CL, Bianciotto C, Pirondini C, et al. Autofluorescence of orange pigment
In summary, the current standard autofluorescent technique overlying small choroidal melanoma. Retina 2007;27:1107–11.
(580 nm excitation, 695 nm barrier filter) provides subtle 18. Shields CL, Pirondini C, Bianciotto C, et al. Autofluorescence of choroidal nevus in 64
information on intrinsic autofluorescence of choroidal mela- cases. Retina 2008;in press.
19. Dorey C, Wu G, Ebenstein D, et al. Cell loss in the ageing retina: relationship to
noma and more obvious information on the related RPE/retinal lipofuscin accumulation and macular degeneration. Invest Ophthalmol Vis Sci
alterations. Future applications might focus on improved 1989;30:1691–9.
imaging of the melanoma itself by using a multispectral 20. Delori FC, Fleckner MR, Goger DG, et al. Autofluorescence distribution associated
technique to identify optimal wavelengths, particularly in the with drusen in age-related macular degeneration. Invest Ophthalmol Vis Sci
2000;41:496–504.
near-infrared range, for visualisation of the tumour and each of 21. Spaide RF. Fundus autofluorescence and age-related macular degeneration.
its related findings. As further information is uncovered, the Ophthalmology 2003;110:392–9.
potential benefits of autofluorescence photography in the early 22. Lois N, Owens SL, Coco R, et al. Fundus autofluorescence in patients with age-
related macular degeneration and high risk of visual loss. Am J Ophthalmol
detection of melanoma will be realised. 2002;133:341–9.
23. Cideciyan AV, Swider M, Aleman TS, et al. ABCA4-associated retinal degenerations
Funding: Support provided by the Retina Research Foundation of the Retina Society in spare structure and function of the human parapapillary retina. Invest Ophthalmol Vis
Cape Town, South Africa to CLS), the Paul Kayser International Award of Merit in Sci 2005;46:4739–46.
Retina Research, Houston, TX (to JAS), a donation from Michael, Bruce, and Ellen 24. Smith RT, Chan JK, Busuoic M, et al. Autofluorescence characteristics of early,
Ratner, New York, NY (to JAS, CLS), Mellon Charitable Giving from the Martha W atrophic, and high-risk fellow eyes in age-related macular degeneration. Invest
Rogers Charitable Trust, Philadelphia, PA (to CLS), the LuEsther Mertz Retina Research Ophthalmol Vis Sci 2006;47:5495–504.
Foundation, New York, NY (to CLS), and the Eye Tumour Research Foundation, 25. Elsner AE, Burns SA, Weiter JJ, et al. Infrared imaging of sub-retinal structures in
Philadelphia, PA (to CLS, JAS). The sponsors did not participate in the design or the human ocular fundus. Vision Res 1995;36:191–205.