Uveal melanoma is fatal in approximately half of patients because of the development of metastatic disease. Identification of high-risk patients who may potentially benefit from more rigorous surveillance or adjuvant therapy is currently under investigation. Clinical factors, such as tumor size, ciliary body involvement, and extraocular extension, and histocytologic factors, such as extravascular matrix patterns, epithelioid morphology, and nuclear grade, are not sufficiently accurate 1. That loss of a chromosome 3 in tumors is associated with the development of metastasis is now well established, and a variety of techniques are being used to test tumors for monosomy-3 and other high-risk chromosomal alterations 2,3. Gene array technology is also being used 4. The predictive accuracy of chromosome-based and gene-based assessments has consistently been shown to be superior to clinical-histocytologic systems 3,4. There are, however, limitations. Obtaining tumor samples can be problematic. Uveal melanoma is usually diagnosed clinically and not histocytologically, and most patients are treated with brachytherapy and not surgery. Fine-needle aspiration (FNA) biopsy at the time of radioactive plaque application is increasingly being used to obtain tumor samples for molecular prognostication, but does present risk, and may not be applicable to all patients (e.g. monocular status, small tumors, and near macula). Because of genetic heterogeneity, one random tumor sample obtained by FNA may not be representative 5,6.
Although progress has been made in the identification of blood biomarkers for monitoring disease progression in patients with uveal melanoma, prognostic blood biomarkers have not been identified. The expression of insulin-like growth factor-1 receptor (IGF1-R), implicated in the progression of liver metastatic tumors, and of human leukocyte antigen (HLA) class I molecules, implicated in regulating immune surveillance, by uveal melanoma tumors have also been associated with the development of metastatic disease 7–10. Like chromosomal and gene expression alterations, both are considered by The American Joint Committee on Cancer to be important prognostic factors in uveal melanoma, even though they are not included in the staging algorithms 11.
IGF1-R and HLA class I molecules do not appear in the circulation, but their components and modifiers do. Blood levels of IGF-1 and its serum-binding protein, IGFBP-3, have been reported to be elevated in patients with cancer, including in patients with uveal melanoma metastatic to the liver 12,13. Blood levels of IGF-1 were significantly higher in patients with scleral invasion when compared with patients without scleral invasion, supporting the possibility that blood IGF-1 levels may have prognostic utility 14. β-2microglobulin (β2M) is a component of the HLA class I molecule light chain. Like the HLA class I heavy chain, tumor β2M expression by immunohistochemistry has been associated with metastasis in uveal melanoma 15. Because it is noncovalently associated, β2M can circulate. Increases in blood levels of β2M have been observed in both hematologic and solid tumor malignancies, where they have been associated with tumor burden and a poor prognosis 16–19. Blood β2M as a prognosticator, however, has not been reported in patients with uveal melanoma. To explore their potential prognostic utility, we examined the relationship between blood levels of IGF-1, IGFBP-3, and β2M with tumor chromosome 3 status in patients with primary uveal melanoma.
Materials and methods
Patients with a clinical diagnosis of uveal melanoma presenting to the Cleveland Clinic Cole Eye Institute between 2009 and 2011 were enrolled. The study was approved by the Institutional Review Board, and all patients provided written, informed consent. At the time of diagnosis, each patient underwent a comprehensive ophthalmic examination with supporting diagnostic studies, including fundus photography and ultrasonography. Tumor size was classified using the Collaborative Ocular Melanoma Study criteria 1. Computed tomography scans of the chest, abdomen, and pelvis were performed to rule out metastatic disease. All patients had normal laboratory evaluations, including liver and renal function tests.
An FNA biopsy was performed in vivo in patients undergoing brachytherapy (plaque radiotherapy) at the time of plaque insertion. If the tumor was pre-equatorial, sampling was performed using a trans-scleral approach. After localization of the tumor by the transilluminator on the scleral bed and placement of tying sutures, a partial-thickness scleral incision (1 mm long) was made close to the center of the tumor. A 25-G needle connected to a 5 mm syringe was introduced to the maximum depth of 50% tumor height. Two passes were made and material from each pass was collected in a single CytoLyt Solution Centrifugation Tube (Cytyc, Marlborough, Massachusetts, USA). If the tumor was postequatorial, sampling was carried out using a transvitreal approach. Fresh tumor cells were obtained by FNA ex vivo from enucleated globes by aspirating with a 25-G needle connected to a 5 mm syringe. Again, two passes were made and material from each pass was collected in a single Cytolite tube.
FNA aspirate samples were entirely rinsed into 2–3 ml of normal saline, visually examined for adequacy, and then entirely transferred to a ThinPrep Processor (Cytyc), which yielded four alcohol-fixed Papanicolaou-stained slides per case and analyzed for cellular features including cell type (pure epithelioid, pure spindle, or mixed) and nuclear grade (1, 2, or 3) 1. Grade 3 nuclei had at least four-fold pleomorphism, hyperchromasia, coarse chromatin, and nucleoli conspicuous from a ×10 objective. Grade 1 nuclei had none of these features. All the other cases were grade 2.
Chromosome 3 status was assessed by fluorescence in-situ hybridization (FISH) using a directly labeled SpectrumOrange enumeration probe for the α centromeric locus of chromosome 3 using a previously described interphase FISH method for touch preparations 20. Probes were hybridized to Carnoy’s fixed fresh cells affixed to touch preparations from frozen tissue. A total of 200 interphase cells were scored using a Zeiss FISH workstation (Oberkochen, Germany) to determine the percentage of signals for each locus. A cut-point of 20% was used to determine monosomy-3 21.
IGF-1, IGFBP-3, and β2M levels
Blood was drawn before therapy into EDTA. Enzyme-linked immunosorbent assay kits were used to measure plasma levels of IGF-1 (normal range, 40–174 ng/ml), IGFBP-3 (normal range, 1430–3311 ng/ml) (R&D Systems, Minneapolis, Minnesota, USA), and β2M (normal range, 1100–2600 pg/ml; Calbiochem, San Diego, California, USA).
Data were summarized as frequency counts and percentages, and medians and ranges. For convenience, a recursive partitioning algorithm was used to identify the optimal cut points for β2M and the largest basal diameter. Monosomy-3 was evaluated in two ways: using the 20% cutoff as described above and by evaluating the absolute percentage of cells showing monosomy-3. For univariable analysis, Fisher’s exact test and the Cochran–Armitage trend test or the χ2-test were used to assess associations between the binary versions of monosomy-3 and β2M and categorical factors that had two or more than two levels, respectively. Similarly, associations with measured data were assessed using the Wilcoxon rank sum test (factors with two levels) and the Jonckheere–Terpstra test or the Kruskal–Wallis test (factors with more than two levels). Spearman rank correlations were used to summarize the overall relationships between the proportion of cells expressing monosomy-3 and absolute β2M levels and the largest basal diameters. A logistic regression model was used to simultaneously assess the association between monosomy-3 and β2M, tumor diameter, and nuclear grade. All tests of statistical significance were two sided, and all analyses were carried out using SAS version 9.2 (SAS Institute Inc., Cary, North Carolina, USA) and StatXAct version 9.0 (Cytel Inc., Cambridge, Massachusetts, USA).
A total of 76 consenting, eligible patients were studied. These included 37 men and 39 women, mean age 61 years (range 23–86). Forty-seven patients (62%) underwent brachytherapy and 29 patients (38%) underwent enucleation. Melanoma was confirmed cytologically in all patients. Clinically, most tumors were large (59%) and located in the choroid; most did not manifest extraocular extension (83%). On FNA cytology, most tumors were of mixed cell type (61%) and nuclear grade 2 (69%). No tumor manifested a pure epithelioid cytology. Overall, 33 patients (43%) were considered to have monosomy-3 on the basis of greater than 20% of cells showing monosomy-3 by FISH. The median proportion of cells showing monosomy-3 was 8.8% (range 0–78%).
Table 1 summarizes the associations between tumor monosomy-3 and clinical and cytologic factors. Related (or at least possibly related) to monosomy-3 were largest basal diameter (all P≤0.07) and nuclear grade (P=0.07 for the association between grade and monosomy-3). Table 2 summarizes the associations between tumor monosomy-3 and blood levels of IGF-1 and IGFBP-3, IGF-1 : IGFBP-3 ratio, and β2M. The vast majority of patients had blood levels of IGF-1 within normal ranges; only three patients (4%) had elevated levels. In contrast, the levels of IGFBP-3 in most patients (67%) were lower than the reported normal range. IGF-1, IGFBP-3, and IGF-1 : IGFBP-3 levels were not related to tumor monosomy-3. In contrast, blood β2M levels, which were above the normal range in 28%, were related to tumor monosomy-3 (all P≤0.02). The overall correlation and the distribution of blood β2M levels in patients with and without tumor monosomy-3 are shown in Fig. 1. When β2M levels of 3000 pg/ml were used as a cut-off point, as determined using a recursive partitioning algorithm, 33% (11 of 33) of the patients with monosomy-3 were considered to have elevated levels to 5% (two of 42) of the patients without monosomy-3 (P=0.002).
In addition to being correlated with monosomy-3, β2M was also correlated with tumor size (Fig. 2; P=0.01), location (Fig. 3a; P=0.03), and nuclear grade (Fig. 3b; P=0.03). In multivariable analysis, which considered β2M, tumor size, and nuclear grade, β2M (≥3000 vs. <3000 pg/ml, P=0.01) and largest basal diameter (≥16 vs. <16 mm, P=0.03), but not nuclear grade, were found to be independently associated with monosomy-3. Combining both factors, 30% (14 of 46) of patients with small/medium-sized tumors (≤16 mm) and ‘low’ β2M levels (<3000 pg/ml) had monosomy-3 compared with 100% (6/6) patients with large tumors (>16 mm) and ‘high’ β2M levels (≥3000 pg/ml); 57% (eight of 14) patients with low β2M levels but large tumors and 71% (five of seven) patients with small/medium tumors and high β2M levels had monosomy-3 (P=0.001, Fig. 2).
Uveal melanoma disseminates hematogenously, and blood biomarkers are a logical as well as a convenient assessment that may be useful not only for the identification of metastasis but also for risk stratification. Monitoring of blood levels of markers, such as osteopontin and S-100β, in conjunction with liver function tests may help to detect metastatic disease in patients with uveal melanoma before the metastases are detectable by imaging 22,23. Prognostic blood biomarkers that may help guide surveillance and adjuvant therapy recommendations have not been identified. We applied the ability of molecular techniques to more accurately predict the risk of metastasis in uveal melanoma patients to facilitate the discovery of prognostic biomarkers. We examined blood correlates of tumor factors that have been associated with a poor prognosis, IGFR-1, and HLA class I molecules. We found that tumor monosomy-3 was associated with elevated blood levels of the HLA class I component, β2M, in patients with primary uveal melanoma.
More than one group has reported that tumor expression of HLA class I antigen is associated with a poor prognosis and metastatic disease 9,10. Loss of heterozygosity at the HLA class I locus on 6p is frequent in uveal melanoma and is correlated with improved patient survival 24. Gene expression profiling has also found that several genes of the HLA complex are overexpressed in the tumors of patients with uveal melanoma who develop metastasis 25. This contrast with cutaneous melanoma, in which a low expression of HLA class I is associated with increased tumor thickness and a poor prognosis 26. How β2M may promote cancer progression is not established. Immune modulation has been implicated. Although necessary for T-cell recognition, β2M renders cells resistant to natural killer cells 27. There is also experimental evidence that β2M may directly promote tumor cell growth, epithelial mesenchymal transition, and metastasis 28. Elevated levels of β2M are used clinically to assess prognosis in multiple myeloma and other hematologic malignancies 16,17. There is evidence that β2M may have predictive value in cutaneous melanoma. Elevated serum β2M values were associated with relapse in patients with cutaneous melanoma administered adjuvant interferon-α-2b 29.
More than one group has also identified tumor IGF-R1 expression, which is implicated in the proliferation, invasion, and metastasis of uveal melanoma cells, as associated with a poor prognosis in uveal melanoma 7,8,30. IGF-I and IGFBP-3, which can mediate both tumor-promoting and tumor-inhibitory effects, have been implicated in the development and progression of many cancers. We did not find an association between blood levels of IGF-1 or IGFBP-3 and tumor monsomy-3. Although elevated levels of IGF-1 have been observed in patients with uveal melanoma 13, including in our study, lower blood levels of IGF-1 have characterized most cancers 31. IGF-1 levels in uveal melanoma patients were significantly lower compared with healthy individuals in one study. During the year before metastatic diagnosis, IGF-1 levels, however, did not change and were not predictive of metastasis 32. An inverse correlation between blood IGFBP-3 levels and disease progression has been reported in cutaneous melanoma 33.
Further prospective, clinical testing will be necessary to establish the measurement of blood β2M levels as a prognostic biomarker in this rare cancer, either alone or with other assessments, and is underway. β2M is not tumor specific. As it is associated with lymphocyte activation and also cleared by the kidney, increases are observed in patients with infections and renal insufficiency 34. Given the complexity of the metastatic process, a combination of several biomarkers may be necessary to achieve adequate sensitivity and accuracy. The time from diagnosis of the primary tumor to discovery of metastasis in patients with high-risk karyotypes or genotypes can range from months to years. It is possible that measurement of β2M and other blood biomarkers in conjunction with tumor molecular markers may help distinguish between slowly and rapidly metastatic tumors. Studies in which β2M levels are being assessed in conjunction with liver function tests and imaging studies as part of systemic surveillance for metastases in patients with defined tumor genotypes are also underway. Finally, measurement of blood levels of β2M and also IFG-1 and IGFBP-3 may also have implications for therapy, as methods of targeting tumor-associated β2M and IGF-1R are under investigation 35,36.
This work was supported in part by RO1CA136776 from the National Cancer Institute, National Institutes of Health, Bethesda, MD; a grant from the Falk Medical Research Trust, Chicago, IL; and an unrestricted grant from Research to Prevent Blindness, Department of Ophthalmology, Cleveland Clinic Lerner College of Medicine.
Conflicts of interest
There are no conflicts of interest.
1. Turell ME, Tubbs RR, Biscotti CV, Singh AD. Uveal melanoma: prognostication. Monogr Clin Cytol. 2012;21:55–60
2. Shields CL, Ganguly A, Materin MA, Teixeira L, Mashayekhi A, Swanson LA, et al. Chromosome 3 analysis of uveal melanoma using fine-needle aspiration biopsy at the time of plaque radiotherapy in 140 consecutive cases: the Deborah Iverson, MD, Lectureship. Arch Ophthalmol. 2007;125:1017–1024
3. Damato B, Coupland SE. Translating uveal melanoma cytogenetics into clinical care. Arch Ophthalmol. 2009;127:423–429
4. Onken MD, Worley LA, Tuscan MD, Harbour JW. An accurate, clinically feasible multi-gene expression assay for predicting metastasis in uveal melanoma. J Mol Diagn. 2010;12:461–468
5. Schoenfield L, Pettay J, Tubbs RR, Singh AD. Variation of monosomy-3 status within uveal melanoma. Arch Pathol Lab Med. 2009;133:1219–1222
6. Dopierala J, Damato BE, Lake SL, Taktak AF, Coupland SE. Genetic heterogeneity in uveal melanoma assessed by multiplex ligation-dependent probe amplification. Invest Ophthalmol Vis Sci. 2010;51:4898–4905
7. All-Ericsson C, Girnita L, Seregard S, Bartolazzi A, Jager MJ, Larsson O. Insulin-like growth factor
-1 receptor in uveal melanoma: a predictor for metastatic disease and a potential therapeutic target. Invest Ophthalmol Vis Sci. 2002;43:1–8
8. Mallikarjuna K, Pushparaj V, Biswas J, Krishnakumar S. Expression of insulin-like growth factor
receptor (IGF-1R), c-Fos, and c-Jun in uveal melanoma: an immunohistochemical study. Curr Eye Res. 2006;31:875–883
9. Blom DJ, Luyten GP, Mooy C, Kerkvliet S, Zwinderman AH, Jager MJ. Human leukocyte antigen class I expression. Marker of poor prognosis
in uveal melanoma. Invest Ophthalmol Vis Sci. 1997;38:1865–1872
10. Ericsson C, Seregard S, Bartolazzi A, Levitskaya E, Ferrone S, Kiessling R, Larsson O. Association of HLA class I and class II antigen expression and mortality in uveal melanoma. Invest Ophthalmol Vis Sci. 2001;42:2153–2156
11. Edge SB, Byrd DR, Compton CC, Fritz AG, Green FL, Trotti A. Malignant melanoma of the uvea. The AJCC cancer staging handbook. 20107th ed. New York Springer:611–622
12. Renehan AG, Zwahlen M, Minder C, O’Dwyer ST, Shalet SM, Egger M. Insulin-like growth factor
(IGF)-I, IGF binding protein-3, and cancer risk: systematic review and meta-regression analysis. Lancet. 2004;363:1346–1353
13. Egan KM, Giovanucci E, Gragoudas ES, Pollak M. Role of serum insulin-like growth factor
(IGF)-1 in the progression of choroidal melanoma [ARVO Abstract]. Invest Ophthalmol Vis Sci. 2000:41 Abstract 2017
14. Topcu-Yilmaz P, Kiratli H, Saglam A, Söylemezoglu F, Hascelik G. Correlation of clinicopathological parameters with HGF, c-Met, EGFR, and IGF-1R expression in uveal melanoma. Melanoma Res. 2010;20:126–132
15. Krishnakumar S, Abhyankar D, Lakshmi SA, Pushparaj V, Shanmugam MP, Biswas J. HLA expression in choroidal melanomas: correlation with clinicopathological features. Curr Eye Res. 2004;28:409–416
16. Bataille R, Durie BG, Grenier J. Serum beta2 microglobulin and survival duration in multiple myeloma: a simple reliable marker for staging. Br J Haematol. 1983;55:439–447
17. Federico M, Bellei M, Marcheselli L, Luminari S, Lopez-Guillermo A, Vitolo U, et al. Follicular lymphoma international prognostic index 2: a new prognostic index for follicular lymphoma developed by the international follicular lymphoma prognostic factor project. J Clin Oncol. 2009;27:4555–4562
18. Tsimberidou AM, Kantarjian HM, Wen S, O’Brien S, Cortes J, Wierda WG, et al. The prognostic significance of serum beta2 microglobulin levels in acute myeloid leukemia and prognostic scores predicting survival: analysis of 1180 patients. Clin Cancer Res. 2008;14:721–730
19. Sabbioni ME, Siegrist HP, Bacchi M, Bernhard J, Castiglione M, Thürlimann B, et al. Association between immunity and prognostic factors in early stage breast cancer patients before adjuvant treatment. Breast Cancer Res Treat. 2000;59:279–287
20. Frater JL, Tsiftsakis EK, Hsi ED, Pettay J, Tubbs RR. Use of novel t(11;14) and t(14;18) dual-fusion fluorescence in situ hybridization probes in the differential diagnosis of lymphomas of small lymphocytes. Diagn Mol Pathol. 2001;10:214–222
21. Turell ME, Tubbs RR, Saunthararajah Y, Biscotti CV, Triozzi PL, Singh AD. Improved detection of monosomy-3 in uveal melanoma using double-target FISH with centromeric and 3p25
probes [ARVO Abstract]. Invest Ophthalmol Vis Sci. 2011;52 E-Abstract 1435
22. Barak V, Kaiserman I, Frenkel S, Hendler K, Kalickman I, Pe’er J. The dynamics of serum tumor markers in predicting metastatic uveal melanoma (part 1). Anticancer Res. 2011;31:345–349
23. Hendler K, Pe’er J, Kaiserman I, Baruch R, Kalickman I, Barak V, Frenkel S. Trends in liver function tests: a comparison with serum tumor markers in metastatic uveal melanoma (part 2). Anticancer Res. 2011;31:351–357
24. Metzelaar-Blok JA, Jager MJ, Moghaddam PH, van der Slik AR, Giphart MJ. Frequent loss of heterozygosity on chromosome 6p in uveal melanoma. Hum Immunol. 1999;60:962–969
25. Singh AD, Sisley K, Xu Y, Li J, Faber P, Plummer SJ, et al. Reduced expression of autotaxin predicts survival in uveal melanoma. Br J Ophthalmol. 2007;91:1385–1392
26. Bröcker EB, Suter L, Brüggen J, Ruiter DJ, Macher E, Sorg C. Phenotypic dynamics of tumor progression in human malignant melanoma. Int J Cancer. 1985;36:29–35
27. Claësson MH, Nissen MH. Binding of human beta 2-microglobulin to murine EL4 thymoma cells upregulates MHC class I heavy-chain epitopes, inhibits IL-2 secretion and induces resistance to killing by natural killer cells. Immunol Lett. 1994;39:195–202
28. Josson S, Nomura T, Lin JT, Huang WC, Wu D, Zhau HE, et al. β2-microglobulin induces epithelial to mesenchymal transition and confers cancer lethality and bone metastasis in human cancer cells. Cancer Res. 2011;71:2600–2610
29. Hofmann MA, Kiecker F, Küchler I, Kors C, Trefzer U. Serum TNF-α, B2M and sIL-2R levels are biological correlates of outcome in adjuvant IFN-α2b treatment of patients with melanoma. J Cancer Res Clin Oncol. 2011;137:455–462
30. Girnita A, All-Ericsson C, Economou MA, Aström K, Axelson M, Seregard S, et al. The insulin-like growth factor
-I receptor inhibitor picropodophyllin causes tumor regression and attenuates mechanisms involved in invasion of uveal melanoma cells. Clin Cancer Res. 2006;12:1383–1391
31. Jacobs ET, Martínez ME, Alberts DS, Ashbeck EL, Gapstur SM, Lance P, Thompson PA. Plasma insulin-like growth factor
I is inversely associated with colorectal adenoma recurrence: a novel hypothesis. Cancer Epidemiol Biomarkers Prev. 2008;17:300–305
32. Barak V, Zloto O, Pe’er J, Kalickman I, Frenkel S. Insulin like growth factor-1 as a predicting biomarker for metastatic uveal melanoma in humans [ARVO Abstract]. Invest Ophthalmol Vis Sci. 2011;52 E-Abstract 1449
33. Panasiti V, Naspi A, Devirgiliis V, Curzio M, Roberti V, Curzio G, et al. Correlation between insulin-like growth factor
binding protein-3 serum level and melanoma progression. J Am Acad Dermatol. 2011;64:865–872
34. Matos AC, Durao MS Jr, Pacheco-Silva A. Serial beta-2-microglobulin measurement as an auxiliary method in the early diagnosis of cytomegalovirus infection in renal transplant patients. Transplant Proc. 2004;36:894–895
35. Yi Q. Novel immunotherapies. Cancer J. 2009;15:502–510
36. Karasic TB, Hei TK, Ivanov VN. Disruption of IGF-1R signaling increases TRAIL-induced apoptosis: a new potential therapy for the treatment of melanoma. Exp Cell Res. 2010;316:1994–2007
Keywords:© 2013 Lippincott Williams & Wilkins, Inc.
blood biomarkers; insulin-like growth factor; prognosis