Acta Ophthalmologica 2015
Immunomagnetic detection of micrometastatic cells in bone marrow of uveal melanoma patients: a paradox Nils Eide,1 Ragnar S. Faye,2,3 Hanne K. Høifødt,3 Berit Sandstad,4 Geir Qvale,1 Rowan Faber,1 Peter Jebsen,5 Gunnar Kvalheim6 and Øystein Fodstad3 1
Eye Department, University Hospital HF, Oslo, Norway Department of Dermatology, Oslo University Hospital HF, Oslo, Norway 3 Department of Tumor Biology, Oslo University Hospital HF and University of Oslo, Oslo, Norway 4 Department of Statistics, Oslo University Hospital HF, Oslo, Norway 5 Division of Pathology, Oslo University Hospital HF, Oslo, Norway 6 Department of Oncology, Oslo University Hospital HF, Oslo, Norway 2
ABSTRACT. Purpose: Our objective was to study survival rates with the bone marrow (BM) results in a cohort of uveal melanoma patients with long follow-up. Methods: Mononuclear cell fractions isolated from BM were examined for tumour cells using our immunomagnetic separation (IMS) method. The patients were classified as BM positive or BM negative. Clinical follow-up, histopathological findings, vital status and cause of death were registered. Results: The study included 328 consecutive patients with uveal melanoma from 1997 to 2006. Tumour cells were found in BM samples in 29% (95% CI, 25–34) at enrolment (96 cases). After a minimum follow-up time of 6 years, 156 (48%) (95% CI, 42–53) melanoma patients had died. The causes were as follows: melanoma metastases 92 (59%), another cancer 20 (13%) and non-cancer 44 (28%). Nine patients were still living with melanoma metastases. Until the latest work-up, 101 (31%) (95% CI, 26–36) patients had developed melanoma metastases. Cyto- or histopathological verification of the metastatic lesions was obtained in 85 cases (84%). In the group with melanoma metastases, 28 tested BM positive at study entry (28%) (95% CI, 19–38). In total, 39 of 101 with metastases tested positive at least once after a maximum of three tests (39%) (95% CI, 29–49). The overall median survival from the first BM test was shorter for the BM negative patients (9.5 years) compared with the BM positive (14.4 years), p = 0.02, log rank test. Conclusion: Ocular melanoma cells detected in BM seem to have a positive prognostic impact on survival in contrast to our original hypothesis. Key words: immunomagnetic detection – micrometastasis in bone marrow – outcome – survival – uveal melanoma
Acta Ophthalmol. 2015: 93: 59–66 ª 2015 Acta Ophthalmologica Scandinavica Foundation. Published by John Wiley & Sons Ltd
doi: 10.1111/aos.12462
Introduction Uveal melanoma is the most common primary intraocular malignant tumour
in adults (Jensen 1963; Raivio 1977), the highest incidence being found in
the Nordic countries (Bergman et al. 2002) with a decreasing north-to-south incidence in Europe (Virgili et al. 2007). At presentation, only 1–4% of patients have disseminated disease (Wagoner & Albert 1982; Pach et al. 1986). The dissemination of uveal melanoma is purely haematogenous, unless the tumour cells perforate the sclera (Dithmar et al. 2000).The liver is the most frequently affected organ in metastatic disease (Jensen 1982; Eskelin et al. 1999). Positron emission tomography–computed tomography (PET/CT) studies reveal a much higher frequency of bone metastases at diagnosis than previously believed and bone is the second most frequent location (Finger et al. 2005). Despite primary enucleation, more than 50% of patients with uveal melanoma develop metastases (Jensen 1963, 1982; Raivio 1977; Kujala et al. 2003). Adjusted for known prognostic parameters, the probability of dissemination seems to be independent of the treatment modality such as enucleation, plaque radiotherapy, proton beam radiotherapy or trans-scleral resection as long as local tumour control is achieved (Seddon et al. 1990; Diener-West et al. 2001; Egger et al. 2001). Cytogenetic studies have disclosed a marked reduction in survival of melanoma patients with tumours carrying certain genetic changes. In
59
Acta Ophthalmologica 2015
particular, the loss of heterozygosity for chromosome 3, gain in 8q or loss in 8p or specific changes in gene expression profiling appear to have a high predictive accuracy (Sisley et al. 1990; Damato et al. 2007; Harbour 2009; Singh et al. 2009; Ewens et al. 2013). Most uveal melanoma metastases occur within the first 5 years after diagnosis. However, the increased risk of death due to this disease is always present, and the cumulative probability of suffering a melanoma-related death increases from 31% after 5 years to 45% after 15 years (Kujala et al. 2003). Tumour dormancy may last up to 20–30 years, but latencies of more than 40 years have been reported (Coupland et al. 1996). Today, there is no effective treatment for uveal melanoma with metastases (Woll et al. 1999). Regardless of treatment, death usually follows within 5–8 months after diagnosis of the metastatic lesion (Gragoudas et al. 1991; Eskelin et al. 2003). A substantial number of patients have subclinical metastases at presentation (Eide et al. 2009). This is supported by the high frequency of circulating melanoma cells detected at diagnosis and at different intervals after primary treatment of patients with primary uveal melanoma (Callejo et al. 2007). Micrometastases may be established years before the tumour is diagnosed and treated. This assumption is supported by sequential observations of metastatic growth and mathematical calculations based on tumour doubling time (Eskelin et al. 2000) and is commented on in a letter discussing these data (Singh 2001). This study has been conducted to detect micrometastases and to explore their possible clinical impact.
approved the study. All patients signed a written consent form before enrolment, which clearly stated that they would get no information about the test results. Further details about the material have been reported previously (Eide et al. 2009). The melanoma cells were harvested from bone marrow (BM) by aspiration from the crista iliaca. The patients were classified as either BM positive or BM negative by finding a rosette: a melanoma cell with five or more beads on its surface (Fig. 1). Two such rosettes are required for a positive sample. The processing of BM and immunomagnetic detection techniques, including sensitivity and validity, have been outlined in previous papers for cutaneous melanomas (Faye et al. 2004), uveal melanomas (Eide et al. 2009), as well as colon cancer, breast cancer and osteosarcoma (Flatmark et al. 2002; Wiedswang et al. 2003; Bruland et al. 2005). During the first 2 years, the patients were evaluated clinically every 3rd or 4th month. Thereafter, medical evaluations, including liver ultrasound and eye examination, were performed at least once annually for 10 years, predominantly at the study centre. Otherwise, local eye departments or general practitioners checked the patient’s eye and general health condition annually. If no follow-up information was automatically received, it was requested by telephone. Computed tomography (CT), magnetic resonance imaging (MRI) and/or PET/CT scan were used to confirm metastases and for staging. Liver function tests were not performed routinely. We always strive for a diagnosis based on pathological tissue, both at initial diagnosis and at dissemination. In co-operation with an oncologist, the indication for not
Material and Methods This prospective study was compiled from 1997 to 2006 at an ophthalmological referral centre in Norway. In total, 328 patients (162 males and 166 females) were included. The mean/median age at inclusion was 63.2/66.0 years (range 8– 94). All patients with a clinical diagnosis of uveal melanoma, independent of location and age, were eligible and asked to participate. Fifteen patients (5%) refused to be included. The Regional Committee for Medical Research Ethics, Southern Norway,
60
harvesting tissue for pathological verification of the metastases was considered. Based on information from the most recent follow-up, we categorized cause of death as due to metastatic melanoma, non-melanoma cancer or a non-cancer related cause. Clinical and histopathological materials available at the time of death were routinely retrieved from clinical departments and general practitioners, and our database was twice compared with the database of the Cancer Registry of Norway. The latest TNM system (7th Edition) staging system is used in this report, (Fig. 3) (Edge et al. 2010). Survival was measured from the time of entry into the study until death or the date of latest observation. Patients still alive without melanoma metastases at the time of statistical work-up were treated as censored. Survival was estimated using the Kaplan–Meier method, and groups were compared with the log rank test. kappa was used for measure of agreement. p-Values 500), (Table 6). The numbers of metastases in the separate groups of 127 samples are listed to illustrate the importance of selecting cut-off points. The numbers of revealed clinical melanoma metastases in the different groups defined by the numbers of rosettes are given. In the material, 23 samples were classified as uncertain, either with only one definite rosette or with a suspect rosette in addition. Both these findings were classified as negative. Verification of finding melanoma cells
Some BM samples were tested with 9.2.27 and three different antibodies in parallel. Comparison of the results obtained with different antibodies binding different epitopes showed good agreement, with kappa test values between 0.69 and 0.76 (Table 7).
Discussion This study is focused on detecting micrometastasis at diagnosis. Our first hypothesis was that finding micrometastatic melanoma cells in BM should be an indicator of prognosis and our second assumption was that the number of BM+ patients should increase with further testing over time, especially those with dissemination. Our results did not confirm any of these hypotheses, although the tumour cells detected and isolated in BM were confirmed to be melanoma cells, as demonstrated by the binding of different fluorescent microspheres coated with two different melanoma antibodies to the surface protein of the cells captured with 9.2.27 immunobeads (Eide et al. 2009). The results were consistent when more antimelanoma antibodies were tested as immunobeads in parallel tubes. The kappa test gives values of 0.69/0.72/0.79 (Table 7). Our hypothesis of an association between the BM results at inclusion and survival seems to be confirmed. The results were, however, opposite to our assumptions (Fig. 2). The BM results could not be used clinically as an indicator of prognosis because the estimate is not accurate enough to predict the outcome for an individual
63
Acta Ophthalmologica 2015
patient. A metastatic index has been elaborated to identify those patients at greatest risk for metastatic death (Damato & Coupland 2008). The staging in the new TNM system is designed to classify patients according to likely survival (Kivel€a & Kujala 2013), and our material shows clear survival differences according to classification as documented by others (Kujala et al. 2013). The detection level of micrometastatic melanoma cells at inclusion was higher than with any other clinical method with 29.3% testing positive. In comparison, PET/CT studies have found the highest frequency of metastases at diagnosis, but the number is lower than 10% (Finger et al. 2005). The number of BM+ samples among the 101 with melanoma metastases increased from 28% to 40% after retesting. The crude survival of our material is similar to reports from other melanoma studies. Approximately 50% of patients with uveal melanoma will die of disseminated disease within 10 years after the diagnosis (Gamel et al. 1993; Singh & Topham 2003). A Finnish group (Kujala et al. 2003) confirms a continuous, although decreasing annual risk for dissemination. In our material, the all-cause mortality was 27% (90/328) (95% CI, 23–33) at 5 years of follow-up and 48% (156/ 328) (95% CI, 42–53) at 10 years. Dissemination without liver affection gives a much better survival, 9.6 years versus 4.6 in our material. That result confirms the findings of an older study (Kath et al. 1993). If our tumours are classified according to the criteria in the COMS study, the small tumours have a probability of 10-year survival of more than 80% in contrast to the medium with 54% and the larger 20%. The importance of tumour localization inside the eye on survival is confirmed with a highly significant effect, p < 0.001. Different cut-off values had a marked influence on the risk for developing metastasis (Table 6). Epithelioid tumour cells, high microvascular density, low pigmentation and type and number of infiltrating macrophages are important parameters that predict outcome, as well as the number of liver metastases and ciliary body involvement (Toivonen et al. 2004; Kodjikian et al. 2005). In our study, cell type or pigmentation in the tumour were not significant predictors.
64
Inside a metastatic lesion from the same primary ET, differences in gene expression profiles have been shown. Some of the metastatic melanoma cells were highly and others poorly invasive (Seftor et al. 2002; Singh et al. 2009). Relatively, few tumour cells were found in the BM samples (Eide et al. 2009). The clinical importance is probably determined by the properties of the malignant cells more than by their numbers. Our results strongly indicate that this assumption is correct. Thus, we found the highest risk in the two groups with few rosettes. In the group with the lowest number of rosettes still defined as BM positive, 44% (95% CI, 27–62) developed metastases, the highest frequency. The frequency of metastasis was the second highest in the group, which was classified as negative 39% (95% CI, 20–61), but there were inconclusive findings according to our definition. The lowest frequency was found in the genuinely negative group. The supposed cascade of the metastatic process may have relevance for some of our data (Hanahan & Weinberg 2000). Some of the detected tumour cells in this study may not have the capacity to establish metastases outside the BM or they may be apoptotic. Therefore, the mere presence of disseminated tumour cells in BM is not sufficient for prognostic or therapeutic purposes in uveal melanoma. Moreover, we cannot rule out the possibility that the metastatic cells in some patients have lost their expression of the epitope still expressed in the ET samples. The fact that the frequency of BM positive patients was higher in small choroidal tumours classified as T1a, T1b and T2 than in older and bigger T3 and T4 tumours, could support this assumption (Eide et al. 2009). The paradox of longer survival of BM positive compared with BM negative patients indicates the same. However, in our material, changing from BM positive status to negative BM was an indicator of a better outcome, as already documented for breast cancer (Wiedswang et al. 2004). Only one patient of eight in this group died of metastases. The numbers are too small to give more than an indication. This influence on survival is not easily explained. However, a positive correlation between local tumour control and longer survival was
documented in a proton beam study with 2435 patients (Egger et al. 2001). In the category changing from BM negative to positive, we found the highest risk for dissemination, six of 13 after two tests or nine of 16 after three tests (56%) (95% CI, 30–80). The difference in mortality between the BM+ and BM in Fig. 2 is astonishing with a shorter survival of the BM . All patients who died of another cancer (20) were BM at inclusion. Excluding these, 20 patients had no influence on the Kaplan–Meier estimates. The most reliable procedure to detect metastasis by BM screening may be to test at diagnosis and retest 2 and 4 years after diagnosis. In our material, metastases are most frequent between 0 and 2 years. The curve for metastases shows a plateau configuration for the first 5 years. After 6 years, there was a marked reduction in the risk for dissemination. An important factor in long-term studies is to make the procedure simple, especially when several samplings at different times are planned to monitor a disease. A test in peripheral blood (PB) is much easier to perform and more convenient for the patient, although BM is relatively easy to sample. Regretfully, the low yield of testing PB (with our method 2% positive) gives this procedure practically no value. A yield of 7% of circulating melanoma cells has been reported (Callejo et al. 2007), but this fraction is much too low to be of clinical interest. Alterations in genes during tumourigenesis allow identification and quantification of circulating tumour-derived DNA (ctDNA)(Madic et al. 2012; Metz et al. 2013) Tumour-specific mutations of GNA11 or GNAQ, which are highly specific for uveal melanoma, have been measured in plasma from metastatic uveal melanoma patients and identified. Most of the ctDNA is derived from tumour tissue cells rather than from circulating ones (Diehl et al. 2005). With a PCR modification, pyrophosphorolysis-activated polymerization plasmatic ctDNA was detected in 20 of 21 metastatic patients and quantification showed correlation with the tumour volume assessed by liver MR (Madic et al. 2012). However, a larger and prospective study is necessary to evaluate the value of this approach in follow-up for patients without diagnosed metastatic disease.
Acta Ophthalmologica 2015
Even if all 13 BM+ patients in the group of non-cancer death were moved to BM+ with melanoma metastasis, thereby increasing this group from 37 to 50 patients, the groups of comparison would be 50/116 (43%) (95% CI, 34–53) versus 66/116 (57%) (95% CI, 47–66). More BM negative patients will still have died of dissemination than BM positive. Therefore, an incorrect death cause classification cannot explain our results. The cohort from 2003 was systematically offered a new sampling and with time an increasing number of patients denied. This fact underscores that clinical studies lasting many years are very complicated to perform; especially when procedures like BM aspiration with some risk for pain is a part of the protocol. One of the main strengths of the present study is that our 328 patients were unselected and collected prospectively, and the inclusion rate was 95%. The follow-up time was more than 6 years for all patients. The same three persons have performed collection of clinical data, sampling and technical processing. This fact and the documentation of both the primary diagnoses and metastases considerably strengthen the material. One of the weaknesses of this study is that there was no regular time interval to BM retesting, no systematic and complete ET testing for the epitope, and the uncertainty of the cause of death with too few autopsies, especially in patients with an acute, unexpected death outside a hospital. Gene testing for prognostication is lacking because our study was planned before this became a routine procedure, although FNAB was performed frequently (Ewens et al. 2013). Our two main hypotheses were not substantiated. The results were opposite to our primary assumption despite the fact that uveal melanoma cells were confirmed to be present in the BM. The immunomagnetic method is simple and inexpensive and shown to work consistently, but does not unveil a high enough fraction of uveal melanoma patients with dissemination for it to be useful in uveal melanoma staging. Therefore, substantial work has still to be performed to develop better tests, especially to pinpoint metastatic melanoma cells in patients with more advanced tumours with a high metastatic potential.
References Bergman L, Seregard S, Nilsson B, Ringborg U, Lundell G & Ragnarsson-Olding B (2002): Incidence of uveal melanoma in Sweden from 1960 to 1998. Invest Ophthalmol Vis Sci 43: 2579–2583. Bruland OS, Høifødt H, Sæter G, Smeland S & Fodstad Ø (2005): Hematogenous micrometastases in osteosarcoma patients. Clin Cancer Res 11: 4666–4673. Callejo SA, Antecka E, Blanco PL, Edelstein C & Burnier MN Jr (2007): Identification of circulating malignant cells and its correlation with prognostic factors and treatment in uveal melanoma. A prospective longitudinal study. Eye (Lond) 21: 752–759. Coupland SE, Sidiki S, Clark BJ, McClaren K, Kyle P & Lee WR (1996): Metastatic choroidal melanoma to the contralateral orbit 40 years after enucleation. Arch Ophthalmol 114: 751–756. Damato B & Coupland SE (2008): Managing patients with ocular melanoma: state of the art. Clin Experiment Ophthalmol 36: 589– 590. Damato B, Duke C, Coupland SE, Hiscott P, Smith PA, Campbell I, Douglas A & Howard P (2007): Cytogenetics of uveal melanoma: a 7-year clinical experience. Ophthalmology 114: 1925–1931. Diehl F, Li M, Dressman D et al. (2005): Detection and quantification of mutations in the plasma of patients with colorectal tumors. Proc Natl Acad Sci USA 102: 16368–16373. Diener-West M, Earle JD, Fine SL, Hawkins BS, Moy CS, Reynolds SM, Schachat AP & Straatsma BR (2001): The COMS randomized trial of iodine 125 brachytherapy for choroidal melanoma, III: initial mortality findings. COMS Report No. 18. Arch Ophthalmol 119: 969–982. Dithmar S, Diaz CE & Grossniklaus HE (2000): Intraocular melanoma spread to regional lymph nodes: report of two cases. Retina 20: 76–79. Edge SB, Byrd DR, Compton CC, Fritz AG, Greene FL & Trotti A (2010): Malignant melanoma of the uvea: at-a-glance. In: Edge SB (ed.). AJCC cancer staging manual, 7th edn. New York: Springer 547–553. Egger E, Schalenbourg A, Zografos L, Bercher L, Boehringer T, Chamot L & Goitein G (2001): Maximizing local tumor control and survival after proton beam radiotherapy of uveal melanoma. Int J Radiat Oncol Biol Phys 51: 138–147. Eide N, Faye RS, Høifødt HK, Overgaard R, Jebsen P, Kvalheim G & Fodstad Ø (2009): Immunomagnetic detection of micrometastatic cells in bone marrow in uveal melanoma patients. Acta Ophthalmol 87: 830–836. Eskelin S, Pyrh€ onen S, Summanen P, Prause JU & Kivel€a T (1999): Screening for metastatic malignant melanoma of the uvea revisited. Cancer 85: 1151–1159. Eskelin S, Pyrh€ onen S, Summanen P, HahkaKemppinen M & Kivel€a T (2000): Tumor
doubling times in metastatic malignant melanoma of the uvea: tumor progression before and after treatment. Ophthalmology 107: 1443–1449. Eskelin S, Pyrh€ onen S, Hahka-Kemppinen M, Tuomaala S & Kivel€a T (2003): A prognostic model and staging for metastatic uveal melanoma. Cancer 97: 465–475. Ewens KG, Kanetsky PA, Richards-Yutz J, Al-Dahmash S, De Luca MC, Bianciotto CG, Shields CL & Ganguly A (2013): Genomic profile of 320 uveal melanoma cases: chromosome 8p-loss and metastatic outcome. Invest Ophthalmol Vis Sci 54: 5721–5729. Faye RS, Aamdal S, Høifødt HK, Jacobsen E, Holstad L, Skovlund E & Fodstad Ø (2004): Immunomagnetic detection and clinical significance of micrometastatic tumor cells in malignant melanoma patients. Clin Cancer Res 10: 4134–4139. Finger PT, Kurli M, Reddy S, Tena LB & Pavlick AC (2005): Whole body PET/CT for initial staging of choroidal melanoma. Br J Ophthalmol 89: 1270–1274. Flatmark K, Bjørnland K, Johannessen HO et al. (2002): Immunomagnetic detection of micrometastatic cells in bone marrow of colorectal cancer patients. Clin Cancer Res 8: 444–449. Gamel JW, McLean IW & McCurdy JB (1993): Biologic distinctions between cure and time to death in 2892 patients with intraocular melanoma. Cancer 71: 2299– 2305. Gragoudas ES, Egan KM, Seddon JM, Glynn RJ, Walsh SM, Finn SM, Munzenrider JE & Spar MD (1991): Survival of patients with metastases from uveal melanoma. Ophthalmology 98: 383–389. Hanahan D & Weinberg RA (2000): The hallmarks of cancer. Cell 100: 57–70. Harbour JW (2009): Molecular prognostic testing and individualized patient care in uveal melanoma. Am J Ophthalmol 148: 823–829. Jensen OA (1963): Malignant melanomas of uvea in Denmark 1943–1952. A clinical, histopatological, and prognostic study. Acta Ophthalmol (Copenh) 43: 1–220. Jensen OA (1982): Malignant melanomas of the human uvea: 25-year follow-up of cases in Denmark, 1943–1952. Acta Ophthalmol (Copenh) 60: 161–182. Kath R, Hayungs J, Bornfeld N, Sauerwein W, H€ o4ffken K & Seeber S (1993): Prognosis and treatment of disseminated uveal melanoma. Cancer 72: 2219–2223. Kivel€a T & Kujala E (2013): Prognostication in eye cancer: the latest tumor, node, metastasis classification and beyond. Eye 27: 243–252. Kodjikian L, Grange JD, Baldo S, Baillif S, Garweg JG & Rivoire M (2005): Prognostic factors of liver metastases from uveal melanoma. Graefes Arch Clin Exp Ophthalmol 243: 985–993. Kujala E, M€akitie T & Kivel€a T (2003): Very long-term prognosis of patients with
65
Acta Ophthalmologica 2015
malignant uveal melanoma. Invest Ophthalmol Vis Sci 44: 4651–4659. Kujala E, Damato B, Coupland SE, Desjardins L, Bechrakis NE, Grange JD & Kivel€a T (2013): Staging of ciliary body and choroidal melanomas based on anatomic extent. J Clin Oncol 31: 2825–2831. Madic J, Piperno-Neumann S, Servois V et al. (2012): Pyrophosphorolysis-activated polymerization detects circulating tumor DNA in metastatic uveal melanoma. Clin Cancer Res 18: 3934–3941. Metz CH, Scheulen M, Bornfeld N, Lohmann D & Zeschnigk M (2013): Ultradeep sequencing detects GNAQ and GNA11 mutations in cell-free DNA from plasma of patients with uveal melanoma. Cancer Med 2: 208–215. Pach JM, Robertson DM, Taney BS, Martin JA, Campbell RJ & O’Brien PC (1986): Prognostic factors in choroidal and ciliary body melanomas with extrascleral extension. Am J Ophthalmol 101: 325–331. Raivio I (1977): Uveal melanoma in Finland. An epidemiological, clinical, histological and prognostic study. Acta Ophthalmol Suppl 133: 1–64. Seddon JM, Gragoudas ES, Egan KM, Glynn RJ, Howard S, Fante RG & Albert DM (1990): Relative survival rates after alternative therapies for uveal melanoma. Ophthalmology 97: 769–777.
66
Seftor EA, Meltzer PS, Kirschmann DA, Pe’er J, Maniotis AJ, Trent JM, Folberg R & Hendrix MJ (2002): Molecular determinants of human uveal melanoma invasion and metastasis. Clin Exp Metastasis 19: 233–246. Singh AD (2001): Uveal melanoma: implications of tumor doubling time. Ophthalmology 108: 829–831. Singh AD & Topham A (2003): Survival rates with uveal melanoma in the United States: 1973–1997. Ophthalmology 110: 962–965. Singh AD, Tubbs R, Biscotti C, Schoenfield L & Trizzoi P (2009): Chromosomal 3 and 8 status within hepatic metastasis of uveal melanoma. Arch Pathol Lab Med 133: 1223–1227. Sisley K, Rennie IG, Cottam DW, Potter AM, Potter CW & Rees RC (1990): Cytogenetic findings in six posterior uveal melanomas: involvement of chromosomes 3, 6, and 8. Genes Chromosom Cancer 2: 205–209. Toivonen P, M€akitie T, Kujala E & Kivel€a T (2004): Microcirculation and tumor-infiltrating macrophages in choroidal and ciliary body melanoma and corresponding metastases. Invest Ophthalmol Vis Sci 45: 1–6. Virgili G, Gatta G, Ciccolallo L, Capocaccia R, Biggeri A, Crocetti E, Lutz JM & Paci E (2007): Incidence of uveal melanoma in Europe. Ophthalmology 114: 2309–2315. Wagoner MD & Albert DM (1982): The incidence of metastases from untreated
ciliary body and choroidal melanoma. Arch Ophthalmol 100: 939–940. Wiedswang G, Borgen E, K aresen R et al. (2003): Detection of isolated tumor cells in bone marrow is an independent prognostic factor in breast cancer. J Clin Oncol 21: 3469–3478. Wiedswang G, Borgen E, K aresen R, Qvist H, Janbu J, Kvalheim G, Nesland JM & Naume B (2004): Isolated tumor cells in bone marrow three years after diagnosis in disease-free breast cancer patients predict unfavorable clinical outcome. Clin Cancer Res 10: 5342–5348. Woll E, Bedikian A & Legha SS (1999): Uveal melanoma: natural history and treatment options for metastatic disease. Melanoma Res 9: 575–581.
Received on June 14th, 2013. Accepted on May 1st, 2014. Correspondence: Nils Eide Eye Department Oslo University Hospital HF Postbox 4950 Nydalen 0424 Oslo Norway Tel.: +47 22118080 Fax: +47 22119989 Emails: [email protected], [email protected]
Immunomagnetic detection of micrometastatic cells in bone marrow of uveal melanoma patients: a paradox Nils Eide,1 Ragnar S. Faye,2,3 Hanne K. Høifødt,3 Berit Sandstad,4 Geir Qvale,1 Rowan Faber,1 Peter Jebsen,5 Gunnar Kvalheim6 and Øystein Fodstad3 1
Eye Department, University Hospital HF, Oslo, Norway Department of Dermatology, Oslo University Hospital HF, Oslo, Norway 3 Department of Tumor Biology, Oslo University Hospital HF and University of Oslo, Oslo, Norway 4 Department of Statistics, Oslo University Hospital HF, Oslo, Norway 5 Division of Pathology, Oslo University Hospital HF, Oslo, Norway 6 Department of Oncology, Oslo University Hospital HF, Oslo, Norway 2
ABSTRACT. Purpose: Our objective was to study survival rates with the bone marrow (BM) results in a cohort of uveal melanoma patients with long follow-up. Methods: Mononuclear cell fractions isolated from BM were examined for tumour cells using our immunomagnetic separation (IMS) method. The patients were classified as BM positive or BM negative. Clinical follow-up, histopathological findings, vital status and cause of death were registered. Results: The study included 328 consecutive patients with uveal melanoma from 1997 to 2006. Tumour cells were found in BM samples in 29% (95% CI, 25–34) at enrolment (96 cases). After a minimum follow-up time of 6 years, 156 (48%) (95% CI, 42–53) melanoma patients had died. The causes were as follows: melanoma metastases 92 (59%), another cancer 20 (13%) and non-cancer 44 (28%). Nine patients were still living with melanoma metastases. Until the latest work-up, 101 (31%) (95% CI, 26–36) patients had developed melanoma metastases. Cyto- or histopathological verification of the metastatic lesions was obtained in 85 cases (84%). In the group with melanoma metastases, 28 tested BM positive at study entry (28%) (95% CI, 19–38). In total, 39 of 101 with metastases tested positive at least once after a maximum of three tests (39%) (95% CI, 29–49). The overall median survival from the first BM test was shorter for the BM negative patients (9.5 years) compared with the BM positive (14.4 years), p = 0.02, log rank test. Conclusion: Ocular melanoma cells detected in BM seem to have a positive prognostic impact on survival in contrast to our original hypothesis. Key words: immunomagnetic detection – micrometastasis in bone marrow – outcome – survival – uveal melanoma
Acta Ophthalmol. 2015: 93: 59–66 ª 2015 Acta Ophthalmologica Scandinavica Foundation. Published by John Wiley & Sons Ltd
doi: 10.1111/aos.12462
Introduction Uveal melanoma is the most common primary intraocular malignant tumour
in adults (Jensen 1963; Raivio 1977), the highest incidence being found in
the Nordic countries (Bergman et al. 2002) with a decreasing north-to-south incidence in Europe (Virgili et al. 2007). At presentation, only 1–4% of patients have disseminated disease (Wagoner & Albert 1982; Pach et al. 1986). The dissemination of uveal melanoma is purely haematogenous, unless the tumour cells perforate the sclera (Dithmar et al. 2000).The liver is the most frequently affected organ in metastatic disease (Jensen 1982; Eskelin et al. 1999). Positron emission tomography–computed tomography (PET/CT) studies reveal a much higher frequency of bone metastases at diagnosis than previously believed and bone is the second most frequent location (Finger et al. 2005). Despite primary enucleation, more than 50% of patients with uveal melanoma develop metastases (Jensen 1963, 1982; Raivio 1977; Kujala et al. 2003). Adjusted for known prognostic parameters, the probability of dissemination seems to be independent of the treatment modality such as enucleation, plaque radiotherapy, proton beam radiotherapy or trans-scleral resection as long as local tumour control is achieved (Seddon et al. 1990; Diener-West et al. 2001; Egger et al. 2001). Cytogenetic studies have disclosed a marked reduction in survival of melanoma patients with tumours carrying certain genetic changes. In
59
Acta Ophthalmologica 2015
particular, the loss of heterozygosity for chromosome 3, gain in 8q or loss in 8p or specific changes in gene expression profiling appear to have a high predictive accuracy (Sisley et al. 1990; Damato et al. 2007; Harbour 2009; Singh et al. 2009; Ewens et al. 2013). Most uveal melanoma metastases occur within the first 5 years after diagnosis. However, the increased risk of death due to this disease is always present, and the cumulative probability of suffering a melanoma-related death increases from 31% after 5 years to 45% after 15 years (Kujala et al. 2003). Tumour dormancy may last up to 20–30 years, but latencies of more than 40 years have been reported (Coupland et al. 1996). Today, there is no effective treatment for uveal melanoma with metastases (Woll et al. 1999). Regardless of treatment, death usually follows within 5–8 months after diagnosis of the metastatic lesion (Gragoudas et al. 1991; Eskelin et al. 2003). A substantial number of patients have subclinical metastases at presentation (Eide et al. 2009). This is supported by the high frequency of circulating melanoma cells detected at diagnosis and at different intervals after primary treatment of patients with primary uveal melanoma (Callejo et al. 2007). Micrometastases may be established years before the tumour is diagnosed and treated. This assumption is supported by sequential observations of metastatic growth and mathematical calculations based on tumour doubling time (Eskelin et al. 2000) and is commented on in a letter discussing these data (Singh 2001). This study has been conducted to detect micrometastases and to explore their possible clinical impact.
approved the study. All patients signed a written consent form before enrolment, which clearly stated that they would get no information about the test results. Further details about the material have been reported previously (Eide et al. 2009). The melanoma cells were harvested from bone marrow (BM) by aspiration from the crista iliaca. The patients were classified as either BM positive or BM negative by finding a rosette: a melanoma cell with five or more beads on its surface (Fig. 1). Two such rosettes are required for a positive sample. The processing of BM and immunomagnetic detection techniques, including sensitivity and validity, have been outlined in previous papers for cutaneous melanomas (Faye et al. 2004), uveal melanomas (Eide et al. 2009), as well as colon cancer, breast cancer and osteosarcoma (Flatmark et al. 2002; Wiedswang et al. 2003; Bruland et al. 2005). During the first 2 years, the patients were evaluated clinically every 3rd or 4th month. Thereafter, medical evaluations, including liver ultrasound and eye examination, were performed at least once annually for 10 years, predominantly at the study centre. Otherwise, local eye departments or general practitioners checked the patient’s eye and general health condition annually. If no follow-up information was automatically received, it was requested by telephone. Computed tomography (CT), magnetic resonance imaging (MRI) and/or PET/CT scan were used to confirm metastases and for staging. Liver function tests were not performed routinely. We always strive for a diagnosis based on pathological tissue, both at initial diagnosis and at dissemination. In co-operation with an oncologist, the indication for not
Material and Methods This prospective study was compiled from 1997 to 2006 at an ophthalmological referral centre in Norway. In total, 328 patients (162 males and 166 females) were included. The mean/median age at inclusion was 63.2/66.0 years (range 8– 94). All patients with a clinical diagnosis of uveal melanoma, independent of location and age, were eligible and asked to participate. Fifteen patients (5%) refused to be included. The Regional Committee for Medical Research Ethics, Southern Norway,
60
harvesting tissue for pathological verification of the metastases was considered. Based on information from the most recent follow-up, we categorized cause of death as due to metastatic melanoma, non-melanoma cancer or a non-cancer related cause. Clinical and histopathological materials available at the time of death were routinely retrieved from clinical departments and general practitioners, and our database was twice compared with the database of the Cancer Registry of Norway. The latest TNM system (7th Edition) staging system is used in this report, (Fig. 3) (Edge et al. 2010). Survival was measured from the time of entry into the study until death or the date of latest observation. Patients still alive without melanoma metastases at the time of statistical work-up were treated as censored. Survival was estimated using the Kaplan–Meier method, and groups were compared with the log rank test. kappa was used for measure of agreement. p-Values 500), (Table 6). The numbers of metastases in the separate groups of 127 samples are listed to illustrate the importance of selecting cut-off points. The numbers of revealed clinical melanoma metastases in the different groups defined by the numbers of rosettes are given. In the material, 23 samples were classified as uncertain, either with only one definite rosette or with a suspect rosette in addition. Both these findings were classified as negative. Verification of finding melanoma cells
Some BM samples were tested with 9.2.27 and three different antibodies in parallel. Comparison of the results obtained with different antibodies binding different epitopes showed good agreement, with kappa test values between 0.69 and 0.76 (Table 7).
Discussion This study is focused on detecting micrometastasis at diagnosis. Our first hypothesis was that finding micrometastatic melanoma cells in BM should be an indicator of prognosis and our second assumption was that the number of BM+ patients should increase with further testing over time, especially those with dissemination. Our results did not confirm any of these hypotheses, although the tumour cells detected and isolated in BM were confirmed to be melanoma cells, as demonstrated by the binding of different fluorescent microspheres coated with two different melanoma antibodies to the surface protein of the cells captured with 9.2.27 immunobeads (Eide et al. 2009). The results were consistent when more antimelanoma antibodies were tested as immunobeads in parallel tubes. The kappa test gives values of 0.69/0.72/0.79 (Table 7). Our hypothesis of an association between the BM results at inclusion and survival seems to be confirmed. The results were, however, opposite to our assumptions (Fig. 2). The BM results could not be used clinically as an indicator of prognosis because the estimate is not accurate enough to predict the outcome for an individual
63
Acta Ophthalmologica 2015
patient. A metastatic index has been elaborated to identify those patients at greatest risk for metastatic death (Damato & Coupland 2008). The staging in the new TNM system is designed to classify patients according to likely survival (Kivel€a & Kujala 2013), and our material shows clear survival differences according to classification as documented by others (Kujala et al. 2013). The detection level of micrometastatic melanoma cells at inclusion was higher than with any other clinical method with 29.3% testing positive. In comparison, PET/CT studies have found the highest frequency of metastases at diagnosis, but the number is lower than 10% (Finger et al. 2005). The number of BM+ samples among the 101 with melanoma metastases increased from 28% to 40% after retesting. The crude survival of our material is similar to reports from other melanoma studies. Approximately 50% of patients with uveal melanoma will die of disseminated disease within 10 years after the diagnosis (Gamel et al. 1993; Singh & Topham 2003). A Finnish group (Kujala et al. 2003) confirms a continuous, although decreasing annual risk for dissemination. In our material, the all-cause mortality was 27% (90/328) (95% CI, 23–33) at 5 years of follow-up and 48% (156/ 328) (95% CI, 42–53) at 10 years. Dissemination without liver affection gives a much better survival, 9.6 years versus 4.6 in our material. That result confirms the findings of an older study (Kath et al. 1993). If our tumours are classified according to the criteria in the COMS study, the small tumours have a probability of 10-year survival of more than 80% in contrast to the medium with 54% and the larger 20%. The importance of tumour localization inside the eye on survival is confirmed with a highly significant effect, p < 0.001. Different cut-off values had a marked influence on the risk for developing metastasis (Table 6). Epithelioid tumour cells, high microvascular density, low pigmentation and type and number of infiltrating macrophages are important parameters that predict outcome, as well as the number of liver metastases and ciliary body involvement (Toivonen et al. 2004; Kodjikian et al. 2005). In our study, cell type or pigmentation in the tumour were not significant predictors.
64
Inside a metastatic lesion from the same primary ET, differences in gene expression profiles have been shown. Some of the metastatic melanoma cells were highly and others poorly invasive (Seftor et al. 2002; Singh et al. 2009). Relatively, few tumour cells were found in the BM samples (Eide et al. 2009). The clinical importance is probably determined by the properties of the malignant cells more than by their numbers. Our results strongly indicate that this assumption is correct. Thus, we found the highest risk in the two groups with few rosettes. In the group with the lowest number of rosettes still defined as BM positive, 44% (95% CI, 27–62) developed metastases, the highest frequency. The frequency of metastasis was the second highest in the group, which was classified as negative 39% (95% CI, 20–61), but there were inconclusive findings according to our definition. The lowest frequency was found in the genuinely negative group. The supposed cascade of the metastatic process may have relevance for some of our data (Hanahan & Weinberg 2000). Some of the detected tumour cells in this study may not have the capacity to establish metastases outside the BM or they may be apoptotic. Therefore, the mere presence of disseminated tumour cells in BM is not sufficient for prognostic or therapeutic purposes in uveal melanoma. Moreover, we cannot rule out the possibility that the metastatic cells in some patients have lost their expression of the epitope still expressed in the ET samples. The fact that the frequency of BM positive patients was higher in small choroidal tumours classified as T1a, T1b and T2 than in older and bigger T3 and T4 tumours, could support this assumption (Eide et al. 2009). The paradox of longer survival of BM positive compared with BM negative patients indicates the same. However, in our material, changing from BM positive status to negative BM was an indicator of a better outcome, as already documented for breast cancer (Wiedswang et al. 2004). Only one patient of eight in this group died of metastases. The numbers are too small to give more than an indication. This influence on survival is not easily explained. However, a positive correlation between local tumour control and longer survival was
documented in a proton beam study with 2435 patients (Egger et al. 2001). In the category changing from BM negative to positive, we found the highest risk for dissemination, six of 13 after two tests or nine of 16 after three tests (56%) (95% CI, 30–80). The difference in mortality between the BM+ and BM in Fig. 2 is astonishing with a shorter survival of the BM . All patients who died of another cancer (20) were BM at inclusion. Excluding these, 20 patients had no influence on the Kaplan–Meier estimates. The most reliable procedure to detect metastasis by BM screening may be to test at diagnosis and retest 2 and 4 years after diagnosis. In our material, metastases are most frequent between 0 and 2 years. The curve for metastases shows a plateau configuration for the first 5 years. After 6 years, there was a marked reduction in the risk for dissemination. An important factor in long-term studies is to make the procedure simple, especially when several samplings at different times are planned to monitor a disease. A test in peripheral blood (PB) is much easier to perform and more convenient for the patient, although BM is relatively easy to sample. Regretfully, the low yield of testing PB (with our method 2% positive) gives this procedure practically no value. A yield of 7% of circulating melanoma cells has been reported (Callejo et al. 2007), but this fraction is much too low to be of clinical interest. Alterations in genes during tumourigenesis allow identification and quantification of circulating tumour-derived DNA (ctDNA)(Madic et al. 2012; Metz et al. 2013) Tumour-specific mutations of GNA11 or GNAQ, which are highly specific for uveal melanoma, have been measured in plasma from metastatic uveal melanoma patients and identified. Most of the ctDNA is derived from tumour tissue cells rather than from circulating ones (Diehl et al. 2005). With a PCR modification, pyrophosphorolysis-activated polymerization plasmatic ctDNA was detected in 20 of 21 metastatic patients and quantification showed correlation with the tumour volume assessed by liver MR (Madic et al. 2012). However, a larger and prospective study is necessary to evaluate the value of this approach in follow-up for patients without diagnosed metastatic disease.
Acta Ophthalmologica 2015
Even if all 13 BM+ patients in the group of non-cancer death were moved to BM+ with melanoma metastasis, thereby increasing this group from 37 to 50 patients, the groups of comparison would be 50/116 (43%) (95% CI, 34–53) versus 66/116 (57%) (95% CI, 47–66). More BM negative patients will still have died of dissemination than BM positive. Therefore, an incorrect death cause classification cannot explain our results. The cohort from 2003 was systematically offered a new sampling and with time an increasing number of patients denied. This fact underscores that clinical studies lasting many years are very complicated to perform; especially when procedures like BM aspiration with some risk for pain is a part of the protocol. One of the main strengths of the present study is that our 328 patients were unselected and collected prospectively, and the inclusion rate was 95%. The follow-up time was more than 6 years for all patients. The same three persons have performed collection of clinical data, sampling and technical processing. This fact and the documentation of both the primary diagnoses and metastases considerably strengthen the material. One of the weaknesses of this study is that there was no regular time interval to BM retesting, no systematic and complete ET testing for the epitope, and the uncertainty of the cause of death with too few autopsies, especially in patients with an acute, unexpected death outside a hospital. Gene testing for prognostication is lacking because our study was planned before this became a routine procedure, although FNAB was performed frequently (Ewens et al. 2013). Our two main hypotheses were not substantiated. The results were opposite to our primary assumption despite the fact that uveal melanoma cells were confirmed to be present in the BM. The immunomagnetic method is simple and inexpensive and shown to work consistently, but does not unveil a high enough fraction of uveal melanoma patients with dissemination for it to be useful in uveal melanoma staging. Therefore, substantial work has still to be performed to develop better tests, especially to pinpoint metastatic melanoma cells in patients with more advanced tumours with a high metastatic potential.
References Bergman L, Seregard S, Nilsson B, Ringborg U, Lundell G & Ragnarsson-Olding B (2002): Incidence of uveal melanoma in Sweden from 1960 to 1998. Invest Ophthalmol Vis Sci 43: 2579–2583. Bruland OS, Høifødt H, Sæter G, Smeland S & Fodstad Ø (2005): Hematogenous micrometastases in osteosarcoma patients. Clin Cancer Res 11: 4666–4673. Callejo SA, Antecka E, Blanco PL, Edelstein C & Burnier MN Jr (2007): Identification of circulating malignant cells and its correlation with prognostic factors and treatment in uveal melanoma. A prospective longitudinal study. Eye (Lond) 21: 752–759. Coupland SE, Sidiki S, Clark BJ, McClaren K, Kyle P & Lee WR (1996): Metastatic choroidal melanoma to the contralateral orbit 40 years after enucleation. Arch Ophthalmol 114: 751–756. Damato B & Coupland SE (2008): Managing patients with ocular melanoma: state of the art. Clin Experiment Ophthalmol 36: 589– 590. Damato B, Duke C, Coupland SE, Hiscott P, Smith PA, Campbell I, Douglas A & Howard P (2007): Cytogenetics of uveal melanoma: a 7-year clinical experience. Ophthalmology 114: 1925–1931. Diehl F, Li M, Dressman D et al. (2005): Detection and quantification of mutations in the plasma of patients with colorectal tumors. Proc Natl Acad Sci USA 102: 16368–16373. Diener-West M, Earle JD, Fine SL, Hawkins BS, Moy CS, Reynolds SM, Schachat AP & Straatsma BR (2001): The COMS randomized trial of iodine 125 brachytherapy for choroidal melanoma, III: initial mortality findings. COMS Report No. 18. Arch Ophthalmol 119: 969–982. Dithmar S, Diaz CE & Grossniklaus HE (2000): Intraocular melanoma spread to regional lymph nodes: report of two cases. Retina 20: 76–79. Edge SB, Byrd DR, Compton CC, Fritz AG, Greene FL & Trotti A (2010): Malignant melanoma of the uvea: at-a-glance. In: Edge SB (ed.). AJCC cancer staging manual, 7th edn. New York: Springer 547–553. Egger E, Schalenbourg A, Zografos L, Bercher L, Boehringer T, Chamot L & Goitein G (2001): Maximizing local tumor control and survival after proton beam radiotherapy of uveal melanoma. Int J Radiat Oncol Biol Phys 51: 138–147. Eide N, Faye RS, Høifødt HK, Overgaard R, Jebsen P, Kvalheim G & Fodstad Ø (2009): Immunomagnetic detection of micrometastatic cells in bone marrow in uveal melanoma patients. Acta Ophthalmol 87: 830–836. Eskelin S, Pyrh€ onen S, Summanen P, Prause JU & Kivel€a T (1999): Screening for metastatic malignant melanoma of the uvea revisited. Cancer 85: 1151–1159. Eskelin S, Pyrh€ onen S, Summanen P, HahkaKemppinen M & Kivel€a T (2000): Tumor
doubling times in metastatic malignant melanoma of the uvea: tumor progression before and after treatment. Ophthalmology 107: 1443–1449. Eskelin S, Pyrh€ onen S, Hahka-Kemppinen M, Tuomaala S & Kivel€a T (2003): A prognostic model and staging for metastatic uveal melanoma. Cancer 97: 465–475. Ewens KG, Kanetsky PA, Richards-Yutz J, Al-Dahmash S, De Luca MC, Bianciotto CG, Shields CL & Ganguly A (2013): Genomic profile of 320 uveal melanoma cases: chromosome 8p-loss and metastatic outcome. Invest Ophthalmol Vis Sci 54: 5721–5729. Faye RS, Aamdal S, Høifødt HK, Jacobsen E, Holstad L, Skovlund E & Fodstad Ø (2004): Immunomagnetic detection and clinical significance of micrometastatic tumor cells in malignant melanoma patients. Clin Cancer Res 10: 4134–4139. Finger PT, Kurli M, Reddy S, Tena LB & Pavlick AC (2005): Whole body PET/CT for initial staging of choroidal melanoma. Br J Ophthalmol 89: 1270–1274. Flatmark K, Bjørnland K, Johannessen HO et al. (2002): Immunomagnetic detection of micrometastatic cells in bone marrow of colorectal cancer patients. Clin Cancer Res 8: 444–449. Gamel JW, McLean IW & McCurdy JB (1993): Biologic distinctions between cure and time to death in 2892 patients with intraocular melanoma. Cancer 71: 2299– 2305. Gragoudas ES, Egan KM, Seddon JM, Glynn RJ, Walsh SM, Finn SM, Munzenrider JE & Spar MD (1991): Survival of patients with metastases from uveal melanoma. Ophthalmology 98: 383–389. Hanahan D & Weinberg RA (2000): The hallmarks of cancer. Cell 100: 57–70. Harbour JW (2009): Molecular prognostic testing and individualized patient care in uveal melanoma. Am J Ophthalmol 148: 823–829. Jensen OA (1963): Malignant melanomas of uvea in Denmark 1943–1952. A clinical, histopatological, and prognostic study. Acta Ophthalmol (Copenh) 43: 1–220. Jensen OA (1982): Malignant melanomas of the human uvea: 25-year follow-up of cases in Denmark, 1943–1952. Acta Ophthalmol (Copenh) 60: 161–182. Kath R, Hayungs J, Bornfeld N, Sauerwein W, H€ o4ffken K & Seeber S (1993): Prognosis and treatment of disseminated uveal melanoma. Cancer 72: 2219–2223. Kivel€a T & Kujala E (2013): Prognostication in eye cancer: the latest tumor, node, metastasis classification and beyond. Eye 27: 243–252. Kodjikian L, Grange JD, Baldo S, Baillif S, Garweg JG & Rivoire M (2005): Prognostic factors of liver metastases from uveal melanoma. Graefes Arch Clin Exp Ophthalmol 243: 985–993. Kujala E, M€akitie T & Kivel€a T (2003): Very long-term prognosis of patients with
65
Acta Ophthalmologica 2015
malignant uveal melanoma. Invest Ophthalmol Vis Sci 44: 4651–4659. Kujala E, Damato B, Coupland SE, Desjardins L, Bechrakis NE, Grange JD & Kivel€a T (2013): Staging of ciliary body and choroidal melanomas based on anatomic extent. J Clin Oncol 31: 2825–2831. Madic J, Piperno-Neumann S, Servois V et al. (2012): Pyrophosphorolysis-activated polymerization detects circulating tumor DNA in metastatic uveal melanoma. Clin Cancer Res 18: 3934–3941. Metz CH, Scheulen M, Bornfeld N, Lohmann D & Zeschnigk M (2013): Ultradeep sequencing detects GNAQ and GNA11 mutations in cell-free DNA from plasma of patients with uveal melanoma. Cancer Med 2: 208–215. Pach JM, Robertson DM, Taney BS, Martin JA, Campbell RJ & O’Brien PC (1986): Prognostic factors in choroidal and ciliary body melanomas with extrascleral extension. Am J Ophthalmol 101: 325–331. Raivio I (1977): Uveal melanoma in Finland. An epidemiological, clinical, histological and prognostic study. Acta Ophthalmol Suppl 133: 1–64. Seddon JM, Gragoudas ES, Egan KM, Glynn RJ, Howard S, Fante RG & Albert DM (1990): Relative survival rates after alternative therapies for uveal melanoma. Ophthalmology 97: 769–777.
66
Seftor EA, Meltzer PS, Kirschmann DA, Pe’er J, Maniotis AJ, Trent JM, Folberg R & Hendrix MJ (2002): Molecular determinants of human uveal melanoma invasion and metastasis. Clin Exp Metastasis 19: 233–246. Singh AD (2001): Uveal melanoma: implications of tumor doubling time. Ophthalmology 108: 829–831. Singh AD & Topham A (2003): Survival rates with uveal melanoma in the United States: 1973–1997. Ophthalmology 110: 962–965. Singh AD, Tubbs R, Biscotti C, Schoenfield L & Trizzoi P (2009): Chromosomal 3 and 8 status within hepatic metastasis of uveal melanoma. Arch Pathol Lab Med 133: 1223–1227. Sisley K, Rennie IG, Cottam DW, Potter AM, Potter CW & Rees RC (1990): Cytogenetic findings in six posterior uveal melanomas: involvement of chromosomes 3, 6, and 8. Genes Chromosom Cancer 2: 205–209. Toivonen P, M€akitie T, Kujala E & Kivel€a T (2004): Microcirculation and tumor-infiltrating macrophages in choroidal and ciliary body melanoma and corresponding metastases. Invest Ophthalmol Vis Sci 45: 1–6. Virgili G, Gatta G, Ciccolallo L, Capocaccia R, Biggeri A, Crocetti E, Lutz JM & Paci E (2007): Incidence of uveal melanoma in Europe. Ophthalmology 114: 2309–2315. Wagoner MD & Albert DM (1982): The incidence of metastases from untreated
ciliary body and choroidal melanoma. Arch Ophthalmol 100: 939–940. Wiedswang G, Borgen E, K aresen R et al. (2003): Detection of isolated tumor cells in bone marrow is an independent prognostic factor in breast cancer. J Clin Oncol 21: 3469–3478. Wiedswang G, Borgen E, K aresen R, Qvist H, Janbu J, Kvalheim G, Nesland JM & Naume B (2004): Isolated tumor cells in bone marrow three years after diagnosis in disease-free breast cancer patients predict unfavorable clinical outcome. Clin Cancer Res 10: 5342–5348. Woll E, Bedikian A & Legha SS (1999): Uveal melanoma: natural history and treatment options for metastatic disease. Melanoma Res 9: 575–581.
Received on June 14th, 2013. Accepted on May 1st, 2014. Correspondence: Nils Eide Eye Department Oslo University Hospital HF Postbox 4950 Nydalen 0424 Oslo Norway Tel.: +47 22118080 Fax: +47 22119989 Emails: [email protected], [email protected]
