Prognostic assessment of choroidal melanoma based on interphase cytogenetics and pathological features

It is now well recognised that in uveal melanoma and specifically choroidal melanoma, monosomy 3 has a stronger association with death caused by metastases than clinical and histological parameters. However, the identification of the presence or absence of monosomy 3 and its correlation with morphological heterogeneity within tissue sections has not been assessed. Furthermore, the role of monosomy 3 in predicting time to death is not fully understood. All these aspects have been investigated in more detail in this thesis. In the introduction to this thesis previously published work concerning new and well- established prognostic factors in choroidal melanoma is reviewed. Diagnostic modalities including cytogenetic studies and treatment of this malignant neoplasm and its metastatic disease are also discussed. In the first study of this thesis the technique of chromosome in situ hybridisation (CISH) was adapted to allow its application to archival choroidal melanoma, which was both formalin and gluteraldehyde fixed. The advantage of this technique compared with other techniques available is that it is applicable to archival material allowing the direct correlation of chromosomal changes with histology. CISH was shown to be an effective method for assessing monosomy 3 in these archival paraffin embedded tissue sections. The efficiency of the CISH technique was compared with fluorescence in situ hybridisation (FISH). The main additional advantage of FISH is that it is possible to assess the copy number of several chromosomes at one time. However, a consistent signal could not be obtained with FISH in archival tissue thus the CISH technique was adopted for further study of archival choroidal melanoma. The CISH technique was then used to assess the presence or absence of monosomy 3 in an archival series of melanoma examining tumours from individuals that had died from metastatic melanoma and comparing them with individuals that had survived. The results confirmed the work of previous studies that monosomy 3 is an important predictor of death in choroidal melanoma. However, this study also showed that disomy 3 does not guarantee survival since a small group of tumours displaying disomy 3 resulted in metastatic death. The reasons for this may be related to the inclusion criteria used in this investigation since this was a retrospective study. Furthermore, with the CISH technique used in this study the probes were centromere specific and regional losses of chromosome 3 might not have been detected. However, regional losses seem to occur only in a minority of cases. Finally, and probably more likely, there are additional alterations that allow invasion and metastases in choroidal melanoma. This study also showed that monosomy 3, in some cases, could also be predicted from histology, particularly by the presence of epithelioid cells and vascular loops. One advantage of this archival series was the long follow up available such that late deaths from metastatic melanoma were represented (the longest time to death for metastatic melanoma was 14 years) as well as those that succumbed to metastases at around 3 years. Additional this study showed that monosomy 3 does not influence time until death as there was no significant difference in this interval between metastatic melanomas, with and without monosomy 3. Another advantage of CISH is that it is possible to assess chromosome copy number in conjunction with cell morphology. This aspect of the technique was assessed in the final study of this thesis. A group of choroidal melanomas with well-defined populations of epithelioid and spindle cells were identified from the original study group. Counting was undertaken within these defined morphological populations and monosomy 3 was heterogeneous in a subpopulation of these choroidal melanomas occurring in epithelioid but not spindle shaped cells. This has implications for cytogenetic screening for monosomy 3 since the genetic information obtained from small samples may not be representative. Therefore, a sample composed purely of spindle cells might carry less prognostic significance than one containing both spindle and epithelioid cells. In conclusion, CISH has proven a valuable additional technique allowing retrospective studies of monosomy 3 in choroidal melanoma therefore including larger numbers than most prospective studies and a longer follow-up. Utilising CISH has provided important additional information concerning the significance of monosomy 3 in choroidal melanoma