Applications of monoclonal antibodies in the investigation, diagnosis, and treatment of retinoblastoma.

In Europe and North America, retinoblastoma has a reported incidence ranging from 1 in 14 000 to 1 in 36 000 live births'-5 and is the most common ocular malignancy affecting children and infants. Early diagnosis allows the application of a number of therapies directed at localised disease, including external beam and plaque radiotherapy, cryotherapy, and photocoagulation, although the principal treatment accounting for the high cure rate ofaround 90% is enucleation, which is still carried out in over half of all cases. The justification for such radical therapy is the poor prognosis once the tumour escapes the eye, and the high mortality from metastases.6 In the developing countries ofAfrica and Asia the picture is different. The higher reported incidence in many areas,7-" coupled with a lower incidence ofocular malignant melanoma in particular racial groups results in retinoblastoma being the commonest ocular tumour in all age groups. Moreover, late diagnosis and treatment lead to a correspondingly high rate of mortality due to metastatic spread, approaching 100% in some areas.'0 Therefore the inability to deal effectively with metastatic retinoblastoma results in a high mortality in areas where medical care is limited and the extensive use of prophylactic enucleation in developed countries. Despite the efficacy of current therapeutic intervention, there remains a need for further investigation into improvements in treatment of metastatic disease. Recent advances in the understanding ofthe cancer process are due largely to the advent ofmolecular genetics and the use of monoclonal antibodies, and perhaps the greatest contribution ofsuch techniques has been in studies of retinoblastoma. The occasional association of the familial predisposition to retinoblastoma with a visible chromosome deletion at 13ql4,'2-'4 and the linkage to the gene coding for esteraseD,'3 11 6 confirmed the genetic basis for this disease. Epidemiological evidence led Knudson to formulate the 'two hit' hypothesis for the tumorigenesis of retinoblastoma, postulating that both copies of the viable gene must be inactivated by two independent mutational events (or 'hits').' 8 An individual with the familial predisposition has inherited one inactive gene, such that only one subsequent mutational event is necessary, during the susceptible stage of embryonic development, to cause a retinoblast to become malignant. Retinoblastoma was the first cancer to be identifie7d with a tumour suppressor gene, or anti-oncogene,'" to have this gene, RB1, located and cloned, and its protein product, pl1 0' ', isolated.4 19 The inheritance is autosomally dominant in individuals with a single germline mutation due to the high probability of a second 'hit' inactivating the single remaining viable copy of the gene in one or more of the vulnerable retinoblasts. This malignancy is recessive at the genetic level, as indicated by the frequent loss of heterozygosity in tumour cells'920 and reversal of the malignant phenotype in tumour derived cell lines fused with RB1 competent cells.2' Reversal of malignancy has also been reported in cell lines transfected with functional BR1 gene constructs. Recently, however, these results have been challenged by other researchers who have found incomplete reversal of malignancy in retinoblastoma cell lines,22 although such effects may be artefacts resulting from additional genetic changes having occurred in cells adapting to tissue culture environments. The importance of RB1 in malignant transformation has also been demonstrated by inactivating its product by association with virus proteins such as adenovirus EIA,23 SV40 large T antigen24 and papillomavirus E7 oncoprotein,2" both by transfection and transgenic manipulations.26 Furthermore, this gene has been implicated in the progression of other cancers, suchas colon and breast carcinomas, and hence its significance as a tumour suppressor extends beyond its role in the tumorigenesis of retinoblastoma. Perhaps the greatest recent clinical advance to emerge from theoretical studies ofretinoblastoma, is in the field ofprenatal and perinatal screening and genetic counselling. The predisposing gene for retinoblastoma may be identifed using DNA probes recognising restriction fragment length polymorphisms (RFLP, variable length fragments of DNA prepared by sequence dependent endonuclease activity) on Southern blot analyses of genomic DNA.'215 This enables predisposed individuals to be identified in utero, as well as recognising carriers of the aberrant retinoblastoma gene in cases of incomplete penetrance. 5 Monoclonal antibodies provide highly specific molecular recognition that can characterise and selectively localise tumour cells. To date their greatest contribution has been in elucidating the cellular origin of tumours, by equating immunohistochemical profiles in neoplastic cells with those of defined cell types in their tissue of origin. Although antibodies are highly specific for their molecular markers, the apparent lack of antigens expressed uniquely by cancer cells has limited their diagnostic and therapeutic applications. By identifying the inadequacies of monoclonal antibodies, researchers are now devising means of optimising their use in the clinical situation, and recent advances in recombinant technology may help in overcoming many oftheir limitations.