Good question. In brief, pRb function is important during the retinal development in early stage. Therefore, defects in pRb function give rise to childhood-associated retinoblastoma.   

Thanks for appreciation, As pRB plays additional role along with cell cycle regulation in retinal cell, the impaired activity of pRB is prone to childhood-associated Retinoblastoma.

Will surf to find similar such additional role of BRCA in mammary development  ..

Thanks again sir, its half decade question in my mind .. is getting clear now.. 

This is an intriguing question. It is probably an important issue especially in the context of model systems we use to understand tumorigenesis. I assume that the cells from which the retinoblastoma originate from in humans are the most susceptible to the loss of one functional RB1 allele. In the mouse, the loss of RB1 does not do much to the same cells most likely due to redundancy with other RB-family members. The RB1 mouse model preferentially shows pituitary tumors and thyroid tumors but to my knowledge no retinoblastomas (unless you drive the loss of RB1 and other factors specifically in this tissue).

The same is true for e.g. NF1 and to some extent the p53-mediated cancer spectrum in humans and mice: it is not really the same. For BRCA1 it is even more striking, baffling scientists so much that they write: "The failure of mutations in Brca1 to lead to increased tumor susceptibility has been documented by several groups, each of which generated mouse lines carrying different Brca1 mutations (7, 8, 9, 10, 11) . Thus the mouse differs markedly from the human, in which the presence of a mutant BRCA1 allele results in an estimated 80–90% overall lifetime risk for developing breast and/or ovarian cancer (3) ." (Mammary Tumor Formation in p53- and BRCA1-deficient Mice. Cell Growth Differ. 1999 Jan;10(1):1-10). Even improved Brca1 mutants did not exhibit primarily breast cancer but again lymphomas (Tumorigenesis in mice carrying a truncating Brca1 mutation. Genes Dev., 15 (2001), pp. 1188–1193).

If you want to make a very harsh statement then you could say that genetic mouse cancer models have contributed very little to our understanding of cancer, to forging novel treatment options or identifying treatment targets. I have not done a comprehensive survey but it may be that most of the progress in cancer research stems simply form carefully analyzing human tumors, determining mutations and altered pathways, and developing drugs against the identified alterations. Mouse models may have played a role in confirming the principal activity of the identified pathways and genes but in general they have not been the driver. I may be wrong but that is my impression.

Therefore, I am not entirely sure what can be learned about the specifics of human retinoblastoma by looking at mouse RB1 models. However, such models allow for a deeper understanding of carcinogenesis since comparison of different mammalian species should reveal how these genes work and give clues how they produce different cancer spectra in different species. I think this is sometimes referred to One medicine or One health or simply comparative biology.

Overall, cancer seems to have the tendency to affect mammals differently. The most common tumors in mice are sarcomas and lymphomas, in dogs histiocytomas, lipomas and adenomas etc. So you could also turn your question towards: why do mice have different tumors than humans, than dogs etc. There seems to be a lot of it in the "fine-print", details and context. The human eye seems to be the weakest link when it comes to RB1 mutations, the human breast when it comes to BRCA1 mutations and so forth. 

However, human retinoblastoma patients have secondary tumors outside the eye that include bone and soft tissue sarcomas. Unfortunately, I am not expert enough to tell whether this is generally due to the treatment of the retinoblastoma (e.g. radiation) and whether these secondary tumors occur frequently in the area where treatment of the primary tumor may have made it more likely to form such secondaries...