The RB1 pathway and the p53 pathway represent important, interconnected biochemical units frequently perturbed in human
cancer. Essential
tumor protective mechanisms, such as cellular growth control and apoptosis, are regulated through these systems. Comprehensive studies of these pathways, including most known pathway components, have not been performed in NHL. We therefore analyzed the involvement of aberrations of these pathways in NHLs from the population-based West-Danish NHL registry, LYFO registry, as well as in a series of
neurofibromatosis 1-related
tumors. The aim of the studies was to obtain information about extent and interrelation of alterations of pathway components, as well as clinical information such alterations might provide. We found that alteration of components of one or both of these pathways are very common, occurring in the vast majority of DLCLs. Our data suggest that the pathways are not entirely linear in lymphomagenesis. The p53 pathway components MDM2 and p53 were frequently altered in the same
lymphoma indicating that the role of MDM2 in lymphomagenesis is not entirely dependent on the downstream target, p53. The linearity of the RB1 pathway was clearer as only 1 of 34 DLCLs showed aberration of more than one of the components
cyclin D3, p16INK4A, and pRB. An intriguing novel observation was that p16INK4A inactivation was associated with increased expression of cdk4, a
kinase target of p16INK4A inhibitory function. This could indicate the existence of a regulatory feedback loop between p16INK4A and cdk4.
Cyclin D3 has yet to be established as an
oncoprotein. Our finding of
cyclin D3 overexpression in a significant number of DLCLs (including all thyroid
lymphomas analyzed), as well as the intimate inverse relation to other RB1 pathway alterations suggest, that
cyclin D3 is important in lymphomagenesis. However, further studies are needed to implicate
cyclin D3 definitively as an
oncoprotein. Our data contain several lines of evidence supporting roles of CDKN2A and MDM2 in progression of neoplastic disease. We found that loss of p16INK4A coincided with transformation of
neurofibromas to
malignant peripheral nerve sheath tumors in
neurofibromatosis 1 patients. Furthermore, one DLCL lost CDKN2A from diagnosis to relapse. MDM2 overexpression was more frequent in aggressive than in indolent
lymphomas, and in follicle center
lymphomas none of our follicle center grade I/II
lymphomas overexpressed MDM2. In contrast, MDM2 was overexpressed in 60% of grade III/diffuse follicle center
lymphomas. Clinical correlations revealed novel and interesting findings. Both p53 disruption and low expression of E2F-1 correlated with poor response of aggressive
lymphomas to treatment. Chemotherapeutic regimens used in
lymphoma treatment are based on apoptosis induction, and as both E2F-1 and p53 are regulators of apoptosis, it is possible that the observed treatment failure is associated with reduced E2F-1- and p53-mediated apoptosis. Survival analyses revealed numerous novel and potentially important findings. Several of the studied cell cycle regulators carried independent prognostic value in various subsets of
lymphomas. In DLCL, both p16INK4A inactivation and reduced E2F-1 expression conferred shortened survival. p53 alteration was associated with poor prognosis of both B-cell and, especially,
T-cell lymphoma. Low expression of p27, a cell cycle regulator haplo-insufficient for
tumor suppression, predicted poor outcome in indolent and aggressive
lymphoma, and overexpression of
cyclin D3 was associated with poor prognosis in indolent
lymphomas. Finally, MDM2 overexpression identified among patients with follicle center
lymphomas, extranodal marginal zone
lymphomas, and
mantle cell lymphomas cases with poor prognosis. While these results must necessarily be confirmed on larger prospective series of patients, the data nonetheless suggest that valuable prognostic information can be provided by studies of these cell cycle regulators.