The ultimate utility of MRD detection is to determine the clinical significance of occult tumor cells in relations to patient relapse, survival, and even cure, using progression-free and overall survival as the clinical end points. Prognostic impact of MRD needs to factor in the genetic profile of the tumor. For example, NB patients with stage-4s disease are well known to have marrow disease, and yet they predictably have favorable outcome. This suggests that the presence of marrow disease may not necessarily be clinically relevant to all NB stages. In fact, a Children's Cancer Group study with 374 patients reported no statistically significant difference between stage-1 and stage-2 patients who had immunocytology-positive vs immunocytology-negative marrow disease at diagnosis (Perez et al. 2000). Interestingly, in a compa rable patient group, a false-positive rate of >30% was found after genetic verification of the immunologically positive DTCs (Mehes et al. 2001). Indeed, an unambiguous identification and quantification of MRD must be a requisite for MRD monitoring. For patients with stage-4 NB undergoing multi-modality treatment, the clinical significance of MRD is highly dependent on when the sampling is being carried out.
Several research groups have reported that among stage-4 NB patients older than 1 year at diagnosis, MRD during and after therapy had a statistically significant impact on patient survival (Table 11.7.2). Some studies have further underscored the adverse effect of the presence of NB cells in BM and PB at diagnosis on clinical outcome (Seeger et al. 2000; Burchill et al. 2001). The implication of this conclusion is that emphasis must be placed on improved therapeutic strategies. It is not surprising to find the presence of MRD at the end of therapy to have prognostic importance, as demonstrated by the MSKCC and UKCCSG studies (Cheung and Cheung 2001; Cheung et al. 2000; Burchill et al. 2001). More relevant to the clinical management of NB is likely the impact of MRD during and after induction (Seeger et al. 2000; Fukuda et al.2001).In BM/PBSC harvest, an adverse effect on survival was demonstrated with > 100 tumor cells per 105 nucleated BM detected by im-munocytology (Seeger et al. 2000), >500 TH transcripts in PBSC by qRT-PCR (Tchirkov et al. 2003), and >5 GD2 synthase transcript units in prepurged BM by qRT-PCR (Cheung et al. 2002). In contrast, Handgretinger et al. reported that patients with CD34+ PBSC grafts containing >2000 tumor cells as measured by immunocytology using chimeric anti-GD2 antibody ch14.18 had a lower risk of relapse than patients with fewer contaminating tumor cells (Handgretinger et al. 2003). It was suggested that a threshold number of tumor cells would elicit an antitumor immune response after autologous transplant, although false positivity remains a possibility. With the advent of novel post-induction therapies, MRD serves as a sensitive surrogate response marker in comparing the efficacy of different adjuvant therapies. The presence of positive GD2 synthase transcript in patients who did not achieve CR/VGPR before the onset of radioimmunotherapy had a higher risk of relapse and death (Cheung et al. 2003a). For patients undergoing another adjuvant therapy with a combination of anti-GD2 antibody 3F8 and GM-CSF, early molecular response was found to have prognostic impact on progression-free survival (Cheung et al. 2003b). The ability to identify a subset of patients who are unlikely to benefit from this adjuvant therapy and are at a great risk of relapse may provide the rationale for a more timely application of alternative treatment options.
Was this article helpful?