Supervisors info:
Μπάμια, Χριστίνα, Αν. Καθηγήτρια, Ιατρική, Ε.Κ.Π.Α.
Σουλιώτης, Βασίλης, Α' Ερευνητής, Εθνικό Ίδρυμα Ερευνών
Μπουρνέτας, Απόστολος, Καθηγητής, Μαθηματικό, Ε.Κ.Π.Α.
Summary:
Background: We studied the molecular changes that occur in the DNA damage response system (DDR) and their possible association with the stages of Multiple Myeloma (MGUS-SMM-MM) and patients’ response to melphalan treatment. Additionally, it was assessed whether the measurements taken from two tissues (peripheral blood lymphocytes and bone marrow plasma cells) are correlated and lead to equivalent conclusions.
Methods: Molecular changes were evaluated using 12 parameters measured in two tissues: peripheral blood lymphocytes (PBMCs) and bone marrow plasma cells (BMPCs). In total, 89 subjects were included in the present study: 20 MGUS, 29 SMM, 15 MM and 25 healthy (only PBMCs measurements). Next, five scores measuring DNA damage response were calculated based on simple sums of 1) the standardized values of the 12 initially measured parameters (scores 1 & 2) and 2) the binary variables (score 3, 4 & 5), as defined by calculating the optimal cut-off values using ROC curve analysis. Finally, the discriminant ability of the above scores with respect to disease stages and patients’ response to treatment was evaluated.
Results: The values of all 12 parameters varied significantly with disease progression (MGUS->SMM->MM). Significant difference in these parameters between responders and non-responders to melphalan treatment was observed. Regarding the direction of variation, the results were similar using either BMPCs or PBMCs measurments. Regarding their discriminant ability, for the BMPCs measurements, score 5 which was based on all 12 parameters seems to be the optimal score in classifiying patients to the disease stages and to responders/non-responders to chemotherapy (MGUS vs. SMM: perfect prediction, SMM vs. MM: AUC=0.935, responders vs. non-responders: perfect prediction). This score seems to also perfectly discriminate between responders and non-responders for the PBMCs measurements. For the PBMCs measurements, scores 1 and score 3 which were calculated using only the 9 parameters with increasing mean value with disease progression, seems to best differentiate among disease stages (score 1: Healthy vs. MGUS AUC=0.822, MGUS vs. SMM AUC=0.9207, SMM vs. MM AUC=0.8851 and score 3: Healthy vs. MGUS AUC=0.876, MGUS vs. SMM AUC=0.9155, SMM vs. MM AUC=0.8425). Variable RNA synthesis seems to have a particularly high discriminant ability in PBMCs (Healthy vs. MGUS: AUC=0.990, MGUS vs. SMM: AUC=0.992, SMM vs. MM: perfect prediction, responders vs. non-responders AUC=0.907).
Conclusions: The DNA repair rate, as measured in the easily accessible blood tissue, can be indicative of patients’ response to therapy at an individual level. Meanwhile, the 12 parameters seem to also be indicative of MM stage. Further investigation is deemed necessary.
Keywords:
Multiple myeloma, DNA damage response, Melphalan, ROC, Factor analysis