Multiple Myeloma (MM) is a lethal hematological malignancy with an incidence of 40–60 per 1.000.000 per year. Over the last decade, the most significant therapeutic improvement has been the introduction of high-dose Melphalan therapy in combination with autologous stem cell support. This strategy has led to an increase in survival from a median of approximately 3 years till now >6 years also in Denmark (unpublished data from Nordic Myeloma Study Group, NMSG #7/98). Despite this important increase in survival, substantial variability still persists among patients in response to high-dose Melphalan. Clinical and genetic heterogeneity is a characteristic feature of MM, and new methodological approaches, such as global gene expression profiling (GEP), are valuable tools in order to link clinical endpoints, such as response to treatment and survival, with the molecular and genetic heterogeneity of the disease (

Shaughnessy, J. et al.,
). Here we present a bioinformatic analysis of the clinical impact of a novel chemo sensitivity index based on GEP of 1) a panel of sixty cell lines (NCI60) that subsequently have been exposed to a range of anti-cancer agents, including Melphalan, and, 2) a retrospective validation of our method from lung, brain and lymphoma patients. Figure 1, presents a retrospective leave-one-out cross-validation of the prediction whether a patient will respond to the treatment given. Lung- and brain cancer patients were treated with Cisplatin, and the predicted sensitivity or resistance to this drug was used to group patients (hazard ratio 2.9, 95% CI 0.7–11). Lymphoma patients were treated with Vincristine and Adriamycine, and the patients were grouped according to their predicted sensitivity or resistance to these drugs. Patient and raw DNA chip data were obtained from: Beer et al, Nature Medicine 8(8):816–824; Pomeroy et al, Nature 415(6870):436–442; Shipp et al, Nature Medicine 8(1):68–74. For each patient, the method may also allow for a quantitative prediction of drug sensitivity towards a range of drugs including Melphalan. Figure 2 shows the drug sensitivity profile for two selected Multiple Myeloma patients (patient #2 and #9 from Agnelli L et al. Haematologica. 2007 Jan; 92(1):56–65). We are currently collecting chip data from a cohort of 70 newly diagnosed biobanked bone marrow samples from MM patients treated with high-dose Melphalan in NMSG trial #9/99 in order to further confirm these results in MM. These data are subsequently being validated using an independent population of bone marrow samples from MM patients also treated with high-dose Melphalan. We demonstrate that GEP can provide new information concerning resistance towards Melphalan and a range of anti-cancer drugs, separately or in combination (multi drug regimen), which may be clinically useful, and serve as a guide for the future development of individualized treatment strategies.

Author notes

Disclosure: No relevant conflicts of interest to declare.