Cancer/Testis Genes in Multiple Myeloma: Expression Patterns and Vaccination Strategies Determined by Microarray Analysis.

Abstract

Cancer-testis (CT) antigens are expressed in testis and malignant tumors, but rarely in non-gametogenic tissues. Due to this pattern, they represent attractive targets for cancer vaccination approaches. The aims of the present study were (1) to assess for the first time the expression of CT genes on a pangenomic basis in multiple myeloma (MM), (2) to provide selection strategies of CT antigens for clinical vaccination trials and (3) to assess the impact of CT gene expression on event-free survival. We report here the expression pattern of CT genes in purified MM cells (MMC) of 64 patients with newly-diagnosed MM, 12 patients with monoclonal gammopathy of unknown significance (MGUS), in normal plasma cell and B cell samples and in 20 MMC lines, using gene expression profiling (GEP). Out of 46 CT genes interrogated by the Affymetrix HG U133 Set arrays, 35 were expressed in MMC of at least one patient, according to the Affymetrix “present” call (frequency range: 2% – 66%). Of these, 24 CT genes were expressed in more than 5% of the MMC samples and 25 are located on chromosome X. MMC of 98% of the patients expressed at least one CT gene, 86% at least two, and 70% at least three CT genes. By using a set of 10 CT genes including KM-HN-1, MAGE-C1, MAGE-A3/6/12, MAGE-A5, MORC, DDX43, SPACA3, SSX-4, GAGE-1–8 and MAGE-C2, a combination of at least three CT genes - desirable to circumvent tumor escape mechanisms and immune tolerance - could be obtained in MMC of 67% of the patients. Thus, gene expression profiling can be used to select CT antigens as vaccination targets in individual patients. In a series of MMC from 111 patients treated with the same high-dose chemotherapy and autologous peripheral blood stem cell transplantation protocol and having a median two-year follow-up, we found that the expression of six CT genes, i.e. CTAG-1B, CTAG-2, MAGE-A1, MAGE-A2, MAGE-A3 and MAGE-A6 was associated with a shorter event-free survival (EFS). Furthermore, considering only the 25 CT genes encoded by chromosome X, a CT-Xhigh cluster comprising MMC of one third of the patients (35 of 111) could be defined using a binary hierarchical clustering based on Affymetrix call. Patients in the CT-Xhigh cluster had a shorter EFS (median 13 months) compared to patients in the CT-Xlow cluster (median 18 months, P = .003). The CT-Xhigh clsuter included more patients with a stage III disease (P = .004). These results confirm data from previous studies indicating that patients expressing some CT genes located on chromosome X have a poor outcome.