Logo: to the web site of Uppsala University

The clinical course of chronic lymphocytic leukemia (CLL) is highly heterogeneous, which has prompted the search for biomarkers that can predict prognosis in this disease. The IGHV gene mutation status and certain genomic aberrations have been identified as reliable prognostic markers of clinical outcome for this disorder. However, the search for more feasible prognostic markers in CLL is still being pursued. Recently, certain single nucleotide polymorphisms (SNPs) in the GNAS1, BCL2 and MDM2 genes and the RNA expression levels of the LPL, ZAP70, TCL1, CLLU1 and MCL1 genes were suggested as novel prognostic markers in CLL.

In papers I-III, we performed genotyping analyses of the GNAS1 T393C, BCL2 -938C>A and MDM2 SNP309 polymorphisms in 268-418 CLL patients and related the genotypes with clinical data. Association studies between the polymorphisms and established prognostic markers (i.e. IGHV mutation status, genomic aberrations, CD38 expression) were also performed. Our studies did not find any significant relationship between these SNPs with either clinical outcome or other known prognostic markers in CLL.

In paper IV, we measured the RNA expression levels of LPL, ZAP70, TCL1, CLLU1 and MCL1 in 252 CLL cases and correlated these levels with clinical outcome. Here, we verified that high expression of all these RNA-based markers, except MCL1, were associated with an unfavourable prognosis. We also confirmed a close relationship between IGHV mutation status and the RNA-based markers, especially for LPL and CLLU1 expression. Among the RNA-based markers, multivariate analysis revealed LPL expression as the strongest independent prognostic marker for overall survival and time to treatment. Furthermore, the RNA-based markers could add further prognostic information to established markers in subgroups of patients, with LPL expression status giving the most significant results.

In summary, data from papers I-III could not verify the GNAS1 T393C, BCL2 -938C>A and MDM2 SNP309 polymorphisms as prognostic markers in CLL. Future SNP markers must hence be confirmed in large, independent cohorts before being proposed as prognostic marker in CLL. In paper IV, we conclude that LPL expression appears to be the strongest among the RNA-based markers for CLL prognostication. Further efforts to standardize LPL quantification are required before it can be applied in the clinical laboratory to predict clinical outcome in this disease.

The 309T>G polymorphism in the promoter region of the MDM2gene, known as SNP309, has recently been suggested as an unfavorable prognostic marker in chronic lymphocytic leukemia (CLL) although this has been questioned. To investigate this further, we analyzed the MDM2 SNP309 genotypes in 418 CLL patients and correlated the results with established CLL prognostic factors, time to treatment and overall survival. In this Swedish cohort, no association existed between any particular MDM2 SNP309 genotype, overall survival and time to treatment. Furthermore, no correlation was shown between the MDM2 SNP309 genotypes and Binet stage, IGHV mutational status and recurrent genomic aberrations. In summary, this study argues against the use of the MDM2 SNP309 as a prognostic marker in CLL.

In this study, we followed minimal residual disease (MRD) in eight children with Philadelphia-positive acute lymphoblastic leukemia (Ph+ ALL) using (i) flow cytometry (FCM), (ii) real-time quantitative PCR of IG/TCR gene rearrangements and (iii) RT-PCR detecting fusion gene transcripts. In six of the eight cases the kinetics of MRD clearance was comparable. One of the two discordant cases could be explained by presence of an alternative fusion transcript. The other discordant case showed high BCR-ABL1 RNA level while the other methods did not detect any MRD. In our limited material quantitative RT-PCR of fusion gene transcripts seemed particularly useful to measure MRD in Ph+ ALL. However, BCR-ABL1 expression may not reflect the percentage of leukemic cells as FCM and IG/TCR rearrangement quantification do, and these methods are thus complementary.

In this Swedish multi-center study of early treatment response in childhood acute lymphoblastic leukemia (ALL), we evaluated the concordance between multicolour flow cytometry (FCM) and real-time quantitative polymerase chain reaction (RQ-PCR) for assessment of minimal residual disease (MRD). Multiple time points (i.e. day 15, 29, 50 and 106) were evaluated with the NOPHO (Nordic Society of Pediatric Hematology and Oncology) ALL 2000 treatment protocol as backbone. During 2002-2006, 334 children were diagnosed with ALL, where 228 had paired samples taken at any of the four time points. With the detection level of 0.1%, the concordance between RQ-PCR and FCM was 90% in the 726 paired samples analyzed. At day 29, the correlation between the methods was greater with MRD levels >0.1% (rs=0.7, p<0.001) than below (rs=0.2, p=0.024). MRD levels higher than 0.1% at day 29 was a significant predictor of higher risk of having a bone marrow relapse. This was true both for BCP ALL and T-ALL analysed with either FCM or RQ-PCR, although RQ-PCR was a better discriminator than FCM in T-ALL. However, using the NOPHO ALL 2000 protocol, our data indicate that a higher cut-off value (0.2%) should be applied in BCP ALL when using RQ-PCR as MRD method. In contrast, MRD levels ≥ 0.1%, analysed with either method late during induction therapy, was not a predictor of isolated extramedullary relapse. We therefore conclude that MRD assessment by RQ-PCR based IG/TCR rearrangement and multicolour FCM monitoring can be used as a clinical tool if the aim is to find childhood ALL cases with increased risk of having bone marrow relapses.

Minimal residual disease (MRD) detection during the early treatment phase has become an important stratification parameter in many childhood acute lymphoblastic leukemia (ALL) treatment protocols. Here we aimed to address the applicability of real-time quantitative polymerase chain reaction (RQ-PCR) of rearranged antigen receptor genes as an MRD method in a multi-centre setting. From a Swedish population-based cohort of 334 ALL cases diagnosed during 2002-2006, we analysed 279 diagnostic samples (84%) by screening for rearranged immunoglobulin (IG) and T-cell receptor (TCR) genes. Allele-specific oligonucleotides were designed for each patient rearrangement, and the sensitivity and quantitative level was determined for each target. The analyses were performed at five different centres while interpretation of the results was performed at consensus meetings. Overall, clonal IG/TCR rearrangements were detected in 97% (236/244) of B-cell precursor ALL (BCP ALL) and 94% (33/35) of T-ALL. A sensitive RQ-PCR analysis (≤ 10^-4) was obtained in 89% (216/244) of BCP ALL and in 74% (26/35) of T-ALL. With the stratification threshold of ≥10^-3 for identification of high-risk patients, 93% of BCP ALL and 86% of T-ALL reached this quantitative range by at least one target gene. Taken together, this national multi-centre study supports the use of RQ-PCR analysis as a robust method for MRD detection in the majority of childhood ALL cases.