Prognostic Impact of the VH Mutation Status

To examine the VH mutation status in a large series (n = 300) of B-CLL patients, the VDJ-rearrangement of the immunoglobulin genes was amplified by PCR from genomic DNA, and the

0 24 48 72 96 120 144 168 192

0 24 48 72 96 120 144 168 192

Months

Months

Fig. 6. Probability of survival in B-CLL patients with mutated and unmutated VH genes according to the 97% cutoff values. (A) The estimated median survival time for the VH homology > 97% group was 79 mo. The last observed death in the VH homology < 97% group was after 152 mo of follow-up time (survival probability 56%). (B) When only patients diagnosed at Binet stage A were evaluated the estimated median survival times for the VH homology > 97% and VH homology < 97% groups were 79 mo vs not reached (last observed death after 152 mo of follow-up time; survival probability 53%). (From ref. 10, with permission.)

0 24 48 72 96 120 144 168 192

0 24 48 72 96 120 144 168 192

Months

Months

Fig. 6. Probability of survival in B-CLL patients with mutated and unmutated VH genes according to the 97% cutoff values. (A) The estimated median survival time for the VH homology > 97% group was 79 mo. The last observed death in the VH homology < 97% group was after 152 mo of follow-up time (survival probability 56%). (B) When only patients diagnosed at Binet stage A were evaluated the estimated median survival times for the VH homology > 97% and VH homology < 97% groups were 79 mo vs not reached (last observed death after 152 mo of follow-up time; survival probability 53%). (From ref. 10, with permission.)

mutation status of the VH genes was determined by DNA sequencing (10,132). Taking the classical cutoff value of 98% homology to the nearest related germline gene to differentiate between mutated and unmutated VH genes, 132 cases (44%) showed mutated and 168 cases (56%) unmutated VH genes. The method of maximally selected log rank statistics was applied to test the prognostic relevance of the VH mutation status. A corrected p value (pcor) for the best possible separation of two subgroups with different survival probabilities was found at a VH homology to the nearest related germline gene of 97% (pcor < 0.001; 95% confidence interval 96-98%) (10). With a cutoff value of 97% homology to the nearest related germline gene, 112 cases (37%) showed mutated and 188 cases (63%) unmutated VH. The Kaplan-Meier estimate of the median survival time in the two VH subgroups differed both for the overall group (n = 300) and within the subgroup of patients in Binet stage A at the time of diagnosis (n = 189) (Fig. 6) (10).

In addition to the VH mutation status, the study of the structure of the VDJ rearrangement and the character of the mutations with respect to biological factors of disease etiology, like antigen selection, is of interest. Until recently, there were only studies available on small numbers of cases (123,125,133,134, and references therein). In a large B-CLL series, at least one clonal VDJ rearrangement of genomic DNA could be amplified in all 300 cases (132). Cases with mutated VH showed a different VDJ rearrangement structure than cases with unmutated VH. Genes of the VH3 and VH4 families were over-represented in the mutated VH subgroup, whereas the VH1 family was found more frequently in the VH unmutated subgroup. Specific VH genes were responsible for the differences, and these imbalances were in line with previous studies (123,125,133,134, and references therein). The mean length of the CDR3 region differed significantly between the VH mutated and unmutated subgroups. The median mutation rates and ratios of replacement/ silent (R/S) mutations were greater within the VH subregions in the CDRs than in the FRs. Cases with less than 98% homology to the nearest related germline gene were examined by means of the algorithm of Chang and Casali (135) for evidence of an antigen selection in the mutation pattern of the VH gene. In 43 cases, mutation patterns consistent with antigen selection were

Was this article helpful?

0 0

Post a comment