Abstract

BACKGROUND: IgVH mutation status is a well accepted prognostic marker in CLL. Typically >98% sequence homology to the most similar germline gene is considered unmutated, and ≤98% is considered mutated. The 98% cutoff was originally chosen to account for undefined IgVH polymorphisms that could be responsible for DNA sequence disparity. Prior data showed that mutated patients have a more benign course of disease vs. unmutated patients. Although it has been presumed that this is due to a structural change in the BCR, specific consideration of the nature of the mutations (silent or replacement) and the relation to clinical course has not been made. We recently showed that IgVH somatic mutation correlates to disease course in a continuous fashion; prognosis improves as the number of mutations increases. Therefore, the present study was designed to determine if replacement mutations were more important than silent mutations or if the “mutation process” (reflected as the total number of mutations, irrespective of their structural impact) was responsible for improved survival

METHODS: We sought to determine if further characterization of mutations (replacement vs. silent) and location thereof (framework (FR) v. complementarity determining region (CDR)) changed with increasing number of mutations. We hypothesized the ratio of replacement-to-silent (R:S) mutations would be greater in those with better prognosis because such changes could significantly alter the structure of the antigen-binding site and that this would be more pronounced in the CDRs. Survival analysis (Kaplan-Meier, logrank test (overall and adjusted pairwise), Cox regression) was used to analyze both outcomes of time to treatment (TTT) and survival (OS). Quartiles of numbers of mutations were used to determine how outcome varied accordingly. The range of mutation in each quartile divided as follows: 1–9; 10–15; 11–20; and 21–39. We examined the ratio of R:S mutations to determine whether the ratios varied according to quartile (Mann-Whitney test).

RESULTS: 353 patients were analyzed. 89 of 334 (27%) had received treatment and 53 of 353 (15%) had died. The lower quartile of mutations (1–9) had a significantly shorter TTT and OS than the three higher quartiles (P< 0.0006, P<0.0014, respectively, logrank test); the three upper quartiles did not differ statistically from one another. Among mutations in the CDR, there was a significant difference in the R:S ratio between the lower quartile of mutation number and the combined 3 upper quartiles. (Mean R:S = 3.25 in Q1, 5.64 in Q2–Q4, P<0.0008). Among mutations in the FR, R:S did not differ significantly between quartile groups (Mean R:S = 2.44 in Q1, 2.19 in Q2–Q4). Among all mutations (FR and CDR), R:S ratio did not differ significantly between quartile groups. (Mean R:S = 2.81 in Q1, 2.55 in Q2–Q4).

CONCLUSION: The association of better prognosis with more IgVH mutations may reflect the higher ratio of R:S mutations in the CDR. These data have implications as to the role of the CLL B-cell receptor (BCR) and its specificity in clinical outcome. Assuming that in vivo BCR stimulation causes CLL cell proliferation, increased replacement mutations in the CDRs would lead to more binding site structural or conformational change and thereby narrow antigen-binding range. This would limit the diversity of antigens capable of delivering a stimulatory signal. That no increase in R:S was found in the FRs is consistent with these regions’ function as a scaffolding to the BCR, the structural alteration of which could render the BCR ineffective.

Patients divided into quartiles according to mutation number 
 Quartile 1 Quartiles 2-4 p-value 
R:S in CDR and FR could not be calculated for those patients with no mutations in the corresponding region. 
Number of Mutations 1–9 10–36  
Hazard Ratio: TUT 2.37  <0.0001 
Hazard Ratio: OS 2.90  0.0003 
Region n Pts Mean R:S ratio n Pts Mean R:S ratio  
CDR + FR 94 2.81 259 2.55 not signif 
FR 80 2.44 259 2.18 not signif 
CDR 66 3.25 258 5.64 0.0008 
Patients divided into quartiles according to mutation number 
 Quartile 1 Quartiles 2-4 p-value 
R:S in CDR and FR could not be calculated for those patients with no mutations in the corresponding region. 
Number of Mutations 1–9 10–36  
Hazard Ratio: TUT 2.37  <0.0001 
Hazard Ratio: OS 2.90  0.0003 
Region n Pts Mean R:S ratio n Pts Mean R:S ratio  
CDR + FR 94 2.81 259 2.55 not signif 
FR 80 2.44 259 2.18 not signif 
CDR 66 3.25 258 5.64 0.0008 

Disclosures: No relevant conflicts of interest to declare.

Author notes

Corresponding author