Background

Consensus has been achieved that standardized molecular quantitative analysis (RQ-PCR) on peripheral blood (PB) is a suitable method for monitoring residual disease in chronic myeloid leukemia (CML). However, BM is still obtained at specific timepoints, and in a number of cases, only bone marrow (BM) sample collected for cytogenetic analysis is available. Being one of the laboratory involved in the standardization process of molecular monitoring for CML patients, we decided to perform a comparative analysis of BM and PB samples in order to evaluate the consistency of the results.

Methods

Between March 2009 and January 2013, 230 consecutive RQ-PCR tests to assess BCR-ABL transcript levels from simultaneously collected PB and BM samples were performed (for a total of 460 analysis) on 77 patients affected by Ph+ CML in chronic phase treated in our center. All samples were analyzed in the same laboratory following international guidelines (Cross N, Leukemia 2012) and results were expressed according to the International Scale; ABL1 was used as control gene. Time from blood-drawn to processing was within 3-4 hours.

Results

Among the 230 pairs, 3 were considered as not evaluable because of inadequate material; for the purpose of this study, the remaining 227 pairs were considered as “evaluable”. 204 pairs were classified as “fit” when both BM and PB ABL amplification resulted in more than 10.000 copies; 23 pairs were considered unfit for ABL1 <10.000 in either one of the two samples (21) or both (2). The mean number of ABL1 copies in all evaluable samples was 35.639 for BM (SD 21.465) and 30.958 for PB samples (SD 18.696). Correlation analysis was performed on the whole population and in 4 subgroups: No Complete Cytogenetic Response (CCyR, 22%), CCyR without Major Molecular Response (MMR), (21.6%), CCyR with MMR (excluding patients with MR4 or better,19.8%), and CCyR with MR4 – MR4.5 (32,6%). Cytogenetic response was not available in 9 BM samples (4%), not included in the subgroup analysis. Spearman correlation of BCR/ABL ratio values between PB versus BM paired samples resulted in a statistically significant correlation in all groups, both for evaluable and fit pairs. Correlation was stronger in samples that were not in MMR or better (table 1 and figure 1). The Wilcoxon test showed that the mean difference of BCR/ABL values between paired PB and BM samples was not significantly different from zero (in evaluable and fit pairs by considering the whole population). Concordance was further analyzed by the K test which resulted in a coefficient equal to 0.627, corresponding to a notable degree of concordance. For patients in CCyR, agreement on classification of response (MMR, MR4, MR4.5) between paired PB and BM samples was observed in 125/168 evaluable pairs; 22 out of the 43 evaluable cases of disagreement were due to technical failures (in 10 BM and 12 PB samples). In 14 of the remaining 21 cases, PB was more sensitive.

Table 1

Spearman correlation of % is between paired PB and BM samples / fit samples

 CoefficientP-value
%IS   Total +0.966 <0.001 
%IS – Subgroup No CCyR +0.921 <0,001 
%IS – Subgroup CCyR-No MMR +0.752 <0.001 
%IS – Subgroup CCyR-MMR +0.498 0.003 
%IS – Subgroup: CCyR-MR4, MR4.5 +0.478 <0.001 
 CoefficientP-value
%IS   Total +0.966 <0.001 
%IS – Subgroup No CCyR +0.921 <0,001 
%IS – Subgroup CCyR-No MMR +0.752 <0.001 
%IS – Subgroup CCyR-MMR +0.498 0.003 
%IS – Subgroup: CCyR-MR4, MR4.5 +0.478 <0.001 
Figure 1

Scatter plot of % is between paired PB and BM samples by cytogenetic response (CCyR) / evaluable pairs

Figure 1

Scatter plot of % is between paired PB and BM samples by cytogenetic response (CCyR) / evaluable pairs

Close modal
Conclusions

In a single center experience of molecular analysis, BCR/ABL ratio was highly consistent in BM and PB samples. In less than 10% of the cases a single test did not reach the required sensitivity of 10.000 ABL copies and the double testing allowed to obtain a valid result. This may be especially valuable in evaluating an early response (i.e. at 3 months), when the amount of disease has prognostic relevance. The analysis will be expanded to include samples coming from different centers to evaluate a possible role of timing and transport on data consistency.

Disclosures:

Saglio:Novartis: Consultancy, Honoraria; Bristol Myers Squibb: Consultancy, Honoraria; ARIAD: Consultancy, Honoraria; Celgene: Consultancy, Honoraria.

Author notes

*

Asterisk with author names denotes non-ASH members.

Sign in via your Institution