Correlation Between Peripheral Blood Counts and Day 14 Bone Marrow Biopsy in Acute Myeloid Leukemia during Induction Chemotherapy

Background

Current NCCN guideline recommends that a bone marrow sample be performed 7-10 days (day 14 bone marrow) after completion of induction therapy in newly diagnosed acute myeloid leukemia (AML). However, the value of day 14 bone marrow has been questioned due to the invasive nature of the procedure and lack of specificity pertaining to complete remission in cases of borderline blasts count and cellularity. We examined peripheral blood count and bone marrow from day 0 to day 14, to see if a reduction of peripheral blood correlated and predicted the day 14 bone marrow morphologic changes and complete remission (mCR).

Methods

We did 10 years retrospective review between year 2004 and 2013 at the Henry Ford Hospital, on patients who had newly diagnosed AML and day 14 bone marrow biopsy. The majority of patients underwent "7+3" or a "7+3"-like regimen for induction chemotherapy. Firstly, we evaluated the relationship of change of peripheral blood count from day 0 to day 14 with blast percentage and cellularity of bone marrow. Spearman correlations coefficients were computed for each pair of characteristics. Peripheral blood count includes neutrophil (ANC), monocyte, white blood cells (WBC), blast, hemoglobin and platelet. Secondly, we investigated the possible correlation of mCR to peripheral blood and bone marrow changes, using binary univariate logistic regression. mCR as defined by blast percentage <5, absolute neutrophil (ANC) >1000/mm3, platelets>100,000/mm3. Thirdly, we explored differences in peripheral blood counts on day 14 among three bone marrow groups, those with blast percentage <5, 5-20, >20.

Results

A total of 200 patients were reviewed and 56 patients met the inclusion criteria. Decrease of ANC/WBC correlated with decrease of bone marrow blast/cellularity from day 0 to day 14 (ANC: Blast P ≤ 0.05; ANC: cellularity P ≤ 0.05; WBC: blast P ≤ 0.001, WBC: cellularity P ≤ 0.01). In other words, a larger reduction in ANC/WBC correlated with larger reduction in both blast and cellularity in bone marrow. However, this correlation with bone marrow change was not found in peripheral blast, monocyte, hemoglobin and platelet. We also found that with increasing age, there was less reduction from day 0 to day 14 in bone marrow blast and cellularity.

Bone marrow blast and cellularity on day 14 is strongly associated with mCR (P<0.01), the reduction of blast (43.7 +/- 22.86, Odds ratio 1.03 (1.01, 1.06), P=0.012) and cellularity (66.21 +/- 29.98, Odds ratio 1.03 (1.01, 1.05), P=0.003) from day 0 to 14 is also predictive for mCR. Interestingly, there is a trending effect that the reduction of ANC from day 0 to 14 may predict mCR, but it is not statistically significant (Odds ratio 1.22 (1.02, 1.66), P=0.097). The reduction of WBC is not associated with mCR.

Furthermore, peripheral blood counts on day 14 are similar among 3 bone marrow groups, those of blast percentage <5, 5-20, and >20% on day 14.

Conclusion

ANC/WBC decrease from day 0 to day 14 correlated with the decrease in bone marrow blast count and cellularity, and can be used as a predictor for bone marrow change on day 14, but the level of day 14 peripheral blood findings are similar among 3 bone marrow groups (blast percentage <5, 5-20, and >20% on day 14), so it could not be used to predict the level of bone marrow change. Our data confirmed that the significant decrease of bone marrow blast percentage and cellularity from day 0 to 14 predicts mCR. Decrease of ANC from day 0 to 14 may also predict mCR although it is not statistically significant. A larger sample size can be studied in the future to further explore the possibility of using peripheral blood to predict bone marrow changes and mCR.

Summary

Our data demonstrates a significant reduction of ANC on day 14 after induction therapy in newly diagnosed AML, which correlates with a decrease in bone marrow cellularity and blast percentage. However, a statistically significant association with blast percentage pertaining to mCR was not obtained. In conclusion, while the current findings do not justify replacement of day 14 bone marrow for predicting mCR, further large scale studies are indicated.