Fr. Dick et al., BONE-MARROW MORPHOLOGY DURING INDUCTION-PHASE OF THERAPY FOR ACUTE MYELOID-LEUKEMIA (AML), Hematologic pathology, 9(2), 1995, pp. 95-106
Sequential bone marrow aspirates and sections from patients with acute
myeloid leukemia (AML) were examined to determine ifa correlation exi
sts between bone marrow morphology during induction phase of therapy a
nd outcome. Of 95 patients of AML diagnosed between July 1987 and Dece
mber 1991, 53 uniformly treated patients (induction therapy with cytos
ine arabinoside and daunorubicin) had sequential bone marrow examinati
ons performed in the 2- to 5-week period following initiation of induc
tion therapy. Four morphologic patterns were recognized in these 53 pa
tients: Group I (22 patients)-hypocellularity or normal regeneration (
greater than or equal to 15% cellularity and <5% blasts) on the initia
l 2-week marrow followed by marrows showing normal regeneration; Group
II (10 patients)-hypocellularity followed by ''reactive myeloblastosi
s'' (greater than or equal to 15% cellularity, 5% to 34% blasts, with
promyelocytes approximate to or > blasts); Group III (12 patients)-res
idual blasts (greater than or equal to 5% blasts with blasts much grea
ter than promyelocytes) in the initial posttherapy marrow; Group IV (9
patients)-atypical patterns not fitting any of the other categories.
Complete remission was achieved in all 32 patients in Groups I and II
without additional induction therapy but was achieved eventually in on
ly 10 of 21 patients in Groups III and IV combined (p <0.005), 15 of w
hom received additional induction therapy Remission duration and actua
rial survival for each group were as follows: Group I: 344/596 days; G
roup II: 443 days/>660 days, Group III and IV combined: 351/311 days (
p = 0.017 for actuarial survival). Seven of 21 patients in Groups III
and IV had unfavorable initial morphology (MO, hypoplastic AML and AML
preceded by myelodysplasia) compared to only 3 of 32 patients in Grou
ps I and II (p = 0.039). It was thus observed that ''reactive myelobla
stosis'' with lip to 34% blasts on the third or fourth week bone marro
w following an initial hypocellular marrow does not require additional
induction therapy to achieve durable remissions or favorable survival
. Also, residual blasts that outnumber promyelocytes, and atypical pat
terns of regeneration correlate with lower remission induction rates,
shortened survival, and unfavorable morphology on the initial diagnost
ic bone marrow.