Macrocytosis and Iron Stores in Patients with Chronic Myeloid Leukemia Being Treated with Tyrosine Kinase Inhibitors.

Since imatinib was approved in 2001, tyrosine kinase inhibitors (TKIs) have become standard of care for first line treatment of Chronic Myeloid Leukemia (CML). Macrocytosis has been observed in some patients being treated with imatinib but the etiology of this phenomenon is unclear. Altered DNA metabolism resulting from the inhibition of c-kit by imatinib may be an explanation but anecdotal correction of macrocytic anemia after iron replacement has prompted us to explore this further. The correlation between bone marrow iron stores, anemia and macrocytosis has not been extensively studied in this population.

The Total Cancer Care (TCC) database was used to identify all patients treated for CML at Moffitt cancer Center (MCC) between 1992 and 2010. All patients treated with TKI's were evaluated. Bone marrow iron stores, macrocytosis and anemia were recorded from the first records available at MCC while patients were taking TKI's. Iron stores, anemia and macrocytosis were recorded as categorical variables, while hemoglobin was recorded as a continuous variable. Descriptive data were reported and chi square test was used for categorical variables. Statistical analysis was done using SPSS statistical software, version 19.

A total of 540 patients were treated for CML at MCC between 1992 and 2010, and 478 received treatment with TKI's. Of these, 53% were male, 47% were female and 70% were under the age of 60 at diagnosis. At the time of data cutoff 68% were still living. Imatinib was the first TKI used in 470 (98.3%) patients, while nilotinib and dasatnib were first line treatment in 6 (1.3%) and 2 (0.4%) patients respectively. Second line TKI's were necessary in 177 (37%) patients.

Anemia was present in 274 (57%) of the 478 patients treated with TKI's. Bone marrow iron stores were increased, adequate or decreased/absent in 38 (8%), 85 (18%) and 289 (60%) respectively. Reports on iron stores were missing in 66 (14%) patients. One hundred eighty nine (40%) patients had macrocytosis compared with 278 (58%) who did not. This data was missing in 11 (2%) patients.

Data correlating iron stores and anemia were available on 410 patients (Table 1). From them, 178 (43%) patients were anemic. Of these, 34 (19%) had adequate iron stores, 131 (74%) had decreased/absent iron stores and 13 (7%) had increased iron stores. There were 232 patients with available data on iron stores who were not anemia. In this group, iron stores were adequate, decreased/absent and increased in 51 (22%), 156 (67%) and 25 (16%) patients respectively.

Data correlating anemia with macrocytosis were available in 467 patients (Table 2) and in this assessment 270 patients were anemic. One hundred nineteen (44%) had a macrocytic anemia while 151 (56%) did not. Of the 197 patients in this evaluation who were not anemic, 70 (36%) had a macrocytosis while 127 (64%) did not.

Lastly, an association between iron stores and macrocytosis was looked at, and there were 407 evaluable patients in this group (Table 3). Macrocytosis was noted in 239 patients and of these, bone marrow iron stores were adequate, decreased/absent, and increased in 32 (19%), 120 (71%) and 16 (10%) patients respectively. Of the 168 patients without macrocytosis, iron stores were adequate in 53 (22%) patients, decreased/absent in 164 (69%) patients and increased in 22 (9%) patients.

Macrocytic anemia is most often caused by an alteration in DNA metabolism due to nutritional deficiencies or medications. Contrary to what would be expected, out data show that over two-thirds of the patients with macrocytosis had decreased iron stores in their bone marrow. Similarly, nearly half of all anemic patients had a macrocytic anemia which raises the question of whether iron deficiency really is the primary etiology of their anemia. Inhibition of kinases involved in iron metabolism may be involved in the etiology of the macrocytosis seen in these patients.

No relevant conflicts of interest to declare.