Increased Bone Marrow Iron at Diagnosis Is Associated with Inferior Prognosis in Patients with Myelodysplastic Syndromes

Introduction: Iron storage in patients (pts) with myelodysplastic syndromes at the time of diagnosis may vary from normal to iron overload. Even before the first blood transfusion, storage iron can be increased due to down-regulation of hepcidin and subsequent increase in duodenal iron uptake. Iron overload is known to worsen the prognosis of MDS patients, partly due to iron-related organ damage after long-term transfusion therapy, and partly due to an increased risk of infections. However, it is unclear whether increased storage iron at the time of diagnosis already has a prognostic influence. We assessed bone marrow iron stores at the time of MDS diagnosis and correlated them with clinical outcome.

Methods: In a retrospective analysis of 3762 adult MDS patients from the Düsseldorf MDS Registry, Prussian blue staining of marrow smears was performed in our cytology lab to assess iron stores according to the following categories: normal or decreased iron stores versus increased iron stores versus iron overload. Patients were followed up for survival and AML evolution until June 2021. Median time of follow-up was 20 months. 67.4% of the patients died during the course of the disease.

Results: The study included 3.762 adult patients who received their initial diagnosis of MDS between 1970 and 2021. 58% were diagnosed as non-blastic MDS ( MDS SLD (RS) (n=240), MDS MLD (RS) (n=350), MDSdel(5q) (n=107), and MDS-U (n=25). Iron stores were decreased in 8% of the patients, normal in 44%, increased in 41%, and strongly increased in 7% (massive iron overload).

In 282 cases, histologic assessment of storage iron was available. When comparing cytologic and histologic assessment, we found a strong correlation (p<0.0005), since 87% of the patients with increased iron on cytomorphology also showed increased iron as assessed by histopathology. However, 37% of the patients who cytologically showed normal iron stores, were reported to have slightly increased iron as assessed by histopathology. Median and mean serum ferritin values of patients with normal or decreased iron stores were 295 and 629 µg/l, respectively, as compared to 548 and 902 µg/l, respectively, in patients with increased iron stores.

The cumulative risk of AML evolution was not associated with the results of iron staining. Regarding survival, we found that patients with decreased or normal storage iron had a median survival of 31 months, whereas those with increased iron had a median survival of 28 months (p=0.007). Focusing on patients with non-blastic MDS, the difference was not significant (46 vs 44 ms). However, patients who presented as EB I (n=435), EBII (n=510), AML MRC (n=264), CMML I (n=254), or CMML II (n=77), showed a prognostic impact of storage iron; patients with increased iron had a median survival of 11 months, as compared to 16 months in patients with normal or decreased iron (p<0.0005).

Conclusion: Increased tissue iron in the bone marrow at the time of diagnosis is associated with inferior survival in patients with MDS, primarily in patients with higher risk MDS. At diagnosis, patients are not yet transfusion-dependent. This suggests that increased iron reflects a prolonged period of increased duodenal iron uptake as a consequence of ineffective erythropoiesis. Therefore, increased marrow iron at the time of MDS diagnosis seems to be a surrogate parameter of hematopoietic insufficiency, which is the real cause of inferior prognosis.