Evaluation of beta Globin mRNA as an Early Marker of Hb Response to Epoetin Treatment.

Background

Anaemia is a very common complication of cancer, and treatment with erythropoietic stimulating agents (ESA) is an important part of the supportive care. However, only about 60% of cancer patients respond to ESA treatment. An early marker of ESA response would be very valuable in order to avoid ineffective treatment. So far, predictors and early markers of response are unreliable in individual patients. We have previously shown that beta globin mRNA increases rapidly after epoetin beta treatment of healthy controls. In the present study we have evaluated whether the level of expression of this marker during the first two weeks of ESA treatment could predict later Hb response in patients with anemia of cancer.

Patients and methods

Twenty patients with malignant disease and Hb <10.5 g/dl, who had no change in anti-tumoral therapy within 6 weeks before entering or during the study, were treated with epoetin beta (NeoRecormon®) 10000 IU s.c. three times per week during 6 weeks. Other causes of anaemia than the cancer itself were excluded by normal levels of S-B12, S-folate, S-ferritin, S-haptoglobin, S-Fe, S-Transferrin (Tf) and Tf saturation. Microcytosis was not allowed and two F-Hb samples were negative. The diagnoses included 4 biliary, 5 pancreatic, 3 gynecologic, 2 breast-, 1 gastric and 1 prostate cancer, 3 multiple myelomas and 3 myelodysplastic syndrome patients.

Samples were then taken twice weekly during the first two weeks of treatment, once weekly during week 3–6 for measurement of Hb, reticulocytes and β-globin mRNA, which was measured quantitatively using PCR via the 5′ nuclease assay.

Results

Eleven patients responded with a Hb increase of >1 g/dL, 9 were nonresponders. All responders increased in beta globin mRNA within two weeks, mean 7.7 × baseline (range 1.4–13.8). However, 6/9 non-responders also increased (2.2–8 × baseline). There was no correlation between the degree of Hb elevation and increase in beta globin mRNA. With a cut-off of an increase of 3 times the baseline value, 1 responder was missed and there were 5 false positives, which means a specificity of 45% and a sensitivity of 91% for the prediction of a later increase of Hb> 1g/dL. If the cut-off was set at 4 × baseline, the specificity increased to 66%, but the sensitivity decreased to 82%.

The reticulocyte count increased in 7/11 responders and in 3/9 non-responders. There was no correlation between the Hb increase and the elevation in reticulocytes and no correlation between increase in beta globin mRNA and reticulocytes.

Conclusions

Beta globin mRNA increases before Hb after epoetin treatment in all responding patients. However, some non-responding patients also show an increase, and there is a trade-off between specificity and sensitivity as the cut-off level for prediction is set at different levels. Compared to reticulocyte count, beta globin mRNA is more reliable in the individual patient, but the clinical usefulness of the assay needs to be evaluated in larger studies.