[ 1 ]Received: 16 December 2016 | Accepted: 19 December 2016
To the Editor:
Although patients with nontransfusion-dependent thalassemia (NTDT) are not dependent on frequent transfusions, these patients are at risk of iron overload, predominantly because of increased intestinal iron [ 2 ] absorption.1 With the wide array of serious morbidities resulting from iron overload, the diagnosis, follow-up, and timely management with iron chelation therapy become of significant importance.2,3 The assess- ment of the liver iron concentration (LIC) by magnetic resonance imag- ing (MRI) remains the recommended means of quantifying iron overload in patients with NTDT to guide iron chelation decisions.4 With the high prevalence of NTDT in resource-poor countries, MRI technology may still not be readily available to a significant proportion of patients with NTDT. Serum ferritin a simple and more affordable tool, has a significant positive correlation with LIC from observational studies and may be used in countries where MRI is not available or affordable.4 In earlier studies, 800 ng/mL were shown to be associated with clinical morbidity in NTDT and to almost invariably predict an LIC >5 mg/g dry weight (dw), the level that is associated with iron-related morbidity at which iron chelation therapy is recommended">.2-6 However, in the current study, close to 50% of patients with serum ferritin levels <800 ng/mL also continue to have LIC levels >5 mg/g dw and thus remain at risk of being denied iron chelation therapy although they need it5 Accordingly, for patients with serum ferritin levels between 300 (upper limit of normal) and 800 ng/mL, it is still recommended that LIC assessment by MRI be used to guide management of chelation therapy and ensure no patients with LIC >5 mg/g dw are missed.7 In this study, we aim at providing the cli- nician, who lacks access to LIC assessment by MRI, with a tool that could facilitate clinical decision making for this unique group of patients with serum ferritin levels of 300 to 800 ng/mL To this aim and in a cross-sectional design, we evaluated 71 patients with serum ferritin levels ranging between 300 and 800 ng/mL attending centers in Italy, Lebanon, Oman, and Thailand. Data was collected as part of now com- pleted clinical studies, which were approved by the Institutional Review Board at all four institutions. All patients had signed an informed con- sent form for participating in the original studies in accordance with the Declaration of Helsinki. Table 1 shows the baseline characteristics of the patients. Their mean age was 33.3 ± 13.9 years (range: 10-71); 53.5% were female and 45.1% were splenectomized. A total of 37 (52.1%) patients in fact had a LIC >5 mg/g dw. Sixty-one of the patients (85.9%) had one of the molecular variants of ß-thalassemia intermedia, three patients (4.2%) had one of the molecular variants of «-thalassemia intermedia, four patients (5.6%) had a molecular diagno- sis of Hemoglobin E/ß-thalassemia, and three patients (4.2%) did not carry a definitive molecular diagnosis. We performed logistic regression analysis with LIC >5 mg/g dw as the dependent variable we estimated FIG U R E 1 Probability curve of LIC > 5 mg/g dw as a function of serum ferritin (ng/mL) when serum ferritin is between 300 and 800 ng/mL in patients with NTDT (LIC: liver iron concentration, dw: dry weight, NTDT: nontransfusion-dependent thalassemia) a probability curve for serum ferritin levels in predicting an LIC >5 mg/g dw (Figure 1). The probability of having a LIC >5 mg/g dw was calculated from serum ferritin using the formula: probability of LIC >5 mg/g dw = 1/(1 + e-[logistic regression z’score]), the z-score from logistic regression being (3.62 - 0.002 * serum ferritin) and using a Lowess smoothing function. To check the validity of the model, a negative binomial regression model showed a correlation between serum ferritin and LIC with an incidence rate ratio of 1.002 (confi- dence interval [1.001 - 1.004] with a P < .001) In addition, we applied the Delong method to check if the addition of age, sex, and splenectomy status improved the logistic regression model. There was tie, statistically significant improvement in the model (P = .526). Instead of categorically assuming patients with serum ferritin levels <800 ng/mL are not at morbidity risk and eligible for chelation, while a good majority in fact will be, our probability curve helps assign a percentage chance of still having LIC >5 mg/g dw for every serum ferritin level between 300 and 800 ng/mL. Based on the probability curve and on other clinical or laboratory factors suggesting an iron overload state, the clinician should be able to make a better informed decision when MRI for assessment of LIC is not available. This curve is applicable to a pool of patients with NTDT with the same admix- ture of thalassemia syndromes found above; further evaluations in other more specific patient subgroups may be needed. [ 3 ]====Correspondence====
Ali T. Taher, MD, PhD, FRCP, Professor of Medicine, Hematology & Oncology, Associate Chair - Research, Department of Internal Medicine, American University of Beirut Medical Center, P.O. Box 11-0236, Beirut 11072020, Lebanon.
AUTHORSHIP
Study design: ANS, KMM, MDC, ATT. Data preparation: ANS, MDC, SD, W. Data analysis: KMM, ANS. Analysis review and manuscript preparation: ANS, KMM, MDC, GG, SD, W, ATT. All authors gave final approval for submission.
ROLE OF THE FUNDING SOURCE
None.
Antoine N. Saliba1, Khaled M. Musallam2,3, Maria D. Cappellini3, Giovanna Graziadei3, Shahina Daar4, Vip Viprakasit5, Ali T. Taher2 1 Department of Medicine, Indiana University School of Medicine, Indianapolis, Indiana, United States 2 Department of Internal Medicine, American University of Beirut Medical Center, Beirut, Lebanon department of Clinical Sciences and Community, University of Milan, IRCCS Co' Granda Foundation Maggiore Policlinico Hospital, Milan, Italy department of Hematology, College of Medicine & Health Sciences, Sultan Qaboos University, Oman 5Department of Pediatrics & Thalassemia Center, Faculty of Medicine Siriraj Hospital, Mahidol University, Bangkok, Thailand
REFERENCES
[1] Musallam KM, Rivella S, Vichinsky E, Rachmilewitz EA Non-transfu-
sion-dependent thalassemias. Haematologica. 2013;98:833-844.
[2] Musallam KM, Cappellini MD, Daar S, et al. Serum ferritin level and
morbidity risk in transfusion-independent patients with beta-
thalassemia intermedia: the ORIENT study. Haematologica. 2014;99:
e218-e221.
[3] Musallam KM, Cappellini MD, Wood JC, et al. Elevated liver iron con-
centration is a marker of increased morbidity in patients with beta tha-
lassemia intermedia. Haematologica. 2011;96:1605-1612.
[4] Taher A Vichinsky E, Musallam K, Cappellini MD, Viprakasit V.
Guidelines for the Management of Non Transfusion Dependent Tha-
lassaemia (NTDT). In: Weatherall D, ed. Thalassaemia International
Federation (c) 2013. Nicosia, Cyprus: Thalassaemia International Fed-
eration, 2013.
[5] Taher AT, Porter JB, Viprakasit V, et al. Defining serum ferritin thresh-
olds to predict clinically relevant liver iron concentrations for guiding
deferasirox therapy when MRI is unavailable in patients with non-
transfusion-dependent thalassaemia. BrJ Haematol. 2015;168:284-290.
[6] Musallam KM, Cappellini MD, Taher AT. Evaluation of the 5mg/g
liver iron concentration threshold and its association with morbidity
in patients with beta-thalassemia intermedia. Blood Cells Mol Dis.
2013;51:35-38.
[7] Taher AT, Viprakasit V, Musallam KM, Cappellini MD. Treating iron
overload in patients with non-transfusion-dependent thalassemia. Am
J Hematol. 2013;88:409-415.