To the Editor

Recently, Makhoul et al. (2005) published a paper on the genetic heterogeneity of β-thalassemia in Lebanon. In this article, the authors refer to a previous study published by our group (Zahed et al. 1997), stating that no notable religious and/or regional specificities had been identified in our study and that our results were based on the analysis of a fairly small number of subjects. While it is true that our study only included 110 carriers of β-thalassemia, it did however suggest that the disorder was most frequent among the Sunni Muslims, followed by the Shiite Muslims and the Maronites. It also noted that the molecular basis of the disease was most heterogeneous among the Sunni Muslims. Similarly, it did recognize that some mutations had a particular geographic distribution; for example, that cd29 was mainly found in the Bekaa valley. More importantly, the authors fail to refer to another wider study published by our group (Zahed et al. 2000), in which a total of 277 β-thalassemia chromosomes were included with haplotype analysis performed on 108 of these. This is surprising, since the authors refer to a later article (Zahed et al. 2002) in which the study published in 2000 was mentioned repeatedly. The findings which we published in 2000 were extremely similar to those of Makhoul et al. This is not surprising, since both studies are based on the analysis of the same population, namely patients or obligate carriers presenting at the Chronic Care Center, a centre dedicated to the treatment and prevention of β-thalassemia in Lebanon. In our study published in 2000, in order to evaluate the actual frequency of the different mutations, those in homozygotes from consanguineous marriages (62% of cases) were counted as one mutation. Just like the study of Makhoul et al., we had found that the majority of patients (72%) were Muslims, and that the Sunni Muslims were the most heterogeneous at the molecular level. While the most common mutations were found in all religious groups, we also found that IVSI-5 (G > C) and cd 8 (−AA) were exclusive to the Sunnis. Similarly, we also reported that two mutations were exclusive to the Shiites, namely cd30 (G > C) and cd 29 (C > T). Analysis of the distribution of mutations by geographic region in our study also showed that cd29 seems to be clustered in the Bekaa valley, while cd30 was found in Southern Lebanon. Haplotype analysis of the mutant β-globin genes also revealed similar findings in both studies. However, in our study we identified a rare 5′ subhaplotype (haplotype 12, previously reported in South Africa and Asia) linked to IVSI-110 in three unrelated Maronite individuals. In addition, Makhoul et al. make some incorrect statements in their article. They state that, according to the syllabus of thalassemia mutations (Huisman et al. 1997), 300 different mutations in the β-globin gene had been described. However, this number actually refers to all thalassemia alleles and Hereditary Persistance of Fetal Hemoglobin (HPFH) determinants, and not only to β-globin mutations (E. Baysal personal communication). The authors also erroneously mentioned that “β-thalassemia is the most predominant genetic defect in Lebanon”. The total number of patients with β-thalassemia registered at the Chronic Care Center is around 550, which more or less represents the great majority of Lebanese β-thalassemia patients residing in the country. Although accurate statistics and frequencies are currently lacking in Lebanon, other autosomal recessive hereditary disorders, such as Familial Mediterranean Fever (over 800 patients on record so far) or non-syndromic hearing loss due to mutations in the GJB2 gene (carrier frequency estimated to be 1/40), have a higher frequency (A. Megarbane personal communication). The authors also state, referring to our study of 1997, that the frequency of carriers of β-thalassemia is highly variable among religious groups, the highest frequency being in the Sunnis. However, what we refer to in our study is not the frequency of carriers of β-thalassemia, but the distribution of affected individuals among different communities, two notions which are nevertheless different. To our knowledge, to date there is no