Transplantation in thalassemia: Revisiting the Pesaro risk factors 25 years later

During the last few years, we have witnessed extraordinary progress in our knowledge of iron pathophysiology, as well as in our understanding of iron toxicity and iron overload. This formula proposes that tissue toxicity arises from both the quantity of toxic iron species (i.e., tissue reactive species = non-transferrin bound iron (NTBI), labile plasma iron (LPI), and enhanced LPI)2 and the duration of exposure. The detrimental effects are further modulated by the individual's anti-oxidant genetics and by environmental factors such as dietary antioxidants, or other metals like copper, or selenium. In the late eighties, Guido Lucarelli and the Pesaro group developed a well recognized scoring system to predict the outcome of allogeneic hematopoietic stem cell transplantation in thalassemia, which is a paradigm for iron overload diseases.3-6 This scoring system stratified patients <16 years old into three categories of risk factors, resulting in statistically significant and clinically remarkable differences in outcome. There were three statistically significant risk factors, that is, hepatomegaly (≤2 cm versus >2 cm from the costal arch), liver fibrosis (absent versus present at any degree), and duration of exposure to iron chelation (Table 1). This last parameter, in the deferoxamine-only era, was defined as life-long deferoxamine, starting no later than 18 months after the first transfusion and administered regularly as an 8-hour continuous infusion for at least 5 days/week. The overall survival and thalassemia-free survival were 94% and 87%, 84% and 81%, and 50% and 47% in Class 1, 2, and 3, respectively. When the same conditioning regimen was used, the difference was almost entirely due to transplant-related mortality (6, 15, and 47% in Class 1, 2, and 3, respectively; P < .0001). These three risk factors were clearly related to iron overload, but no other conventional direct or indirect iron overload related factors (i.e., age, transfusions, serum ferritin, liver iron concentration, splenomegaly, splenectomy, chronic hepatitis, etc.) achieved statistical significance by univariate and multivariate analyses.3, 6 Moreover, this scoring system did not predict risk in adult patients,7 and received criticism for its nonquantitative nature and for the susceptibility to intraobserver and interobserver variability, particularly regarding hepatomegaly and the definition of adequate chelation. Moreover, several colleagues questioned the absence of impact by factors directly related to iron overload (i.e., serum ferritin and liver iron concentration). Despite these criticisms, this scoring system has notably influenced transplantation strategy for the last 25 years.8 What is the relationship between the iron toxicity formula mentioned above1 and the Pesaro scoring system? If we closely analyze the Pesaro score under the magnifying glass of this formula, it is evident that all the three risk factors were not quantitative direct markers of iron overload “per se”, but indirect measures of intensity and extent of tissue exposure to toxic iron. Adequacy of chelation clearly means consistent and sufficient suppression of tissue reactive iron species (NTBI/LPI) over time, based on the practice applied 25 years ago. Liver fibrosis is definitely a marker of toxic iron exposure and environmental factors (i.e., viral infections) in the liver. Hepatomegaly reflects the extent of iron deposition and the time-averaged exposure to transferrin bound iron and toxic reactive iron9 as loading of the liver significantly increases when NTBI/LPI enters through ion channels and transporters in addition to the regulated entry through transferrin receptor-1 and 2-mediated mechanisms normally operative in the liver. Thus, not only the magnitude of iron overload, but the duration of exposition to toxic iron is important. Based on the discussion above, it is easy to understand why despite several studies no definitive conclusion has been reached regarding the relation between transplant-related mortality/morbidity and iron outside the field of thalassemia.11-15 In all the studies, only the correlation between direct or indirect estimates of iron overload (mainly serum ferritin) and outcome parameters was explored while the duration of exposure to toxic iron species was not taken into account. Of course, this relation would be much more difficult to study in disorders like leukemia that are outside the homogenous field of thalassemia, and where factors like neoplastic disease confound the outcomes. Reinterpreting transplant predictive factors in the light of the current advances in understanding of iron homeostasis further supports the concept that the key to successful transplantation in thalassemia is regular and life-long chelation therapy to consistently suppress tissue reactive iron species and prevent tissue damage. Because the ongoing trials on gene therapy in thalassemia are performed using autologous transplantation, it is likely that the same concept could be relevant to the gene therapy approach. Moreover, the role of toxic iron should be considered in all transplantations in general as most patients who undergo marrow transplantation have had significant transfusion exposure before and during the procedure.11-15 However, based on our understanding of iron toxicity,1, 9 chelation of toxic iron could be done throughout transplantation to quickly reduce NTBI/LPI and block the amplifying effect of reactive forms of iron on inflammation.18 Finally, Guido Lucarelli should be recognized for his clinical observations and establishment of risk factors for transplantation in thalassemia as they are now clearly supported 25 years later by the huge progress in iron pathophysiology understanding. The authors would like to acknowledge Martine Torres, PhD for her editorial assistance.