Changing patterns of splenectomy in transfusion‐dependent thalassemia patients

The need of splenectomy in thalassemia major is more likely where the disease is not suppressed efficiently by transfusion treatment. The aim of this report has been to evaluate the proportion of patients for whom splenectomy has been avoided or delayed in a large cohort of thalassemic patients during a 40-year span. A series of 872 regularly transfused β thalassemia patients born between 1960 and 1999 was pooled from the records of WebThal, a thalassemia dedicated software in five Italian Centers. For each patient's date of birth, first transfusion and splenectomy were considered. Kaplan–Meier and Wilcoxon tests for group comparison were applied. Age at splenectomy and date of splenectomy correlated positively (r = 0.73, P < 0.001). The probability to undergo surgery within the first 10 years of life was 57, 22, 6, and 7%, respectively, for patients born in the 1960s, 1970s, 1980s, and 1990s. For thalassemic patients on standard treatment, the chance to be splenectomized is today low during childhood and young adulthood. Further studies are needed to quantify the specific contribution of the presence of the spleen to the prolonged survival and quality of life in well-treated thalassemic patients. Splenomegaly and hypersplenism are common clinical features of thalassemia major. Splenectomy is the recommended intervention to reduce excessive blood consumption and consequent severe iron overload. However, many complications are associated with the absence of spleen. The risk of infection, most commonly sepsis from encapsulated organisms (Streptococcus pneumonia, Haemophilus influenzae and Neisseria meningitides), varies among studies, from more then 30-fold in comparison with normal population [1] to a smaller but yet relevant risk in more recent studies [2]. Other negative effects of splenectomy include a raised risk of thrombotic complications and pulmonary hypertension, possibly due to the raised number of circulating platelets and immature red blood cells with alteration of the endothelial function, enhanced platelet activation, decreased levels of proteins C and S [3, 4]. A progressive increase of awareness of these negative effects may have contributed to lower the tendency to perform splenectomy in thalassemia. The aim of this report has been to evaluate the proportion of patients for whom splenectomy has been avoided or delayed in a large cohort of thalassemic patients during a 40-year span. From 1,055 patient records, 162 (15.4%) and 21 (2%) were excluded for incomplete data and first transfusion after the age of 6 years, respectively. The final population consisted of 872 patients and 20,181 years of observation, with 19,076 years of regular transfusion (94.5%). Mean age was 23.1 ± 7.1 years (range 1.6–41.5; Supporting Information graph). Mean age at first transfusion was 1.3 ± 1.2 years (range 0.1–5.9; Supporting Information graph). Mean pretransfusion hemoglobin levels have been 72 g/L during the 1960s, 86 g/L in the 1970s, 97 g/L in the 1980s, and 98 g/L in the 1990s. Two hundred and eighty-four patients of 872 (32.6%) had been splenectomized, at a mean age of 10.0 ± 5.1 (range 1.7–31.1; Supporting Information graph). We found a positive correlation between the age at splenectomy and the date of splenectomy (r = 0.73, P < 0.001; Fig. 1), whereas no correlation was found with the date of birth. Correlation between date of splenectomy and age at splenectomy. (284 cases; r=0.73; P < 0.0001). [Color figure can be viewed in the online issue, which is available at wileyonlinelibrary.com.] The probability to be splenectomized was significantly different in different cohorts of birth (Fig. 2). For patients born in the 1960s, 1970s, 1980s, and 1990s, the probability to be splenectomized within the first 10 years of life was 57, 22, 6, and 7%, respectively. No patient less than 5 years of age was splenectomized after 1987, whereas previously 39 children were splenectomized before 5 years of age and 9 before 3 years of age. Cumulative proportion of splenectomy-free survival (Kaplan-Meier) in different cohorts of birth (1960-1969, 1970-1979, 1980-1989, and 1990-1999). During the period of observation, 284 patients (32.6%) have been splenectomized, whereas 588 patients (67.4%) still had their spleen in at the end of the study. (γ2=139.483, P<0.0001). This study confirms an important trend toward a delay or prevention of splenectomy in the transfusion-dependent thalassemic population. A link between the improvement of global management and the lowering of the need of spleen removal had been found in the past [5], but a precise description of the extent and age distribution of this lowering had not been described in previous works. Using shared software for clinical management, we had the chance to merge from five thalassemia Centers a series of homogeneously treated patients that represents a significant portion of the thalassemia major patients living in Italy, and is similar to other national registries such as those from the USA [6], England [7], and France [8]. Overall, 67.4% of our patients were splenectomy free, but the probability to undergo surgery was strongly influenced by the date of birth. For patients born before 1970, this was 57% at 10 years of age and 89% at 20 years of age, whereas it was less than 22% at 10 years of age for patients born after 1970. This picture is significantly different from series described in the past literature. A frequent bias is given by considering in this type of analysis thalassemia intermedia patients, which means a milder phenotype and a lower probability of the need of splenectomy [9]. Being aware of this, we excluded from our study transfusion-dependent patients that had their first transfusion after the age of six years; in our study population, 89% of cases had the first transfusion before the age of three (supplementary online material). Another feature of our series is the treatment homogeneity due to the long-lasting cooperation among our Centers with the use of a common tool as WebThal software and common protocols for transfusion. This is slightly different from most of the published data, where factors like limited access to optimal care and large variability in patients' conditions may have facilitated the need of spleen removal. Nevertheless, the observational nature of our study does not allow any speculation on a causal relationship between the improvement in thalassemia treatment and the prevention of splenectomy. A better understanding of the relationship between splenic function and transfusional schemes and indices may be derived from specific studies on this topic. The increase of mean age at splenectomy has been progressive in our series, as shown in Fig. 1. This may represent an important protection against the risk of severe infections, particularly from encapsulated microorganisms that is not fully covered by active immunization against Pneumococcus, Meningococcus, and group B Haemophilus influenzae [10], and is higher in children below 5 years of age. As the risk of infection after splenectomy lasts for life, and survival and quality of life in thalassemia through the last decades improved progressively, splenectomy should be considered more cautiously today [2]. Thromboembolic complications are frequent in thalassemia, and even more frequent in splenectomized patients [11]. One of the main factors is the procoagulant effect of anionic phospholipids on the surface of altered red blood cells and erythroblasts [12]. As the number of these circulating cells is dramatically triggered by the absence of the spleen, it is likely that the prevention of splenectomy will reduce the risk of thromboembolic complications in thalassemia. Similar considerations should be taken into account for another long-term complication of thalassemia that has been described more recently: pulmonary hypertension [4, 13, 14]. After splenectomy, together with the number of altered red blood cells, also the number of platelets and their activation increases. The prevention of splenectomy, minimizing these factors so important in the pathogenesis of pulmonary hypertension in thalassemia, should have a beneficial effect on perspectives of life of these patients. As additional value, this study underlines the great potential of dedicated digital tools in facilitating the research on rare diseases. This study is part of a large cooperative project among Thalassemia Centers on clinical management of the disease. Since January 2000, five major Italian Centers (Cagliari, Turin, Genoa, Brindisi, and Milan) shared a computerized clinical record for thalassemia. A dedicated software (Webthal®) was developed by us and by Most srl, Turin and made available on the internet to be used with the common browsers [15]. Secure socket layer system and passwords were used to assure data safety and confidentiality. Each local Ethics Committee gave the approval and each individual gave written informed consent to the data collection. All patients followed at the participating centers were considered. Basic data, data on transfusion, iron chelation, supportive treatment and follow-up testing were entered. For this study, each center extracted from the database the content of six fields: patient identifier, date of birth, diagnosis, date of diagnosis, date of first transfusion, and date of splenectomy. The mean pretransfusional hemoglobin for each decade considered in the study has been calculated by each Center. Inclusion criteria were: (i) diagnosis of β thalassemia major; (ii) date of birth between January 1, 1960, and December 31, 1999; (iii) transfusion on a regular basis; (iv) starting of regular transfusion treatment before 6 years of age. Exclusion criteria were: (i) diagnosis of thalassemia intermedia, sickle cell anemia, or other hemoglobinopathies; (ii) lack of any of the previous data. The study was closed on December 17, 2001. This date or a previous date of bone marrow transplantation or death was used as study exit date. Four cohorts of birth were done based on the decade of birth (1960–1969, 1970–1979, 1980–1989, and 1990–1999). A preliminary descriptive analysis was performed and sent to each center for a review of incomplete data or inconsistencies. For the final analysis, Shapiro–Wilks W test for normality, linear correlation, Kaplan-Meier analysis, and Wilcoxon test for group comparison were used. Statistical analysis was performed by Statistica v.5.5, StatSoft, Tulsa, USA. Additional Supporting Information may be found in the online version of this article. Please note: The publisher is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article. Antonio Piga antonio.piga@unito.it*, Melania Serra*, Filomena Longo*, Gianluca Forni , Giovanni Quarta , Maria D. Cappellini§, Renzo Galanello¶, * Università di Torino, Italy, Centro della Microcitemia e delle Anemie Congenite Ospedale Galliera, Genoa, Italy, Divisione di Ematologia, Ospedale Perrino Brindisi, Italy, § Università di Milano, Policlinico Foundation IRCCS, Italy, ¶ Università di Cagliari, Italy.