Splenectomy in β‐thalassemia patients: Practices and risks in a nationwide study

β-thalassemia is a common hemoglobinopathy characterized by a genetically determined defective production of the β-globin chain of hemoglobin (Hb), resulting in chronic anemia and ineffective erythropoiesis.1 Reducing transfusion requirements to better control iron overload is a major goal in patient care. Several approaches have been developed to that end, including drugs to decrease ineffective erythropoiesis and improve anemia,2-4 and splenectomy to decrease spleen red blood cell trapping. The indication for splenectomy remains, however, challenging to define due to its arguable risk-to-benefit ratio: while surgery may improve transfusion requirements, this benefit can be outweighed by an increased risk of infectious and thrombotic complications. This retrospective study aims to further analyze the incidence of spleen-related complications, namely thrombotic events (TE) and pneumococcal complications, in a large national cohort of patients with β-thalassemia followed longitudinally in the French Thalassemia Registry (NaThalY) to better characterize the risk associated with splenectomy. As of December 2022, a total of 834 patients with β-thalassemia were enrolled in the registry, including 544 (65%) with thalassemia major (TM) and 290 (35%) with thalassemia intermedia (TI) (Supporting Information Methods). Overall, 322 (39%) patients underwent splenectomy with a significantly greater proportion and at a younger age in TM patients. Focusing on the main common known risks of splenectomy, all data entries in the registry of infectious or thrombotic nature were considered and analyzed during the follow-up period for all patients. The observation period was 20 200 patient/years (PY) with a median follow-up for all patients of 20.0 years. Further characteristics and general outcome are detailed in Table 1A. 31.8 (8.9; 79.4) 61.0 (19.5; 79.4) 31.4 (8.9; 55.6) 28.5 (8.9; 51.2) 33.9 (21.3; 79.4) A sharp decrease was evidenced over time in the proportion of splenectomized patients (from 80% in patients born in the 60 s to 30% in those born in the 2000s) along with a decrease in median age at splenectomy (17.9 years vs. 6.5 years) in both thalassemia phenotypes (Figure 1A). Regarding infectious complications, four cases due to encapsulated bacteria were documented (two pneumococcal meningitis, one pneumococcal pneumonia, and one pneumococcal sepsis), all in splenectomized patients (3 TM, 1 TI), yielding an incidence of 0.04 PY. Median time of onset after splenectomy was 15.4 years (min 5; max 22). Additionally, 27% (n = 13/48) of causes of death reported during the study period were due to infections (with no identified pathogen available), with 92% (n = 12) involving splenectomized patients. The occurrence of TE was analyzed according to spleen status and phenotype. Distribution is shown in Table 1B. A total of 92 TE was observed in 59 patients, including 55 (93%) who underwent splenectomy, showing a significantly higher rate than in nonsplenectomized patients (P < 0.0001). The overall TE incidence dramatically increased by a 13-fold following splenectomy (0.79 vs. 0.06 PY; P < 0.0001). Those TE included 27 venous thrombosis (31.4%), 19 pulmonary embolisms (22.1%), 15 cases of pulmonary hypertension (17.4%), 11 portal thrombosis (12.8%), 6 thrombosis associated with a central device (7%), 6 strokes (7%), and 2 arterial thrombosis (2.3%). Multiple TE occurred in 30% (n = 17) of splenectomized patients and included up to 6 episodes in two. Of the four non-splenectomized patients who experienced TE, three were TI with a first event occurring at a mean age of 67 years. The total number of TE was significantly higher in TI patients compared to TM, with an incidence of 0.64 PY and 0.31 PY, respectively (P = 0.0008). TI patients also displayed a higher rate of multiple events. Following splenectomy, the highest overall incidence was evidenced in patients with TI (1.39 PY) compared to TM patients (0.48 PY). Median (Q1; Q3) time between splenectomy and first TE was 18.7 (8.6; 24.1) years, similarly in TM and TI patients (Figure 1B). At first TE, 22 patients (37%) had comorbidities known to be associated with an increased thrombotic risk, such as heart failure, arrhythmia, diabetes, cancer, or cirrhosis. Other known TE risk factors were identified in 5 (8.5%) patients (factor V Leiden and factor II mutations, protein C and S deficiency, antiphospholipid syndrome (APS), or oestroprogestative contraception). Only one death was related to a TE: a 30-year-old female died from a pulmonary embolism in a context of heart failure and diabetes, 20 years after splenectomy. Multivariate analysis, including β-thalassemia phenotype, spleen status, and age, showed that splenectomy was the most important risk factor for TEs, with a risk increased by 20. As expected, TI phenotype came second, followed by age (Table 1C). Transfusion therapy did not appear to be fully protective of TE, with 64% (n = 38) of patients who experienced TE being regularly transfused at the time of first occurrence. This study, focusing on post-splenectomy events on a nationwide basis over multiple decades, allows to demonstrate two key findings: a significant decline in the use of splenectomy among patients with β-thalassemia over the last 20 years; and a high incidence of thrombotic complications following splenectomy, across all thalassemia phenotypes, with the highest incidence, consistent with previous evidence, observed in TI patients.5-7 Abnormal, prematurely senescent, or adherent red cells circulating in thalassemia patients are normally cleared by specialized macrophages within the spleen. Following splenectomy, these abnormal red cells along with an increased platelet count are found in the peripheral circulation altogether increasing the thrombotic risk.8 Historically, splenectomy was indicated to improve anemia in TI and reduce transfusion needs in TM, so we can reasonably assume that was the case for the patients included in this cohort. Nevertheless, presumably due to growing awareness of the high incidence of complications, indications for splenectomy in TM patients have decreased over time in France, similar to other countries, with the prompt initiation of iterative transfusions in severely anemic patients, reducing the occurrence of hypersplenism.9, 10 In fact, splenectomy should rarely be indicated in well-transfused patients as regular transfusions suppress ineffective erythropoiesis, extra medullary erythropoiesis, and splenomegaly. In TI patients, who are the most prone to post-splenectomy TE, a discussion on the benefit/risk ratio of splenectomy versus regular transfusions should be undertaken. Access to medications that alleviate anemia and reduce ineffective erythropoiesis, such as pyruvate kinase activators, activin ligand traps, or HbF inducers, will be of potential benefit in those patients.11 Expectedly, venous thrombosis was the most frequent event. Portal thrombosis merits special attention as it was mainly found upon systematic screening in otherwise asymptomatic patients, sometimes during a distant follow-up from splenectomy and found with a higher frequency (3.4%) than in other conditions described in the literature.12 As it may additionally impact liver function, it appears particularly important to systematically monitor this complication with portal echography screening post-surgery. Comorbidities such as cardiovascular disease or diabetes, which are known thrombotic risk factors, are also frequent in β-thalassemia patients due to iron overload. In this study, more than one-third of patients who experienced TE had comorbidities or identified risk factors, thus highlighting the need for these high-risk patients to be screened and closely monitored. Regular transfusions reduce the risk of TE by suppressing ineffective erythropoiesis and diluting abnormal red cells in the circulation. In this study, ten patients classified as TI were regularly transfused at first event, suggesting that the protecting effect of transfusion is offset by the effect of splenectomy in the long term, particularly in TI patients.5 Overall, prophylactic anti thrombotic treatment seems reasonable to recommend given the high rate of TE and the hypercoagulable state in beta thalassemia patients, particularly in TI patients.13 Very few studies have previously evaluated the timing of thrombotic complications after splenectomy and to our knowledge, none have done so in both TM and TI population over an extended period.14-16 One notable feature here is the ability to assess this long-term risk. With 6734 cumulated years of follow-up after splenectomy, complications occurred up to 42 years after surgery with no significant difference between TM and TI, underscoring the need for continued follow-up. Higher rates of infections have long been reported after splenectomy in general.17 While the incidence may have been underestimated by the inability to identify the pathogen and missing data, notably regarding infectious causes of death, the low incidence of post-splenectomy sepsis caused by encapsulated bacteria remains an interesting feature in this study compared to the incidence observed in the context of trauma-related splenectomy.17 This study has several limitations. As is often the case, the data extracted from a registry may result in biases and/or be subject to missing information within or outside the collected dataset, such as indication for splenectomy, post-splenectomy TE prevention, and additional risk factors. Also, the older age of splenectomized patients in this study may have overestimated the association between splenectomy and TE, while conversely, the incidence of TE in nonsplenectomized patients may rise with time. Finally, in this study, as in others, a strong association between variables does not permit us to infer causality. Altogether, these findings add to the growing body of evidence that splenectomy is associated with a high risk of thrombotic complications in β-thalassemia patients, particularly in TI patients. They support the need for a rigorous individualized assessment of risk factors and highlight the importance of lifelong prevention and monitoring of TE in these patients. Hopefully, the development of new therapies that alleviate the burden of anemia and improve the quality of life for affected individuals will further reduce the need for splenectomy. We wish to thank all physicians who participated in the follow-up of patients and the team supporting NaThalY. The following investigators contributed to NaThalY, composing the NaThalY Network: Catherine Mathey, CH du Pays d'Aix, Aix en Provence; Louisa Bey, CH Notre-Dame de la Miséricorde, Ajaccio; Marie-Josée Lucchini, CH Notre-Dame de la Miséricorde, Ajaccio; Valérie Li Thio Te, CHU Amiens Picardie, Amiens; Reda Garidi, CHU Amiens Picardie, Amiens; Loïc Garcon, CHU Amiens Picardie, Amiens; Corentin Orvain, CHU d'Angers, Angers; Jean-François Brasme, CHU d'Angers, Angers; Claire Reynes, CH Annecy Genevois, Annecy; Marie Belloy, CHI Robert Ballanger, Aulnay-Sous-Bois; Julie Durin, CHI Robert Ballanger, Aulnay-Sous-Bois; Oula Alhindy, CH Henri Mondor, Aurillac; Lara Sabbagh, CH Henri Mondor, Aurillac; Borhane Slama, CH Henri Duffaut, Avignon; Anthony Sanchez, CH de Bastia, Bastia; Alice Moulin, CH de Bastia, Bastia; Danielle Belgodere, CH de Bastia, Bastia; Lise Benjemia, CHRU Besançon, Besancon; Pauline Simon, CHRU Besançon, Besancon; Eric Oziol, CH de Béziers, Beziers; Nathalie Aladjidi, Groupe hospitalier Pellegrin, Bordeaux; Yoann Huguenin, Groupe hospitalier Pellegrin, Bordeaux; Charlotte Jubert, Groupe hospitalier Pellegrin, Bordeaux; Edouard Forcade, Groupe hospitalier Pellegrin, Bordeaux; Pierre Duffau, Groupe Hospitalier Saint André, Bordeaux; Abdallah Maakaroun, CH Jacques-Coeur, Bourges; Florence Lachenal, Hôpital Pierre Oudot, Bourgoin-Jallieu; Charline Pegon, Hôpital d'Estaing, Clermont-Ferrand; Marie-Helene Odievre, CH Louis Mourier, Colombes; Charlotte Petitdidier, CHSF, Corbeil-Essonnes; Sai Atmani, CH Laennec, Creil; Florence Raqbi, CH Laennec, Creil; Corinne Pondarre, CHI Créteil, Créteil; Jenny Youn, CHI Créteil, Créteil; Annie Kamdem, CHI Créteil, Créteil; Dora Bachir, CHU Henri Mondor, Créteil; Pablo Bartolucci, CHU Henri Mondor, Créteil; Gonzalo De Luna, CHU Henri Mondor, Créteil; Anoosha Habibi, CHU Henri Mondor, Créteil; Elena Fois, CHU Henri Mondor, Créteil; Antoine Leblanc, CHSF, Evry; Pierre-Yves Dides, CH de Grasse, Grasse; Caroline Makowski, CHU Grenoble, Grenoble; Corinne Armari, CHU Grenoble, Grenoble; Narcisse Elenga, CH Ouest guyanais, Guyane; Christelle Chantalat-Auger, Hôpital Bicêtre, Le Kremlin-Bicêtre; Françoise Driss, Hôpital Bicêtre, Le Kremlin-Bicêtre; Ferielle Zenchri, Hôpital Bicêtre, Le Kremlin-Bicêtre; Corinne Guitton, Hôpital Bicêtre, Le Kremlin-Bicêtre; Mathilde Jehanne, CHU de La Réunion, La Réunion; Cecile Stoven, CHU de La Réunion, La Réunion; Axel Chaminade, CHU de La Réunion, La Réunion; Stephanie Gourdon, CHU de La Réunion, La Réunion; Yves Reguerre, CHU de La Réunion, La Réunion; Edouard Randriamalala, CHU de La Réunion, La Réunion; Véronique Paumier-Bonjour, CH Jacques Monod, Le Havre; Anne Besancon-Bergelin, CH Le Mans, Le Mans; Anne Lambilliotte, CHRU Jeanne de Flandre, Lille; Benedicte Bruno, CHRU Jeanne de Flandre, Lille; Fanny Gonzales, CHRU Jeanne de Flandre, Lille; Melissa Barbati, CHRU Jeanne de Flandre, Lille; Benjamin Carpentier, Hopital Saint Vincent de Paul, Lille; Christophe Piguet, CHU Dupuytren, Limoges; Mohamed Touati, CHU Dupuytren, Limoges; Arnaud Hot, CHU Edouard Herriot, Lyon; Fiorenza Barraco, CHU Edouard Herriot, Lyon; Giovanna Cannas, CHU Edouard Herriot, Lyon; Emilie Virot, CHU Edouard Herriot, Lyon; Romain Fort, CHU Edouard Herriot, Lyon; Grégoire Ducoux, CHU Edouard Herriot, Lyon; Hélène Desmurs Clavel, CHU Edouard Herriot, Lyon; Kamila Kebaili, Institut d'hématologie et d'oncologie pédiatrique, Lyon; Nathalie Garnier, Institut d'hématologie et d'oncologie pédiatrique, Lyon; Cécile Renard, Institut d'hématologie et d'oncologie pédiatrique, Lyon; Yves Bertrand, Institut d'hématologie et d'oncologie pédiatrique, Lyon; Anne-Laure Peugnet, Institut d'hématologie et d'oncologie pédiatrique, Lyon; Carine Halfon-Domenech, Institut d'hématologie et d'oncologie pédiatrique, Lyon; Judith Diakiesse, CH François Quesnay, Mantes La Jolie; Kokou-Placide Agbo-Kpati, Grand hôpital de l'Est Francilien, Marne la Vallée; Nathalie Ravet, Grand hôpital de l'Est Francilien, Marne la Vallée; Isabelle Thuret, Hôpital de la Timone, Marseille; Sarah Szepetowski, Hôpital de la Timone, Marseille; Estelle Jean, Hôpital de la Timone, Marseille; Benjamin De Sainte Marie, Hôpital de la Timone, Marseille; Benoit Meunier, Hôpital de la Timone, Marseille; Benoit Faucher, Hôpital de la Timone, Marseille; Matias Muszlak, CH Mayotte, Mayotte; Abdourahim Chamouine, CH Mayotte, Mayotte; Alain Garou, CH Mayotte, Mayotte; Caroline Masserot-Lureau, Grand hôpital de l'Est Francilien, Meaux; Jean-Marc Schneider, CH Bel-Air Metz Thionville, Metz-Thionville; Veronique Dorvaux, CHR Metz Thionville, Metz-Thionville; Philippe Carassou, CHR Metz Thionville, Metz-Thionville; Honorat Andriamihaja, Centre Hospitalier de l'Agglomération Montargoise, Montargis; Muriel Lalande, Hôpital Arnaud de Villeneuve, Montpellier; Perrine Mahe, Hôpital Arnaud de Villeneuve, Montpellier; Anne Sirvent, Hôpital Arnaud de Villeneuve, Montpellier; Patricia Aguilar Martinez, Hôpital Arnaud de Villeneuve, Montpellier; Ivan Bertchansky, Hôpital Arnaud de Villeneuve, Montpellier; Jean-Claude Eisenmann, CH Mulhouse Emile Muller, Mulhouse; Agathe Masseau, CHU de Nantes, Nantes; Marie-Laure Couec, CHU de Nantes, Nantes; Fanny Rialland, CHU de Nantes, Nantes; Pierre-Simon Rohrlich, CHU Archet, Nice; Sandrine Baron-Joly, CHU de Nîmes, Nimes; Mohamed Conde, CHR Orléans, Orleans; Antoine Dossier, Bichat, Paris; Valentine Brousse, Hôpital Robert Debré, Paris; Karima Yakouben, Hôpital Robert Debré, Paris; Malika Benkerrou, Hôpital Robert Debré, Paris; Ghislaine Ithier, Hôpital Robert Debré, Paris; Florence Missud, Hôpital Robert Debré, Paris; Jean-Hugues Dalle, Hôpital Robert Debré, Paris; Bérengène Koehl, Hôpital Robert Debré, Paris; Laurent Holvoet, Hôpital Robert Debré, Paris; Jean-Antoine Ribeil, Hôpital Necker, Paris; Laure Joseph, Hôpital Necker, Paris; Marina Cavazzana, Hôpital Necker, Paris; Marianne De Montalembert, Hôpital Necker, Paris; Benedicte Neven, Hôpital Necker, Paris; Despina Moshous, Hôpital Necker, Paris; Mélissa Taylor, Hôpital Necker, Paris; Slimane Allali, Hôpital Necker, Paris; Bruno Varet, Hôpital Necker, Paris; Gerard Socie, CHU Saint Louis, Paris; Alienor Xhaard, CHU Saint Louis, Paris; Régis Peffault De La Tour, CHU Saint Louis, Paris; Thierry Leblanc, CHU Saint Louis, Paris; Nicolas Boissel, CHU Saint Louis, Paris; Marie Robin, CHU Saint Louis, Paris; Nathalie Dhedin, CHU Saint Louis, Paris; Robert Girot, CHU Tenon, Paris; Sarah Mattioni, CHU Tenon, Paris; François Lionnet, CHU Tenon, Paris; Jean Donadieu, CHU Trousseau, Paris; Guy Leverger, CHU Trousseau, Paris; Sylvie Fasola, CHU Trousseau, Paris; Jean-Benoit Arlet, Hôpital Européen George Pompidou, Paris; Corina Manoli, Hôpital Européen George Pompidou, Paris; Maryse Etienne-Julan, CHU de la Guadeloupe, Pointe à Pitre; Emmanuelle Bernit, CHU de la Guadeloupe, Pointe à Pitre; Justine Gellen-Dautremer, CHU la Milétrie, Poitiers; Stéphanie Eyssette, Hôpital NOVO, Pontoise; Gilles Blondin, Centre Hospitalier de Cornouaille, Quimper; Claire Pluchart, American Memorial Hospital, Reims; Stanislas Nimubona, Hôpital Pontchaillou, Rennes; Jacinthe Bonneau, Hôpital Sud, Rennes; Sophie Bayart, Hôpital Sud, Rennes; Sophie Pertuisel, Hôpital Sud, Rennes; Benedicte Collet, Hôpital Victor Provo, Roubaix; Laurence Detourmignies, Hôpital Victor Provo, Roubaix; Julie Machin, Hôpital Victor Provo, Roubaix; Mathieu Wemeau, Hôpital Victor Provo, Roubaix; Laurence De Tourmigni, Hôpital Victor Provo, Roubaix; Jean-Pierre Vannier, CHU Charles Nicolle, Rouen; Bruno Filhon, CHU Charles Nicolle, Rouen; Cecile Dumesnil, CHU Charles Nicolle, Rouen; Agnes Lahary, CHU Charles Nicolle, Rouen; Pascale Schneider, CHU Charles Nicolle, Rouen; Stéphanie Ngo, CHU Charles Nicolle, Rouen; Aurore Malric, Hôpital Delafontaine, Saint Denis; Maud Schoeny, CH Saint Dizier, Saint Dizier; Claire Berger, CHU Nord Saint Etienne, Saint-Etienne; Isabelle Guichard, CHU Nord Saint Etienne, Saint-Etienne; Aurélie Desbree, CHU Nord Saint Etienne, Saint-Etienne; Jean-Louis Stephan, CHU Nord Saint Etienne, Saint-Etienne; Audrey David, Hôpital de Hautepierre, Strasbourg; Catherine Paillard, Hôpital de Hautepierre, Strasbourg; Alexandra Salmon, Hôpital de Hautepierre, Strasbourg; Shanti Ame, Nouvel Hôpital Civil, Strasbourg; Philippe Gestas, Clinique Cordella, Tahiti; Marianne Besnard, Clinique Paofai, Tahiti; Marie-Pierre Castex, CHU des Enfants, Toulouse; Geneviève Plat-Wilson, CHU des Enfants, Toulouse; Pierre Cougoul, CHU des Enfants, Toulouse; Marlène Pasquet, CHU des Enfants, Toulouse; Jill Serre, CHU Tours, Tours; Marion Yvert, CHU Tours, Tours; Jean-Baptiste Valentin, CHU Tours, Tours; Francis Duchene, Hôpital Nord Franche-Comté, Trévenans; Suzanne Mathieu, CHU de Nancy - Hopital Brabois, Vandoeuvre Les Nancy; Marie Drouilly, CHU de Nancy - Hopital Brabois, Vandoeuvre Les Nancy; Marie Detrait, CHU de Nancy - Hopital Brabois, Vandoeuvre Les Nancy; Cécile Pochon, CHU de Nancy - Hopital Brabois, Vandoeuvre Les Nancy; Aurélie Phulpin, CHU de Nancy - Hopital Brabois, Vandoeuvre Les Nancy; Sylvie Nathanson, Centre Hospitalier de Versailles, Versailles. Audrey Benoit: Conceptualization; investigation; data curation; Justine Sarah and and Isabelle and Le Catherine and Valentine Conceptualization; NaThalY Network: The no of The French national registry is by the This study also from a patients were that data collected be for upon for in NaThalY. This no The data that support the findings of this study are from the upon reasonable The is not for the or of supporting information by the than missing should be to the for the