Portal de Periódicos da CAPES

Carta Acesso aberto Produção Nacional Revisado por pares

BIBTEX RIS

Chronic lymphocytic leukemia in Brazil: A retrospective analysis of 1903 cases

2017; Wiley; Volume: 92; Issue: 8 Linguagem: Inglês

10.1002/ajh.24779

1096-8652

Matheus Vescovi Gonçalves, Celso Arrais-Rodrigues, Irene Lorand‐Metze, Marcelo Pitombeira de Lacerda, Maria de Lourdes Chauffaille, Alita Azevedo, Cíntia G. Machado, Carlos Chiattone, Sérgio Costa Fortier, Leila Perobelli, Maura Rosane Valério Ikoma, Nelma Clementino, Nelson Hamerschlak, Vivia Machado Sthel, Larissa Veloso Mendes Ommati, Danielle Leão Cordeiro de Farias, Fernando Barroso Duarte, Valéria Buccheri, Ana Paula de Azambuja, Denise Ramos de Almeida, V.L.P. Figueiredo, Mihoko Yamamoto,

Immunodeficiency and Autoimmune Disorders

Despite previously described epidemiological differences in Chronic Lymphocytic Leukemia (CLL) incidence and prognosis,1-4 CLL has not been assessed in larges series in multicultural developing countries. Economical and health access inequalities in these countries have not been represented in clinical trials or patient cohorts. In Brazil, no official nationwide data on the incidence and lethality of CLL is available. Our population is essentially interethnic, and has experienced a progressive ageing process, with probable effects on CLL prevalence. The Brazilian Registry of CLL was started in 2004 as a prospective noninterventional data collection tool. All participant centers register their patients online (www.llc.unifesp.br) and each center's ethics committee approved the study. Inclusion criteria for enrollment followed the IWCLL guidelines.5 From January 2004 to December 2016, 1903 patients from 28 centers were enrolled. Epidemiological and clinical characteristics are presented in Table 1. The median age at diagnosis was 66 years (range 28–106), and 56% were male (male: female ratio =1.26:1). Median age at diagnosis was identical between genders. However, male gender was associated with more advanced disease (Binet B/C, male vs. female: 41% vs. 33%; P = .001). Public university centers were responsible for 52% of the cases, public non-university institutions for 36% and private centers for 12%. The most frequently reported prognostic marker was CD38 expression. At a cutoff of 30%, 28% of patients were CD38-positive. Cytogenetic studies were performed in only 309 patients (16.2%): fluorescence in situ hybridization (FISH) in 241 cases (12.5%) and karyotyping in 74 (4.5%). The frequencies of abnormalities by FISH were: del(13q) = 40%; +12 = 22%; del(11q) = 16%; and del(17p) = 12%. FISH tests at diagnosis (N = 169) were more often normal than at treatment (38% × 18%, P = .002), with lower rates of de13q (35% × 51%), del11q (10% × 29%) and del17p (9% × 21%) (P < .01). Among 1903 patients, 1687 had sufficient data for survival analyzes. Median follow-up time was 36 months (range, 3–155). Treatment-free survival (TFS) at 3 and 5 years was 44% and 32%, respectively. Binet staging had a strong correlation with TFS (Table 1). Male patients had inferior TFS (38% vs. 51%, P < .0001). A multivariate analysis showed that both advanced stage [hazard ratio (HR)=3.21, 95% CI= 2.82–3.66] and male sex (HR = 1.20, 95% CI = 1.06–1.37) were independently associated with worse TFS. Younger patients (<65 years) presented with more advanced Binet stage (A = 55%, B = 28%, and C = 17% of cases, versus A = 63%, B =18%, and C = 19% in older patients; P = .0004) and had worse TFS (40% vs. 48%, P = .02). FISH with no abnormality or isolated del13q had superior 3-year TFS than patients with other abnormalities, alone or combined (62% vs. 42%, P < .0001). Patients with del17p patients had inferior TFS (38% vs. 58%, P = .03). Unexpectedly, CD38 expression did not correlate with TFS. During follow-up, 845 patients received treatment. Chlorambucil (alone or in combination, at any time) was received by 65%, fludarabine-based regimens were received by 41%, and only 22% received rituximab in combination with other drug(s). CHOP/CVP-like regimens were received by 19% of treated patients, and 5% received CHOP/CVP as the only treatment protocol, combined with rituximab in only 1% of cases. Fludarabine was predominantly used among patients younger than 65 years (younger vs. older: 59% vs. 24%, P < .001), as well as rituximab (28% vs. 17%, P < .001). Chlorambucil was mainly used in older patients (younger vs. older: 49% vs. 80%, P < .001), however 25% of younger patients received chlorambucil as the only treatment regimen. Other treatment modalities were seldom used, such as allogeneic stem cell transplantation, bendamustine, ibrutinib or venetoclax. Overall survival (OS) was 89% at 3 years and 82% at 5 years. While Binet A had superior survival, no differences were noted between Binet B and C (Binet A 3-year OS = 94%; B/C = 85%, P < .0001). Patients 65 years had 3-year OS estimates of 93% vs. 87% and 5-year OS of 89% vs. 81% (P = .009). Binet A patients <65 years had mortality rates very similar to the general Brazilian population (3% vs. 4.5%). However, young Binet B/C patients had much higher mortality rates: 12.9% in 3 years and 20% in 5 years. For older CLL patients, probability of death in 3 and 5 years was, respectively, 13.8% and 21.5%, versus 8% and 14.2% in the general population (www.ibge.gov.br). After a multivariate analysis, both advanced Binet stage (HR = 2.49, 95% CI = 1.80–3.43) and older age (HR = 2.26, 95% CI = 1.58–3.24) remained independent risk factors for OS. This is the first study of this kind in our country and one of the largest in the developing world. Many limitations should be considered in interpreting our results, especially in the absence of official data. The observed median age at diagnosis of 66 years was slightly younger than other series.1, 5 A lower male: female ratio (1.26) could be explained by a more significant gap between life expectancy for men and women in Brazil (71 vs. 79 years) compared to Germany (79 vs. 83) or the United States (77 vs. 82).6 The use of prognostic markers was very low. Only 16% of patients were evaluated for cytogenetic abnormalities, compared to 49% of US patients outside of clinical trials.7 The choice of treatment in our registry should be evaluated in the context of low availability of rituximab and very low availability of other monoclonal antibodies and small molecules (ibrutinib, idelalisib or venetoclax). Nevertheless, the use of fludarabine (FC, FCR) was rather low (41%), even among young patients (59%). Use of regimens with low evidence of benefit, such as CHOP, CVP or chlorambucil monotherapy, was observed in a significant proportion of patients (45%). Finally, the registry is also intended to promote greater scientific integration. The Brazilian Group of CLL (GBLLC) was created in 2013, with periodic meetings, recruitment for clinical trials and publication of recommendations for CLL diagnosis and treatment.8 Cooperative efforts in Brazil and other developing countries are in need for a more representative scenario of CLL, and the adequate assessment of epidemiology, outcomes and unmet needs in diagnosis, prognostic evaluation and treatment. The authors would like to thank the investigators of the Brazilian Group of Chronic Lymphocytic Leukemia for providing information for the Registry database. The authors gratefully acknowledge the collaboration of the following centers (in alphabetical order) that submitted data for this analysis: Casa de Saúde Santa Marcelina (São Paulo), Centro de Pesquisa Oncológica – CEPON (Florianópolis), Hemocentro Regional de Uberaba (Uberaba), Hemocentro de Pernambuco (Recife), Hospital Israelita Albert Einstein (São Paulo), Hospital do Servidor Publico Estadual (São Paulo), Hospital Amaral Carvalho (São Paulo), Hospital Brigadeiro/Hospital de Transplantes Euryclides de Jesus Zerbini (São Paulo), Hospital Evangélico de Cachoeiro de Itapemirim (Cachoeiro de Itapemirim), Hospital Haroldo Juaçaba/Inst. do Câncer do Ceará (Fortaleza), Hospital Santa Lúcia (Brasília), Hospital São Vicente de Paulo (Passo Fundo), Hospital Sírio Libanês (São Paulo), Instituto de Imunologia e Oncologia (São Paulo), Instituto de Previdência dos Servidores do Estado de Minas Gerais (Belo Horizonte), Instituto do Câncer do Estado de São Paulo (São Paulo), Santa Casa de Misericórdia de Santos (Santos), Santa Casa de Misericórdia de São Paulo (São Paulo), Universidade Estadual de Campinas (Campinas), Universidade Estadual de Londrina (Londrina), Universidade Federal de Goiás (Goiânia), Universidade Federal de Minas Gerais (Belo Horizonte), Universidade Federal de São Paulo (São Paulo), Universidade Federal do Ceara (Fortaleza), Universidade Federal do Paraná (Curitiba), Universidade Federal do Rio de Janeiro (Rio de Janeiro). We would also like to thank José Márcio Duarte and the staff from Departamento de Informática em Saúde - UNIFESP for providing technical support. The authors have no conflicts of interest to declare. Additional Supporting Information may be found online in the supporting information tab for this article. Supporting Information 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.