Adult T-cell leukemia/lymphoma (ATLL) is a hematological malignancy caused by the retrovirus human T-cell lymphotropic virus type I (HTLV-1). Worldwide, 5–20 million individuals are estimated to be carriers, with infection being endemic in Japan, intertropical Africa, the Caribbean, and South America [1]. Among carriers, approximately 5–7% of men and 2–4% of women develop ATLL, usually after a latency of four to six decades, during which time somatic mutations are thought to drive malignant transformation [1–3].

ATLL is classified into four clinical subtypes, smoldering, chronic, lymphoma, and acute/leukemic, according to the Shimoyama classification, based on clinical and laboratory features [4]. The smoldering and chronic forms are considered indolent, in contrast to the aggressive lymphoma and acute types, which have distinct prognostic and therapeutic implications [2,4]. Further refinements subdivided chronic disease into favorable and unfavorable groups and recognized a primary cutaneous tumoral form [5].

Management of ATLL is particularly difficult. The disease shows high resistance to chemotherapy, low sustained response rates, and frequent relapses [6]. Antiviral therapy with zidovudine and interferon is recommended for indolent forms and may benefit some patients with acute disease [7] though results remain debatable [8]. Immunotherapy has been less effective than in other lymphomas: mogamulizumab, an anti-CCR4 antibody, has produced variable responses with limited survival benefit [9–11]. Hematopoietic stem cell transplantation (HSCT) is the only treatment with curative potential and is recommended for all eligible patients with aggressive disease [12]. Overall, prognosis is poor, with survival in aggressive subtypes measured in months [6,13].

In Brazil, the prevalence of HTLV-1 infection in 2012 was estimated at 83–222 cases per 100,000 blood donors [14], corresponding to 300–800 thousand people living with the virus [1]. Based on a 4% lifetime risk and a life expectancy of 75 years, 427–1333 new ATLL cases are expected annually [15]. Yet, national cancer registry data (1986–2016) documented only 369 cases over 30 years, an average of 12 per year, highlighting substantial underdiagnosis and underreporting [15]. Even with underreporting, it is estimated that ATLL represents the second most frequent peripheral T cell lymphoma in Latin America. [16]

Improved characterization of ATLL in Brazil is essential to enhance diagnostic accuracy, treatment strategies, and patient monitoring. In this context, the present study was designed to describe the clinical profile of patients with ATLL diagnosed at a major university cancer center in Brazil.

A retrospective analysis, based on the review of medical records, was conducted of patients diagnosed with ATLL treated at the Hospital das Clínicas of the School of Medicine, University of São Paulo (HC-FMUSP) and the Instituto do Câncer de São Paulo (ICESP) between 1994 and 2022. HC-FMUSP is the largest hospital complex in Latin America and features an outpatient clinic for monitoring asymptomatic HTLV-1 carriers, where 720 patients have been followed since 1994. This study was approved by the Ethics Committee of FMUSP, under the protocol number CAAE: 69,983,723.5.0000.0068.

A total of 55 patients with documented HTLV-1 infection (predominantly confirmed by anti-HTLV-1/2 serology) and fulfilling diagnostic criteria for ATLL were included in this study. The median age at diagnosis was 55 years, with aggressive forms being diagnosed at a younger age compared to indolent forms (median: 49 vs. 57 years; p-value = 0.037) (Table 1). Women comprised the majority of patients (58%), except for the acute subtype, where men accounted for 63% of patients. With the exception of one Japanese individual, all patients were Brazilian. Assessment of ethnic background was hampered by a high proportion of missing data (40% of patients without information on race). Nevertheless, at least 14% of patients were of Asian descent. Importantly, while most patients were born in the state of Sao Paulo, a significant proportion of patients came from the state of Bahia (17%) or other Northeastern states of Brazil (summing up 34% of the patients).

Regarding clinical presentation, skin involvement was common, observed in 31 of 51 patients (69%) with available data and was evenly distributed across Shimoyama subgroups (Table 1). In many cases, a skin biopsy was the procedure that led to HTLV-1 diagnosis. None of the patients presented HTLV-1-Associated Myelopathy/Tropical Spastic Paraparesis (HAM/TSP) and only one patient presented uveitis.

Thirteen laboratory parameters were analyzed, including blood counts and biochemistry (Table 2). As expected, lymphocyte counts and the ratio of LDH to the upper limit of normality were significantly higher in the acute subtype, as was the case of neutrophil counts. Alkaline phosphatase levels were also elevated in the aggressive disease groups, but, unexpectedly, showed no correlation with calcium levels (r = 0.089; p = 0.5739). None of the other variables remained statistically significant after adjustment for multiple comparisons.

According to Shimoyama's classification, 15 patients (27.3%) were diagnosed with smoldering ATLL, 14 (25.5%) with chronic, seven (12.7%) with lymphoma, and 19 (34.5%) with acute disease (Table 1). The majority (19/29; 65.5%) of patients with an indolent form (smoldering/chronic) were identified through the outpatient HTLV-1 carrier monitoring program at the hospital. Nevertheless, despite the availability of this clinic, most patients (51%) were only diagnosed with HTLV-1 infection at the time of ATLL diagnosis. This proportion was even higher (77%) among those presenting initially with aggressive forms (lymphoma/acute). Among patients with a prior HTLV-1 diagnosis (24/55; 44%), the median interval between viral detection and ATLL diagnosis was 9.9 years (IQR: 3.1–17.1 years).

Of the 29 patients with indolent forms, 16 (55%) received treatment, most of whom (12/16; 75%) had chronic ATLL. Treatment was indicated primarily for cutaneous disease (15/16 cases), with zidovudine plus Interferon-alpha (AZT-IFN) being the most frequently used regimen (15/16). Despite treatment, ten patients (62.5%) progressed to an aggressive disease. Among the 13 patients who did not receive therapy, smoldering disease predominated (11/13, 84.6%). The poorer survival observed in treated patients (Figure 1) likely reflects the worse prognosis of chronic compared with smoldering subtypes suggesting that treatment does not overcome this intrinsic difference. Overall, 14 patients progressed from smoldering/chronic disease to aggressive disease, with a median time to progression of 2.1 years (IQR: 0.9–4.3 years). After progression, outcomes were dismal, with a median overall survival of only 55 days (Figure 2).

Figure 1.

Survival by indication for treatment in indolent disease forms. Patients with indolent adult T-cell leukemia/lymphoma forms (Smoldering and Chronic) were treated mainly based on the presence of symptoms (usually, cutaneous manifestations). The patients requiring treatment were mainly from the chronic subgroup, in opposition to a majority of patients with smoldering disease in the untreated groups. Having an indication for treatment was associated with worse survival, and treatment did not seem to prevent disease progression to more aggressive forms.

Figure 2.

Survival after progression to aggressive forms. Median overall survival after progression from an Indolent form to an aggressive form was 55 days, with 12.5% survival at 2 years.

Median overall survival (OS) for the entire cohort was 21 months, but outcomes differed significantly according to Shimoyama's classification at the initial ATLL diagnosis. Patients with the smoldering form had the longest median survival (68 months), followed by those with the chronic form (41 months), underscoring the prognostic distinction between these two indolent forms. In contrast, patients with aggressive disease had markedly shorter survival, with a median of 13 months for the lymphoma form and 12.5 months for the acute subtype (Figure 3).

Figure 3.

Overall Survival by Shimoyama group. Shimoyama Classification captured the difference in prognosis across groups, with median overall survival of aggressive forms around one year after diagnosis.

A total of 39 patients were diagnosed with or progressed to aggressive ATLL. Of these, 36 patients received chemotherapy, most commonly (72%) the CHOEP (Cyclophosphamide, Doxorubicin, Vincristine, Etoposide and Prednisone) regimen. The complete response rate to chemotherapy was low, achieved in only eight of 36 patients (22%). Among responders, half subsequently underwent allogeneic HSCT. Post-transplant survival was as follows: two patients died at 1.5 and 3.5 years after the procedure while two remain alive at the last follow-up (3.0 and 7.5 years, respectively). Of the four patients who achieved complete remission but did not undergo HSCT, three relapsed, including one very late relapse after ten years of follow-up, which ultimately led to death in all cases.

This study also explored whether additional laboratory values beyond those included in the Shimoyama classification and those used to classify chronic ATLL into favorable and unfavorable forms (urea and albumin) were associated with mortality in the cohort. In particular, the prognostic relevance of red cell distribution width (RDW) recently proposed as a prognostic factor for survival in ATLL, was investigated [18]. As shown in Figure 4, patients with high RDW (≥14.5%, the most commonly accepted threshold for normality) had significantly shorter survival compared with those with low RDW. However, in the multivariate Cox proportional hazards model, including age, Shimoyama group, and eosinophil count, RDW was not independently associated with survival (Table 3). Conversely, higher eosinophils counts were associated with reduced survival in the multivariate model.

Figure 4.

Prognosis of RDW in adult T-cell leukemia/lymphoma patients. Patients with higher red cell distribution width (RDW) showed inferior survival (median: 0.9 years; 95% confidence interval: 0.4–2.5) compared to those with normal RDW (median: 4.8; confidence interval: 3.4-NA; p = 0.0014).

Brazil harbors one of the largest populations of HTLV-1 carriers [1], yet data on infection monitoring and secondary illnesses remain scarce. Even as a referral center for HTLV-1 infection and ATTL, about half of cases in the current study were diagnosed already in aggressive clinical forms, with many patients only learning of their HTLV-1 status at ATLL onset, highlighting the underrecognition of this condition in the country [15].

The present cohort showed an overrepresentation of patients of Asian descent (14.5%), compared to ∼1% in the general Brazilian population, a finding consistent with the higher HTLV-1 prevalence in this community [20]. A high proportion of patients also originated from Bahia, a state with large African-descendant populations and the highest burden of ATLL in Brazil [5]. Women predominated overall, but men were more frequent in the acute leukemic group, in line with prior reports in Latin America [5,16,21,22]. Median age at diagnosis was 55 years-of-age, which was younger than in Japan [13], but similar to other Latin American series [16,22], and in line with a recently published analysis comparing different world regions [23]. The median age at death was 52 years, which is well below the national life expectancy of 75 years reported in Brazil in 2022.

Few patients with Smoldering disease (27%) and most with Chronic disease (86%) received treatment, usually AZT-IFN for skin lesions. Unlike previous reports [7,24)], most treated patients (62.5%) progressed rapidly to aggressive forms, with poor survival (Figure 1); no effect of AZT-IFN regimen in preventing disease progression and was observed. Aggressive types showed dismal outcomes, with median survival of roughly one year for lymphoma/acute cases and two months for progression cases (Figures 2 and 3). We hypothesize that prior AZT-IFN exposure in 71% of patients with documented disease progression may have selected for chemoresistant clones, thereby explaining their exceptionally short survival compared to patients with newly diagnosed, de novo advanced disease.

The CHOEP regimen, with or without concomitant antiviral treatment, was the most common treatment for advanced disease, which is in line with patterns of treatment in Latin America., [23] Complete response rates were low (22%), as commonly reported for ATLL [6,21]. These well-documented features of aggressiveness and treatment resistance explain why ATLL carries an exceptionally poor prognosis, even among T-cell lymphomas [16], a group of malignancies that is already notoriously challenging to treat.

Given the low response to chemotherapy and high relapse rate, allogeneic HSCT is a fundamental procedure in ATLL management [12,25]. However, data on HSCT procedures and outcomes in low- and middle-income countries remains very limited [22,26]. Indeed, in a recent report including 364 Latin American patients with ATLL, only one underwent HSCT [23], with a reported disease free survival of five years [27]. In the current cohort, Half of the patients (n = 4) who responded to chemotherapy underwent allogeneic HSCT. Although numbers were small, three of the four transplanted patients remained alive at three years, supporting its benefit when remission is achieved [25], and suggesting that expanding access to this treatment can represent an important movement in patient management in Latin America. However, as few patients are considered candidates for HSCT, newer treatment strategies are urgent in ATLL. Given the rarity of the disease and its concentration in underserved populations in most developed countries, an effort to identify and include patients from low and middle-income countries in clinical trials is necessary. Consequently, initiatives like the Epidemiology of Lymphomas in Latin America and the Brazilian T-Cell Project will play an important role in the improvement of ATLL in the region.

Prognostic markers apart from the Shimoyama classification were also assessed. Due to sample size, only RDW and eosinophil count were included. High RDW, previously reported in ATLL [18], correlated with worse outcomes, consistent with findings in T-cell lymphomas [28] and other diseases [29]. Eosinophilia (>500 cells/mm³) is common in ATLL [30] and occurred in 18% of cases; Although survival was not significantly different when using this dichotomous cutoff, higher eosinophil counts (analyzed as a continuous variable) were independently associated with an increased mortality risk in our multivariable model (Table 3). These parameters may provide prognostic value when standard classification is incomplete.

The several limitations of this study stem primarily from its retrospective design. This led to missing data on demographic characteristics and some laboratory values, as some patients did not undergo a standardized sample collection. Thus, the analyses of bone marrow infiltration and beta-2 microglobulin levels were not possible. Flow cytometric protocols changed along the study timeline, rendering the comparison of the percentage of abnormal cells across study groups unfeasible. In addition, the small sample size, especially when considering the lymphoma subgroup, limited the power to find differences between the groups and correlations among variables. Finally, as a single center study, the generalizability of the findings may not apply to other centers in Brazil, even though many findings, including limitations in the availability of treatment were similar to what has been previously published in the region [5,21,23]. These limitations underscore the need of a coordinated effort to include ATLL patients in prospective multicenter or international registries of the disease, as already stated in the current work.

ATLL in Brazil remains underdiagnosed, with half of patients presenting in aggressive stages. Antiviral therapy did not prevent progression, chemotherapy responses were poor, and allogeneic HSCT was feasible only in few patients, though potentially curative. These findings highlight the urgent need for effective therapies and reinforce the importance of nationwide HTLV-1 screening and transmission prevention as the most impactful strategies to reduce disease burden.

FAM, KSM and YN study design. FAM, JCKS, and JVRO data collection. JCKS, JGS, LJO data analysis. JCKS and FAM initial manuscript draft. JC, JP, VR and YN supervised, final data analysis, and preparation of the final manuscript, ensuring its scientific accuracy.

This study was reviewed and approved by the Research Ethics Committee of the Hospital das Clínicas, Faculty of Medicine, University of São Paulo (Approval Number: 6.127.685; CAAE: 69983723.5.0000.0068). The study is in compliance with the ethical principles established by the Declaration of Helsinki and with the guidelines of the Brazilian National Health Council Resolution CNS n° 466/12.