Key points

Introduction

Adult T-cell leukemia/lymphoma (ATLL) is a severe, yet fascinating, disease that develops in a small proportion of individuals who are chronically infected with human T-cell lymphotropic virus type 1 (HTLV-1). ATLL is epidemiologically unique because of this association with HTLV-1, which is the first described human retrovirus. Clinically, ATLL has protean manifestations, with each of 4 variants possessing a distinct constellation of symptoms. The aggressive variants have an unrelenting progressive course, with primary refractory disease proving to be the norm.

This article reviews the epidemiology of ATLL, as well as the clinical manifestations, diagnostic considerations, traditional approach to therapy, and recent data on novel therapeutic agents.

Introduction

Adult T-cell leukemia/lymphoma (ATLL) is a severe, yet fascinating, disease that develops in a small proportion of individuals who are chronically infected with human T-cell lymphotropic virus type 1 (HTLV-1). ATLL is epidemiologically unique because of this association with HTLV-1, which is the first described human retrovirus. Clinically, ATLL has protean manifestations, with each of 4 variants possessing a distinct constellation of symptoms. The aggressive variants have an unrelenting progressive course, with primary refractory disease proving to be the norm.

This article reviews the epidemiology of ATLL, as well as the clinical manifestations, diagnostic considerations, traditional approach to therapy, and recent data on novel therapeutic agents.

Epidemiology

The geographic distribution of ATLL is rooted in the epidemiology of HTLV-1. HTLV-1 infection is common, with an estimated 15 million to 20 million people worldwide living with chronic infection. These individuals predominantly live in endemic regions. The areas with an HTLV-1 seroprevalence of at least 1% to 5% include Japan, Iran, the Caribbean islands, Honduras, Brazil, Peru, Ecuador, and many west African nations ( Fig. 1 ). Specific communities have hyperendemic HTLV-1; notably the Shikoku and Kyushu islands of Japan, as well as the Okinawa Prefecture, have seroprevalences of up to 37%. These regions stand out even compared with other high-prevalence regions, such as Jamaica, which has a rate of 5%. In contrast, France has a very low seroprevalence, ranging from 0.004% to 0.007% in blood donors, as does the United States with a seroprevalence of 0.018%.

Geographic distribution of HTLV-1.

( From Goncalves DU, Proietti FA, Ribas JGR, et al. Epidemiology, treatment, and prevention of human T-cell leukemia virus 1-associated diseases. Clin Microbiol Rev 2010;23(3):579; with permission.)

HTLV-1 persists in these endemic regions primarily because of vertical transmission of the virus from mother to child via breastfeeding, with a transmission rate of roughly 20%. HTLV-1 can also be transmitted via blood transfusion, sexual contact, and parenterally via shared needles. Most patients who develop ATLL acquire HTLV-1 via breastfeeding, which is likely related to the early age of infection. Individuals who develop HTLV-associated myelopathy/tropical spastic paresis are more likely to have acquired the virus as a blood-borne pathogen. In Nagasaki, an HTLV-1 screening program that included prenatal testing and counseling on breastfeeding avoidance for HTLV-1–positive mothers decreased transmission of the virus by 80%. Blood transfusion was a major source of HTLV-1 transmission before universal screening measures.

In contrast with HTLV-1 infection, ATLL is a rare malignancy. The lifetime risk of a patient with HTLV-1 developing ATLL is 2% to 4%, and the average latency period from infection to ATLL onset is 60 years in Japan and 40 years in Jamaica. Worldwide, there were 2100 cases of ATLL in 2008, and most of these cases were clustered in the HTLV-1–endemic regions described earlier. ATLL is most prevalent in the Kyushu region of Japan, with an age-standardized incidence of 2 per 100,000 individuals in 2008. ATLL accounted for 36.8% of all lymphoma cases from 2003 to 2008 in Kyushu. ATLL is very common in the Kyushu region, even compared with Japan as a whole (0.3 per 100,000 individuals), and particularly compared with the United States where ATLL is exceedingly rare (0.02 per 100,000 individuals). In addition, ATLL accounts for only 0.2% of all lymphoma cases in the United States. However, the rate of ATLL seems to be increasing. From 2002 and 2011, there was an annual percentage change (APC) of 6.2%, which is a statistically significant increase. In Kyushu, where screening programs have helped decrease transmission of HTLV-1, the APC is 0%. The increasing trend in the United States is driven by increased emigration from endemic regions, with certain cities on the eastern seaboard reporting a higher incidence of ATLL than the rest of the country.

Mechanism

HTLV-1 is a human retrovirus that has the capacity to directly transform CD4-positive T cells. Mechanistic research, coupled with extensive epidemiologic, clinical, and experimental (animal) research, has led the International Agency for Research on Cancer (IARC) to classify HTLV-1 as a group 1 carcinogen, meaning this exposure is carcinogenic to humans.

Extensive basic research has identified that the viral transactivator protein, Tax, plays a key role in carcinogenesis. Tax activates the viral promoter, as well as the cyclic AMP and nuclear factor kappa-B pathways. In addition, antiapoptotic proteins are upregulated and p53 is suppressed. Most recently, researchers have shown that Tax can induce the epigenetic reprogramming of more than half the cellular genes via activation of enhancer of zeste homolog 2 (EZH2) and subsequent alteration of trimethylation at histone H3Lys27. Ultimately, a T cell is transformed into a cancer cell, and there is expansion of the malignant clone.

Tax protein expression is often undetectable in circulating ATLL cells, whereas the HTLV-1 basic leucine zipper factor (HBZ) is consistently expressed. In the study described earlier, the researchers were able to reproduce their findings via Tax transduction into normal T lymphocytes, whereas HBZ was not able to induce these epigenetic changes. In addition, the epigenetic profile persisted into the late stages of disease. Therefore, it seems that Tax plays an ongoing role in driving ATLL, but it is unclear whether this is caused by an initial activating event or continuous Tax expression, which may be possible if the methods of detection are lacking in sensitivity. Ultimately, these findings may have clinical implications because agents targeting Tax and/or EZH2 have shown preliminary efficacy.

Clinical presentation

The clinical presentation of ATLL is highly variable, but correlates with various subtypes of disease that were first described by Shimoyama ( Table 1 ). The 4 subtypes of ATLL are (1) acute, (2) lymphomatous, (3) chronic, and (4) smoldering. The chronic subtype is further divided into unfavorable and favorable, based on the presence or absence of several factors. The unfavorable risk factors are a serum lactate dehydrogenase (LDH) level greater than the upper limit of normal (ULN), serum blood urea nitrogen level greater than the ULN, and serum albumin level lower than the lower limit of normal. The smoldering, chronic, and acute types of ATLL can be viewed on a continuum of leukemic involvement, with the smoldering subtype representing the mildest form of the disease, and the acute type representing the most aggressive form of the disease. The lymphomatous subtype is also very aggressive, but lacks a leukemic component by definition. The smoldering subtype has atypical lymphocytes, termed flower cells ( Fig. 2 ), but does not have a lymphocytosis (absolute lymphocyte count [ALC] <4000 mcL). In addition, these patients have no lymphadenopathy (LAD), splenomegaly, or significant increase in LDH level. The chronic subtype differs from the smoldering type in that there is an increased ALC greater than or equal to 4000, and patients may have mild LAD, splenomegaly, or increased LDH level. The acute type has an atypical lymphocytosis and often presents with diffuse LAD, splenomegaly, significantly increased LDH level, and hypercalcemia. Lastly, the lymphomatous subtype often presents with rapidly growing LAD and splenomegaly without a leukemic component. The LDH level is typically markedly increased in the lymphomatous subtype, but hypercalcemia is less frequent. All subtypes of ATLL can have variable dermatologic manifestations ( Fig. 3 ).

Flower cells.

( From Jain P, Prabhash K. Flower cells of leukemia. Blood 2010;115:1668; with permission.)

Dermatologic manifestations of ATLL. Dermatologic findings include erythema ( A ) and papules ( B ).

( From Ohshima K, Jaffe ES, Kikuchi M. Adult T-cell leukaemia/lymphoma. In: Swerdlow SH, Campo E, Harris NL, et al, editors. WHO classification of tumours of haematopoietic and lymphoid tissues. 4th edition. Lyon (France): International Agency for Research on Cancer (IARC); 2008; with permission.)

The acute and lymphomatous subtypes are often grouped together as aggressive ATLL. Progression is rapid in these patients, and they often present in organ failure or respiratory compromise from rapidly enlarging lymph node masses and visceral organ invasion. Even after therapy is initiated, aggressive ATLL often progresses, resulting in primary refractory disease. In addition, patients with aggressive ATLL are at risk for central nervous system (CNS) involvement, and any neurologic deficits at presentation warrant further evaluation. As in other aggressive lymphomas, tumor lysis syndrome (TLS) is a concern, both spontaneously and after initiation of therapy.

A distinctive finding in ATLL is a profoundly increased calcium level at presentation. Calcium levels more than 20 mg/mL are common in acute-type ATLL. Such high calcium levels are less common in the lymphomatous, chronic, and smoldering forms of the disease. Pathologic studies have revealed that patients with ATLL with hypercalcemia often have an increased number of osteoclasts, as well as increased bone resorption. Parathyroid hormone–related protein has been suggested as a driver of bone resorption in ATLL, although increased transcription of receptor activator of nuclear factor kB (RANK) ligand seems to correlate more precisely in patients with ATLL with hypercalcemia. More recently, researchers have found that ATLL cells overexpress Wnt5a, which induces osteoclastogenesis, and this could be responsible for both the osteolytic lesions and hypercalcemia found in many patients with ATLL.

All patients with ATLL are immunosuppressed and at risk for infections, which frequently complicates therapy. In his initial description of the ATLL subtypes, Shimoyama quantified this risk. In the acute type, 27% of patients had an infection at diagnosis. The infection was most commonly bacterial (12%), followed by fungal (8%), parasitic (5%), and viral (3%). The nematode Strongyloides stercoralis is commonly identified in patients with ATLL, and causes severe, disseminated disease. Because of the risk of severe infection in patients with ATLL, prophylaxis against bacterial, fungal, and viral pathogens is appropriate and prompt diagnosis and management of infectious complications are paramount.

Making the diagnosis

The diagnosis of ATLL begins with the historical evaluation. Although it is possible for someone who is not from an endemic country to become infected with HTLV-1, it is highly unlikely; almost all patients with ATLL were born in an endemic country. Therefore, in evaluating a patient with either a leukemic or lymphomatous presentation, ATLL must be considered for any patient emigrating from an endemic country. A profound hypercalcemia (>20 mg/mL), although not 100% specific, can point toward a diagnosis of ATLL. In addition, about 20% of patients with ATLL have an eosinophilia, which can be severe. If there is a leukemic component, the diagnosis can be made if flower cells are visualized on the peripheral smear because these cells are pathognomonic for ATLL. Any case suspicious for ATLL must have HTLV-1 serologic testing. Although a positive test does not confirm ATLL, a negative test does rule-out ATLL. Flow cytometry sent on a peripheral blood sample is the confirmatory test of choice for any ATLL with a leukemic component. The characteristic flow cytometry pattern is positive for CD2, CD3, CD4, CD5, and CD25. Typically, CD7, CD8, and CD56 are negative, and CD30 is variable (positive in 20%–50%) ( Table 2 ). Lymphomatous presentations depend on an excisional lymph node biopsy for diagnosis, which should have flow cytometry and immunohistochemical (IHC) testing. Additional IHC tests include anaplastic lymphoma kinase (ALK), paired box 5 (PAX5) and terminal deoxynucleotidyl transferase (TdT) which are all negative in ATLL. The Ki-67 proliferation index is very high in aggressive ATLL. Bone marrow aspiration and biopsy may be performed to obtain a diagnosis or to complete the staging.

Further work-up

In addition to the complete blood count with differential, complete metabolic panel, and peripheral blood smear, all patients with ATLL should have a LDH test performed, as well as a TLS panel, including uric acid, phosphate, calcium, potassium, and creatinine levels. Glucose-6-phosphate dehydrogenase (G6PD) testing should also be sent with the initial work-up in order to evaluate for the presence of a hereditary deficiency that would preclude the use of the recombinant urate oxidase enzyme rasburicase. A viral panel consisting of human immunodeficiency virus, hepatitis B virus, and hepatitis C virus serology should be sent for all patients with leukemia/lymphoma. In addition, all patients with aggressive ATLL should have human leukocyte antigen (HLA) typing of their siblings in preparation for allogeneic stem cell transplant. This process should start immediately on diagnosis of aggressive ATLL, because the process can take time and remissions do not last long in this disease.

All patients should have imaging to evaluate the extent of lymphadenopathy, splenomegaly, organ infiltration, and skeletal involvement. Even patients with acute-type ATLL can have massive LAD that can impair organ function. Imaging with either computed tomography (CT) with intravenous contrast or PET is acceptable. Note that because of the rapid progression of this disease, treatment should not be delayed to obtain PET imaging unless it is readily available. In this situation, CT imaging is prudent. Aggressive ATLL often invades the CNS. Therefore, all patients newly diagnosed with either the acute or lymphomatous types of ATLL should have brain imaging (CT or MRI), as well as a lumbar puncture (LP) sent for cytology and flow cytometry. Intrathecal chemotherapy should be given at the time of the initial LP.

Prognosis

The prognosis for ATLL depends on the subtype, with the smoldering and chronic variants having improved survival compared with the aggressive subtypes ( Fig. 4 ). In addition, the chronic variant has been divided into favorable and unfavorable types, with the unfavorable type correlating with a poorer survival. Importantly, a recent study reexamined the original Shimoyama prognostic groups and found that the median survival times were 8.3, 10.6, 31.5, and 55 months for the acute, lymphomatous, chronic, and smoldering types, respectively. Four-year overall survival (OS) rates were 11%, 16%, 36%, and 52%, respectively. These data show an improvement in the 4-year OS for the aggressive subtypes, but the smoldering type had worse than expected survival.

Survival by ATLL subtype.

( From Shimoyama M. Diagnostic criteria and classification of clinical subtypes of adult T-cell leukaemia-lymphoma. Br J Haematol 1991;79:428–37; with permission.)

Efforts have been made to apply prognostic scores to ATLL. One study applied the commonly used International Prognostic Index (IPI) for lymphoma to patients with ATLL. This score takes into account 5 factors (age >60 years, stage III/IV, Eastern Collaborative Oncology Group [ECOG] score ≥2, increased LDH level, >1 extranodal region), with prognosis worsening with the presence of a greater number of factors. Most patients evaluated in the ATLL study had the lymphomatous subtype (89%). More specific prognostic scores have been evaluated, including the simplified ATLL prognostic index (ATL-PI). This study reviewed data from 807 patients with newly diagnosed acute and lymphomatous type ATLL. Five variables were identified that correlated with worse survival: stage III/IV (2 points), ECOG score 2 to 4 (1 point), age greater than 70 years (1 point), serum albumin level less than 3.5 g/dL (1 point), and soluble interleukin 2 receptor (sIL-2R) level greater than 20,000 U/mL (1 point). A low score (0–2 points) correlated with a median OS of 16.2 months, an intermediate score (3–4 points) correlated with a median OS of 7 months, and a high score (5–6 points) correlated with a median OS of 4.6 months.

In addition to prognostic scores, several studies have identified various other poor prognostic factors. Takasaki and colleagues evaluated the impact that visceral and cytologic abnormalities have on survival. In this study of 168 patients with ATLL, bone marrow involvement, skin involvement, and monocytosis correlated with poorer survival in patients with aggressive ATLL. Bone marrow involvement coupled with additional visceral organ involvement worsened prognosis further. Eosinophilia has also been associated with a poor prognosis.

CD30 positivity not only has therapeutic implications but also has prognostic importance in the acute and unfavorable chronic types. A study evaluating 68 patients with ATLL found that 22.1% of these patients were positive for CD30. In the acute/unfavorable chronic group, the median survival of CD30-positive patients was 10.1 weeks, whereas CD30-negative patients had a median survival of 33.7 weeks. CD30 was not a significant prognostic marker in patients with lymphomatous ATLL.