Acute lymphoblastic leukemia (ALL) accounts for 20% of all acute leukemias seen in patients over 20 years of age, and affects approximately 2 persons per 100,000 in the United States annually. Despite its relative rarity, ALL continues to generate considerable interest because of its high mortality when untreated, and because of the biologic and therapeutic lessons learned from studying this disease.
The success of therapy in childhood ALL has also fueled a quest for similar cure rates in adults, using intensive remission induction chemotherapy and the concepts of post-remission consolidation and maintenance therapies. So far, treatment for adult ALL has yielded inferior, but increasingly promising, complete remission rates of 75% to 90% and 3-year survival rates of 25% to 50%.1 Moreover, recent studies have contributed to an understanding of the biology of ALL and its prognostic factors [2-4]. Importantly, specific therapies directed at relatively homogeneous subgroups of ALL are now emerging. Parallel improvements in supportive care, such as antibiotic prophylaxis, the appropriate use of bone marrow transplantation, and perhaps the use of hematopoietic growth factors have also contributed to better survival rates.
The diagnosis of ALL continues to rely on morphology and cytochemistry. However, with the increasing availability of immunophenotyping by flow cytometry, this technique has become an important part of the diagnostic evaluation (Table 1). Thus, in the initial assessment of the acute leukemia patient, a relatively limited panel of monoclonal antibodies will allow patients to be divided into those with ALL derived from either B lymphocytes (B-lineage) or T lymphocytes (T-lineage) and those with acute myeloid leukemia (AML). The distinction between ALL and AML is the critical first step in the selection of therapy, and when this division is not correctly made, poor results are common.
Other diseases may be confused with ALL, the most common of which are listed in Table 2. Minimally differentiated AML (MO by the French-American-British Cooperative Group [FAB] criteria) and acute undifferentiated leukemia lack lymphoid surface markers, and are rarely cured with typical ALL treatment.
The diagnosis of AML-MO is often difficult to make, as it relies on morphology, cytochemistry, and immunophenotyping. These cells usually have FAB-L2 morphology and are negative for myeloperoxidase or Sudan Black B reactivity. Thus, patients with AML-MO are often mistakenly entered into ALL clinical trials, where their outcome has been poor. However, these cases, by definition, do not express lymphoid-specific markers. They may be positive for terminal deoxynucleotidyl transferase (TdT) reactivity and CD7 expression, but these markers are not specific for lymphoblasts. In contrast, AML-MO cells are positive for CD13, CD33, or other myeloid markers, suggesting a minimal level of differentiation along the granulocytic pathway.
Occasional patients may have hybrid acute leukemia, wherein the blast cells express both myeloid and lymphoid surface markers. They may be either bilineal, when such features are seen in separate cell populations, or biphenotypic, when they are seen on the same cell. In the former case, therapy is usually based on the dominant pattern.
In approximately 20% of cases of typical ALL in adults, the individual lymphoblasts express both myeloid and lymphoid antigens. These cases are more commonly B-lineage than T-lineage in origin. The myeloid antigen-positive ALL immunophenotype does not appear to be associated with a poor outcome in children, but the data are less clear in adults . However, recent improvements in treatment may have overcome the poor prognosis formerly associated with myeloid antigen expression in adult ALL [4,6].
Together, the use of morphology, cytochemical stains, immunophenotyping, and electron microscopy can reliably differentiate between AML and ALL in more than 95% of patients. Most cases of ALL are strongly positive for terminal deoxynucleotidyl transferase. Lymphoblasts that are negative for terminal deoxynucleotidyl transferase often have FAB-L3 morphology, and correspond to mature B-cell ALL. Such cases are sometimes called Burkitt cell ALL, because of their similarities with Burkitt's lymphoma. The lymphoblasts of these patients generally express surface membrane immunoglobulins, and the 8;14 translocation or one of its variants [t(2;8) or t(8;22)] is usually present.
Cytogenetic evaluation has become a critical part of the pretreatment evaluation of patients with ALL . Indeed, the most predictable clinical outcomes occur when patients are classified according to recurring cytogenetic abnormal- ities.4 The most common of these involves the Philadelphia chromosome . Originally described in chronic myelogenous leukemia, this rearrangement involves the translocation of the ABL proto-oncogene from chromosome 9 to the breakpoint cluster region (BCR) gene on chromosome 22: t(9;22)(q34;q11). In chronic myelogenous leukemia, this results in the production of the hybrid protein p210, but in Philadelphia-positive ALL, either a p210 or a smaller p190 protein results.
The importance of recognizing this subgroup lies in the considerably shorter survival observed in both childhood and adult Phildelphia-positive ALL cases. Although fewer than 5% of childhood ALL cases are positive for this chromosome, the frequency of positivity increases steadily with age, and approximately 30% of adults with ALL have Philadelphia-positive disease [3,8]. Indeed, the poorer prognosis of adult ALL overall may be due, in part, to the proportionately higher number of Philadelphia-positive cases seen among adults.
Other karyotypes that occur in ALL and have important prognostic significance include t(8;14), t(4;11), and t(1;19). The poorer prognosis associated with certain karyotypes has dictated that different approaches be taken with these patients, as discussed below.
Table 3 lists the adverse prognostic factors that have a major influence on complete remission rates, remission duration, and survival in patients with ALL [1,4]. In multivariate analyses, patients presenting with white blood cell counts > 30,000/µL have had a significantly shorter duration of remission than patients with lower leukocyte counts. However, among patients with T-cell ALL, extreme leukocytosis does not negatively affect outcome .
Older age (> 60 years) is another adverse characteristic. Remission duration and overall survival have decreased in almost every adult ALL trial as the ages of the patient groups have increased. Minor prognostic factors, or those that have had some significance with certain treatment regimens, are the percentage of circulating blast cells; the degree of bone marrow involvement; the presence of hepatomegaly, splenomegaly, or lymphadenopathy; lactate dehydrogenase levels; central nervous system (CNS) involvement at presentation; and the time required to achieve complete remission (eg, > 4 to 6 weeks).