Approximately half of all adults with acute lymphoblastic leukemia now survive long term. This article summarizes the current approaches to treating acute lymphoblastic leukemia in adults, with a focus on a pragmatic approach to decision making. Coupled with a particularly punishing and often complex combination chemotherapy treatment regimen, treatment-related morbidity and mortality are frequent, and this article focuses on these situations. The field will change significantly over the next few years with many ongoing clinical studies and molecular insights which will be translated into providing prognostic information and novel therapeutic targets.
Overview
Approximately half of all adults with acute lymphoblastic leukemia (ALL) now survive long term. An elegant study of unselected registry data, in which point estimates of survival were made for two 5-year time periods, 20 years apart, demonstrated highly significant 14% to 20% survival improvements for each age group except the over-60 group, in which no significant improvement in outcome had occurred. This article summarizes the current approaches to treating ALL in adults, with a focus on a pragmatic approach to decision making, based on available data. A major problem in treating adults with ALL is that few physicians or institutions have a large personal practice, because the disease is rare. Coupled with a particularly punishing and often complex combination chemotherapy treatment regimen, treatment-related morbidity (TRM) and mortality are frequent and individual patient decisions on how to best balance efficacy with toxicity can be difficult. This article focuses on such situations. As examples, dealing with the toxicity of induction regimens and treating older people with ALL are both scenarios that vex clinicians and areas in which there are few conclusive answers in the literature. In many situations, it can be concluded that there is still “no right answer.” Thankfully, there is a vibrant academic interest in ALL, both scientifically and clinically. The field will change significantly over the next few years as many ongoing clinical studies report and molecular insights are translated into providing prognostic information and novel therapeutic targets. Monoclonal antibodies are likely to make a considerable contribution to the treatment of ALL and are discussed in a separate article.
Diagnosis
ALL is a medical emergency. It should be diagnosed and treated without delay. A bone marrow aspirate should be examined morphologically by an expert hematopathologist. Bone marrow—or peripheral blood, if the blast count is high—should be examined using a panel of monoclonal antibodies to T-cell–associated and B-cell–associated antigens, which identify almost all cases of ALL. Aberrant expression of myeloid antigens is not uncommon and should not deflect from the correct diagnosis. In the differential diagnosis, blastic transformation of chronic myeloid leukemia should be specifically ruled out by morphologic examination. Trephine biopsy examination is sometimes helpful, but the result is not required before starting treatment. Cytogenetic examination of the blast cells is mandatory and should comprise both examination of metaphases and fluorescence in situ hybridization with specific probes (eg, for B CR -ABL and MLL-AF4L). Screening by polymerase chain reaction (PCR) for the potential B CR -ABL transcripts, p190 and p210, should be performed. MLL-AF4 translocations can also be sought by PCR. Standardized primer sets are specified. It is important that a specimen also be examined by molecular methods for the detection of patient-specific immunoglobulin and T-cell receptor (Ig/TCR) rearrangements or by flow cytometry to detect a specific immunophenotype, both tests can be used for minimal residual disease (MRD) quantification. If this is not performed at diagnosis, the opportunity to quantify MRD after therapy is lost unless a patient has a specific marker, such as B CR -ABL. At present, quantification of Ig/TCR rearrangements is the only standardized method for detection of MRD. If bone marrow transplant is a possible part of a patient’s future therapy, tissue typing of any patient siblings who are willing to be typed should also be performed at diagnosis. If there are no HLA-matched siblings, consideration should be given to prompt initiation of an unrelated donor search.
Diagnosis
ALL is a medical emergency. It should be diagnosed and treated without delay. A bone marrow aspirate should be examined morphologically by an expert hematopathologist. Bone marrow—or peripheral blood, if the blast count is high—should be examined using a panel of monoclonal antibodies to T-cell–associated and B-cell–associated antigens, which identify almost all cases of ALL. Aberrant expression of myeloid antigens is not uncommon and should not deflect from the correct diagnosis. In the differential diagnosis, blastic transformation of chronic myeloid leukemia should be specifically ruled out by morphologic examination. Trephine biopsy examination is sometimes helpful, but the result is not required before starting treatment. Cytogenetic examination of the blast cells is mandatory and should comprise both examination of metaphases and fluorescence in situ hybridization with specific probes (eg, for B CR -ABL and MLL-AF4L). Screening by polymerase chain reaction (PCR) for the potential B CR -ABL transcripts, p190 and p210, should be performed. MLL-AF4 translocations can also be sought by PCR. Standardized primer sets are specified. It is important that a specimen also be examined by molecular methods for the detection of patient-specific immunoglobulin and T-cell receptor (Ig/TCR) rearrangements or by flow cytometry to detect a specific immunophenotype, both tests can be used for minimal residual disease (MRD) quantification. If this is not performed at diagnosis, the opportunity to quantify MRD after therapy is lost unless a patient has a specific marker, such as B CR -ABL. At present, quantification of Ig/TCR rearrangements is the only standardized method for detection of MRD. If bone marrow transplant is a possible part of a patient’s future therapy, tissue typing of any patient siblings who are willing to be typed should also be performed at diagnosis. If there are no HLA-matched siblings, consideration should be given to prompt initiation of an unrelated donor search.
Prognostic factors
Many factors that can be identified at—or soon after—diagnosis have a bearing on outcome ( Table 1 ). These prognostic factors often form the basis for treatment decisions in ALL, although there is little convincing evidence that currently available therapeutic strategies other than allogeneic haematopoietic stem cell transplantation (alloHSCT) are able to overcome the adverse factors. When planning matching a potentially effective—but toxic—therapy, however, such as alloHSCT, a high risk of death as a result of ALL can be balanced against a more risky treatment strategy. Decisions of this nature are used in clinical practice and as tools to stratify patients within clinical trials. In contrast to pediatric practice, examination of prognostic factors cannot yet define a set of adults who have a particularly good prognosis. Hence, there is currently no strategy with which to limit therapy for adults on the basis of an expectation of particularly good outcome.
| Factor | Detail | Selected References |
|---|---|---|
| At diagnosis | ||
| Age | Worse outcome with advancing age—no clear age cutoff in adults | |
| Presenting white blood cell count | >30 × 10 9 /L (B), >100 × 10 9 /L (T) | |
| Immunophenotype | In adults, T ALL can have a better outcome than B ALL CD20 expression has been associated with a less good outcome | |
| Cytogenetics | Poor: t(9;22), t(4;11), complex (>5 abnormalities), low hypodiploidy near triplody | |
| Specific molecular abnormalities | JAK2 IKFZ1 PAX5 | |
| Response to therapy | ||
| Steroid responsiveness | Response to steroids has clear relationship with outcome in childhood ALL. Less well defined and tested in adult ALL. | |
| Speed of initial response | Rapid initial response—CR within 4 weeks predicts better outcome. Not uniformly demonstrated. | |
| Minimal residual disease | Clear relationship between MRD at protocol-specific time points and outcome in several studies. | |
Minimal Residual Disease
The measurement of MRD deserves a specific discussion due to its pivotal role in the management of ALL. Long accepted as vital for management in pediatric practice, it has been conclusively demonstrated as carrying the same important prognostic information in adult ALL. Its adoption into standard practice, however, has been less quick.
MRD can be quantified by both molecular quantification of patient-specific Ig/TCR rearrangements : by quantification of specific molecular abnormalities, such as B CR -ABL or MLL-AF4, or by flow cytometry, where leukemia-specific immunophenotypes can be identified and quantified after initial treatment.
The molecular techniques are sensitive and specific and have a sensitive range of approximately 1 in 100,000 cells. They can be accurately quantitative at the level of 1 in 10,000 cells. Molecular and flow cytometric techniques are patient specific and require a good-quality diagnostic sample. Both techniques, when performed properly, are labor intensive and expensive to carry out. Both techniques should be performed in reference laboratories according to standardized protocols. All techniques require a diagnostic sample. For treating clinicians, their clinical meaning is protocol dependent. MRD in adult ALL is an excellent example of what can be achieved by interlaboratory cooperation; the European Study Group overseen by VanDongen is a successful and thriving association of more than 30 laboratories from different countries that run standardized protocols for molecular MRD determination and are also working on joint strategies for standardized flow cytometric techniques.
A particular attraction of determination of MRD in adult ALL is that it can give information on patients d otherwise classified as “standard risk.” An impressive illustration of this concept was provided by the German Multicenter Study Group for Adult Acute Lymphoblastic Leukemia (GMALL). Bruggeman and colleagues demonstrated that quantification of MRD separates those who will have a poor outcome from those who will do better at several time points along the standard treatment protocol. It is a protocol-specific choice, however, as to which time point should be chosen for an intervention or intensification of therapy, such as bone marrow transplantation. At 1 year into treatment, for example, MRD-positive patients fare poorly but it is likely too late to intervene.
Another potential use of MRD is in post-therapy monitoring. The concept of molecular relapse is a realistic one in ALL, as shown by a recent study from Raff and colleagues, also on behalf of the GMALL. Because the median time from MRD reappearance to clinical relapse was 9.5 months, an opportunity for intervention was offered. It remains to be seen what interventions would be appropriate in this situation.
Strategies during induction treatment
Induction Therapy—General Points
The aim of initial treatment is to achieve complete remission (CR), which is currently defined on a morphologic basis as less than 5% blasts in the bone marrow in the presence of overall hematopoietic recovery. Because treatments of ALL have evolved, an increasing intensity of therapy has been applied to achieve CR. As a result, serious adverse effects are commonplace and treatment-related mortality is a significant risk. Hence, the highest standard of supportive care is imperative.
Supportive care
Before starting any treatment, patients should be well hydrated and receive appropriate medication to prevent urate nephropathy. ALL can be exquisitely sensitive to small doses of treatment and tumor lysis syndrome is not uncommon, especially in patients with bulky extramedullary disease. Allopurinol should be started 24 hours before induction chemotherapy and should be continued for a minimum of 5 days. Rasburicase should be considered an alternative to allopurinol if the white blood cell count is high (ie, >100 × 10 9 /L) if a patient has bulky disease (eg, large mediastinal mass or elevated urate at diagnosis). Adults with ALL receive high doses of myelosuppressive chemotherapy along with steroids. They are, therefore, at high risk of infectious complications particularly during the induction phase of treatment. Adherence to strict anti-infective strategies is important. All patients should receive prophylaxis against herpes simplex virus and varicella-zoster virus reactivation. Generally, patients are given acyclovir throughout therapy, although local policies may be followed.
All patients need prophylaxis against Pneumocystis jeroveci from the start of therapy. The recommended prophylaxis is co-trimoxazole (twice a day for 2 or 3 days each week), avoiding any day on which methotrexate (MTX) is given. In the event of the allergy to co-trimoxazole, alternative prophylactic agents include nebulized pentamidine or dapsone.
Antifungal prophylaxis is strongly recommended for all patients on ALL therapy from the time of induction. Azoles must be avoided when a patient is on vincristine due to potentiation of neurotoxicity. There is no clear evidence to suggest which antifungal prophylaxis regimen should be used in this situation. An international randomized controlled trial comparing liposomal amphotericin with placebo is under way. Antifungal prophylaxis is not generally required when a patient is on maintenance therapy unless that patient is deemed at high risk for fungal disease.
The use of granulocyte colony-stimulating factor is strongly recommended for all patients to hasten neutrophil recovery, particularly after induction.
Steroid prephase
Corticosteroids are among the most important drugs in the treatment of ALL, and recent trials in pediatric ALL patients have suggested that the use of dexamethasone, as opposed to prednisolone, may improve outcome. This is based on data suggesting that dexamethasone has greater in vitro antileukemia activity than prednisolone, has better penetration of the central nervous system (CNS), and causes fewer thromboembolic events. Randomized trials have demonstrated improved survival in children receiving dexamethasone as opposed to prednisolone, although this has not been shown in every study. Based on these data, many adult regimens substitute a discontinuous schedule of dexamethasone for prednisolone during induction.
Induction therapy
The primary goal of induction therapy is a complete eradication of ALL cells from blood, bone marrow, and CNS or other extramedullary sites (when initially involved). This should be achieved as early as possible and with as few toxic side effects as possible to start rapidly the postremission therapy. Most regimens use phases of induction, often called phases 1 and 2 or parts a and b, with the second part applied regardless of CR after induction 1. Many examples of highly effective induction protocols have been reported, all of which result in CR rates of 90% or more. Due to the complexity of existing induction regimens and the fact that it is already possible to obtain a CR in 90% to 95% or more of unselected adult patients, evaluation of any new treatment elements during induction requires either the documentation of an improvement in event-free survival (EFS) or in overall survival (OS) or evaluation of an alternative, precisely quantitative, endpoint such as MRD.
Induction regimens in adult ALL are composed of steroid, vincristine, anthracycline (often daunorubicin or doxorubicin) and L-asparaginase (L-asp). Cyclophosphamide and cytarabine may also be included. Because more anthracycline is used in adult practice than in pediatric practice, prolonged myelosuppression is common. The intensity of anthracycline for maximal benefit during induction has already been reached and studies attempting to improve outcome by further intensification have not shown enhanced outcomes.
Dealing with Problems and Delays During Induction Therapy
L-asparaginase and coagulation problems
Induction treatment is frequently complicated by problems that relate to a particular drug used. The most common reported cause of severe side effects of therapy relate to the bacterial enzyme L-asp. L-asp is regarded as one of the most important drugs in the treatment of ALL but is also one of the most difficult drugs to dose and manage. Expert panel guidance has recently been made, although the evidence base for the management of L-asp toxicity in adults is scant.
A major problem with L-asp is the occurrence of abnormalities of coagulation. The management of the coagulation issues during treatment with L-asp varies widely worldwide and there is little clear evidence base particularly in the management of adult ALL. Although L-asp is associated with deranged coagulation as measured in the laboratory, thrombosis is the most commonly encountered clinical problem. Data indicate that thrombosis occurs in 10% to 20% of adults who are receiving L-asp therapy. There is little or no evidence that infusion of fresh frozen plasma can appropriately correct regulation abnormalities nor can it prevent thrombosis or bleeding events. In the French CAPELAL study, the mean antithrombin and fibrinogen levels increased from 61% to 88% and from 1 g/L to 1.4 g/L after infusion of antithrombin or fibrinogen, respectively, whereas both levels remained unchanged after the infusion of fresh frozen plasma. Additionally, there is some evidence that infusion of fresh frozen plasma can contribute to the replenishment of the asparginase pool that the drug is used to deplete. Hence, using fresh frozen plasma to replenish coagulation factors does not seem a sensible option. Fibrinogen concentrates can be used to elevate fibrinogen levels more successfully. A median dose of 0.03 g/kg was required in the CAPELAL study to increase levels from 1.0 ± 0.3 g/L to 1.4 ± 0.45 g/L. Infusion of fibrinogen, however, was not associated with a lower incidence of bleeding complications and—unlike thrombotic events—hemorrhagic events were not associated with any difference in patients’ survival. There is some evidence that antithrombin concentrate can reduce the number of thromboses that occur during L-asp therapy. In a clinical study of adults undergoing ALL treatment, the occurrence of thrombosis was associated with a less good outcome, even though none of the thrombotic events, in and of themselves, was fatal. One explanation for this may be that thrombosis resulted in delay to or reduction in dose of therapy. When patients treated for ALL experience a thrombosis, it can be difficult to decide whether to continue with L-asp treatment. Because the greatest risk of thrombosis is during induction therapy, cessation of one of the major components of the treatment strategy at that stage is likely to have an adverse impact on outcome. Strategies to reduce the risk of thrombosis are important. The presence of a central venous catheter considerably increases the risk of thrombosis and many treating centers delay placing indwelling central venous catheter until patients have reached CR at the end of the first phase of induction therapy. When catheter-associated thrombosis occurs, low-molecular-weight heparin is the treatment of choice. When a patient experiences a central nervous system thrombosis, the use of low-molecular-weight heparin is more contentious—patients treated with heparin should receive appropriate monitoring, because a proportion of patients have heparin resistance due to the depletion of antithrombin. When patients experience thrombosis, replacement of antithrombin is likely valuable. In the French CAPELAL study, patients who received antithrombin concentrates less frequently had delays or omissions to L-asp and a lower rate of thrombosis.
There are several preparations of L-asp, which are variously licensed and marketed in different countries. In pediatric practice, a pegylated version of the Escherichia coli enzyme is commonly used and there is increasing evidence of its usefulness. The advantage of this agent is that it has a long half-life; thus, administration results in prolonged asparagine depletion. A phase 2 study demonstrated that this drug can be given successfully to adult patients and was able to adequately deplete asparagine. Increasing age, however, was associated with a less successful asparagine depletion as a result of fewer doses and more side effects. The role of pegylated L-asp and the optimal dose are not established in adults and are the subject of more than one clinical trial.
Liver function abnormalities and pancreatitis
It common for patients to develop transaminitis and hyperbilirubinemia during induction therapy. Again, L-asp is often implicated in the pathogenesis. Experiencing severe sepsis exacerbates the problem. Most patients recover from this liver insult given time, but severe delays to therapy often result, compromising overall efficacy.
Pancreatitis is a life-threatening complication of L-asp treatment, seen more often in children than adults. It is a contradindication to further use of the drug.
Postremission treatment
CNS-Directed Prophylaxis
The presence of leukemic blasts in the CNS is a more frequent occurrence in patients with ALL than acute myeloid leukemia and the importance of prophylactic treatment to prevent the development of progression to the CNS has long been recognized. The role of prophylactic cranial irradiation in the era of combined intrathecal and high-dose systemic therapy has been questioned in recent studies, with the intent of reducing the risk of late sequelae. This is of particular interest to pediatricians because the late effects of cranial irradiation in children are well documented. Trials in children have already demonstrated that CNS irradiation can be eliminated without worsening overall outcome. In adults, several trials have reported CNS recurrence rates of less than 10% with the combined use of high-dose systemic and intrathecal chemotherapy without the use of cranial irradiation; thus, it is not considered an essential component of therapy.
MTX is the most common chemotherapy drug used for CNS-directed prophylaxis. It is by no means without specific toxicity, however, even when delivered at low doses via the intrathecal route. MTX-related encephalopathy is a particular problem, although less common in adults than in children. It presents with fits, focal neurologic deficit, or impaired consciousness and typically occurs within 1 day to approximately 3 weeks of exposure to intrathecal MTX. Full recovery is usual but the clinical presentation is dramatic and can be terrifying for patients. Diagnosis is often one of exclusion; other causes of CNS events should be considered and ruled out, such as cerebrovascular events, sagittal sinus thrombosis, infections, or CNS involvement with ALL. There may be typical findings on MRI scanning. In the event of this complication, the role of future MTX should be questioned. Future MTX should be discontinued when patients are also receiving cytarabine systemically. Safe rechallenge in other circumstances may possible without recurrence of toxicity, but if recurrence happens, the intrathecal regimen should be changed to cytarabine (50 mg) in association with hydrocortisone (12.5 mg). Local policy may be followed for the hydrocortisone dose if necessary.
Consolidation and Maintenance Therapy
Consolidation therapy typically consists of several cycles of treatment similar to but often less intensive than those given during induction. Consolidation is typically better tolerated than induction, although specific practices during consolidation are poorly studied. In line with pediatric style approaches, consolidation often contains one or more blocks of delayed intensification, in which the intensity of therapy is enhanced. Again, this approach has been adopted to a variable extent, although systematic study of intensification of therapy in consolidation has showed little or no evidence for benefit in adults, suggesting that a focus on improvements to initial therapy holds more promise.
Maintenance therapy remains obligatory in those not undergoing alloHSCT. Daily mercaptopurine, weekly MTX, and pulses of vincristine and steroids for 18 to 24 months after consolidation are standard. The composition and duration of maintenance therapy has never been the subject of a specific study in adults and has not changed in approach for many years. There is now randomized controlled trial evidence, however, of the value of prolonged chemotherapy compared with a shorter, more-intensive approach in adult ALL. In the Medical Research Council (MRC) UKALL12/E2993, patients without a matched sibling donor were eligible for randomization between high-dose therapy with etoposide and total body irradiation (TBI) with autologous HSCT (autoHSCT) rescue and maintenance therapy. The intent-to-treat analysis of 456 randomized patients showed that those randomized to prolonged chemotherapy had significantly superior EFS (41% vs 32%; P = .02) and OS (46% vs 37%; P = .03) at 5 years compared with those randomized to autoHSCT. The TRM did not differ between the groups. AutoHCST simply provided less-adequate disease control than prolonged chemotherapy.
Allogeneic Hematopoietic Stem Cell Transplant
alloHSCT has been extensively investigated in the treatment of ALL. Most physicians recommend alloHSCT for patients with high-risk ALL who are of suitable age and performance status where a donor is available. Patients receiving fully matched unrelated donor stem cells probably have an equivalent outcome in terms of toxicity and TRM as those receiving sibling stem cells from matched siblings—several studies have addressed this and although disease-free survival (DFS) is hard to compare with studies of sibling alloHSCT, TRM seems similar. In a single-center study of 84 patients with high-risk ALL, a considerable proportion of whom were beyond CR1, in which almost all patients received a TBI-based conditioning regimen, TRM did not differ significantly and in which almost all patients received either sibling or unrelated donor stem cells. The median age of patients in this study, however, was only 23 years and the study included children. A National Marrow Donor Program study of 127 patients in 46 centers demonstrated a high TRM for unrelated donor alloHSCT. The low relapse rate, however, resulted in a 37% ± 13% DFS for patients in CR1, which compared favorably at the time to results obtained with chemotherapy alone and those after HLA-identical sibling alloHSCT. A similar study in a smaller number of centers included a larger number of patients (N = 221, 72 of whom were Philadelphia positive [Ph+]) and a median age more representative of an adult population also showed no difference in TRM between matched sibling and unrelated donor alloHSCT. Again, patients beyond CR1 were included, which is likely to adversely influence the TRM, which was high (43% for sibling alloHSCT and 50% for unrelated donor alloHSCT).
The succinct conclusion, however, that adults with high-risk ALL should be offered alloHSCT in ALL—although true in overarching conceptual simplicity—oversimplifies the situation considerably and does little to help with clinical management of individual patients, because as new risk factors emerge, transplant conditioning regimens and supportive care change, chemotherapy protocols develop and improve, and novel agents become available. At the same time, past data and trials involving alloHSCT cannot take into account current insights. As an example, what constitutes high risk now includes those who are MRD positive at certain time points within the protocol. Conversely, the high risk of B CR -ABL positivity may be mitigated to some extent by the use of tyrosine kinase inhibitor (TKIs)—in children, this has provoked interest in managing pediatric patients who are B CR -ABL positive without alloHSCT (discussed later, in section on “Philadelphia-positive ALL”). Furthermore, a “suitable age” for alloHSCT is also the subject of debate. As the results for chemotherapy alone regimens continue to improve outcomes for adolescents and young adults, there is increased potential for better long-term OS without the toxicity of alloHSCT. It is against this fluid background that data on alloHSCT must be interpreted and decisions made for individual patients. Many trials have been conducted incorporating autoHSCT and alloHSCT into the treatment algorithm. There are several studies of myeloablative therapy followed by sibling alloHSCT. In all, the data indicate that in selected individuals, DFS or OS seems better than expected with treatment with chemotherapy alone. The strongest direct experimental support for alloHSCT in ALL, however, comes from 3 large studies from France, the United Kingdom and United States, and the Dutch-Belgian Hemato-Oncology Cooperative Group (HOVON). In all the studies, the existence of an allogeneic donor among those eligible for alloHSCT was independently predictive of remission duration; demonstrations of how so-called biologic randomization or donor versus no donor analysis can be used.
The largest study of alloHSCT, UKALL12/E2993, evaluated the outcome of Philadelphia-negative patients assigned to HLA-matched sibling alloHSCT compared with patients randomized to autologous SCT or chemotherapy. In a comparison of 389 patients with a donor to 530 patients without a donor, the donor groups had superior EFS (50% vs 41%; P = .009) and OS (53% vs 45%; P = .02). A similar statistically significant benefit was seen when the no donor group was restricted to those who were randomized to the chemotherapy arm only. This benefit was primarily seen in the standard-risk patients (OS 63% for donor vs 51% for no donor patients; P = .01) but not in high-risk patients (OS 40% vs 36%; P = .6). The lack of difference in outcomes between donor and no donor patients in the high-risk group was related to a high nonrelapse mortality of 39% at 2 years (20% at 2 years for the standard risk patients), which in large part was seen in older patients.
Meta-analysis of abstract data also shows an advantage for myeloablative alloHSCT. An additional justification for considering alloHSCT in high-risk patients first in CR (CR1) is the dismal outcome of patients who relapse from CR1 (discussed later).
Currently, the role of myeloablative alloHSCT in adults with ALL could be considered fluid and decisions may change as new data emerge. At the time of this writing, for patients in their early 20s and 30s in whom a sibling donor is available, results achieved can be superimposed on those achieved using a young adult chemotherapy approach. Data from 2 recent studies illustrate this point. In a Spanish study group Programa para el Estudio de la Terapéutica en Hemopatía Maligna (PETHEMA) trial, a pediatric-style approach was used for adults ages 19 to 30 with standard-risk disease—a CR rate of 98% was obtained and the 3-year OS was 63%. An almost identical, superimposable outcome for patients of the same age group (15–35) was shown in the UKALL12/E2993 study, in which patients received adult induction followed by sibling alloHSCT; the CR rate was 95% and 5-year OS was 62%.
Choice of Myeloablative Conditioning Regimen
The Stanford conditioning regimen of high-dose etoposide and TBI is often used because, when published, it demonstrated superior results in alloHSCT for ALL. There are no randomized controlled studies, however, comparing conditioning regimens; the optimal conditioning regimen has not been demonstrated. A retrospective analysis of data from selected centers suggested that TBI-based conditioning likely results in improved DFS by comparison with non-TBI, busulfan-containing regimens. The most commonly used regimens in conjuction with TBI are cyclophosphamide or etoposide. A small study suggested no difference between these 2 TBI-based regimens in childhood ALL. To date, the best available evidence comes from a retrospective analysis of International Blood and Marrow Transplant Research (IBMTR) data. No difference in relapse risk or OS was demonstrated between conditioning regimens containing either cyclophosphamide or etoposide when alloHSCT was performed in CR1. There was a modest advantage to using etoposide and higher doses of TBI when the alloHSCT was in second CR (CR2). The major acute toxicity of the etoposide-containing regimen is severe mucositis. One aspect of the UKALL14 study will examine whether amelioration of the mucositis by using the keratinocyte growth factor pallifermin might have a beneficial effect on graft-versus-host disease by facilitating delivery of the full dose of MTX prophylaxis. A role for T-cell–depletion unrelated-donor alloHSCT has not been defined. A British Society of Blood and Marrow Transplantation study shows an excellent outcome for high-risk ALL treated with alemtuzumab in vivo as part of the conditioning regimen. Conversely, an IMBTR study showed a less good outcome for those who received T-cell depletion as part of their therapy. Whether the use of alternative donor options, such as haploidentical stem cells or umbilical cord blood, is justified is a relevant question in high-risk ALL, but a full discussion is beyond the scope of this article.
Nonmyeloablative, Reduced-Intensity Conditioning
No prospective studies of transplant using reduced-intensity conditioning (RIC) have been reported to date. Published reports are subject to the considerable bias. To compound interpretation of the data, many series include patients beyond CR1. Only one study has compared the outcome of patients receiving reduced intensity as opposed to myeloablative conditioning—multivariate analysis showed that conditioning intensity did not affect transplantation-related mortality ( P = .92) or relapse risk ( P = .14), adding considerable justification for the further study of this approach. There are a plethora of reports of RIC alloHSCT in adult ALL. The European Group for Blood and Marrow Transplantation reported 97 patients who received various different RIC regimens, many of which were delivered in conjunction with some form of T-cell depletion. A 2-year OS of 52% for those transplanted in CR1 was reported. A GMALL examined the outcome of a mixed group of 22 patients with high-risk ALL receiving nonmyeloablative alloHSCT, 11 of whom had Ph+ disease. Half were beyond CR1. Few patients survived long term and mortality was high, but for many patients it was a second alloHSCT after relapse, making a good outcome unlikely. Another retrospective study included 27 patients from 4 different studies who had undergone nonmyeloablative alloHSCT. More than 80% of the patients whose median age was 50 years were beyond CR1. Two-year OS was 31%. Treatment-related mortality was modest for such a high-risk population, at 23%. A City of Hope study reported on 24 patients with adults with high-risk ALL treated with fludarabine and melphalan conditioning without T-cell depletion. Approximately half of the patients were over 50 years of age; there was a 2-year OS of 61.5% and DFS of 61.5%, with a treatment-related mortality of 21.5%. Bachanova and colleagues reported a 3-year OS of 50% among 22 patients, median age 49, all with high-risk ALL. Patients received a uniform reduced-intensity approach of fludarabine, cyclophosphamide, and low-dose TBI in the University of Minnesota Transplant Program. Nonmyeloablative allogeneic HSCT approaches are promising but require careful prospective study required to define their role in Ph+ ALL. The forthcoming study from the UK National Cancer Research Institute, UKALL14, will assign all patients with ALL aged 40 or older to a nonmyeloablative approach with fludarabine, melphalan, and alemtuzumab in an attempt to reduce the high incidence of graft-versus-host disease (86%), which occurred in the City of Hope report, in which fludarabine and melphalan was used without T-cell depletion. A recent report from the Fred Hutchinson Cancer Research Center Seattle underscores an important point that the level of disease at the time of alloHSCT might be of more relevance when RIC is used. In the series of 51 patients with high-risk ALL conditioned with a low-dose TBI approach, the 3-year OS rate was 62%; for the subgroup without evidence of MRD at transplantation, the OS was 73%.
Specific scenarios in ALL therapy
CNS Involvement
When the CNS is involved at diagnosis, specific therapy needs to be given. Systemic chemotherapy, cranial irradiation, and intrathecal therapy can all be used to help control CNS ALL. Although cranial irradiation as CNS-directed prophylaxis is arguably dispensable in the presence of prolonged intrathecal therapy, it is undoubtedly an effective form of therapy for established CNS leukemia and should be given strong consideration.
MTX, cytarabine, and glucocorticoids can all be safely given by the intrathecal route. Intrathecal therapy should be given once or twice per week until resolution of any CNS signs and symptoms and clearance of blasts from the cerebrospinal fluid (CSF) has occurred. There have been no benefits demonstrated by giving triple intrathecal therapy compared with MTX alone. When administering intrathecal medication there are some practical considerations: it should be given in sufficient volume to distribute well throughout the CSF and patients should remain recumbent for at least 1 hour (some recommend 6 or even 12 hours) after treatment. Because cytotoxic concentrations are not maintained in the CSF for long, frequent intrathecal administrations are needed, which can present technical difficulties. A liposomal preparation of cytarabine is available, which results in an extended concentration of cytarabine in the CSF. This should not be given during treatment with high-dose, CNS-penetrating treatment because there are reports of serious neurotoxicity. Whatever standard drug or route of intrathecal administration is given, therapy should be given once or twice per week until resolution of any CNS signs and symptoms and clearance of blasts from the CSF.
High doses of MTX, cytarabine, and glucocortocoids cross the blood-brain barrier. Many regimens have used MTX at doses ranging from 0.5 g/m 2 to 5 g/m 2 and meta-analysis of 43 randomized controlled trials showed a benefit to this approach in terms of reduction in relapse risk and improved event-free survival athough only a small effect on CNS disease. Rescue with folinic acid is required at an appropriate time (not too early to reduce efficacy) as is careful determination of systemic MTX levels to minimize systemic toxicity, in particular nephrotoxicity. Because systemic steroids also cross the blood-brain barrier, they may also constitute good agents for inclusion in the therapy for CNS disease—it was reported that dexamethasone was superior to prednisolone in reducing CNS relapse risk. This may have been due, however, to lack of dose equivalence between the 2 agents.
What is the effect on outcome when the CNS is involved at diagnosis? In the MRC UKALL12/E2993 trial, 77 of 1508 (5%) patients had CNS involvement at diagnosis. The incidence was higher in patients with T-cell ALL. Of these 77 patients, 69 (90%) achieved CR. This study demonstrated that although long-term DFS is attainable in patients who present with CNS involvement, OS at 5 years was inferior at 29% compared with 38% for patients without CNS involvement ( P = .03).
ALL in Younger Adults
Several retrospective studies have shown that an intensive pediatric approach to therapy with high doses of steroid and L-asp can confer survival superior to an adult approach in older teenagers and young adults. Although such retrospective studies are subject to considerable selection bias, the weight of evidence favors evaluating such a therapeutic approach carefully in older adults. As a consequence, the field has tended to become somewhat polarized, with chemotherapy and bone marrow transplant viewed as opposing therapeutic candidates. The outcome of this debate and shift in emphasis in terms of long-term OS and long-term toxicity remain to be determined and it is likely both approaches will play a role in continued improvement of outcomes.
It is becoming clearer that pediatric-style intensive therapy can be delivered to patients who are by no means adolescents—at least up to age 30—with a reasonably good outcome of approximately 60% OS at 5 years. This outcome is superimposable, however, on the outcome seen for patients of the same age and risk status with a sibling donor in the UKALL12/E2993 donor versus no donor analysis. To date, there are currently no data to suggest that the poor prognostic relevance of being over 40 can be overcome by a pediatric therapeutic approach but this is an active and important area of study for both chemotherapy and RIC alloHSCT approaches. Data from the French Group for Research on Adult Acute Lymphoblastic Leukemia (GRAALL) indicate a 22% chance of death in remission when the intensive approach was used in over-40s. A US intergroup study is currently evaluating a pediatric chemotherapy approach whereas the UKALL14 study from the United Kingdom evaluates RIC alloHSCT in all patients over 40 with a suitable donor.
ALL in Older Adults
Large studies have typically not included patients over age 65 and this may be one of the reasons that patients of this age are the only group in whom survival has not improved over time. The median age of adults with ALL is over 60 years, however. In the United Kingdom, the age-specific annual incidence of ALL rises from 0.45/100,000 in adults aged 35 to 39 to 0.78/100,000 in patients 60 to 64 to 1.2/100,000 cases in the over-85 group. The predicted rise in the proportion of people over 65 in the United Kingdom from 16% of the population in 2009 to 23% by 2034 (UK Office for National Statistics) will result in a UK national increase in the number of older patients with ALL in the future.
There is no standard regimen for the treatment of older patients with ALL worldwide. UKALL XII shows that older patients fared poorly, with less than 15% 5-year OS. For older patients with ALL, deemed unsuitable for trial entry either because of eligibility criteria or fitness, treatment is scheduled only according to prevailing local practice. It is not known whether treatment is with curative or palliative intent. It is not known whether long-term DFS equating to cure is a realistic goal. Moreover, there are no data investigating quality of life in elderly patients with ALL and its relationship with the intensity of treatment regimens and response. In the Southwest Oncology Group study, the median survival of 40 patients aged 50 years or older was 1 month versus 17.1 months for the whole cohort of 168 patients. A CR rate of just 77% was reported on CALGB 9111 in 35 patients aged 60 years or older versus 85% in the whole group. Median survival in the over-60s was 12 months compared with 23 months for the cohort as a whole. In an MD Anderson Cancer Center study of patients treated with hyperfractionated cyclophosphamide, vincristine, doxorubicin, and dexamethasone (hyper-CVAD), only 79% of the 44 patients aged 60 years or older achieved a CR compared with 91% in the 204 total patient cohort. Five-year survival in this study was only 17% in the over-60s versus 39% in the under 60s. Recent data from the UKALL12/E2993 study also confirm significantly inferior CR, EFS, and OS rates in patients over age 55. This study reported significantly more infections during phase I induction in patients aged 55 years and older versus those younger than 55 years (67% vs 45% respectively; P <.0001) and a need for more drug dose reductions during induction in the patients 55 years and older (47% vs 27%; P = .0006) compared with those younger than 55 years old. Crucially, infection during induction in addition to adverse karyotype and high white blood cell count predicted for worse EFS and this was especially significant in those who had infection in both phases of induction (6% vs 38%; P = .007). Overall, published data indicate a 3-times to 10-times higher high rate of induction deaths in older individuals after treatment of ALL.
Two population-based registry studies (from the Northern region of the United Kingdom) report even lower response and survival rates. Taylor and colleagues and Moorman and colleagues reported CRs in only 10 of 49 and 20 of 39 older patients with ALL, respectively. The contrast between outcomes reported in studies and population-based surveys suggests a significant bias toward including only younger and fitter older patients in therapy trials. It is likely that the overall outcome for older patients with ALL may be poorer than are reported from published studies. The Dutch-Belgian group, HOVON, recently trialed an intensive regimen in patients up to 70 years. In the HOVON 71 study, patients between 40 and 70 years were included. CR rate in the pilot study was 85% and OS at 3 years in patients over 60 was 54% (median follow-up 30 months) but EFS was 41%. It is not clear how to predict which older persons will most benefit from intensive approaches to treatment. Formal studies of attempts to increase intensity of specific aspects of therapy, such as anthracylines, have not shown benefit in any subgroup of older people.
For older patients with Ph+ ALL, there has been some recent, real progress in improving at least short-term outcomes. The understanding that CR might be achieved with less toxicity when TKIs are added to therapy has emerged from 2 studies conducted in older people. One of the early studies, by the GMALL, was performed in patients older than 55 years of age and randomized participants to receive imatinib or multiagent chemotherapy for initial induction. In this randomized comparison of the 2 approaches, the overall CR in the imatinib arm was 96%, whereas in the chemotherapy arm it was 50%. The suggestion that imatinib may offer good initial responses with less toxicity is also borne out by the imatinib and steroid combination results reported in 30 elderly patients (median age 69 years) by the Gruppo Italiano Malattie Ematologiche dell’Adulto, with all achieving hematologic CR with a median survival from diagnosis of 20 months, although continuing drops in the survival curves suggest that there are unlikely to be any long-term disease-free survivors. Remarkably, within this study, most patients did not require admission to hospital.
Balancing toxicity of treatment against potentially a more aggressive disease phenotype is an enormous challenge in the management of older patients with ALL—attempts to increase intensity of therapy undoubtedly result in significant increases in toxicity. Emerging novel agents, such as monoclonal antibodies, have the tremendous potential of adding independent antileukemia effect without the addition of substantial toxicity. The challenge, however, is how best to integrate these novel agents into ALL protocols for older individuals for whom a standard of care is not defined.
Philadelphia-Positive ALL
The treatment of Ph+ ALL has been revolutionized by the addition of TKIs to treatment. A detailed description of the considerable literature exploring the use of TKIs in Ph+ ALL is beyond the scope of this article—readers are referred to recent review articles by the author and others. There are many studies published that document a higher rate of CR when the B CR -ABL–specific TKI is added to combination chemotherapy. When used as a single agent in conjunction with steroids, CR has been demonstrated in all patients without treatment-related mortality—this was particularly impressive in a study of older patients, most of whom were treated as outpatients. This, however, is not a long-term curative option for such patients, who eventually all relapse without substantive therapy. An accumulating data set demonstrates the existence of small clones with pre-existing resistance to TKIs as a result of B CR -ABL kinase domain mutations, which undoubtedly contribute to loss of response to TKIs. For this reason, B CR -ABL quantitation showing a good initial reduction in B CR -ABL in response to TKI therapy does not correlate with a good long-term outcome in the same way it was originally reported in the pre-TKI era. Although there have been no randomized controlled trials comparing chemotherapy with chemotherapy plus imatinib in this disease, OS when imatinib is included as part of therapy has been compared with historical controls in large data sets in several studies. There is little doubt from these studies that the overall outcome of therapy is superior with the inclusion of TKIs. It is not clear, however, whether this ultimately relates to the higher rate of alloHSCT, which also been demonstrated in these and other studies. When imatinib and chemotherapy combinations have been used followed by subsequent alloHSCT, reports of 3-year OS of up to 60% are emerging. Although there is some evidence that Ph+ ALL in children can be treated successfully with chemotherapy and TKI combinations alone, this conclusion is based on a small study that was not designed to ask this specific question.
In adults, for whom alloHSCT remains the mainstay of therapy, there are many questions of how best to use TKIs after alloHSCT. It is not clear whether TKIs are needed at all. A German study is currently investigating in a randomized fashion the relative effectiveness of an expectant approach, giving imatinib only if the B CR -ABL PCR is positive, with an approach in which all patients are given imatinib starting at 3 months post-alloHSCT. Imatinib is not always well tolerated post-alloHSCT. Outside a clinical trial, administration of imatinib in this setting might be best confined to situations when a B CR -ABL signal is detected.
The agent dasatinib is less specific for brc-abl and also blocks Src kinases, which, unlike in CML, are known to play a role in Ph+ ALL. In theory, this agent should result in a better outcome. It is less well tested than imatinib in Ph+ ALL but phase 2 studies have evaluated its efficacy has also been used both alone and in combination with chemotherapy. Initial studies demonstrated responses to single-agent dasatinib in relapsed disease—the drug was active in situations of imatinib resistance due to B CR -ABL kinase domain mutations. This agent also crosses the blood-brain barrier. The occurrence of the T315I mutation confers resistance to all currently licensed TKIs and requires an experimental approach to treatment either using novel TKIs, such as ponatinib, or the bispecific antibody blinatumomab. The role of monoclonal antibodies in ALL treatment is discussed in a subsequent article in this issue. There is no clear evidence to date in the adult setting that Ph+ ALL can be optimally managed without alloHSCT because, even in the presence of TKIs, most patients eventually relapse without the definitive therapy for alloHSCT.
Related posts:
Acute Leukemias
Clinical Implications of Novel Mutations in Epigenetic Modifiers in AML
Current Therapeutic Strategies in Adult Acute Lymphoblastic Leukemia
Cellular Therapies in Acute Lymphoblastic Leukemia
Diagnostic and Prognostic Value of Cytogenetics in Acute Myeloid Leukemia
Prognostic Factors in Adult Acute Leukemia
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
