Minimal Residual Disease Following Allogeneic Hematopoietic Stem Cell Transplantation

Abstract

Minimal residual disease (MRD), both before and after transplant, is a clinically important yet relatively poorly defined aspect of allogeneic hematopoietic stem cell transplantation (alloHSCT). The clinical relevance of MRD in the context of alloHSCT has been demonstrated by its association with the development of clinical relapse. However, with the possible exception of chronic myeloid leukemia, the specific techniques, timing, frequency and clinical utility, relative to improvement in patient outcomes, for monitoring MRD in the setting of alloHSCT has yet to be clearly defined. A concise overview of monitoring techniques for detecting MRD, as well as treatment strategies and biologic and clinical research initiatives for MRD suggested by the National Cancer Institute 1st International Workshop on the Biology, Prevention, and Treatment of Relapse after Allogeneic Hematopoietic Stem Cell Transplantation, is covered in this paper.

INTRODUCTION

Minimal residual disease (MRD), in the setting of allogeneic hematopoietic stem cell transplantation (alloHSCT), poses several interesting questions and complex challenges. The relevance of these questions and challenges is personified by the relationship between MRD and the risk of relapse, which is primary cause of treatment failure and death after alloHSCT [1]. The clinical relation of post-transplant MRD with relapse, particularly in relationship to chronic myeloid leukemia (CML), was recognized early with development of cytogenetic and molecular techniques of detection [2]. The clinical relevance of MRD has been further recognized with the increased use of non-myeloablative and reduced-intensity conditioning regimens, with which relapse is even a greater clinical problem [3,4].

Despite the clear association of MRD with relapse, the clinical relevance of MRD in the AlloHSCT setting remains to be determined. First and foremost, the definition of MRD needs to be defined for each disease, and needs to be distinguished from what we currently refer to as “remission” or “relapse”. The detection of persistent disease post-transplant by immunophenotypic measures has significantly different implications for patients with acute lymphocytic leukemia (ALL) as compared to someone with persistent chronic lymphocytic leukemia (CLL) [5,6]. Similarly the molecular detection of a cytogenetic abnormality in the post-transplant is markedly different for a patient transplanted with CML as compared to a patient with acute myeloid leukemia (AML) [7]. Second, when and how often we should be using available techniques for a specific disease remains to be defined. This not only applies to the post-transplant setting but also to the pre-transplant setting, where multiple studies have demonstrated the prognostic significance of MRD prior to conditioning [8]. As the majority of relapses occur within the first six months after transplant [1], it is important to determine the frequency of monitoring for recurrent disease within this post-transplant period. If we can determine when and how often, the next question is what tests should we performing and are those tests adequately sensitive, specific, reproducible, practical and economical. Finally, and most importantly, does monitoring for MRD make a clinical difference? There is sufficient evidence that detection of MRD provides prognostic information. However, does this information result in clinical decisions, relative to choice of conditioning regimen or stem cell product relative to detection of pre-transplant MRD or intervention (e.g. withdrawal of immune suppression or donor lymphocyte infusion) that result in improved outcomes? These remain essential questions for which there are relatively limited data and recommendations with the possible exceptions of CML and ALL, and even with these diseases, there remains a need for further investigation.

This manuscript attempts to provide a concise overview of many of these issues. Specifically it attempts to address methods for monitoring of MRD and strategies to clinically manage patients once MRD is detected. In addition a brief summary is provided on the National Cancer Institute 1st International Workshop on the Biology, Prevention, and Treatment of Relapse after Allogeneic Hematopoietic Stem Cell Transplantation, which attempted to address in a formal manner many of the issues described above.

MONITORING MINIMAL RESIDUAL DISEASE AFTER ALLOGENEIC STEM CELL TRANSPLANTATION

Improved supportive care, the introduction of reduced intensity conditioning regimen and careful donor selection has substantially decreased the non-relapse mortality after alloHSCT in the recent years and therefore relapse has become the leading cause of death following alloHSCT. Furthermore, as inferred above, relapse remains the primary cause of death among patients surviving more than two years after alloHSCT [9]. Despite improved understanding of the biology that underlies the graft-versus-leukemia/tumor (GVT) effect the relapse rate did not decrease over the past 20 years [10,11]. It is obvious that relapse after alloHSCT evolves from residual disease which escaped the preceding conditioning regimen as well as the graft versus malignancy effect.

New methodological and technological advances allow sensitive detection of MRD and early recognition of recurrence after alloHSCT. This is of clinical importance since intervention prior to florid relapse improves outcome for certain haematological malignancies [12,13]. Standard diagnostic criteria that are widely employed in the definition of relapse for the different haematological malignancies, which are based on morphological bone marrow investigations, imaging, and/or specific laboratory findings. After alloHSCT more sensitive methods, such as tumor-specific molecular primers, molecular genetics, fluorescence in situ hybridisation (FISH), flow cytometry, and/or chimerism analysis, are commonly used to monitor patients with respect to relapse (table 1).

Table 1.

Diagnostic methods to monitor residual disease and relapse after allogeneic stem cell transplantation (qPCR: quantitative real-time PCR). (modified after 6 )

Detection of residual disease (MRD)	Karyotyping	FISH	Flow cytometry	Antigen receptor PCR	Translocation or other mRNA PCR	Chimerisms: XY FISH	Chimerism: qPCR/VNTR-PCR
Utility	Subset of all types of neoplasms with chromosomal abnormalities	Subset of all types of neoplasms with know chromosomal abnormality	ALL; most AML; CLL; Myeloma	ALL; lymphoma, Myeloma	CML; Subset of ALL; subset of AML; subset of lymphoma	All types of neoplasms (sex mismatched SCT) Disadvantage not specific for MRD	All neoplasms (precondition differences in donor/recipient polymorphisms) Disadvantage: not specific for MRD
Sensitivity	10−1	10−2	10−3 – 10−4	10−4 – 10−5	10−3 – 10−6	10−2	10−3 – 10−6

Broadly, two different approaches are mainly used for the post-transplant surveillance of disease status: characterization of chimerism and specific detection of minimal residual disease. The latter approach measures the malignant clone directly, whereas chimerism assessment characterizes the origin of post-transplant hematopoiesis. For chimerism as well as for specific detection of residual disease, a variety of techniques are available, though in general there have been more studies looking directly at markers of residual tumor than of chimerism [14] Despite the increasing sensitivity by the described methods of chimerism determination, due to its low specificity this method is not a reliable means of detecting MRD. The specificity is higher is diseases which originate from a stem- or progenitor cell (e.g., AML, CML), whereas in B-cell lymphoma or multiple myeloma, which originate from a late B cell stage of development, the specificity of chimerism to detect MRD or relapse is low. The lack of specificity might be overcome partly by performing lineage-specific chimerism in some diseases such as multiple myeloma [15].

A paradigm for the importance of minimal molecular disease and prediction of relapse after alloHSCT is CML. Here, it is now well established that the detection of the chimeric BCR-ABL mRNA transcript by reverse transcriptase polymerase chain reaction (RT-PCR) is a powerful predictor of subsequent relapse [16]. The use of quantitative PCR has greatly increased the clinical value of monitoring MRD. It could be demonstrated that the kinetics of BCR-ABL level over time described impending relapse and response to donor lymphocyte infusion (DLI). Low or absence of residual BCR-ABL was associated with a very low risk of relapse (1%), compared to 75% relapse rate in CML patients with increasing or persistently high BCR-ABL levels [17]. The activating mutation V617F of the JAK2 gene is an obvious target for monitoring MRD in patients with myeloproliferative disorders undergoing alloHSCT. There are emerging data suggesting that, similar to BCR-ABL in CML, PCR negativity for JAK2-V617F correlates with prolonged remission and that reappearance of a detectable JAK2-V617F clone is associated with relapse [18].

However, the utility of the available tools in the monitoring of disease status after alloHSCT has not yet been fully elucidated across all hematologic malignancies. In AML and myelodysplastic syndromes, several studies demonstrated the relevance of chimerism, and especially its kinetics, for the prediction of relapse. A variety of genetic markers are available for MRD in AML such as rearrangements t (15;17)/PML-RARA, inv(16)/CBFB-MYH11, and t(8;21)/RUNX1-RUNX1T1, NPM1, FLT3 or MLL-PTD but have not been studied in a larger cohort of patients.

Methods for MRD monitoring in B- or T-lymphoid malignancies include PCR techniques aiming to quantitatively detect disease specific T cell receptor (TCR) or immunoglobulin (Ig) gene rearrangements. Multiple studies support the independent prognostic value of MRD measurements in pediatric and adult patients with B- and T-lineage ALL. Furthermore, the risk of relapse appears to be proportional to the level of MRD, which in some studies was found to be the most powerful prognostic factor for relapse in multivariate analyses [13]. Similarly, detection of pre-transplant MRD in pediatric and some adult studies is highly predictive of relapse following alloHSCT and, coupled with post-transplant MRD evaluation, may guide early post-transplant intervention such as early withdrawal of immunosuppression, administration of DLI, or addition of post-transplant maintenance therapy (e.g., targeted tyrosine kinase inhibition for Ph+ ALL).

In CLL two main approaches of MRD assessment have been followed: Flow cytometry, taking advantage of the unique immunophenotype of CLL, and PCR-based strategies using the clonal rearrangement of the hypervariable region of the V_H part of the immunoglobulin heavy chain gene (CDR3 region). Several studies showed that MRD assessment after alloHSCT is predictive for durable freedom from CLL progression if: 1) MRD levels are below 1 × 10⁻⁴ one year post-transplant; or 2) show decreasing or stable kinetics within the quantitative range. The clinical impact of MRD detection in different lymphomas is not identical.

Specific chromosomal translocations detectable by PCR amplification, particularly t(11;14) and t(14;18) translocation, are present in mantle cell lymphoma and follicular lymphoma, respectively, but t(14;18) translocation is also detectable by PCR at low levels in 10–25% of healthy individuals. For Hodgkin lymphoma, neither cytogenetics, flow cytometry, nor molecular testing are helpful for assessing residual disease [19].

In multiple myeloma minimal residual disease can be detected by PCR using patient specific primers derived from the re-arrangement of immunoglobulin heavy-chain genes. It could be shown that durable PCR-negativity after allografting had a cumulative risk of relapse at five years of 0%, in comparison to 33% for PCR-mixed patients and 100% for patients who never achieved PCR-negativity [20]

Ongoing and further clinical trial investigate whether sensitive MRD detection will allow for earlier therapeutic intervention, and it is hoped that treatment prior to overt relapse may improve outcome of allogeneic stem cell transplantation for hematologic malignancies.

STRATEGIES AND OPTIONS FOR RECURRENT DISEASE FOLLOWING ALLOGENEIC STEM CELL TRANSPLANTATION

The clinical significance of MRD after alloHSCT is different among diseases. MRD has been extensively studied using the qualitative PCR method during the early 1990s. Detection of BCR-ABL by PCR in the first year after alloHSCT for CML patients disappears in the majority of patients, secondary to ongoing GVT effects; however, detection of MRD after alloHSCT for Philadelphia chromosome-positive acute lymphoblastic leukemia (Ph+ALL) is indicative of imminent hematologic relapse [21–24]. In the case with t(8;21) AML, MRD after chemotherapy does not always indicate eventual clinical relapse. In the last decade, quantitative PCR machines are widely available, and sequential and quantitative tests of leukemic genes have become available. With this technique, a rise in the amount of leukemic genes strongly suggests clinical relapse in the near future. Also, several investigators have tried to find thresholds for the amount of genes that are predictive of clinical relapse. However, due to a lack of standardization of this technique, hitherto universal threshold has not been clarified at any leukemia with the possible exception of CML.

Clinical Intervention

Due to the limitation of quantitative PCR as mentioned above, clinical intervention upon the emergence of MRD has not been well established. Clinical interventions for early relapse and minimal residual disease after alloHSCT are performed as two ways; one is adoptive immunotherapy including DLI and vaccination, and another is administration of new agents, which are expected to preserve normal hematopoietic cells. Several questions are raised in this clinical setting. First, does early intervention have more clinical effects than the intervention performed at hematological relapse? Second, does clinical intervention affect the other parameters such as graft-versus-host disease (GVHD), related adverse events and the subsequent alloHSCT. Third, which is the better way, prophylactic administration or intervention upon MRD for patients with a high risk of relapse?

Adoptive Immunotherapy

DLI was first developed for relapsed patients. Although they are dramatically effective for CML, DLI remains limited of limited utility for patients with other diseases because of inadequate responses and toxicity related to GVHD, which occurs in one-third of patients. Strategies to reduce the incidence and severity of GVHD while preserving the GVL effect, tumor-specific DLI are proposed [25]. A protocol to generate hematopoietic cell-specific minor antigens (e.g. HA-1, HA-2, ACC-1) specific T-cell lines from mHag-negative donors was studied for adoptive immunotherapy. Warren et al. conducted a phase I/II study to test the toxicity and effectiveness of CTL clones specific for minor H antigens [26]. However, this strategy using cloned antigen-specific T cells has been shown to be ineffective mostly because these cells could not survive long enough to execute their cytotoxic ability in vivo. This problem could be overcome by: 1) infusion of a relatively young and small number of memory T cells without extensive expansion in vitro, and 2) infusion of autologous peripheral blood T cells transduced retrovirally with T cell receptor alpha and beta cDNA cloned from tumor/minor antigen specific T cell clones [27]. The latter approach has been shown to be promising in the setting of melanoma treatment in studies conducted by Rosenberg and colleagues at the National Cancer Institute [28]. Thus, T cells armed with TCR specific for WT-1, HA-1, HA-2 and ACC-1 would be great candidates for adoptive immunotherapy in the very near future. Another approach studied intensively in the clinical hematology field is a vaccination using epitope peptides such as WT-1, PR3, MUC-1, NY-ESO-1 and BCR-ABL fusion polypeptides. In particular, WT-1 is one of the most promising tumor antigens because WT1 vaccination-driven immunological responses and clinical responses, including reduction of leukemic cells, and the reduction of the M-protein amount in myeloma, have been reported. Further enhancement of the efficacy of the WT1 peptide vaccine can be expected by co-administration of WT1-specific helper peptide, Th1-inducing adjuvant or immunosuppressive chemotherapy prior to vaccinations to take advantage of inhibition of regulatory T cells and facilitation of homeostatic expansion of desired T cells. Adoptive immune therapies as prophylaxis or preemptive therapy would be performed in the near future

New Agents

Chemotherapy for the patients with recurrent disease is hampered by the fact that these agents impel the normal hematopoietic cells as well as tumor cells and tumor specific agents have long been desired. Recently, a new molecular-specific targeting agent has been developed. The specific manner of these new agents prompts us to use them for earlier interventions. Nevertheless, most of these tumor-specific agents exert some effects on normal hematopoietic cells and interfere with immunological functions after alloHSCT.

Tyrosine kinase inhibitors

Philadelphia chromosome positive ALL is associated with highly aggressive disease. Although alloHSCT is at present the only curative treatment option, hematological relapse still remains a major obstacle. Recently, there have been some reports of post-transplant imatinib administration, but its efficacy and administration methods are still controversial. Nishiwaki and colleagues compared prophylactic administration of imatinib with intervention upon molecular relapse to evaluate the effect of post-transplant imatinib administration. MRD became positive in both groups, leading to hematological relapse [29]. It was therefore concluded that post-transplant imatinib administration may not be an ideal prophylactic treatment for Ph+ALL patients. In contrast, Ottmann et al. demonstrated that all Ph+ALL patients who received imatinib upon appearance of BCR-ABL and promptly achieved molecular response remained in remission for the duration of imatinib treatment [30].

Bortezomib

Recently, both conventional chemotherapy and both autologous and alloHSCT have been combined with new agents, such as thalidomide, lenalidomide and bortezomib have improved the depth of response and survival of multiple myeloma patients. However, after transplantation most patients still harbor residual disease. Ladetto et al. reported the effect of post-transplant consolidation including bortezomib on MRD detected by PCR using tumor-clone–specific primers [31]. Molecular remissions were achieved in 3% of patients after autologous HSCT and 18% after consolidation with bortezomib. It has been proposed that bortezomib increases the expression of Fas and DR5 and enhances GVT effects, and that this agent also suppresses the activity of NFkB resulting in reduction of inflammatory cytokines related to graft-versus-host activity [32].

Lenalidomide

Lenalidomide is an immunomodulatory drug (IMiD) that has multiple effects on myeloma cells and their microenvironment. Administration of IMiDs for post-autologous HSCT maintenance resulted in prolonged progression-free survival even in patients who achieved very good partial response or complete response before lenalidomide administration. In the alloHSCT setting, lenalidomide plus low-dose dexamethasone combination therapy have shown significant disease and chronic GVHD control for myeloma patients, who relapsed after transplantation [33]. GVHD control with IMiD is still controversial but a very attractive issue for investigation [34].

Hypomethylating agents

Low-dose 5-azacitidine (5-Aza) was used by investigators at the M.D. Anderson Cancer Center for patients with AML/MDS as a maintenance therapy or salvage therapy upon relapse after alloHSCT; an overall survival rate of 90% at 1 year was reported [35]. Additive effects of DLI to 5-Aza were also reported. The administration of 5-Aza was not associated with an increased incidence of GVHD. Sanchez-Abarca et al. reported that 5-Aza inhibits T cell proliferation and activation, blocking the cell cycle in the G0 to G1 phase and decreasing the production of proinflammatory cytokines such as tumor necrosis factor-alpha and interferon-gamma [36]. They also reported that administration of 5-Aza after transplantation prevented the development of GVHD, leading to a significant increase in survival in a fully mismatched bone marrow transplantation mouse model. Recently, decitabine, another DNA hypomethylating agent, was reported to be used in patients experiencing cytogenetic relapse after alloHSCT [37].

Humanized monoclonal antibodies

Rituximab (anti-CD20 monoclonal antibody) was used for 9 chronic lymphocytic leukemia patients who had persistent disease after alloHSCT underwent immuno-manipulation to augment GVT effects including immunosuppression withdrawal and donor lymphocyte infusion with rituximab treatment and 8 patients had a complete response [38]. Alemtuzumab (anti-CD52 monoclonal antibody), as well as anti-thymocyte globulin, has been used as a T-cell depletion method in alloHSCT. Since it is reported that the majority of precursor B-ALL blasts express CD52 and CD52 is expressed on other ALL cells, alemtuzumab is considered to potentially contribute to the eradication of minimal residual disease [39].

Summary on the Treatment of MRD

For decades, interventions for relapsed patients have been performed using DLI and chemotherapies; however, they are a two-edged sword, hampering normal hematopoietic cells as well as tumor cells. Recently, the emergence of new strategies using tumor-specific DLI and tumor-specific new agents has prompted us to use these methods before clinical relapse. Some of them are used as prophylaxis and some of them are used upon tumor emergence at molecular level. Trials confirming these strategies are just beginning and there is a need for the definition of MRD. Thus, it is becoming more and more important that the measurement of MRD becomes standard practice; otherwise, clinical studies will be somewhat meaningless.

NATIONAL CANCER INSTITUTE 1ST INTERNATIONAL WORKSHOP ON THE BIOLOGY, PREVENTION, AND TREATMENT OF RELAPSE AFTER ALLOGENEIC HEMATOPOIETIC STEM CELL TRANSPLANTATION

As stated above, there is a strong association of MRD with relapse following alloHSCT. The growing recognition of relapse as one of the most significant post-transplant problems led to the organization and convening of the National Cancer Institute 1st International Workshop on the Biology, Prevention, and Treatment of Relapse after Allogeneic Hematopoietic Stem Cell Transplantation [40]. The primary objectives of the Workshop were to reviewing the current “state-of-the-science” relative to the biology, natural history, prevention and treatment and identify the most important biologic and clinical questions that need to be addressed relative to relapse following alloHSCT.

The Workshop which took place on November 2 and 3, 2009 in Bethesda, Maryland, USA and brought together an international group of more than 200 basic and clinical researchers. Over 50 formal presentations were made by the Workshop committee members that addressed both GVT and non-GVT biology, relapse epidemiology and natural history, strategies and therapies for prevention, disease-specific methods and strategies for monitoring, and disease-specific treatment of relapse following alloHSCT. These presentations are available for viewing at https://ccrod.cancer.gov/confluence/display/NCIRelapse/Presentations+from+Workshop. Each of the six workshop committees subsequently prepared a “state of the science” manuscript, which contained their commended research priorities; these manuscripts were published sequentially during 2010 in the Biology of Blood and Marrow Transplantation [1,14,19,41–44]. The central Workshop theme was that in its most simplistic form, relapse occurs because tumor cells are first able to resist the cytotoxic effects of the conditioning regimen. These surviving cells either never respond to initial GVT or they subsequently escape from GVT effects after initial control.

Central and recurrent research themes included the necessity to establish biorepositories to collect and store tumor samples before transplant when possible and after transplant, store samples from allografts for analysis, and collect blood and serum samples at set post-transplant time points and at the time of relapse for study of immunology related to relapse. Second, there is a need for more careful study of the natural history of relapse for specific diseases, particularly in regard to MRD. In order to perform such studies there needs to be international acceptance of standard definitions and techniques; it is hoped that the definitions and techniques proposed by the Workshop would be considered for this purpose. Finally, there needs to be multi-institutional collaboration in regard to prevention and treatment of relapse after alloHSCT. A formal summary of the workshop recommendations will be presented during the 2011 Tandem Transplant Meetings Educational Sessions.