Molecular Characterization of Mobilized Apheresis Hematopoietic Progenitor Cells Designated for Autologous Stem Cell Transplantation in Multiple Myeloma

ABSTRACT

Introduction

Multiple myeloma is the second most common malignancy of the lymphoid system and accounts for about 2% of cancer deaths. Treatment generally consists of chemotherapy and autologous stem cell transplantation. Clinical trial results are equivocal as to whether autologous stem cell transplantation in the treatment of multiple myeloma prolongs overall survival. One possible etiology for this conundrum could be that the transplanted autologous stem cells may also contain residual tumor cells/tumor stem cells.

Methods

Residual autologous stem cells from patients undergoing stem cell transplantation were examined for cells containing monoclonal immunoglobulins and clonotypic DNA detected in diagnostic bone marrow to determine if malignant myeloma cells may have been transplanted along with hematopoietic stem cells.

Results

In all eight cases of autologous stem cell transplants examined from multiple myeloma patients, the infused cells included cells containing monoclonal immunoglobulins of the same type as were noted in patients' sera at the time of diagnosis. This was not seen in allogeneic cells from donors without cancer. In all four cases in which a dominant clone was identified by molecular analysis in the diagnostic bone marrow, the same clone was detectable in the autologous stem cells.

Conclusions

We posit that lack of prolongation of overall survival with autologous stem cell transplantation in multiple myeloma is in part due to infusion of cells containing tumor cells that repopulate the recipient. This finding may warrant novel collection and processing strategies to mitigate the infusion of malignant cells. Until such methods become available, the use of autologous stem cell transplantation should be considered in the context of the conundrum reported here.

Trial Registration:

The authors have confirmed clinical trial registration is not needed for this submission

Abbreviations

ASCT

autologous stem cell transplantation

CAR‐T

chimeric antigen receptor (CAR)‐T cell

HSC

hematopoietic stem cells

MM

multiple myeloma

SIFE

serum immunofixation electrophoresis

SPEP

serum protein electrophoresis

  

“You are going to give me my own cells back?”

Question from a patient undergoing autologous stem cell collection.

1. Introduction

Neoplastic disorders of mature B‐lymphocytes, plasma cells, consist of monoclonal gammopathy of undetermined significance (MGUS), smoldering/asymptomatic multiple myeloma (SMM), and the malignant entity of multiple myeloma (MM). MM is the second most common malignancy of the hematopoietic system in adults and accounts for about 2% of cancer deaths [1]. Advances in chemotherapy have improved survival, although the disease remains incurable [2]. At diagnosis, patients are evaluated for suitability for autologous stem cell transplantation (ASCT). About 70% of MM patients receive ASCT [3, 4]. The current paradigm for the treatment of MM involves chemotherapy to induce remission and subsequent collection of peripheral stem cells following mobilization with drugs including granulocyte colony‐stimulating factor and Plerixafor [5]. The mobilized stem cells are stored frozen. The patient is treated with a myeloablative regimen (myeloablative dose of Melphalan and/or radiation therapy), and bone marrow is reconstituted by infusion of the stored autologous peripheral hematopoietic stem cells (HSC). Maintenance chemotherapy is continued following successful reconstitution of the bone marrow [6, 7]. Open‐label trials have shown improvement in progression‐free and overall survival [6, 7, 8, 9]. However, in randomized clinical trials, ASCT has been shown to improve progression‐free survival but not overall survival [10, 11]. Improvement in quality of life with ASCT has been evaluated with mixed results [11]. Historically, a similar lack of effectiveness in prolonging survival was also observed in treating metastatic breast cancer by ASCT [12]. These results call into question the efficacy of ASCT and indicate that additional considerations must apply, including holding this treatment in abeyance until modified collection and processing of stem cells show improvement in overall survival.

2. Methods

2.1. Analysis for Monoclonal Immunoglobulin

Taking a cue from statements by patients undergoing stem cell collection questioning the use of autologous cells, we analyzed residual hematopoietic progenitor cells collected for ASCT for the presence of monoclonal immunoglobulin‐containing cells. HSC from 17 allogeneic sources and 8 autologous sources that were slated for disposal were examined under an institutional review board‐approved protocol. All of the patients for whom these HSC preparations were intended or infused were deceased. All 8 autologous and 17 allogeneic HSC were analyzed.

For immunoglobulin analysis, stem cells from preparations used for patients were collected by centrifugation of the thawed HSC stored below −150°C in 10% DMSO. Cells were lysed by addition of deionized water and by subjecting the preparation to vortexing. The preparation was then subjected to freezing at −30°C for 10 min to ensure lysis. The lysed preparations were centrifuged at 13,300 g for 10 min, and the supernatant was analyzed by the standard procedure used for serum protein immunofixation electrophoresis (SIFE) (Helena Laboratories, Beaumont, TX) [13]. The SIFE gels were stained with antisera to IgG, IgM, and IgA kappa and lambda. In two patients with light chain myelomas, the SIFE was stained with antisera to free light chains as well. The free light chain‐specific antisera were obtained from Sebia Inc [13].

2.2. Analysis for Malignant Clones

Bone marrow aspirate cells collected at the time of diagnosis were stored frozen in a cryoprotectant containing DMSO. Frozen diagnostic bone marrow aspirates from six different patients were sequenced to identify dominant clonotypes associated with malignant plasma cells. Clonal analysis of the diagnostic bone marrow and autologous cells was performed in a clinical diagnostic laboratory by Adaptive Technologies. In five of the six bone marrow samples, a dominant clone could be identified on DNA analysis. The bone marrow specimen in the sixth case was too dilute to provide a reliable clonotype. An aliquot collected from the peripheral blood‐derived mononuclear cell isolates designated for ASCT for five patients was subjected to analysis for minimal residual disease (MRD) using the ClonoSEQ assay developed by Adaptive Technologies. Because an apheresis product from the sixth patient was not available for analysis; thus, interpretable results were available in four patients.

3. Results

3.1. Monoclonal Immunoglobulins

Monoclonal immunoglobulins were detectable in all eight cases of autologous stem cells collected from MM patients, but in none of the 17 allogeneic stem cells collected from healthy donors (Figure 1). The type of immunoglobulin detected in autologous cells was consistent with the type of monoclonal immunoglobulin in each patient, including in two cases of light chain myeloma. The distribution of monoclonal immunoglobulins included four cases of IgG kappa and one each of IgG lambda and IgA kappa, and one each of free kappa and lambda light chains. The findings demonstrate that monoclonal immunoglobulin‐containing cells were present in the mobilized HSC that were reinfused in all cases of ASCT. As expected, only polyclonal immunoglobulins were detectable in HSC from allogeneic sources.

FIGURE 1.

Immunofixation electrophoresis of representative patients with IgG kappa, IgG lambda, and IgA kappa MM is presented. The fourth set of lanes labeled “ALLOGENEIC” was from a patient with IgG kappa MM who received allogeneic HSC. Cell lysate supernatants from residual autologous HSC (first three electropherograms) and allogeneic HSC were electrophoresed and stained with antisera to IgG, IgA, and IgM kappa and lambda immunoglobulins. The patterns showing monoclonal IgG K, IgG L, and IgA kappa are in accord with the MM lesions in the respective patients. No monoclonal bands were detected in any of the 17 preparations of allogeneic HSC; a representative example is shown in the fourth set of lanes (marked “allogeneic”), which shows polyclonal staining only. (Cell debris creates a band in all lanes at the point of application of the sample.)

As all 17 specimens from allogeneic sources were negative for monoclonal immunoglobulins and all eight specimens from autologous sources contained monoclonal immunoglobulins, statistical analysis was considered to be redundant. However, the two groups were significantly different, by Fisher's exact test with a p‐value < 0.00001.

3.2. Clonal Marker Analysis

Six patient samples from bone marrow aspiration, which had been cryopreserved at the time of collection, were submitted for clonality analysis (in certain cases, the aspirate was derived from post‐ASCT relapse, as noted in Table 1). Among these, in one sample, no dominant monoclonal species was detectable, likely due to the hemodilute nature of the aspirate specimen. For a second patient, autologous peripheral stem cell collection was unavailable at the time of analysis. For the four remaining patients, residual plasma cells with monoclonality at the IgH locus constituting MRD could be identified, ranging in limit of detection from 10⁻⁴ to 10⁻⁶ (Table 1). In all analyzable cases, the signal clonal sequence could be detected at a frequency from 8 to 50 × 10⁻⁴.

TABLE 1.

Patients all presented with IgG‐kappa multiple myeloma diagnosis. Time of aspirate refers to pre‐ASCT versus at relapse, with plasma cell % composition for determination of IGH clonotype/clonal marker and uniqueness score of clonotype indicated. Frequency of clonotype/clonal marker identification is indicated in apheresis product.

Diagnosis	Time of aspirate	Age at transplant	Sex	Race	Plasma cell %	MRD locus	Uniqueness score	MRD freq
IgG K	pre	50	M	W	60%	IGH	3.8E‐44	0.000116
IgG K	post	68	M	AA	50%	IGH	3.34E‐18	0.006485
IgG K	post	65	M	W	20%	IGH	4.93E‐31	0.005606
IgG K	post	49	M	AA	30%	IGH	5.29E‐34	0.000743

4. Discussion

Based on the detection of monoclonal immunoglobulins and relevant malignant clones in the mobilized autologous stem cells in MM, it could be inferred that, along with normal HSC, tumor cells were also infused into MM patients receiving ASCT. This evokes a conundrum addressing the medical utility, ethics, and fiscal stewardship of ASCT in MM. The lack of improvement in overall survival following ASCT could be due to reinfusion of tumor cells. This consideration has been reported using multiple approaches [14, 15]. While previous studies examining the clonality of apheresis HPC products relied on multicolor flow‐cytometric methods, in the current study, clonality of ASCT apheresis samples has been demonstrated using a molecular evaluation by sequence analysis at the immunoglobulin heavy chain locus.

While this conundrum raises a shortcoming with this approach to therapy, there may be other approaches to mitigate this concern. One approach to improve the usefulness of ASCT could be by developing techniques to deplete tumor cells in autologous HSC preparations. Similar to the approach taken in CAR‐T treatment, tumor cells/tumor stem cells could be purged by various strategies, including oncolytic virus therapy [16, 17], or negative selection with plasma cell‐specific antibodies, such as B‐cell maturation antigen, CD 138, or other tumor‐associated antigens, before infusion. In vitro CAR‐T treatment of stem cell preparations could be considered. One of the risks in targeting cell surface markers by CAR‐T is the concomitant risk of destroying/damaging normal cells, especially in the brain [18]. Even if such antigens are present in brain or other vital structures, such a risk could be mitigated by antibodies and CAR‐T cells directed at myeloma tumor/stem cells prior to infusion.

There are a number of shortcomings in this study, with the most important being the limited number of cases. However, 100% concordance with the monoclonal immunoglobulin type and clonotype/clonal marker of the malignant cells present in infused autologous stem cells makes a strong case for concluding that malignant cells were present in the HSC infused into the patient and probably contributed to the recurrence of disease. However, while we observed markers of malignant cells in the HSC infused into the patients, we did not establish that the HSC contained tumor stem cells capable of seeding the bone marrow and proliferating to cause recurrence of the tumor.

Earlier studies have also shown the presence of cells with markers of malignant cells in the transfused preparations. These studies used flow cytometry to detect plasma cells [19, 20]. In our study, despite the small number of cases, clonotypic identity/clonoal marker of the diagnostic marrow and infused cells makes a strong case for concluding that the infused HSC contributed to the recurrence of disease.

Previous interventions to remove the tumor/tumor stem cells were not successful in improving outcomes, and the interventions proposed by us may not be successful [16, 20]. However, since infusion of ASCT/HSC did not improve overall survival outcomes, we propose that ASCT treatment should be held in abeyance till the development of more effective treatments rendering the ASCT useful in prolonging overall survival.

Author Contributions

G.S. conceived the project. G.S. and R.J.B. collected the data, and collaborated in drafting and revising the manuscript. K.K. assisted in the revision of the manuscript.

Funding

The authors have nothing to report.

Ethics Statement

This study was approved by the Institutional Review Board of Augusta University.

Consent

Requirement for consent was waived by the IRB, as the patients whose cells were used were all deceased at the time of the study.

Conflicts of Interest

Dr. Gurmukh Singh serves as a consultant to Helena Laboratories, Sebia Inc., Diazyme Laboratories, and Nexicart‐2 IRC, and to Warm Springs (GA) Medical Center, and has applied for a patent for the FLC‐Modified SIFE method. The other authors declare no conflicts of interest.

Data Availability Statement

Data are available upon request.