Multiple myeloma (MM) is an incurable disease with a variety of treatment options. In those capable of handling aggressive treatment, often high dose chemotherapy (HDT) and autologous stem cell transplant (ASCT) are used in first remission to improve progression free (PFS) and overall survival (OS). It has been theorized and demonstrated that propagation of the disease following HDT/ASCT is partially due to the return of the disease in the autograft. What is less clear is whether or not a threshold of disease in the autograft correlates with a higher risk of earlier recurrence. This study hypothesized that such a threshold exists.


MM patients in this single institution study were prospectively observed following initial diagnosis and consent from 1998 to 2014. Basic patient demographics collected included age, sex, International Staging (ISS), type of induction therapy and response. Patients were observed for primarily time to progression following HDT/ASCT. OS was a secondary endpoint. At diagnosis, a bone marrow aspirate aliquot was used to create patient-specific primers for detection of their unique MM variable region sequence (VDJ). These primers were subsequently used during quantitative PCR on each patient's autograft at time of collection and reinfusion (i.e. post thawing). Cut-point analysis was used to determine a threshold for which detection of minimal residual disease in the autograft may translate into poorer survival. Cox proportional hazards multivariate analysis took into consideration age at transplant, sex, ISS, type of induction and response to induction.


Between 1998 and 2014, 89 patients were accrued to study. 77 patients are included here for analysis. 12 patients were not included due to an inability to properly sequence patient specific primers (10 of which harbored light chain disease). Basic demographics included median age at diagnosis of 58 years (range 36-71), 61 male, and a majority of patients having ISS stage 3 (n=31; n=24 for ISS 2). Given the time points involved, induction therapy was variable including bortezomib-containing (n=31), doxorubicin-based (n=33), melphalan (n=4), and dexamethasone (n=9). A threshold of detection at ≥ 1 in 100,000 circulating white blood cells in the autograft was associated with poorer PFS following HDT/ASCT (median 1.7 vs 2.6 years, hazard ratio [HR] 1.78, p=0.03). A trend to poorer OS was also seen in this group (median 5.6 vs. 15.8 years, HR 2.65, p=0.08). This detriment remained significant when taking into consideration age, sex and ISS staging. Inclusion of induction and response was not done due to the heterogeneity of treatment regimens and will require further study. Further, the aggregate number of malignant cells in the reinfused autograft was the same as the number prior to cryopreservation - indicating that essentially all malignant cells in the autograft survive the cryopreservation process and thawing.


This is the first study of which we are aware that demonstrates an association between the degree of autograft contamination with MM cells and survival. Whether this phenomenon continues to hold in the current era where more focused treatments continue to enter pre-ASCT remains to be seen. However, as HDT/ASCT remain standard treatment, our results may point toward a renewed focus on autograft purging as potential means of improving at least PFS. This study remains open and continues to accrue patients.


Belch: Amgen, Celgene, Takeda: Honoraria. Venner: Celgene, Janssen: Research Funding; Celgene, Janssen, Takeda, Amgen, Merck: Honoraria.

Author notes


Asterisk with author names denotes non-ASH members.

This icon denotes a clinically relevant abstract