A Global Strategy to Monitor M-Protein in Patients with Multiple Myeloma By Mass Spectrometry

Background:

Multiple myeloma (MM) produces monoclonal immunoglobulins (Igs) and free light chains (LCs), which serve as biomarkers for treatment response and relapse. Effective therapies can drastically reduce circulating monoclonal immunoglobulin (Ig) levels, rendering them undetectable using conventional methods. This lack of sensitivity in assessing minimal residual disease (MRD) is critical, as MRD positivity correlates strongly with poorer progression-free survival.

Currently, MRD relies on bone marrow biopsy, which is an invasive procedure. Although informative, these blind biopsies can miss focal disease areas, leading to discrepancies in the assessment of low-burden MM. There is a need for sensitive, specific, and noninvasive methods to accurately measure and monitor MRD. We addressed this gap by developing two complementary mass spectrometry (MS) approaches targeting patient-specific LCs and specific sample and data processing to minimize the limitations of tryptic clonotypic peptide analysis and the nucleotide sequencing step.

Methods:

Our pilot study involved 10 serum samples collected from patients with MM at diagnosis and various treatment points in Parisian hospitals (Pitié-Salpêtrière & Saint-Antoine). We used a two-step approach :

1. Intact LC enrichment Specific nanobeads were used to enrich intact LCs, enabling accurate detection using a “middle-down” strategy.

2. Clonotypic peptide identification: The enriched LCs were enzymatically digested and the resulting clonotypic peptides were identified using a “bottom-up” strategy. This approach enhances the detection limits, especially for low-abundance peptides within monoclonal proteins.

De novo sequencing complemented protein database searches to identify patient-specific clonotypic peptides. The selection criteria were:

• Overlap at least partly with a hypervariable (CDR) region and conserved “constant” region in the IGBMT protein database.

• A BLAST search of the human protein library revealed no significant homology with other human proteins.

• Demonstrably decreased peptide signal intensity between diagnostic and post-treatment blood samples.

• The de novo software identification score exceeded 70%.

Results:

In both approaches, an intact specific monoclonal LC was observed when it was no longer detectable using conventional methods. The accuracy of mass measurements increases the specificity of detection. Our study showed that tryptic clonotypic peptide monitoring is not a universal strategy for accurate and sensitive MRD follow-up. However, an efficient sequencing process makes the characterization of peptides of interest even more robust and sensitive. One patient was monitored over time from the initial diagnosis through the treatment phase until considered negative and after relapse. In all monitored serum samples, intact LC was detected at 22,604 Da, which was consistent with the clonal nucleotide sequence obtained by RACERepSeq. However, among the 11 theoretical tryptic peptides generated from this 106 amino acid sequence, most are not suited for the bottom-up strategy (mass above 4000 Da), with at most two clonotypic candidates after total proteolysis. None could be used for CDR1, one peptide could cover CDR2 at 1557.8 Da, and one could cover only one-third of CDR3 at 878.5 Da. Although well detected in the initial diseased state and after relapse, neither of the two candidates could be identified at the end of the treatment and before relapse, resulting in an unambiguous false-negative result. Changing the proteolysis experimental conditions allowed us to cover CDR1, CDR2, and CDR3 more efficiently with at least four clonotypic peptide candidates. Consequently, at least 3 sequences covering CDR2 (3145.7 Da and 2982.6 Da) and CDR3 (2253.2 Da) were identified in all serum samples.

Conclusions:

Our optimized toolbox aimed to significantly improve the sensitivity and specificity of MRD detection in patients with MM. Utilizing blood samples allows for accurate, frequent, highly sensitive, and specific MRD monitoring. Personalized clonotypic profiling is crucial for a robust analysis. This approach can reduce the risks associated with undetected disease persistence and enable the earlier detection of relapse. Ongoing studies are comparing this method with bone marrow assessment. We anticipate that further technological advances will also increase the analytical throughput.

Choquet:Kite-Gilead: Honoraria. Garderet:BMS: Honoraria; Janssen: Honoraria; Sanofi: Honoraria.