Appraisal of IgM Kappa/IgM Lambda Variations Using HevyLite® After Rituximab As Consolidation Therapy in Patients with Waldenström's Macroglobulinemia

In Waldenström's Macroglobulinemia (WM), response rates to rituximab (RTX) range from 30% to 75%. RTX is also known to generate a transient IgM flare in some patients. Our study aimed to characterize this phenomenon using the newly Hevylite® assay developed by The Binding Site that specifically quantifies IgM-kappa(K) and IgM-lambda(L).

Between 2007 and 2009, 33 WM patients received a 6-course fludarabine-based regimen, 32 as first and 1 as second line. Partial response (PR), defined as a reduction ≥50% but <90% in baseline monoclonal peak (BMP) on serum protein electrophoresis (SPE), was the minimal goal. Two of the 32 patients treated at first line needed further treatment with CHOP regimen to reach the minimal PR. All received 4 consecutive weekly IV injections of 375 mg/m2 of RTX as consolidation. All were tested using the SPAplus analyzer for residual IgM-K and IgM-L at first and fourth injection. Monoclonal IgM-K or IgM-L level variations and IgM-K/IgM-L ratios were analyzed. SPE and immunofixation (IF) were performed 3 months after the fourth RTX injection to evaluate the impact of consolidation on response.

Of the 33 tested, at a variation threshold ≥20% relative to baseline IgM levels, 3 subgroups were defined: 15 or 45.5%, the no-flare patients, showed an increase or a decrease <20%, 11 or 33.3%, the flare patients, showed an increase ≥20% while 7 or 21.2%, the anti-flare patients, showed a decrease ≥20%. In the no-flare subgroup, there was no significant difference in monoclonal K (9) and L (6) light chains while in the 18 but 1 patients of the merged flare and anti-flare subgroups, monoclonal K light chain was overrepresented as there were 17 K vs 1 L (p<0.05). A highest initial IgM-K/L ratio (p<0.005) characterized the flare subgroup. No association was seen with demographic and hematologic characteristics and RTX pharmacokinetics. Regarding Fc gamma receptor (FcγR) IIIa-158 polymorphism, 10 patients were defined as VF and 5 as FF in the no-flare subgroup, 3 as VF, 3 as FF and 1 as VV in the anti-flare subgroup, 3 as VF and 8 as FF in the flare subgroup, with no statistically significant difference between the last 2 subgroups (p=0.22).

We also analyzed the data without defining a critical threshold in monoclonal IgM level variations. Patients were, then, separated in 2 subgroups: 17 flare patients and 16 anti-flare patients. Monoclonal K light chain was overrepresented in the 2 subgroups, with 13 K vs 4 L in flare patients and 13 K vs 3 L in anti-flare patients; no statistically significant difference was found between them regarding monoclonal K and L light chains (p=0.32). No association was seen with demographic and hematologic characteristics and RTX pharmacokinetics. Regarding FcgR IIIa-158 polymorphism, there was a statistically significant difference (p=0.03) as 12 patients were of the FF subtype and 5 of the VF one in the flare subgroup while 4 were of the FF subtype, 11 of the VF one and 1 of the VV one in the anti-flare subgroup.

At response evaluation 3 months after the fourth injection of RTX, all patients were screened for SPE and 22 for IF. When we compared the 17 patients of the flare subgroup to the 16 ones of the anti-flare subgroup, no patient vs 1 reached complete remission defined as disappearance of BMP on SPE and negativity of IF, 1 vs 6 advanced to very good partial response defined as a decrease ≥90% in BMP on SPE and positivity of IF and the remaining 16 vs 9 improved their partial response by a mean additional reduction of 20% (range: 0–30%) vs 18.75% (range: 0–50%) in BMP, but there was no statistically significant difference in this regard (p=0.79).

Using Hevylite® assay, monoclonal IgM up and down variations of at least 20% following RTX occurred in up to 54.5% of our patients and this raised to 100% when no threshold value was considered. Such assay provides a reliable and highly sensitive quantitative assessment, especially in patients whose monoclonal IgM was at low concentration.

Patients of the flare subgroup were predominantly of the FF subtype of FcgR IIIa-158 polymorphism while patients of the anti-flare subgroup were predominantly of VF subtype, which was statistically significant. It is tempting to hypothesize a difference in RTX mechanism of action, with ADCC predominating in the anti-flare subgroup and CDC in the flare subgroup. It would be of interest to assess such an hypothesis on a larger cohort and to analyze its impact on response profile.

Pietrantuono:The Binding Site: Employment.