Variability of definition of high‐risk multiple myeloma across phase III clinical trials

Abstract The definition of high‐risk multiple myeloma (HRMM) is evolving. Use of a clear definition of HRMM in clinical trials was not previously studied. We explored the definition of HRMM in completed phase III clinical trials. There is extreme variability in the definition and cutoffs used to define HRMM, with a significant number of studies lacking a clear definition. Our study provides a quantification of the variability in defining HRMM and suggests a need to better define HRMM in future clinical trials to enable more consistent treatment recommendations.

staging system in their peer-reviewed publications. A total of 79 trials were conducted after 2005, in which R-ISS was used in 14 trials (17.7%) and ISS was used in 50 trials (63.3%). A total of 54 trials(56.3%) included only patients with newly diagnosed MM. Table 1 shows the characteristics of the included trials. As shown in Table 2, del(17p), t(4; 16), and t (14;16) were the most reported aberrations although the reporting of thresholds used to define positivity was only reported in 18%, 17.6%, and 18.2, respectively. Missing cytogenetics data were reported in 66 trials with a median of 22.6% of patients per trial having missing cytogenetics data (IQR: 36.7%). A trend of lower missing cytogenetics in more recent trials was noted. Peer-reviewed publications, a larger patient enrollment, and more recent enrollment and publication dates were all significantly associated with HRMM definition reporting (p < 0.001). However, the location of the study (United States vs. others) and funding source (industry vs. nonindustry) did not predict whether the definition of HRMM was reported (p = 0.071 and 0.452, respectively).
None of these factors were significant in the binary logistic regression model (p > 0.05). Finally, both a later starting date of the trial and the publication year were correlated significantly with a lower missing cytogenetics data (p < 0.05).

DISCUSSION
The findings of our study suggest that the definition of HRMM is underreported in interventional phase III MM trials. Although almost all the trials that defined HRMM used a cytogenetics-based definition, a specific threshold and/or exact definitions were lacking. Such information is important as different high-risk cytogenetics aberrations can influence the disease progression and response to treatment differently [12]. For example, patients with t(4; 14) and del(17p) may benefit from prolonged proteasome inhibitor treatment [13,14].  [17].
We also observed the use of different thresholds in determining the presence of cytogenetic abnormality. The presence of del(17p) is not prognostic in all patients, and the size of the clone carrying the abnormality is important [18,19]. Patients presenting with del(17p) less than <60% of their plasma cells did not show a specific poor outcome. What is even more problematic is the underreporting of the used threshold in phase III clinical trials. High inter-lab and intra-lab variabilities in cytogenetics testing by different labs were previously reported [20].
The number of cytogenetic abnormalities is important. In MASTER trial [21], the presence of two or more cytogenetic abnormalities (defined as ultrahigh-risk disease) was associated with worse outcomes, which did not occur in patients with zero or one cytogenetic abnormalities. The

GMMG-CONCEPT trial investigated isatuximab plus KRd in HRMM
patients based on the presence of at least one cytogenetic abnormality [22]. The OPTIMUM/MUKnine trial also used the same high-risk stratification criteria as the MASTER trial but only included ultrahighrisk patients and primary plasma cell leukemia patients [23]. Although n/a n/a n/a n/a Abbreviations: IQR; Interquartile range, n/a; not available/applicable. cytogenetics-based HRMM definition is the most common criteria to be used in HRMM-only trials, GEP is another robust non-cytogeneticsbased criterion that is commonly used in Total Therapy trials by the University of Arkansas for Medical Sciences group. TT5 only included patients with high-risk GEP profiles [24]. In these patients with highrisk GEP profiles, cytogenetic abnormity was not associated with survival disadvantage.
It is important to note that the definition of HRMM continues to evolve. A second revision (R2-ISS) classified patients into four risk groups that predict different survival outcomes. Compared with the R-ISS, the new R2-ISS includes the independent poor prognostic factors 1q gain (3 copies of 1q) or amplification (≥4 copies of 1q) [25].
It is expected that the definition of HRMM will continue to evolve. To