Page 199 - Haematologica - Vol. 105 n. 6 - June 2020

HRQoL with thalidomide and lenalidomide
followed by maintenance therapy. Importantly, there was a clinically relevant decrease in pain, defined as large by Cocks et al., in both arms. Clinically meaningful diarrhea developed in approximately 30% of patients during lenalidomide treatment, being of medium clinical effect and significantly higher than in patients treated with thalidomide. Importantly, in patients who reached main- tenance, there was no clinically relevant further deteriora- tion of diarrhea during MPR-R therapy. In contrast, clini- cally meaningful peripheral neuropathy developed in approximately 60% of patients during thalidomide treat- ment, being significantly higher than in lenalidomide- treated patients, both during induction and maintenance.
In general, the improvement in HRQoL subscales reached clinical relevance after six and 12 months of main- tenance therapy only, with the exception of global QoL, future perspective and pain, which improved early during induction therapy and sustained during treatment. A sub- analysis of patients who started maintenance therapy and were treated for at least three months, showed that lenalidomide resulted in a clinically meaningful improve- ment in global QoL and role functioning over time, with- out any clinically meaningful deteriorations. In contrast, there was no clinical benefit of thalidomide maintenance treatment, only clinically relevant worsening of peripheral neuropathy occurred.
Importantly, patients with a poor WHO performance status of ≥2 at baseline reached similar HRQoL during treatment, compared to patients with a better WHO per- formance status at baseline, irrespective of the treatment arm. Although this could be explained by regression to the mean, indicating that the most pronounced improvement can be achieved in patients with the worst HRQoL, it is reasonable to suppose that the performance status can be negatively affected by the disease and therefore that treat- ment also improves HRQoL of the physically most com- promised patients. Therefore, treatment should not be automatically withheld from those patients.
Our findings are in line with previous data on HRQoL describing improvement in QoL during treatment.6 In addition, we provide data about HRQoL during up to 1 year maintenance therapy with lenalidomide and thalido- mide, being rarely reported in the current literature.10-13,16 In view of the continuous lenalidomide treatment approaches, our data showing clinically meaningful improvement of global QoL during maintenance therapy is important for clinical practice. In the FIRST trial, no such improvement was reported; the global QoL remained sta- ble, which might have been attributed to the higher dose of lenalidomide and the continuation of dexamethasone, known to be associated with side effects that can hamper QoL.9 In the MM-015 trial, a statistically significant, but not clinically relevant deterioration in global QoL after six months of maintenance therapy was observed.11 This can- not be accounted for by a difference in the percentage of patients starting maintenance, the dose of lenalidomide, response rates or different QoL evaluation methods. The fact that thalidomide maintenance has only limited impact on global QoL is in line with the data of the HOVON 49 and the MRC IX data. The latter even show an inferior global QoL at three months with a trend to further detri- ment during maintenance treatment.8,12
A limitation of our and HRQoL studies of patients with MM in general, is the fact that firstly, long term data reflect a subset of patients who tolerate remaining in treat-
ment. Secondly, we collected no data after discontinuation of the study although such results would rather reflect the outcome of subsequent therapies.29 Therefore, biases are common, and comparisons between therapies with different toxicities and discontinuation rates are difficult. Especially, as missing data are not missing at random when related to toxicity and cannot be corrected for.30 Such bias might be present in our sub-analyses of patients starting maintenance therapy, showing benefit for patients treated with lenalidomide only. Therefore, this can only be concluded for those patients who did benefit from MPR induction and were able to continue lenalidomide. In addition, the low number of patients continuing thalido- mide might be the cause of a lack of finding statistically significant changes in their HRQoL during maintenance.
Interestingly, the higher incidence of peripheral neu- ropathy with thalidomide, both clinically meaningful to patients and according to CTCAE reported by physicians, did not translate into an inferior global QoL, neither in our study nor the FIRST trial.9 Via several analyses, we exclud- ed a bias in global QoL evaluation due to missing HRQoL questionnaires from patients who discontinued treatment because of peripheral neuropathy. The fact that peripheral neuropathy did not negatively affect HRQoL was surpris- ing. The opposite has been reported.31 Our observation might be explained by response shift, which is a well- known phenomenon in longitudinal QoL research. It reflects the probability that a patient’s standards and val- ues change over time.32 Patients with MM might adapt to their worsening function and increased symptoms and thereby not allow these aspects to affect their global QoL.33 However, it cannot be excluded that the global QoL scale of the EORTC QLQ-C30 questionnaire has lim- itations in detecting the negative impact of toxicity on HRQoL. This discussion was addressed in a meta-analysis by Schuurhuizen and colleagues,34 who also reported sim-
Figure 5. Patient- versus investigator-reported peripheral neuropathy. Patient- reported peripheral neuropathy compared to investigator-reported peripheral neuropaty assessed by National Cancer Institute Common Terminology Criteria for Adverse Events (CTCAE) version 3.0. The surface of the circles reflects the absolute number of patients plus investigators.
haematologica | 2020; 105(6)
1657