DETERMINING MEANINGFUL CHANGE IN THE MYELOFIBROSIS SYMPTOM ASSESSMENT FORM (MFSAF) V2.0 USING A COMBINATION OF DISTRIBUTION- AND ANCHOR-BASED APPROACHES IN THE COMFORT-I TRIAL
Author(s): ,
Amylou Dueck
Affiliations:
Mayo Clinic,Scottsdale, AZ,United States
,
Robyn Scherber
Affiliations:
Oregon Health and Science University,Portland, OR,United States;Mayo Clinic,Phoenix, AZ,United States
,
Heidi Kosiorek
Affiliations:
Mayo Clinic,Scottsdale, AZ,United States
,
Blake Langlais
Affiliations:
Mayo Clinic,Scottsdale, AZ,United States
,
Leslie Padrnos
Affiliations:
Mayo Clinic,Phoenix, AZ,United States
,
Jeanne Palmer
Affiliations:
Mayo Clinic,Phoenix, AZ,United States
,
Srdan Verstovsek
Affiliations:
MD Anderson,Houson, TX,United States
,
Jeff Sloan
Affiliations:
Mayo Clinic,Rochester, MN,United States
,
Geyer Holly
Affiliations:
Mayo Clinic,Phoenix, AZ,United States
Ruben Mesa
Affiliations:
Mayo Clinic,Phoenix, AZ,United States
(Abstract release date: 05/18/17) EHA Library. Dueck A. 05/18/17; 181107; E1331
Dr. Amylou Dueck
Dr. Amylou Dueck
Contributions
Abstract

Abstract: E1331

Type: Eposter Presentation

Background
Symptom response was defined in the COMFORT-I trial as a 50% improvement from baseline at week 24 in the Myelofibrosis Symptom Assessment Form (MFSAF) v2.0 total symptom score (TSS; Mesa [J Clin Onc, 2013]; 0 to 60 scale where 60 represents the worse symptom experience imaginable) with no minimum score requirement at baseline.

Aims
In this analysis of the phase III placebo-controlled COMFORT-I study we used distribution- and anchor-based approaches to investigate whether alternative change scores in the MFSAF v2.0 TSS could be meaningful relative to patient-reported quality of life (QOL).

Methods
One third and one half of the pooled standard deviations (SD) of scores and change scores (raw and percentage change) were used as distribution-based estimates. The anchor-based approach estimated meaningful changes (raw and percentage change) relative to the patient’s change in global health status/QOL (GH/QOL; 0=worst, 100=best) as measured by the EORTC QLQ-C30 where a decrease of 12.1 or more points was considered as deterioration; an increase of 7.6 or more points was considered as improvement; and all other changes were considered as stable based on change scores established in a multiple myeloma population (Kvam et al., Eur J Hem, 2011). Analysis of covariance (ANCOVA) was used to investigate whether estimated meaningful changes were consistent across the spectrum of observed baseline TSS. This model of TSS changes at week 24 included a continuous term for baseline TSS, a 3-level grouping factor for GH/QOL change (deterioration vs stable vs improvement), and an interaction term between baseline TSS and the GH/QOL grouping factor.

Results
301 patients randomized to ruxolitinib [N=149] or placebo [N=152] completed TSS at baseline (45% female, median age 68 [range 40-91]). Median baseline TSS was 16.6 (range 0 to 52.7). Pooled SD at baseline and week 24 in TSS was 11.4 and 11.6, respectively, resulting in estimated meaningful changes of 3.8-5.8 points. For change and percentage change from baseline at week 24 in TSS, the pooled SDs were 9.8 and 75%, respectively, resulting in estimated meaningful changes of 3.3-4.9 points or 25%>38%. Among patients with TSS and QLQ-C30 data at baseline and week 24, 51 (23%) patients had deterioration, 61 (27%) were stable, and 110 (50%) had improvement based on QLQ-C30 GH/QOL changes. Mean (95% CI) changes in TSS for the three groups were 0.8 (-2.5 to 4.2), -1.4 (-3.6 to 0.8), and -6.8 (-9.0 to -4.6), and for percent changes 20% (-6% to 46%), 17% (-11% to 44%) and -34% (-45% to -22%). ANCOVA revealed that baseline TSS statistically significantly impacted meaningful change estimates (p=0.02, Figure 1) resulting in estimated mean (95% CI) changes in TSS for the improved group of -20.8 (-26.4 to -15.1), -11.7 (-14.3 to -9.0), and -2.6 (-5.1 to -0.1) for baseline TSS of 50, 30, and 10.

Conclusion
Distribution- and anchor-based approaches suggest that changes as small as 3-6 points on a 0-60 scale of the MFSAF v2.0 TSS may be meaningful to patients. However, estimates of meaningful change appear to increase in magnitude for higher baseline scores, though in a way that a static percentage change criterion would either require too much change for lower baseline TSS or not enough change for higher baseline TSS. All analyses suggest that some changes in symptoms which do not meet a 50% improvement may still be meaningful to patients.

Session topic: 16. Myeloproliferative neoplasms - Clinical

Keyword(s): Quality of Life, Myelofibrosis

Abstract: E1331

Type: Eposter Presentation

Background
Symptom response was defined in the COMFORT-I trial as a 50% improvement from baseline at week 24 in the Myelofibrosis Symptom Assessment Form (MFSAF) v2.0 total symptom score (TSS; Mesa [J Clin Onc, 2013]; 0 to 60 scale where 60 represents the worse symptom experience imaginable) with no minimum score requirement at baseline.

Aims
In this analysis of the phase III placebo-controlled COMFORT-I study we used distribution- and anchor-based approaches to investigate whether alternative change scores in the MFSAF v2.0 TSS could be meaningful relative to patient-reported quality of life (QOL).

Methods
One third and one half of the pooled standard deviations (SD) of scores and change scores (raw and percentage change) were used as distribution-based estimates. The anchor-based approach estimated meaningful changes (raw and percentage change) relative to the patient’s change in global health status/QOL (GH/QOL; 0=worst, 100=best) as measured by the EORTC QLQ-C30 where a decrease of 12.1 or more points was considered as deterioration; an increase of 7.6 or more points was considered as improvement; and all other changes were considered as stable based on change scores established in a multiple myeloma population (Kvam et al., Eur J Hem, 2011). Analysis of covariance (ANCOVA) was used to investigate whether estimated meaningful changes were consistent across the spectrum of observed baseline TSS. This model of TSS changes at week 24 included a continuous term for baseline TSS, a 3-level grouping factor for GH/QOL change (deterioration vs stable vs improvement), and an interaction term between baseline TSS and the GH/QOL grouping factor.

Results
301 patients randomized to ruxolitinib [N=149] or placebo [N=152] completed TSS at baseline (45% female, median age 68 [range 40-91]). Median baseline TSS was 16.6 (range 0 to 52.7). Pooled SD at baseline and week 24 in TSS was 11.4 and 11.6, respectively, resulting in estimated meaningful changes of 3.8-5.8 points. For change and percentage change from baseline at week 24 in TSS, the pooled SDs were 9.8 and 75%, respectively, resulting in estimated meaningful changes of 3.3-4.9 points or 25%>38%. Among patients with TSS and QLQ-C30 data at baseline and week 24, 51 (23%) patients had deterioration, 61 (27%) were stable, and 110 (50%) had improvement based on QLQ-C30 GH/QOL changes. Mean (95% CI) changes in TSS for the three groups were 0.8 (-2.5 to 4.2), -1.4 (-3.6 to 0.8), and -6.8 (-9.0 to -4.6), and for percent changes 20% (-6% to 46%), 17% (-11% to 44%) and -34% (-45% to -22%). ANCOVA revealed that baseline TSS statistically significantly impacted meaningful change estimates (p=0.02, Figure 1) resulting in estimated mean (95% CI) changes in TSS for the improved group of -20.8 (-26.4 to -15.1), -11.7 (-14.3 to -9.0), and -2.6 (-5.1 to -0.1) for baseline TSS of 50, 30, and 10.

Conclusion
Distribution- and anchor-based approaches suggest that changes as small as 3-6 points on a 0-60 scale of the MFSAF v2.0 TSS may be meaningful to patients. However, estimates of meaningful change appear to increase in magnitude for higher baseline scores, though in a way that a static percentage change criterion would either require too much change for lower baseline TSS or not enough change for higher baseline TSS. All analyses suggest that some changes in symptoms which do not meet a 50% improvement may still be meaningful to patients.

Session topic: 16. Myeloproliferative neoplasms - Clinical

Keyword(s): Quality of Life, Myelofibrosis

By clicking “Accept Terms & all Cookies” or by continuing to browse, you agree to the storing of third-party cookies on your device to enhance your user experience and agree to the user terms and conditions of this learning management system (LMS).

Cookie Settings
Accept Terms & all Cookies