The Effects of Azacitidine on Quality of Life Measured Longitudinally In MDS Patients Treated at a Tertiary Care Center

Myelodysplastic syndromes (MDS) are clonal hematopoetic stem cell disorders characterized by ineffective hematopoiesis and a propensity to develop AML predicted by conventional scoring systems such as the International Prognostic Scoring System (IPSS). Azacitidine (AZA), a hypomethylating agent is indicated for high and high intermediate IPSS scores based on survival and leukemia-free survival benefits demonstrated in randomized trials (Silverman 2002, 2006, Fenaux 2009). Additionally, improvements in fatigue, dyspnea, physical functioning, affect and psychological distress were demonstrated in the CALGB study (Kornblith 2002).

We previously showed that most symptom and functional domains of quality of life (QOL) are impaired in MDS patients measured by several instruments and are primarily determined by Hb and transfusion dependence (Buckstein 2009). With the exception of the CALGB paper, there is a paucity of data assessing the ‘real world' QOL in MDS patients treated with AZA longitudinally. All consented patients with MDS followed at our center have QOL assessed every 3–4 months as part of routine care. We present the QOL scores of patients on AZA as assessed by the EORTC QLQ-C30, EQ-5D and a global fatigue scale.

Clinically significant score differences were considered to be 10 points for the EORTC, and 0.05 for the EQ-5D. Linear regression analysis was used to detect each QOL change over time. Log-transformation was applied for all QOL scores to normalize the distribution. To search for significant predictive factors of each QOL, linear regression analysis (for continuous predictive factors) or Analysis of Variance (for binary predictive factors) was conducted at baseline. A two-sided p-value less than 0.05 was considered statistically significant.

30 patients in our database were/are currently treated with AZA. The median age was 73 years, with 63% being male. Of the 26 patients with measureable IPSS scores, 54% were high/high intermediate risk. Seventy percent had a Hb <100 at the time of baseline QOL prior to AZA, 50% had a ferritin ≥ 1000 ug/L, 65% were transfusion dependent (TD) as defined by the WPSS (Malcovati 2007). Sixty-seven percent of patients were on AZA for ≥ 4 cycles of treatment.

Of the 19 TD patients only 3 became transfusion independent (TI) on AZA and 3 patients who were TI at baseline became TD.

Of the 30 patients, 20 have QOL data available for analysis with a median follow up time of 10 weeks (range 0–80) and an interval duration between QOL assessments of 15.5 weeks. Fourteen out of twenty patients have serial QOL assessments, 5 with two, 9 with three or more. The only clinically significant improvements were observed with the EORTC physical functioning and fatigue subscales but constipation scores were higher and global health status/QOL deteriorated over time (Figure 1). At baseline assessment ferritin ≥ 1000 ug/L was negatively associated with physical functioning (p=.0007), cognitive functioning (p=.0012), global QOL (p=.0048) and global fatigue (p=.0003) while transfusion dependence was not predictive of QOL scores. No significant clinical improvements were detected by linear regression or ANOVA over time, but constipation worsened using both models. The health utilities (determined by the summary score of the EQ-5D) are seen in table 1.

Figure 1:

View largeDownload slide

EORTC QLQ-30 domains with clinically significant changes in longitudinal assessment.

Figure 1:

View largeDownload slide

EORTC QLQ-30 domains with clinically significant changes in longitudinal assessment.

Close modal

Many clinically important function and symptom domains of 3 different QOL instruments have not changed significantly over time in our patients receiving AZA. This is likely explainable by the limited sample size and serial number of assessments in our patients. Our present patient population is higher risk than that tested in the CALGB study and our previous report on the MDS patients in our database. Furthermore, we have yet to see the rates of transfusion independence that might be associated with improved QOL. The negative association of increased ferritin with numerous symptom and function scores may simply be a surrogate for the extent of transfusion dependence. We hope that with longer follow up and larger sample size, we will be able to reproduce the QOL benefits observed in the pivotal CALGB trial (Kornblith 2002).

No relevant conflicts of interest to declare.