Wu E, et al. ESMO 2017:1119P

A trial-based EuroQoL EQ-5D health utility analysis in patients with classical Hodgkin lymphoma


In the KEYNOTE-087 trial, pembrolizumab showed high responses in patients with relapsed/refractory (R/R) classic Hodgkin lymphoma (cHL).1 An analysis of health-related quality of life (HRQoL) for patients in this trial were presented at the European Society of Medical Oncology (ESMO) 2017 Congress.2

Study design

  • KEYNOTE-087 is an ongoing multicentre, single-arm, phase II trial evaluating pembrolizumab 200 mg via intravenous infusion every three weeks in patients with R/R cHL in three cohorts:
    • Cohort 1: R/R cHL after autologous stem cell transplant (ASCT) and subsequent brentuximab vedotin (BV);
    • Cohort 2: R/R cHL in patients ineligible for ASCT and who have relapsed after treatment with or failed to respond to BV; and
    • Cohort 3: R/R cHL after ASCT but not treated with BV post-ASCT.
  • HRQoL data were collected at baseline, at treatment cycles 1–5, and every 12 weeks until 30 days following treatment discontinuation.
  • HRQoL was assessed using EuroQoL-5 Dimension (EQ-5D) and European Organization for Research and Treatment of Cancer Quality of Life Questionnaire- C30 (EORTC QLQ-C30) instruments.
    • EQ-5D 3-Level covers five attributes: mobility, usual activity, pain/discomfort, anxiety/depression, and self-care.
    • Each attribute was measured by three levels: no problem, some problems, and major problems.
  • Data cutoff for this study was September 2016.
  • Published algorithms were used to convert the generic health statuses from both instruments to population-based utility values.
    • U.K.- and U.S.-based scoring was used for U.K. and U.S. patients, respectively.
    • For all other patients, E.U.-based scoring was applied.
  • HRQoL data was presented by response and disease progression status.
    • Response was defined by International Working Group criteria and was investigator-assessed.
  • HRQoL scores obtained at all visits before progression were used to estimate utility for progression-free state.
  • HRQoL scores at on- versus off-treatment period for progression-free state were estimated.
  • Health disability outcomes were stratified by the patients who experienced grade ≥3 adverse events (AEs) and by Eastern Cooperative Oncology Group performance status (ECOG PS).
    • Time points associated with grade 3–5 AEs were identified for each patient.
    • HRQoL scores collected at these time points were used to estimate health utility by comparing patients with and without AEs.
    • Utility of progression-free state without AEs was estimated using the HRQoL scores of patients without AEs at any time points.
  • This analysis measured statistically significant differences, with a difference of 0.08 (U.K.-based scores) and 0.06 (U.S.-based scores) being the threshold for clinically meaningful differences.3

Key findings

  • The majority of patients in the trial were male (53.8%), younger than 65 years of age (91.4%), had an ECOG PS of 0 or 1 (99.5%), and were previously treated with BV (83.3%).
  • Of the 210 patients on trial, HRQoL data were collected for 207 patients.
  • The mean HRQoL score at baseline for the total population was 0.759, and there was no significant difference in HRQoL scores between cohorts.
  • A clinically meaningful difference in HRQoL scores (0.082) at baseline between patients with an ECOG of 0 (n = 102; mean score: 0.804) and ≥1 (n = 107; mean score: 0.722) was observed.
  • In the total population, the difference in mean HRQoL scores between responders (0.829) and nonresponders (0.764) was considered clinically meaningful (>0.06). (Figure 1)
  • In patients with an ECOG PS ≥1, the difference in mean HRQoL scores between responders and nonresponders was considered clinically meaningful. (Figure 2)
  • The mean utility value for time spent prior to progression (0.821, 95% CI: 0.809–0.833) decreased post disease progression (0.808, 95% CI: 0.784–0.833).
  • There was no statistical difference in HRQoL scores by disease progression status, except in cohort 3.
    • In cohort 3, mean HRQoL for patients who were progression-free was 0.840 (95% CI: 0.821–0.858) and for patients with progressive disease was 0.795 (95% CI: 0.745–0.845).
  • The mean health utility scores for progression-free patients who experienced grade ≥3 AEs (n = 17) versus those who did not were 0.827 (95% CI: 0.814–0.839) and 0.847, respectively.
  • Experiencing a grade ≥3 AE reduced the HRQoL score to 0.737 (95% CI: 0.677–0.796), irrespective of disease state.
  • The following study limitations exist:
    • The response duration and progression-free duration were not taken into account (a repeated measure mixed model will be considered for further investigation); and
    • Due to having only a single follow-up 30 days after treatment discontinuation, HRQoL scores for progressive disease state may be overestimated.

Figure 1. Mean HRQoL scores and 95% CI by response status

Figure 2. Mean HRQoL scores and 95% CI by response and ECOG status

Key conclusions

  • Response to pembrolizumab was associated with clinically meaningful improvement in HRQoL.
  • HRQoL was not sensitive to disease progression, potentially due to the short post-treatment follow-up.
  • Utility estimates from this study will be useful for economic evaluations of treatments in patients with R/R cHL.

References: 1. Chen RW, Zinzani PL, Fanale MA, et al. Pembrolizumab for relapsed/refractory classical Hodgkin lymphoma (R/R cHL): phase 2 KEYNOTE-087 study. J Clin Oncol (ASCO Annual Meeting) 2016;34(Suppl):abstr 7555. 2. Wu E, Liao J, Balakumaran A. A trial-based EuroQOL EQ-5D health utility analysis in patients with classical Hodgkin lymphoma. ESMO Annual Congress Abstracts 2017:1119P. 3. Pickard AS, Neary MP, Cella D. Estimation of minimally important differences in EQ-5D utility and VAS scores in cancer. Health Qual Life Outcomes 2007;5:70–7.