1755 - Evaluation of minimal important difference (MID) for the European Organisation for Research and Treatment of Cancer (EORTC) Pancreatic Cancer Module (PAN26) in patients with surgically resected pancreatic adenocarcinoma

Background

PAN26 was developed as a supplement to the EORTC QLQ-C30 to assess pancreatic cancer–specific health-related quality of life (HRQoL) issues. Despite its use in clinical trials, PAN26 MID has not been established. In undertaking the most comprehensive evaluation to date of PAN26’s measurement properties in patients with surgically resected pancreatic cancer receiving adjuvant therapy, we used distribution-based minimum detectable change (MDC) and anchor-based methods to determine PAN26 scale MIDs using the data collected during the Adjuvant Pancreatic Adenocarcinoma Clinical Trial (APACT).

Methods

PAN26 and QLQ-C30 scores were assessed in 12-week intervals at screening (BL), middle and end of treatment, and up to 2.5 years of follow-up. MDC values were calculated via BL standard deviation (SD) and reliability-biased standard error of measurement approaches. Anchor-based values were determined via a linear mixed model in which QLQ-C30 overall health (OH; Item 29) served as a serial patient global impression of severity anchor for predicting PAN26 scores. Pattern-mixture models (PMM) were fit to assess the impact of attrition on estimates.

Results

MDC values were in the 12 to 15 range (Table). These findings were consistent with SD-based values inferred from earlier studies, although our results extended to standalone items. Anchor-based MIDs were almost twice as sensitive as MDC, with values in the 5 to 9 range (PMM values almost identical). Anchor values could not be inferred for the PAN26 bowel, hepatic, and healthcare scales and gas item due to their low (r < 0.30) correlation with OH.Table:

687P PAN26 anchor (QLQ-C30 overall health) and distribution-derived MID

BLSD, baseline standard deviation; SE, standard error; SEM, standard error of measurement (based on intraclass correlation).

Low r (< 0.3) with overall health.

Conclusions

The MID estimates for PAN26 subscales can help clinicians and researchers interpret the changes in HRQoL and determine sample sizes in the design of future clinical trials.

Clinical trial identification

Editorial acknowledgement

Pharmerit International.

Legal entity responsible for the study

The authors.

Funding

Celgene Corporation.

Disclosure

M. Reni: Non-remunerated activity/ies, Personal Fees: Celgene, Baxalta, Shire, Eli Lilly, Pfizer, Novocure, Novartis, AstraZeneca. J. Braverman: Full / Part-time employment: Celgene Corporation. T. Macarulla Mercade: Honoraria (self): Roche, Sanofi, Tesaro, Shire, Genzyme; Advisory / Consultancy: Baxalta, Celgene, H3B, QED, Shire; Speaker Bureau / Expert testimony: Celgene, Sanofi/Aventis, Shire; Research grant / Funding (self): Agios, Aslan, AstraZeneca, Bayer, Celgene, Genetech, Hallozyme, Immunomedics, Lilly, Merimarck, Millennium, Novartis, Novocure, Pfizer, Pharmacyclics, Roche; Travel / Accommodation / Expenses: Bayer, H3B, Merck, Sanofi. D. Oh: Advisory / Consultancy: Genentech/Roche, AstraZeneca, Novartis, Merck Serono, Bayer, Taiho, ASLAN, Halozyme, Zymeworks; Research grant / Funding (institution): AstraZeneca, Novartis, Array, Eli Lilly. H. Riess: Non-remunerated activity/ies, Personal Fees: Celgene, Roche, Shire. M.A. Tempero: Advisory / Consultancy: AbbVie, Advance Medical, BioPharm Communications, BMS, Celgene, Eisai, Ignyta, Pharmacyslics, Pharmcyte Biotech, Tocagen; Advisory / Consultancy: AstraZeneca, CPRIT, Immunovia; Research grant / Funding (institution): Halozyme. B. Lu: Shareholder / Stockholder / Stock options, Full / Part-time employment: Celgene Corporation. J. Marcus: Full / Part-time employment: Pharmerit international; Advisory / Consultancy, Research grant / Funding (self): Celgene Corporation. N. Joshi: Full / Part-time employment: Pharmerit International; Advisory / Consultancy: Celgene Corporation. M. Botterman: Honoraria (institution): Celgene, Bayer, Daiichi, BMS. All other authors have declared no conflicts of interest.