Longitudinal opioid utilisation trajectories and domain-specific quality of life in advanced cancer pain: A prospective cohort study using group-based trajectory modelling
Kavshik K B, Rakesh M, Prince Alex Abraham, Nivetha SBackground
Opioid therapy is the cornerstone of advanced cancer pain management, yet requirements vary considerably between patients over time. Identifying distinct longitudinal opioid utilisation phenotypes and their relationship with quality of life (QoL) may help detect patients at risk of poor outcomes.
Objective
To identify longitudinal opioid utilisation phenotypes in advanced cancer pain using Group-Based Trajectory Modelling (GBTM) and evaluate their association with domain-specific QoL.
Methods
In this prospective cohort study, 150 adults with stage III or IV cancer receiving opioids for cancer-related pain were followed monthly for six months. Opioid use was quantified as oral morphine milligram equivalents per day and modelled using GBTM. QoL was assessed with the EORTC QLQ-C15-PAL. Linear mixed-effects models examined QoL changes across trajectory groups, and multinomial logistic regression identified predictors of unfavourable phenotypes.
Results
Four phenotypes were identified: Low-Stable, Moderate-Stable, Escalating, and High-Stable. Participants in the Escalating and High-Stable trajectories had significantly lower physical, emotional, role, and social functioning, and higher pain and fatigue scores than those in the Low-Stable and Moderate-Stable groups (all P < 0.001). Stage IV disease, poor performance status (ECOG ≥3), neuropathic pain, older age, and higher baseline depression scores independently predicted unfavourable trajectories.
Conclusion
Distinct opioid utilisation phenotypes are associated with clinically meaningful QoL differences in advanced cancer pain. Early identification of escalating or persistently high opioid requirements may enable timely individualised pain management, psychosocial support, and specialist palliative care to improve patient-centred outcomes.