Neurotoxic Adverse Effects of Morphine and Oxycodone in Continuous Subcutaneous Infusion for Treatment of Pain in Terminal Patients With Diminished Renal Function: a Randomized Controlled Trial
Neurotoxic Adverse Effects of Morphine and Oxycodone in Continuous Subcutaneous Infusion for Treatment of Pain in Terminal Patients With Diminished Renal Function: a Randomized Controlled Trial
Significant pain is a common condition in dying patients. Continuous subcutaneous infusion (CSCI) of opioids is the cornerstone in treatment of pain in this last phase of life. Although morphine is the most frequent used opioid in this respect, burdensome adverse effects, like delirium and allodynia/hyperalgesia, can occur in dying patients, due to accumulation of morphine metabolites in decreasing renal function. Oxycodone seems preferable in this situation, as central effects of circulating metabolites of oxycodone are negligible. However, studies of sufficient quality investigating the clinical effect of this hypothesis are lacking at the moment.
This study investigates whether there is a difference in occurrence of delirium and allodynia/hyperalgesia between oxycodone and morphine. Residents of hospices and somatic or psychogeriatric (PG) wards of nursing homes in the Netherlands, who are eligible for start of CSCI of an opioid for the treatment of pain in the terminal phase of life, are randomly assigned to one of two groups. One group receives CSCI of oxycodone and the other group CSCI of morphine. 117 patients per group are needed. Occurrence of delirium and allodynia/hyperalgesia is assessed three times a week until death of the participant. Quality of dying, as perceived by the patient's relatives, is assessed in an interview with a relative after death.
INTRODUCTION AND RATIONALE:
Significant pain is a common condition in dying patients. Continuous subcutaneous infusion (CSCI) of opioids is the cornerstone in treatment of pain in this last phase of life. Although morphine is the most frequent used opioid in this respect, its main metabolites start to accumulate when renal function decreases, as is frequently the case in dying patients. The accumulation of one of these metabolites, morphine-3-glucuronide (M3G), is associated with neurotoxic adverse effects like delirium, allodynia and hyperalgesia. These adverse effects can highly effect the patient's comfort in the last days of life. The central effects of circulating metabolites of oxycodone, on the other hand, are negligible. On theoretical considerations CSCI of oxycodone for the treatment of pain in dying patients with a diminished renal function should therefore result in a reduced occurrence of the neurotoxic adverse effects delirium and allodynia/hyperalgesia in comparison to morphine. However, studies of sufficient quality investigating the clinical effect of this hypothesis are lacking at the moment.
OBJECTIVES:
The objective of this study is to investigate whether and to what extent the occurrence of the neurotoxic adverse effects delirium and allodynia/hyperalgesia differs between morphine and oxycodone, administered by continuous subcutaneous infusion (CSCI), for the treatment of pain in dying patients with a diminished renal function.
Primary Objective: To compare the prevalence of delirium between oxycodone and morphine, administered by CSCI, for the treatment of pain in dying patients with a diminished renal function.
Secondary Objective: To compare the prevalence of allodynia/hyperalgesia between oxycodone and morphine, administered by CSCI, for the treatment of pain in dying patients with a diminished renal function.
Hypothesis: CSCI of oxycodone for treatment of pain in dying patients with a diminished renal function results in a reduced occurrence of the neurotoxic adverse effects delirium and allodynia/hyperalgesia in comparison to morphine.
STUDY DESIGN:
An intervention study, designed as a randomized, controlled, observer blinded, multicenter, superiority trial (RCT) with two parallel groups of either CSCI with morphine or oxycodone, with an 1:1-allocation-ratio. The total time of follow up is from start of CSCI with the assigned opioid until death.
STUDY POPULATION:
The study will be conducted in hospices and both somatic and psychogeriatric (PG) wards of nursing homes in the southern part of the Netherlands. 15 nursing homes and 1 hospice are participating. The total number of beds in these locations is around 875 (350 somatic beds, 519 PG beds and 6 hospice beds), dived over 54 wards (24 somatic wards, 29 PG wards, 1 hospice). In order to be eligible to participate in this study, an adult subject must be in the dying phase, i.e. death in the near future is expected by the treating physician, and start of CSCI with an opioid for the treatment of pain must be indicated by the treating physician.
STUDY PROCEDURE:
After written informed consent has been obtained, randomization between CSCI with morphine or oxycodone is performed. Outcome measures will be gathered 3 times a week until death of the participant. A venipuncture for the purpose of the determining the actual renal function at baseline is performed at the first visit. Two weeks after a participant's death, and only when a separate written informed consent has been obtained, a significant other or legal representative will be contacted for an interview in which the perceived quality of dying is assessed by using the Dutch translation of the Quality of Dying and Death (QoDD) Questionnaire.
Randomization: In order to prevent an uneven distribution of relevant prognostic factors, allocation will be stratified by means of minimisation for type of ward (somatic or psychogeriatric, representing absence or presence of a clinical relevant stage of dementia and thereby indirectly also status of mental and communicative capabilities) and for presence or absence of opioid use at baseline.
Sample size calculation: Sample size calculation is based on a two-tailed test. Based on the available literature, the percentage of delirium in terminal patients is estimated at 86%. A difference of 15% is considered to be clinically relevant. The probability of a type 1 error is fixed at 5%. Groups are analyzed according to the intention to treat principle. In order to achieve a power of 80% for detecting a clinically relevant difference, 117 patients per group are needed.
- Statistical analysis: All analyses will be performed according to the intention-to-treat principle. Imputation techniques will be used for supplementation of incomplete data, thereby guaranteeing analysis of all participants in the group they were assigned to by randomization. The method we will use will depend on the proportion of missing values and on the assumptions that can be made about the missingness mechanism after collection of the data: In case the percentage of incomplete cases is less than, or equal to, 5%, we will use single stochastic imputation to impute the dataset, as the difference in precision due to not taking between-imputation variance into account is likely to be negligible. If the proportion of incomplete records exceeds 5%, multiple imputation will be used. In that case, the number of imputations will be set to 10. For both imputation strategies, the imputed values will be drawn using predictive mean matching with a fully conditionally specified model. If, after collection of the data, we expect that data are missing not at random, we will impute using a missing not at random mechanism (again, using single or multiple imputation), and perform a sensitivity analysis to see whether our conclusions are robust.
ETHICS:
The study will be conducted according to the principles of the 7th Declaration of Helsinki (Fortaleza 2013) and in accordance with the Medical Research Involving Human Subjects Act. Written informed consent has to be obtained before participation for all participants. Monitoring will be performed by an independent institution which follows the international ICH-GCP guidelines.
Data will be entered in electronic Case Report Forms (eCRFs), employing data entry validation. This electronic data management system is hosted by an external institution and complies with all applicable regulations regarding data security. Only authorized members of the research staff involved in the study will have access to the system. A full audit trail will be kept. Data storage is secured by regular back-ups by the hosting institution. Data will be stored for 15 years.
No identifiable reference to subjects will be made in analysis, reports or publications. To protect the confidentiality of subjects, all data will only be linked to a subject by a numeric identification code, of which the key will be safeguarded by the investigator. Blood samples for determining the eGFR will be destroyed by the laboratory after 1 week and will not be used for any other purpose than described in the study protocol.
An interim analysis will be performed for the primary and secondary study parameters after the end of follow up, i.e. death, of the first 30 participants in each group. A difference between the intervention and control group of more than 50% for the primary study parameter is considered to be an indication that one of the groups is significantly harmed more than the other. When this difference can't be contributed to other factors, like an uneven distribution of baseline characteristics, especially other medication use, it is considered unethical to withhold one of the groups from the superiority of one of the opioids. Therefore, this potential situation is regarded as a stopping rule.
Objection by incapacitated subjects: Since the major part of dying patients experiences a decline in cognitive functions and are not able to respond adequately anymore, it is essential that this incapacitated population is not being excluded in our study. When a participant is or has become incapacitated, his or her legal representative decides whether the subject will be participating in the study or not, unless the subject previously has stated explicitly otherwise. Signs of consistent objection or resistance to any of the study procedures after inclusion have to be considered as an expression of the participant's wish to not participate in the study any longer. Consequently he or she will be withdrawn from the study immediately. In case a incapacitated participant shows signs of objection or resistance to the continuous subcutaneous route of administration (for example by removal of the infusion tube or needle) the treating physician will decide whether continuation of CSCI of an opioid is a medical necessity to avoid severe suffering from pain. When there is no medical necessity or when adequate pain control can be achieved otherwise, it is not allowed for the subject to participate in the study any longer and he/she will be withdrawn from the study.
Inclusion Criteria:
Exclusion Criteria:
Subjects with an eGFR >50 ml/min/1.73m2 should not be included in the study, because accumulation of metabolites is considered to be irrelevant in this range of renal functions. Despite this fact, we cannot formulate an unaffected renal function as an exclusion criterion prior to allocation, because the time needed for assessment of the renal function could lead to either an unethical delay in treatment of pain or occurrence of death even before the lab results are known. Therefore renal function is assessed after inclusion. In case a subject turns out to exceed the threshold of 50 ml/min/1.73m2, this will be considered as meeting an extended exclusion criterion and the subject concerned will be replaced by a new subject.