Some text in this manuscript has been reproduced from Stone et al.1 This is an Open Access article distributed in accordance with the terms of the Creative Commons Attribution (CC BY 4.0) licence, which permits others to distribute, remix, adapt and build upon this work, for commercial use, provided the original work is properly cited. See: https://creativecommons.org/licenses/by/4.0/. The text below includes minor additions and formatting changes to the original text.
Rationale and background
Fatigue is a common problem in patients with advanced cancer receiving palliative care.2 It has major effects on quality of life (QoL),3 and patients report that it affects them more than other symptoms, such as pain or nausea.4 The causes of cancer-related fatigue are multifactorial.5 Occasionally, treatment directed at a clearly identified underlying problem can alleviate the fatigue (e.g. blood transfusions for anaemia6). For the majority of patients, no underlying cause is clearly identified and management has to be symptom-directed.7 There is some evidence that exercise8,9 and (to a lesser extent) psychological therapies10,11 can help to alleviate fatigue in these circumstances. However, these therapies are not always appropriate or effective in patients with advanced cancer. Methylphenidate (MPH) is a psychostimulant drug that is widely used as part of the management of people with attention deficit hyperactivity disorder.12,13 It has also been evaluated as a potential treatment for cancer-related fatigue. Some trials and meta-analyses have suggested that this medication may be effective,14–18 but the evidence is mixed with many trials showing no benefit.19–25 In the context of ongoing uncertainty about its role, this research was conducted in response to a commissioned call by the NIHR (HTA no 15/46/02). This synopsis summarises the work we undertook to address the commissioning brief; the main findings have been published in the Journal of Clinical Oncology.1
Objective
To estimate clinical effectiveness of MPH versus placebo for cancer-related fatigue in patients receiving palliative care.
Primary aim
To compare Functional Assessment of Chronic Illness Therapy – Fatigue Scale (FACIT-F) score in patients with advanced cancer receiving individually titrated doses of MPH with patients receiving placebo after 6 weeks’ treatment.
Secondary aims
To compare secondary outcomes [other measures of QoL, adverse events (AEs), activities of daily living, appetite, anxiety, depression, satisfaction of patients, survival and need for other medication] between patients receiving MPH and placebo. To compare adverse effects in patients receiving MPH and placebo.
Methods
The study methods are summarised below. The Study Protocol is available as Supplementary Information (Report Supplementary Material 1). A summary of the major amendments to the protocol and the rationale for the changes are shown in Appendix 1, Table 4.
Design
This was a prospective, double-blind, 1 : 1 randomised, parallel-group, placebo-controlled, multicentre trial.
Participants
Participants were adults with advanced incurable cancer (all types), receiving generalist or specialist palliative care with moderate or severe fatigue [> 3/10 on a numerical rating scale (NRS)]. Exclusion criteria were: pregnancy; breastfeeding; sensitivity to MPH; glaucoma; pheochromocytoma; planned general anaesthesia; concomitant treatment with psychostimulants, clonidine, warfarin, monoamine oxidase inhibitors or modafinil; severe mood disorders; psychosis; hypertension (> 160/100 mmHg); uncontrolled heart failure or angina; arterial occlusive disease; congenital heart disease; cardiomyopathies; myocardial infarction (MI) or stroke (within last year); life-threatening arrhythmias and channelopathies; cerebral aneurysm; cerebrovascular abnormalities; seizures; hyperthyroidism; haemoglobin < 80 g/l; platelets < 50 × 103/μl; white blood count < 1.5 × 109/l; estimated glomerular filtration rate (eGFR) < 45 ml/minute/1.73 m²; alanine aminotransferase (ALT) > 2 or bilirubin > 1.5 × upper limit of normal; infection; substance/alcohol dependency; participation in another study; insufficient English-language skills; or inability to swallow medication.
Settings
Participants were recruited from hospital, community and hospice palliative care services in England.
Study intervention
Initially, participants were prescribed 1 tablet twice daily (=MPH 10 mg/day) or matching placebo. Participants were contacted every week, and the dose of medication was adjusted (either up or down) over the course of the first 6 weeks of the study (dose titration phase), up to a maximum of 12 tablets/day (= 60 mg/day). Thereafter, for the next 2 weeks (dose maintenance phase), the dose of medication was not increased any further (although it could be reduced in response to adverse effects). During the next week (dose tapering phase), the dose was reduced and then the medication was stopped completely for the last week of the trial.
Dose titration schedule
During each of the first 6 weeks of the study, principal investigators (PIs) were permitted to increase (by up to 10 mg/day) the dose of study medication if the participant did not feel fatigue was ‘adequately controlled’, provided there were no dose-limiting AEs (see below). In the presence of dose-limiting AEs, the dose of medication either remained stable or was decreased (depending on the clinical judgement of the PI). If participants reported that fatigue was ‘adequately controlled’, then the dose of study medication was not changed.
Study procedures
After obtaining written informed consent, participants were randomised to receive either MPH or matching placebo tablets in a double-blind fashion. Participants were contacted every week either in person or by telephone. Initially, the assessments at baseline, week 3, week 6 and week 10 occurred face to face, with the other assessments occurring by telephone. However, in response to the COVID pandemic, the trial was modified to permit all assessments, after the baseline visit, to occur remotely if necessary. A flow diagram to summarise the assessments undertaken at each week is shown in Figure 1.
FIGURE 1 Methylphenidate versus placebo for fatigue in advanced cancer flow diagram of study assessments. ECOG, Eastern Cooperative Oncology Group; EORTC QLQ-C15-PAL, European Organisation for Research and Treatment of Cancer Quality of Life Core 15 Palliative Care; EQ-5D-5L, EuroQol-5 Dimensions, five-level version; HADS, Hospital Anxiety and Depression Scale. (Figure not reproduced here; see the original article.)
Primary outcome assessment
The 40-item FACIT-F questionnaire consists of a 27-item FACIT-General QoL questionnaire and a standalone 13-item fatigue questionnaire, known as FACIT-F.26 The primary outcome of this study was fatigue at 6 weeks (± 2 weeks) measured using the FACIT-F score. Each of the 13 items on this questionnaire can be answered on a 5-point scale: ‘Not at all’, ‘A little bit’, ‘Somewhat’, ‘Quite a bit’, ‘Very much’. Scores can range between 0 and 52, with higher scores representing lower fatigue levels. FACIT-F scores were measured at baseline and every week of the study. FACIT-F scores were completed by patients themselves (paper records) at face-to-face visits or responses were collected over the telephone for remote assessments. If FACIT-F scores were not available at week 6, then the scores at weeks 7, 8, 5 or 4 (in that order) were used instead for the primary outcome assessment.
Secondary outcome assessments
Other measures of fatigue
FACIT-F scores were obtained at each study week, and all such measures, except those obtained at week 6 (± 2 weeks), were regarded as secondary outcomes. Fatigue as a secondary outcome was also assessed using the fatigue subscale of the European Organisation for Research and Treatment of Cancer Quality of Life Core 15 Palliative Care (EORTC QLQ-C15-PAL)27 described in more detail below. Scores were obtained at baseline and at weeks 3, 6 and 10.
Quality of life
Quality of life was assessed using the EORTC QLQ-C15-PAL27 and the EuroQol-5 Dimensions, five-level version (EQ-5D-5L).28 The EORTC QLQ-C15-PAL is a 15-item core QoL questionnaire which provides a global QoL score (1 item), a physical functioning (3 items) and emotional functioning (2 items) score; and symptom scores for fatigue (2 items), pain (2 items), nausea (1 item), anorexia (1 item), dyspnoea (1 item), constipation (1 item) and insomnia (1 item). Each item has four possible responses: ‘Not at all’; ‘A little’; ‘Quite a bit’ and ‘Very much’. On QoL and functional scales, higher scores represent better levels of functioning, and on symptom scales, higher scores represent greater severity.
The EQ-5D-5L28 measure provides a simple descriptive profile and a single preference-based index value for health status. It assesses five dimensions of QoL (mobility, self-care, usual activities, pain/discomfort and anxiety/depression), each of which can be scored on a 5-point scale ranging from no problems to inability to do an activity or extreme severity of symptoms. The second part of the measure is a vertical visual analogue scale (VAS) numbered 0–100, with 0 representing ‘The worst health you can imagine’ and 100 representing ‘The best health you can imagine’.
Scores on both scales were obtained at baseline and at weeks 3, 6 and 10.
Activities of daily living
Activities of daily living were assessed using the physical functioning subscale of the EORTC QLQ-C15-PAL (see above) at baseline and at weeks 3, 6 and 10.
Appetite
Appetite was assessed using the anorexia item on the EORTC QLQ-C15-PAL (see above) at baseline and at weeks 3, 6 and 10.
Anxiety and depression
Anxiety and depression were assessed using Hospital Anxiety and Depression Scale (HADS).29 This is a 14-item screening tool consisting of separate 7-item scales for anxiety [Hospital Anxiety and Depression Scale – Anxiety subscale (HADS-A)] and depression [Hospital Anxiety and Depression Scale – Depression subscale (HADS-D)]. Scores on each scale can vary between 0 and 21, with higher scores representing more severe symptoms. HADS scores were assessed at baseline and at weeks 3, 6 and 10.
Satisfaction of patients
Satisfaction of patients was assessed using a study-specific 5-point global benefit score (GBS) and assessed weekly. Respondents were asked to rate their degree of satisfaction by choosing a response to the statement, ‘Overall with regard to fatigue, during the last week, I found that . . . ’. Responses were ‘Things have got much better’, ‘Things have got a little better’, ‘There has been no change’, ‘Things have got a little worse’ and ‘Things have got a lot worse’.
Need for other medication
Participants’ use of other medication (specifically steroids, antidepressants, anxiolytics and analgesics) was recorded weekly.
Survival
Survival of participants was recorded up until the end of the study period (and until at least 75 days after the patient had been randomised).
Blood pressure and pulse
Blood pressure (BP) and pulse were measured at baseline and at weeks 3, 6 and 10.
Adverse events
Adverse events were actively sought by asking participants about the presence or absence of the following known potential side effects of MPH: cough, sore throat, other airway symptoms, abdominal pain, diarrhoea, nausea, vomiting, dry mouth, other gastrointestinal (GI) symptoms, anxiety, depression, irritability, aggression, mood swings, abnormal behaviour, other mood and mental state symptoms, hair loss, itch, skin rashes, other skin and hair symptoms, loss of appetite, loss of weight, feeling heart racing, awareness of abnormal rhythms, headaches, dizziness, drowsiness, difficulty sleeping, abnormal muscle movements (twitching), abnormal activity levels (pacing), joint pain, fever, cold or flu-like symptoms, other symptoms.
Participants were asked to rate whether the AEs were mild, moderate or severe, and PIs were asked to determine whether or not any reported AEs should be regarded as dose-limiting. A dose-limiting AE was defined as one which prevented an increase (or necessitated a reduction) in the dose of MPH. Non-serious adverse events (non-SAEs) were recorded weekly and assessed for severity. SAEs and suspected unexpected serious adverse reactions (SUSARs) were additionally assessed for causality and expectedness and were reported to the sponsor within 24 hours.
Patient satisfaction
Patient satisfaction was measured on a weekly basis using the 5-point GBS30,31 described above.
Statistical methods
The statistical methods are summarised below. The full statistical analysis plan is available as Supplementary Information (see Report Supplementary Material 2). For clarity of presentation, for secondary outcomes only, we have described results which would conventionally be regarded as statistically significant, as being ‘nominally statistically significant’. This is to reflect the issue that although such findings may have reached a conventional level of statistical significance, they were not the primary outcome, and it would therefore be potentially misleading to describe such results as ‘statistically significant’ without further qualification.
Sample size
Randomising 162–230 participants and attaining 130–172 evaluable patients (20–25% attrition) would result in this study having 80–90% power to detect a difference of 5 points on FACIT-F score (effect size 0.5) at 6 (± 2) weeks between groups at 5% significance (two-sided). This was regarded as representing a minimal clinically important difference (MCID).
Randomisation and concealment
Randomisation (1 : 1) was online with random permuted blocks, stratified by: site, receipt of palliative treatment, high depression score (> 10 on HADS-D) and severity of fatigue (> 7/10 on NRS). Randomisation was performed by the PI and undertaken using an independent data management company (‘Sealed Envelope’). Following randomisation, PIs were provided with a randomisation code. Site pharmacies held the randomisation code list which was required to dispense the correct study medication bottles. The study medication bottles’ main labels were blinded; however, they had a tear-off unblinded portion of the label which stated either active investigational medicinal product (IMP) or placebo. Prior to dispensing, the site pharmacy removed the unblinded portion of the tear-off label and filed this in the pharmacy file so that the bottles dispensed were fully blinded. The research team did not have access to the pharmacy file containing the randomisation code list. Other than the dispensing pharmacist, all staff and participants were blind to study allocation. Data were analysed blind to allocation.
Analysis
The trial was analysed by intention to treat. Analysis of the primary outcome used a mixed-effects linear model. Each participant provided two values, one for baseline and one for follow-up, with random intercepts for participant. Fixed effects were time point (baseline or follow-up), randomised treatment and the stratification factors (except site and fatigue severity). Secondary outcomes were analysed in a similar way using mixed-effects linear regression. The GBS, categorised as ‘stayed the same’ or ‘got worse’ versus ‘got better’, was analysed using logistic regression, including randomised group only, because of lack of power. Survival of participants was analysed using Cox proportional hazards regression controlling for receipt of palliative treatment, high depression score (stratification factors). Kaplan–Meier survival curves were generated to display survival from date of randomisation in the two study arms. Frequency and percentage of SAEs, severe and other AEs were analysed descriptively. We examined the sensitivity of the primary outcome to the allocation of any participants who were randomised in error, repeating the primary analysis, including them in the group they were allocated. In addition, we undertook a threshold analysis, in which participants with missing data in the MPH group were attributed a score at the highest 10% of placebo values, and participants with missing data in the placebo group were attributed the median value for the placebo participants. We undertook subgroup analyses with the primary outcome; we first analysed data with an interaction between randomised group and stratification factors (except site) separately. Then we analysed data from each subgroup separately to determine the effect of randomisation on the participants in the given subgroup.
Results
The main results have previously been presented in our linked publication in the Journal of Clinical Oncology.1
The sites at which participants were recruited and the PIs at each site are shown in Appendix 2, Table 5. Across the 17 active sites, 162 participants were randomised {73 men; mean 65.8 [standard deviation (SD) 10.3] years}. The first patient was enrolled on 29 June 2018, and the last patient was randomised on 27 April 2023. The last visit for the last patient was 3 July 2023, which is when the trial ended.
The Consolidated Standards of Reporting Trials diagram to show the flow of participants through the trial is shown in Figure 2. Three participants were excluded from analysis (two patients in the methylphenidate group were randomised in error, and one participant in the placebo group never took the allocated pills). Seventy-seven were allocated and received placebo [baseline FACIT-F score = 22 (SD 10)]; and 82 were allocated and received MPH [FACIT-F score = 20 (SD 9)]. After 6 weeks, the median daily dose of MPH/placebo was six tablets (=MPH 30 mg).
FIGURE 2 Consolidated Standards of Reporting Trials diagram showing flow of study participants. Reproduced from Stone et al.1 This is an Open Access article distributed in accordance with the terms of the Creative Commons Attribution (CC BY 4.0) licence, which permits others to distribute, remix, adapt and build upon this work, for commercial use, provided the original work is properly cited. See: https://creativecommons.org/licenses/by/4.0/. The figure includes minor additions and formatting changes to the original text. (Figure not reproduced here; see the original article.)
Data for analysis of the primary outcome [FACIT-F score at 6 (± 2) weeks] were available for 75 in the MPH group and 72 in the placebo group. Three participants in the MPH group did not have FACIT-F score outcome data at week 6 but were nonetheless included in the analysis of the primary outcome (1 provided usable data at week 5; 2 provided usable data at week 4; no additional data were used at week 7 or 8). Five participants in the placebo group did not have FACIT-F score outcome data at week 6 but were nonetheless included in the analysis of the primary outcome (1 provided usable data at week 5; 4 provided usable data at week 4; no additional data were used at week 7 or 8).
Baseline characteristics
The screening/baseline characteristics of participants are shown in Table 1. The participants in the randomised groups were balanced for key characteristics. Of those participants for whom performance status data were available, the majority in both groups had ECOG status 1 (ambulatory and able to carry out work of a light or sedentary nature) or 2 (unable to carry out any work activities; up and about more than 50% of waking hours). The participants in the MPH arm were slightly more fatigued at baseline than those in the placebo arm.
| Placebo | MPH | |||
|---|---|---|---|---|
| Characteristic | n/N or mean | % or (SD) | n/N or mean | % or (SD) |
| Screening data | ||||
| Male | 35/77 | 45 | 38/82 | 46 |
| Age at randomisation | 62.6 | (11.8) | 64.7 | (11.9) |
| ECOG performance status | ||||
| 0 | 3 | 4 | 3 | 4 |
| 1 | 30 | 39 | 38 | 46 |
| 2 | 27 | 3 | 21 | 26 |
| 3 | 3 | 4 | 5 | 6 |
| 4 | 0 | 0 | 0 | 0 |
| Missing data | 14 | 18 | 15 | 18 |
| Primary diagnosis | ||||
| Breast | 20 | 26 | 21 | 26 |
| Lung | 15 | 20 | 11 | 14 |
| Prostate | 14 | 18 | 9 | 11 |
| Lower GI | 7 | 9 | 11 | 14 |
| Urogenital | 5 | 7 | 5 | 6 |
| Upper GI | 4 | 5 | 5 | 6 |
| Other | 3 | 4 | 3 | 4 |
| Haematological | 2 | 3 | 6 | 7 |
| Gynaecological | 2 | 3 | 5 | 6 |
| Head and neck | 0 | 0 | 1 | 1 |
| Neurological | 1 | 1 | 1 | 1 |
| Unknown primary | 1 | 1 | 0 | 0 |
| Two primary diagnoses | 1 | 1 | 3 | 4 |
| Rare tumour groups | 0 | 0 | 0 | 0 |
| Missing data | 2 | 3 | 1 | 1 |
| Sites of metastases | ||||
| Bone | 32 | 42 | 31 | 38 |
| Lung | 23 | 30 | 20 | 25 |
| Nodal | 17 | 22 | 17 | 21 |
| Liver | 12 | 16 | 16 | 20 |
| Other | 11 | 14 | 15 | 19 |
| None | 6 | 8 | 6 | 7 |
| Adrenal | 2 | 3 | 3 | 4 |
| Malignant pleural effusion | 2 | 3 | 1 | 1 |
| Renal | 1 | 1 | 2 | 2 |
| Brain | 1 | 1 | 2 | 2 |
| Malignant ascites | 0 | 0 | 1 | 1 |
| Unknown | 0 | 0 | 1 | 1 |
| Missing data | 1 | 1 | 1 | 1 |
| Blood results | ||||
| Thyroxine (pmol/l) | 14.7 | (2.9) | 15.1 | (3.2) |
| Thyroid-stimulating hormone (mU/l) | 1.7 | (1.1–2.5) | 1.8 | (1.3–3.0) |
| Haemoglobin (g/l) | 121 | (17) | 123 | (15) |
| Platelets median (× 109/l) | 237 | (189–310) | 248 | (206–298) |
| White blood count (× 109/l) | 6.2 | (4.6–7.9) | 5.8 | (4.4–7.3) |
| eGFR (ml/minute/1.73 m²) | 72 | (60–90) | 69 | (60–90) |
| ALT (U/l) | 16 | (12–31) | 18 | (13–29) |
| Bilirubin (µmol/l) | 7 | (5–9) | 6 | (5–10) |
| Stratification factors | ||||
| Current, recent or scheduled disease-modifying treatment | 62 | 81 | 62 | 76 |
| Depression > 10 on HADS-D | 17 | 22 | 25 | 30 |
| Fatigue > 7/10 on NRS | 26 | 34 | 30 | 37 |
| Baseline data | ||||
| FACIT-F score | 22 | (10) | 20 | (9) |
| HADS depression score median (IQR) | 6 | (4–9) | 7 | (5–11) |
| HADS anxiety score median (IQR) | 4 | (2–9) | 6 | (3–9) |
| EORTC QLQ-C15-PAL | ||||
| Pain | 53 | (20) | 52 | (19) |
| Physical functioning | 50 | (17) | 50 | (18) |
| Emotional functioning | 42 | (17) | 44 | (17) |
| Fatigue | 72 | (18) | 74 | (15) |
| QoL | 31 | (14) | 30 | (13) |
| Nausea | 39 | (22) | 44 | (22) |
| Loss of appetite | 52 | (26) | 52 | (25) |
| Shortness of breath | 53 | (24) | 52 | (21) |
| Constipation | 41 | (21) | 47 | (24) |
| Sleep | 51 | (27) | 56 | (25) |
| EQ-5D-5L utility | 0.65 | (0.18) | 0.62 | (0.21) |
| Systolic blood pressure (SBP) | 126 | (16) | 130 | (16) |
| Diastolic blood pressure (DBP) | 76 | (10) | 77 | (10) |
| Pulse rate | 81 | (13) | 83 | (13) |
| Concomitant medication | ||||
| Strong opioids | 31 | 40 | 31 | 38 |
| Other analgesia | 39 | 51 | 34 | 41 |
| Benzodiazepines | 5 | 6 | 4 | 5 |
| Antidepressants | 22 | 29 | 32 | 39 |
| Steroids | 16 | 21 | 21 | 26 |
| None of the above | 4 | 5 | 2 | 2 |
Primary outcome
The results of the statistical analyses for the primary outcome are shown in Table 2. There was no statistically or clinically significant difference in primary outcome between groups. After 6 (± 2) weeks, FACIT-F scores were 1.97 points [95% confidence interval (CI) −0.95 to 4.90; p = 0.186] higher (better) on MPH than placebo. A sensitivity analysis (including the two participants randomised in error) similarly showed a non-significant difference between groups (FACIT-F score coefficient 2.05, 95% CI −0.85 to 4.95; n = 161).
| Outcome | Estimate | 95% CI |
|---|---|---|
| Primary outcome | ||
| Week 6 (± 2) (n = 159) | 1.97 | (−0.95 to 4.90) |
| Including predictors of missingness (n = 157) | 1.89 | (−1.05 to 4.84) |
| Including those randomised in error (n = 161) | 2.05 | (−0.85 to 5.95) |
| Threshold analysis (n = 159) | 3.15 | (0.26 to 6.04) |
| Stratification factors | ||
| Disease-modifying treatment – yes (n = 35) | 0.12 | (−6.20 to 6.43) |
| Disease-modifying treatment – no (n = 124) | 2.40 | (−0.77 to 5.58) |
| p-value for interaction | 0.002 | |
| Baseline HADS-D > 10 (n = 117) | 2.31 | (−1.05 to 5.67) |
| Baseline HADS-D ≤ 10 (n = 42) | 0.76 | (−5.42 to 6.95) |
| p-value for interaction | 0.552 | |
| Fatigue > 7/10 on NRS (n = 56) | 1.28 | (−4.12 to 6.67) |
| Fatigue ≤ 7/10 on NRS (n = 103) | 2.32 | (−1.01 to 5.66) |
| p-value for interaction | 0.067 | |
| Secondary outcomes | ||
| FACIT-F score (coefficient) (n = 159) | ||
| Week 1 | 0.06 | (−2.41 to 2.53) |
| Week 2 | 3.25 | (0.53 to 5.97) |
| Week 3 | 3.24 | (0.37 to 6.11) |
| Week 4 | 3.16 | (0.16 to 6.16) |
| Week 5 | 3.18 | (0.21 to 6.15) |
| Week 6 | 3.11 | (0.16 to 6.05) |
| Week 7 | 2.72 | (−0.51 to 5.94) |
| Week 8 | 3.47 | (0.40 to 6.54) |
| Week 9 | 1.91 | (−1.12 to 4.95) |
| Week 10 | −0.55 | (−3.56 to 2.47) |
| All weeks | 2.20 | (0.39 to 4.01) |
| EORTC QLQ-C15-PAL (coefficients) | ||
| Pain (n = 159) | ||
| Week 3 | 0.19 | (−4.85 to 5.24) |
| Week 6 | 1.63 | (−3.66 to 6.93) |
| Week 10 | 1.77 | (−3.96 to 7.50) |
| Physical functioning (n = 159) | ||
| Week 3 | −0.31 | (−4.05 to 3.44) |
| Week 6 | −2.67 | (−7.28 to 1.94) |
| Week 10 | 1.70 | (−2.93 to 6.34) |
| Emotional functioning (n = 158) | ||
| Week 3 | 0.75 | (−3.97 to 5.48) |
| Week 6 | 1.71 | (−3.18 to 6.59) |
| Week 10 | 3.40 | (−2.04 to 8.84) |
| Fatigue (n = 159) | ||
| Week 3 | −4.74 | (−9.78 to 0.29) |
| Week 6 | −2.50 | (−7.79 to 2.80) |
| Week 10 | 2.86 | (−2.69 to 8.41) |
| QoL (n = 159) | ||
| Week 3 | 0.71 | (−2.87 to 4.30) |
| Week 6 | −2.09 | (−5.91 to 1.74) |
| Week 10 | −0.25 | (−4.20 to 3.69) |
| Nausea (n = 159) | ||
| Week 3 | 4.05 | (−2.22 to 10.31) |
| Week 6 | 4.30 | (−2.07 to 10.66) |
| Week 10 | 6.14 | (−0.53 to 12.82) |
| Appetite (n = 159) | ||
| Week 3 | 0.85 | (−7.72 to 6.03) |
| Week 6 | 6.71 | (−0.59 to 14.00) |
| Week 10 | 4.04 | (−3.61 to 11.68) |
| Shortness of breath (n = 159) | ||
| Week 3 | 0.03 | (−6.06 to 6.11) |
| Week 6 | −6.17 | (−12.78 to 0.44) |
| Week 10 | −4.71 | (−11.59 to 2.17) |
| Constipation (n = 158) | ||
| Week 3 | 3.24 | (−3.25 to 9.74) |
| Week 6 | 4.64 | (−2.02 to 11.29) |
| Week 10 | 0.41 | (−6.36 to 7.18) |
| Sleep (n = 159) | ||
| Week 3 | −6.55 | (−13.74 to 0.64) |
| Week 6 | −5.49 | (−12.96 to 1.98) |
| Week 10 | −6.85 | (−13.97 to 0.28) |
| EQ-5D-5L utility (coefficient) | ||
| Week 3 (n = 147) | 0.025 | (−0.027 to 0.076) |
| Week 6 (n = 138) | 0.028 | (−0.031 to 0.086) |
| Week 10 (n = 130) | 0.001 | (−0.052 to 0.055) |
| Overall difference in quality-adjusted life-years (n = 159) | 0.009 | (−0.006 to 0.024) |
| HADS anxiety (coefficient) (n = 159) | ||
| Week 3 | −0.70 | (−1.61 to 0.22) |
| Week 6 | −0.39 | (−1.40 to 0.62) |
| Week 10 | −0.32 | (−1.42 to 0.77) |
| HADS depression (coefficient) (n = 159) | ||
| Week 3 | −0.73 | (−1.65 to 0.19) |
| Week 6 | −1.35 | (−2.41 to −0.30) |
| Week 10 | −0.39 | (−1.45 to 0.66) |
| GBS (OR) | ||
| Week 1 (n = 157) | 0.68 | (0.34 to 1.34) |
| Week 2 (n = 152) | 0.77 | (0.41 to 1.45) |
| Week 3 (n = 147) | 0.80 | (0.42 to 1.55) |
| Week 4 (n = 147) | 0.66 | (0.34 to 1.26) |
| Week 5 (n = 140) | 1.48 | (0.76 to 2.88) |
| Week 6 (n = 139) | 1.27 | (0.64 to 2.51) |
| Week 7 (n = 135) | 1.06 | (0.53 to 2.12) |
| Week 8 (n = 134) | 1.31 | (0.64 to 2.70) |
| Week 9 (n = 129) | 1.29 | (0.62 to 2.70) |
| Week 10 (n = 132) | 2.11 | (0.96 to 4.61) |
| Died at any time (OR) (n = 159) | 0.86 | (0.45 to 1.61) |
| Time to death (HR) (n = 159) | 0.98 | (0.66 to 1.47) |
| Increased or started between baseline and 6 weeks (OR) | ||
| Strong opioids (n = 158) | 0.50 | (0.18 to 1.34) |
| Other analgesia (n = 159) | 1.70 | (0.48 to 6.07) |
| Benzodiazepines (n = 157) | 0.46 | (0.04 to 5.21) |
| Antidepressants (n = 158) | 1.95 | (0.35 to 10.95) |
| Steroids (n = 158) | 0.48 | (0.15 to 1.51) |
The predictors of missingness for study data are shown in Supplementary Information (see Report Supplementary Material 3). Patients with lung cancer were more likely to have missing data than other participants, but including predictors of missingness in the statistical modelling did not significantly affect the results of the statistical modelling for the primary outcome. A threshold analysis, in which participants with missing data in the MPH group were assumed to have low levels of fatigue and participants with missing data in the placebo arm were assumed to have average levels of fatigue, was nominally significant (FACIT-F score coefficient 3.15, 95% CI 0.26 to 6.04; n = 159).
Interaction between stratification factors and randomised group
None of the subgroups based on the stratification factors (high/low baseline fatigue; high/low baseline depression; or disease-modifying therapies yes/no) showed significant improvements in FACIT-F score (see Table 2). There was, however, a nominally significant interaction between disease-modifying treatment and primary outcome, with participants who were in receipt of disease-modifying therapy having higher fatigue levels (lower FACIT-F scores) than participants who were not in receipt of such therapy.
Secondary outcomes
Summary data for the primary and secondary outcomes for each week of the study (by randomised group) are shown in Appendix 3, Tables 6–15. Statistical modelling results for secondary outcome data are shown in Table 2.
Other measures of fatigue
There were nominally statistically significant increases in FACIT-F score (lower fatigue) at all study weeks, except for week 1, week 7 (first week of maintenance), week 9 (dose tapering) and week 10 (off study medication). There was also a nominally significant reduction in fatigue across the whole 10-week period (2.20, 95% CI 0.39 to 4.01). There were no significant differences in EORTC-QLQ-C15-PAL-fatigue scores at weeks 3, 6 or 10.
Quality of life
There were no significant differences in EORTC-QLQ-C15-PAL-QoL or EQ-5D-5L utility scores at weeks 3, 6 or 10.
Activities of daily living
There were no significant differences in EORTC-QLQ-C15-PAL-physical functioning scores at weeks 3, 6 or 10.
Appetite
There were no significant differences in EORTC-QLQ-C15-PAL-anorexia scores at weeks 3, 6 or 10.
Anxiety and depression
There was a nominally significant reduction in HADS-D scores at week 6 (−1.35, 95% CI −2.41 to −0.31) but not at week 3 (−0.73, 95% CI −1.65 to 0.19) or week 10 (−0.39, 95% CI −1.45 to 0.66). There were no significant differences in HADS-A scores at weeks 3, 6 or 10.
Patient satisfaction
There were no significant differences in the ORs (stayed the same/got worse vs. got better) for GBSs at any week of the study.
Need for other medication
There were no differences between the groups at week 6 with regard to the likelihood that they needed to start a new concomitant medication (opioid, other analgesic, benzodiazepine, antidepressant or other), or to have the dose of an existing such medication increased (see Table 2).
Survival
There were no significant differences in the number of participants dying, nor in the time to death. The Kaplan–Meier survival curves are shown in the Supplementary Information (see Report Supplementary Material 4).
Blood pressure and pulse
Mean BP and pulse rates were similar in MPH (BP 128/77 mmHg; pulse 82 b.p.m.) and placebo (BP 124/75 mmHg; pulse 84 b.p.m.) groups at week 6.
Adverse events
There were 25 SAEs in 20 participants receiving MPH and 25 SAEs among 16 participants receiving placebo. There were no SUSARs. Six participants receiving MPH and two receiving placebo died within 75 days of randomisation (Fisher’s exact p-value 0.278).
Adverse events in each group at each week of the study are shown in Appendix 4, Tables 16–22. There was no pattern to suggest consistent differences in adverse effects between groups.