��������� ������

��������������������� ����� ��������� ���� ����� ������ ���������������������

��������!"��������#�$�%#��&�� $#�����$�#�'� ��&"�(���#��%#��&��$#������ ��'"���)���#��**+#�$�%#�+���� �+"����,� #�*��-(� �.#�(�����"�+� ���#$�%#�$������'"�+������#�$�%#�$������� ��#�$�%#��&��$#�%�����"������/#��&��$#�$(

$00� +1223�4567-�3.11�89��

%:0� �1"�1�9;<"<��� ��,,� "61�3"14"115

&����� ��� '$+�2289

����������� � Journal of Pain and Symptom Management

&�������%��� �4�:��=���61�7

&������%��� 64�������61�3

�������%��� ��������61�3

$��������������������������������� !#�(����'&#���)�����'#����,� �+��+#�+� ����(�#�+�������$'# ���$#������/�%�#���������������������� ����� ��������� ���� ����� ������ ����������������������#���������� �������������������-61�3.#������1"�1�9;<"<��� ��,,� "61�3"14"115"

����������$%����������� �� ����,� ��������������=�� ��������������=������ "�����������������������,����/������������ ������������������ �������,� ������"�����,� �������/����� ����������� �#������� �#�� ������/������������� ��������=�������������=������� ������ ������,"�$����� ��������� ������������ ����������������,���=�����������/��������������������� � #�� ����������������,������������������<��� �������� "

MANUSCRIP

T

ACCEPTED

ACCEPTED MANUSCRIPT

1

Original Article 14-00543R1

The Effect of Methylphenidate on Fatigue in Advanced Cancer: An Aggregated N-of-1

Trial

Geoffrey K. Mitchell, PhD, FRACGP, FAChPM, Janet R. Hardy, MD, FRACP, Catherine J.

Nikles, MBBS, PhD, Sue-Ann S. Carmont, BA (Hons), Hugh E. Senior, PhD, Philip J.

Schluter, PhD, Phillip Good, PhD, FRACP, and David C. Currow, FRACP, MPH

School of Medicine (G.K.M., C.J.N., S.-A.S.C., H.E.S.), University of Queensland, Brisbane;

Department of Palliative and Supportive Care (J.R.H., P.G.), Mater Health Services, South

Brisbane; Mater Research (J.R.H., P.G.) -University of Queensland, South Brisbane,

Queensland, Australia; School of Health Sciences (P.J.S.), University of Canterbury,

Christchurch, New Zealand; St. Vincent’s Private Hospital (P.G.), Brisbane, Queensland;

Cancer Institute of New South Wales (D.C.C.), Sydney, New South Wales; and Department

of Palliative and Supportive Care (D.C.C.), Flinders University, Adelaide, South Australia,

Australia

Address correspondence to:

Geoffrey K. Mitchell, PhD, FRACGP, FAChPM

University of Queensland School of Medicine

Ipswich Campus

Salisbury Road

Ipswich, 4305, Queensland, Australia.

E-mail g.mitchell@uq.edu.au

MANUSCRIP

T

ACCEPTED

ACCEPTED MANUSCRIPT

2

Abstract

Context. Fatigue is common in life-limiting cancer. Methylphenidate (MPH), a

psychostimulant, may be a useful therapy. Gathering evidence in patients with advanced

cancer can be challenging.

Objectives. To determine if MPH improves cancer-related fatigue in people with

advanced cancer.

Methods. N-of-1 trials are multi-cycle, double-blind, randomized, controlled

crossover trials using standardized measures of effect in individuals. They are normally used

to assess treatment effects in individuals. Aggregated N-of-1 trials from participants with

end-stage cancer suffering fatigue were used to assess the group effect of MPH, producing an

estimate of equivalent power to a parallel group randomized controlled trial (RCT) but

requiring less than half of the sample size. Up to three cycles of MPH 5mg twice daily (three

days) versus identical placebo (three days) capsules were offered to participants. Primary

outcome was improvement in fatigue as measured by the Functional Assessment of Chronic

Illness Therapy-Fatigue Scale and the Wu Cancer Fatigue Scale. Analysis employed

Bayesian statistical methods using intention-to-treat principles.

Results. Forty-three participants completed 84 cycles of MPH and placebo in random

order – exceeding sample size estimates. Overall, MPH did not improve fatigue (mean

difference 3.2; 95% credible interval: -2.0, 9.0; posterior probability of favorable effect

0.890). Eight participants showed important improvement and one showed important

worsening of fatigue on MPH. There were no features that distinguished participants whose

fatigue responded to MPH compared with those who did not.

Conclusion. MPH does not improve fatigue in the population of patients with end-

stage cancer. Aggregated N-of-1 trial methodology is feasible and produces population-based

sample estimates with less than half the sample size required for the equivalent parallel group

MANUSCRIP

T

ACCEPTED

ACCEPTED MANUSCRIPT

3

RCT. It also identified individuals who did and did not respond to MPH, which is a feature

difficult to achieve in a standard RCT. The study was registered with the Australian Clinical

Trials Registry (ACTRN 12609000794202).

Key Words: methylphenidate, fatigue, N-of-1 trial, palliative care, advanced cancer

Running Title: Aggregated N-of-1 Trial of MPH for Fatigue

Accepted for publication: April 1, 2015

MANUSCRIP

T

ACCEPTED

ACCEPTED MANUSCRIPT

4

Introduction

Fatigue is “a distressing, persistent, subjective sense of tiredness or exhaustion related

to cancer or cancer treatment that is not proportional to recent activity and interferes with

usual functioning” (1). The fatigue of patients with cancer, particularly those with co-existing

anaemia, is significantly worse than that of non-cancer patients (2, 3). In advanced cancer,

the prevalence of fatigue is 60-90% and can be related to the treatment or the disease itself (4,

5).

Methylphenidate hydrochloride (MPH) is a central nervous system stimulant. It

blocks the dopamine transporter in the pre-synaptic cell membrane, thereby increasing

extracellular dopamine (D2) levels (6). The U.S. National Cancer Care Network (NCCN)

Guidelines on cancer-related fatigue recommend the use of psychostimulants for fatigue,

where the cause cannot be found, or depression, where a rapid response is required (1). When

the drug is used in carefully titrated doses, side effects are usually mild and reverse rapidly

with cessation (6). There was no treatment x time interaction over eight weeks for MPH in

people with brain tumors undergoing radiation therapy (7).

A systematic review on the effect of psychostimulants on cancer fatigue demonstrated

a small effect on fatigue (8). However, the largest of the five reported trials, which

demonstrated a positive effect, was in chemotherapy-related fatigue (9). The other four,

including two in advanced cancer, had negative results (10, 11). A 2010 Cochrane review of

treatments for fatigue in end-stage disease (12) found two trials of MPH (7, 10). The meta-

analysis of these trials showed a small benefit from MPH. One further trial of

dexamphetamine in patients with advanced cancer showed no benefit (11). Subsequent to

2010, a trial of MPH combined with a nurse telephone intervention demonstrated no

improvement with either intervention (13). Similarly, a crossover trial of MPH in a

heterogeneous cancer population where primary treatment had been completed showed no

MANUSCRIP

T

ACCEPTED

ACCEPTED MANUSCRIPT

5

benefit for fatigue, but improvements in cognition and hours worked (14). In contrast, studies

of patients with advanced prostate cancer (15) and a heterogeneous palliative care population

(16) showed improvements in fatigue with MPH. Thus the evidence supporting the

effectiveness of MPH in advanced cancer is inconclusive.

Gathering evidence in people at the end of life is challenging, with many trials failing

through inability to achieve the required sample size in parallel group randomized controlled

trials (RCTs) (17). An alternative trial design that may overcome these difficulties is the

aggregated N-of-1 trial design. N-of-1 trials are multi-cycle, double-blind, controlled

crossover trials using standardized measures of effect (Fig. 1). While normally used as a test

to provide the strongest possible evidence of effectiveness of a treatment in an individual

(18), aggregation of multiple N-of-1 trials also provides the robustness of a normal parallel-

arm RCT design. In addition, each participant contributes multiple sets of perfectly matched

data to both arms of the trial (19, 20). If a participant withdraws early, any completed pairs of

treatment can still contribute to the final analysis. Further, participants can resume a new

cycle if it is possible to stabilize their condition. N-of-1 trials can be used if the treatment has

a short half-life, with a rapid onset and offset of action, and is treating but not altering the

natural history of the target symptom (20). A limitation is the need for the patient’s

underlying condition to remain stable within each cycle. MPH meets all the conditions of a

treatment amenable to the N-of-1 trial design (20).

The primary aim of this study was to determine a population estimate of the efficacy

of MPH compared with placebo in alleviating fatigue in patients with advanced cancer

compared with placebo, using an N-of-1 trial design. Secondary aims were: to determine the

effect of MPH on depression and global well-being; to assess the frequency and severity of

adverse events associated with MPH therapy; and to assess the feasibility of conducting N-of-

1 trials in palliative care populations.

MANUSCRIP

T

ACCEPTED

ACCEPTED MANUSCRIPT

6

Methods

Each N-of-1 trial comprised a series of three double-blind, crossover RCTs of MPH

versus placebo (identical in appearance and taste). Each cycle comprised a pair of three-day

periods, thus the full trial was 18 days. The washout of medicine from one period to the next

requires a time period equivalent to five half-lives (21). Because the half-life of MPH is four

hours (22), data from the first day of each three-day period were not analyzed. Completed

cycles of all individual N-of-1 trials were aggregated to achieve a population estimate of

effect.

Patients who experienced an important change in their clinical condition during the

trial had the option of restarting the trial with a new cycle once their condition stabilized. A

significant change in clinical condition was defined as a change of ≥≥≥≥10 on the Australian

Karnofsky Performance Scale (AKPS) (23), the development of a new symptom requiring a

significant new clinical intervention, or a significant change in the dose of a current

medication.

The trial was discontinued at the end of a cycle if the participant ceased the trial

medication, developed intolerable side effects, the clinical condition deteriorated

significantly, or the patient or physician judged that cessation was in the best interests of the

patient.

Ethical approval for this study was given by the University of Queensland (no.

2009001088) and the Ethics Committees of the participating hospitals. All participants

provided written informed consent. The study was registered with the Australian Clinical

Trials Registry (ACTRN 12609000794202).

Participant Selection

Participants were drawn from six specialist palliative care services in two states

(Queensland and New South Wales) of Australia.

MANUSCRIP

T

ACCEPTED

ACCEPTED MANUSCRIPT

7

Inclusion criteria included: age ≥≥≥≥18 years; a diagnosis of advanced cancer; an AKPS

score of ≥40, screening fatigue score ≥≥≥≥4/10 (NCCN fatigue guideline single screening

question (8)); a stable treatment regimen (including steroids) for at least 48 hours; no plan for

treatment likely to influence fatigue (e.g., chemotherapy, blood transfusion) during the trial;

and no change in thyroxine, antidepressant therapy or other drugs with sympathomimetic

potential for three weeks prior to recruitment. Patients with hepatic or renal dysfunction were

not excluded.

Patients were excluded if they were unable to comprehend written English; were

confused or had a Mini-Mental State Examination score <24 (24); had unstable symptoms or

disease; symptoms of anemia warranting blood transfusion; a history of severe ischemic heart

disease; uncontrolled arrhythmia or hypertension; erythropoietin therapy in the last two

weeks; or electrolyte imbalances where attempt at correction was being considered.

Randomization and Masking

Treatments were MPH and placebo, in capsules identical in appearance and taste.

Randomization of treatment order was conducted using a computer-generated randomization

schedule, held by the pharmacy. The randomization order was revealed at the end of each N-

of-1 trial so a clinical report could be generated by the study coordinator, who was not a

clinician and who took no part in the selection of patients. If a participant discontinued the

trial prior to completion and then elected to restart, new medication packs were prepared for

the uncompleted cycle(s).

Measures of Fatigue and Depression

The Functional Assessment of Chronic Illness Therapy-Fatigue (FACIT-F) fatigue

subscale is a 13-question, 5-point Likert scale with higher scores indicating less fatigue

(score range 0-52) (25). It is a reliable and valid measure of fatigue in cancer (14, 15), and

MANUSCRIP

T

ACCEPTED

ACCEPTED MANUSCRIPT

8

sensitive to detecting the impact of fatigue in cancer patients. It correlates with self-reported

capacity to perform everyday activities (25).

The Wu Cancer Fatigue Scale (WCFS) is a valid and reliable nine-question, 5-point

scale, with a score range of 5 to 45 (26). A lower score indicates less fatigue.

The Edinburgh Depression Scale (EDS) is a self-assessment scale comprising ten

items, each rated on a 4-point scale. In patients with metastatic cancer receiving palliative

care, sensitivity is 70% and specificity 80% (27).

Patient performance status was assessed using the AKPS scale (23). This scale has

been validated in palliative care patients and is sensitive to changes near death. Both the EDS

and AKPS were recorded at baseline, and at the end of each three-day period.

Assessment of Fatigue

Patients completed a daily diary recording symptom scores and any side effects. At

the end of the trial, the within-cycle order of medications was unmasked and the treatment

order was compared with the patient’s observations. Where the results of all three cycles

favored MPH treatment periods over placebo, the participant was reported as being a

responder; two cycles of three as possible response; and one or no cycles favoring MPH as

non-response. The clinical importance of the result was described by comparing the result to

a predetermined clinically important change of 8 on the FACIT-F fatigue subscale, the

difference noted in a previous study of palliative care patients (10).

Dose Selection

A dose-ranging study was conducted previously to determine the dose of MPH to be

tested (28). The trial tested an escalating range of doses from 5mg daily to 30mg daily, to

determine the dose beyond which an increase in dose gave progressively less net benefit. The

dose selected was 5mg twice daily.

Sample Size

MANUSCRIP

T

ACCEPTED

ACCEPTED MANUSCRIPT

9

The FACIT-F fatigue subscale was the primary outcome measure. In a previous

double-blind RCT trial of patients with life-limiting cancer (10), the mean baseline FACIT-F

fatigue score was 17.0 (standard deviation [SD] 7.9) and the mean change in FACIT-F

fatigue scores from baseline for patients treated with the placebo was 7.5 (SD 11.3). For a

conventional RCT, the sample size required to detect a difference in effect of 8 on the

FACIT-F fatigue subscale between MPH and placebo with 5% significance level and 80%

power, using a two-sided test, is 33 per treatment group, inflated to 47 per group (94 in total)

to allow for 30% attrition. Using the same information, assuming no period effect or

treatment x time interaction, computer simulation of size N = 10,000 in SAS v. 9.2 (SAS

Institute Inc., Cary, NC) was used to model the required sample size for the equivalent N-of-

1 trial design. If 60% of recruited patients complete the first cycle, 50% complete the first

two cycles, and 45% complete all three cycles, then 21 patients would be needed to satisfy

the same significance and power requirements. We aimed for a sample of 40 patients, to

account for possible patient clustering and conservative estimates of dropout.

Reporting of Adverse Events

We recorded all side effects and reportable adverse events, graded according to

standard NCCN Clinical Practice Guidelines in Oncology criteria (6). Reportable adverse

events were reviewed by an independent Data Safety and Monitoring Committee to assess

possible relationships between the events and the trial treatment. The dose ranging study

minimized the risk of serious adverse events.

Statistical Analysis

The data generated from this study were used in two ways. The first was to provide an

assessment of the effectiveness of MPH for fatigue in each participant, comparing the result

against a benchmark of a change of 8 on the FACIT-F fatigue subscale (10).

MANUSCRIP

T

ACCEPTED

ACCEPTED MANUSCRIPT

10

Second, we generated a population estimate of effect by aggregating individual N-of-

1 results. We used hierarchical Bayesian methods using normal likelihood distributions, as

proposed by Zucker et al. (29), or non-normal likelihood distributions, such as that proposed

by Schluter and Ware.(30). This method provides probabilistic estimates of effect for

individuals and patient groups. We employed non-informative prior and hyper-prior

distributions. These distributions (which are invariant under transformation) give no

preference to the domain of possible outcome measures. The results of all complete cycles

were combined to produce a posterior probability of the overall difference between the MPH

cycles and the placebo cycles, commensurate with a standard RCT. We defined a positive

response to treatment to have occurred when the posterior probability of the mean difference

favoring the treatment exceeded 0.975 and a negative response when the posterior probability

of the mean difference favoring the treatment fell below 0.025. The clinical characteristics of

responders and non-responders to MPH were compared with nonparametric methods

(including Fisher’s exact test and Wilcoxon’s rank-sum test).

As this was an intention-to-treat analysis, all recruited participants were included in

the analysis. However, participants who did not complete any cycles contributed no data to

the analysis, and their resultant posterior distributions ended up being a weighted

combination of group estimates and non-informative prior values.

Results

We recruited 43 patients between June 2008 and December 2011 (Table 1). These

participants completed 84 cycles: 33 (76.7%) participants completed one cycle, 29 (65.1%)

completed two cycles, and 24 (55.8%) completed three cycles. Ten (23.3%) patients did not

complete any cycles, of whom six withdrew before collecting any data, and one withdrew

because of a side effect judged to have a possible relationship to trial medication. On both

MANUSCRIP

T

ACCEPTED

ACCEPTED MANUSCRIPT

11

numbers of patients and proportion of participants completing each cycle, the sample size

targets were exceeded (Fig. 2).

Effect of MPH on Fatigue

The mean population estimate was calculated using the results of all 43 participants,

including the non-informative results from the ten patients who did not complete one cycle.

The results for the 33 patients who completed at least one cycle are shown in Fig. 3. Neither

the mean population estimate between MPH and placebo for the FACIT-F fatigue subscale

(3.2, 95% credible interval [CI] -2.0, 9.0; posterior probability of favorable effect [PP] 0.890)

nor the WCFS (-3.2, 95% CI -7.0, 0.6; PP 0.952) showed an important difference. Eight of

the participants had important individual improvements in fatigue on both the FACIT-F

fatigue subscale and WCFS scores, with seven having a positive effect greater than the

clinically important difference of 8 on the FACIT-F fatigue score. A further three

demonstrated a favorable effect on the WCFS alone. One patient had an important worsening

in fatigue on MPH compared with placebo in both scales. There were no differences in the

clinical or demographic characteristics or the number of completed cycles of those who

responded favorably to MPH on both the FACIT-F fatigue subscale and the WCFS compared

to those who either had no response or an unfavorable response (Table 2, Fig. 3).

Effect of MPH on the Edinburgh Depression Scale

The population mean estimate showed no important improvement between treatments

in the EDS (-0.7, 95%CI -1.8, 0.4; PP 0.907). Seven patients had important individual

improvements in the EDS, and one had an important worsening in depression. This person

was a different person from the one whose fatigue responded unfavourably to MPH. Six of

the seven patients with important individual improvements in EDS scores also had important

individual improvements in fatigue scores.

Adverse Effects

MANUSCRIP

T

ACCEPTED

ACCEPTED MANUSCRIPT

12

Six events leading to withdrawal required reporting to the Data Safety and Monitoring

Committee. Of these, one level 2 event and two level 1 events were judged to have possible

relationships to trial medication (Table 3).

Change in Functional Status Over the Trial

Over the course of the full trial, the median change in AKPS functional status was 0

(interquartile range -10, 0), with a range of -40 to +40. There was no relationship between the

direction of change of AKPS and changes in fatigue or depression scores.

Discussion

Principal Findings

In this study, MPH had no effect on fatigue in a population of patients with advanced

cancer, supporting previous studies showing that MPH does not improve fatigue in advanced

cancer or end-of-life populations (10, 11). While there was no important group response to

MPH in this population, this trial method identified that about one-quarter of individual

participants had a clinically important benefit, and one had a clinically important worsening

of their fatigue.

We also demonstrated the feasibility of deriving a population estimate of effect using

N-of-1 trial methodology, gaining appropriate power with a cohort 45.7% of the sample size

of an equivalent RCT.

Strengths and Weaknesses of the Study, and in Relation to Other Studies

Problems with N-of-1 Design and Potential Sources of Bias. There are potential

pitfalls with the N-of-1 trial design, and the method is not universally applicable (20). N-of-1

trials will only work when the underlying condition is stable, and where the treatment does

not alter the underlying condition (for example, as antibiotics do). The test treatment must

have a rapid onset, and a rapid withdrawal of effect when the treatment is withdrawn. This

precludes treatments with a cumulative effect or where the effect is not completely dose-

MANUSCRIP

T

ACCEPTED

ACCEPTED MANUSCRIPT

13

dependent (for example, antidepressants). The treatment has to possess a short half-life for

the method to be practical. Finally, there needs to be a validated measure of the desired

effect.

Potential non-sampling and confounding factors associated with recruitment and

retention of the participant group over the full trial implies there is a risk they are not

representative of the broader population under study, and an evaluation of representativeness

is desirable when interpreting the results. However, the small sample size limits the utility of

such an evaluation, a problem common to most oncological studies. In this study, there was

no difference in the proportion of responders or probable responders regardless of the number

of cycles completed. Those who withdrew early did so for a variety of reasons. Of those who

withdrew because of adverse events, none was definitely related to the trial medication.

The feedback of results to patients at the time each completes participation, and

before the overall trial’s end, may unblind the treating clinician to particular characteristics of

responders, and might influence selection of future participants. Ideally, the clinician

recruiting to the study is different to the one receiving the results and treating the patient. In

this trial, there was no difference in demographic or clinical characteristics between the

responders and non-responders. The impact of this potential problem is further reduced by

the study design where all participants received both test and placebo conditions in random

order, and the primary observations were those of the participant, not the clinician.

It is possible that the short cycle times mask a progressively positive methylphenidate

effect. However, this possibility is contradicted by a recent trial (11), where a similar

stimulant (dexamphetamine) over eight days showed a positive effect on fatigue on day 2, but

no effect at the end of study (day 8). However, improvement at 14 days (and with dose

titration) was reported in a trial of MPH reported in 2012 (16). It is also possible that patients

who responded stayed in the trial, whereas those who did not withdrew early. Early

MANUSCRIP

T

ACCEPTED

ACCEPTED MANUSCRIPT

14

withdrawals from the study (Fig. 2) were for a variety of reasons, so this source of potential

bias is unlikely.

Our study involved participants with a wide range of cancer types, and the possibility

of within-group similarity and between-group differences exists. The small sample size

precludes subgroup analysis to identify these differences. Further, the focus of palliative care

research is management of symptom itself, not the disease from which it arose. MPH has

been recommended as a potential treatment for fatigue at the end if life, regardless of disease

state (1). Several studies of symptom management at the end of life take this approach (10,

31-33).

Meaning of the Study

Although MPH does not improve fatigue in patients with advanced cancer, the use of

N-of-1 trial methodology made it possible to identify individuals who did respond or become

worse on the active treatment. Rational clinical decisions were made immediately after trial

participation on the basis of these individual findings. In future clinical practice, routine N-

of-1 tests could be a useful decision-making tool for treatments that may have adverse

consequences or costs. By comparison, clinicians using RCT evidence interpret the

population estimate to assess the likely benefit or otherwise to a given patient, thereby

potentially missing on an opportunity to offer effective treatment to responders in a

population.

The ability of aggregated N-of-1 trials to deliver adequately powered results has

obvious advantages in patient populations where gathering evidence is difficult to achieve

(20). Patients receiving palliative care typify populations where participants are hard to

recruit, and retain on trial (34). Where the treatment characteristics allow the method to be

used, this may be the design of choice in this setting.

Future Research

MANUSCRIP

T

ACCEPTED

ACCEPTED MANUSCRIPT

15

This study was not powered to determine characteristics that predict response, and

determining this would be clinically important. Further N-of-1 trials of treatments for

symptoms in palliative care populations would both produce evidence, and allow gathering of

further evidence about the place of the method in palliative care research.

Disclosures and Acknowledgments

This study was funded by a grant from the National Health and Medical Research

Council (no. 456223). Jane Nikles was funded by an NHMRC post-doctoral research

fellowship (no. 401780). The authors declare no financial relationships with any

organizations that might have an interest in the submitted work in the previous three years,

and no other relationships or activities that could appear to have influenced the submitted

work. All authors have completed the Unified Competing Interest form, which are available

on request from the corresponding author.

The authors acknowledge the clinicians and research staff at each site, in particular:

Angela Tapuni, Mater Adult Hospital, South Brisbane; Kerri-Anne Dooley, St. Vincent’s

Hospital, Kangaroo Point; Rohan Vora, Gold Coast Hospital, Robina; Meera Agar, Braeside

Hospital, Sydney; Katherine Clark, Calvary Hospital, Newcastle; Carol Douglas, Royal

Brisbane and Women’s Hospital, Herston; and Judith McEniery, Ipswich Hospital, Ipswich.

The authors also acknowledge early input into this project from Associate Professor Michael

Yelland.

References

1. Cancer Related Fatigue panel, Berger AC. NCCN Clinical Practice Guidelines in

Oncology- Cancer-Related Fatigue. Fort Washington, PA: National Comprehensive Cancer

Network, 2015.

2. Cella D, Lai JS, Chang CH, Peterman A, Slavin M. Fatigue in cancer patients

compared with fatigue in the general United States population. Cancer. 2002;94:528-38.

MANUSCRIP

T

ACCEPTED

ACCEPTED MANUSCRIPT

16

3. Fosså S, Dahl A, Loge J. Fatigue, anxiety, and depression in long-term survivors of

testicular cancer. J Clin Oncol 2003;21:1249-1254.

4. Piper RE, Woo P, Beiss M, et al. The revised Piper Fatigue Scale: Psychometric

evaluation in women with breast cancer. Oncol Nurs Forum 1998;25:677-684.

5. Vogelzang NJ, Breitbart W, Cella D, et al. Patient, caregiver, and oncologist

perceptions of cancer-related fatigue: results of a tripart assessment survey. Semin Hematol

1997;34:S(2) 4-12.

6. National Comprehensive Cancer Network. NCCN clinical practice guidelines in

oncology- adult cancer pain. Fort Washington, PA: NCCN, 2010.

7. Butler JJ, Case L, Atkins J, et al. A phase III, double-blind, placebo-controlled

prospective randomized clinical trial of d-threo-methylphenidate HCl in brain tumor patients

receiving radiation therapy. Int J Radiat Oncol Biol Phys 2007;69:1496-1501.

8. Minton O, Richardson A, Sharpe M, Hotopf M, Stone PC. Psychostimulants for the

management of cancer-related fatigue: a systematic review and meta-analysis. J Pain

Symptom Manage 2011;41:761-767.

9. Lower EE, Fleishman S, Cooper A, et al. Efficacy of dexmethylphenidate for the

treatment of fatigue after cancer chemotherapy: a randomized clinical trial. J Pain Symptom

Manage 2009;38:650-662.

10. Bruera E, Valero V, Driver L, et al. Patient-controlled methylphenidate for cancer

fatigue: a double-blind, randomized, placebo-controlled trial. J Clin Oncol 2006;24:2073-

2078.

11. Auret KA, Schug SA, Bremner AP, Bulsara M. A randomized, double-blind, placebo-

controlled trial assessing the impact of dexamphetamine on fatigue in patients with advanced

cancer. J Pain Symptom Manage 2009;37:613-621.

MANUSCRIP

T

ACCEPTED

ACCEPTED MANUSCRIPT

17

12. Peuckmann V, Elsner F, Krumm N, Trottenberg P, Radbruch L. Pharmacological

treatments for fatigue associated with palliative care. Cochrane Database Syst Rev 2010;

11:CD006788.

13. Bruera E, Yennurajalingam S, Palmer JL, et al. Methylphenidate and/or a nursing

telephone intervention for fatigue in patients with advanced cancer: a randomized, placebo-

controlled, phase II trial. J Clin Oncol 2013;31:2421-2427.

14. Escalante CP, Meyers C, Reuben JM, et al. A randomized, double-blind, 2-period,

placebo-controlled crossover trial of a sustained-release methylphenidate in the treatment of

fatigue in cancer patients. Cancer J 2014;20:8-14.

15. Roth AJ, Nelson C, Rosenfeld B, et al. Methylphenidate for fatigue in ambulatory

men with prostate cancer. Cancer 2010;116:5102-5110.

16. Kerr CW, Drake J, Milch RA, et al. Effects of methylphenidate on fatigue and

depression: a randomized, double-blind, placebo-controlled trial. J Pain Symptom Manage.

2012;43:68-77.

17. Shelby-James TM, Hardy J, Agar M, et al. Designing and conducting randomized

controlled trials in palliative care: a summary of discussions from the 2010 clinical research

forum of the Australian Palliative Care Clinical Studies Collaborative. Palliat Med

2012;26:1042-1047.

18. Guyatt GH, Haynes RB, Jaeschke RZ, et al. Users' guides to the medical literature:

XXV. Evidence-based medicine: principles for applying the users' guides to patient care.

JAMA 2000;284:1290-1206.

19. Zucker DR, Schmid CH, McIntosh MW, et al. Combining single patient (N-of-1)

trials to estimate population treatment effects and to evaluate individual patient responses to

treatment. J Clin Epidemiol 1997;50:401-410.

MANUSCRIP

T

ACCEPTED

ACCEPTED MANUSCRIPT

18

20. Nikles J, Mitchell GK, Schluter P, et al. Aggregating single patient (n-of-1) trials in

populations where recruitment and retention was difficult: the case of palliative care. J Clin

Epidemiol 2010;64:471-480.

21. Greenblatt DJ. Elimination half-life of drugs: value and limitations. Annu Rev Med

1985;36:421-427.

22. Novartis. Ritalin hydrochloride product information sheet. East Hanover, NJ:

Novartis, 2010.

23. Abernethy AP, Shelby-James T, Fazekas BS, Woods D, Currow DC. The Australia-

modified Karnofsky Performance Status (AKPS) scale: a revised scale for contemporary

palliative care clinical practice [ISRCTN81117481]. BMC Palliat Care 2005;4:7.

24. Folstein MF, Folstein SE, McHugh PR. "Mini-mental state". A practical method for

grading the cognitive state of patients for the clinician. J Psychiatr Res 1975;12:189-198.

25. Cella D. The Functional Assessement of Cancer Therapy- Anaemia (FACT-An) scale:

a new tool for the assessment of outcomes in cancer anaemia and fatigue. Sem Hematol

1997;34:13-19.

26. Wu H, Wyrwich K, McSweeney M. Assessing fatigue in persons with cancer: further

validation of the Wu Cancer Fatigue Scale. J Pain Sympt Manage 2006;32:255-265.

27. Lloyd-Williams M, Friedman T, Rudd N. Criterion validation of the Edinburgh

postnatal depression scale as a screening tool for depression in patients with advanced

metastatic cancer. J Pain Symptom Manage 2000;20:259-265.

28. Hardy JR, Carmont SA, O'Shea A, et al. Pilot study to determine the optimal dose of

methylphenidate for an n-of-1 trial for fatigue in patients with cancer. J Palliat Med

2010;13:1193-1197.

MANUSCRIP

T

ACCEPTED

ACCEPTED MANUSCRIPT

19

29. Zucker DR, Ruthazer R, Schmid CH. Individual (N-of-1) trials can be combined to

give population comparative treatment effect estimates: methodologic considerations. J Clin

Epdemiol 2010;63:1312-1323.

30. Schluter PJ, Ware RS. Single patient (n-of-1) trials with binary treatment preference.

Stat Med 2005;24:2625-2636.

31. Hardy J, Daly S, McQuade B, et al. A double-blind, randomised, parallel group,

multinational, multicentre study comparing a single oral dose of ondansetron 24 mg with

placebo and metoclopramide 10 mg tds po in the treatment of opioid-induced nausea and

emesis in cancer patients. Support Care Cancer 2002;10:231-236.

32. Abernethy AP, McDonald CF, Frith PA, et al. Effect of palliative oxygen versus room

air in relief of breathlessness in patients with refractory dyspnoea: a double-blind,

randomised controlled trial. Lancet 2010; 376:784-793.

33. Hardy J, Quinn S, Fazekas B, et al. Randomized, double-blind, placebo controlled

study to assess the efficacy and toxicity of subcutaneous ketamnine in the management of

cancer pain. J Clin Oncol 2012 30:3611-3617.

34. Currow DC, Plummer JL, Kutner JS, Samsa GP, Abernethy AP. Analyzing phase III

studies in hospice/palliative care. a solution that sits between intention-to-treat and per

protocol analyses: the palliative-modified ITT analysis. J Pain Symptom Manage

2012;44:595-603.

MANUSCRIP

T

ACCEPTED

ACCEPTED MANUSCRIPTTable 1. Demographic Characteristics of Sample (n=43) Median (min, max)

Age (yrs) 71 (36, 91)

AKPS score at baseline 60 (40, 90)

n (%)

Gender

Female 20 (47.6)

Cancer type

Lower gastrointestinal 11 (25.6)

Prostate 9 (20.9)

Upper gastrointestinal 6 (14.0)

Lung 5 (11.6)

Breast 4 (9.3)

Genitourinary 4 (9.3)

Other 4 (9.3)

MANUSCRIP

T

ACCEPTED

ACCEPTED MANUSCRIPT

Table 2. Characteristics of Responders and Non-Responders to Methylphenidate

(FACIT-F scale)

Non-Responder Responder

n (%) n (%) P-

value

Gender 0.99a

Female 11 (42.3) 4 (50.0)

Male 15 (57.7) 4 (50.0)

Number of completed cycles 0.99a

1 5 (19.2) 2 (25.0)

2 3 (11.5) 1 (12.5)

3 18 (69.2) 5 (62.5)

Median (Q1, Q3) Median (Q1, Q3)

Age (yrs) 70.8 (59.6, 82.3) 72.9 (67.5, 76.7) 0.76b

FACIT-F score at baseline 7 (6, 8) 8 (6.5, 9) 0.24b

AKPS score at baseline 60 (50, 70) 70 (70, 75) 0.07b

aFisher’s exact test. b Wilcoxon’s rank-sum test.

MANUSCRIP

T

ACCEPTED

ACCEPTED MANUSCRIPT Table 3. Clinical Reasons for Dropout / Adverse Events

Event

Day of Withdrawal

from Trial

Grade

Relation of

Event to Trial

Pain increased (hospital admission) 14 4 Unrelated

Fall secondary to general

deterioration

5 4 Unrelated

Breathless 3 2 Possible

Confusion 2 1 Unrelated

Leg swelling 11 1 Possible

Drowsy 1 1 Unlikely

General deterioration 9,9,14 - -

Family distress 1 - -

Adverse event but trial completed

Nausea 7 days after trial

completion

1 Possible

Joint pain Present on day 18 1 Unrelated

Palpitations 3 days after cessation 3 days after cessation - Unrelated

MANUSCRIP

T

ACCEPTED

ACCEPTED MANUSCRIPT

Period 1 Period 2

Cycle 1 Treatment Placebo

Cycle 2 Placebo Treatment

Cycle 3 Treatment Placebo

Figure 1. Typical N of 1 study.

The order of treatment and placebo are randomly assigned for each cycle.

MANUSCRIP

T

ACCEPTED

ACCEPTED MANUSCRIPT

Figure 2: Patient flow chart, showing timing of, and reasons for, withdrawal

Number recruited to trial = 43

Number completed cycle 1 = 33

Number completed cycle 2=29

Number completed cycle 3=24

New symptom developed 3General deterioration 1Non-compliance with trial 2Other, including decision to withdraw 4

New symptom developed 1General deterioration 1Other, including decision to withdraw 2

New symptom developed 1General deterioration 1Other, including decision to withdraw 3

MANUSCRIP

T

ACCEPTED

ACCEPTED MANUSCRIPT

Figure 3a. Mean difference (95% credible intervals) between methylphenidate(MPH) compared to placebo on individual fatigue scores (FACIT-F) for eachpatient (circle) and the overall group (square).Note: the solid black circles designate positive responders, the hollow circlesdesignate non-responders, and the solid gray circle designates a negativeresponder.(Note repeated on each graph but only needs to be stated once only in figure 3)

Figure 3a

MANUSCRIP

T

ACCEPTED

ACCEPTED MANUSCRIPT

Figure 3b. Mean difference (95% credible intervals) between methylphenidate(MPH) compared to placebo on individual Wu Cancer Fatigue Scale (WCFS)scores for each patient (circle) and the overall group (square).Note: the solid black circles designate positive responders, the hollow circlesdesignate non-responders, and the solid gray circle designates a negativeresponder.

(Note repeated on each graph but only needs to be stated once only in figure 3)

Figure 3b

MANUSCRIP

T

ACCEPTED

ACCEPTED MANUSCRIPT

Figure 3c. Mean difference (95% credible intervals) between methylphenidate(MPH) compared to placebo on individual Edinburgh Depression Scale scores foreach patient (circle) and the overall group (square).Note: the solid black circles designate positive responders, the hollow circlesdesignate non-responders, and the solid gray circle designates a negativeresponder.

(Note repeated on each graph but only needs to be stated once only in figure 3)

Figure 3c