Is it time to abandon methylphenidate?

Journal of Clinical Oncology has published a large (for palliative care), randomized, placebo-controlled trial of long-acting methylphenidate for cancer-related fatigue.  It continues the trend of mixed-to-negative findings (in controlled studies) of methylphenidate for fatigue (see the last big-ish RCT here; a negative RCT here for 'prophylactic' methylphenidate; a brow-furrowing meta-analysis of the whole topic here;  some grey literature on the effectiveness of modafinil for severe fatigue in cancer patients undergoing chemotherapy), and raises the question is it time to abandon methylphenidate?

This was a randomized trial involving ~150 patients (mean age 60 years, 65% currently on chemotherapy, 70% with stage III/IV cancer) with cancer related fatigue (4+/10 on a 0-10 fatigue scale).  The patients had to have an ECOG performance score of 2 or less and a life expectancy of more than 6 months (they don't say how they defined that).  Patients were randomized to either long-acting methylphenidate (Concerta) 54mg a day (patients worked up to after 2 weeks, starting at 18mg a day) or matching placebo.  They were kept on 54mg/day for 2 weeks.  

In the main analysis (which included looking at patient self related fatigue, and several health related quality of life indicators), methylphenidate did nothing compared to placebo.  Both arms improved modestly during the trial, and equally so...there is not even anything you could call a 'hopeful' trend in favor of methylphenidate. 

However, adverse events were markedly worse with methylphenidate, particularly with nervousness and appetite loss. 

They did a few secondary analyses of their data, which temper the 'negative' findings to an extent.  When looking only at patients with stage III/IV cancer, fatigue improved ~20 points (on a 0-100 scale) with methylphenidate compared to 2/100 with placebo.   The N for this secondary analysis is about 100 patients.   They make the curious comment that patients with early stage disease actually had less fatigue improvement with methylphenidate than with placebo.  In addition, they note those with the worst fatigue (8+/10) at baseline improved 26/100 with methylphenidate vs 16/100 with placebo (not statistically significant).

To summarize: in an undifferentiated group of cancer patients with stages I-IV of cancer, most of whom are on chemotherapy, and with both moderate to severe fatigue, long-acting methylphenidate did not seem to do anything other than cause nervousness.  However:  for patients with advanced-stage disease, and patients with severe fatigue, methylphenidate seems to be helping.  I'm not entirely sure how it ends up that when they do a secondary analysis which nonetheless involves over 2/3 of their patients they find a modest-but-significant treatment effect for methylphenidate unless their suggestion that methylphenidate takers with early stage disease were worse off (NS) than placebo takers...ie methylphenidate actually worsened fatigue (or, more accurately, improved fatigue less than placebo) in the early stage patients. 

All this is why my call to 'abandon' methylphenidate up top was a bit of bluster.  It's no wonder-drug either.   Some observations however about the last decade of psychostimulant research in cancer, particularly comparing this to the last big controlled study on methylphenidate by Bruera & colleagues. 

• Bruera's study was of immediate acting methylphenidate, and generally involved much lower doses than this one (~20mg daily vs 50).
• That is why, perhaps, side effects were measurably higher in this study than in Bruera's - it does not appear than in the 20-30mg/day range methylphenidate causes significant side effects (vs placebo) in an advanced cancer population.
• One can imagine that there are differences in LA and immediate release methylphenidate in tolerability as well.  Speaking from my *completely* idiosyncratic and anecdotal experience with short acting and long acting opioids, patients can sometimes react in surprising ways to long-acting opioids (e.g. not tolerate 60mg bid of oxycodone ER when they can easily tolerate 30mg q3h of oxycodone IR).  
• This study, similar to a modafinil study, suggests that the only patients who, as a group, measurably benefit from psychostimulants, are those who report severe fatigue at baseline.  An accompanying editorial suggests there is some (yet to be published outside of abstract) additional data to support that observation.  
• This leads me to conclude what would be most helpful would be a study restricted to advanced cancer patients with severe fatigue, and no one else.  Or at least a meta-analysis of prior studies restricted to that population.  
• I think though we can begin to conclude that, for instance, me the 'healthy' cancer patient with, say, early stage breast cancer receiving adjuvant chemotherapy with moderate fatigue is not a good candidate for psychostimulant therapy.  Exercise, education, and emotional support yes; methylphenidate no.  
• My question is, what if the RCT I'm hoping for above is negative (for methylphenidate)?  Will you stop prescribing it?