Does Improvement in Symptoms of Attention Deficit Hyperactivity Disorder (ADHD) Mediate an Effect of Long-Acting OROS-Methylphenidate on Cigarette Smoking? A Secondary Analysis of CTN-0029

1
Does Improvement in Symptoms of Attention Deficit
Hyperactivity Disorder (ADHD) Mediate an Effect
of Long-Acting OROS-Methylphenidate on Cigarette
Smoking?A Secondary Analysis of CTN-0029

• Edward V Nunes1, Lirio Covey1, Mei-Chen Hu1,
  Martina Pavlicova1, Eugene Somoza2, Theresa
  Winhusen,2
• 1Columbia University Medical Center (CTN Long
  Island Node),
• 2University of Cincinnati (CTN Ohio Valley
  Node),
• Funded by U10 DA13035 (NIDA-CTN) and K24 DA022412
  (Nunes)

2
ABSTRACT CTN-0029 was a multisite,
randomized, double-blind, placebo-controlled
trial of long-acting OROS-methylphenidate
(OROS-MPH) for treatment of patients with both
nicotine dependence and attention deficit
hyperactivity disorder (ADHD). The principal
outcome analysis found, as expected, a robust
beneficial effect of OROS-MPH in improving
symptoms of ADHD, but no clear effect of OROS-MPH
on cigarette smoking outcomes. Since OROS-MPH is
primarily a treatment for ADHD, it is reasonable
to hypothesize that a beneficial effect on
smoking outcome might occur only among those
patients who experience a substantial improvement
in their ADHD during treatment. We therefore fit
a linear model in which prolonged abstinence at
week 10 (end of the acute treatment phase) was
modeled as a function of medication treatment
(OROS-MPH versus Placebo), the change in the ADHD
symptom severity score between baseline and end
of study, and the interaction between ADHD
improvement and treatment. The interaction was
significant, suggesting that among those patients
with the greatest improvement in ADHD symptoms
during treatment, OROS-MPH was superior to
Placebo in promoting prolonged abstinence from
nicotine. This secondary analysis suggests that
OROS-MPH may be effective in promoting smoking
cessation among the subset of patients whose ADHD
responds well to methylphenidate treatment.
3

• Specific Aim
• To explore, among smokers with Attention Deficit
  Hyperactivity Disorder (ADHD), whether a
  beneficial effect of OROS-Methylphenidate
  (OROS-MPH) on smoking cessation is mediated by
  improvement in symptoms of ADHD
• Rationale
• Stimulant treatment, such as with
  OROS-Methylphenidate (OROS-MPH), is effective for
  symptoms of ADHD.
• A multisite, placebo-controlled CTN trial of
  OROS-MPH among ADHD adults with cigarette smoking
  found the expected effect of OROS-MPH to improve
  ADHD symptoms, but no main effect of OROS-MPH on
  cigarette smoking
• However, a beneficial effect of stimulant
  treatment on smoking outcome may occur only among
  those patients whose ADHD symptoms improve
• This would be consistent with the self-medication
  hypothesis
• If among patients with ADHD, nicotine is used
  for its stimulant properties to improve attention
• Then stimulant treatment should increase
  nicotine abstinence only among those patients
  whose ADHD gets much better in response to the
  stimulant

4

• Background
• ADHD is a neuropsychiatric condition that begins
  in childhood and, in many cases, persists to
  adolescence and adulthood.
• In the U.S., ADHD is estimated to affect 2 to
  18 of children and adolescents, and about 4.4
  of adults.
• ADHD has been linked with nicotine dependence.
• Persons with ADHD are more likely to become
  regular smokers, begin smoking earlier, smoke
  more heavily, and experience greater difficulty
  to stop smoking, compared to persons without
  ADHD.
• Evidence that nicotine ameliorates
  inattentiveness and performance deficits and that
  nicotine can reduce deficits in dopaminergic
  function related to inattentiveness suggests a
  self-medication rationale for greater tobacco
  use among persons with ADHD.

5

• Methods
• A randomized, double-blind, placebo-controlled
  trial comparing OROS-MPH vs. placebo as
  adjunctive smoking cessation treatment (nicotine
  patch and counseling).
• Whites 202 Non-whites 51 (African-American1
  5, Hispanic 16, Other20).
• Males56.5, mean age37.8 years (s.d. 10.0),
  mean years school 14.4 years (s.d. 2.4).
• Six study sites, located in Cambridge
  Massachusetts, Columbus Ohio, New York City New
  York (2 sites), Portland Oregon, and Rochester
  Minnesota, recruited participants. Study approved
  by Institutional Review Board at each study site.
• The trial included an 11- week treatment phase
  (OROS-MPH vs. placebo) consisting of a four-week
  pre-quit phase and seven weeks of a planned
  abstinence period. All participants received
  brief counseling(11 weeks) and nicotine patch
  (during Weeks 4-11).
• Conducted in CTN between December 2005 and
  January 2008

6

• Outcome and Data Analysis
• Abstinence outcome Prolonged abstinence for four
  weeks at end of medication treatment (weeks 7 to
  10)
• Standard categorical measure of treatment success
  recommended by Society for Research on Nicotine
  and Tobacco
• Mediator Change from baseline to end of study in
  the ADHD symptom severity score (ADHD Rating
  Scale (DuPaul, Power, et al., 1998))
• A continuous score, where higher change scores
  indicate greater improvement in ADHD during
  treatment.
• Analysis A mixed effect model was fit, modeling
  prolonged abstinence as a function of
• Treatment (OROS-MPH vs Placebo)
• Mediator (change in ADHD symptom severity during
  treatment)
• Cigarettes per day at baseline as covariate
• Graph of the model Although the mediator was a
  continuous covariate, for the purposes of
  graphical presentation the score was divided into
  four quartiles, from low to high change.

7
Table 1 Baseline characteristics
8
Model A Effect of Medication Treatment without
Accounting for Improvement in ADHD
Term in the model D.F Chi-Square P Comment
Site 5 18.79 .003
Treatment 1 0.05 .824 No Treatment Effect
Cigarettes per day at baseline 1 6.66 .010 Abstinence is less likely for heavier nicotine use
9
Model B Effect of Medication Treatment
Improvement in ADHD included in the Model
Term in the Model D.F. Chi-Square P Comment
Site 5 11.04 .051
Treatment 1 3.53 .061
Cigarettes per day at baseline 1 6.06 .014 Abstinence is less likely for heavier nicotine use
Change in ADHD 1 6.26 .013 More improvement in ADHD associated with more abstinence
Interaction Treatment by ADHD change 1 6.83 .009 Effect of treatment on abstinence differs by level of ADHD change
10
Percent of patients achieving prolonged
abstinence, as a function of medication
treatment, and improvement in ADHD during
treatment --When ADHD improvement is high,
medication is superior to placebo in promoting
abstinence --When ADHD improvement is lower,
medication trends to being worse than placebo in
abstinence
11
Summary of Findings

• There was a significant interaction of treatment
  with change in ADHD score during treatment
• This suggests that ADHD change during treatment
  mediates the impact of OROS-MPH on outcome
• When OROS-MPH produces a large improvement in
  ADHD symptoms, it improves smoking outcome
  compared to placebo
• When OROS-MPH produces little improvement in ADHD
  symptoms, the trend is that smoking may get worse

12
Implications

• Stimulant treatment, such as with OROS-MPH, may
  be an effective intervention for smoking
  cessation among smokers with ADHD--if the
  stimulant produces a large improvement in ADHD
  symptoms.
• This is consistent with the self-medication
  hypothesis, which would predict that unless ADHD
  symptoms get very much better, patients will
  continue to smoke to get the attention-improving
  effects of the nicotine

13
Clinical Implications

• ADHD responds quickly to stimulant
  treatment--within a few weeks, depending on how
  long it takes to adjust the dose
• Several different stimulants, and stimulant
  formulations, as well as non-stimulant
  medications may be effective for ADHD, and
  individuals differ in their response
• Therefore, an effective strategy for cigarette
  smokers with ADHD may be to treat the ADHD
  aggressively, and change medications if needed
  until a robust improvement in ADHD is achieved
• More research is needed to test algorithmic
  approaches such as this for the treatment of
  patients with ADHD and nicotine dependence, and
  perhaps for patients with co-occurring
  psychiatric disorders and addictions in general

14
Implications for Research on Treatment of
Co-Occurring Disorders

• Addictions are associated with increased rates of
  many psychiatric disorders, and the co-occurrence
  carries a poor prognosis
• Depression, bipolar disorder, PTSD, other anxiety
  disorders, ADHD
• Many research studies have sought to improve
  addiction treatment outcome by treating
  co-occurring disorders
• These studies may be inherently underpowered if
  the effect of treatment on addiction outcome
  resides only in the subgroup of patients whose
  co-occurring disorder responds well.
• For most co-occurring psychiatric disorders
  different patients respond to different
  treatments, and only a minority of patients will
  respond to any one treatment.
• Clinically, it may be that treatment of the
  co-occurring disorder needs to be adjusted
  aggressively until a good response is achieved
• In designing studies of treatment of co-occuring
  psychiatric and addictive disorders, we should
  consider studying algorithms that seek to
  maximize response of the co-occurring disorder,
  by adjusting doses and changing treatments,
  rather than testing single treatment strategies.