Interpreting the CPASS Trial: Do Not Shift Motor Therapy to the Subacute Phase

Abstract

The Critical Periods After Stroke Study (CPASS, n=72) showed that, compared to controls, an additional 20 hours of intensive upper limb therapy led to variable gains on the Action Research Arm Test depending on when therapy was started post-stroke: the subacute group (2–3 months) improved beyond the minimal clinically important difference and the acute group (0–1 month) showed smaller but statistically significant improvement, but the chronic group (6–9 months) did not demonstrate improvement that reached significance. Some have misinterpreted CPASS results to indicate that all inpatient motor therapy should be shifted to outpatient therapy delivered 2–3 months post-stroke. Instead, however, CPASS argues for a large dose of motor therapy delivered continuously and cumulatively during the acute and subacute phases. When interpreting trials like CPASS, one must consider the substantial dose of early usual customary care (UCC) motor therapy that all participants received. CPASS participants averaged 27.9 hours of UCC OT during the first 2 months and 9.8 hours of UCC OT during the 3^rd and 4^th months post-stroke. Any recovery experienced would therefore result not just from CPASS intensive motor therapy but the combined effects of experimental therapy plus UCC. Statistical limitations also did not allow direct comparisons of the acute and subacute group outcomes in CPASS. Instead of shifting inpatient therapy hours to the subacute phase, CPASS argues for preserving inpatient UCC. We also recommend conducting multi-site dosing trials to determine whether additional intensive motor therapy delivered in the first 2–3 months following inpatient rehabilitation can further improve outcomes.

The phase IIb Critical Periods After Stroke Study (CPASS) randomized 72 participants with mild to moderate upper limb impairment to receive an additional 20 hours of intensive upper limb therapy, on top of their usual and customary care motor rehabilitation (UCC), delivered starting in the acute (0–1 month), subacute (2–3 months), or chronic (6–9 months) phases post-stroke¹. CPASS was designed to closely mirror rodent studies showing enhanced efficacy of forelimb training started within one-month post-stroke combined with environmental enrichment^2,3. In CPASS the 20 hours of participant-directed intensive upper limb therapy was, similarly, not delivered in isolation, but in addition to UCC delivered as standard of care. The primary outcome measure was the Action Research Arm Test (ARAT) assessed over the first 12 months post-stroke. CPASS showed that, compared to controls receiving only UCC, the subacute group showed recovery beyond the minimal clinically important difference⁴ (ARAT difference +6.78 ± 2.63, P = 0.009), the acute group showed significant recovery (ARAT difference +5.25 ± 2.59, P=0.043), and the chronic group did not show significantly greater recovery than the control group (ARAT difference +2.41 ± 2.25, P = 0.29). Thus, CPASS was the first study to provide evidence in humans for a critical period early after stroke when patients may be more responsive to rehabilitation therapies.

Given this first signal for post-stroke critical period plasticity in humans, the ensuing excitement has led to some misinterpretation of our findings that requires further clarification. Most troubling is the suggestion that early inpatient motor rehabilitation should be shifted to the subacute phase. Reporters writing about CPASS concluded that “the best time for rehabilitation after stroke may actually be 2–3 months later”^5,6. We want to clarify that the CPASS results do not suggest shifting inpatient motor rehabilitation to the subacute phase: crucially, since the participants in the three treatment groups received both UCC and additional CPASS intensive motor therapy at different timepoints post-stroke, one must consider the details of this combination as part of the CPASS results. Instead, we propose that CPASS revealed that upper limb motor therapy should be increased and enhanced in the acute and subacute phases post-stroke as compared to current standard systems of care, and also calls for future dosing trials to determine the optimal amount of therapy that should be provided during these time windows.

Foremost among the reasons to reject shifting UCC to the subacute phase is that all participants in the CPASS trial received a substantial amount of UCC during early rehabilitation. To provide details about this UCC, we determined precisely how much occupational therapy (OT, which tends to focus on the upper limb) participants received as part of UCC over the 12-months of their post-stroke rehabilitation. Participants received an average of 27.9 hours of UCC (combined inpatient and outpatient OT) during the first 2 months post-stroke (figure 1) and an average of 9.8 hours of UCC (outpatient OT only) during months 3 and 4 post-stroke. Thus, most of the therapy (UCC and CPASS intensive motor therapy) for the acute group was clustered early in the rehabilitation process, whereas for the subacute group the total dose of therapy was spread more evenly over time. Our results show that outcomes over 12 months must be considered not just with regard to the timing of intensive CPASS motor training, but also in combination with UCC, which was most concentrated early after stroke.

Figure 1.

Average hours of CPASS intensive upper limb therapy per participant in each respective study group (A), and average hours of inpatient and outpatient usual customary care (UCC) occupational therapy for all participants (B). Note the thin black band in month 1 (denoted by black arrow) representing 5 total hours of outpatient UCC therapy delivered among all the participants.

In addition, there are important differences in the type of therapy provided during the CPASS intensive experimental therapy as compared with UCC that may have had a crucial impact on the study results. UCC focuses on performing activities important for transitioning to the next level of care (e.g. self-care tasks and functional transfers), whereas CPASS intensive motor therapy focused on upper limb activities selected by and deemed most important to the participant for their recovery (e.g. cooking dinner, driving, or reading the Bible). UCC sessions are also known to include relatively little task-oriented training⁷, whereas the CPASS intensive motor therapy focused solely on task-oriented training. Psychological studies suggest that giving participants more autonomy and focusing on the end goal instead of the component parts leads to enhanced motor learning⁸. The CPASS study incorporated the basic principles of motor learning in that successful completion of tasks results in positive reinforcement and willingness to take on more challenging tasks. The progression of task demands from simple to more complex is designed to facilitate the successful completion of the desired activities. This process strengthens the coupling of goals to actions.

Finally, there are statistical considerations that suggest we should not shift all UCC to the subacute phase. CPASS was designed to test whether there were any differences between the groups over 12 months. The null hypothesis was rejected when the acute and subacute groups each demonstrated significant improvement over control in the repeated measures model. However, the CPASS study was not designed to determine whether the improvement for the subacute group was significantly greater than that for the acute group. Further research is necessary to determine the critical period windows with greater resolution in humans.

If shifting inpatient rehabilitation hours to the subacute phase is not recommended, then what are the clinical implications of CPASS? Given that CPASS was a phase II study, the results should not change current rehabilitation practice until further studies are conducted. However, the CPASS results do hint at many exciting future directions for research. CPASS should be replicated in a phase III trial to more definitively establish the existence of a post-stroke critical period in humans and clarify how long this window lasts. In addition, investigators should perform trials to establish the optimal dose of upper limb therapy to deliver in the acute and subacute phases. Assuming the findings are replicated in trials of larger sample size, CPASS may have uncovered a way to enhance our current systems of post-stroke care – namely, that we should increase and perhaps alter the nature and amount of upper limb therapy delivered during the first 3 months. Indeed, over the last 2 decades the average inpatient rehabilitation stay in the U.S. declined from 18 to 12 days^9,10, and as we demonstrate here, the dose of UCC drops off dramatically in the outpatient setting. We suspect that stroke patients who are either discharged directly home or to skilled nursing facilities following their acute hospitalization would also benefit most from upper limb therapy delivered in the first 3 months, but this requires further study.

In conclusion, the CPASS trial shows the important benefits of intensive upper limb motor therapy continuously and cumulatively provided during the acute and subacute phases; but there are many reasons why the comparison between these two early phases is less clear and further research is necessary. CPASS was designed to closely emulate rodent studies that combined forelimb training and environmental enrichment; a later study showed that this combination was more effective than either intervention alone¹¹. Rather than shifting inpatient therapy to the subacute phase, we recommend preserving inpatient UCC and conducting future trials to determine how combining this with additional intensive motor therapy delivered in the first 3 months following inpatient rehabilitation can further improve outcomes.