In 2011 Cooper et al published the DECRA trial examining the use of decompressive craniectomy in patients with traumatic brain injuries (TBI) and elevated intracranial hypertension refractory to medical management (1). Shockingly what was until then thought of as a lifesaving procedure was found to be inferior to conservative medical management. Many have discredited the results reported by the DECRA authors stating a number of issues, the most glaring being the criteria for the decompressive craniectomy was an ICP greater than 20 for only 15 minutes. RESCUEicp was intended as the solution to these methodological flaws. And while its results certainly confirm the criteria utilized for craniectomy in the DECRA cohort may have been heavy-handed, whether RESCUEicp’s results support the use of a more deliberate strategy is unclear.
Published in the NEJM by Hutchinson et al, over a 10-year period, the authors of RESCUEicp randomized 408 patients with TBI to either medical management or decompressive craniectomy with the intention of controlling ICP (2). Patients ranged from 10-65 years years of age with intracranial hypertension refractory to medical management. Only after failing tier 1 and tier 2 medical therapies, were patients randomized to either decompressive craniectomy or continued medical management with the addition of a barbiturate coma.
Like DECRA, the patients in RESCUEicp randomized to decompressive craniectomy, underwent the procedure shortly after randomization (2.3 hours in DECRA vs 2.2 hours in the RESCUEicp). Delayed decompressive craniectomy was performed in 37.2% of patients in the RESCUEicp cohort and 18% of the patients in the DECRA trial. Unlike DECRA, which found similar mortality between the two groups (19% and 18% respectively), the authors of RESCUEicp reported a significant decrease in 6-month mortality in patients who underwent decompressive craniectomy when compared to medical management alone (26.9% and 48.9% respectively). They also reported an 8.2% absolute increase in the number of patients with a favorable neurological outcome at 6 months (defined using the extended Glasgow Outcome Scale (GOS-E)). And while on first glance RESCUEicp appears to validate the complaints of DECRA’s detractors, a closer inspection suggests that success is all in how you define it.
Despite some statistical complexities in the manner the authors attempted to describe their findings, there is an obvious difference in the distribution of functional outcomes between the two groups. A brief glance at fig. 2 illustrates these differences. The shift in neurological outcomes was primarily observed in patients moving from death to severe disability. In these instances the portion of patients in a vegetative state at 6-months was 8.5% vs 2.1% in the decompressive craniectomy and medical management groups respectively. 21.9% vs 14.4% of patients had lower severe disability and 15.4% vs 8.0% had upper severe disability. 42.8% were alive with a favorable neurological outcome at 6-months in the decompressive craniectomy vs 34.6% in the medical management group (p=0.12). The definition of “favorable” neurological outcome utilized by the authors of RESCUEicp was somewhat atypical. The authors included patients with upper severe disability (functionally independent at home, but requires assistance in public) at 6-months as a favorable outcome. While the dichotomous cutoff that traditionally delineates a favorable or unfavorable outcome separates moderate disability and good recovery from upper severe disability. In fact, if the same dichotomous cutoff used in the DECRA cohort was utilized in RESCUEicp, then 27.4% and 27.0% of the respective groups would have been considered to have had a favorable neurological outcome.
There is certainly a strong signal of harm. DECRA randomized 155 patients with TBI and refractory intracranial hypertension to either decompressive craniectomy or medical management. Specifically, the authors examined the clinical benefit of early decompressive craniectomy for the management of refractory hypertension. As such, their inclusion criteria only required patients to experience an ICP > 20 mm Hg for 15 minutes to be eligible for randomization. This is in contrast to RESCUEicp, which required patients have an ICP > 25 mm Hg for greater than 1-hour. As such the DECRA cohort was likely to enroll some patients who would have responded to more aggressive non-invasive measures and may have not required a decompressive craniectomy for ICP control. It is likely this early and aggressive use of decompressive craniectomy was the reason for their poor outcome when compared to medical management. Like the RESCUEicp cohort, DECRA demonstrated that patients randomized to decompressive craniectomy experienced worse neurological outcomes, only in this case it was at the price of neurologically intact survivors. These findings suggest that early aggressive decompressive craniectomy is in fact likely to harm a subset of patients.
This is not to say a signal of clinically important benefit is entirely absent. While there was a significant increase in the number of patients left in a vegetative state in the decompressive craniectomy group, there was also a significant increase in the amount of patients found to have upper severe disability. This obviously becomes an important issue of quality of life for each patient and their family to address individually.
It is an unfortunate and messy truth that reality is never as black and white as we would like. This study reminds us, as with most therapies in medicine, there is a J shaped curve that delineates harm from benefit. DECRA certainly illustrates the harms of such interventions when performed on the wrong patient population. RESCUEicp offers a small signal of benefit in an otherwise negative trial. The exact patients that will benefit from decompressive craniectomy is still unclear. Perhaps a tool as blunt and clumsy as ICP is not appropriate to demarcate such a fine line. RESCUEicp was a tremendous effort of quality and insight, but I fear its results have not made the bedside decision to perform a decompressive craniectomy any clearer.
Sources Cited:
- Cooper DJ, Rosenfeld JV, Murray L, et al. Decompressive craniectomy in diffuse traumatic brain injury. N Engl J Med 2011; 364:1493-502
- Hutchinson PJ, Kolias AG, Timofeev IS, et al. Trial of decompressive craniectomy for traumatic intracranial hypertension. N Engl J Med.
- EM Nerd-The Case of the Partial Cohort - May 24, 2020
- EM Nerd: The Case of the Sour Remedy Continues - January 20, 2020
- EM Nerd-The Case of the Adjacent Contradictions - December 23, 2019
pediatric neurosurgeon – very well written commentary on a paper that looks to confirm the preconceived notions of many experts – if you don’t like DC, then your take on it is music. If you like DC, then you might cherry pick the cut-point of “success” (like including upper severe disability). If you are not sure and think there might be a patient population that will benefit, but we need to refine our ability to find them – this will check that box perfectly. Thanks for the succinct summary of the paper.
Thank you! I appreciate the support.
You sir are a spectacular writer, thank you for your commentary
Thank you! you are far too kind.