Decompressive craniectomy for malignant MCA infarction

Decompressive craniectomy is a dramatic solution for some of the most desperate complications of "Acute ceberebrovascular injury", an L1 topic from Section 2.1.8 in the second edition of the CICM Syllabus for the Second Part Examination. It had come up in Question 2 from the second paper of 2014 and then again in Question 24 from the first paper of 2025.  The first time, apart from neurosurgical options the College wanted their candidates to discuss blood pressure management, therapeutic hypothermia, likely prognosis and quality of life issues. Judging from the model answer, some detailed knowledge of the published trial evidence was expected. Ten years later, the SAQ writers wanted their trainees to "discuss" the role of this intervention (a glossary term that suggests some level of analysis was expected), as well as "in your answer, include the evidence for this practice". Though the practice is not listed individually in the syllabus, these are clear signals that the examiners want to put emphasis on this condition.

A section of Oh's manual is dedicated to this exact issue. Chapter 51 ("Acute  cerebrovascular  complications") by Bernard  Riley  and  Thearina  de  Beer contains within it a few paragraphs on decompressive craniectomy, on page 571. It would be worth pointing out that the key issues raised in these paragraphs were also raised in the model answer, and one would be well advised to review this material in order to produce an answer which appeases the examiners. For all we know, Riley or de Beer wrote that SAQ.

Apart from the brief entry in Oh's, several excellent resources exist for this topic:

• A 2012 article about conservative management, from Current opinion in critical care
• A 2015 meta-analysis of decompressive craniectomy from The Surgeon
• A summary of guidelines for management of MMCAS from Neurocritical care (not free)
• A summary of the 2014 AHA Guidelines, specific for stroke with mass effect.

If one were to discuss or critically evaluate the use of decompressive craniectomy in stroke, one would go about it in a systematic fashion:

Definition and epidemiology of malignant MCA infarction

"Malignant" anything sounds like a pretty serious problem, but after reviewing uncountable multitudes of definitions one comes to the conclusion that "malignant" usually means something more than just "really, really bad". There really does not appear to be an "official" definition, but  Treadwell & Thanvi (2010) pin it down to a specific timeframe:

"neurological deterioration ... rapidly within 24 h of symptom onset, and when associated with involvement of the whole MCA territory has been termed ‘malignant MCA infarction"

Another good method for finding an official-sounding definition is to see how clinical trials of the condition had selected patients for enrolment. 

• DESTINY asked for a NIHSS score >18-20 and an infarction of 2/3 of the MCA territory
• HAMLET demanded NIHHS of  ≥ 16 for rightsided and  ≥ 21 for leftsided lesions, an dalso 2/3rds of the MCA territory infracted, but also called for a  drop in GCS
• DECIMAL was happy with a NIHSS of merely >16, and only 50% of the territory

In short, its a MCA stroke large enough to produce enough cerebral oedema to damage other non-infarcted brain regions by the local pressure effect and midline shift. These are rare (apparently 1-2/100,000) but carry an extremely poor prognosis (~80% mortality when managed conservatively), which is the reason for the intrusive surgical option.

Rationale

• MCA infarction has a substantial mortality and morbidity.
• Factors which produce this effect include:
  • Large volume of infarcted brain tissue, leading to significantly increased ICP.
  • Significant risk for haemorrhagic transformation.
  • Significant midline shift, with pressure on medial cerebral structures .
  • Temporal herniation, with compression of the posterior cerebral artery.
  • Poor perfusion of the contralateral cerebral hemisphere due to raised intracranial pressure.
• Many of these factors are related to the increased intracranial pressure.
• Decompressive craniectomy can decrease intracranial pressure by increasing cranial compliance.
• Ergo, decompressive craniectomy should be able to improve mortality and morbidity from acute MMCAS by vastly improving the perfusion in the penumbra of the stroke shortly after the craniectomy.

Advantages and disadvantages

• Advantages:
  • A mortality improvement is consistently seen in the literature
  • For younger patients, an improved functional outcome is also possible
  • The association between raised ICP and mortality seems to be genuine, eg according to Jung et al (2024) 
  • The procedure appears to be safe even following thrombolysis and while on antiplatelet agents
  • The idea has a strong theoretical foundation (by saving more brain tissue, one should give the chance for a better recovery)
• Disadvantages:
  • It is a very intrusive thing to do to a person who is  often fundamentally incapable of giving consent for it. Gatos et al (2024) described it as an "aggressive amputative procedure".
  • Apart from disabled survivorship, the procedure itself has a lot of complications, not the least of which include seizures, hygroma, local wound and CNS infections and bone graft resorption. 
  • The longer ICU stay usually prevents these patients from accessing early stroke rehab
  • Liu et al (2025) bemoaned the lack of clear and consistent diagnostic criteria

Controversies and risks

• Criteria for being offered craniectomy are unclear. For example Shen et al (2024) found that a poor GCS prior to decompression was a strong risk factor for poor outcome after decompression, and yet some of the protocols insist on a drop in GCS to qualify for decompression.
• We cannot agree on the timing. It is not clear whether this is something that needs to be performed immediately (eg. within 48 hrs, as in all the trials) or whether progression of the stroke and deteriorating neurology should be the trigger for action.
• There is no data from developing nations, i.e it is not clear whether this procedure benefits anybody when there is no strong ICU/rehab safety net
• The effect of age on functional recovery is unclear - we suspect that age decreases the likelihood of a good outcome, but nobody knows to what extent.

Evidence

The college model answer lists three landmark studies worth referencing, which form the three famous European hemicraniectomy trials. They were HAMLET, DESTINY and DECIMAL.

DESTINY trial (2007):

• Prospective, multicenter RCT in Germany
• 32 patients were enrolled; then a statistically significant mortality reduction was found, and the study was terminated. The projected sample size was calculated to 188 patients, but the steering committee decided to terminate this trial anyway because of the results of the three other European decompressive craniectomy trials.
• Raw data suggests improved survival for the craniectomy group: 88% vs 47%.

DECIMAL trial (2007):

• Prospective, multicenter RCT in France
• After randomization of 38 patients, the data safety monitoring committee recommended stopping the trial because of slow recruitment.
• Absolute reduction of 52.8% in the death rate in the surgery group.
• Survival was 22% in the "conservative management" group, and ~ 75% in the surgery group

HAMLET trial (2009):

• Prospective, multicenter RCT in the Netherlands
• 64 patients were randomised
• Again, survival was better with surgery (absolute risk reduction was 38%)

Pooled analysis of the European studies

A pooled analysis of the first three studies, including 93 patient cases, came to a fairly positive conclusion:

"...after decompressive surgery the probability of survival increases from 28% to nearly 80% and the probability of survival with an mRS of ≤3 doubles."

(mRS here being the score of the modified Rankin scale, equating to a disability where one requires some help, some of the time, with some things - but is otherwise able to walk unassisted).

In addition to the above studies, the college mentions that people in the over-60s age group are also being investigated as potential candidates for decompressive craniectomy.

They were probably referring to the DESTINY II Trial (2014):

• 112 patients, older than 60 years (median age was 70)
• Primary outcome measure was survival without severe disability; this was improved: 38% in the hemicraniectomy group, as compared with 18% in the control group.
• Survival in general also had lower mortality in the surgery group (33% vs. 70%).
• However, in contrast to the young patients, practically none of the survivors has an outcome as good as an mRS score of 3. The vast majority of the post-operative survivors were severely disabled.

Protocol

Progression of an MCA infarct to a "malignant" MCA infarct:

As mentioned above, these are the findings from the three abovementioned European trials.

• MCA territory stroke of >50% or >66% on CT
• Perfusion deficit of >66% on CT
• Infarct volume >82 mL within 6 hours of onset (on MRI)
• Infarct volume of >145mL within 14 hours of onset (on MRI)

Patient selection

• Age <60 years.
• Within 48 hours of symptom onset.
• It seems the benefit of craniectomy was lost after 96 hours; presumably all the salvageable penumbra has died, and mass effect is maximal.

Procedural issues

• Craniectomy has to be large enough to extend past the margins of the infarct.
• Well tolerated even after thrombolysis( though apparently antiplatelet drugs tend to increase the risk of bleeding).
• There is no difference in outcome whether dominant or non-dominant hemispheres are involved.
• If there is a haemorrhagic transformation, a craniectomy and evacuation of clot may be required even if the patient did not meet MMCAS criteria as above. This is particularly helpful for posterior fossa pathology, where space is anatomically limited.

Intracranial pressure monitoring

• Does not work. ICP monitoring has not been proven to change outcome.
• EVD insertion may still be required for drainage of secondary hydrocephalus resulting from haemorrhagic stroke

Why might this be the case, one might ask? Generally, everybody attributes the morbidity to the rise in ICP and midline shift. pressure must clearly play a role: in fact compression is clearly held accountable for the poor outcomes, to the extent that de-compression is recommended as the rescue therapy. So why no EVD or ICP-guided management, if the situation is so similar to TBI? 

Yun & Ding (2020) asked the same question, pointing to the fact that the decision not to offer ICP monitoring to these people is based on some very tenuous data. For example, everyone seems to refer to Frank (1995), but that's a paper from a single neurosurgeon describing 19 patients, who had mixed ICP values in the first 12 hrs, and in whom the seubsequent deterioration did not seem strongly associate with ICP. Poca et al (2010) is another cohort of 19, in whom Codman's monitors failed to detect an ICP elevation that nonetheless must have been present, because twelve of them had uncal herniation at the time. On the other hand, Jung et al (2024) and Alhamdan et al (2025) found ICP elevations were related to some of the worst outcomes, with a CPP inflection point of around 80 mmHg (below this, they seem to do poorly). In short, ICP-guided management for these patients may have a real benefit, and we simply have not seen this because they are fewer in number than TBI patients, and none of the work so far has met the criteria for statistical reliability.

Outcomes

• NNT for survival is 2
• NNT for severe disability is 6.
  • Malignant MCA infarct has a mortality of 70%
  • Craniectomy reduces this to 30%, but with residual deficit.