The University of Chicago Journal Club, August 2010
Editor, B Roitberg, MD
This is the first publication of the “University of Chicago Journal Club” series. The format is new, and some explanation is in order.
- “Journal club” is a popular educational tool, familiar to many readers in some form. Residents, students and faculty meet to discuss one or more articles, trying to put the paper in perspective of previously published literature, practice, and personal experience. This is also an opportunity to critique study design, understand strengths and limitations of the various articles. Many formats can be used for a journal club, but most often a student or resident would review an article and discuss it, followed by comments and questions by faculty and other residents or students. For me, a journal club always felt like a much better way to learn that by just reading a few articles on my own. Even just listening to a journal club was illuminating. The goal of the new “journal club” series at SNI is to share our journal club at the University of Chicago Neurosurgery with the readers. The format is a work in progress, and we welcome comments and feedback.
This particular journal club was designed to review several related articles, discussing the current understanding of the role of decompressive surgery in the management of patients with malignant infarction of the middle cerebral artery. Each resident was assigned to review the articles in the context of the others, with the option of adding further insight based on additional literature review. After the presentations, participating faculty and other residents had the opportunity to add comments. These comments were recorded and are presented here. Although the comments were originally given as part of a lively discussion, here they are summarized for the sake of readability.
Topic of Journal Club: ‘Early decompressive surgery in malignant infarction of the middle cerebral artery’
1) Vahedi et al, on behalf of the DECIMAL investigators. Sequential-Design, multicenter, randomized, controlled trial of early decompressive craniectomy in malignant middle cerebral artery infarction (DECIMAL trial). Stroke 2007;38;2506-2517
2) Hifmeijer et al, for the HAMLET investigators. Surgical decompression for space-occupying cerebral infarction (the Hemicraniectomy after Middle Cerebral Artery infarction with life Threatening edema Trial [HAMLET]): a multicenter, open, randomized trial. Lancet neurol 2009;8:326-33
3) Vahedi et al, for DECIMAL, DESTINY and HAMLET investigators. Early decompressive surgery in malignant infarction of the middle cerebral artery: a pooled analysis of three randomized controlled trials. Lancet Neurol 2007;6:215-22
Resident discussants: Dr Mahua Dey, Dr Nasser Mansour, Dr Sherise Ferguson
Faculty comments: Dr Issam Awad, Dr Jeffrey Frank, Dr Ben Roitberg
Additional resident comments: Dr Joseph Hsieh
DEY: Introduction and overview: Large MCA strokes are often devastating and can be fatal even with best medical management. Management ranges from supportive care to aggressive measures like very early decompressive hemicraniectomy. There are several reported cases and case series that report benefit of decompressive hemicraniectomy in the setting of large MCA stroke. In this Journal club we review three European multicentre, randomized, controlled clinical trials: DECIMAL, DESTINY and HAMLET, which were undertaken to assess the effect of decompressive surgery in patients with space- occupying MCA infarction. Clinical assessments were based on two types of standard evaluations: the National Institutes of Health stroke scale (NIHSS) in the setting of acute stroke, and Modified Rankin Scale (mRS) as a measure of disability. The NIHSS is an 11 point standard examination focusing on basic motor, sensory and language parameters. The NIHSS ranges from 0=No Stroke to 21-42, Severe Stroke scores. ))=No stroke, 1-4=Minor Stroke, 5-15= Moderate Stroke, 15-20=Moderate to Severe Stroke. The mRS provides a disability score from 0 (no symptoms) to 6 (dead). In between are mRS 1 – No significant disability despite symptoms; able to carry out all usual duties and activities; 2 - Slight disability; unable to carry out all previous activities, but able to look after own affairs without assistance; 3 - Moderate disability; requiring some help, but able to walk without assistance; 4 - Moderately severe disability; unable to walk without assistance and unable to attend to own bodily needs without assistance; 5 - Severe disability; bedridden, incontinent and requiring constant nursing care and attention.
All three trials had similar inclusion and exclusion criteria. Major inclusion criteria involved age 18–60 years, clinical deficits suggestive of infarction in the territory of the MCA , a score on the National Institutes of Health stroke scale (NIHSS) >15, decrease in the level of consciousness to a score of 1 or greater on item 1a of the NIHSS, signs on CT of an infarct of at least 50% of the MCA territory with or without additional infarction in the territory of the anterior or posterior cerebral artery on the same side, or infarct volume >145 cm3 as shown on diffusion-weighted MRI, written informed consent by the patient or a legal representative. Major exclusion criteria were: pre-stroke score on the mRS (modified Rankin Scale) ≥2, two fixed dilated pupils, contralateral ischemia or other brain lesion that could affect outcome, space-occupying hemorrhagic transformation of the infarct (≥ parenchymal hemorrhage grade 2), life expectancy <3 years, other serious illness that could affect outcome, known coagulopathy or systemic bleeding disorder, contraindication for anaesthesia, pregnancy. After interim analyses all the trials were stopped prematurely and a pooled analysis was done. Primary outcomes for all the trials were scored on the modified Rankin Scale (mRS) at 1 year, dichotomised between favorable (mRS 0 to 4) and unfavorable (mRS 5 and death) and secondary analyses further subdivided by mRS, in which favorable outcome was defined as a score of 0–3 and unfavorable outcome as a score of 4 to death, and case fatality at 1 year.
MANSOUR: Summary of the pooled analysis: 93 patients were included in the pooled analysis. More patients in the decompressive-surgery group than in the control group had an mRS≤4 (75% vs 24%; pooled absolute risk reduction 51% [95% CI 34–69]), an mRS≤3 (43% vs 21%; 23% [5–41]), and survived (78% vs 29%; 50% [33–67]), indicating numbers needed to treat of two for survival with mRS≤4, four for survival with mRS≤3, and two for survival irrespective of functional outcome. The effect of surgery was highly consistent across the three trials. Although the inclusion and exclusion criteria of the three trials were largely similar, there were differences: a longer interval allowed from stroke onset to start of treatment in HAMLET (99 h) than in DECIMAL (30 h) and in DESTINY (36 h). For the pooled analysis, a maximum time window from stroke onset to randomization of 45 h (ie, 48 h to treatment) was adopted. Neuroimaging criteria were too different between the three trials to be included in the pooled analysis. These criteria were an infarct volume on diffusion-weighted MRI of more than 145 cm3 in DECIMAL, brain CT ischemic changes affecting more than two-thirds of the MCA territory and including the basal ganglia in DESTINY, and brain CT ischemic changes affecting at least two-thirds of the MCA territory with space-occupying edema in HAMLET. Randomization: In DECIMAL, patients were centrally randomized in blocks of four using a pre-established randomization list. In DESTINY, randomization was done according to a central computer-generated randomization list for each participating centre. In HAMLET, randomization was done centrally with a computerized algorithm in which an element of chance was added to the treatment decision of minimization.
The trials used largely similar outcome measures. In DECIMAL, outcomes were assessed by a neurologist unaware of treatment allocation; in DESTINY, outcome was assessed unblinded; and in HAMLET, the score on the mRS was determined independently by three investigators masked to treatment allocation on the basis of a narrative written by an unblinded independent study nurse and, if necessary, this process was followed by a consensus meeting. DECIMAL was designed to include a maximum of 60 patients, 30 in each group. Recruitment stopped after inclusion of 38 patients in March 2006, because of slow enrolment, a significant difference in mortality favoring decompressive surgery, and the opportunity of a pooled analysis with DESTINY and HAMLET. A planned interim analysis of DESTINY including 32 patients showed a significant benefit of surgery on 30 day mortality. and the study was stopped definitively after a revised sample size projection indicated that 188 patients would be needed to show a significant difference in the primary endpoint (mRS 0–3 vs 4 to death at 6 months).
Outcome and analysis: In the pooled analysis, surgery was beneficial (p<0.01) in all predefined subgroups (age [above and below 50 years], presence of aphasia, and time to randomization [above and below 24 h]), as measured by mRS of 4 or less at 12 months, with no significant subgroup-treatment effect interactions. After decompressive surgery the probability of survival increased from 28% to nearly 80% and the probability of survival with an mRS of ≤3 doubled. However, the probability of surviving in a condition requiring assistance from others (mRS of 4) increases more than ten times, although the risk of very severe disability (mRS of 5) is not increased. The choice of performing decompressive surgery in an individual patient with space-occupying hemispheric infarction will therefore depend on the willingness to accept survival with serious disability. Information about quality of life of survivors is essential for guiding such decisions. Previous studies on quality of life after decompressive surgery for space-occupying infarction have reported divergent results. Even patients with aphasia may improve significantly. Information about quality of life will be provided in separate future publications of the trials – this will be very interesting to read. In the three trials under study, patients were excluded if they were older than 55 years or 60 years of age. The results can probably not be generalized to patients who are older. In a systematic review of uncontrolled studies on decompressive surgery, 80% of the patients older than 50 years were dead or remained severely disabled compared with 32% of the patients aged 50 years or younger. Moreover, quality of life can remain impaired, especially in older patients. In the studies we discuss no difference in outcome was found between patients treated on the first and those treated on the second day. However, they were not designed to find such a difference.
FERGUSON: Further comments:
Patient Age: Even though among younger patients malignant MCA infarction is more common, overall, ischemic cerebral stroke is a condition of older individuals. More specifically, more than 60% of patients are older than 50 years, and 40% are older than 60 years old (Hacke W et al. Arch Neurol 1996). Despite this, the DECIMAL and DESTINY trials only investigate the benefits of hemicraniectomy in patients younger than age 60. They had a reasonable basis for concentrating on surgical benefit for younger patients. A 2004 meta-analysis by Gupta et al., investigated the predictors of outcome following hemicraniectomy after malignant MCA infarction in138 patients. The authors found that younger age was the only pre-operative clinical determinant of survival with good functional outcome (Gupta et al., Stroke 2004). There are several other studies that come to a similar conclusion (Chen et al., J of Clin Neuroscience 2007; Curry et al., Neurosurgery 2005; Walz et al., J Neurol 2002). However, the interpretation of such studies should be done cautiously. Most of the available data were collected in studies that were retrospective, uncontrolled and not blinded, allowing for potential bias. For example, physicians and nurses may treat an elderly patient less aggressively than a younger one, which would significantly affect the outcome of these studies. The participants in the journal club concluded neither DECIMAL, DESTINY nor HAMLET adequately address this issue of age. In the pooled analysis, there was no statistical difference comparing patients older or younger than 50 years. However, the number of patients older than 50 years was very small. Thus, the dilemma of offering hemicraniectomy to older patients (>60 years) remains unresolved. Of note, the planned DESTINY II-trial will study patients older than 60 years and will hopefully provide more information on this matter.
Timing of surgery: Practices vary widely among centers. The benefits of waiting mainly center around the notion that not all large MCA infarctions result in malignant cerebral edema. Furthermore, some patients may respond well to medical management for increased ICP and edema and may not need to be subjected to a hemicraniectomy. The clear drawback is that late surgery may be too late to help. Early surgery, on the other hand, may prevent irreversible damage to adjacent brain tissue following cerebral infarction. Furthermore, it may be beneficial to surgically treat malignant infarction upfront before aggressive medical measures are needed and thereby avoid the complications of prolonged intubation. The timing of hemicraniectomy was limited to less than 48 hours after the stroke onset in DECIMAL and DESTINY and less than 96 hours for HAMLET. Additionally, in the pooled analyses, there was no difference in outcome if hemicraniectomy was performed in the first 24 hours compared with later time window up to 48 hours. Hence the benefit of late “salvage” surgery in initially medically managed patients is still uncertain.
Outcome: It is clear that surgery for MCA infarction does save lives but at a cost. Should simply being alive be considered a positive outcome? Even though surgery saves lives, many patients who survive surgery end up unable to walk or care for themselves (mRS 4). For certain individuals, this option may not be preferable to death. What are the ethics behind trading death for functional disability? Furthermore, the question arose regarding whether the mRS (modified Rankin scale) is the most appropriate scale for assessing outcome; in that it only assesses motor capabilities and dependency. It is arguable that the addition of a more comprehensive scale (i.e. quality of life) would be more useful. The combination of such a massive cerebral insult and a big surgery commits the patient and family to several months of inpatient and outpatient rehabilitation, not to mention the need for nursing care, wheelchair, transport, medication etc. With this in mind social support, education, and socioeconomic status are examples of factors that may influence outcome.
Conclusion: The trials presented and the pooled analysis are well designed and give remarkable insight into the outcome of MCA infarction following hemicraniectomy. Even though there are still questions to be answered (i.e. patient age and timing of surgery), overall it gives physicians useful guidelines in order to better counsel family members on how to proceed when a loved one is faced with this devastating condition.
Additional resident questions and comments:
HSIEH: Most patients were done in a few institutions, so it is hard to generalize. Who had the authority to stop the trial?
MANSOUR: Will an increase in the number of operations on patients with MCA infarctions stretch the available neurosurgical services in parts of the US? Are there enough neurosurgeons in the community?
FERGUSON: What to do if there is no family to give consent?
ROITBERG: I have a comment and a question. First, a comment regarding equipoise. This is a relatively new concept, but one that is increasingly recognized as central to studies involving surgical treatment. In many clinical situations, doctors have a clear opinion regarding what has to be done. Patients in those situations should not be randomized, for many reasons. For example, if a patient whose surgeon believes that an operation should be performed is randomized to receive non-operative management, chances are that the operation will still be performed, and the patient will be recorded as a ‘cross-over’, diminishing the value of the study. The study design must take existing practices and opinions into account. The clinical situation of a patient who is a candidate for randomization should be such that a serious doubt exists in the mind of treating physicians regarding the best course of treatment, or at least they believe that both approaches offered as part of the study are acceptable. This situation is called ‘equipoise’. A randomized clinical trial assumes that equipoise exists in a substantial number of cases, and aims to provide data to help resolve actual clinical dilemmas. I believe the study authors recognize the importance of equipoise and tried to exclude extreme cases. However, it is always difficult to adequately address equipoise while including enough patients.
On a different note, the non-operative care was not defined in the studies. Some of the patients were apparently not treated with maximal medical care. Are we dealing with a comparison of maximally aggressive therapy and less aggressive therapy, rather than just surgical or medical management?
AWAD: When to operate is a central question. It is clear that waiting for herniation to occur is too late. But how to decide? Where is the clinical equipoise? The trials discussed here did not address it. Frequent CT scans can be done to look for signs of swelling and brain shift, but are not practical in the community hospital setting. In this case it is better to err on the side of inclusion. Empirically, we may state that operating within 48 hours avoids operating too late. The level of non-operative care can vary significantly. The outcome in a specialized intensive care unit may be much better. Would the clinical equipoise change depending on whether excellent neuro-ICU care is available or not? Is the next equipoise about local surgery versus transfer to a specialized ICU?
Let’s also discuss issues of consent, study design and comparability of data across various institutions. For example, different countries have different consent procedures applied, and a different cultural context regarding end of life decisions. The surgical procedure should also be defined. How large is the flap? If the large flap is defined as 12 cm across, where do you put it? It is important to include the temporal fossa. A stellate incision in the dura helps avoid brain injury by the dural edge as the brain expands. How are dural patching or closure defined and performed? The advent of Duragen and similar products changed the methods.
Dr Hsieh asked who has the authority to stop a trial. This would be the DSMB – Data and Safety Monitoring Board. It is a committee of experts not participating in the study. The labels are open for them.
Regarding Dr Ferguson’s comment about converting patients who would otherwise die to those who are profoundly disabled. Looking at the outcomes, it is clear that rather than convert patients who would die to a vegetative state, the overall outcome was improved by surgical intervention. This is an important result of the study. However, the number of patients with Rankin scores 3 and 4 increased. Whether this is a good outcome remains open to personal interpretation.
Finally, if there is nobody available to give consent, for a patient younger than 60, operate if you think it is necessary.
FRANK: When to operate is indeed a central question. We do not want to wait for intracranial pressure to increase before offering operative treatment. However, it is OK to wait and evaluate the pace and degree of cerebral swelling. Operative intervention also has serious complications. The approach has to be individualized by age, collateral circulation. Some data exists to help guide us regarding prognosis. In the Head First study here we did frequent CT scans (every 24 hours). When midline shift was greater than 7 mm at the anteroseptal location, or greater than 4 mm pineal shift, within 96 hours, we randomized to offer surgery. Among patients who did not meet the criteria, there was no mortality. Another possible prognostic factor may be available – early increase in ICP may be associated with high mortality. This statement is empirical and controversial, though.
Now - some questions and criticism. How was the consent for the study obtained in every case? Was it by the surgeon, or by designated personnel who did not perform the operation? Non-operative management was not uniform. Mortality and outcome of conservative management would be strongly influenced by standardization of the approach. If some of the non-operative patients did not receive appropriate intensive care, the effect on outcome would be large. As an aside, but important to note – Even with a left MCA infarction and global aphasia, many patients will recover receptive language. This is an important consideration, not always appreciated when discussing prognosis.
“Rankin 3 or 4” actual and perceived outcome depends on socioeconomic status, the availability of resources to support the disabled person. However, the surrogate’s decision should be based on the patient’s likely desires, not on the surrogate’s preferences. Is a Rankin 4 outcome reasonable? It is for many, and we should not avoid bringing up the surgical option even if the likely outcome is serious disability.
Hacke W, Schwab S, Horn M, Spranger M, De Georgia M, von
Kummer R.'Malignant' middle cerebral artery territory
infarction: clinical course and prognostic signs. Arch
Neurol. 1996 Apr;53(4):309-15.
Gupta R, Connolly ES, Mayer S, Elkind MS. Hemicraniectomy
for massive middle cerebral artery territory infarction: a
systematic review. Stroke. 2004 Feb;35(2):539-43.
Chen CC, Cho DY, Tsai SC. Outcome of and prognostic factors
for decompressive hemicraniectomy in malignant middle
cerebral artery infarction. J Clin Neurosci. 2007 Apr;14
Curry WT Jr, Sethi MK, Ogilvy CS, Carter BS. Factors
associated with outcome after hemicraniectomy for large
middle cerebral artery territory infarction. Neurosurgery.
Walz B, Zimmermann C, Böttger S, Haberl RL.Prognosis of
patients after hemicraniectomy in malignant middle cerebral
artery infarction. J Neurol. 2002 Sep;249(9):1183-90.