Articles

Early surgery versus initial conservative treatment in

patients with spontaneous supratentorial lobar intracerebral
haematomas (STICH II): a randomised trial
A David Mendelow, Barbara A Gregson, Elise N Rowan, Gordon D Murray, Anil Gholkar, Patrick M Mitchell, for the STICH II Investigators

Summary
Background The balance of risk and benefit from early neurosurgical intervention for conscious patients with Lancet 2013; 382: 397–408
superficial lobar intracerebral haemorrhage of 10–100 mL and no intraventricular haemorrhage admitted within 48 h Published Online
of ictus is unclear. We therefore tested the hypothesis that early surgery compared with initial conservative treatment May 29, 2013
[Link]
could improve outcome in these patients.
S0140-6736(13)60986-1
This online publication has been
Methods In this international, parallel-group trial undertaken in 78 centres in 27 countries, we compared early corrected. The corrected version
surgical haematoma evacuation within 12 h of randomisation plus medical treatment with initial medical treatment first appeared at [Link]
alone (later evacuation was allowed if judged necessary). An automatic telephone and internet-based randomisation on August 2, 2013, and further
corrections have been made on
service was used to assign patients to surgery and initial conservative treatment in a 1:1 ratio. The trial was not
September 16, 2021
masked. The primary outcome was a prognosis-based dichotomised (favourable or unfavourable) outcome of the
See Comment page 377
8 point Extended Glasgow Outcome Scale (GOSE) obtained by questionnaires posted to patients at 6 months. Analysis
Copyright © Mendelow et al.
was by intention to treat. This trial is registered, number ISRCTN22153967. Open Access article distributed
under the terms of CC BY-NC-ND
Findings 307 of 601 patients were randomly assigned to early surgery and 294 to initial conservative treatment; 298 and Newcastle University,
291 were followed up at 6 months, respectively; and 297 and 286 were included in the analysis, respectively. 174 (59%) Neurosurgical Trials Unit,
of 297 patients in the early surgery group had an unfavourable outcome versus 178 (62%) of 286 patients in the initial Newcastle upon Tyne, UK
(Prof A D Mendelow FRCS[SN],
conservative treatment group (absolute difference 3·7% [95% CI –4·3 to 11·6], odds ratio 0·86 [0·62 to 1·20]; p=0·367). B A Gregson PhD,
E N Rowan PhD,
Interpretation The STICH II results confirm that early surgery does not increase the rate of death or disability at P M Mitchell FRCS); Edinburgh
6 months and might have a small but clinically relevant survival advantage for patients with spontaneous superficial University, Centre for
Population Health Sciences,
intracerebral haemorrhage without intraventricular haemorrhage. Medical School, Edinburgh,
UK (Prof G D Murray PhD); and
Funding UK Medical Research Council. Royal Victoria Infirmary,
Newcastle upon Tyne, UK
(A Gholkar FRCR)
Introduction haemorrhage in the internal capsule or basal ganglia.
Correspondence to:
Spontaneous supratentorial intracerebral haemorrhage Also some clots are too small or the patient is too well to Dr Barbara A Gregson,
is a heterogeneous disorder with clinical manifestations consider intervention. The hypothesis in the present Neurosurgical Trials Unit,
that range from none to rapid death. It affects 4 million STICH II study was based on the results of a subgroup 3–4 Claremont Terrace,
patients worldwide each year and median case fatality at analysis from the first STICH trial that accorded with Newcastle upon Tyne NE2 4AE,
UK
1 month is 40%.1 Many survivors remain severely disabled these ideas.10 [Link]@[Link]
and therefore are an enormous burden on stroke services Several prospective randomised controlled trials11–19
with only a quarter having a good outcome.2 were undertaken during the previous century, culmin-
Surgery has the potential to reduce the volume of ating in the first large trial of early surgery for spon-
intracerebral haemorrhage and there is clinical and taneous supratentorial intracerebral haemorrhage,20 the
experimental evidence that mass removal might reduce results of which were neutral. This outcome seemed to
nervous tissue damage, possibly by relieving local occur because some groups of patients did worse with
ischaemia3–6 or removal of noxious chemicals.7–9 Never- surgery (those with deep-seated bleeds or with intra-
theless, responses to surgery do not seem to be ventricular haemorrhage and hydrocephalus) and some
homogeneous, with trial data, expert opinion, and better (patients with superficial lobar haematomas
mechanistic reasoning all indicating that early surgery without intraventricular haemorrhage).10 The same effect
benefits only some clots. For example, large, surgically was noted in a meta-analysis of other studies: a benefit
accessible clots exerting a mass effect might benefit with surgery that was not significant.21
from early surgery; whereas inaccessible clots, with These findings led to the STICH II trial, designed to
surgical approach paths that cross eloquent speech and find out whether early surgery would improve outcomes
motor regions probably do not. Therefore, most neuro- compared with initial conservative treatment in patients
surgeons would remove a large frontopolar intracerebral with superficial lobar supratentorial intracerebral
haemorrhage with recent deterioration in conscious- haemorrhage without intraventricular haemorrhage. The
ness and would not remove a small intracerebral hypothesis was that early surgery could improve outcome

[Link] Vol 382 August 3, 2013 397

 Articles

in conscious patients in whom there is a superficial Randomisation and masking

intracerebral haemorrhage of 10–100 mL and no evidence The Centre for Healthcare Randomised Trials, Aberdeen,
of intraventricular haemorrhage. UK, provided an automatic telephone and internet-based
randomisation service to assign patients in a 1:1 ratio.
Methods Patients were stratified by country group and planned
Trial design and participants operation (craniotomy or other) and within these strata
STICH II was an international, multicentre, prospective, they were allocated according to a minimisation
randomised, parallel group, pragmatic trial as described algorithm based on age (<60 years, 60–69 years, or
For the trial protocol see http:// in the protocol.21 129 neurosurgical units in 39 countries ≥70 years) and neurological deficit in the worst affected
[Link]/ completed all regulatory requirements and registered for arm or leg (normal, weak, or paralysed) with a random
content/12/1/124
participation in this trial. component such that there was a 20% chance of the
For the UK, ethics approval was obtained from the allocation being reversed.
Scotland Multicentre Research Ethics Committee and Patients, relatives, and site investigators were aware of
the Newcastle and North Tyneside Local Research which treatment the patient had been allocated to; however,
Ethics Committee. Elsewhere, appropriate local ethical at the coordinating centre only the data manager was
approval was sought from each participating centre aware of the allocation. Throughout the study, data broken
and the trial was undertaken according to the UK down by treatment assignment were never provided to
Medical Research Council’s good clinical practice investigators or to the study team. Outcome was assessed
guidelines and local ethical and research and develop- with questionnaires that were completed by the patients or
ment procedures. their relatives. If it was necessary to administer these
Patients were eligible if they had a spontaneous lobar questionnaires to the patients then the interviewer was
intracerebral haemorrhage on CT scan (≤1 cm from the masked to treatment allocation. Outcome assignment,
cortical surface of the brain) with a volume of between data cleaning, and CT assessment were all completed
10 mL and 100 mL, were within 48 h of ictus, had a best before the treatment assignment was unmasked.
motor score on the Glasgow Coma Score (GCS) of 5 or 6,
and had a best eye score of 2 or more (ie, were conscious Outcomes
at randomisation). They were ineligible if the haemor- The primary outcome was a prognosis-based favourable
rhage was due to an aneurysm or angiographically or unfavourable outcome dichotomised from the
proven arteriovenous malformation; was secondary to Extended Glasgow Outcome Scale (GOSE) at 6 months
tumour or trauma; involved the basal ganglia, thalamic, after randomisation. GOSE was computed from the
cerebellar, or brainstem regions; or if there was any answers to 14 questions in a postal questionnaire
intraventricular blood. Patients were also ineligible if completed by patients or their relatives.22 The postal
they had any severe pre-existing physical or mental questionnaires were translated into the appropriate
disabilities or comorbidities that could interfere with the languages and mailed to patients or their relatives or
assessment of the outcome. carers at 5 months after randomisation and, if needed,
followed with a reminder at 6 months. In the centres
Interventions where there were problems with the postal systems,
Patients were randomly allocated to either early surgery there were likely to be literacy or language problems, or
or initial conservative treatment. In the early surgery when there had been no response to a reminder,
group, surgeons were expected to endeavour to undertake questionnaires could be completed by a social worker or
evacuation of the haematomas within 12 h. In the initial research nurse, who did not know the treatment
conservative treatment group, delayed evacuation was allocation, in interview with the patient or relative.
permitted if judged clinically appropriate. All patients The prognosis-based outcome was designed as a
were given the best medical treatment as defined differential type of outcome with two different levels of
according to their hospital settings. success.23 The prognostic score was based on GCS, age,
Information about the preintracerebral haemorrhage and haemorrhage volume at randomisation, and the
status and early progress of all patients (including GCS algorithm had been developed from patients with a broad
and focal signs for the first 5 days) was gathered by the range of different intracerebral haemorrhages. It was
local investigator at 2 weeks, discharge, or death, whichever calculated as:
was earliest. Surviving patients had an additional CT scan
at 5 days (give or take 2 days) to assess changes in the 10 × GCS – age – 0·64 × volume
haematoma size with and without surgery. Patients’
general practitioners (UK) or their local investigators The predefined cutoff of 27·67224 for supratentorial
(outside the UK) were then contacted at 4 months after intracerebral haemorrhage was used to divide the
randomisation to confirm that the patients were alive, their patients into a poor prognosis group and a good
place of residence, and to request completion of a major prognosis group. The outcome was judged favourable if
events form. GOSE was good recovery or moderate disability;

398 [Link] Vol 382 August 3, 2013

 Articles

additionally, in the poor prognosis group, upper severe surgery (two-sided p<0·05) with 80% power. We therefore
disability. This disability represents patients who are proposed a sample size of 600 patients to allow for
completely self-caring within their homes but who withdrawals and crossovers.
are unable to shop or use public transport without The independent data monitoring committee reviewed
assistance. data from the study after 50, 100, 200, and 400 patients
Secondary outcomes were mortality, time to death, had been recruited. These interim reviews were con-
prognosis-based dichotomised Rankin (appendix p 4), fidential, with only the data manager and the data See Online for appendix
and GOSE, and Rankin and EuroQoL; all measured at monitoring committee having access to them. The For EuroQoL see [Link]
6 months. committee did not plan any formal interim analyses but [Link]/eq-5d-products/eq-
[Link]
We also report the crossover and major event rates in worked on the principle that a difference of at least
each treatment group. 3 SEs in an analysis of a major outcome event (eg, death
from all causes or independent survival at 6 months)
Statistical analysis would be needed to justify halting or modifying the study
Based on findings from previous work that a prognosis- before the planned recruitment was completed.
based favourable outcome would be noted in 37% of the The detailed analysis plan has been reported pre-
conservative treatment group, a sample size of 566 (283 in viously.24 Analysis was undertaken on an intention-to-
each group) was needed to show a 12% benefit from treat basis by treatment allocation. Outcomes were

3984 patients reported Unknown number of patients

on screening logs not reported on logs

313 patients eligible 221 consented

92 did not consent 601 randomly assigned 380 consented

307 allocated to early surgery group 294 allocated to initial conservative treatment group

2 excluded* 2 excluded*

305 early surgery group (baseline assessment) 292 initial conservative treatment group (baseline assessment)

4 withdrew

2 week follow-up: 2 week follow-up:

288 received allocated intervention 229 received allocated intervention
13 did not receive allocated intervention 62 did not receive allocated intervention
1 no data available for analysis

1 withdrew 1 lost to follow-up

2 lost to follow-up

298 followed up at 6 months 291 followed up at 6 months

1 alive but status 5 alive but status

not known not known

297 analysed for primary outcome at 6 months 286 analysed for primary outcome at 6 months

Figure 1: Trial profile

*One site recruited one patient but had undertaken surgery before randomisation (the patient was allocated to initial conservative treatment); another site recruited
three patients, and two of these had surgery before randomisation (one allocated to early surgery and one to initial conservative treatment); because of the severe
breach of protocol all four patients were excluded.

[Link] Vol 382 August 3, 2013 399

 Articles

reported as odds ratios (OR) with 95% CI. Absolute Secondary analysis consisted of a Kaplan-Meier sur-
differences with 95% CI were also reported. vival curve with log-rank test, χ² test for mortality at
Primary outcome analysis was a simple categorical 6 months, and 6 month prognosis-based Rankin.
frequency comparison by use of the χ² test for prognosis- Additionally, GOSE, Rankin, and EuroQoL were reported
based favourable and unfavourable outcome on GOSE. by treatment allocation.
Logistic regression was undertaken to adjust for co- We also did a sensitivity analysis based on a proportional
variates, age, GCS, volume of haematoma, and neuro- odds model.
logical deficit. Prespecified subgroup analyses were also undertaken
for age (<65 years, ≥65 years), volume of haematoma
Early surgery Initial conservative (≤35 mL, >35 mL), GCS (8–12, 13–15), time from ictus
group (n=305) treatment group to randomisation (<21 h, ≥21 h), and severity of neuro-
(n=292) logical deficit in worse limb (normal, weak, paralysed).
Age (years) An additional analysis of the two prognosis groups was
Median (IQR; range) 65 (55 to 74; 65 (56 to 74; undertaken.
17 to 90) 23 to 94) This trial is registered, number ISRCTN22153967.
Mean (SD) 63·9 (13·0) 63·9 (13·7)
<60 105 (34%) 106 (36%)
60–69 89 (29%) 70 (24%) Early surgery Initial conservative
group (n=305) treatment group
≥70 111 (36%) 116 (40%) (n=292)
Sex
(Continued from previous page)
Male 174 (57%) 166 (57%)
Localising arm
Female 131 (43%) 126 (43%)
Normal 84 (28%) 82 (28%)
Preintracerebral haemorrhage Rankin*
Weak 129 (42%) 116 (40%)
0 240 (80%) 236 (81%)
Paralysed 92 (30%) 94 (32%)
1 41 (14%) 37 (13%)
Localising leg
2 17 (6%) 11 (4%)
Normal 94 (31%) 96 (33%)
3 2 (<1%) 5 (2%)
Weak 141 (46%) 121 (41%)
4 1 (<1%) 2 (<1%)
Paralysed 70 (23%) 75 (26%)
Preintracerebral haemorrhage mobility*†
Neurological deficit
Able to walk 200 m 283 (94%) 275 (95%)
Normal 80 (26%) 81 (28%)
Able to walk indoors 17 (6%) 13 (4%)
Weak 129 (42%) 112 (38%)
Unable to walk 1 (<1%) 2 (<1%)
Paralysed 96 (31%) 99 (34%)
Glasgow Coma Score, eye
Medical history
2 26 (9%) 27 (9%)
Documented hypertension 204 (68%) 196 (67%)
3 69 (23%) 65 (22%) (>140 mm Hg/90 mm Hg)*
4 210 (69%) 200 (68%) On antihypertensive medication* 144 (48%) 150 (52%)
Glasgow Coma Score, verbal Previous myocardial infarction* 18 (6%) 14 (5%)
1 40 (13%) 44 (15%) Previous stroke* 30 (10%) 33 (11%)
2 36 (12%) 25 (9%) Medication before intracerebral haemorrhage*
3 37 (12%) 35 (12%) Anticoagulant drugs 22 (7%) 20 (7%)
4 93 (30%) 96 (33%) Antiplatelet drugs 42 (14%) 31 (11%)
5 99 (32%) 92 (32%) Thrombolytic drugs 4 (1%) 2 (<1%)
Glasgow Coma Score, motor Prognostic score
5 83 (27%) 71 (24%) Median (IQR; range) 42·5 (20 to 62; 41·2 (16 to 63;
6 222 (73%) 221 (76%) –57 to 120) –53 to 121)
Glasgow Coma Score, total Mean (SD) 38·6 (30·7) 39·0 (31·7)
8 12 (4%) 4 (1%) Prognostic score category
9 9 (3%) 15 (5%) Poor 104 (34%) 105 (36%)
10 23 (8%) 21 (7%) Good 201 (66%) 187 (64%)
11 32 (10%) 32 (11%)
For continuous variables, data are median (IQR; range) and mean (SD); for
12 32 (10%) 34 (12%) categorical variables, data are number (%). *Data were missing for four patients
13 46 (15%) 43 (15%) in the early surgery group who withdrew after randomisation and for one patient
in the initial conservative treatment group for whom no 2 week data were
14 70 (23%) 68 (23%)
obtained. †One patient in the initial conservative treatment group did not
15 81 (27%) 75 (26%) provide this information.
(Continues on next page)
Table 1: Baseline characteristics of patients

400 [Link] Vol 382 August 3, 2013

 Articles

Role of the funding source Four patients withdrew from the study after allocation
Neither the sponsor nor the funder of the study had any to early surgery group because they or their relatives
role in study design, data gathering, analysis, and inter- refused surgery and further involvement, and all data for
pretation, or writing of the report. The corresponding
author and ENR had full access to all the data in the study Early surgery group Initial conservative
and all members of the writing committee had respon- (n=288) treatment group (n=62)
sibility for the decision to submit for publication. Surgery
Craniotomy 284 (99%) 59 (95%)
Results Craniectomy 1 (<1%) 3 (5%)
601 patients from 78 centres in 27 countries were randomly Minimally invasive* 3 (1%) ··
assigned between Jan 11, 2007, and Aug 15, 2012: 307 to Any other procedure 15 (5%) 6 (10%)
early surgery and 294 to initial conservative treatment; Paralysed and sedated 20 (7%) 7 (11%)
recruitment by centre is shown in the appendix p 2. Four Preoperative Glasgow Coma Score, eye†
patients were excluded because they were recruited by two 1 7 (3%) 20 (36%)
centres that randomly assigned patients after evacuating 2 23 (9%) 20 (36%)
the haematoma: a serious protocol violation (figure 1). All 3 62 (23%) 11 (20%)
other patients were included in the analysis irrespective of 4 176 (66%) 4 (7%)
the decision of the central CT reading committee about Preoperative Glasgow Coma Score, verbal†
their eligibility (the CT committee’s findings will be 1 35 (13%) 24 (44%)
reported later in a separate paper). This analysis, therefore,
2 30 (11%) 11 (20%)
includes 305 patients assigned to early surgery and 292 to
3 33 (12%) 12 (22%)
initial conservative treatment (figure 1). Table 1 shows
4 83 (31%) 6 (11%)
details of the patients’ age, sex, previous medical history,
5 87 (33%) 2 (4%)
and neurological status. The two groups were well matched
Preoperative Glasgow Coma Score, motor†
at baseline. 57% were men and the median age of the
1 ·· ··
patients was 65 years (range 17–94; table 1). Patients were
2 ·· 1 (2%)
randomly assigned within 48 h of ictus and a quarter
3 1 (<1%) 11 (20%)
(76 [25%] of 305 in the early surgery group and 73 [25%] of
4 5 (2%) 14 (25%)
292 in the initial conservative treatment group) were
5 69 (26%) 24 (44%)
assigned within 12 h (median 21·6 h [IQR 12·0–31·5] and
6 193 (72%) 5 (9%)
21·0 [12·0–32·0], respectively). 50% of the patients in the
early surgery group and 49% in the initial conservative Preoperative arm†

treatment group had a GCS of 14 or 15 at randomisation Normal 75 (28%) 2 (4%)

(table 1). The planned method of evacuation in 98% of all Weak 116 (43%) 10 (19%)
cases was craniotomy. Table 2 shows the haematoma Paralysed 76 (28%) 40 (77%)
characteristics reported by site investigators at random- Preoperative leg†
isation. The median volume of the haematoma (with the Normal 82 (31%) 2 (4%)
Broderick method25) was 36 mL (23·0–55·5) and the Weak 129 (48%) 12 (24%)
median depth from the cortex surface was 1 mm (0–2). Paralysed 56 (21%) 38 (73%)
Preoperative speech†
Normal 121 (48%) 5 (15%)
Early surgery group Initial conservative Dysarthric 60 (24%) 14 (41%)
(n=305) treatment group (n=292)
Aphasic 70 (28%) 15 (44%)
Volume (mL) Time randomisation to surgery (h)
Median (IQR; range) 38 (24–54; 10–100) 36 (22–58; 10–100) Median (IQR; range) 3·3 (1·9–5·7; 0·2–34·6) 22·8 (5·6–68·8; 0·6–249·6)
Mean (SD) 41·4 (21·2) 41·0 (22·9) Mean (SD) 4·5 (4·3) 45·5 (56·0)
Depth (mm) Surgery within 12 h of randomisation 281 (98%) 23 (37%)
Median (IQR; range) 1 (0–2; 0–10) 1 (0–2; 0–10) Time from ictus to surgery (h)
Mean (SD) 1·6 (2·4) 1·6 (2·5) Median (IQR; range) 26 (15·3–35·3; 2·5–69·0) 46 (21·7–81·9; 5·0–266·5)
Side of haemorrhage* Mean (SD) 26·7 (13·6) 64·2 (58·9)
Left 158 (52%) 149 (51%) Surgery within 12 h of ictus 48 (17%) 9 (15%)
Right 147 (48%) 142 (49%)
For continuous variables, data are median (IQR; range) and mean (SD); for categorical variables, data are number (%).
For continuous variables, data are median (IQR; range) and mean (SD); for *Minimally invasive procedures were burrhole with endoscopic evacuation in two patients and keyhole evacuation in one
categorical variables, data are number (%). *Data were missing for one patient in patient. †These questions were not answered for patients who were paralysed and sedated at that time and data were not
the initial conservative group for whom no forms were returned. recorded for some other patients.

Table 2: Characteristics of haematomas Table 3: Surgery details

[Link] Vol 382 August 3, 2013 401

 Articles

one patient in the initial conservative treatment group had complete follow-up at 6 months for the primary
were lost by the centre (figure 1). Thus, process data were outcome analysis.
available for 301 patients in the early surgery group and Of the 301 assessable patients in the early surgery group,
291 in the initial conservative treatment group. Another 288 (96%) had surgery and 281 (93%) had surgery within
patient withdrew (self-discharged) in the early surgery 12 h (table 3). 13 (4%) patients did not have surgery because
group between 2 weeks and the 6 month follow-up their families refused (n=6), or they had a rebleed or
because the patient did not want to have surgery intraventricular haemorrhage (3), cardiac problem, respir-
(figure 1). Nine patients were lost to follow-up (although atory problem, or fever (3), and logistical problems (1). Of
six were known to be alive at 6 months and could be 291 assessable patients in the initial conservative treatment
included in survival analyses; figure 1). Thus, 583 patients group, 62 (21%) had surgery (table 3). Reasons for these
patients requiring operation were deterioration in GCS
(n=36), oedema (1), rebleed (3), deterioration and oedema
Early surgery group Initial p value Absolute
conservative difference (7), deterioration and rebleed (5), deterioration, oedema,
treatment group (95% CI) and rebleed (4), rise in intracranial pressure (2), surgeon
Primary outcome 297 286 error (1), family request (2), and an underlying cause (1).
Prognosis based 0·367* 3·7% Craniotomy was the most commonly used surgical
(–4·3 to 11·6) technique for evacuation in 343 (98%) of 350 of cases.
Unfavourable 174 (59%) 178 (62%) ·· ·· Comparison of patients in the initial conservative
Favourable 123 (41%) 108 (38%) ·· ·· treatment group who had surgery with those who did
Secondary outcomes 298 291 not showed that they were more likely to undergo
Mortality at 6 months 0·095* 5·6% surgery if they had a paralysed limb at randomisation
(–1·0 to 12·2) (34 [55%] of 62 vs 65 [28%] of 229; p<0·0001), lower GCS
Dead 54 (18%) 69 (24%) ·· ·· (median 13 [IQR 10–14] vs 14 [12–15]; p<0·0001), larger
Alive 244 (82%) 222 (76%) ·· ·· haematoma (54 mL [35–74] vs 32 mL [20–50]; p<0·0001),
Prognosis-based modified 0·456* 3·1% or were in the poor prognosis group (34 [55%] vs 71 [31%];
Rankin (–5·0 to 11·2) p=0·0005). Patients who crossed over to surgery were
Unfavourable 155 (53%) 158 (56%) ·· ·· more likely than were those allocated to early surgery to
Favourable 140 (47%) 126 (43%) ·· ·· have a GCS of at least one point lower just before their
GOSE 0·091*; ·· operation than at randomisation (47 [82%] of 57 vs
0·075†
50 [19%] of 269; p<0·0001).
Dead 54 (18%) 69 (24%) ·· ··
At 2 weeks, the status of patients was classified as dead,
Vegetative 0 0 ·· ·· still on a neurosurgical ward, transferred to another unit
Lower severe disability 64 (22%) 66 (23%) ·· ·· or hospital, or discharged. There were significant
Upper severe disability 72 (24%) 59 (21%) ·· ·· differences between the early surgery group and initial
Lower moderate disability 20 (7%) 15 (5%) ·· ·· conservative group (p=0·02). In the early surgery group,
Upper moderate disability 32 (11%) 35 (12%) ·· ·· 16 (5%) of 301 patients had died, 85 (28%) were still on a
Lower good recovery 37 (12%) 26 (9%) ·· ·· neurosurgical ward, 80 (27%) were transferred, and
Upper good recovery 18 (6%) 16 (6%) ·· ·· 120 (40%) were discharged, whereas in the initial
Rankin 0·128*; ·· conservative treatment group 29 (10%) of 291 patients
0·147†
had died, 102 (35%) were still on a neurosurgical ward,
0 20 (7%) 16 (6%) ·· ··
60 (21%) were transferred, and 100 (34%) were
1 54 (18%) 57 (20%) ·· ·· discharged. Few patients in either group (four [1%] of 301
2 58 (20%) 41 (14%) ·· in the early surgery group and one [<1%] of 291 in the
3 35 (12%) 32 (11%) ·· ·· initial conservative treatment group) were treated with
4 40 (14%) 28 (10%) ·· ·· factor VIIa. Patients in the initial conservative group
5 34 (12%) 41 (14%) ·· ·· were more likely to have an angiogram (108 [37%] of 291
Dead 54 (18%) 69 (24%) ·· ·· vs 87 [29%] of 301; p=0·034). However, clinically
EuroQoL Index 235 210 0·751‡ ·· significant lesions were only found in ten patients: three
Median (IQR; range) 0·64 (0·20 to 0·85; 0·69 ·· ·· with an arteriovenous malformation and three with an
–0·59 to 1·00) (0·08 to 0·82;
aneurysm in the early surgery group, and three patients
–0·59 to 1·00)
with an arteriovenous malformation and one with an
Data are number or number (%), unless otherwise indicated. EuroQol utility index was calculated with UK weightings aneurysm in the initial conservative treatment group.
provided by the EuroQol Group Foundation. Absolute differences (95% CIs) are provided for binary outcomes. Rankin
Post-randomisation adverse events reported during the
was not available for three patients in the early surgery group and for seven in the initial conservative group. GOSE was
not available for one patient in the early surgery group and five patients in the initial conservative group. hospital stay before the 2 week point were similar in the
GOSE=Extended Glasgow Outcome Scale. *χ2 test. †Proportional odds model. ‡Mann-Whitney test. two groups: overall 37 (6%) of 592 patients had a further
intracerebral haemorrhage increasing the volume by at
Table 4: Prespecified outcomes at 6 months
least 20% (14 in the early surgery group and 23 in the

402 [Link] Vol 382 August 3, 2013

 Articles

initial conservative treatment), six (1%) had an ischaemic 95% CI 0·26–0·92; p=0·02; figure 4). By contrast, there
stroke (five and one, respectively), six (1%) a pulmonary was no advantage for surgery in the good prognosis
embolism (one and five, respectively), 16 (3%) a major group (1·12, 0·75–1·68; p=0·57).
cardiac event (nine and seven, respectively), and 71 (12%) There were differences in the causes of death between
pneumonia (31 and 40, respectively). the two groups. Patients in the early surgery group were
With the prognosis-based dichotomy of GOSE, 123 (41%) more likely to die from cardiac events (14 [26%] of 54 vs
of 297 patients in the early surgery group had a favourable five [7%] of 69) and less likely to die from intracerebral
outcome at 6 months compared with 108 (38%) of 286 haemorrhage or rebleed (eight [15%] vs 20 [29%]), chest
patients in the initial conservative treatment group
(OR 0·86, 95% CI 0·62 to 1·20; p=0·367). Early surgery
had an absolute benefit of 3·7% (table 4) and a relative 1·0 Early surgery
Initial conservative treatment
benefit of 9·7% (–11·4 to 30·8). Adjustment for the
covariates age, GCS, haemorrhage volume, and neuro-
logical deficit made little difference to the prognosis-based
outcome (0·85, 0·59 to 1·22; p=0·384). 0·8

The mortality rate at 6 months was 18% in the early

surgery group and 24% in the initial conservative
treatment group (table 4; OR 0·71, 95% CI 0·48 to 1·06;
0·6
Probability of survival

p=0·095); absolute difference in favour of early surgery

was 5·6% (table 4) and the relative difference was 7·3%
(–1·3 to 16·0). The actual survival advantage during the
first 6 months with early surgery was not significant
0·4
(figure 2). 27 (9%) of 298 patients died at 30 days and
43 (14%) at 90 days in the early surgery group, whereas
43 (15%) of 291 patients died at 30 days and 63 (22%) at
90 days in the initial conservative treatment group. 0·2
Table 4 shows the full extended GOSE, Rankin, and
log-rank p=0·073
EuroQoL by treatment group. The prognosis-based
Rankin showed favourable outcome in 47% of the
patients in the early surgery group and in 44% of those in 0
the initial conservative treatment group (p=0·46; table 4); 0 30 60 90 120 150 180
Days since randomisation
the absolute difference in favour of early surgery was Number at risk
3·1% (table 4) and the relative difference was 7·0% Early surgery 300 271 260 255 251 231 169
Initial conservative 292 252 232 237 222 202 148
(95% CI –11·4 to 25·3). The actual distribution of GOSE treatment
was more favourable for the early surgery group than for
the initial conservative treatment group (figure 3), Figure 2: Kaplan–Meier survival curve
although the difference was not significant (p=0·091;
table 4). The proportional odds model analysis of these Dead Lower severe disability Lower moderate disability Lower good recovery
Upper severe disability Upper moderate disability Upper good recovery
data (OR 0·77, 95%CI 0·58 to 1·03; p=0·075) was
consistent with the χ² trend analysis.
Figure 4 shows prespecified subgroup analyses and
analyses for the poor and good prognosis groups. No Early surgery
prespecified subgroups showed heterogeneity of treat-
ment response.
At 6 months, 79 (39%) of 203 patients in the poor
prognosis group died and 67 (33%) had lower severe
disability, whereas 44 (12%) of 380 patients died in the
good prognosis group and 63 (17%) had lower severe Initial
disability. Patients in the good prognosis group were conservative
much more likely to have a good recovery (85 [22%] of treatment

380) or moderate disability (89 [23%]) than were those in

the poor prognosis group (12 [6%] of 203 and 13 [6%],
respectively). Subgroup analysis of the prognosis-based
0 20 40 60 80 100
prediction group showed significant heterogeneity
Percentage of cases
(I²=79%, p=0·03). Patients in the poor prognosis group
were more likely to have a favourable outcome with early Figure 3: Extended Glasgow Outcome Scale at 6 months
surgery than with initial conservative treatment (OR 0·49, Proportional odds model p=0·075.

[Link] Vol 382 August 3, 2013 403

 Articles

Early surgery Initial conservative treatment Odds ratio* (95% CI)

Events Total Events Total
Age of patient (years)
<65 72 145 68 140 1·04 (0·66–1·66)
≥65 102 152 110 146 0·67 (0.40–1·11)
I2=39%, p=0·20

Haematoma volume (mL)

≤35 68 136 77 141 0·83 (0·52–1·33)
>35 106 161 101 145 0·84 (0·52–1·36)
I2=0%, p=0·98

Randomisation GCS
8–12 67 103 76 105 0·71 (0·39–1·28)
13–15 107 194 102 181 0·95 (0·63–1·43)
I2=0%, p=0·42

Time from ictus to randomisation

<21 h 82 143 95 143 0·68 (0·42–1·10)
≥21 h 92 154 83 143 1·07 (0·68–1·70)
I2=45%, p=0·18

Worst limb deficit

Normal 32 78 31 80 1·10 (0·58–2·08)
Weak 76 127 67 107 0·89 (0·52–1·51)
Paralysed 66 92 80 99 0·60 (0·31–1·18)
I2=0%, p=0·44

Country
Czech Republic 7 12 5 9 1·12 (0·20–6·41)
Egypt 12 23 16 22 0·41 (0·12–1·42)
Germany 27 48 29 44 0·67 (0·29–1·55)
India, Sri Lanka, Nepal 22 42 19 38 1·10 (0·46–2·65)
Latvia 7 15 7 17 1·25 (0·31–5·07)
Lithuania 12 15 9 13 1·78 (0·32–10·01)
Poland 5 11 6 11 0·69 (0·13–3·72)
Romania, Macedonia, Turkey 24 38 21 32 0·90 (0·34–2·40)
Spain 9 12 5 8 1·80 (0·26–12·50)
UK 28 39 27 43 1·51 (0·59–3·83)
USA 10 16 15 17 0·22 (0·04–1·33)
Other countries 11 26 19 32 0·50 (0·18–1·43)
I2=0%, p=0·69

Previous anticlotting agent

No 134 240 144 240 0·84 (0·59–1·21)
Yes 40 57 33 45 0·86 (0·36–2·04)
I2=0%, p=0·97

Prognosis group
Poor 64 99 82 104 0·49 (0·26–0·92)
Good 110 198 96 182 1·12 (0·75–1·68)
I2=79%, p=0·03

0·1 0·2 0·5 1·0 2·0 5·0 10·0

Favours early surgery Favours initial conservative

Figure 4: Subgroup analysis Forest plot

GCS=Glasgow Coma Score.

404 [Link] Vol 382 August 3, 2013

 Articles

infection (13 [24%] vs 20 [29%]), or a pulmonary embolism respectively, seizures in ten and five, respectively, and
(two [4%] vs nine [13%]) than were those in the initial cardiac events in four and eight, respectively.
conservative treatment group. There were an additional
139 serious adverse events reported in 107 patients Discussion
(appendix p 1), with no differences between the treatment In STICH II, using the prognosis based outcome, we
groups. The most common adverse events were did not find significant evidence to support our
respiratory infection in eight patients in the early surgery hypothesis that early surgery compared with initial
group and 12 in the initial conservative group, intra- conservative treatment (with delayed surgery if the
cerebral haemorrhage or rebleed in nine and 12, patient deteriorates) improves outcome in conscious
respectively, neurological deterioration in nine and 17, patients in whom there is a superficial intracerebral

A
Surgery Control Peto odds ratio (95% Cl)
McKissock, et al (1961)
15
71/89 60/91 2·00 (1·04–3·86)
Auer, et al11 (1989) 28/50 37/50 0·46 (0·20–1·04)
Juvela, et al14 (1989) 25/26 21/26 4·39 (0·81–23·65)
Batjer, et al12 (1990) 6/8 11/13 0·55 (0·06–4·93)
Chen, et al19 (1992) 40/64 31/62 1·66 (0·82–3·34)
Morgenstern, et al16 (1998) 9/15 11/16 0·69 (0·16–2·94)
Zuccarello, et al17 (1999) 4/9 7/11 0·48 (0·09–2·69)
Chen, et al13 (2001) 86/263 97/230 0·67 (0·46–0·96)
Teernstra, et al18 (2001) 33/36 29/33 1·51 (0·32–7·12)
Hosseini, et al32 (2003) 0/1 0/1 Not estimable
Hattori, et al33 (2004) 60/121 82/121 0·47 (0·28–0·79)
Mendelow, et al20 (2005) 346/468 378/496 0·89 (0·66–1·19)
Pantazis, et al34 (2006) 36/54 49/54 0·24 (0·10–0·60)
Wang, et al35 (2009) 87/194 120/181 0·42 (0·28–0·63)
Mendelow, et al (2013) 174/297 178/286 0·86 (0·62–1·20)
Total (95% CI) 1695 1671 0·74 (0·64–0·86)

Total events: 1005 (surgery), 1111 (control)

Test for heterogeneity: χ²=39·29, df=13 (p=0·0002), I²=66·9%
Test for overall effect: Z=4·00 (p<0·0001)

0·1 0·2 0·5 1·0 2·0 5·0 10·0

Favours surgery Favours control

B
Surgery Conservative Fixed odds ratio (95% Cl)

Lobar haematomas with no intraventricular haemorrhage (assuming Auer intraventricular

haemorrhage distributed evenly between treatment groups)
Auer, et al11 (1989) 8/21 12/18 0·31 (0·08–1·15)
Juvela, et al14 (1989) 3/3 0/3 49·00 (0·74–3236·99)
Zuccarello, et al17 (1999) 1/3 1/3 1·00 (0·03–29·81)
Chen, et al13 (2001) 7/11 9/13 0·78 (0·14–4·27)
Teernstra, et al18 (2001) 15/16 7/9 4·29 (0·33–55·59)
Mendelow, et al20 (2005) 66/112 90/128 0·61 (0·36–1·03)
Mendelow (STICH II), et al (2013) 190/297 194/286 0·84 (0·60–1·19)
Subtotal (95% CI) 463 460 0·78 (0·59–1·02)

Total events: 290 (surgery), 313 (conservative)

Test for heterogeneity: χ²=8·43, df=6 (p=0·21), I²=28·8%
Test for overall effect: Z=1·81 (p=0·07)

0·1 0·2 0·5 1·0 2·0 5·0 10·0

Favours surgery Favours conservative

Figure 5: Updated meta-analysis of 15 trials of surgery in patients with intracerebral haemorrhage (A) and individual patient data in cases of lobar
haematomas without intraventricular haemorrhage (B)
Data are n/N, unless otherwise indicated. Reported outcomes were unfavourable outcome in (A) and death or disability in (B). df=degrees of freedom.

[Link] Vol 382 August 3, 2013 405

 Articles

haemorrhage of 10–100 mL and no evidence of The absence of a significant difference between the two
intraventricular haemorrhage. groups in this trial might be because of the case mix of
The prespecified recruitment target was attained in patients: more than half were fully conscious or just
the STICH II trial and a follow-up rate of 98% was confused; such patients perhaps could be safely observed
achieved at 6 months. Historically, recruitment and and delayed surgery undertaken only if their state
power are difficult and ubiquitous issues in surgical deteriorates. There is evidence that the lack of effect could
trials. The proportional odds model is sensitive to be a consequence of surgery being beneficial for some and
differences across the outcome scale rather than just to not for other patients with an overall average showing
those in the primary dichotomy (favourable or little difference. In secondary analyses, a survival advan-
unfavourable). That this analysis gives a lower p value tage (with no vegetative survivors) was noted for surgery
than the primary outcome suggests insufficient power throughout the 6 month follow-up (figure 2), but it was
might have been an issue; however, this trial was not significant. Mortality rate remained lower with early
powered to detect a 12% difference in outcome and took surgery than with initial conservative treatment through-
67 months to reach the target sample size. If the 4% out. This pattern contrasts with the Kaplan-Meier plot
difference was regarded as clinically significant it would from the first STICH trial in which a more diverse group
require many more patients. of patients with intracerebral haemorrhage was examined
Interpretation of many surgical trials is also complicated and there was no difference.20 Because we had dichoto-
by crossovers from conservative treatment to surgery as mised patients’ prognoses as part of the primary outcome
commonly seen in spinal, cardiac, and neurosurgical in the STICH II trial, we looked at the post hoc effects of
trials.20,26,27 In STICH II, 62 (21%) of 291 patients assigned surgery in the two prognostic groups. Patients in the
to initial conservative treatment went on to have delayed STICH II trial with a poor prognosis did better with early
surgery. At the time of the delayed surgery, the patients surgery, whereas those with a good prognosis did not
were in deeper coma with worse neurological deficits than (figure 4). The interaction test was significant (p=0·03),
were those in the early surgery group (table 3). The but this result needs to be interpreted cautiously because
crossover to surgery from initial conservative treatment the analysis was not prespecified. However, the result
might therefore have rescued these patients from what accords with some surgeons’ preference to initially
otherwise might have been a fatal outcome, but because of observe patients with a good prognosis thus avoiding
the intention-to-treat analysis they remained in the initial surgery unless a patient deteriorates later.
conservative treatment group. Two-thirds of patients in this trial had documented
hypertension before the intracerebral haemorrhage.
Panel: Research in context Almost a quarter of these hypertensive patients were not
on antihypertensive medication at the time of their ictus.
Systematic review In STICH II, almost all the patients undergoing
Previously reported meta-analyses36,37 were updated with the addition of data from STICH II surgery had craniotomy. Other operative procedures
after confirmation that there had been no other reports of trials of surgery versus initial include decompressive craniectomy and minimal access
conservative treatment. STICH II is the 15th pragmatic randomised controlled trial of surgery procedures. Use of decompressive craniectomies have
for intracerebral haemorrhage compared with conservative treatment to be reported. been reported for patients with intracerebral haemor-
STICH II (n=597) is of moderate size compared with occlusive stroke trials but in intracerebral rhage28 but no prospective randomised controlled trials
haemorrhage it is second only in size to STICH (n=1033). Incorporation of the results from have yet been undertaken to compare their effect with
STICH II (583 patients) with the previous meta-analysis of 14 trials of surgery37 gives a total conservative treatment. Other minimally invasive tech-
sample size of 3366. The result shows a significant advantage for surgery with an odds ratio niques are being tested in ongoing trials, in particular
of 0·74 (95% CI 0·64–0·86; p<0·0001), although there is significant heterogeneity stereotactic delivery of tissue plasminogen activator to
(p=0·0002) because the studies included different patient groups and different types of clots to dissolve them29 and tissue plasminogen activator
surgery. Addition of the STICH II data to a previous individual patient data meta-analysis assisted clearance of ventricular haemorrhage.30 Almost
subgroup of patients with a lobar intracerebral haemorrhage and no intraventricular half of all intracerebral haemorrhages are associated with
haemorrhage38 did not show evidence of heterogeneity (p=0·21) but there was still not a intraventricular haemorrhage and preliminary results
significant benefit from surgery (n=923; 0·78, 0·59–1·02; p=0·07). from CLEAR IVH,31 the precursor to CLEAR III, have
Interpretation shown a benefit for treatment with intraventricular tissue
The results of the meta-analyses suggest that there is a role for surgery in patients with plasminogen activator. In the STICH and STICH II trials,
intracerebral haemorrhage, but that there is still some uncertainty about which patients craniotomy was assessed whereas in the CLEAR III and
benefit most. The STICH studies differ from the other trials in that early surgery was MISTIE trials minimal access techniques were assessed.
compared with the option of delayed surgery for patients who later deteriorate. The survival Figure 5 shows the updated meta-analyses for the trials
advantage for conscious patients with lobar haematomas seems to be greatest when the of surgery for intracerebral haemorrhage and for the
prognosis is poorer (Glasgow Coma Score 9–12) and when randomly assigned within 21 h. subgroup of patients with lobar intracerebral haemor-
This slight advantage is lost in patients with a better prognosis perhaps because there is rhage and no intraventricular haemorrhage; the results
time to observe them initially and only operate in those that later deteriorate. are discussed in the panel. However, minimal access
techniques might be more beneficial for deeper clots

406 [Link] Vol 382 August 3, 2013

 Articles

and intraventricular haemorrhage. Further research is India A Agrawal, A Kakani (Acharya Vinoba Bhave Rural Hospital,
warranted to assess the benefits of these other surgeries. Maharashtra), S A V Rao, N K Venkataramana, A L Naik (BGS Global
Hospitals, Bangalore), S S Grewal, B Gupta (Christian Medical College and
The results of STICH II confirm that early surgery does Hospital, Ludhiana), R Bhattacharya (AMRI Hospitals, Dhakuria),
not increase the rate of death or disability at 6 months T V R Murty (Care Hospitals, Hyderabad), B Indira Devi, G Jagath Lal
and might have a clinically relevant survival advantage (National Institute of Mental Health and Neurosciences, Bangalore),
for patients with spontaneous superficial intracerebral P S Chandra, M Tripathi, B S Sharma (All India Institute of Medical
Sciences, New Delhi), K Sridhar, R Sengupta (National Neurosciences
haemorrhage without intraventricular haemorrhage. Centre, Calcutta), S Nair, G Menon (Sree Chitra Tirunal Institute for
Contributors Medical Sciences and Technology, Trivandrum), P V Ramana,
ADM, BAG, PMM, GDM, and AG conceived the idea for and designed P M Jagannath (Care Hospital, Visakhapatnam); Israel L Levi, M Zaaroor
the study. All listed investigators contributed to enrolment of patients (Rambam Hospital, Haifa); Italy R Delfini, A Pichierri (University Hospital
and interpretation of the data. BAG, ENR, and GDM analysed the study. Sapienza, Rome); Japan A Morita, T Kimura (Hospital NTT Medical
ADM, ENR, and BAG drafted the report, but all listed contributors Center Tokyo, Tokyo); Latvia K Auslands (Riga East University Hospital,
edited and revised the report. Clinic Gailezers, Riga), E Valeinis, R Mikijanskis (Pauls Stradins Clinical
University Hospital, Riga); Lithuania A Gvazdaitis, K Jacikevicius,
STICH II Investigators
D Liutkus (Klaipeda University Hospital, Klaipeda); Macedonia K Lozance,
Writing Committee: A D Mendelow, B A Gregson, E N Rowan,
A Chaparoski, I Pangovski (University Neurosurgical Clinic Skopje,
G D Murray, A Gholkar, P Mitchell. Steering Committee: P Sandercock,
Skopje); Malaysia N Rahman (Hospital Sultanah Aminah, Johor Bahru),
G Ford, D Barer, A Strong, P M Mitchell, A R Gholkar, G D Murray,
J M Abdullah, S K Sim (Hospital Universiti Sains Malaysia, Kubang
A D Mendelow, B A Gregson. Data Monitoring Committee: D Hanley,
Kerian); Mexico S Romero-Vargas, D Mendez-Rosito (Instituto Nacional de
D T Hope, A Skene, H M Fernandes. Radiology Committee: S Metcalfe,
Neurologia y Neurochirugia, Mexico City), J Ruiz-Sandoval (Hospital Civil
A Iqbal, A Gholkar, K S M Prasad. Management Committee:
de Guadalajara, Guadalajara); Nepal Y B Roka (B P Koirala Institute of
A D Mendelow (principal investigator), B A Gregson (principal
Health Sciences, Dharan), K Sharma (B & B Hospital, Lalitpur); Pakistan
investigator and trial director), E N Rowan (data manager 2008–09, trial
K Mahmood, T Salahuddin (Lahore General Hospital, Lahore), M T Khan,
manager 2009–13), G M Kenyon (trial administrator 2007–13), L Chilton
F F Khan (Northwest General Hospital and Research Centre, Peshawar);
(data manager 2006–08), Z Liao (data manager 2009–10), A Andras (data
Poland S Nowak, B Sokol (Karol Marcinkowski University of Medical
manager 2010–11), R Francis (data manager 2012), L Bailey (trial
Sciences, Poznan), P Szydlik, Z Mariak, J Kochanowicz (Medical
administrator 2006–07). National Coordinators: Armenia Ruben
University Hospital, Bialystok); Romania I S Florian, P A Kiss (Cluj County
Fanarjyan; Australia Andrew Kaye; Egypt Abd-Elhafiz Shehab-Eldien;
Emergency Hospital, Cluj-Napoca), H Ples, M A Angelescu,
England, UK Andrew King; Germany Hansdetlef Wassmann; Greece
M S Hanas (County Hospital, Timisoara); Russia A Krivoshapkin
George Stranjalis; India Bhawani S Sharma; Israel Leon Levi; Latvia Egils
(Novosibirsk State Medical University, Novosibirsk); Saudi Arabia
Valeinis; Russia Alex Krivoshapkin; Scotland, UK Sam Eljamel; Spain
E Elgamal (King Khalid University Hospital, Riyadh); Singapore
Alfonso Vazquez-Barquero; Turkey Orhan Barlas; USA
D K S Choy, K J Teo (National University Hospital, Singapore); South Africa
Christopher Loftus. Physician Champions: Armenia Ararat Minasyan;
S Mokgokong (Steve Biko Academic Hospital, Pretoria); Spain R S Sarabia,
Australia Stephen Davis; Egypt Nabil Kitchener; England, UK
I A Reganon (Hospital Universitario Rio Hortega, Valladolid), M Galarza
Philippa Tyrrell; Germany Thorsten Steiner; Greece Kostas Vemmos;
(University Hospital Murcia, Murcia), A Vazquez-Barquero, I Pinto Rafael
India Kameshwar Prasad; Israel Sagi Har Nof; Latvia Andrejs Millers;
(University Hospital Marques de Valdecilla, Santander), J Garibi,
Russia Tatyana Makhovskaya; Scotland, UK Rustam Al-Shahi Salman;
I Pomposo (Cruces University Hospital, Bilbao), R Sarabia (Hospital
Spain Ruben Martin Laez; Turkey Sara Bahar; USA Benjamin Eidelman;
Clinico Universitario, Valladolid), J Ibanez, E Gonzalez (Son Espases
Wales, UK Anne Freeman.
University Hospital, Palma de Mallorca), C Dominguez, L Muñoz
Centre Investigators (recruited at least one patient or completed screening
(Germans Trias I Pujol Hospital, Badalona-Barcelona); Sri Lanka
logs; appendix pp 2–3): Australia S Davis, P Hand (Royal Melbourne
H S Kularathne, S D Perera, P Kamani (National Hospital of Sri Lanka);
Hospital, Melbourne, VIC); Austria G Kleinpeter (Rudolfstiftung Wein,
Turkey O Barlas, N Y Barlas (Istanbul University Faculty of Medicine,
Vienna); Canada M Findlay (University of Alberta Hospital, Edmonton,
Istanbul); UK J Timothy, R Mathew (Leeds General Infirmary, Leeds),
ALB); China Y Zhao (Beijing Tiantan Hospital, Beijing), Y Sin, J Hu
R Strachan, S Metcalfe (James Cook Univesity Hospital, Middlesbrough),
(Huashan Hospital, Shanghai); Czech Republic T Krejci, S Poticny (Faculty
A King, H Patel (Salford Royal, Salford), B A Bell, T L Jones (Atkinson
Hospital of Ostrava, Ostrava), V Benes, O Bradac, M Mohapl (Charles
Morley Department of Neurosurgery, London), D Bulters, A Belli, S Ross
University and Military University Hospital, Prague), M Smrcka, V Juran,
(Southampton University Hospital, Southampton), S Eljamel, F Falcone
K Svoboda (University Hospital, Brno), P Buchvald, V Benes 3rd (Liberec
(Ninewells Hospital and Medical School, Dundee), G Critchley (Hurstwood
Regional Hospital, Liberec); Egypt O S Abdelaziz, I Zidan (Alexandria
Park Neurosurgical Centre, Brighton), P Kirkpatrick (Addenbrooke’s
University Hospital, Alexandria), A-E Shehab-Eldien, E M Kandil,
Hospital, Cambridge), J Crossman, A D Mendelow, P Mitchell, N Ross,
H M Taher, M F El-Faresy (Mansoura International Specialised Hospital,
P Bhattathiri, S Metcalfe, D Holliman (Royal Victoria Infirmary,
Mansoura); Germany A Sepehrnia (Clemens Hospital, Münster), J Kiwit,
Newcastle), P Bhatt, M Kamel (Aberdeen Royal Infirmary, Aberdeen),
S Schreiber, F Youssef (Helios Klinikum Berlin Buch, Berlin),
P Eldridge, M Javadpour (Walton Centre, Liverpool), N Gurusinghe,
M Buchfelder, F Swozil (Universitatsklinikum Erlangen, Erlangen);
N Kumarasinghe (Lancashire Teaching Hospital NHS Trust, Preston),
A Kleindienst, M Marin, M Megele (Hospital Klinikum Amberg, Amberg),
R A Salman, I Whittle (Western General Hospital, Edinburgh);
D Hänggi, K Beseoglus (Medical Faculty, Heinrich-Heine-University,
USA I Kureshi, L Hosig (Hartford Hospital, Hartford), M Weaver,
Düsseldorf); K Kiening, B Orakcioglu, P Schiebel (University Hospital
F Sultan, D Laske, P Connolly (Temple University Hospital, Philadephia),
Heidelberg, Heidelberg), M Holling, H Wassmann (University Hospital
G Zipfel (Washington University School of Medicine, St Louis), J German
Münster, Münster), K Schwerdtfeger, J Szczygielski (Saarland University
(Albany Medical Centre, Albany), M Schneck, C Loftus (Loyola University
Medical Centre, Homburg-Saar), G Nowak, S Spuck (University
Hospital, Chicago), K M Cockroft (Penn State Hershey Medical Centre,
Schleswig-Holstein, Lübeck), K Sadowy (Neurochirurgische Klinik,
Hershey), B H Eidelman, J F Meschia (Mayo Clinic, Jacksonville),
Dessau), H W S Schroder, C Müller (Greifswald University, Greifswald),
S Amin-Hanjani, K Slavin (University of Illinois Hospital and Health
G F Hamann, R Schönmayr (Dr Horst Schmidt Kliniken, Wiesbaden),
Sciences System, Chicago).
E Juettler, J Woitzik H Neugebauer (Charite—University Medicine Berlin,
Berlin), A Waschke, R Kalff (Universitatsklinikum Jena, Jena), Y Chehade Conflicts of interest
(Asklepios Klinik Altona, Hamburg); Greece P Tsitsopoulos (Ippokration ADM is a director of Newcastle Neurosurgical Foundation, which is a
General Hospital, Thessaloniki), G Stranjalis, L Stavrinou (Evangelismos non-profit organisation for academic research and education, and he is
Hospital, Athens); Hungary A Buki, L Szapary Pecs (University Hospital, an adviser to Stryker (craniofacial surgery committee). BAG and ENR
Pecs), J Dobai (Borsod County and University Teaching Hospital, Borsod); received salary support from the STICH II grant but not from any

[Link] Vol 382 August 3, 2013 407

 Articles

commercial organisations. The other authors declare that they have no 18 Teernstra O, Evers S, Lodder J, Leffers P, Franke C, Blaaw G.
conflicts of interest. All surgeons in fee-for-service health-care systems Stereotactic treatment of intracerebral hematoma by means of a
receive additional fees for undertaking surgery. plasminogen activator: a multicenter randomized controlled trial
(SICHPA). Stroke 2003; 34: 968–74.
Acknowledgments 19 Chen X, Yang H, Cheng Z. A prospective randomised trial of
STICH II was funded by the UK Medical Research Council (MRC; grant surgical and conservative treatment of hypertensive intracerebral
number G0501444), now managed by the National Institutes of Health haemorrhage. Acta Acad Shanghai Med 1992; 19: 237–40.
Research (NIHR; grant number 09-800-18). This report is independent 20 Mendelow AD, Gregson BA, Fernandes HM, et al, for the STICH
research funded by the MRC and managed by the NIHR on behalf of the investigators. Early surgery versus initial conservative treatment in
MRC and NIHR partnership. The views expressed in this report are patients with spontaneous supratentorial intracerebral haematomas
those of the authors and not necessarily those of the MRC, National in the International Surgical Trial in Intracerebral Haemorrhage
Health Service, NIHR, or the Department of Health. (STICH): a randomised trial. Lancet 2005; 365: 387–97.
21 Mendelow AD, Gregson BA, Mitchell PM, et al. Surgical Trial in
References
Lobar Intracerebral Haemorrhage (STICH II) Protocol. Trials 2011;
1 van Asch CJJ, Luitse MJA, Rinkel GJE, van der Tweel I, Algra A, 12: 124.
Klijn CJM. Incidence, case fatality, and functional outcome of
intracerebral haemorrhage over time, according to age, sex, and 22 Wilson JT, Edwards P, Fiddes H, Stewart E, Teasdale GM. Reliability
ethnic origin: a systematic review and meta-analysis. Lancet Neurol of postal questionnaires for the Glasgow Outcome Scale.
2010; 9: 167–76. J Neurotrauma 2002; 19: 999–1005.
2 Feigin VL, Lawes CMM, Bennett DA, Anderson CS. Stroke 23 Murray GD, Barer D, Choi S, et al. Design and analysis of phase III
epidemiology: a review of population-based studies of incidence, trials with ordered outcome scales: the concept of the sliding
prevalence, and case-fatality in the late 20th century. Lancet Neurol dichotomy. J Neurotrauma 2005; 22: 511–17.
2003; 2: 43–53. 24 Gregson BA, Murray GD, Mitchell PM, Rowan EN, Gholkar AR,
3 Siddique MS, Fernandes HM, Arene NU, Wooldridge TD, Mendelow AD. Update on the Surgical Trial in Lobar Intracerebral
Fenwick JD, Mendelow AD. Changes in cerebral blood flow as Haemorrhage (STICH II): statistical analysis plan. Trials 2012; 13: 222.
measured by HMPAO SPECT in patients following spontaneous 25 Broderick JP, Brott TG, Grotta JC. Intracerebral hemorrhage
intracerebral haemorrhage. Acta Neurochir Suppl 2000; volume measurement. Stroke 1994; 25: 1081.
76: 517–20. 26 Fairbank J, Frost H, Wilson-MacDonald J, Yu LM, Barker K,
4 Nehls DG, Mendelow DA, Graham DI, Teasdale GM. Experimental Collins R. Randomised controlled trial to compare surgical
intracerebral hemorrhage: early removal of a spontaneous mass stabilisation of the lumbar spine with an intensive rehabilitation
lesion improves late outcome. Neurosurgery 1990; 27: 674–82. programme for patients with chronic low back pain: the MRC spine
5 Mendelow AD, Bullock R, Teasdale GM, Graham DI, McCulloch J. stabilisation trial. BMJ 2005; 330: 1233.
Intracranial haemorrhage induced at arterial pressure in the rat: 27 Weinstein JN, Tosteson TD, Lurie JD, et al. Surgical vs nonoperative
part 2. Short term changes in local cerebral blood flow measured by treatment for lumbar disk herniation: the Spine Patient Outcomes
autoradiography. Neurol Res 1984; 6: 189–93. Research Trial (SPORT): a randomized trial. JAMA 2006; 296: 2441–50.
6 Mendelow AD. Mechanisms of ischaemic brain damage with 28 Fung C, Murek M, Z’Graggen WJ, et al. Decompressive
intracerebral haemorrhage. Stroke 1993; 24 (suppl I): I115–I7. hemicraniectomy in patients with supratentorial intracerebral
7 Xi G, Keep RF, Hoff JT. Mechanisms of brain injury after hemorrhage. Stroke 2012; 43: 3207–11.
intracerebral haemorrhage. Lancet Neurol 2006; 5: 53–63. 29 Hanley DF. MISTIE III: Minimally Invasive Surgery plus rTPA for
8 Keep RF, Xi G, Hua Y, Hoff JT. The deleterious or beneficial ICH Evacuation Phase III. 2013. [Link]
effects of different agents in intracerebral hemorrhage: think big, mistie-iii-about (accessed May 17, 2013).
think small, or is hematoma size important? Stroke 2005; 30 Hanley DF. CLEAR-III Clot lysis: Evaluating Accelerated Resolution
36: 1594–96. of Intraventricular Hemorrhage Phase III. 2013. http://
9 Xi G, Wagner KR, Keep RF, et al. Role of blood clot formation on [Link]/clear-about (accessed April 12, 2013).
early edema development after experimental intracerebral 31 Naff N, Williams MA, Keyl PM, et al. Low-dose recombinant
hemorrhage. Stroke 1998; 29: 2580–86. tissue-type plasminogen activator enhances clot resolution in brain
10 Bhattathiri PS, Gregson B, Prasad KS, Mendelow AD, STICH hemorrhage: the Intraventricular Hemorrhage Thrombolysis Trial.
Investigators. Intraventricular hemorrhage and hydrocephalus after Stroke 2011; 42: 3009–16.
spontaneous intracerebral hemorrhage: results from the STICH 32 Hosseini H, Leguerinel C, Hariz M, et al. Stereotactic aspiration of
trial. Acta Neurochir Suppl 2006; 96: 65–68. deep intracerebral hematomas under computed tomographic
11 Auer LM, Deinsberger W, Niederkorn K, et al. Endoscopic surgery control: a multicentric prospective randomised trial. Cerebrovas Dis
versus medical treatment for spontaneous intracerebral hematoma: 2003; 16S: 57.
a randomized study. J Neurosurg 1989; 70: 530–35. 33 Hattori N, Katayama Y, Maya Y, Gatherer A. Impact of stereotactic
12 Batjer HH, Reisch JS, Allen BC, Plaizier LJ, Su CJ. Failure of hematoma evacuation on activities of daily living during the chronic
surgery to improve outcome in hypertensive putaminal period following spontaneous putaminal hemorrhage:
hemorrhage. A prospective randomized trial. Arch Neurol 1990; a randomised study. J Neurosurg 2004; 101: 417–20.
47: 1103–06. 34 Pantazis G, Tsitsopoulos P, Mihas C, et al. Early surgical treatment
13 Chen XC, Wu JS, Zhou XP, et al. The randomized multicentric vs conservative management for spontaneous supratentorial
prospective controlled trial in the standardized treatment of intracerebral hematomas: a prospective randomized study.
hypertensive intracerebral hematomas: the comparison of surgical Surg Neurol 2006; 66: 492–501.
therapeutic outcomes with conservative therapy. 35 Wang WZ, Jiang B, Liu HM, et al. Minimally invasive
Chin J Clin Neurosci 2001; 9: 365–68. craniopuncture therapy vs conservative treatment for spontaneous
14 Juvela S, Heiskanen O, Poranen A, et al. The treatment of spontaneous intracerebral hemorrhage: results from a randomized clinical trial
intracerebral hemorrhage. A prospective randomized trial of surgical in China. Int J Stroke 2009; 4: 11–16.
and conservative treatment. J Neurosurg 1989; 70: 755–58. 36 Prasad K, Mendelow AD, Gregson B. Surgery for primary
15 McKissock W, Richardson A, Taylor J. Primary Intracerebral supratentorial intracerebral haemorrhage.
haematoma: a controlled trial of surgical and conservative treatment Cochrane Database Syst Rev 2008; 4: CD000200.
in 180 unselected cases. Lancet 1961; 278: 221–26. 37 Mendelow A, Gregson B. Surgery for intracerebral hemorrhage.
16 Morgenstern LB, Frankowski RF, Shedden P, Pasteur W, In: Mohr J, Wolf P, Grotta JC, Moskowitz M, Mayberg M,
Grotta JC. Surgical treatment for intracerebral hemorrhage von Kummer R, eds. Stroke: pathophysiology, diagnosis and
(STICH): a single-center, randomized clinical trial. Neurology management. 5th edn. Philadelphia: WB Saunders, 2011.
1998; 51: 1359–63. 38 Gregson BA, Broderick JP, Auer LM, et al. Individual patient data
17 Zuccarello M, Brott T, Derex L, et al. Early surgical treatment for subgroup meta-analysis of surgery for spontaneous supratentorial
supratentorial intracerebral hemorrhage: a randomized feasibility intracerebral hemorrhage. Stroke 2012; 43: 1496–504.
study. Stroke 1999; 30: 1833–29.

408 [Link] Vol 382 August 3, 2013