Clinical condition of 120 patients alive at 3 years after poor-grade aneurysmal subarachnoid hemorrhage

KUH, one of the five university hospitals in Finland, is an academic, non-profit, publicly funded tertiary center, which serves a defined population (about 842,000 in 2015) in Eastern Finland. The KUH area contains four central hospitals, each with 24/7 neuroacutology, CT services, intensive care, neurology, and neurorehabilitation. The KUH area is served by ambulance and (since 2002) by helicopter transfer. At KUH Neurosurgery, at least two neurosurgeons are on duty at all times. During the study period from 2005 to 2015, all cases of SAH were acutely transferred to KUH for neurointensive care, neuroradiology (4-vessel angiography and/or CT angiography), and neurosurgery [21]. Neurointensive care was provided to virtually all patients regardless of the age or condition on admission, including H&H 4–5 patients [21]. A dedicated team of neurointensivists, neurosurgeons, and interventional neuroradiologists coordinated the aSAH treatment. The KUH Neurovascular Group provided microsurgical or endovascular occlusion of the ruptured aneurysm and evacuated significant ICHs with immediate microsurgery. The KUH aSAH neurointensive care protocol in 2005–2015 followed international recommendations in detail [9, 27, 29, 44, 46], and it was presented in our previous study of organ donors after aSAH [21, 45]. Briefly, the protocol aimed to prevent further brain damage due to rebleeding, increased intracranial pressure (ICP), hydrocephalus, electrolyte disturbances, seizures, cardiac and pulmonary dysfunction, fever, hyperglycemia, and development of delayed brain ischemia. The protocol included, when appropriate, e.g., external ventricular drainage (EVD), parenchymal ICP monitoring, endovascular procedures, and intra-arterial nimodipine infusion in case of delayed brain ischemia, as well as decompressive craniectomy (DC).

The database, prospective since 1995, contains all cases of unruptured and ruptured IAs admitted to KUH since 1980. A dedicated, full-time nurse administrates the database, interviews all new IA patients, including their family history, and arranges the follow-ups. The clinical data, including prescribed medicines, hospital diagnosis, and causes of death, have been fused from the national registries, using the Finnish personal codes. We have characterized the aSAH patients, e.g., for the 14-day mortality and organ donation [21], 12-month [22] and long-term excess mortality [17], shunt-dependent hydrocephalus [1], depression [19], epilepsy [18], pain [33], anti-psychotics [39], diabetes mellitus [31], hypertension [25, 32], as well as, polycystic kidney disease [36]. Three first-degree relatives with a diagnosed sIA disease form an sIA family.

A total of 769 consecutive aSAH patients were acutely admitted to the KUH Neurointensive Care Unit from 2005 to 2015 (Fig. 1). The six patients lost to follow-up were excluded. A total of 269 (35%) aSAH patients had H&H 4 (n=145) or H&H 5 (n=124; extension to pain) on arrival or before intubation before arrival to the KUH (Fig. 1). Their clinical lifelines were re-constructed from their clinical data in the Kuopio database and from the national clinical registries until death (n=149) or July 2019 (Table 1). The patients who deteriorated from H&H 1–3 on admission to H&H 4–5 during the neurointensive care were excluded.

A total of 120 H&H 4 (n=88) or H&H 5 (n=32) patients were alive at 3 years after admission (Tables 1 and 2). Their 120 primary digital CT scans were analyzed for the presence or absence of ICH, IVH blood clot, IVH sediment, and hydrocephalus (Table 2). A representative CT slice of each patient was compiled together and sorted according to the modified Rankin Scale (mRS 0 to 5) at 3 years. The ICH volumes were calculated from the CT scans using the formula π × (a × b × c) / 6 where a, b, and c are the three perpendicular diameters of ICH. All available clinical data were obtained from the KUH electronic health care records, purchases of prescripted drugs, and all hospital and primary health care diagnoses from the national databases. By our definition, the use of anti-epileptic, anti-depressive, or anti-psychotic medications excluded mRS 0.

Firstly, PubMed was searched in November 2020 for English articles on human aSAH published between 2010 and 2020 with the words (Aneurysm* AND “subarachnoid hemorrhage”) AND ((poor-grade) OR (poor grade)) AND (long-term AND (mortality OR survival OR outcome OR prognosis)) AND (12 months OR twelve-month). This gave 14 hits.

Secondly, we identified the original clinical aSAH patient cohorts, excluding case reports, duplicate publications, systematic reviews, and meta-analyses.

Thirdly, we approved the aSAH cohorts fulfilling the following criteria:

Finally, none of the articles fulfilled the criteria.

The categorical variables were expressed in proportions, and the χ2 test was used in comparisons. The continuous variables were expressed in medians, quartiles, and ranges, and the non-parametric tests were used in comparisons. The Kaplan-Meier analysis was used to calculate the cumulative mortality rates, and the log-rank test was used to test for differences between groups. Independent risk factors for the clinical condition at 3 years of the 120 patients were searched with logistic regression analysis. P values < 0.05 were considered significant. We used the SPSS 27 statistical software (SPSS, Inc., Chicago, IL).

A total of 269 consecutive aSAH patients were in poor condition (H&H 4, n=145; H&H 5, n=124) on admission to the KUH Neurointensive Care Unit (flowchart in Fig. 1, Table 1). Table 1 presents their clinical characteristics on admission, as well as those alive at 12 months and 3 years. Of the 269 patients, 155 (58%) were alive at 14 days, 125 (46%) at 12 months, and 120 (45%) at 3 years, significantly more often the H&H 4 patients (Fig. 2). Decompressive craniectomy (DC) was performed in 35 (13%) patients in a median of 2 (IQR 1–4) days; 13 (37%) of them died within 14 days (Table 1).

At 3 years, 120 (45%) patients were alive, 88 (73%) after H&H 4 and 32 (27%) after H&H 5 on admission (Table 1). Their clinical condition at 3 years is distributed as follows (Table 2, Fig. 3): 14% mRS 0; 31% mRS 1; 17% mRS 2; 13% mRS 3; 19% mRS 4; and 6% mRS 5 (nursing home or hospital). By 3 years, 33 (28%) patients had started new anti-epileptic (AE) medication, 33 (28%) new anti-depressive (AD) medication, and 16 (13%) new anti-psychotic (AP) medication (Table 2). A total of 55 (46%) patients carried a ventriculoperitoneal shunt (Tables 1 and 2, Fig. 3).

Figure 3 presents one representative CT slice of each patient, together with the surgical interventions (see below), arranged according to the patient’s modified Rankin Scale (mRS) from 0 to 5 at 3 years. There were 71 (59%) intracerebral hematomas (ICH) with a median volume of 22 cm³ (IQR 11–56), 47 (66%) from the middle cerebral artery (MCA) sIA, and 12 (17%) from the anterior communicating artery (AComA) sIA. There were 31 (26%) intraventricular blood clots (IVH), 15 (48%) from the AComA sIA, and 10 (32%) from the MCA sIA. Mere intraventricular blood sediments were seen in 59 (49%). Acute hydrocephalus was present in 76 (63%) patients.

Extraventricular drainage (EVD) was started in 105 patients (88%). Table 3 cross-tabulates the 12 combinations of microsurgical evacuation of ICH (n=29), microsurgical clipping (n=55) or endovascular occlusion (n=64) of ruptured sIA, and decompressive craniectomy (n=16). In each of the 12 patient groups, median mRS at 3 years is presented. Figure 3 shows the representative CT slices for each of the 12 patient groups. Intraventricular blood clots, present in 31 (26%) patients (Fig. 3), were not evacuated microsurgically or endoscopically; one patient had intraventricular alteplase (t-PA) thrombolysis.

A total of 55 (46%) patients had a ventriculoperitoneal shunt (Fig. 3), inserted in a median of 17 days (IQRs 9 and 64) after aSAH. In the 55 shunted patients, the ventricles on admission contained IVH clot in 18 (33%) (Fig. 3), mere blood sedimentation in 34 (62%), and no blood in 3 (5%). Until 3 years, 32 (58%) patients had no shunt revisions (median mRS 1). A total of 23 (42%) patients had shunt revisions (median mRS 3), 11 once, and 12 two to five times. At 3 years, the 65 non-shunted patients had a median mRS of 2.

DC was performed in 16 (13%) patients (Table 1 and 3, Fig. 3) in a median of 2 (IQR 0–5) days after admission. In 14 of the 16 DC patients, cranioplasty was performed in a median of 94 (IQR 46–127) days after DC, always with the own frozen bone flap; 4 flaps were later replaced with an artificial flap.

At 3 years, 54 patients (median 55 years) had mRS 0 or 1, 45% of the 120 three-year survivors but only 20% of all the 269 original aSAH patients with H&H 4–5 on admission (Table 2). As many as 17 (31%) patients had no symptoms (mRS 0; no AE, AD, or AP drug use) despite minor ICH in five, IVH blood clot in five, or shunt in six (Table 2, Fig. 3).

At 3 years, 30 patients (median 57 years) had mRS 4 or 5, 25% of the 120 three-year survivors but only 11% of the 269 original H&H 4–5 aSAH patients (Table 2). As many as 24 (80%) patients had ICH (14 evacuated) while only 6 had no ICH at all, 7 (23%) had an IVH clot, and 15 (50%) had a shunt (Tables 2 and 3, Fig. 3). In multivariate testing of available data, the ICH volume stood out as a significant predictor of mRS at 3 years.

After acute aSAH, “poor condition” or “poor grade” on admission to the first hospital for CT and after transfer to the neurointensive care predicts early mortality so high that aSAH patients brain dead within 14 days are a significant group of organ donors [21, 37, 45]. The acute effect may be so devastating that some aSAH patients do not reach alive the first hospital and diagnostic CT [30]. The long-term mortality, often presented at 12 months [47], is also high: 37% in H&H 4 and 73% in H&H 5 in our present study (Fig. 2). A minority of H&H 4–5 or WFNS 4–5 patients will become true long-term survivors, in the present study 120 (45%) patients alive at 3 years (Table 2), albeit being stroke-risk carriers and prone to further vascular events [16]. They have remained an unrecognized group among stroke survivors for obvious reasons: prospective databases with long recruitment and outpatient follow-up periods would be virtually impossible to maintain, while much of literature is focused on “aneurysm treatment” [43]. We have retrospectively analyzed their clinical condition during very long follow-ups: excess mortality (median 12 years) [17]; epilepsy (median 6 years) [18]; anti-depressants (median 9 years) [19]; anti-psychotics (median 9 years) [39]; and shunt dependency (median 8 years) [1].

In the present study, we constructed the clinical date point calendaric lifelines for the 269 consecutive aSAH patients in poor condition (54% with H&H 4 and 46% with H&H 5) on admission to the Neurointensive Care Unit of the Kuopio University Hospital (KUH) from the defined KUH catchment population from 2005 to 2015 (Fig. 1, Table 1). For the lifeline analysis, each patients’ medical data, using the Finnish personal identification code, was obtained from the nationwide registries, and fused into the Kuopio Intracranial Aneurysm Patient and Family Database. The cross-sectional study population consisted the 120 consecutive patients alive at 3 years, 73% with H&H 4 and 27% with H&H 5 on admission (Table 1, Table 2). We found it impossible, using sentences and risk factors only, to present a comprehensive view of the associations of multiple intertwining factors during the neurointensive care vs. the clinical conditions at 3 years. For the same reason, the patients who deteriorated from H&H 1–3 to H&H 4–5 during the neurointensive care were excluded. For visual estimation and browsing, one representative slice of each of the 120 primary CT scans (ICH; IVH; hydrocephalus) and the interventions (extraventricular drainage; sIA occlusion; ICH evacuation; decompressive craniectomy; shunt) were compiled together, according to the mRS (0 to 5) at 3 years (Fig. 3) (see also 45 asymptomatic meningiomas by Näslund et al. in 2020 in Acta Neurochirurgica [35]). Furthermore, the interventions and the presence or absence of ICH were cross-tabulated into 12 groups (Table 3).

Of the 120 H&H 4 or 5 patients alive at 3 years, 71 (59%) had an intracerebral hematoma (ICH) with a median volume of 22 cm³, mostly (47; 66%) from a middle cerebral artery (MCA) sIA, and (12; 17%) from an anterior communicating artery (AComA) sIA. Aneurysmal ICH, depending on its volume and location, is the single most damaging and deadly complication of acute aSAH [51]. In the present study, ICH volume was a significant predictor of mRS at 3 years (Table 2). Arterial blood yet from the ruptured aneurysm wall tears and enters the adjacent brain tissue, forming a permanent brain cavity, filled with ICH clot. MRI (diffusion/perfusion/tractography) may give further prognostic information on brain edema [42], early brain ischemia [11, 14], and injury to white matter tracts and nuclei, e.g., in relation to motor function [4] or impaired consciousness [20]. The local injury of the brain connectome is permanent; i.e., no axonal connections will redevelop across the ICH cavity that later in neuroimaging appears as a CSF cavity. Convalescence, e.g., from acute hemiparesis or aphasic disorder, depends on the functional re-organization of the connectome around the cavity.

In our series, 71 (59%) of the 120 patients presented with ICH, and 29 ICHs (median volume 56 cm³) were microsurgically evacuated. Acute evacuation relieves expansion and pressure against the cavity wall, but the extent of brain tear remains unchanged. Only gentle subtotal evacuation is realistic as aneurysmal ICHs are rooted locally among cortical and perforating arteries [49]. Nevertheless, the ICH clot inducing neuroinflammation becomes smaller [53]. However, there is not yet concrete evidence on long-term benefits of acute aneurysmal ICH evacuation [14]. Notably, in spontaneous ICHs and IVHs, there is increasing interest in acute endoscopic and stereotactic evacuation [3, 34].

In our series, DC was performed in 35 (13%) of the 269 aSAH patients with H&H 4 or 5 on admission: 16 of them (46%) were among the 120 three-year survivors. After this series, we abandoned the frozen bone flaps for artificial ones, because 19% of own bone flaps were removed due to complications in a Finnish study [24]. In a systematic review of 407 aSAH patients with WFNS 4 or 5, the effect of DC on functional outcome remained unknown because of the lack of robust control groups [2].

Aneurysmal IVHs of various volumes are frequent in acute aSAH patients [38, 44]. In our basic series, IVHs ranged from mere blood sediments (38%) in the occipital horns to clots (34%), some casting one or both lateral ventricles. IVH clots may co-exist with adjacent ICH, in 64 (24%) cases in our series, adding instant brain tissue injury to multiple potentially harmful effects of IVH. With pathobiology poorly understood, they include neuroinflammation [10]; alteration in ependymal cells, ciliary beat, arachnoid villi, CSF production and resorption, and glymphatic circulation; and clinically manifest hydrocephalus (acute; subacute; latent). IVH with enlarged ventricles often require prolonged EVD with ICP and CSF outflow monitoring, with the risk of catheter occlusions and exchanges, and eventual bacterial meningitis [1]. In principle, it would make sense to reduce clots and casts in the lateral ventricles, even III and IV, with neuroendoscopy as soon as feasible. In spontaneous IVH, recent meta-analysis suggested that endoscopy improved the survival and prognosis with the lowest rate of shunts and infection [34].

Secondary hydrocephalus is frequent among aSAH survivors, and several risk scores have been published [1]. The pathobiology is poorly understood but neuroinflammation is a candidate [10], activating in the acute phase and possibly exerting a long-term dysfunction in the CSF environment, with tendency of valves and catheters to occlude, as well as shunt infection. EVD is an instant indicator of shunt dependency, and concomitant bacterial meningitis or ventriculitis increases the risk [1]. Surprisingly, in our series and other cohorts, even normal ventricle size and sediments of IVH only could be followed by latent shunt dependency [1, 8, 52]. The long-term outcome and quality of life of shunted aSAH survivors is virtually unknown according to our literature review. Long-term follow-up studies will show whether shunt-dependent post-aSAH hydrocephalus is a degenerative brain disease, such as idiopathic normal pressure hydrocephalus (iNPH) after temporary improvement with a shunt [23].

Our pilot study, aimed to analyze a significant group of long-term aSAH survivors, did not include enough data (e.g., previous conditions of patients; ICP monitoring; EVD outflow; acute or delayed ischemic brain injury; electrolyte disturbances; cardiopulmonary complications; CNS or systemic infections; and complications of management) for computerized (e.g., machine learning) prediction on admission or during neurointensive care for subsequent mortality and outcome. Still, some aspects concerning triage on admission can be extracted from the 120 H&H 4–5 survivors at 3 years, as visualized in Fig. 3 and Table 3.

Firstly, in 49 patients (41%) of 120 patients with H&H 4 or 5 on admission, the primary CT scan did not show immediate brain injury by ICH. Their mRS range at 3 years from 0 to 5 indicates that coma and extension to pain are unreliable triage predictors of dismal outcome [28], interfered, e.g., by seizures, hydrocephalus, brain herniation, sedation, intubation, and ventilator care. Acute MRI would add information on possible acute ischemic brain injury [14].

Secondly, large ICHs in eventual mRS 3 to 5 survivors (maximum 200 cm³ in our series) may question chances of survival at the initial triage for neurointensive care. Large ICHs raise the question whether lives saved with neurointensive care are “worth living” with mRS 4 to 5 in the long run, i.e., whether such logistics would be ethically acceptable in neuroacutology. With the opt-out system for brain death and organ donation in Finland since 2010, we also admit “hopeless” aSAH patients, provided that the National Transplantation Service does not by phone exclude the possibility of organ donation due to concomitant diseases or conditions [21, 45]. The donors of kidneys included brain dead aSAH patients even at their 80’s. In Finland, the annual costs of dialysis (about 40,000 euros) far exceed the costs after the first year of a kidney transplant patient [15]. Eventually, evaluating CT scans in Fig. 3, the lives saved with large ICH and dismal (“unacceptable”) long-term condition (mRS 4 or 5) were few: 23 (9%) of the 269 H&H 4 or 5 patients on admission (Fig. 3).

Thirdly, it was surprising, e.g., how the six mRS 0 or 1 patients (median 58 years) with ICH over 27 cm³ (the largest quartile) managed to recover so well at 3 years (Tables 1 and 2, Fig. 3). This unexpected resilience to brain injury requires further investigations.

The strengths derive from the free tax-paid Finnish health care system, as well as the automatic archival of clinical data, using the Finnish identity codes, in the national clinical registries. Finland is divided into exclusive catchment areas among the 5 university hospitals which allows the creation of disease cohorts that are minimally selected and biased. The Kuopio Intracranial Aneurysm Patient and Family Database reliably reflects aSAH in the Eastern Finnish population and allows reconstruction of the clinical date point lifelines of all diagnosed aSAH patients, also using data from the national clinical registries. Our study is retrospective although the database prospectively collected all aSAH patients through the study period.

There are also limitations. The evaluation of the clinical condition of the 120 three-year survivors was based on the available case reports and the data available in the national registries. In this pilot study, the patients who deteriorated from H&H 1–3 on admission to H&H 4–5 later were excluded from the final study population. Finally, the extensive digital neurointensive care monitoring data, available in the national database, was not used [41]; such data are indispensable in further studies with machine learning algorithms.

Firstly, retrospective re-reconstruction of clinical lifelines for individual aSAH patients, or any neuroacutology patients, is arduous patchwork at present, for clinical research or quality control. The IT systems of the hospital catchment area should automatically provide individual digital timelines (in minute scale) from the first contact through all phases to, e.g., the neurointensive care. Such quality control practice would aid to detect and prevent delays and deviations in aSAH logistics that might risk the final outcome, including time of intubation or release of tentorial herniation.

Secondly, the long-term course of consequences of acute aSAH on the brain has not been adequately studied. Long-term understanding would require consecutive MRI scans (diffusion/perfusion and tractography), at acute phase (acute hemorrhagic injuries, severing of tracts, hydrocephalus, and ischemia), at subacute and late subacute phases, at 12 months (permanent hemorrhagic and ischemic injuries), and after a few years (neurodegeneration).

Thirdly, neuroinflammation might function as an umbrella for cellular and molecular biology research of several calamities of the aSAH brain: early brain ischemia; ICH; IVH; ischemic events during occlusive IA therapy; delayed ischemia; shunt-dependent hydrocephalus—also postulated neurodegeneration in the post-aSAH brain [5].