From October 2017 to February 2024, we collected data on all aSAH patients admitted to our institution, including age, sex, medical history (hypertension, diabetes, and smoking), aneurysm location, treatment modality, and the severity of subarachnoid hemorrhage (measured by modified Fischer [mFS] grade and Hunt-Hess grade). Patients who had persistent CVS despite treatment following our management algorithm for DCI were defined as having medically refractory CVS. Those who responded well to IA nimodipine were classified into the conventional IA nimodipine group, while those who did not respond and underwent CIAN were allocated into the CIAN group. We recorded the symptoms, timing, and duration of rescue treatments, as well as any complications, infarctions, and the modified Rankin scale (mRS) scores at discharge. Cerebral infarctions associated with vasospasm were categorized as minor (<1/3 of a vessel territory) and major (≥1/3 of a vessel territory). An mRS score of 2 or less was considered a favorable outcome.

All patients who presented with aSAH underwent initial computed tomography angiography (CTA) and digital subtraction angiography (DSA). The ruptured aneurysm was obliterated as soon as possible by either surgical or endovascular treatment. The choice of treatment modality was determined by aneurysm-related factors such as location, size, and shape, as well as patient-specific factors including age, medical history, and initial mental status. If hydrocephalus or brain swelling were observed on the initial cranial CT, an external ventricular drainage (EVD) was performed, and a lumbar drain was inserted in all patients. In the neurocritical care unit (NCU), invasive blood pressure monitoring was used to maintain normotension (systolic pressure between 100-140 mmHg), and for patients with continuous intracranial pressure (ICP) monitoring via EVD, cerebral perfusion pressure (CPP) was maintained above 60 mmHg. During the initial fasting period, nimodipine (Moditop^® injection, 10 mg/50 mL; Samjin Pharm, Seoul, Korea) was administered at a rate of 5-10 mL/hr. Once blood pressure stabilized, and oral or enteral nutrition was possible, oral nimodipine was given at a dose of 60 mg every 4-8 hours. Daily TCD was performed to detect delayed CVS. If clinical deterioration occurred or TCD suggested severe vasospasm (middle cerebral artery [MCA] >150 cm/sec, anterior cerebral artery [ACA] >120 cm/sec), perfusion CTA (PCTA) or diffusion-weighted magnetic resonance imaging (DWI) was performed. If vessel irregularity or narrowing of more than 50% compared to baseline, impairment of hemodynamic parameters (cerebral blood flow and volume, mean transit time, and time to peak), or diffusion restriction was detected, systolic blood pressure was elevated to 140-180 mmHg through induced hypertension. If medical treatment was ineffective for several hours, we defined this condition as refractory CVS. In such cases, DSA was performed to confirm the presence of CVS, indicating the need for IA nimodipine infusion.

For single-session IA nimodipine infusion, a microcatheter (Excelsior^® SL-10; Stryker, Fremont, CA, USA) tip was placed in the parent artery of the most spastic vessel. However, if a perfusion defect was present without a significantly spastic vessel, the catheter tip was positioned in the distal ICA. Nimodipine was diluted with normal saline at a 1 : 1 ratio to achieve a concentration of 10 mg/100 mL. The total dosage was adjusted based on symptoms and angiographic findings, with a maximum dose of 9 mg. If symptoms or angiographic CVS persisted, CIAN was initiated.

Vascular access was obtained via the common femoral artery using a 6 F introducer sheath, and the distal tip of the microcatheter was positioned through an Envoy DA guiding catheter (Cerenovus, Fremont, CA, USA) (Fig. 1). To prevent dislodgement and infection, the system was covered with an occlusive dressing using an iodine-impregnated adhesive surgical drape (Ioban; 3M, St. Paul, MN, USA). The administration rate of nimodipine, with the same concentration as described, was adjusted between 0.4 and 2 mg/hr. Thromboembolic complications were mitigated by administering heparin (3 IU/mL) into the guiding catheter at a rate of 60 IU/hr. Additionally, the activated clotting time (ACT) was measured every 4 hours, with a target range of 1.5-2 times the initial ACT. This was achieved using intravenous heparin administered as a bolus of 1000-2000 IU.

To compare the characteristics of patients with and without medically refractory CVS, the Mann-Whitney U-test, Fisher’s exact test, and Cochran-Armitage trend test were applied. The same three tests were used to compare the characteristics and clinical outcomes (TCD abnormal days, infarction, length of stay [LOS] in the NCU, and LOS in the hospital) between the CIAN and the conventional IA nimodipine groups. Logistic regression analysis was additionally performed to assess the impact of CIAN implementation on mRS scores. Following univariate analysis, a stepwise reduced model was created, and multivariate analysis was conducted using the variables that remained. All statistical analyses were performed using R software (version 4.4.1; R Foundation for Statistical Computing, Vienna, Austria), with statistical significance set at p<0.05 and a 95% confidence interval.

Out of the total cohort of 274 patients, a subset of 15 patients (5.5%) required IA nimodipine infusion as a rescue treatment for medically refractory CVS. Additionally, five patients (1.8%) were treated with CIAN for angiographically refractory CVS. The mean age was 55.9 years in patients with medically refractory CVS and 60.2 years in those without medically refractory CVS. In the group of patients with medically refractory CVS, more females (87%, 13/15) were observed compared to the group without medically refractory CVS (66%, 171/259). No significant differences were observed in other factors such as hypertension, diabetes mellitus, or smoking. Patients with medically refractory CVS had a higher severity of initial aSAH (mFS ≥3 : 100%, 15/15) compared to those without medically refractory CVS (mFS ≥3 : 83%, 215/259) (Table 1).

All patients who underwent CIAN had an mFS grade of 3 or higher, indicating a significant SAH. CIAN was initiated between 8-10 days after the onset of aSAH and lasted for 21-81 hours (median duration, 53 hours).

Two complications were observed in patients receiving CIAN. In one case (case 1 in Table 2), severe brain edema developed. CIAN was initiated after the diagnosis of a major infarction but was discontinued to perform a decompressive craniectomy. In another case (case 5 in Table 2), suspected heparin-induced thrombocytopenia (HIT) developed after CIAN started. The platelet counts progressively decreased after each subsequent DSA. Following CIAN, the platelet count dropped below 50000, leading to the discontinuation of CIAN, after which the platelet count recovered.

In follow-up CT scans performed at the outpatient clinic, one patient was found to have a major infarction, while four patients had minor infarctions. However, the sizes of the infarctions showed no changes compared to pre-CIAN CT scans, indicating that no new lesions had developed after CIAN. Two patients continued vasopressor therapy without an increase in requirements, two patients maintained hemodynamic stability without the need for vasopressors throughout, and one patient demonstrated a reduction in vasopressor requirements. At discharge, one patient had an mRS score of 5, while four patients showed a favorable outcome with an mRS score of 2 or below.

A 28-year-old female patient presented with an aSAH due to a ruptured anterior choroidal artery aneurysm, with an mFS grade of 3. The aneurysm was treated surgically with clipping (Fig. 2A). During the early post-operative period, chemical prophylaxis with oral nimodipine was initiated, in addition to the maintenance of euvolemia and normotension.

On day 6 post-aSAH, the patient exhibited motor aphasia. DWI revealed a minor infarction in the left MCA territory, leading to the diagnosis of DCI (Fig. 2B). DSA showed irregularities in the left ICA, the M1 segment of the MCA, and the A1 segment of the ACA. A single-session IA nimodipine infusion of 9 mg was administered via the left cervical ICA. Following treatment, both the patient’s symptoms and the angiographic CVS improved. TCD performed the next day also showed improvement in the mean flow velocity (MFV) of the left MCA.

However, on day 9 post-aSAH, the patient developed right hemiparesis, with a muscle strength grade of 3. A follow-up DWI scan revealed an additional infarction in the border zone of the left MCA and posterior cerebral artery. Consequently, CIAN was initiated, with a microcatheter placed in the distal ICA. The patient’s symptoms improved after 34 hours of treatment. DSA and TCD performed 4 days later confirmed the resolution of CVS, allowing the removal of the catheter. The patient was discharged a month later with no neurological deficits.

There were five patients in the group who received CIAN due to the ineffectiveness of single-session IA nimodipine infusion, and 10 patients in the group treated only with conventional IA nimodipine. The mean age of the CIAN group was 46.2 years, which was lower than the 60.7 years in the conventional IA nimodipine group, although this difference was not statistically significant. The severity of aSAH (Hunt-Hess grade, mFS grade) showed no significant differences between the two groups. The time required for the MFV on TCD to recover tended to be longer in the CIAN group (CIAN : 16.0±10.1 days vs. conventional IA nimodipine : 9.4±7.9 days, p=0.188). Major infarctions occurred in 20% of patients in both groups, with no significant difference. Both NCU LOS and hospital LOS were longer in the CIAN group, but the differences were not statistically significant (NCU LOS : CIAN 17.4±10.1 days vs. conventional IA nimodipine 14.1±7.0 days, p=0.472; hospital LOS : CIAN 46.6±28.7 days vs. conventional IA nimodipine 29.5±13.2 days, p=0.262) (Table 3). At discharge, 80% of patients (4/5) in the CIAN group achieved favorable outcomes (mRS ≤2), compared to 70% (7/10) in the conventional IA nimodipine group (Fig. 3). In the logistic regression analysis for mRS, the reduced model identified Hunt-Hess grade and CIAN implementation as the remaining significant variables. A Hunt-Hess grade of 4 was associated with an odds ratio (OR) of 39.59 (p=0.001) for increased mRS, and a grade of 5 was associated with an OR of 93.29 (p=0.002), indicating that Hunt-Hess grade was the strongest predictive factor. CIAN implementation demonstrated a clinically significant trend, with an OR of 0.42 (p=0.073) for reduced mRS, suggesting its potential beneficial effect (Table 4).

After the diagnosis of DCI, DSA, the gold standard for the diagnosis of CVS, is performed to address potential CVS. In a study by Kim et al. [14], IA nimodipine infusion improved angiographic CVS in 42 out of 53 procedures and improved the GCS score in 23 out of 53 procedures, with a maximum dose of 5 mg. In a pharmacological study utilizing a two-compartment model [21], it was observed that IA nimodipine infusion could achieve a higher concentration in the cerebral arteries more rapidly with a lower infusion rate than that of intravenous or oral nimodipine. However, after discontinuation, the concentration showed an immediate decline. TBA, a mechanical dilation technique, involves stretching and dilating vasospastic arteries. The recurrence rate of CVS following the use of IA vasodilators is higher than that of TBA [5,10]. TBA is used for localized vasospasm of proximal vessels and has a success rate of 87% [7]. However, it has its drawbacks, including serious complications such as embolism, thrombosis, dissection, and vessel rupture, which can occur in up to 5% of patients. Additionally, it is challenging to perform in tortuous vessels [11].

In cases where angiographic CVS persisted or symptoms did not improve despite the administration of IA nimodipine, we proceeded with CIAN. To attain a higher intravascular concentration of nimodipine, CIAN was first described by Mayer et al. [16]. Following the initial report, many studies have elaborated upon the effects and complications of CIAN. The protocols varied across studies, particularly in terms of the use or nonuse of tirofiban, the administration route of heparin, and the rate of nimodipine, which ranged from 0.5-2.0 mg/hr [2,3,13,15,23]. In a study by Kramer et al. [15], the median onset of CIAN was 9 days after aSAH, with a median duration of 5 days. In another study [13], the mean onset was similar at 8 days, but the duration was longer, with a mean of 10.5 days. We adjusted the rate of CIAN administration between 0.4 and 2 mg/hr.

Prior to our study, analyses solely focused on the outcomes and complications of the CIAN patient group [13,15] or compared the CIAN group with a control group that did not receive any IA vasodilator [3]. In Bele et al. [3], when comparing the control group that used oral nimodipine and induced hypertension with the experimental group that additionally used CIAN, the experimental group showed a significantly higher proportion of favorable functional outcomes, 76% compared to 10%. In terms of radiologic outcomes, at the time of discharge, CT scans showed infarctions in 42.6% of the experimental group and 75% of the control group, demonstrating a significant difference. However, the comparison with a such a control group has limitations as it does not reflect the current trends where IA vasodilator is the fundamental treatment for medically refractory CVS. Other studies have compared CIAN with conventional single-session IA nimodipine; however, in one study, the initial status of the group that received CIAN was significantly worse, making it difficult to observe the treatment effects of CIAN [23]. Another study utilized a before-and-after design, which could not control for the various monitoring tools and medical treatments used in the management of aSAH [2]. To clearly evaluate the effectiveness of CIAN, comparing the CIAN group with a conventional IA nimodipine group that exhibits angiographically refractory CVS during the same period has merits, ensuring that the initial status does not significantly differ. Through strict surveillance, we aimed to quickly identify patients requiring DSA and to select those suitable for CIAN. This allowed us to compare patients who underwent CIAN with those who only received conventional IA nimodipine without significant differences in their initial statuses.

The persistence of CVS despite conventional IA nimodipine treatment is associated with a poor prognosis. In such cases, there are currently no effective rescue treatments. Our protocol, which implemented CIAN, has additional value as a rescue treatment because it can prevent further deterioration in the prognosis of patients exhibiting angiographic CVS after conventional IA nimodipine. Additionally, in poor-grade aSAH patients who are neurologically unresponsive, multimodal monitoring—such as partial brain tissue oxygen monitoring, microdialysis, and cerebral blood flow monitoring—can be important tools to assess whether CVS improvement persists or deteriorates after IA nimodipine infusion. However, many clinical settings may not allow for such monitoring. In these cases, reliance on serial, non-continuous tests such as TCD and PCTA is unavoidable. If CIAN is maintained during these intermittent assessments, treatment can be adjusted based on the TCD or PCTA results, potentially serving as a valuable strategy to minimize DCI in poor-grade aSAH patients where multimodal monitoring is not feasible.

Complications related to CIAN have included cardiac arrest, HIT, ventilator-associated pneumonia, catheter tip thrombosis, and ICA dissection [13,23]. However, it remains uncertain whether all of these were directly procedure-related complications. Moreover, Kapapa et al. [12] reported that complications and adverse events associated with CIAN were not significantly correlated with outcomes based on the GOS. Kramer et al. [15] successfully prevented thromboembolism with the administration of IA heparin and intravenous tirofiban. In our study, we utilized both intravenous and IA heparin. The intravenous bolus of heparin was adjusted to achieve twice the initial ACT value, while the rate of IA heparin was fixed. No thromboembolic complications were observed in our study. However, severe brain edema occurred in one case, which started after a major infarct. IA nimodipine infusion is known to increase ICP (IICP) due to its vasodilatory effects, which lead to an increase in cerebral blood volume [13]. Similarly, intravenous nimodipine infusion can also induce IICP, with reflex vasodilation following CPP depression further contributing to IICP [6]. While the exact pathophysiological mechanism of IICP during CIAN remains unclear, IA infusion offers an advantage over intravenous administration by minimizing reflex vasodilation. Although CIAN has been associated with ICP elevation, a study has shown that ICP remains within normal limits during the procedure [13]. Other potential causes of IICP after aSAH include hydrocephalus, intracranial hemorrhage, global cerebral edema, and cerebral infarction [1]. In our case, where CIAN was terminated due to severe IICP, cerebral infarction appeared to be the primary cause of IICP rather than the CIAN itself. Therefore, we recommend ICP monitoring prior to initiating CIAN in patients with such underlying pathologies. In another case, thrombocytopenia was observed but did not meet the criteria for HIT, which requires a timeline of over 5 days from the initial administration. In cases of HIT, direct thrombin inhibitors should be used to prevent thrombosis after platelet activation, allowing CIAN to continue [19].

Finally, we proposed a protocol (Fig. 4) for the prevention and treatment of DCI, which includes CIAN as the final rescue treatment. For dose titration, obtain TCD measurements prior to treatment, as well as at 2 and 6 hours after starting treatment. Thereafter, daily TCD assessments are recommended. The CIAN infusion rate is initially set at 1 mg/hr and adjusted based on TCD findings within a range of 0.4 to 2 mg/hr. For patients with ICP monitoring, the goal is to maintain CPP above 70 mmHg. If ICP exceeds 25 mmHg, the following factors should be evaluated : 1) whether blood pressure is excessively high, requiring adjustments such as increasing nimodipine or reducing norepinephrine; 2) secondary causes, such as intracerebral hemorrhage, acute subdural hematoma, or cerebral infarction; and 3) the vasodilatory effects of nimodipine. If the mean arterial pressure falls below 100 mmHg, appropriate interventions include increasing norepinephrine, decreasing nimodipine, or addressing potential cardiac origins.

To prevent catheter-related infections and displacement, the use of occlusive, transparent, and adhesive dressings is recommended. Monitoring for signs of infection or hematoma-such as swelling, redness, or discharge-is essential. Additionally, assess distal limb pulsation, sensory and color changes to detect potential limb ischemia caused by arterial injury or occlusion. Bedside doppler sonography can assist in identifying catheter related thrombus. While various approaches to anticoagulation exist; in our protocol, IA heparin at a fixed rate of 60 U/hr. ACT levels are monitored every 4 hours, aiming to double the proper initial ACT level with IV heparin administered at 1000-2000 U as need.

Our study is constrained by its retrospective nature. Nevertheless, we believe that our cohort of consecutive patients treated according to a standardized treatment protocol over a contemporaneous time frame helps to partially address this limitation. Also, the limited sample size restricts our analyses to non-parametric statistical methods. Therefore, additional investigations utilizing a prospective cohort design are warranted to ascertain the consistency of our results. Additionally, for a more precise analysis, comparing the CIAN group with a cohort exhibiting angiographically refractory CVS that did not receive CIAN is imperative. However, ethical considerations may present substantial challenges to the feasibility of such a study.