Effect of cisternal drainage after early operation for ruptured intracranial aneurysms

Impact of ventriculo-cisternal irrigation on prevention of delayed cerebral infarction in aneurysmal subarachnoid hemorrhage: a single-center retrospective study and literature review

OBJECTIVE

The aim of this study was to evaluate the effectiveness of ventriculo-cisternal irrigation (VCI) in preventing vasospasms and delayed cerebral infarction (DCI) by washing out subarachnoid clots earlier after aneurysm surgery.

METHODS

We retrospectively identified 340 subarachnoid hemorrhage (SAH) patients with ruptured intracranial aneurysms treated with postoperative VCI at our institution between December 2010 and January 2020. As VCI therapy, a ventricular drain/cisternal drain was placed during aneurysm surgery, and lactated Ringer's solution was used for irrigation until day 4 of SAH, followed by intracranial pressure control at 5-10 cmH2O until day 14.

RESULTS

The median age was 65 years (interquartile range 52-75), with 236 female patients (69%). The World Federation of Neurosurgical Societies grade distribution was as follows: grade I or II, 175 patients (51%); grade III or IV, 84 (25%); and grade V, 81 (24%). With VCI management in all patients, total vasospasm occurred in 162 patients (48%), although the DCI incidence was low (23 patients [6.8%]). Major drainage-related complications were observed in five patients (1.5%). Early surgery, performed on SAH day 0 or 1, was identified as a preventive factor against DCI occurrence (odds ratio (OR) 0.21, 95% confidence interval (CI) 0.07-0.67; P = 0.008), while additional surgery (4.76, 1.62-13.98; P = 0.005) and dyslipidemia (3.27, 1.24-8.63; P = 0.017) were associated with DCI occurrence.

CONCLUSION

Managing vasospasms with VCI after SAH is considered a safe and effective method to prevent DCI. Early surgery after SAH may be associated with a decreased risk of DCI with VCI therapy.

The Role of Cisternostomy and Cisternal Drainage in the Treatment of Aneurysmal Subarachnoid Hemorrhage: A Comprehensive Review

Aneurysmal subarachnoid hemorrhage (aSAH) provokes a cascade reaction that is responsible for early and delayed brain injuries mediated by intracranial hypertension, hydrocephalus, cerebral vasospasm (CV), and delayed cerebral ischemia (DCI), which result in increased morbidity and mortality. During open microsurgical repair, cisternal access is achieved essentially to gain proximal vascular control and aneurysm exposition. Cisternostomy also allows brain relaxation, removal of cisternal clots, and restoration of the CSF dynamics through the communication between the anterior and posterior circulation cisterns and the ventricular system, with the opening of the Membrane of Liliequist and lamina terminalis, respectively. Continuous postoperative CSF drainage through a cisternal drain (CD) is a valuable option for treating acute hydrocephalus and intracranial hypertension. Moreover, it efficiently removes the blood and toxic degradation products, with a potential benefit on CV, DCI, and shunt-dependent hydrocephalus. Finally, the CD is an effective pathway to administer vasoactive, fibrinolytic, and anti-oxidant agents and shows promising results in decreasing CV and DCI rates while minimizing systemic effects. We performed a comprehensive review to establish the adjuvant role of cisternostomy and CD performed in cases of direct surgical repair for ruptured intracranial aneurysms and their role in the prevention and treatment of aSAH complications.

The positive impact of cisternostomy with cisternal drainage on delayed hydrocephalus after aneurysmal subarachnoid hemorrhage

BACKGROUND

Hydrocephalus is one of the major complications of aneurysmal subarachnoid haemorrhage (aSAH). In the acute setting, an external ventricular drain (EVD) is used for early management. A cisternal drain (CD) coupled with the micro-surgical opening of basal cisterns can be an alternative when the aneurysm is clipped. Chronic hydrocephalus after aSAH is managed with ventriculo-peritoneal (VP) shunt, a procedure associated with a wide range of complications. The aim of this study is to analyse the impact of micro-surgical opening of basal cisterns coupled with CD on the incidence of VP shunt, compared to patients treated with EVD.

METHODS

The authors conducted a retrospective review of 89 consecutive cases of patients with aSAH treated surgically and endovascularly with either EVD or CD between January 2009 and September 2021. Patients were stratified into two groups: Group 1 included patients with EVD, Group 2 included patients with CD. Subgroup analysis with only patients treated surgically was also performed. We compared their baseline characteristics, clinical outcomes and shunting rates.

RESULTS

There were no statistically significant differences between the two groups in terms of epidemiological characteristics, WFNS score, Fisher scale, presence of intraventricular hemorrhage (IVH), acute hydrocephalus, postoperative meningitis or of clinical outcomes at last follow-up. Cisternostomy with CD (Group 2) was associated with a statistically significant reduction in VP-shunt compared with the use of an EVD (Group 1) (9.09% vs 53.78%; p < 0.001). This finding was confirmed in our subgroup analysis, as among patients with a surgical clipping, the rate of VP shunt was 43.7% for the EVD group and 9.5% for the CD group (p = 0.02).

CONCLUSIONS

Cisternostomy with CD may reduce the rate of shunt-dependent hydrocephalus. Cisternostomy allows the removal of subarachnoid blood, thereby reducing arachnoid inflammation and fibrosis. CD may enhance this effect, thus resulting in lower rates of chronic hydrocephalus.

Oculomotor nerve palsy presumably caused by cisternal drain during microsurgical clipping

Background:

Oculomotor nerve palsy can be caused by diverse etiologies, but no report has yet documented its association with a cisternal drain.

Case Description:

A 35-year-old woman presented with severe headache. The patient did not exhibit oculomotor nerve palsy at presentation. Cranial computed tomography (CT) revealed diffuse subarachnoid hemorrhage. The patient underwent open microsurgical clipping of a ruptured middle cerebral artery aneurysm. During surgery, a cisternal drain was placed in the basal cistern at the medial aspect of the clinoidal portion of the internal carotid artery. The patient presented with the left oculomotor nerve palsy immediately after surgery. CT revealed displacement of the cisternal drain to the lateral aspect of the anterior clinoid process. The patient’s mydriasis and sluggish light reaction recovered after 7 days, while extraocular movements persisted for 50 days. The constructive interference steady-state sequence detected the left oculomotor nerve coursing adjacent to the clinoidal internal carotid artery.

Conclusion:

Oculomotor nerve palsy can be caused by collision with a thin silastic tube placed during surgery for aneurysmal subarachnoid hemorrhage. Withdrawal of the drain as early as possible is recommended when drain-associated oculomotor nerve palsy is suspected.

Early surgery of multiple versus single aneurysms after subarachnoid hemorrhage: an increased risk for cerebral vasospasm?

OBJECT

As many as 33% of patients suffering from subarachnoid hemorrhage (SAH) present with multiple intracranial aneurysms (MIAs). It is believed that aneurysm surgery has the potential to increase the risk of cerebral vasospasm due to surgical manipulations of the parent vessels and brain tissue. Consequently, 1-stage surgery of MIAs, which usually takes longer and requires more manipulation, could even further increase the risk of vasospasm. The aim of this study is to define the correlation between vasospasm and the operative treatment of single intracranial aneurysms versus MIAs in a 1-stage operation.

METHODS

The authors analyzed a database including 1016 patients with SAH, identified retrospectively between 1989 and 1996 and prospectively collected between 1997 and 2004. Exclusion criteria were endovascular treatment, surgery after SAH Day 3, and, in patients with MIAs, undergoing more than 1 operation. Cerebral vasospasm was diagnosed by transcranial Doppler (TCD) ultrasonography and was defined as a maximum mean blood flow velocity > 120 cm/second. The diagnosis of symptomatic vasospasm was made if a new neurological deficit occurred that could not be explained by concomitant complications.

RESULTS

A total of 643 patients who experienced 810 aneurysms were included. Four hundred twenty-four patients were female (65.9%) and 219 were male (34.1%) with an average age of 53.1 years. One hundred twenty-one patients (18.8%) were diagnosed with MIAs. Maximum mean flow velocities measured by TCD were 131 cm/second in patients with MIAs and 129.5 cm/second in patients with single intracranial aneurysms. The incidence of TCD vasospasm (p = 0.561) as well as of symptomatic vasospasm (p = 0.241) was not significantly different in the 2 groups.

CONCLUSIONS

Clipping of more than 1 aneurysm in a 1-stage operation within 72 hours after SAH can be performed without increasing the risk of cerebral (TCD) vasospasm and symptomatic vasospasm.

The Utility and Benefits of External Lumbar CSF Drainage after Endovascular Coiling on Aneurysmal Subarachnoid Hemorrhage

OBJECTIVE

Cerebral vasospasm still remains a major cause of the morbidity and mortality, despite the developments in treatment of aneurysmal subarachnoid hemorrhage. The authors measured the utility and benefits of external lumbar cerebrospinal fluid (CSF) drainage to prevent the clinical vasospasm and its sequelae after endovascular coiling on aneurysmal subarachnoid hemorrhage in this randomized study.

METHODS

Between January 2004 and March 2006, 280 patients with aneurysmal subarachnoid hemorrhage were treated at our institution. Among them, 107 patients met our study criteria. The treatment group consisted of 47 patients who underwent lumbar CSF drainage during vasospasm risk period (about for 14 days after SAH), whereas the control group consisted of 60 patients who received the management according to conventional protocol without lumbar CSF drainage. We created our new modified Fisher grade on the basis of initial brain computed tomography (CT) scan at admission. The authors established five outcome criteria as follows : 1) clinical vasospasm; 2) GOS score at 1-month to 6-month follow-up; 3) shunt procedures for hydrocephalus; 4) the duration of stay in the ICU and total hospital stay; 5) mortality rate.

RESULTS

The incidence of clinical vasospasm in the lumbar drain group showed 23.4% compared with 63.3% of individuals in the control group. Moreover, the risk of death in the lumbar drain group showed 2.1% compared with 15% of individuals in the control group. Within individual modified Fisher grade, there were similar favorable results. Also, lumbar drain group had twice more patients than the control group in good GOS score of 5. However, there were no statistical significances in mean hospital stay and shunt procedures between the two groups. IVH was an important factor for delayed hydrocephalus regardless of lumbar drain.

CONCLUSION

Lumbar CSF drainage remains to play a prominent role to prevent clinical vasospasm and its sequelae after endovascular coiling on aneurysmal subarachnoid hemorrhage. Also, this technique shows favorable effects on numerous neurological outcomes and prognosis. The results of this study warrant clinical trials after endovascular treatment in patients with aneurysmal SAH.

CONTROVERSIES IN THE SURGICAL TREATMENT OF RUPTURED INTRACRANIAL ANEURYSMS

THE MANAGEMENT OF aneurysmal subarachnoid hemorrhage has evolved over time, including the use of the microscope for aneurysm clip application, improved imaging modalities, endovascular methods for aneurysm treatment, dedicated neurointensive care units, and more aggressive therapy for cerebral vasospasm. Although these advancements have reduced the morbidity and mortality associated with aneurysmal subarachnoid hemorrhage, outcomes for this patient population continue to leave much room for improvement. This work highlights controversial adjuvant techniques, maneuvers, and therapies surrounding the surgical treatment of ruptured cerebral aneurysms that currently lack a consensus opinion. These treatments include centralized care in high-volume centers, as well as the use of antifibrinolytic therapy, routine cerebrospinal fluid diversion, intraoperative hypothermia, temporary clip application, neuroprotective drugs, intraoperative angiography, and decompressive hemicraniectomy. Although definitive answers will only be possible through future multicenter collaboration, we review the controversy surrounding these adjuncts and report the consensus opinion from a highly experienced audience.

Advances in vasospasm treatment and prevention

Outcome after aSAH depends on several factors, including the severity of the initial event, perioperative medical management, surgical variables, and the incidence of complications. Cerebral vasospasm (CV) is ure to consistently respond to treatment, emphasizing the need for further research into the underlying mechanisms of SAH-induced cerebrovascular dysfunction. To this end, our paper reviews the relevant literature on the main therapies employed for CV after aSAH and discusses possible avenues for future investigations. Current management of this condition consists of maximal medical therapy, including triple H regimen and oral administration of calcium antagonists, followed by endovascular balloon angioplasty and/or injection of vasodilatory agents for refractory cases. As the precise pathophysiology of CV is further elucidated, the development of promising investigational therapies will follow.

Fenestration of the Lamina Terminalis as a Valuable Adjunct in Aneurysm Surgery

OBJECTIVE:

Hydrocephalus, vasospasm, and frontobasal injury are common complications after aneurysmal subarachnoid hemorrhage (SAH) from anterior communicating artery aneurysms. Previous studies have suggested that fenestration of the lamina terminalis (FLT) during surgery may be associated with reduced rates of shunt-dependent hydrocephalus and vasospasm. We report 106 patients affected by anterior communicating artery aneurysms and Fisher Grade 3 aneurysmal SAH and the affect of FLT on shunt-dependent hydrocephalus, vasospasm, and frontobasal injury.

METHODS:

During a 3-year period, 53 patients underwent FLT and 53 did not. We prospectively evaluated admission and discharge clinical grades, hydrocephalus at admission, occurrence of clinical vasospasm, need for interventional vasospasm therapy, frontobasal hypodensity incidence, and permanent ventriculoperitoneal shunting requirement. Follow-up ranged from 3 to 35 months (mean, 17.9 mo).

RESULTS:

Shunting incidence after aneurysmal SAH with hydrocephalus was 4.25% in patients who underwent FLT and 13.9% in patients who did not (P&lt; 0.001). Clinical cerebral vasospasm occurred in 29.6% of patients who underwent FLT and in 54.7% of patients who did not (P&lt; 0.001). Frontobasal hypodensity was identified postoperatively in 0% of patients who underwent FLT and in 5% of patients who did not. Good outcome was reported in 69.81% of patients who underwent FLT and in 33.96% of patients who did not (P&lt; 0.001). Poor outcome was associated with higher Hunt and Hess grades, need for ventricular drainage, elevated intracranial pressure, and multiple interventional vasospasm therapies. No complications were linked to FLT.

CONCLUSION:

FLT was associated with statistically significant decreases in shunting rates, incidence of vasospasm, and better outcomes. We recommend its routine use in patients with Fisher Grade 3 anterior communicating artery aneurysmal SAH.

Effectiveness of the head-shaking method combined with cisternal irrigation with urokinase in preventing cerebral vasospasm after subarachnoid hemorrhage

OBJECT

The head-shaking method combined with cisternal irrigation has been proposed to be effective in preventing cerebral vasospasm after subarachnoid hemorrhage (SAH) by facilitating rapid washout of the clot from the subarachnoid space. This study was conducted to evaluate the effectiveness of this method.

METHODS

The inclusion criteria included the following: 1) Fisher Grade 3 SAH on admission computerized tomography (CT) scans; 2) aneurysm secured within 48 hours of SAH onset; and 3) no focal deficit and ability to obey commands within 24 hours postsurgery. Two hundred thirty patients treated between 1994 and 2002 fulfilled the criteria. Because only one machine was available and it required I month of maintenance every other month, 114 patients underwent irrigation combined with the head-shaking method (head-shaking group), whereas the remaining 116 patients received cisternal irrigation alone (control group). There were no significant differences in sex, age, site of aneurysm, or preoperative grade between the two groups. The incidence of symptomatic vasospasm with or without infarction, cerebral infarction on CT scans, and permanent ischemic neurological deficit was 25.7, 17.7, and 8.8%, respectively, in the control group and 15.2, 4.5, and 2.7% in the head-shaking group. The difference was statistically significant for symptomatic vasospasm, cerebral infarction, and permanent ischemic neurological deficit (p < 0.05). In a multivariate backward stepwise logistic regression analysis, absence of head shaking was the only variable that was predictive of permanent ischemic neurological deficit (p = 0.061). The outcomes evaluated using the modified Rankin Scale were better in the head-shaking group (p = 0.051).

CONCLUSIONS

The head-shaking method significantly reduced the incidence of symptomatic vasospasm, cerebral infarction, and permanent ischemic neurological deficit and improved the clinical outcomes in patients who underwent cisternal irrigation therapy after aneurysmal SAH.

Incidence of cerebral vasospasm after endovascular treatment of acutely ruptured aneurysms: report on 69 cases

Cerebral vasospasm is the most common cause of morbidity and mortality in patients admitted to the hospital after suffering aneurysmal subarachnoid hemorrhage (SAH). The early surgical removal of subarachnoid clots and irrigation of the basal cisterns have been reported to reduce the incidence of vasospasm. In contrast to surgery, the endovascular treatment of aneurysms does not allow removal of subarachnoid clots. In this study the authors measured the incidence of symptomatic vasospasm after early endovascular treatment of acutely ruptured aneurysms with Guglielmi detachable coils (GDCs). Sixty-nine patients classified as Hunt and Hess Grades I to III underwent occlusion of intracranial aneurysms via GDCs within 72 hours of rupture. The amount of blood on the initial computerized tomography (CT) scan was classified by means of Fisher's scale. Symptomatic vasospasm was defined as the onset of neurological deterioration verified with angiographic or transcranial Doppler studies. Hypertensive, hypervolemic, hemodilution therapy, with or without intracranial angioplasty, was used to treat vasospasm after GDC placement. Symptomatic vasospasm occurred in 16 (23%) of 69 patients. The clinical grade at admission and the amount of blood on the initial CT were both associated with the incidence of subsequent vasospasm. At 6-month clinical follow-up examination, 12 of these 16 patients experienced a good recovery, two were moderately disabled, and two patients had died of vasospasm. In conclusion, the 23% incidence of symptomatic vasospasm in this series compares favorably with that found in conventional surgical series of patients with acute aneurysmal SAH. These results indicate that endovascular therapy does not have an unfavorable impact on cerebral vasospasm.

Early treatment of subarachnoid hemorrhage after preventing rerupture of an aneurysm

Twelve patients with Hunt and Hess neurological Grades III to V underwent thrombosis of aneurysms using cellulose acetate polymer within 23 hours of aneurysm rupture. On computerized tomography (CT), nine of these patients had difuse or localized thick subarachnoid blood clots, two had diffuse thin clots and one had intraventricular clots. Immediately after thrombosis, all patients were administered tissue plasminogen activator (TPA) through spinal or ventricular catheters. The pressure of the lumbar cerebrospinal fluid was maintained at 100 to 150 mm H2O. The TPA was given as multiple injections of 2 mg on Day 0 and 1 to 2 mg on the following 1 to 2 days. In two patients the second injection of TPA was not given because of severe brain damage resulting from the initial subarachnoid hemorrhage. Ten patients showed complete clearance of the cisternal clot on CT within 72 hours after thrombosis. Seven partially thrombosed aneurysms and five multiple aneurysms were clipped during delayed surgery. Only one patient experienced mild vasospasm as shown on the follow-up angiogram. Eight patients improved clinically and had a good recovery, two had severe disability, and two died. Urgent thrombosis of a ruptured aneurysm followed by immediate postthrombotic administration of TPA may be a safe and reasonable means of preventing vasospasm and improving patient outcome.

Effects of prophylactic intrathecal administrations of nicardipine on vasospasm in patients with severe aneurysmal subarachnoid haemorrhage

Calcium antagonists are currently most widely used for chronic cerebral vasospasm after aneurysmal subarachnoid haemorrhage (SAH). However, the vasodilatory effects of systemically administered calcium antagonists can be limited secondary to hypotension. We previously compared intrathecal and intravenous routes of administration of nicardipine. Intrathecal administration of nicardipine significantly dilated spastic basilar arteries on day 7 in a two-haemorrhage canine model of vasospasm. In the present communication, the effects of prophylactic, serial administration of intrathecal nicardipine on vasospasm was examined in 50 patients. Patients were classified as Fisher SAH group 3 and all had their aneurysms clipped within 3 days of SAH. Following placement of a cisternal drain, 2 mg of nicardipine was injected, three times each day for an average of 10 days. The control group consisted of 91 similar patients with cisternal drainage not treated with nicardipine. Intrathecal administration of nicardipine decreased the incidence of symptomatic vasospasm by 26%, angiographic vasospasm by 20% and increased good clinical outcome at one month after the haemorrhage by 15%. Postoperative angiograms revealed that patients in the nicardipine group showed less vasospasm of major cerebral arteries, near the tip of a drain in the basal cistern, but vasospasm in the A2 and M2 segments was not decreased. Radio-isotope cisternography suggested that nicardipine might not reach the subarachnoid space around A2 and M2 segments. Nine patients complained of headache probably secondary to nicardipine induced vasodilation. Two patients suffered from meningitis, both were successfully treated. Intrathecal administration nicardipine appears to be effective in the treatment of vasospasm, but side effects were significant.

Preventive therapy against delayed cerebral ischaemia after aneurysmal subarachnoid haemorrhage: trials of thromboxane A2 synthetase inhibitor and hyperdynamic therapy

The effects of thromboxane A2 synthetase inhibitor and hyperdynamic therapy on delayed cerebral ischaemia following aneurysmal subarachnoid haemorrhage were evaluated in a series of twenty eight patients, who underwent aneurysmal clipping with 72 hours after subarachnoid haemorrhage. Postoperatively, 13 patients were treated with thromboxane A2 synthetase inhibitor, Xanbon [sodium (E)-3-[p-(1H-imidazol-1-ylmethyl)phenyl]-2-propenoate]. Hyperdynamic therapy with dobutamine was given to the remaining 15 patients. Of the 13 patients treated with Xanbon, nine patients (69%) developed delayed cerebral ischaemia and cerebral infarcts occurred in eight patients (62%). On the other hand, of the 15 patients treated with hyperdynamic therapy, only three patients (20%) manifested delayed cerebral ischaemia and two patients (13%) developed cerebral infarcts. In the present study, the patients treated with hyperdynamic therapy met an expected incidence of ischaemic events after subarachnoid haemorrhage by today's standards, while those treated with thromboxane A2 synthetase inhibitor did not.

Prospective study on the prevention of cerebral vasospasm by intrathecal fibrinolytic therapy with tissue-type plasminogen activator

The authors have evaluated the efficacy of postoperative intrathecal injections of tissue-type plasminogen activator (tPA) in preventing cerebral vasospasm in patients with a diffuse thick subarachnoid hemorrhage (SAH). The present study examined 105 patients who underwent direct surgery within 48 hours of SAH and whose computerized tomography (CT) findings were classified as Fisher CT Group 3. Patients showing diffuse thick subarachnoid blood clots on CT with greater than 75 Hounsfield units (HU) were included in the tPA therapy group and those with below 75 HU comprised the control group. The surgical method was the same in both groups, and both groups had cisternal drainage instituted. On the day following the operation, the tPA group was given an intrathecal injection of tPA (2 mg), and this was continued for several days until all of the cisterns exhibited low density on CT scans. Follow-up angiography showed that 26 cases (87%) in the tPA group had no vasospasm, three (10%) had moderate vasospasm, and one (3%) had severe vasospasm. All four patients showing spasm on angiography were asymptomatic, and there were no cases of delayed ischemic neurological deficits (DIND). In contrast, there were 11 cases (15%) with DIND in the control group. In the tPA group, there was one case of SAH caused by drainage catheter removal, one with a small epidural hematoma, and one with subgaleal fluid accumulation; all of these were treated conservatively with favorable results. Overall, there were no infectious complications related to cisternal drainage and intrathecal injection of tPA. These results indicate that multiple intrathecal injections of small doses of tPA are effective and safe in preventing vasospasm. On the basis of this experience, the authors conclude that injection of 2 mg of tPA daily for 5 days (a total of 10 mg) is effective in preventing the development of vasospasm.

Effect of continuous cisternal drainage on cerebral vasospasm

The effect of continuous cisternal drainage on cerebral vasospasm was studied under strict criteria in 140 patients with ruptured intracranial aneurysms. The degree of subarachnoid haemorrhage (SAH) on the computed tomography scan was graded from I to IV. The patients were classified according to the total amount of cisternal drainage into three groups, regardless of the duration of the drainage and whether or not it was accompanied by irrigation; i.e., those with less than 500 mL (group 1: 57 cases), those with 500-3000 mL (group 2: 44 cases), and those with 3000-9500 mL (group 3: 39 cases). While correlations could be found between both clinical and SAH grades with the severity of vasospasm, closer correlation could be found in the SAH grades. In analyzing the cases with subarachnoid haemorrhage grades III-IV (severe clots), the angiographic vasospasm was less severe in groups 2 and 3 than in group 1, and the incidences of permanent symptomatic vasospasm and low-density area on computed tomography were lower in groups 2 and 3 than in group 1. Regarding the surgical outcome in cases with SAH grades III-IV, the mortality rate was lower in groups 2 and 3 (22% and 19%) than in group 1 (33%). Further, the rate of good recovery was higher in groups 2 and 3 (61% and 57%) than in group 1 (28%). However, there were no differences between groups 2 and 3 in cerebral vasospasm or in surgical outcome. As a shortcoming of continuous cisternal drainage, the need for shunt operation was higher in groups 2 and 3 than in group 1.