Introduction 

   Unruptured intracranial aneurysms (UIAs) may be discovered before they rupture either because they produce symptoms such as new headache or cranial neuropathies or because they appear as incidental findings in imaging studies. Since the optimal method to decrease the mortality and morbidity associated with the rupture of aneurysms is to discover them and operate before rupture, early recognition and appropriate treatment of these lesions has been the most important neurosurgical concern in aneurysmal management. But the treatment of UIAs presents a management dilemma for both clinician and patient because all treatment modalities which currently used carry risks. It ultimately depends on the relative risk of subsequent rupture of aneurysm if left untreated versus the risk of treatment. It is well known that the significant factors influencing the rupture of UIAs include the size, location, presence of hypertension, age, and symptomatic presentation of aneurysms. Therefore, when we incidentally find UIAs, we should consider the natural course of UIAs and whether they should be treated or not, and the treatment modalities. 
   We report the surgical results of 49 patients with 52 UIAs, discuss the risk factors related to the rupture of UIAs, and would like to suggest the appropriate treatment options availuable for UIAs. 

Materials and Methods 

   A retrospective review of all patients surgically treated after diagnosis of intracranial aneurysms with or without subarachnoid hemorrhage (SAH) from January 1999 to December 2005 was performed by review of the medical records and radiologic studies. During this period, 318 operations were performed on 299 patients presenting with 371 aneurysms. Of these aneurysms, 52 aneurysms (14%) among 49 patients (16.4%) were the UIAs and 319 aneurysms (86%) among 264 patients (83.6%) were the ruptured aneurysms. Forty-four of the 264 patients undergoing a operation for ruptured aneurysms had 51 additional UIAs which were simultaneously operated on with ruptured aneurysmone. For the purpose of indirectly evaluating the factors concerning about the rupture of UIAs, we analysed the ruptured cases operated on our hospital excluding above 51 additional UIAs. 
   The patients with UIAs was divided into 4 subgroups. Group Ⅰincluded 18 patients (36.7%) with 19 asymptomatic UIAs incidentally found during workup of unrelated problems (Table 3). Group Ⅱ included 15 patients (30.6%) with 17 asymptomatic UIAs and they had a history of SAH from contralateral aneurysms and a previous operation (Table 4). Group Ⅲ included 7 patients (14.3%) with 7 symptomatic UIAs (Table 5), and group Ⅳ 9 patients (18.4%) with 9 asymptomatic ones which were incidentally found during workup of other intracranial problems such as cerebral infarction or seizure (Table 6). Group Ⅱ in the current study is coincident with group 2 in the International Study of Unruptured Intracranial Aneurysms (ISUIA).⁴¹⁾ 
We defined a size of aneurysm as the long diameter of aneurysm and classified the size of UIAs as follows, less than 5 mm was consistent with a small aneurysm, 5 to 15 mm medium aneurysm, and greater than 15 mm large aneurysm. The modified Glasgow Outcome Scale (m-GOS)2 was used at one month after operation to assess the clinical outcomes in each patient. Surgical mortality was defined as death without regard to the cause. Morbidity included a newly developed neurological deficit occurring after operation . 
   For the statistical analysis, chi-square testing was used for analysing the influence of hypertension and smoking on the rupture of aneurysms. A p value less than 0.05 indicated statistical significance. Data analysis was performed with SPSS (version 12.0 SPSS Inc, Chicago, IL). 

Results 

   The characteristics of patients with UIAs and aneurysms treated in our hospital are presented in Table 1 while the characteristics of patients with ruptured aneurysms and aneurysms are presented in Table 2. Although Wiebers et al insisted that the natural course of UIAs could not be extrapolated from evaluation of patients with ruptured aneurysms,¹⁸⁾ we proceeded to analyse the characteristics of ruptured aneurysms to enhance our understanding of the natural course of UIAs. Reports contradictory to Wiebers and colleagues showed that all aneurysms that had been left untreated and subsequently ruptured were larger at the time of rupture than at the inception of the study and that the rupture of the aneurysm might be the result of acute stretching of the aneurysmal wall.²³⁾⁴⁵⁾ 
   The rate of occurrence in males and females and mean age was similar in both groups. In the case of UIAs, the male/female ratio was 12 (24.5%) : 37 (75.5%) and the mean age was 55.2 years old (range 36~75 years old). In the case of ruptured aneurysms, the male/female ratio was 98 (37.1%) : 166 (62.9%) and the mean age was 53.9 years old (range 19~77 years old). In UIA cases, aneurysm of MCA was shown to have high rate of occurrence, while ruptured aneurysms frequently occurred in the AComA. The location of UIAs were as follows: the distal anterior cerebral artery (ACA) in 3 cases (5.8%), the anterior communicating artery (AComA) in 5 cases (9.6%), the middle cerebral artery (MCA) in 23 cases (44.2%), the internal carotid artery (ICA) in 11 cases (21.2%), the posterior communicating artery (PComA) in 8 cases (15.4%), and the basilar artery in 2 cases (3.8%). The location of ruptured aneurysms was as follows: the distal ACA in 7 cases (2.6%), the ACA in 6 cases (2.2%), the AComA in 98 cases (36.2%), the MCA in 82 cases (30.6%), the ICA in 12 cases (4.5%), the PComA in 50 cases (18.7%), and the posterior circulation in 13 cases (4.8%). The mean size of the UIAs and ruptured aneurysms was 7.6 mm (range 2~20 mm) and 6.15 mm (range 3~25 mm) retrospectively. In UIAs, 44 of 52 aneurysms (84.6%) were in the category of medium size. However, the frequency of small aneurysms was more prominent in ruptured aneurysms than the UIAs, occurring 100 of 268 ruptured aneurysms (37.3%). 
   In patients with UIAs, 53 procedures were performed to 52 UIAs. Neck clipping was achieved in 46 aneurysms (86.8%), wrapping in 2 (3.8%), trapping in 1 (1.9%), coil embolization in 3 (5.6%), and balloon occlusion in 1 (1.9%). In case number 3 patient included in group Ⅲ, the patient was underwent one procedure and one operation. First, balloon occlusion of the proximal ICA was performed. The facial pain was completely relieved immediately after the procedure. However, the patient complained that ptosis and the facial pain were progressively aggravated during the next two weeks. On the brain magnetic resonance image (MRI at two weeks after procedure, it was found that the aneurysm was restored to the previous size. We thought that it might be the collateral flow from the AComA and the PComA. At second stage operation, we ligated the distal ICA through craniotomy approach. The symptom of facial pain improved immediately after the second operation, but the ptosis did not improve until three months after operation. In patients with ruptured aneurysms, neck clipping was achieved in 249 aneurysms (92.9%, wrapping in 10 (3.7%, trapping in 1 (0.4%, and coil embolization in 8 (3.0%. 
   In all patients with UIAs, postoperative angiography was performed and complete obliteration of the aneurysm was documented. The residual neck of greater than 1 mm was not seen. 
   There was no operative deaths among patients with UIAs. In patients with UIAs, there were only minor complications in three patients (12.5%: ptosis in 2 cases and spinal subdural hematoma in 1 case. The spinal subdural hematoma was occurred in case number 8 of group Ⅰ. In that case, the patient had not clearly identified predisposing factors such as bleeding tendency, anticoagulation therapy, or spinal tapping. We had been already reported that cases.24 Of the two cases of ptosis, one case occurred on patients of group Ⅲ and resulted from the excessive retraction of the supraorbital branch of facial nerve during operation. Another cases was found in a patient of group Ⅳ which possibly resulted from intraoperative manupulation. All complications were temporary and completely relieved within three months after the operation. Most patients of group Ⅳ and two patients of group Ⅱ (case 3 & 8) had preoperative motor deficits which were unchanged or partly improved after operation. These cases were not include in the surgical morbidity. The complications of surgery were not related with the size of the aneurysms. Finally, the surgical mortality was 0% and morbidity rate was 2%. 
   Although we divided the UIAs into four subgroups according to patient's characteristics, we could not find any difference in the outcomes according to the patient's group. There was also no relation of predisposition to rupture with hypertension (p=0.523) and cigarette smoking (p=0.872) in both patients with ruptured aneurysms and patients with UIAs. 

Discussion 

   The reported prevalence of UIAs in autopsy series and radiologic studies ranges from 0.2% to 8.9% of the general population and the rupture of UIAs can cause SAH. Most intracranial aneurysms remain intact through one's whole life. As of yet, there are no clinical or paraclinical features which can be used to predict which unruptured aneurysms will rupture after their diagnosis.^{5)16)22)26)31)38)43)} Although the natural course of asymptomatic UIAs is not precisely understood, it has been estimated that the annual rupture rate of asymptomatic aneurysms is between 0.1% to 2.7%. It is known that 10% to 50% of the UIAs eventually bled.^{3)12)18)21)30)46)} 
   Several factors can be considered indicators of the prognosis of a UIA. It is well known that aneurysm size and location have a significant role in determining the risk of future rupture.^{3)18)21)30)32)34)36)41)45)} In 1998, the International Study of Unruptured Intracranial Aneurysms (ISUIA) suggested that rupture was rare below 10 mm in diameter.⁴¹⁾ But the low rupture rate and relatively large critical size for aneurysm rupture apparently contradicted surgeons' experiences with ruptured aneurysms. Substantially, many authors advocated that the critical size for rupture was about 5 mm and recommended surgery for asymptomatic UIAs larger than 5 mm in the absence of other significant adverse factors.^{22)25)29)39)40)47)} Several authors emphasized the risk of rupture of small UIAs of even less than 5 mm.³⁾⁴¹⁾ Furthermore, some neurosurgeons addressed that regardless of its size, an UIA should be considered for surgical obliteration since the literature has clearly shown that the risks of surgical intervention are far less compared with the risks of the natural history of an aneurysm with SAH.²⁰⁾³²⁾ In reports of Crompton and Suzuki et al, UIAs that grew as large as 4 or 5 mm formed the weak points which might be responsible for the rupture.⁸⁾⁴⁰⁾ Based on these clinical and pathological studies, it is reasonable to believe that the critical size for rupture of UIAs begins at 4 or 5 mm in diameter. In our hospital, we recommended surgical treatment if the UIAs were larger than 5 mm if patients were in good medical condition for general anesthesia. For UIAs smaller than 5 mm in diameter. close observation with regular follow-up was advised. But, if the patients complained of any warning signs such as repeated headache or cranial nerve palsy or the enlargement of aneurysm during follow-up, surgical treatment is recommended even though the size of UIAs was small. We thought that the management modality of the aneurysms cannot be determined by aneurysm size only. In our series, 100 of 268 ruptured aneurysms were less than 5 mm in diameter. This means that asymptomatic UIAs smaller than 5 mm were not free of the risk of rupture. 
   Separately these notions about critical size for rupture, Dickey and Kailasnath asserted the diameter-cube hypothesis: the rupture probability varied as third power of aneurysm diameter, a relationship that predicted a continuous increase in rupture probability with increasing size rather than the existence of a threshold diameter that separated low- from high-risk aneurysms. They addressed that the pattern of nonuniform wall thickness was present in essentially all aneurysms, especially those greater than 3 mm in diameter and there was little variation in ultrastructure of the aneurysm. Because it was in the weak sites that aneurysms ruptured, the likelihood of an aneurysm rupture should therefore be proportional to the number of weak spots in an aneurysm wall.¹⁰⁾ 
   Another key prognostic factor for predicting aneurysm rupture and postoperative mortality and morbidity of asymptomatic UIAs is the location of aneurysms.^{3)18)30)36)41)} In general, MCA aneurysms usually preponderate in autopsy material, whereas in clinical material, AComA aneurysms are more frequent than MCA aneurysms and the mean diameter of aneurysms at the time of rupture is smaller in AComA aneurysms than MCA aneurysms.^{1)12)16)32)33)} These findings suggest that a critical size for aneurysm rupture undoubtedly exists but unlike to be the same for each aneurysms. In a report by Ohashi et al, the reason that AComA aneurysms tend to be rupture at a in smaller diameter more easily than MCA aneurysms is that the diameter of the AComA and distal ACA is smaller than that of MCA and ICA.³³⁾ They insisted that the size of aneurysms at the time of rupture may in part be determined by the original thickness and diameter of the parent arteries. Yasui et al reported that all aneurysms that had been left untreated and subsequently ruptured were larger at the time of rupture than at the inception of the study.⁴⁵⁾ Kataoka et al reported that the rupture of an aneurysm might be the result of acute stretching of the aneurysmal wall, and about half of the small aneurysms they examined showd partial damage in the wall of the aneurysm.²³⁾ They asserted that if the acutely stretched, fragile wall of the aneurysm did not rupture, a repair process might begin. These repair process subsituted the hyaline tissue for the dense collagen layer or vascular smooth muscle cell. Fig. 1 shows the difference in histological findings between the UIAs and the ruptured aneurysms. Because the thickness and diameter of parent arteries might be influenced upon these processes, the aneurysms developed from small arteries such as distal ACA or AComA were more easily ruptured before they grew to the critical size for rupture. In our series of ruptured aneurysms, the mean diameter of AComA aneurysms was smaller than the one of MCA and ICA aneurysms (Table 7). Therefore, even though the size of UIAs was smaller than critical size for rupture, if UIAs were located on a susceptible area, the rupture possibility was increased. 
   Besides the size and the location of aneurysms, several factors can be influenced on the formation and rupture of aneurysms: race, age of the patient, hypertension, multiplicity, shape of UIAs, presence of symptoms other than rupture, and a current status of smoking.^{3)7)21)30)33)41)42)45)} The relation between age and the risk of rupture is controversial. Some authors reported that age was adversely related to the risk of rupture³⁾²¹⁾ while others stated there was no relation.⁴¹⁾ Taking into account the relative high rates of annual incidence of SAH, even though there was no difference in the prevalence of UIAs between Caucasian and Japanese, the genetic background of race should also be considered when analysing the rupture risks associated with UIAs.³⁰⁾⁴⁶⁾ 
   Hypertension and smoking are also controversial factors for aneurysm formation and rupture. Some authors reported a positive relation where inadequately or non-tre