Endovascular Treatment of Very Small Unruptured Aneurysms
Rate of Procedural Complications, Clinical Outcome, and Anatomical Results
Background and Purpose—The strategy of treatment of small unruptured intracranial aneurysms is complex because of their presumably low risk of rupture. A precise knowledge of the perioperative complications in this specific subgroup is mandatory. The purpose of this study was to compare the results of the endovascular treatment of aneurysms ≤3 mm and aneurysms >3 mm included in the Analysis of Treatment by Endovascular Approach of Nonruptured Aneurysms study.
Methods—The study included 626 patients harboring 682 unruptured aneurysms. Perioperative adverse events and clinical outcome were analyzed in patients treated for aneurysms ≤3 mm (51 patients, 51 aneurysms) and in patients treated for aneurysms >3 mm (575 patients, 631 aneurysms).
Results—Endovascular treatment failed more often in aneurysms ≤3 mm (13.7%) compared to aneurysms >3 mm (3.3%; P=0.003). The rate of intraoperative rupture for aneurysms ≤3 mm (3.9%; 95% CI, 0.5–13.5) did not significantly differ compared to aneurysms >3 mm (2.4%; 95% CI, 1.2–3.6; P=0.37). Thromboembolic events were not significantly different in both groups (3.9%; 95% CI, 0.5–13.5 in very small aneurysms and 7.1%; 95% CI, 5.1–9.1 in larger aneurysms; P=0.57). One month morbidity/mortality was not significantly different for patients with very small aneurysms (2.0%; 95% CI, 0.05–10.45) and for patients with larger aneurysms (3.3%; 95% CI, 1.8–4.8; P=0.60).
Conclusions—The risks of endovascular treatment are similar in patients with very small or with larger aneurysms. Because the risk of spontaneous rupture is lower in very small aneurysms, their management will include follow-up MRI and active treatment in case of morphological modification.
The management of very small aneurysms is still unclear, especially when they are unruptured. According to the International Study of Unruptured Intracranial Aneurysms, the risk of spontaneous aneurysm rupture is related to aneurysm size and aneurysm location.1 Accordingly, the treatment of small and especially very small unruptured aneurysms is still controversial. However, the decision to treat or not to treat a very small aneurysm must also take into account the risks of treatment. The 2 most frequent complications of endovascular treatment (EVT) of unruptured or ruptured aneurysm are thromboembolic events and intraoperative rupture. In the Analysis of Treatment by Endovascular Approach of Nonruptured Aneurysms (ATENA) series, the rate of thromboembolic events in the global population was 7.1%, and that of intraoperative rupture was 2.6%. Analysis of the global population of the ATENA study showed that the risk of aneurysm rupture was significantly higher in small aneurysms (3.7% for 1–6 mm vs 7% for 7–15 mm; P=0.008) and that the risk of thromboembolic events was lower (4.6% for 1–6 mm vs 9.9% for 7–15 mm; P=0.005).2
Several series have evaluated the results of EVT in the management of very small aneurysms.3–8 However, most of these were retrospective and monocentric, mixing ruptured and unruptured aneurysms. Given the limitations of these previous series, ATENA involved a comparison of the rate of adverse events and clinical outcome in very small and larger unruptured aneurysms. The purpose of the present study was to compare the results of EVT in aneurysms ≤3 mm and aneurysms >3 mm.
Materials and Methods
ATENA Study Protocol
The ATENA protocol has been described previously.2 Briefly, patients were prospectively and consecutively included from 27 Canadian and French neurointerventional centers. Inclusion criteria included patients harboring unruptured, untreated intracranial aneurysms with diameters <15 mm. In each center, indication for treatment and its modality was decided by a local multidisciplinary team, including neurosurgeons, neurologists, neuroanesthesiologists, and neuroradiologists. Fusiform and dissecting aneurysms were excluded, as well as aneurysms associated with brain arteriovenous malformations. In cases of recent subarachnoid hemorrhage (<1 month) related to another aneurysm, patients were not included. The protocol was approved by the Ethics Committee of Reims, and informed consent was obtained from all patients for participation in the study.
The rationale for selecting EVT as the treatment option was not requested in the database. However, it should be emphasized that in most centers in France, EVT is usually proposed as the first-line treatment in unruptured intracranial aneurysms.
Between June 2005 and October 2006, 649 patients harboring 729 unruptured and untreated intracranial aneurysms were included in the ATENA study. Because the goal of the present analysis was to compare the rate of therapeutic complications, clinical outcome, and anatomic results in patients having very small aneurysms and patients having larger aneurysms, patients having both types of aneurysms were excluded from the analysis. Ultimately, 23 patients (3.5%) having both ≤3 mm and >3 mm aneurysms, for a total of 47 aneurysms (6.4%), were excluded.
Clinical and procedural data were collected and entered by the local investigator via an electronic website (Kika-Medical). Data regarding patients (age, gender, hypertension, and smoking), aneurysms (number, location, size, neck size), EVT procedures (number, modalities, anatomic result, adverse events), and patient evolution at 1 month (mortality and morbidity) were collected.
Specific and nonspecific adverse events were evaluated in the present study. Specific adverse events were thromboembolic events, intraoperative rupture, and coil-related complication (stretching, premature detachment). Thromboembolic events were diagnosed intraoperatively by angiography regardless of type (clotting near the neck of the aneurysm, clotting in the distal branches, and parent vessel occlusion). Postoperative thromboembolic events were diagnosed by MRI and/or digital subtraction angiogram performed in cases of sudden neurological compromise. Intraoperative rupture was diagnosed by the exit of the tip of the coil or the microcatheter outside the limit of the aneurysmal sac and/or extravasation of contrast media. Nonspecific adverse events were events related to the vascular procedure, but not specifically to aneurysm treatment (for example, groin hematoma). Adverse events were reported even if no clinical modification was associated with them.
All clinical and procedural data were reviewed by 1 of 2 interventional neuroradiologists (L.P., L.S., each having >15 years of experience in interventional neuroradiology). Each time the data reported by investigators were insufficient or imprecise, or in the case of clinically important adverse events, medical records were requested and evaluated by the 2 interventional neuroradiologists by consensus.
The location of the aneurysm was classified into 1 of 4 groups: internal carotid artery, anterior cerebral artery and anterior communicating artery, middle cerebral artery, and the vertebrobasilar system. The dome and neck of the aneurysm were measured using 3-dimensional angiography. The size of the aneurysm was defined based on the greatest length of the aneurysmal sac.
Clinical status was evaluated at 1 month by a neurologist or neurosurgeon using the modified Rankin scale. Morbidity was defined as a modified Rankin scale score of 2 to 5. When preoperative modified Rankin scale was >1, morbidity was defined by any increase in modified Rankin scale. Any death within 30 days was considered a treatment-related death.
The degree of aneurysm occlusion was defined using the Jean Raymond simplified 3-point classification scale (complete occlusion, neck remnant, and aneurysm remnant)9 and was evaluated by the local investigator. A 2-point classification scale was also used: adequate occlusion (complete occlusion or neck remnant) and aneurysm remnant. Aneurysm occlusion was evaluated postoperatively. Follow-up angiographic results were not evaluated in the present publication.
Definition of Groups
Patients and aneurysms were classified into 2 groups: patients with aneurysms ≤3 mm (group 1) and patients with aneurysms >3 mm (group 2). For the study of intraoperative perforation, another analysis was performed with aneurysms classified into 3 groups: ≤3 mm in diameter, between 3 and 7 mm in diameter, and ≥7 mm in diameter. Intraoperative rupture was evaluated in these 3 groups because the analysis of the whole ATENA population showed that rate of intraoperative rupture was significantly higher in aneurysms ≤6 mm.2 Clinical data (preoperative and postoperative) and rate of adverse events were compared between the 2 groups of patients. Anatomic results were compared between the 2 groups of aneurysms.
Categorical variables were compared using the χ2 or Fisher exact test. Continuous variables were compared using the Student t test; 95% confidence intervals of percentage were calculated by approximation to normal distribution or to binomial distribution, as appropriate. Confidence intervals were given with 1 decimal, except when the inferior limit was close to 0.0%; in these cases, 2 decimals were given. The significance threshold was set at P=0.05, except for analysis of perioperative aneurysmal perforation. For this analysis, Bonferroni correction has been applied because of multiple comparisons (3 statistical tests have been performed) and the significant threshold has been set at P=0.016. Statistical analysis was performed using SAS version 9.1 software (SAS Institute).
A total of 626 patients were included in this study: 51 (8.1%) with only aneurysms ≤3 mm and 575 (91.9%) with only aneurysms >3 mm. Table 1 presents a comparison of the clinical characteristics of patients according to the diameter of their aneurysms.
The 626 included patients harbored a total of 682 aneurysms: 51 aneurysms ≤3 mm in diameter (7.5%) and 631 aneurysms >3 mm in diameter (92.5%). Table 2 gives a comparison of aneurysmal characteristics according to their diameter.
The 682 aneurysms were treated during 665 procedures: 51 (7.7%) with only aneurysms ≤3 mm (procedure group 1) and 614 (92.3%) with only aneurysms >3 mm (procedure group 2). All patients in group 1 had only 1 procedure. In group 2, a total of 539 patients (93.7%) had 1 procedure, 33 (5.7%) had 2 procedures, and 3 (0.5%) had 3 procedures. The number of treated aneurysms by procedure was 1 in group 1 and 1.05±0.23 (1–3 aneurysms) in group 2 (P<0.0001). Selective treatment (occlusion of the aneurysm with no voluntary occlusion of the parent vessel) was performed in 51 aneurysms (100.0%) for group 1 and in 620 aneurysms (98.3%) for group 2 (P=0.34). In group 2, 11 aneurysms (1.7%) were treated by parent vessel occlusion. For selective treatment, coils alone were used in 33 aneurysms (64.7%) for group 1 and in 339 aneurysms (54.7%) for group 2 (P=0.17). The balloon remodeling technique was used in 17 aneurysms (33.3%) for group 1 and in 225 aneurysms (36.3%) for group 2 (P=0.67). Intracranial stenting was used in 1 aneurysm (2.0%) for group 1 and in 53 aneurysms (8.5%) for group 2 (P=0.11), and Trispan was used only for aneurysms in group 2 in 3 cases (0.5% vs 0.0%; P=0.62). EVT failed (procedure abandoned) in 7 aneurysms (13.7%) in group 1 and in 21 aneurysms (3.3%) in group 2 (P=0.003).
Adverse Events Related to EVT
Table 3 reports the types of specific and nonspecific complications according to aneurysm size. Five specific complications (9.8%; 95% CI, 1.5–18.1) were observed in aneurysm group 1 and 79 (12.5%; 95% CI, 9.9–15.1) were observed in aneurysm group 2 (P=0.57). The rate of perioperative aneurysmal perforation was not significantly different according to size of aneurysm (≤3 mm vs >3 mm). The difference in the rate of aneurysm rupture for aneurysms between 3 and 7 mm in diameter and aneurysm rupture for other aneurysms approached statistical significance (n=13/332; 3.9%; 95% CI, 1.8–6.0 vs n=4/350; 1.1%; 95% CI, 0.03–2.26, respectively; P=0.02). Among patients with only aneurysms ≤3 mm, the rate of specific complications for patients with aneurysms discovered after rupture of another aneurysm was not significantly higher than the rate of specific complications for patients with aneurysms discovered for other reasons (n=2/19; 10.5%; 95% CI, 1.3–33.1 vs n=3/32; 9.4%; 95% CI, 2.0–25.0, respectively; P=0.89). No nonspecific complication was observed in aneurysm group 1, but there were 15 (2.4%; 95% CI, 1.2–3.6) in aneurysm group 2 (P=0.62).
Adverse events were associated with a postoperative modification of clinical status in 38 patients: 2 (3.9% of the patients; 95% CI, 0.5–13.5) in patient group 1 and 36 (6.3% of the patients; 95% CI, 4.3–8.3) in patient group 2 (P=0.76). Death was observed for 1 patient (2.0%; 95% CI, 0.05–10.45) in patient group 1 and for 7 patients (1.2%; 95% CI, 0.3–2.1) in patient group 2 (P=0.49). Permanent neurological deficit was observed for 1 patient (2.0%; 95% CI, 0.05–10.45) in patient group 1 and for 17 patients (3.0%; 95% CI, 1.6–4.4) in patient group 2 (P=0.68). Transient neurological deficit was observed for no patients (0.0%) in patient group 1 and for 12 patients (2.1%; 95% CI, 0.9–3.3) in patient group 2 (P=0.61).
Clinical status of patients at 1 month according to the size of their aneurysm is reported in Table 4. The 1-month mortality rate was 2.0% (95% CI, 0.05–10.45; n=1) for patient group 1 and 1.4% (95% CI, 0.4–2.4; n=8) for patient group 2 (P=0.54). In group 2, 1 patient died in the month after the treatment, after hospital discharge. The 1-month morbidity rate was 0.0% (n=0) for patient group 1 and 1.9% (95% CI, 0.8–3.0; n=11) for patient group 2 (P=0.32). Some patients in groups 1 and 2 had a permanent deficit in the postoperative period, improved in the month after the treatment, and had a modified Rankin scale 0 or 1 at 1 month.
The postoperative aneurysmal occlusion results were as follows: 33 cases of complete occlusion (64.7%) for aneurysm group 1 and 361 cases (57.2%) for aneurysm group 2 (P=0.30); 7 cases of neck remnant (13.7%) for aneurysm group 1 and 146 cases (23.1%) for aneurysm group 2 (P=0.12); and 11 cases of aneurysm remnant (21.6%) for aneurysm group 1 and 124 cases (19.7%) for aneurysm group 2 (P=0.74). The rate of adequate occlusion did not differ between groups: 78.4% (n=40) in group 1 and 80.3% (n=507) in group 2 (P=0.74).
Very small aneurysms are heterogeneously defined in the literature. Yasargil initially defined “baby” aneurysms as aneurysms <3 mm.10 In their recent series, Ioannidis et al,7 referring to the Yasargil work, also defined very small aneurysms as aneurysms <3 mm. In more recent series, very small aneurysms were defined as aneurysms ≤3 mm, a threshold used in the current study.
The rate of failure of EVT was significantly higher in very small unruptured aneurysms compared to larger aneurysms (13.7% vs 3.3%, respectively). This outcome is probably related to several factors. First, because of the very small volume of some aneurysms, microcatheterization of the sac is sometimes impossible. Second, the smallest coils available were 2 mm in diameter (and now 1.5 mm), and for aneurysms between 1 and 2 mm, it is sometimes impossible to stabilize a coil inside the aneurysm sac. Finally, because the risk of rupture of very small aneurysms is probably less than that for larger aneurysms, the treatment of such lesions is probably less “aggressive.”
The rate of complications in EVT of very small aneurysms is quite heterogeneous. Suzuki et al6 reported that a series of 21 patients with very small aneurysms was successfully treated endovascularly without procedural complications. However, in most series, the rate of intraoperative rupture in very small aneurysms has been high: 4.4%,7 7.7%,3 and 11.7%.4 In these series, all or most of the treated aneurysms were ruptured. In the meta-analysis performed by Brinjikji et al,5 the intraprocedural rupture rate in unruptured very small aneurysms was 5.0% compared with 10.7% in ruptured very small aneurysms. The rate of intraoperative rupture was similar in our group of very small unruptured aneurysms (3.2%).5
Results regarding the comparison of intraoperative rupture rate in very small and larger aneurysms are also quite heterogeneous. In the Nguyen et al4 series, the rate of intraoperative rupture was 5-times higher in very small aneurysms (11.7%) compared with larger aneurysms (2.3%). In the Van Rooij et al3 series, the rate of intra-operative rupture was also significantly higher in very small aneurysms compared with larger aneurysms (7.7% and 3.6%, respectively). In our series, the rate of intra-operative rupture was also higher, but not significantly, in very small aneurysms compared with larger aneurysms (3.9% and 2.4%). In the initial published findings in the ATENA population, the rate of intra-operative rupture was reported as significantly higher in small aneurysms compared to larger aneurysms (1–6 mm, 3.7%; 7–15 mm, 0.7%; P=0.008).2 These data are not confirmed for very small aneurysms. When comparing the rate of intra-operative rupture for aneurysms >3 mm and <7 mm and other aneurysms (3.9% vs 1.1% in other aneurysms, respectively), the difference was close to statistical significance (P=0.02). This finding is probably related to the fact that the treatment of very small unruptured aneurysms is less aggressive than for greater aneurysms, as shown by the high rate of treatment failure (13.7%) in very small aneurysms, but it may also be a statistical artifact.
The rate of thromboembolic events in very small aneurysms has not been evaluated as thoroughly. Thromboembolic events occurred in 3.1% in the Ioannidis et al7 series. The morbidity rate attributable to thromboembolic events was 2.1% in the van Rooij et al series3 (2.1% in ruptured aneurysms) and 1.3% in unruptured aneurysms in the Brinjikji et al5 meta-analysis. In all of these series, comparison with larger aneurysms was not provided. In our series, the rate of thromboembolic events with or without morbidity was lower in very small aneurysms (3.9% vs 7.1% in larger aneurysms), but not significantly. In the initial analysis of the ATENA series, the rate of thromboembolic events in aneurysms <7 mm was significantly lower compared to that for aneurysms >7 mm (4.6% vs 9.9% for 1–6 mm vs 7–15 mm; P=0.005).2
In our series, morbidity at 1 month was not significantly different according to aneurysm size (0.0% for very small aneurysms compared to 1.9% for larger aneurysms). Similarly, mortality at 1 month was not significantly different between the groups (2.0% in very small aneurysms and 1.4% in larger aneurysms). In the van Rooij et al3 series, procedural morbidity also was not significantly different between the 2 groups (2.1% in very small aneurysms vs 3.4% in larger aneurysms), and neither was procedural mortality (1.1% in very small aneurysms and 2.1% in larger aneurysms). The meta-analysis performed by Brinjikji et al5 showed that morbidity and mortality related to procedural complications and postprocedural hemorrhage were lower in unruptured very small aneurysms (2.5% and 2.1%, respectively, compared to 4.0% and 5.5% in ruptured aneurysms).
The anatomic results are quite similar in very small aneurysms and larger aneurysms. In our series, the rate of adequate occlusion was 78.4% in very small aneurysms and 80.3% in larger aneurysms. Similar results were obtained in the van Rooij et al3 series (adequate occlusion in 94.9%). In the Brinjikji et al5 meta-analysis, the rate of adequate occlusion in very small aneurysms was similar (87.8% in unruptured aneurysms and 95.3% in ruptured aneurysms).
The management of very small aneurysms has to take into account the low risk of rupture of such lesions and the risk of complications, morbidity, and mortality, as reported in this series. More knowledge is probably needed to precisely define the strategy of treatment of such lesions. A good option is probably to have a regular MRI follow-up to detect any modification in size or shape of the aneurysm, which will lead to an active treatment of the aneurysm.
There are several limitations to this study. First, it is a selected population and the results may not be applicable to all patients with small unruptured aneurysms. Second, the number of patients in group 1 is low, as is the number of adverse events in both subgroups. It is probably the reason why some comparisons failed to reach statistical significance. Third, there were significant differences between patients with very small and larger aneurysms, which are likely to have had an effect on complication rates.
The rates of procedural complications, clinical outcome, and anatomic results are similar in very small unruptured aneurysms and larger aneurysms. Because the risk of rupture of very small aneurysms is low, an active treatment is probably indicated in a limited number of cases and anatomic follow-up is probably the best option for most patients.
Participating Centers and Investigators
1. CHU Larrey, Angers, France, Anne Pasco; 2. CHU Jean Minjoz, Besançon, France, Jean-François Bonneville; 3. CHU Pellegrin, Bordeaux, France, Xavier Barreau, Jérôme Berge; 4. CHU de la Côte de Nacre, Caen, France, Patrick Courthéoux, Suzana Saleme; 5. CHU Gabriel Montpied, Clermont Ferrand, France, Emmanuel Chabert, Jean Gabrillargues; 6. CHU Louis Pasteur, Colmar, Alain Tournade; 7. CHU Dijon, France, Frédéric Ricolfi; 8. CHU A. Michallon, Grenoble, France, Pierre Bessou; 9. CHU Roger Salengro, Lille, France, Xavier Leclerc, Jean-Pierre Pruvo, Christian Taschner; 10. CHU La Timone, Marseille, France, Olivier Lévrier; 11. CHU Gui de Chauliac, Montpellier, France, Alain Bonafé; 12. CHUM Notre-Dame, Montréal, Canada, Jean Raymond, Daniel Roy, Alain Weill; 13. CHU Hôpital Neurologique, Nancy, France, René Anxionnat, Serge Bracard, Luc Picard; 14. CHU G. & R. Laënnec, Nantes, France, Hubert Desal, Axel de Kersaint-Gilly; 15. CHU, Kremlin-Bicêtre, France, Hortensia Alvarez, Pierre Lasjaunias, Augustin Ozanne; 16. CH Foch, Suresnes, France, Anne Boulin, Georges Rodesch; 17. CHU La Pitié Salpêtrière, Paris, France, Alessandra Biondi, Fabrice Bonneville, Betty Jean, Nader Sourour; 18. CHU Mondor, Créteil, France, Raphaël Blanc, André Gaston; 19. CH Fondation Rothschild, Paris, France, Guido Lazzarotti, Manoel Maia Filho, Jacques Moret, Charbel Mounayer, Michel Piotin, Christiana Queiroz, Laurent Spelle; 20. CH Sainte-Anne, Paris, France, Sylvie Gordon-Hardy, Jean-François Meder, Denis Trystram; 21. CHU de la Milétrie, Poitiers, France, Jacques Drouineau; 22. CHU Maison Blanche, Reims, France, Sophie Gallas, Laurent Pierot; 23. CHU Bellevue, Saint-Etienne, France, Fabrice-Guy Barral, Luis Manera; 24. CHU Hautepierre, Strasbourg, France, Rémy Beaujeux, Fazel Boujan; 25. CH Sainte-Anne, Toulon, France, Charles Arteaga; 26. CHU Purpan, Toulouse, France, Christophe Cognard, Anne-Christine Januel, Philippe Tal; 27. CHU Bretonneau, Tours, France, Denis Herbreteau.
Sources of Funding
The ATENA study was financially supported by the French Society of Neuroradiology (SFNR).
L.P. is a consultant for Boston Scientific and EV3.
- Received April 27, 2010.
- Accepted September 7, 2010.
Ishibashi T, Murayama Y, Urashima M, Saguchi T, Ebara M, Arakawa H, Irie K, Takao H, Abe T. Unruptured intracranial aneurysms: incidence of rupture and risk factors. Stroke. 2009; 40: 313–316.
Pierot L, Spelle L, Vitry F, ATENA investigators. Immediate clinical outcome of patients harboring unruptured intracranial aneurysms treated by endovascular approach: results of the ATENA study. Stroke. 2008; 39: 2497–2504.
van Rooij WJ, Keeren GJ, Peluso JP, Sluzewski M. Clinical and angiographic results of coiling of 196 very small (≤3 mm) intracranial aneurysms. AJNR Am J Neuroradiol. 2009; 30: 835–839.
Brinjikji W, Lanzino G, Cloft HJ, Rabinstein A, Kallmes DF. Endovascular treatment of very small (3 mm or smaller) intracranial aneurysms: report of a consecutive series and a meta-analysis. Stroke. 2010; 41: 116–121.
van Rooij WJ, Sluzewski M, Beute GN, Nijssen PC. Procedural complications of coiling of ruptured intracranial aneurysms: incidence and risk factors in a consecutive series of 681 patients. AJNR Am J Neuroradiol. 2006; 27: 1498–1501.
Raymond J, Guilbert F, Weill A, Georganos SA, Juravsky L, Lambert A, Lamoureux J, Chagnon M, Roy D. Long-term angiographic recurrences after selective endovascular treatment of aneurysms with detachable coils. Stroke. 2003; 34: 1398–1403.
Yasargil MG. Pathological considerations in microneurosurgery. Stuttgart, Germany: Thieme Medical Publishers; 1987: 280–281.