J Neurosurg 66:30-34, 1987 Natural history of postoperative aneurysm rests ISAAC FEUERBERG, M.D., CHRISTER LINDQUIST, M.D., PH.D., MELKER LINDQVIST, M.D., PH.D., AND LADISLAU STEINER, M.D., PH.D. Departments of Neurosurgery and Neuroradiology, Karolinska Institute, Stockholm, Sweden v" In a series of 715 patients operated on by microsurgical techniques for intracranial saccular aneurysms between 1970 and 1980, part of the aneurysmal sac was not obliterated in 28 aneurysms in 27 patients (3.8% of 715 cases). Clinical follow-up evaluation for 8 years (range 4 to 13 years) and angiographic follow-up studies for 6 years (range 2 to 10 years) in these 27 cases revealed that one aneurysm rest increased in size and bled twice, five were spontaneously obliterated, two decreased in size, 13 remained unchanged, and in seven cases no late follow-up angiography was performed. The incidence of rebleeding from an aneurysm rest was 3.7% of the 27 in whom the sac was not obliterated and 0.14% of all 715 patients who were operated on. KEY WORDS 9 aneurysm 9 aneurysm rest 9 subarachnoid hemorrhage T HE prognostic significance of postoperative aneurysm rests is still not sufficiently documented, and their assessment and management remains controversial. Few studies have specifically addressed this problem. Drake and coworkers ~'5'6 have concluded from their investigations that aneurysm rests are dangerous and that, if possible, another operation should be performed to eradicate them. However, other reports indicate that the chances of spontaneous obliteration of the aneurysm rest are good TM and the risk of a repeat rupture is not high. ~ In a questionnaire sent to 16 neurosurgeons* we asked if they perform angiography following aneurysm surgery and how they deal with an aneurysm rest. The consensus among these surgeons seemed to be that an aneurysm rest is hazardous. However, among these surgeons only one-half routinely perform postoperative angiography and were prepared to reoperate on the patient. The other half of the group seems to be aware that they sometimes leave a rest, either intentionally or unintentionally, but they were of the opinion that a second surgical attempt would not produce better results. Because it is crucial to determine how dangerous aneurysm rests really are, we address this issue in this paper. * The following neurosurgeons provided helpful information: L. Auer, C. Drake, B. Guidetti, H. Handa, R. Heros, N. Kassell, A. Konovalov, H. Nornes, B. Pertuiset, H. Pia, R. Sengupta, B. Stein, K. Sugita, L. Symon, B. Weir, and G. Ya~argil. Clinical Material and Methods Between 1970 and 1980, 715 patients were operated on for intracranial aneurysms at the Karolinska Hospital using microsurgical technique, and in all cases early postoperative angiography was performed. The aneurysm rests found in these studies form the basis of this report. An aneurysm rest was defined as a remnant of the aneurysm proximal to the clip which still filled at angiography. The size of the aneurysm rest was defined as the maximal diameter measured on an angiogram corrected for the magnification factor. Results According to neuroradiological evaluation of the postoperative angiograms, there were 33 aneurysm rests in 32 patients. Follow-up angiography, however, revealed five of the rests to be a false-positive finding (Fig. 1). Misdiagnosis was due to spasm in two cases and to a loop in the parent vessel in three. After excluding these five cases, there remained 28 true aneurysm rests in 27 patients, giving an incidence of 3.8% of the 715 operated cases. In 12 cases the surgeon was aware that an aneurysm rest remained; reasons for not totally clipping the aneurysm included avoiding occlusion of the parent artery in eight cases, a broad base in two cases, and the presence of atherosclerotic plaque in two cases. In two additional cases the surgeon stated that he left an aneurysm rest but did not explain why. Partial clipping in these 14 cases was confirmed on the initial postopera- 30 J. Neurosurg. / Volume 66/January, 1987
i Natural history of postoperative aneurysm rests Case Preop Postop Fallow up ] Case Preop Poata; Fallow Up I0 IS U FnG. 1. Tracings of the best projections from the pre- and postoperative and from the follow-up angiograms in five cases of false-positive neuroradiological diagnosis of aneurysm rests. The causes of the incorrect finding were a loop in the parent artery in Cases 8, 17, and 20 and spasm in Cases 10 and 14. TABLE 1 Location of the 28 aneurysm rests in 27 cases Location of Aneurysm No. of Rests anterior communicating artery 12 middle cerebral artery 10* posterior communicating artery 2 internal carotid bifurcation 1 ophthalmic artery 1 pericallosal artery 1 posterior inferior cerebellar artery 1 total rests 28 * One patient (Case 1) had two aneurysm rests. tive angiogram. In six other cases the surgeon reported total clipping of the aneurysm, but subsequent arteriograms revealed that this was not so. In eight cases the operative report did not reveal whether the surgeon was aware of leaving an aneurysm rest or not. The locations of incompletely clipped aneurysms are given in Table 1. The diameter of the aneurysm rests was less than 4 mm in 15 and 4 mm or larger in 13. Drawings derived from the preoperative, postoperative, and follow-up angiograms of the patients with aneurysm rests are shown in Figs. 2 to 5. Follow-up angiography revealed that five rests had disappeared (Fig. 2), two had decreased in size (Fig. 3), one had increased in size (Fig. 4), and 13 had remained unchanged (Fig. 5). In seven FIG. 2. Drawings of the changes observed on angiography in five cases in which the postoperative aneurysm rests no longer filled on the follow-up angiograms. Case Preop Poatop Follow up FIG. 3. Tracings of changes in two cases where the aneurysm rests decreased in size on the follow-up angiograms. FIG. 4. Tracings of a right middle cerebral artery aneurysm showing that the postoperative aneurysm rest has increased in size. cases (Cases 25 to 31), late follow-up angiography could not be performed. Table 2 indicates the changes in rest size compared to aneurysm location. Table 3 shows the change in size of the aneurysm rest between the first postoperative angiogram and the follow-up angiogram. J. Neurosurg. / Volume 66~January, 1987 31
I. Feuerberg, et al. TABLE 2 Location and changes in size of aneurysm rests Follow-Up Size of Rest Location of Aneurysm Disap- Dimin- Un- Inpeared ished changed creased anterior communicating 0 I 8 0 artery middle cerebral artery 1 0 5 1 ophthalmic artery 1 0 0 0 posterior communicating 1 1 0 0 artery internal carotid bifurcation 1 0 0 0 pericallosal artery 1 0 0 0 TABLE 3 Size of aneurysm rests on late follow-up angiography compared to their initial postoperative size Initial Size Follow-Up Size of Rest Disap- De- Un- Inpeared creased changed creased Total <4mm 4 0 4 1 9 _> 4mm 1 2 9 0 12 The case material is summarized in Table 4. Two illustrative cases are given below. Illustrative Case Reports Case 5 This 51-year-old man underwent surgery on a right middle cerebral artery (MCA) aneurysm (Fig. 5). Nine years postoperatively subarachnoid hemorrhage (SAH) recurred. Angiography at that time revealed no change in the size or shape of the aneurysm rest (Fig. 5). At an operation performed on Day 13 following the SAH, there was no sign of hemorrhage from the aneurysm rest and the clip was left untouched. One year later the patient was in good condition. The SAH in this case obviously occurred from an unknown source in a patient with an aneurysm rest from previous surgery. Case 15 This 35-year-old woman was operated on for a right MCA aneurysm. Five years postoperatively new SAH occurred, and angiography disclosed enlargement of the aneurysm rest (Fig. 4). At surgery a rift was identified proximal to the old clip. A new aneurysm neck was formed by bipolar coagulation and a clip applied. During the postoperative course the patient developed a temporal lobe hematoma, presumably from the aneurysm site, and this was evacuated. The aneurysm was then excluded from the circulation by trapping. This resulted in a dense left-sided hemiparesis that has partially resolved. Discussion The neuroradiological evaluation of postoperative angiograms in 32 patients with 33 aneurysm rests gave false-positive results in five instances. This suggests that the identification of aneurysm rests can be difficult. Spasm and loops in the parent vessels may simulate an aneurysm rest. The frequency of errors could be decreased by delaying the postoperative angiography or repeating it in cases with spasm. The 3.8% incidence of aneurysm rests in the 27 cases in this series following microsurgical aneurysm clipping compares favorably with the results presented in the first article on the subject by Allcock and Drake. ~ However, their study was published 23 years ago and Dr. Drake's surgical results are among the best today. From our correspondence with 16 leading aneurysm surgeons, we have not been able to obtain specific information on the incidence of aneurysm rests in their cases. Estimates ranging from 1% to 5% are given by those among the 16 who perform postoperative angiography. However, eight of the 16 surgeons did not find postoperative angiography to check the results of clipping to be justified, but it was implicit in their answers to our questionnaire that complete clipping is sometimes not achieved. For example, one surgeon stated: "I perform a combination of clipping and coating with fibrin glue." Another wrote: "We have not observed on angiograms enough significant rests to justify a new operation." A third stated: "We routinely open the aneurysm at surgery to make sure that it has been clipped satisfactorily." Our present study shows that establishing the presence of an aneurysm rest is not easy for the surgeon or the neuroradiologist. Opening the aneurysm following clipping does not safeguard against aneurysm rests, since this procedure has routinely been carried out at our institution. If the surgeon wants to confirm an impression that no remnant of the aneurysm has been left, it is necessary to perform angiography. The radiological assessment is often difficult and requires a competent examiner. A wealth of previous observations points to a relationship between aneurysm size and the risk of rupture. 3,4,9'1~ The propensity for an aneurysmal sac to rupture is related to the stress on the aneurysm wall and to factors promoting thrombosis. Stress on the aneurysm wall increases with intra-aneurysmal pressure, with the radius of the aneurysm, and with a decrease of the minimal thickness of the wall. 7 Thrombosis is more likely to occur in an aneurysm that has a larger diameter than that of its orifice. Specific sites on the aneurysm sac seem to be more prone to rupture? In Crawford's series, 3 the rupture occurred at the apex of the aneurysm in 64% of autopsied cases, at the body in 17%, and at the neck in 2%. Consequently, a clip placed proximal to the apex should decrease the risk of rebleeding, but it is obviously more effective when placed proximal to the site of rupture. In our series, spontaneous rest obliteration in all cases but one occurred in association with small rests. On the other hand, the only rest that increased in size and rebled occurred in a case in this group. Thus, a benign 32 J. Neurosurg. / Volume 66/January, 1987
Natural history of postoperative aneurysm rests Preep Pustop aad folluw up [c*su Preep Peste; and (ellew up? 5 22 ~3 3~ FIG. 5. Drawings in 13 cases in which the aneurysm rests remained unchanged in the postoperative and follow-up angiograms. prognosis was not related to the size of the rest, even though the small rests became obliterated spontaneously more frequently than the larger ones. This observation suggests a difference between the natural history of the postoperative aneurysm rests as compared to that of the untreated ruptured and even unruptured aneurysms. The cases of postoperative aneurysm rests can be considered as symptomatic aneurysms converted into asymptomatic aneurysms. The risk of rupture of an asymptomatic aneurysm is estimated at roughly 1% per year. 8"13 Among our cases, we observed recurrent hemorrhages in only one patient during a total observation time of 266 years. If we consider only the rests that are still patent, the corresponding observation time is 126 years. Thus, the risk of rebleeding from an aneurysrn rest is 0.38% to 0.79% per year during the observation period (4 to 13 years). These figures could be compared to the risks of reoperation following "failed aneurysm surgery" as given by Drake, et al., 5 in a recent publication; 7 that is, an operative morbidity rate of 7% and a mortality rate of 5%. In contrast, it must be kept in mind that the observation period in our series is relatively short and the final balance of the natural history of aneurysm rests is not yet available. In conclusion, although the incidence of aneurysm rests following microsurgery is small, it is not negligible. During a total postoperative observation time of 266 years in these cases, rebleeding occurred in only one aneurysm rest. Thus, the risk of rupture of an incompletely clipped aneurysm would seem low and reoperation hardly justified. Nevertheless, a risk of rupture of 0.38% to 0.79% per year makes the cumulative risk one of importance, at least in younger patients. The truism that natural history should be weighed against surgical skill is obviously particularly valid in these cases. J. Neurosurg. / Volume 66/January, 1987 33
I. Feuerberg, et al. TABLE 4 Summary of clinical data in 32 cases with aneurysm rests* Case Aneurysm Angiographic Findings Angiography New Clinical Follow-Up Follow-Up No. Location Postop Follow-Up? Period (yrs) SAH Period (yrs) 1 rt MCA 5-mm rest unchanged 10 0 13 It MCA 2-mm rest not visualized 10 0 13 2 ACoA 3-mm rest unchanged 10 0 12 3 ACoA 4-ram rest unchanged 9 0 12 4 ACoA 3-ram rest unchanged 9 0 11 5 rt MCA 5-mm rest unchanged 9 1 10 6 ACoA 7-ram rest unchanged 8 0 10 7 It PCoA 4-ram rest 2-mm rest 8 0 8 8 rt MCA false-positive not visualized 7 0 9 9 It ophthalmic 3-mm rest not visualized 6 0 9 10 It MCA false-positive not visualized 6 0 9 11 ACoA 5-ram rest unchanged 6 0 8 12 rt MCA 4-ram rest unchanged 5 0 7 13 ACoA 3-ram rest unchanged 5 0 7 14 It MCA false-positive not visualized 5 0 7 15 rt MCA "little" rest increased 5 2 6 16 It pericallosal 3-mm rest not visualized 4 0 7 17 ACoA false-positive not visualized 4 0 6 18 It MCA 5-ram rest unchanged 4 0 6 19 ACoA 5-ram rest 4-mm rest 3 0 6 20 ACoA false-positive not visualized 3 0 6 21 rt int car bifurc 9-mm rest not visualized 3 0 6 22 ACoA 5-mm rest unchanged 3 0 5 23 It MCA 3-ram rest unchanged 2 0 5 24 It PCoA 2-mm rest not visualized 2 0 4 25 rt MCA 3-ram rest -- 0 13 26 rt MCA 3-ram rest -- 0 11 27 It MCA 3-ram rest -- 0 11 28 ACoA 3-mm rest -- 0 10 29 ACoA 4-ram rest -- 0 8 30 ACoA 3-ram rest -- 0 5 31 It PICA 3-mm rest -- 0 6 32 ACoA 5-mm rest unchanged at reop 89 * MCA = middle cerebral artery; ACoA = anterior communicating artery; PCoA = posterior communicating artery; int car bifurc = internal carotid bifurcation; PICA = posterior inferior cerebellar artery; SAH = subarachnoid hemorrhage. t Six patients (Cases 25 to 30) refused follow-up angiography, and the patient in Case 31 died of a pulmonary embolus prior to follow-up angiography. References i. Allcock JM, Drake CG: Postoperative angiography in cases of ruptured intracranial aneurysm. J Neurosurg 20: 752-759, 1963 2. Bonnal J, Stevenaert A: Thrombosis of intracranial aneurysms of the circle of Willis after incomplete obliteration by clip or ligature across the neck. J Neurosnrg 30: 158-164, 1969 3. Crawford T: Some observations on the pathogenesis and natural history of intracranial aneurysms. J Neurol Neurosurg Psychiatry 22:259-266, 1959 4. Crompton MR: Mechanism of growth and rupture in cerebral berry aneurysms. Br Med J 1:1138-I 142, 1966 5. Drake CG, Friedman AH, Peerless S J: Failed aneurysm surgery. Reoperation in 115 cases. J Neurosurg 61: 848-856, 1984 6. Drake CG, Vanderlinden RG: The late consequences of incomplete surgical treatment of cerebral aneurysms. J Neurosurg 27:226-238, 1967 7. Ferguson GG: Physical factors in the initiation, growth, and rupture of human intracranial saccular aneurysms. J Neurosurg 37:666-677, 1972 8. Heiskanen O: Risk of bleeding from unruptured aneu- rysms in cases with multiple intracranial aneurysms. J Neurosurg 55:524-526, 1981 9. Kassell NF, Torner JC: Size of intracranial aneurysms. Neurosurgery 12:291-297, 1983 10. McCormick WF, Acosta-Rua G J: The size of intracranial saccular aneurysms. An autopsy study. J Neurosurg 33: 422-427, t970 11. Sato S, Suzuki J" Prognosis in cases of intracranial aneurysm after incomplete direct operations. Acta Neurochir 24:245-252, 1971 12. Suzuki J, Ohara H: Clinicopathological study of cerebral aneurysms. Origin, rupture, repair, and growth. J Neurosurg 48:505-514, 1978 13. Winn HR, Berga SL, Richardson AE, et al: Long-term evaluation of patients with multiple cerebral aneurysms. Ann Neurol 10:106, 1981 (Abstract) Manuscript received November 21, 1985. Accepted in final form June 5, 1986. Address for Dr. Feuerberg: Hospital de Clinicas, Caracas, Av. Alameda, San Bernardino, Caracas 1010, Venezuela. Address reprint requests to: Ladislau Steiner, M.D., Ph.D., Department of Neurosurgery, Karolinska Hospital, S-104 01 Stockholm, Sweden. 34 J. Neurosurg. / Volume 66/January, 1987