Department of Neurological Surgery, The George Washington University Medical Center, Washington, District of Columbia
OBJECTIVE: In this report, we evaluate the recurrence- and progression-free survival of patients with meningiomas involving the cavernous sinus.
METHODS: The rates of tumor recurrence and progression in 119 patients with meningiomas involving the cavernous sinus treated between 1983 and 1993 were analyzed.
RESULTS: The mean follow-up period was 33.8 months. There were seven recurrences of completely resected tumors (7 of 73, 10%) and seven progressions of incompletely resected tumors (7 of 46, 15%). Life table analysis of recurrence- and progression-free survival was performed. The recurrence-free survival rate was 94% at 3 years and 81% at 5 years. In contrast, the progression-free survival rate was 87% at 3 years and 62% at 5 years (P = 0.0456). There were two patterns of recurrence or progression. The first group of tumors had an aggressive biological behavior with growth at multiple areas of resection; the second group showed growth at or near the margins of resection. Functional status continued to improve slowly during the follow-up period; however, there was no statistically significant difference between the Karnofsky scores obtained 3 to 12 months postoperatively and the follow-up Karnofsky scores (obtained >1 yr postoperatively).
CONCLUSION: Surgery for meningiomas involving the cavernous sinus seems to provide excellent tumor control. Our experience suggests that for the majority of these tumors, complete tumor resection can be performed with acceptable risks of morbidity and mortality and seems to increase the duration of recurrence-free survival. The long-term quality of life of surviving patients is satisfactory, with the great majority of the patients being independent.
(Neurosurgery 39:915920, 1996)
Key Words: Cavernous sinus, Meningioma, Recurrence
Despite the large number of recent publications on cavernous sinus surgery, only a few authors have evaluated the problem of recurrence. Most of the previously published results present a recurrence rate irrespective of the follow-up period. Meningiomas are the most common tumors affecting the cavernous sinus. Patients with meningiomas involving the cavernous sinus can be followed up for long periods of time, because these tumors are benign. Meningiomas are known to have late recurrence, so it is important to study the recurrence- and progression-free survival in these patients.
Recurrence was defined as the radiological appearance of the tumor after a radiographically confirmed gross total surgical resection. Progression was defined as the radiological documentation of an increase in size of an incompletely resected tumor. Tumors were classified according to their extension as "confined" or "extensive." Confined tumors involve the cavernous sinus and areas immediately adjacent to it and measure less than 3 cm at their greatest diameter. Extensive tumors involve multiple areas of the cranial base and measure at least 3 cm at their greatest diameter. Tumors were also classified according to the degree of cavernous sinus invasion into five grades (11). Grade I tumors involve only one area of the cavernous sinus. Grade II tumors involve multiple areas of the cavernous sinus and either displace the intracavernous internal carotid artery (ICA) or encase it partially. Grade III tumors encase the intracavernous ICA totally. Grade IV tumors encase and narrow the intracavernous ICA. Grade V tumors involve both cavernous sinuses with varying degrees of vascular encasement.
A standard life table analysis was used to compute the risk of recurrence and progression (1, 3). The log-rank test of survival time was used to test for statistical difference in survival curves. The Mann-Whitney U test was used to compare the Karnofsky performance scale scores.
|Criteria||No. of Patients||Percent|
|Lost to follow-up||1||1|
|Tumor size and extension|
|Extent of tumor resection|
|Cavernous sinus grade|
|Deaths during follow-up||14||12|
A gross total resection (defined as complete excision according to the surgeon's impression and postoperative magnetic resonance imaging scans) was accomplished in 61% of the patients. Among the patients with confined tumors, a gross total resection was accomplished in 72%; of the patients with extensive tumors, a gross total resection was accomplished in 58%. A gross total removal was accomplished in 32 (55%) of the 58 patients without any prior therapy, 32 (62%) of the 52 patients with prior surgery, and 5 (56%) of the 9 patients with prior radiotherapy. There was no statistically significant relationship between prior surgery, prior radiation, and the degree of tumor resection.
The most frequent complication was cerebrospinal fluid leakage, which occurred in 25 patients (21%). Six patients (5%) suffered cerebral infarction, two of which were related to the intracavernous ICA (one because of vein graft occlusion and one related to preoperative ICA occlusion and subsequent embolism). Four patients (3%) suffered brain hematoma or contusion. Five patients (4%) suffered infection. Seventeen patients (14%) had pituitary dysfunction, usually transient. Complications were more frequent in patients who had undergone surgery for their tumors and in patients with extensive tumors.
Of 119 patients, 1 was lost to follow-up. Fifty-eight percent of the patients had been followed for at least 3 years, and 21% of the patients had been followed for at least 5 years. The mean follow-up period was 33.8 months. Fourteen patients died during the follow-up period, but only four of the deaths were a result of tumor progression (Table 2). Seven of 73 patients suffered tumor recurrence after complete resection (10%), and 7 of 46 patients suffered a tumor progression after incomplete resection (15%).
|Cause of Death||No. of Patients|
|Disabilities after operation||2|
|Acquired immune deficiency syndrome (not related to surgery)||1|
|Liver failure (not related to surgery)||1|
The recurrence-free survival rate was 94% at 3 years and 81% at 5 years, and the progression-free survival rate was 87% at 3 years and 62% at 5 years (Fig. 1). This difference was statistically significant (P = 0.0456). In 50% of the patients with recurrent or progressive disease, the growth was at multiple areas of the previous resection (Fig. 2). In the other patients, the growth occurred at the margins of the previous resection (Fig. 3). Of the patients who suffered a recurrence or progression of their disease, 86% had extensive tumors initially. However, there was no significant difference between extensive and confined tumors and recurrence or progression rates. The extent of tumor resection and recurrence or regrowth according to cavernous sinus grade is shown in Table 3. A smaller percentage of total resections were accomplished in patients with more advanced grades of tumor (ICA completely encased or bilateral cavernous sinus invasion). Patients with more advanced cavernous sinus grades tended to have more extensive tumors, and the lower percentage of total resections reflects this as well. The recurrence/progression rates were not statistically different in these patients.
|Cavernous Sinus Grade||No. of Patients||Resection (%)||Recurrence or Progression (%)|
|III||26||Total: 18 (69)||2 (11)|
|Incomplete: 8 (31)||1 (13)|
|IIIV||93||Total: 55 (59)||5 (9)|
|Incomplete: 38 (41)||6 (16)|
Postoperative radiotherapy was administered to 17 patients (14%), all of whom had incomplete tumor resection. Seven of the 17 patients experienced tumor regrowth. Of the seven patients who received external beam radiation, four died as a result of tumor regrowth. One of the seven patients who underwent gamma knife radiosurgery experienced regrowth, and two of the three patients who underwent LINAC radiosurgery experienced regrowth. The follow-up for this group had a mean of 35 months and a median of 31 months (range, 472 mo). One patient with a malignant meningioma received postoperative chemotherapy.
An ICA to middle cerebral artery or a petrous to supraclinoid ICA saphenous vein bypass graft was used in 20 patients. Nineteen of these were for tumors with encasement of the intracavernous ICA, with or without narrowing (Grades IIIV); the other was for a patient with an extensive Grade II tumor in whom the intracavernous ICA was injured during tumor resection. A gross total resection was accomplished in 16 of these patients (80%). There have been no recurrences in the patients with gross total resection, but two of the four patients with incomplete resection have experienced tumor recurrence. Both of these tumors were biologically aggressive tumors that regrew in multiple areas of the cranial base.
The median preoperative Karnofsky score was 90 (range, 40100), and the median postoperative Karnofsky score was 80 (range, 40100). The median Karnofsky score during the follow-up period was 80 (range, 40100), including the patients who suffered recurrence or progression. Although many patients continued to improve slowly during the follow-up period, there was no statistically significant difference between the postoperative and follow-up Karnofsky scores.
The recurrence rate of meningiomas after surgical treatment depends on the extent of the surgical resection and the rate of tumor regrowth (6). Meningiomas that do not totally encase the intracavernous ICA (Grades III) are much easier to resect, because they can be peeled from it, whereas tumors that do not extensively encase the intracavernous ICA (Grades IIIV) often require a vein graft with subsequent excision of the intracavernous ICA (11). Using life table survival analysis, the series from Massachusetts General Hospital showed recurrence or progression rates for parasellar meningiomas and sphenoid ridge meningiomas to be 19 and 34%, respectively, at 5 years (9). This represents an 81% recurrence- or progression-free survival for parasellar meningiomas and a 66% recurrence- or progression-free survival for sphenoid ridge meningiomas at 5 years. Although this series has a long follow-up, it includes only 28 parasellar meningiomas and 36 sphenoid ridge meningiomas. Mirimanoff et al. (9) found that the extent of resection and the location of the tumor were the factors most associated with recurrence. Sphenoid ridge and parasellar meningiomas were two of the most common tumors with high risk of recurrence or progression because of the high rate of incomplete resection. They obtained a 28% complete resection in sphenoid ridge meningiomas and a 57% complete removal in parasellar meningiomas. DeMonte et al. (4) reported three recurrences in their series of 28 completely resected cavernous sinus meningiomas. They did not use life table survival analysis.
Continued improvements in operative techniques and intraoperative neurophysiological monitoring have enabled the complete removal of the majority of cavernous sinus tumors with acceptable morbidity, thus decreasing the risk of recurrence. In our series, we obtained complete resection in 61% of the patients. DeMonte et al. (4) reported a 76% complete resection rate. Any study that reports recurrence-free rates should have a minimum follow-up that exceeds the expected doubling time of the tumor analyzed (10). The volume-doubling time of benign intracranial meningiomas ranges from 138 to 1045 days (median, 298 d; mean, 415 d) (7). Periodic radiographic evaluation is necessary to assess recurrence, because even completely resected tumors can recur in the long run. Also, subtotally resected meningiomas can continue to grow despite radiation therapy.
Our series has the largest reported number of patients with meningiomas involving the cavernous sinus and has a mean follow-up of 33.8 months. We had seven recurrences and seven progressions of the disease. Life table survival analysis of our series shows that surgery for meningiomas involving the cavernous sinus seems to provide excellent tumor control that is much better than that reported in previously published series of tumors in this area. However, precise comparison is impossible because of many uncontrolled variables, such as tumor size, prior treatment, tumor biology, and the nature of the follow-up, especially the failure to use MR imaging.
During our follow-up, the 14 patients had two patterns of tumor recurrence or progression. Half of them showed tumor growth at multiple areas of the previous resection. This suggests an aggressive biological behavior that was confirmed surgically as having been caused by the tumor extension, although the pathological findings were not changed. The other half showed growth only at the margins of resection. This suggests that either tumor was left behind at the margins or that the cytogenetic process creating tumor growth was still active. DeMonte et al. (4) also suggested a change in the tumor biology to a more aggressive form in those cases of early recurrence. Leonetti et al. (8) have described three patterns of tumor recurrence (medial, inferior, and posterior) in relation to tumor location along the cranial base and the neuro-otological and cranial nerve deficit presentation.
Duma et al. (5) reported that in six patients with meningiomas of the cavernous sinus for whom radiosurgery was the only treatment method, no tumor growth was found in a 26-month follow-up period. They suggested stereotactic radiotherapy as an effective adjuvant or alternative therapy to the microsurgical removal of cavernous sinus meningiomas. We are also using radiosurgery as an alternative or adjuvant modality of treatment, based on the physiological condition, symptoms, and preferences of the patient. However, the results of stereotactic radiosurgery will need to be reevaluated after a longer follow-up is achieved. Our results show that the progression-free survival decreases after 5 years. In our patients, recurrence or progression was treated by subsequent surgery alone or in combination with radiotherapy. In one patient, the initial cell biology appeared aggressive, so chemotherapy was also administered. This patient's tumor was infiltrating the bone, muscle, and lymphatic and perineural tissue.
DeMonte et al. (4) do not recommend carotid artery excision for meningiomas involving the cavernous sinus. Their argument is that carotid artery resection does not improve the oncological removal, because microscopic tumor deposits may remain along the cranial nerves. Our experience using vein grafts in these patients suggests that there is improvement in their long-term recurrence-free survival. In 95% of the patients in whom vein grafting was used, tumor was encasing or stenosing the intracavernous ICA. In these patients, we have had no recurrences and only two progressions (manifested in two biologically aggressive tumors).
How to treat recurrent tumors is an important issue. In our group, we have treated them by surgery, radiotherapy, and combinations thereof. One patient with a very aggressive biology was treated with chemotherapy. In the future, an initial assessment of tumor biology may be a better guide to further treatment.
Received, March 5, 1996.
Accepted, June 12, 1996.
Reprint requests: Laligam N. Sekhar, M.D., F.A.C.S., Department of Neurological Surgery, The George Washington University Medical Center, 2150 Pennsylvania Avenue NW, Suite 7-420, Washington, DC 20037.
The first conclusion made by the authors is that "Surgery for meningiomas involving the cavernous sinus seems to provide excellent tumor control." It will take a much longer period of follow-up to confirm this statement. Already in this relatively short follow-up period, 10% of the patients experienced tumor recurrence after radiographic confirmation of total removal. Among those who underwent incomplete resection, 15% showed growth. We do not know how many of these tumors had documented evidence of growth before surgery.
The second conclusion reported is, "Our experience suggests that for the majority of these tumors, complete tumor resection can be performed with acceptable risks of morbidity and mortality " When surgery is indicated for these tumors, is complete tumor removal with internal carotid artery resection and bypass, in some cases, better than extensive but incomplete removal that spares cranial nerves and the internal carotid artery? The authors report only the total number of serious complications, which included a 5% incidence of cerebral infarction and a 3% incidence of brain hematoma and contusion. We do not know how many of the complications were related to complete tumor resection.
The third conclusion is that complete tumor resection " seems to increase the duration of recurrence-free survival." The statistical analysis shows a trend in this direction, but a longer follow-up of more patients is needed to establish this point. We do not know whether this interval of recurrence-free survival will be long enough to outweigh the risks of complete resection.
The last conclusion is, "The long-term quality of life of surviving patients is satisfactory, with the great majority of the patients being independent." However, only Karnofsky scores are provided, and they did not change over the follow-up period. The most common complication of cavernous sinus surgery is extraocular muscle impairment, and it can significantly impact quality of life, as can loss of function in the trigeminal nerve causing neurotrophic keratitis.
Reports such as this are important in increasing our knowledge about how to manage patients with cavernous sinus meningiomas. As more information about patients being followed up becomes available, we will have a better idea of the natural history of these tumors and will be able to compare these results to those of radiation therapy.
Robert G. Ojemann
The authors report an exceptionally large experience with meningiomas involving the cavernous sinus. Although the title of the article implies a long-term follow-up of these patients, the mean follow-up is only a little longer than 2.5 years; unfortunately, this article does not define the absolute recurrence and progression rates of this tumor. It does, however, serve as a benchmark for aggressive surgical treatment for meningiomas of the cavernous sinus followed up during this period of time. It will be important to continue to follow up this group of patients to find out what the absolute 5- and 10-year recurrence and progression rates are.
There is little discussion in this article regarding complications in this group of patients, and when considering absolute numbers of progression and recurrence, one must also consider the complications involved with various choices of treatment. The 16 patients in this group who underwent radiation therapy are not a large enough number to compare with the remaining nonradiated patients. It will take a number of years to determine whether aggressive surgical treatment, including carotid bypass grafting, is superior overall to less aggressive surgical treatment combined with some form of focused radiation therapy.
William F. Chandler
Ann Arbor, Michigan
De Jesús et al. evaluate tumor recurrence and tumor progression in 119 patients with cavernous sinus meningiomas after aggressive surgical resection. Life table analysis is used to calculate recurrence-free survival in patients with completely resected tumors and progression-free survival in those with incompletely resected tumors. In a mean follow-up period of 34 months, the 5-year recurrence-free survival rate was 81%, and the progression-free survival rate was 62%.
This series also provides useful data regarding overall functional status after surgery. Karnofsky scores are useful as an objective measure of functional independence. However, the surgical morbidity rate (especially cranial nerve deficit) is not presented. The impact of new cranial nerve deficits on "quality of life" is difficult to measure. An "acceptable risk of morbidity" varies with the particular clinical situation.
Although a large number of patients were included in this study and 58% were followed up for longer than 3 years, it is not clear how recurrent/progressive disease was detected. Most neurosurgeons evaluate patients with incompletely resected tumors for disease progression at more frequent intervals. This could significantly bias the determination of the progression-free interval. Such uncertainty exists in any retrospective study, but this must be kept in mind when making conclusions about the effectiveness of aggressive surgical resection or stereotactic radiosurgery.
As the authors discuss, stereotactic radiosurgery is frequently used as an adjuvant or alternative treatment. The preference at the University of Vienna over the past several years has been surgical resection of most cranial base meningiomas, without aggressive attempts at total removal of tumor within the cavernous sinus. Stereotactic radiosurgery is then used for the treatment of the intracavernous tumor. No conclusions can be made in a 4-year follow-up period other than that treatment-related cranial nerve deficits were significantly reduced.
Wolfgang T. Koos
The authors present a study of patients during the recurrence- and progression-free intervals after surgical treatment of parasellar meningiomas. A series of more than 100 meningiomas treated surgically during a period of 10 years offers a large enough sample for study, with respect to recurrence- and progression-free time intervals. It is evident that this kind of study is not of value unless the volume of patients is large enough and the follow-up period long enough. Unfortunately, these two factors were not considered to be important in the past, and confusion arose in the field of surgical treatment of parasellar meningiomas.
In the presented series, only half of the patients were monitored for 3 years after treatment and only one-fifth of them for 5 years. The mortality rate among the patients operated on for parasellar meningiomas in the presented series is high, although it is stated that most of the patients did not die as a result of tumor progression. I think that grading meningiomas in the parasellar areas allows ambiguity; however, any grading used in a larger series and by the same author might be useful. Another questionable factor is grading patients with parasellar meningiomas according to Karnofsky, because most patients have local (transient) deficits.
An important question, which is addressed in this article, is that of encasement of the ICA. The authors' statement that the en bloc resection of the encased and "stenotic" ICA and grafting of the ICA are superior to the partial resection of the tumor around the ICA is very important and should be carefully studied in future.
Many questions still remain unanswered, i.e., when radiosurgery should be used after incomplete resection, what incomplete resection is, and how to treat recurrent meningiomas (by which modality or combination of modalities).
However, it is certain that total resection of a meningioma in this area, like anywhere else, is superior to any other modality. With the accumulated knowledge of and experience with surgical anatomy, morbidity after direct surgical treatment is becoming less and less significant and the percentage of patients without postoperative morbidity is steadily increasing.
Vinko V. Dolenc
Return to Table of Contents
Return to the Neurosurgery Home Page