Therapy of meningeal melanocytomas

Authors


Abstract

BACKGROUND

Meningeal melanocytomas are rare, benign central nervous system lesions with a high probability of recurrence. To the authors' knowledge, approximately 100 cases have been reported since 1972, when the entity first was described. In the current study, four therapies were compared with regard to local control and survival to identify which is best.

METHODS

All reported cases were reviewed with regard to extent of resection, radiotherapy, local control, and survival. If published data were incomplete, the authors were contacted for additional data. Patients were categorized by therapy: complete resection (CTR), complete resection followed by radiotherapy (CTR-RT), incomplete resection (ITR), and incomplete resection followed by radiotherapy (ITR-RT). Local control and survival were calculated using the Kaplan–Meier method. A multivariate analysis was performed including age, gender, tumor location, confirmation of extent of resection, and treatment schedule.

RESULTS

A total of 89 patients were included: 46 with CTR, 3 with CTR-RT, 23 with ITR, and 17 with ITR-RT. The 5-year local control rate was 80% after CTR, 100% after CTR-RT, and 72% after ITR-RT versus 18% after ITR (P < 0.001). The 5-year survival rate was 100% after CTR, CTR-RT, and ITR-RT, respectively, versus 46% after ITR (P < 0.001). Multivariate analysis demonstrated that therapy was the only variable that affected local control and survival significantly. In 14 patients treated with ITR-RT, RT was restricted to the tumor region. The radiation dose was 30–40 grays (Gy) in 6 patients, 45–55 Gy in 7 patients, and unknown in 1 patient. The 5-year local control rate was 86% after a dose of 45–55 Gy versus 27% after a dose of 30–40 Gy (P = 0.1).

CONCLUSIONS

CTR was found to be significantly superior to ITR with regard to both local control and survival. Outcome was significantly improved by RT after ITR. Doses of 45–55 Gy appeared to be more beneficial than doses of ≤ 40 Gy. Cancer 2004. © 2004 American Cancer Society.

In 1972, Limas and Tio1 introduced the term “meningeal melanocytoma.” These very rare central nervous system lesions appear macroscopically as well circumscribed or encapsulated pigmented tumors. On light microscopy, they are comprised of variably shaped cells, arranged in bundles, nests, or fascicles. The cellular cytoplasm contains melanin pigment. Nuclei are large with prominent nucleoli. Mitotic figures are very rare or absent and necrosis or hemorrhage usually are not observed.2–17 Ultrastructural studies revealed that meningeal melanocytomas are derived from leptomeningeal melanocytes. Melanosomes appear at different stages of maturation with absent or mild cellular or nuclear pleomorphism. The zonula adherens rarely is present, and the basal lamina around neoplastic cells is poorly developed.7, 8, 15, 18–20

Staining for HMB-45 (antimelanocytic antibody), Melan A (anti-a-melanocytic protein product of gene mart-1), vimentin, and S-100 protein usually is positive. Staining for glial fibrillary acidic protein (GFAP), neuron-specific enolase (NSE), epithelial membrane antigen (EMA), and cytokeratin usually is negative.2, 5–13, 15, 17–19, 21–23

On computed tomography (CT) scans, meningeal melanocytomas appear as well defined, isodense to slightly dense masses with homogeneous enhancement of contrast medium. Magnetic resonance imaging (MRI) findings are nonspecific with isointensity (with gray matter) or increased signal intensity on T1-weighted images and low intensity, isointensity, or high intensity on T2-weighted images.10, 15, 24–30

Meningeal melanocytomas are generally considered benign with a high probability of recurrence. To our knowledge, < 100 cases have been reported in the literature since 1972.1 This retrospective analysis compares different therapeutic approaches for local control and survival to define the optimal treatment for this very uncommon tumor.

MATERIALS AND METHODS

All cases reported since 19721 were reviewed with regard to age, gender, location, extent of resection, radiotherapy, local control, and survival. If the published data were incomplete, the authors were contacted for additional data, which were available in approximately 60% of the series.

The patients were divided into four groups by therapy: complete tumor resection alone (CTR), complete tumor resection followed by radiotherapy (CTR-RT), incomplete tumor resection alone (ITR), and incomplete tumor resection followed by radiotherapy (ITR-RT). Minimal follow-up for inclusion was 12 months. Local control and survival were calculated for the four groups using the Kaplan–Meier method.31 The log-rank test was used to compare the outcome of the various therapeutic options.

Furthermore, univariate and multivariate analyses (multiple logistic regression analysis, P value for inclusion of 0.05 and P value for exclusion of 0.10) were performed to investigate the possible impact on local control and survival of age, gender, location of meningeal melanocytoma (brain vs. spinal cord), confirmation of extent of resection (CT/MRI vs. surgeon's impression alone), and treatment schedule (CTR, CTR-RT, and ITR-RT vs. ITR).

RESULTS

Eighty-nine patients (84 from the literature and 5 patients from our institutions) were identified who met the criteria for inclusion in the current analysis, 46 of whom were treated with CTR, 3 with CTR-RT, 23 with ITR, and 17 with ITR-RT.2–30, 32–58 Included were 49 females and 40 males. The median age of the patients was 45 years (range, 9–75 years). Forty-six tumors were located in the spinal cord and 43 were located in the brain. Intracranial tumors were most commonly located in the Meckel cave (n = 8) and in the posterior fossa (n = 7), whereas spinal tumors were located for the most part in the thoracic spinal cord (n = 22) and in the cervical spinal cord (n = 17). The extent of resection was confirmed by CT or MRI in 72 patients (81%).

The overall recurrence rates were 37% (33 of 89 patients) in the entire series, 24% (11 of 46 patients) after CTR, 0% (none of 3 patients) after CTR-RT, 78% (18 of 23 patients) after ITR, and 24% (4 of 17 patients) after ITR-RT. The median follow-up in the patients without local recurrence within 5 years was 60 months (range, 12–420 months) in the entire series, 54 months (range, 12–120 months) after CTR, 42 months (range, 34–420 months) after CTR-RT, 72 months (range, 24–120 months) after ITR, and 79 months (range, 12–153 months) after ITR-RT. Patient characteristics related to the four therapies are summarized in Table 1.

Table 1. Patient Characteristics Related to the Four Therapies
 CTRCTR-RTITRITR-RT
  1. CTR: complete tumor resection alone; CTR-RT: complete tumor resection followed by radiotherapy; ITR: incomplete tumor resection alone; ITR-RT: incomplete tumor resection followed by radiotherapy; CT: computed tomography; MRI: magnetic resonance imaging.

Median age (range) (yrs)45 (16–71)43 (27–53)49 (9–75)44 (13–73)
Gender, no./total    
 Female23/461/313/2312/17
 Male23/462/310/235/17
Location, no./total    
 Brain28/461/38/236/17
 Spinal cord18/462/315/2311/17
Extent of resection confirmed,  no./total 
 By CT or MRI38/462/319/2313/17
 By surgeon alone8/461/34/234/17

According to the Kaplan–Meier analysis,31 CTR, CTR-RT, and ITR-RT were associated with better local control and survival compared with ITR (Figs. 1 and 2). The 5-year local control rates were 80% after CTR, 100% after CTR-RT, and 72% after ITR-RT versus 18% after ITR (P < 0.001, log-rank test) (Fig. 1). The survival rates at 5 years were 100% after CTR, 100% after CTR-RT, and 100% after ITR-RT versus 46% after ITR (P < 0.001, log-rank test) (Fig. 2).

Figure 1.

Comparison of the local control achieved with complete tumor resection alone (CTR), complete tumor resection followed by radiotherapy (CTR-RT), incomplete tumor resection alone (ITR), and incomplete tumor resection followed by radiotherapy (ITR-RT) calculated using the Kaplan–Meier method and the log-rank test.

Figure 2.

Comparison of the survival achieved with complete tumor resection alone (CTR), complete tumor resection followed by radiotherapy (CTR-RT), incomplete tumor resection alone (ITR), and incomplete tumor resection followed by radiotherapy (ITR-RT) calculated using the Kaplan–Meier method and the log-rank test.

The extent of resection was confirmed by the surgeon's impression alone in 17% of the patients (8 of 46 patients) after CTR, in 33% of the patients (1 of 3 patients) after CTR-RT, in 17% of the patients (4 of 23 patients) after ITR, and in 24% of the patients (4 of 17 patients) after ITR-RT, respectively.

According to the multivariate analysis, only treatment schedule was found to significantly affect local control (P < 0.001) and survival (P < 0.001). The multivariate analysis did not determine a significant impact for age (P = 1.0 for local control and P = 0.5 for survival), gender (P = 0.7 and P = 1.0, respectively), location of meningeal melanocytoma (P = 0.9 and P = 0.7, respectively), and the method used to confirm the extent of resection (P = 0.7 and P = 0.9, respectively). These findings were mirrored in the univariate analysis.

In the CTR-RT group, RT was restricted to the tumor region in all three patients (local irradiation). In the ITR-RT group, 14 patients received local irradiation, 2 patients received whole brain irradiation, and 1 patient received irradiation of the entire craniospinal axis. Of the 14 patients receiving local irradiation, the total dose was 30–40 grays (Gy) in 6 patients, 45–55 Gy in 7 patients, and was unknown in 1 patient. To identify a possible dose-effect relation, patients receiving 30–40 Gy were compared with those patients receiving 45–55 Gy for local control. A comparison for survival was not performed because no tumor-related deaths were observed after postoperative irradiation. The Kaplan–Meier analysis31 revealed a trend toward better local control for radiation doses of 45–55 Gy versus doses of ≤ 40 Gy (P = 0.1) (Fig. 3). The local control rates at 5 years were 86% and 27%, respectively.

Figure 3.

Comparison of the local control achieved with a radiation dose of 30–40 grays (Gy) compared with a dose of 45–55 Gy calculated using the Kaplan–Meier method and the log-rank test.

DISCUSSION

The current analysis was performed to identify the optimal treatment for meningeal melanocytomas. Based on data from the literature and additional data from the authors, four different therapies were compared with regard to local control and survival. Additional data were available for approximately 60% of the patients. This approach provided a more detailed analysis and a longer follow-up (median follow-up of 60 months) compared with obtaining the data from the literature alone. Because the majority of the authors providing us with additional data were still in contact with the patients who were diagnosed with such a rare tumor, the data appeared to be reliable. Thus, this approach did not appear to hinder the quality of the final data set.

Biases may have been introduced because of the retrospective nature of the current study, the wide time frame of therapy, and the large number of institutions contributing data. Furthermore, the cases did not undergo a central pathologic review. Therefore, the possibility of misclassification cannot be excluded completely. These facts have to be taken into account when interpreting the results of the current study. However, the methodology applied appeared to be the only way to accumulate such a large series of patients with this very uncommon tumor. The rate of confirmation of the extent of resection by the surgeon alone varied between 17–33% in the 4 compared groups. Although the multivariate analysis did not demonstrate a significant impact on the results with regard to the method of confirming the extent of resection, a possible bias could not be excluded completely.

The current analysis demonstrated that CTR was associated with better local control and survival compared with ITR. Therefore, a CTR should be performed whenever possible, either for intracranial or intraspinal tumors. If only an ITR could be achieved, RT appeared to improve both local control and survival. Local control is an important endpoint because tumors allowed to recur may transform from meningeal melanocytoma to melanoma.4, 11, 37 In addition, metastasizing meningeal melanocytomas have been reported previously.18, 39 In these patients, the median time to recurrence was 12 months (range, 12–48 months) after ITR. In patients in whom malignant transformation or metastases were not observed, the median time to recurrence after ITR was longer (median of 24 months; range, 3–119 months).

Furthermore, subsequent resection at the time of recurrence entails significant risk. Thus, RT should be considered strongly after ITR of a meningeal melanocytoma. The value of RT after CTR was quite difficult to evaluate because of the small number of patients treated in this manner. The extent of resection depends on the size of a recurrent tumor. Therefore, surveillance MRI or CT scans should be performed at regular intervals (e.g., every 6 months).

Furthermore, we investigated a possible dose-effect relation in patients treated with ITR-RT. The data demonstrated a trend toward better local control for doses of 45–55 Gy versus doses ≤ 40 Gy. However, statistical significance was not achieved, most likely because of the small number of patients in each dose subgroup. A dose of 45 Gy is considered reasonably safe because this is well below the reported tolerance dose of 60 Gy (5% risk of severe complications at 5 years) for irradiation of up to approximately one-third of the brain.59 When administering RT, we would advise the use of localized fields using modern planning techniques to minimize hot spots and to limit the volume of affected normal nervous tissue. In regions that include specific dose-limiting structures (e.g., the eye, visual nervous tissue, spinal cord, and brain stem), more caution should be used in treatment planning.

Radiosurgery may be an alternative therapeutic option in the future, but to our knowledge only limited experience with this type of treatment for meningeal melanocytoma has been reported to date.60 To our knowledge there also is a lack of data concerning the role of chemotherapy in the treatment of meningeal melanocytoma.16 Therefore, we believe it cannot be recommended.

The results of the current study revealed that CTR was associated with better local control and survival compared with ITR. Patients in whom only ITR could be performed appeared to benefit from postoperative RT. An RT dose of 45–55 Gy was reportedly associated with a better outcome compared with a dose of ≤ 40 Gy.

Acknowledgements

The authors would like to thank the first authors and coauthors of the previously published articles, who provided us with additional data, for their kind cooperation and support.

Ancillary