Epidemiology and clinical features
Primary intracranial melanoma is a rare tumor in the central nervous system (CNS), which was first described by Virchow in 1859 (5). The estimated incidence is 0.5 cases per 10, 000, 000 person-years in the literature (6). Malignant intracranial melanomas can be divided into two types including primary and secondary subtypes. The diagnosis of PIMMs should be consistent with that outlined by Willis (4). In this report, perioperative examination demonstrated all patients had no melanoma or non-brain melanoma surgical history in the other parts of body, and all of 15 patients were diagnosed with PIMM. While the reports of sex predominance in literature are inconsistent, and some studies showed male predominance (5, 6). There was male predominance (male: female=11:4) in our series, which is consistent with the literature (7). The mean age of onset was 37.9 years (19-61 years), which is younger than that in the literature (1).
PIMMs were divided into diffuse meningeal tumors and solitary melanomas by Gibson in 1957 (3). Although PIMMs can occur throughout the central nervous system, they are more likely to develop in the posterior cranial fossa, Meckel cave, and spinal cord, and often present with mass effect (5). In this study, 5 cases located Infratentorial, and 1 case located in middle and posterior cranial fossa. The clinical symptoms of PIMM are nonspecific and consistent with those of other malignant brain tumors. The most common clinical symptoms were intracranial hypertension and hydrocephalus (43.2%) in the literature (5), and the diffuse type has a higher possibility of intracranial hypertension and hydrocephalus than the solitary type (3). While the 14 cases were the solitary type tumors in this report. Symptoms at presentation include headache, vomiting, and focal neurological deficits. The most common clinical symptoms were headache, vomiting due to intracranial hypertension. The mean duration of symptoms was 6.3 months, longer than common intracranial malignant tumors.
CT and MRI are very important for preoperative differential diagnosis of PIMMs. Typical PIMMs may show high density in CT scan, which needs to be differentiated from hemorrhage (Figure1a, Figure2a, Figure3a, Figure4a). But some PIMMs also display equal or low density in CT scan, and generally lack specificity. MRI is still the gold imaging diagnosis standard of PIMMs. MRI scans can reveal typical features of most PIMMs, which are high signal on T1 weighted images and low signal on T2 weighted images. It is different from that of other common intracranial tumors including meningioma, schwannoma, metastasis, and glioma. There is a bipolar dipole interaction between the unpaired electrons of the stable melanin organic radicals and the aquaporin, resulting in the shortening of T1 and T2 relaxation time and the generation of typical MRI features. The decrease of relaxation time is directly proportional to the content of melanin in melanoma (8).
Most of tumors with hyperintensity on T1 weighted images indicate hemorrhage, fat, or melanin. Furthermore, hypointensity on T2 weighted images may be a clue to distinguish lipoma from melanoma. Typical T1 and T2 weighted signals may provide clues for diagnosis of PIMMs. In this study, 14 tumors were mainly hyperintensity on T1 weighted images, hypointensity on T2 weighted images, and had no or mild enhancement. But only three patients were accurately diagnosed with melanoma before surgery. The definite diagnosis was not made until a typical black tumor was found during the operation. Intratumor hemorrhage which leads to MRI signal confusion and the rarity of melanoma make the correct initial diagnosis very difficult. However, the enhancement of the recognition to MRI features of melanoma can improve the preoperative diagnosis rate.
Histopathology and differential diagnosis
Pathological examination is standard protocol for diagnosis. Malignant melanomas can be diagnosed by routine H&E and immunohistochemically staining techniques. The cells of malignant melanomas are densely pleomorphic, spindle -shaped cells with mitosis and abundant cytoplasm with melanin deposits (4). In this report, the tumor tissue rich in melanin (Figure 1f), heteromorphous large cells with obvious nucleoli (Figure 2g) and giant tumor cells can be seen under a light microscope.
Immunohistochemically staining can differentiate malignant melanomas from other tumors. S-100 is highly sensitive (95%) to malignant melanomas, and HMB-45 is another highly sensitive and specific pathological marker for diagnosing malignant melanomas (5, 9). Vimentin (VIM) is a mesenchymal tumor marker, which can provide a complementary effect in the diagnosis and the differential diagnosis of PIMMs combined with other markers (5, 9). In this report, the positive rates of S-100, HMB-45 and vimentin were 93.3%%, 86.7% and 86.7%, respectively, which were consistent with the literature.
Treatments and outcomes
Most authors (1, 5, 7, 10, 11) agree gross total resection is the most important treatment for melanoma of the central nervous system. The incomplete removal increases the risk of recurrence and poor prognoses. The prognoses of the patients that received total resection are better than those of the patients who received incomplete resection (3, 11). Total resection combined with postoperative radiotherapy seems to be the preferred treatment for eliminating mass effect, improving preoperative symptoms, and achieving a histological diagnosis (3). Rodriguez (7) reported the mean survival in patients who received total removal of their tumor (19.6 months) was significantly longer than that in patients who underwent partial resection or biopsy (9.3 months). Man (6) reported complete surgical resection could increase the survival rate, while the age of less than 19 years and intracranial tumor were independent factors of poor prognosis. Subtotal resection or biopsy plus radiotherapy (chemotherapy) cannot change the survival of patients (5, 9, 12). Our current study further validates the results. GTR with RT appeared to extend progression free survival. The average overall survival time of the GTR group was significantly higher than that in the STR group (25 versus 9 months, p=0.000). Surgical resection is based on the patients’ symptoms, the location, size and number of the lesion. Total resection should be attempted with microsurgical techniques on the basis of protecting nerve function. But it is difficult to achieve total resection because of the occult onset of the tumor, the abundant blood supply of the tumor and the close relationship between the tumor and the important neurological structure. In this group, GTR was achieved in 80% cases, and STR was achieved in 20% cases because the tumors had very close relationship with pyramidal tract.
Yamane (13) reported the mean survival in patients with solitary tumors was 20.7 months. Man (6) reported that the 1-year, 2-year, 3-year and 5-year survival rates of primary CNS melanoma were 89.3%, 75.6%, 65.2%, 37.7% respectively, and the median survival rate was 15 months. The median survival time was 23 months in this group. The treatment effect is basically the same as that reported in the literature, but due to the small number of cases, it is necessary to further increase the number of cases to analyze the treatment results.
Although many scholars reported that melanoma is not sensitive to the commonly used doses radiotherapy, some studies also showed that the addition of adjuvant radiotherapy to surgery significantly reduces the local recurrence risk compared with resection alone (11, 14-17). The combined application of WBRT and SRS is more effective than that of WBRT and SRS alone (11, 15). The prognosis of patients undergoing microsurgery combined with SRS and / or WBRT is better than that of patients undergoing microsurgery or WBRT (11, 15). In this study, 12 patients with GTR received adjuvant RT postoperatively, and other 3 patients with STR received CKRS. The average overall survival time of the GTR combined with RT group was significantly higher than that of STR combined with CKRS group.
The role of adjuvant chemotherapy for PIMMs is controversial. There is little evidence that chemotherapy carries significant effect for PIMMs. Chemotherapy drugs have not efficacy because they cannot penetrate the blood-brain barrier (11). some authors reported when the tumors grow within the brain parenchyma, the blood-brain barrier is damaged in structure and function, increasing the permeability (11). Adjuvant chemotherapy has shown limited effect in the management of metastatic melanoma (11). There was no patient received postoperative chemotherapy in our series. Some authors (11, 18-22) reported that immunotherapy could prolong the overall survival of metastatic melanoma patients. Some clinical investigators (11, 23, 24) were also exploring the effect of gene therapy (targeted therapy) on PIMMs. Gene therapy (targeted therapy) combined with immunotherapy could improve the prognosis of metastatic melanoma better than immunotherapy combined with radiotherapy (25). These results also need further evaluation.
A Solitary tumor with leptomeningeal enhancement could be have in a manner similar to diffuse leptomeningeal melanosis, which is considered be a benign feature, but with a very poor prognosis and a mean survival of 6.7 months (1). In this series we excluded the patients with diffuse leptomeningeal enhancement who had no surgical indication. More patients and studies are required to confirm the role and efficacy of surgery and adjuvant therapy for PIMM. The incidence rate of this tumor is low, most of which are reported in the form of case report in literature. It is difficult to compare our report with other series in the literature.