Brain Metastases from Melanoma

Historical Role of Surgery for Melanoma Brain Metastases

A review of the surgical literature reveals, as expected, a steadily increasing median survival over the last couple of decades, reducing overall morbidity and a gradual trend toward expansion of the indications for surgical resection of melanoma brain metastases.

Sampson et al. (1999) reviewed 702 patients with melanoma brain metastases and showed a doubling of survival time if a metastasis was resected when compared to WBRT alone (8.2 months vs 4.2 months). Around 50% reported improvement of symptoms. However the neurological deficit rate following surgery was 22.4% and the risk of death within 30 days was 8.6%.

reported a median survival of 6.7 months after surgical resection in a smaller series of 91 patients. The vast majority of patients had a single lesion ( n =76). Median hospital stay was 14 days. However, in the last 3 years of the study, hospital stay had fallen to 7 days (1991–1994). Central nervous system (CNS) complications occurred in 18.7% of patients and 14.2% died within 30 days of surgery.

reviewed 147 patients. Again, the vast majority had a single metastasis (84%). Median survival was 8.5 months, comparable to the previous 2 studies. The striking difference was their surgical morbidity of 8% and 30 day mortality rate of 2%. Around 78% of patients reported improvement in their symptoms. This group described their surgical technique and stated that in the last 5 years of their study neuronavigation was used routinely, reducing postoperative complications and reducing hospital stay from 14 days to 6 days.

In recent times, there have been few surgical studies pertaining exclusively to melanoma brain metastases. However, surgical techniques for resection of all brain metastases are largely interchangeable. published the results of surgical management of 208 patients with brain metastases treated between 1998 and 2002. Both the 30 day mortality rate and the neurological complication rate were low (1.9% and 6% respectively). Median survival time in this mixed group was 8 months.

A study reported in 2004 by Stark et al. is also worthy of note. Again it was a review of a mixed group of tumors. However, the series does provide some insight into expanding patient selection to include the elderly, those with multiple metastases with significant extracranial disease and patients with recurrence of brain metastases. A total of 177 patients were reviewed. None of them had melanoma. Perioperative mortality for those >70 years of age was 18.2%, while survival was significantly lower for those >70 and for those who had >3 lesions. However, reoperation for recurrence was associated with a statistically significantly higher survival.

The surgical literature described above indicates that median survival rates have remained fairly static, at around 8–9 months after surgery. Surgical resection has been repeatedly proven to be superior to WBRT alone. The relatively static survival figures must therefore be due to the lack of progress in controlling extracranial disease. All the studies showed a clear-cut advantage for those with single lesions and controlled systemic disease. Although those with multiple lesions and the elderly still have a dismal prognosis, there have certainly been incremental gains with regard to neurological deficits, hospital stay and perioperative mortality as surgical technique and precision have improved.

Current Indications for Surgical Management of Melanoma Brain Metastases

In recent times the decision-making climate has changed considerably for the treatment of melanoma brain metastases. Traditionally, patients were usually referred for neurosurgical management when symptoms of the problem became apparent. However, high risk patients now routinely undergo whole body imaging and it is not uncommon for incidental, asymptomatic brain metastases to be discovered. Lesions <1 cm in diameter are generally not appropriate for surgical resection due to limitations of stereotactic navigation systems. These small lesions are therefore best targeted with SRS. However, with the advent of several new drug trials for systemic disease, the surgeon may be called upon to resect these lesions in order for the patient to be rendered disease free in the brain, and thereby become eligible to enrol in a trial. Patients with multiple small metastases are often treated with SRS. With increased use of SRS, we are now seeing adverse effects of this treatment more commonly. Surgery is often necessary to deal with these events. The complications include radionecrosis and post-SRS hemorrhage ( ), both of which can cause mass effect and neurological deficit. Progression of disease after SRS is another indication for surgical resection in the current climate. We have recently entered the era of targeted therapies and immune-mediated therapies that may cross the blood–brain-barrier (BBB) and penetrate the CNS, thereby providing an additional modality of treatment. This may reduce the need for surgery for some tumors. However, the complication profile of these new agents is yet to be defined ( ) and surgery remains the gold standard for treatment of melanoma metastases in the brain.

Technical Aspects of Surgical Resection of Melanoma Brain Metastases

The single most important factor when considering surgery for metastatic disease in the brain is its precise location. Surgery in eloquent areas of the brain can result in significant neurological deficit that impairs the individual’s quality of life, which is of paramount concern since life expectancy is generally limited. Most metastases that cause symptoms such as headache, focal deficit or seizure produce a considerable amount of vasogenic edema with mass effect. It is therefore important that the patient be stabilized with high dose dexamethasone, and anticonvulsants if needed. In those who have not experienced seizures, there appears to be no role for the use of prophylactic anticonvulsants ( ).

Once the patient is stabilized, adequate imaging is required to accurately delineate the location of the lesion(s). The gold standard is MRI examination with and without contrast. This not only defines the location but also confirms the true number of lesions. Adjuncts to standard MRI are functional MRI (fMRI) and diffusion tensor imaging (DTI). With considerable reliability fMRI can map out the motor cortex and speech areas and their relationship to the tumor ( ). DTI visualizes the deep white matter tracts and their relationship to the lesion. In addition, a multiplanar rapid acquisition gradient echo (MPRAGE) sequence is essential for intraoperative neuronavigation. Furthermore, fMRI and DTI images may be fused with MPRAGE images and used intraoperatively to avoid eloquent areas during resection of metastatic lesions ( ).

Most patients may be admitted on the day of surgery. With the use of stereotactic navigation, incorporating the above imaging techniques, it is possible to resect metastatic lesions in a minimally invasive fashion using a small linear incision and mini craniotomy. This technique facilitates the resection of numerous lesions, often under the same anesthetic ( ). After inducing general anesthesia, the patient is positioned on the operating table with the head rigidly secured in pins. The frameless stereotactic system with the preoperative images is then registered. Using the navigation probe, the position of the lesion(s) can be marked on the scalp, followed by an appropriately sized linear incision. A minimal shave is performed, the head is prepared and draped, and sterile components of the frameless navigation system are attached.

A linear incision is made and the pericranium is stripped away. A mini craniotomy is then performed and the dura is opened with a cruciate or curved incision. The operating microscope is then introduced to facilitate microsurgical dissection of the lesion. Stereotactic navigation is indispensable at this stage as it ensures a direct trajectory to the lesion via a safe route that avoids eloquent areas. A microsurgical technique is used to circumferentially dissect the lesion and remove it whole. Melanoma is often highly vascular. Internal debulking therefore can lead to significant blood loss and so should be avoided if possible. However, in lesions into which hemorrhage has occurred, it is often advantageous to first evacuate the clot within and outside the lesion to facilitate circumferential dissection. The most important advantage of circumferential dissection and en bloc resection is that it allows identification and cautery of arterial feeders and draining veins. The plane between brain and tumor is usually quite distinct.

After excision of the lesion, meticulous attention is paid to hemostasis. It is often necessary to line the tumor cavity with Surgicel to achieve hemostasis. In the hemostatic phase of the operation, the anesthetist is also instructed to raise the patient’s systolic BP to a level that is approximately 20 mm Hg higher than the resting BP. Closure of the dura is then carried out and the bone flap is secured with titanium plates. Any large areas of bony defect can be further sealed with acrylic cranioplasty to give a better cosmetic result. Finally, the skin is opposed with galeal sutures, and the skin closed with a subcuticular absorbable suture.

When anesthesia is reversed at the conclusion of the procedure, the patient is assessed for possible deficits. The patient is observed in a high dependency environment for the next 24 h, and a CT or MRI performed at this stage to assess for postoperative hematoma, edema and most importantly to confirm complete resection. Most patients who undergo uncomplicated surgery as described above may be discharged home with 2–5 days of their surgery.

Complications Following Surgical Resection of Melanoma Brain Metastases

Complications are rare but can be devastating when they do occur. Postoperative mortality may be caused by postoperative hemorrhage in the tumor bed or in a distant fashion from a separate metastasis. Neurological deficit may occur after the resection of tumors in eloquent areas. However, generally these tumors caused a deficit preoperatively and the deficit improves postoperatively as the mass effect is relieved. Occasionally temporary worsening is noted. Surgery itself can provoke seizures. Again, the seizure tendency is higher preoperatively due to edema and mass effect. Other postoperative complications include infections, hydrocephalus, and general systemic complications that may be associated with operative procedure.