You are here
Intracerebral metastases of malignant melanoma and their recurrences—A clinical analysis
Clinical Neurology and Neurosurgery, 9, 115, pages 1721 - 1728
Brain metastases (BM) commonly occur in patients with metastatic malignant melanoma (MM). Prognosis is poor even with maximal therapy. The aim of the current study was to retrospectively evaluate patients with BM of MM who were treated neurosurgically with respect to clinical presentation, recurrent disease, survival and factors affecting survival.
Patients and methods
Thirty-four patients (19 f/15 m) with BM of MM were treated in our hospital between 2000 and 2010. Patient data were analysed, survival was examined using Kaplan–Meier-estimates and factors affecting prognosis were evaluated using uni- and multivariate analysis.
Twenty-two patients (64.7%) had a single BM, whereas twelve patients (35.3%) revealed two or more lesions. Median survival for patients with a single BM was 13.0 months (95%-CI 9.3–16.7 months), this was significantly (p = 0.014) better than for patients with two or more BM (median 5.0, 95%-CI 3.4–14.6 months).
Nineteen patients (55.9%) developed an intracranial relapse after microsurgical resection of a first lesion. Patients with an isolated intracerebral relapse survived significantly (p = 0.003) longer than those with systemic progression (median 6.0, 95%-CI 0.0–15.3 months vs median 3.0, 95%-CI 1.7–4.3 months). Similarly, patients with a high performance status showed significantly (p = 0.001) prolonged survival (median 7.0, 95%-CI 0.0–19.9 months vs median 1.0, 95%-CI 0.0–2.2 months). Eleven out of nineteen patients (57.9%) underwent either another microsurgical resection (n = 6) or stereotactic radiosurgery (n = 5). These patients remained on a high performance status even after aggressive therapy.
Even though the prognosis for patients with BM of MM is generally poor, patients with a single BM can benefit from microsurgical resection. However, there is a high risk of intracranial relapse. In selected patients with a good performance status and recurrent intracranial disease, recurrent local therapy can be justified and useful.
Keywords: Malignant melanoma, Brain metastasis, Recurrence, Prognostic factors, Neurosurgery, Reoperation.
Malignant melanoma (MM) is the third most common diagnosis among patients with brain metastases (BM) after lung and breast cancer  . Of all malignant tumours, MM has the highest propensity to metastasize to the brain  . Approximately 10–13% of patients presenting with regional disease (American Joint Committee on Cancer (AJCC) stage III) are at risk for brain metastases  and  and 18–46% of stage IV patients will develop central nervous system (CNS) involvement  and  with a prevalence of 55–75% at autopsy , , , and . The prognosis after diagnosis of BM of MM is poor with a median survival of only a few months , , , and . Fifty-four percent of all deaths in MM patients are attributed to symptomatic BM  .
The central nervous system (CNS) is a common site of relapse in patients with systemic remission  . Relapse rates of 45–55% after surgery of single BM of MM are reported in the literature , , and . However, to the best of our knowledge no detailed information on patients treated for intracerebral relapse has been published so far.
The aim of the current study was to retrospectively analyse patients with BM of MM, who were treated neurosurgically at our hospital, with a focus on their clinical presentation, rate of recurrence, survival and possible prognostic factors. In the second part of the analysis we concentrated on those patients with intracerebral relapse who received repeated local treatment.
2. Materials and methods
2.1. Patient selection and patient characteristics
All patients undergoing surgical treatment for BM of a MM at the Neurosurgical Department at Essen University Hospital (Germany) between 2000 and 2010 were identified from records in the Departments of Neurosurgery, Radiology and Dermatology. Of the 36 patients identified, 34 patients were included in the study and their medical records were reviewed retrospectively. Two patients with a second metastatic malignancy were excluded from the study.
Information on age and sex of the patients, date of initial diagnosis, anatomical localization and histological characteristics of the primary MM, initial stage according to the AJCC classification of 2009  , date of radiographic diagnosis of BM, number, size, location and clinical presentation of BM, time interval between diagnoses of primary MM and BM, presence or absence, number and location of systemic melanoma metastases at the time of diagnosis of a BM, stable or progressing extracranial disease, date and number of surgical treatments, date and number of radiosurgical treatments if applicable, adjuvant treatment and response (radiotherapy, chemotherapy, immunotherapy), pre- and postsurgical Karnofsky score (KPS), date of last follow-up, performance status at last follow-up and date of death were obtained. Patients were categorized according to the three prognostic classes (recursive partitioning analysis, RPA) developed by the Radiation Oncology Group (RTOG)  . RPA class I included patients with a KPS ≥ 70, less than 65 years of age with controlled primary and no extracranial metastases; class III included patients with a KPS < 70, class II included all other patients.
2.2. Clinical management strategy
Patients were treated according to the interdisciplinary tumour board counselling at Essen University Hospital. Twenty-nine patients (85.3%) were treated surgically when BM were discovered. Indication for surgery included new neurological symptoms, intracerebral haemorrhage or new diagnosis of a single BM during tumour staging in patients with otherwise no or minimal tumour burden. The remaining five patients (14.7%) received stereotactic radiosurgery, as they showed very small, deeply or eloquently located lesion. They were treated surgically after intracerebral disease progression. Patients were closely followed with brain imaging studies generally every three months or sooner if symptomatic. Adjuvant therapy varied depending on number of BM, status of extracranial disease and performance status. Patients with recurrent intracranial disease received local therapy (microsurgery or radiosurgery) in cases of limited disease and good performance status. Whole brain radiotherapy (WBRT), chemotherapy or palliative therapy was performed in all other cases depending on the extent of tumour burden and performance status.
2.3. Statistical analysis
Survival was calculated from the time of diagnosis of BM, unless otherwise specified. Clinical follow-up time was defined as the date of last follow-up minus the date of diagnosis of BM.
Survival curves and median survival with confidence intervals (95%-CI) were analysed by the Kaplan–Meier method. Potential factors influencing survival were explored using the log-rank test. Variables proven significant in the univariate analyses were included in a multivariate analysis using the Cox proportional hazards model.
The effect of the following variables on survival was analysed: sex, age and AJCC stage at first diagnosis of MM, occult or amelanotic primary tumour, time span from primary diagnosis to diagnosis of a BM, age at diagnosis of a BM, number and size of BM, haemorrhage, supra- or infratentorial location, preoperative neurological signs and symptoms and KPS, extracranial tumour status and RPA class, pre- and postsurgical therapy.
The effect of the following variables on survival after diagnosis of an intracerebral recurrence was analysed: time span from diagnosis of a BM until the diagnosis of a recurrent BM, age at diagnosis of a recurrent BM, number and size of BM, haemorrhage, supra- or infratentorial location, preoperative neurological signs and symptoms and KPS, extracranial tumour status and RPA class, postsurgical therapy.
Statistical significance was established at probability levels less than 0.05. All statistical calculations were performed using SPSS software (IBM SPSS Statistics 18.0).
Nineteen female (55.9%) and 15 male (44.1%) patients were included in the study. Mean clinical follow up after treatment of a BM was 13.4 months (range 0.5–72 months). Median overall survival after diagnosis of MM was 38.0 months (95%-CI 13.8–62.2 months, mean 51.3 months), and median survival after diagnosis of a BM was 8.0 months (95%-CI 4.2–11.8 months, mean 16.8 months) ( Fig. 1 and Table 1 ).
|Number of patients||Percentage (%)||Survival time after diagnosis of BM (months)||p-Value (log-rank test)|
|Age at primary diagnosis|
|AJCC stage at primary diagnosis|
|I or II||9||26.5||13.0||4.2–21.8||26.6||0.291|
|Time from primary diagnosis to BM (months)|
3.1. Primary diagnosis
Mean age at the time of first diagnosis of a MM was 55.3 years (median 53.5 years, range 32–80 years), 79.4% of patients were younger than 65 years at the time of first diagnosis. The primary tumour was located on the lower extremities in 13 cases (38.2%), on the trunk in seven cases (20.6%), in the head and neck region in six cases (17.4%) and on the upper extremities in one case (2.9%). In seven patients (20.6%), the primary site of MM remained unknown. There were five cases of amelanotic MM (14.7%); in two of these patients the primary site of the MM remained undetected.
At first diagnosis, nine patients (26.5%) were in clinical stage I or II according to the AJCC 2009 classification. They showed a progression to stage IV after a mean time of 67.7 months (range 12–188 months) and presented with a BM 69.4 months (mean) after first diagnosis (range 12–188 months).
Nine patients (26.5%) had lymph node metastases (stage III) at first diagnosis. They progressed to stage IV after a mean time of 28.0 months (range 6–93 months) and presented with a BM after a mean time of 32.0 months (range 6–93 months).
Twelve patients (38.2%) showed metastatic disease (stage IV) at time of first diagnosis. In nine of these patients, BM were already present, in three other patients BM developed after a mean time of 15.7 months (range 1–34 months).
In four cases (11.8%), no data were available on AJCC stage at first diagnosis, they presented with a BM after a mean time of 44.3 months (range 7–84 months).
Of the 30 patients with determinable AJCC stage, all patients with amelanotic MM presented with stage IV disease at initial diagnosis, whereas only 30.8% of all other MM patients presented with stage IV disease. Six of the seven patients (85.7%) with occult primary tumour presented with stage IV disease, the remaining patient presented with stage III disease.
Sex, age or AJCC stage at first diagnosis had no significant influence on survival after diagnosis of BM. There was also no significant survival difference for patients with an occult primary tumour or an amelanotic MM. Patients who developed a BM within one year after primary diagnosis of MM tended to show a shorter survival after the diagnosis of BM (median 7 months, 95%-CI 3.3–10.7 vs median 12 months, 95%-CI 4.9–19.1 for all others). However, this did not reach statistical significance. Patients’ characteristics and survival times are presented in Table 1 .
3.2. Cerebral metastases
Mean age at the time of diagnosis of BM was 58.0 years (median 59.5 years, range 32–80 years), 70.6% of patients were younger than 65 years. Twenty-two patients (64.7%) presented with a single BM, eight patients (23.5%) presented with two to four BM and three patients (8.8%) had multiple lesions. In one emergency patient, magnetic resonance imaging (MRI) was not available to determine the number of lesions. Twenty-four of thirty-four patients (70.6%) underwent microsurgical resection of one BM as first line treatment. Five patients (14.7%) received stereotactic radiosurgery as initial treatment and were treated with microsurgery for recurrent disease. In three cases (8.8%), two neighbouring lesions were resected at the same time. In another two patients with multiple lesions only a stereotactic biopsy was taken.
The diameter of the operated lesion was available in 29 cases. Mean diameter was 32.5 mm (range 9–63 mm). The majority of tumours (91.2%) were located supratentorially. Nineteen (61.3%) of the 31 supratentorial tumours were located in the left hemisphere. An intra- or peritumoural haemorrhage was documented on preoperative images in 18 of 34 cases (52.9%).
Prior to surgery, 14 patients (41.2%) presented with a focal neurological deficit, four patients (11.8%) presented with seizures, three patients (8.8%) had a focal neurological deficit and seizures and three individuals (8.8%) were referred to the emergency department with severe intracerebral haemorrhage. Ten patients (29.4%) were free of neurological deficits or showed unspecific symptoms such as headache only. Preoperative KPS score was ≥70% in 27 cases (79.4%).
Median survival for patients with a single BM was 13.0 months (95%-CI 9.3–16.7 months, mean 22.2 months); this was significantly (p = 0.014) better compared to patients with two or more BM (median survival 5.0 months, 95%-CI 3.4–14.6 months, mean 7.5 months) ( Table 2 ). The subgroup analysis also showed a significant survival difference when dividing into patients with single, two to four or multiple lesions (p = 0.047) (median survival of 5.0 months, 95%-CI 0.8–9.2 months, mean 8.1 months and 2.0 months, 95%-CI 0.4–3.6 months, mean 5.7 months for patients with two to four lesions and for patients with multiple lesions, respectively ( Fig. 2 and Table 2 )). Patients with a KPS ≥ 70% showed a significantly (p = 0.015) improved one-year-survival rate with 51.9% of patients still alive as compared to 14.3% of patients with a KPS < 70% ( Fig. 3 ). However, the influence of performance status on overall survival was not statistically significant ( Table 2 ). Age (</≥65 years) at diagnosis of the BM, neurological signs and symptoms or supratentorial location did not influence survival significantly. Emergency patients, patients with larger tumours or haemorrhage tended to do worse as compared to the other groups. However, this difference was not statistically significant. Results are summarized in Table 2 .
|Number of patients||Percentage (%)||Survival time after diagnosis of BM (months)||p-Value (log-rank test)|
|Age at diagnosis of BM|
|Single BM b|
|Number of BM b|
|Tumour diameter >3 cm c|
|KPS ≥ 70|
|FND and seizures||3||8.8||6.0||0.0–14.0||6.7|
|RPA class I|
a Statistically significant.
b In one emergency case, no MRI was available to evaluate the number of BM-MM.
c Information on tumour diameter was only available in 29 cases.
3.3. Extracranial tumour status
Eleven patients (32.4%) did not show any signs of extracranial disease prior to surgery for a BM of MM. Of those, only six patients (54.5%) showed a single BM, whereas five patients had two to three lesions. Thirteen MM patients (38.3%) were operated despite extracranial disease. Most of them (84.6%) presented with a single BM. Reasons for surgery were symptomatic haemorrhage in six cases (46.2%), new focal neurological deficit in two cases (15.4%) and limited extracranial disease with a good prognosis as judged by our interdisciplinary tumour board in the remaining cases (38.5%). In seven patients (20.6%), who presented with intracerebral lesions with unknown primary, the site of the primary tumour remained occult even after recurrent staging. Three of the seven patients (42.9%) presented with symptomatic haemorrhage, one of them as an emergency case. In the other four patients (57.1%), surgery was indicated to get a histological diagnosis. In three cases (8.8%), data about preoperative staging were not available.
Eight of thirty-four patients were classified as RPA class I, 20 patients were classified as class II and six patients were classified as class III.
RPA class I patients tended to show a better 24-months survival rate (50%) as compared to class II or III patients (11.5%). However, this was not statistically significant in the log-rank analysis (p = 0.068). Similarly, patients without extracranial disease showed an improved 24-month survival rate (36.4% vs 16.7%). Again, this was not significant (p = 0.244). RPA class did not influence overall survival significantly ( Table 2 ).
3.4. Postsurgical therapy
After surgery, twelve patients (35.3%) were treated with radiation therapy, ten (29.4%) received chemotherapy and three individuals (8.8%) were treated with interferon therapy. Four of these patients (11.8%) received a combination therapy. Thirteen patients (38.2%) did not receive a specific therapy postsurgically as they were free of tumour postsurgically (n = 10) and had already received radiosurgery or refused radiotherapy, others showed a rapid decline of their general condition (n = 3) that leaded to supportive care only.
The three patients receiving postsurgical interferon therapy showed a significantly (p = 0.039) improved survival (median 50.0 months, 95%-CI 24.4–75.6, mean 53.0 months vs median 7.0 months, 95%-CI 3.8–10.2 months, mean 12.7 months). However, these results have to be interpreted with care as there might be a selection bias towards patients with a generally better prognosis. Patients without specific postsurgical therapy showed a median survival of 7.0 months (95%-CI 2.3–11.7 months, mean 8.5 months), whereas patients with specific postsurgical therapy survived for a median of 12.0 months (95%-CI 4.7–19.3 months, mean 21.5 months) reaching almost significance (p = 0.053). Results are summarized in Table 3 .
|Number of patients||Percentage (%)||Survival time after diagnosis of BM (months)||p-Value (log-rank test)|
|Any of the above||21||61.8||12.0||4.6–19.3||21.5|
3.5. Recurrent disease and treatment after recurrence
Nineteen of thirty-four patients (55.9%) developed an intracranial recurrence after surgery. Eleven patients (57.9%) developed a BM at a new site and seven cases (36.8%) developed a local recurrence. In one individual treated with whole brain radiation in a different hospital, data on the site of the new BM were not available. Mean time interval from first surgery to the development of an intracerebral relapse was 5.5 months (range 1–25 months). In 63.2% of cases (n = 12) relapse occurred within six months of the first operation. The brain was the only site of relapse in ten cases (52.6%), eight patients (42.1%) showed systemic progression, in one case data on extracranial status were missing. The diagnosis of an intracranial relapse led to local retreatment in 11 of the 19 patients (57.9%). Local retreatment consisted of repeated surgery in six cases and stereotactic radiosurgery in five cases. The other seven patients received whole brain radiation, chemotherapy or supportive treatment depending on tumour status and performance status. In patients selected for local treatment, mean pre-treatment KPS was 86.0% (95%-CI 74.2–97.8%) and mean post-treatment KPS was 86.0% (95%-CI 75.8–96.2%). Patients who did not receive any local treatment showed a mean KPS of 51.4% (95%-CI 29.1–73.7%). KPS during last follow up in the local treatment-group was 68.2% (95%-CI 51.0–85.4%).
Three patients (8.8%) developed a CSF dissemination during the course of their disease. The first patient developed three intracerebral metastases five months after the first operation and was treated with WBRT. A CSF dissemination was documented one month later. The second patient showed a single intracranial recurrence and new pulmonary metastases four months after the first operation and was treated with chemotherapy. A CSF dissemination was documented three months later. The third patient developed a single recurrent intracranial metastasis 25 months after the first operation. Due to a good performance status and no signs of extracranial disease, stereotactic radiosurgery was performed. The patient developed a CSF dissemination another 24 months later. All patient died shortly after diagnosis of a CSF dissemination despite palliative therapy.
Median survival after the diagnosis of an intracerebral recurrence was 4.0 months (95%-CI 1.9–6.1 months, mean 8.2 months). Patients with an isolated intracerebral relapse survived significantly longer (p = 0.003, median 6.0 months, 95%-CI 0–15.3 months, mean 13.1 months) than those with systemic progression (median 3.0 months, 95%-CI 1.7–4.3 months, 3.1 months) ( Fig. 4 ). Those patients who did not receive any specific therapy after their first surgery showed a significantly (p = 0.014) shorter survival than patients receiving radiotherapy, chemotherapy or immunotherapy (median 2.0 months; 95%-CI 0.3–3.7 months, mean 10.7 months vs median 6 months, 95%-CI 2.5–9.5 months, mean 2.8 months). Patients who received local therapy (microsurgery or radiosurgery) for their recurrence showed a significantly (p = 0.011) improved survival compared to patients who received whole brain radiotherapy, chemotherapy or supportive therapy (median 6.0 months, 95%-CI 3.8–8.2 months, mean 11.3 months vs median 3.0 months, 95%-CI 1.8–4.2 months, mean 2.9 months). Similarly, patients with a KPS ≥ 70% lived significantly longer (p = 0.001, median 7.0, 95%-CI 0.0–19.9 months, mean 12.2. months vs median 1.0, 95%-CI 0.0–2.2 months, mean 2 months) ( Fig. 5 ). Results are summarized in Table 4 . When performing a Cox regression analysis, the lack of a systemic progression of disease (p = 0.005) and a high performance status with a KPS ≥ 70% (p = 0.04) remained highly significant prognostic factors for survival with a factor of 8.248 and 10.254, respectively.
|Number of patients||Percentage (%)||Survival time after diagnosis of recurrent BM (months)||p-Value (log-rank test)|
|Recurrence after 6 months|
|Radiation before first operation|
|Adjuvant therapy after first operation|
|Radiotherapy after first operation|
|Chemotherapy after first operation|
|Isolated intracerebral relapse b|
|Local recurrence b|
|KPS ≥ 70 prior to retreatment b|
a Statistically significant.
b Data on systemic progression, site of relapse and performance status were missing in single patients.
4.1. Patient characteristics with impact on survival
The diagnosis of BM in patients with MM is generally associated with a poor prognosis. Our cohort showed a median survival of 8 months after diagnosis of a BM. This is in keeping with other surgical series that report a median survival after diagnosis of a BM of 8.5 months  and . A large retrospective study with 702 unselected patients was already published in 1998  and four more retrospective studies involving patient recruitment from 1985/1986 to 2007 with 330–692 patients followed recently , , , and . In these studies, median survival after diagnosis of a BM was 3.8–5.0 months , , , , and . In one larger prospective study of 89 patients median survival was 5.8 months  . The above-mentioned studies used rather unselected patient populations, whereas our patient population represents a surgical series preselected prior to reaching our department due to the assumption that the individual patient might profit from neurosurgical treatment. Our patients might therefore represent a more homogenous group with a generally better outcome. Median age and time span from primary diagnosis of MM to the development of BM was in concordance with the larger studies.
We could identify the number of BM as the main prognostic factor in univariate analysis. Similarly, the number of BM influenced survival in the majority of larger studies , , , , and .
Patients with a KPS ≥ 70% showed improved one-year-survival rates. Eigentler et al., who also identified KPS as a prognostic factor on univariate analysis, published similar observations  . Sampson et al.  , Fife et al.  and Zakrzewski et al.  identified the absence of extracranial disease as one factor for improved survival. We could show a trend towards improved 2-year-survival in our series for patients without extracranial disease and those grouped in RPA class I, however, this did not reach statistical significance. One explanation might be that only 54.5% of the 11 patients without extracranial disease showed a single BM, whereas 84.6% of the 13 patients with known extracranial disease presented with a single BM.
4.2. Postsurgical therapy
A small proportion of our patients received immunotherapy using high-dose interferon postsurgically. All of these patients were long-term survivors with a median survival of 50 months after diagnosis of the BM. Care must be taken in the interpretation of these data, as the proportion of patients who received immunotherapy in our study was very small. Moreover, there might be a selection bias, as patient receiving interferon therapy tended to show very limited disease. A recent report on 41 patients with BM of MM treated with surgery and immunotherapy (mainly interferon), however, showed that a systemic response to immunotherapy was associated with increased duration of survival  , supporting our findings. A current report on immunotherapy with the monoclonal antibody ipilimumab in patients with BM also showed promising results  .
35.3% of our patients were treated with postsurgical radiation therapy. Indication for postsurgical radiotherapy depended on the patient's postsurgical performance status and previous radiotherapy. Radiotherapy in patients with BM is discussed controversially in the literature, because MM has traditionally been viewed as a radio-resistant cancer. Radiotherapy options include stereotactic radiosurgery with or without WBRT or WBRT alone. Several retrospective studies have shown the efficacy of radiosurgery with or without WBRT in BM of MM, leading to local control rates of 49–90% after one year , , , , and . In a nonrandomized trial on postsurgical whole brain radiation, patients received either 2400–4000 cGy in 200–300 cGy fractions or no radiation therapy. Median survival at 9 months did not differ between the two groups; however, the median time to CNS recurrence was 26 months for those who received radiation after surgery vs. 5.7 months for those treated only with surgical resection. 85% of patients not receiving radiation died of CNS disease as compared to 24% of those who were irradiated. In addition, the relapse rate in the CNS was 37% in the irradiated group and 69% in the non-irradiated group  . Our data and other reports  did not show any influence of whole brain radiation on recurrence or length of survival.
29.4% of our patients were treated with chemotherapy postsurgically. Chemotherapy regimens depended on the year of treatment and the hospital they were treated in as some of the patients received chemotherapy outside Essen University Hospital. Therefore, based on our data a clear statement concerning the influence of chemotherapy on survival cannot be made. CNS responses to traditional chemotherapeutic agents in advanced MM remain infrequent (for a review see Ref.  ). Genetic analysis and advances in targeted therapy have led to new treatment options  . Recent reports on novel agents such as the kinase inhibitor dabrafenib revealed some promising results even in patients with disease progression after local therapy  . As only the part of the study population, who were treated in more recent years, was subjected to standardized mutational analysis, this factor was not included in our analysis.
Patients without specific postsurgical therapy tended to have shorter survival spans in our cohort. This might be a selection bias as this group of patients might include patients too sick for further treatment. Therefore the absence of postsurgical therapy cannot be seen as an independent prognostic factor in our study.
4.3. Intracranial relapse
Intracranial relapse is a major concern in patients with MM. Intracerebral recurrence rate ranges from 48% to 55% in the literature with local recurrence rates of 20–40% , , and . This is in accordance with our own findings of intracranial relapse in 55.9% of cases and a local recurrence rate of 36.8%. Salvati et al., Wronski et al. and Zacest et al. reported resurgeries in 25–75% of patients , , and . We reoperated 6/19 individuals (31.6%) with intracranial relapse and performed radiosurgery in 26.3% (5/19) of our patients. We could show that local retreatment (microsurgery or radiosurgery) was associated with longer survival; this was highly significant. Similarly, Wronski et al.  reported significantly prolonged survival in patients with local retreatment. However, information on factors influencing survival after intracranial relapse is lacking in this study. We performed a thorough analysis in our patients with recurrent intracranial disease with respect to possible factors influencing survival in this difficult patient group. We could demonstrate in multivariate analysis, that the absence of systemic progression and a high performance status were significant and independent prognostic factors for survival after intracranial relapse in our patient cohort. Even though median survival in these more favourable patient groups was 6–7 months only, local retreatment was warranted as these patients remained on a very high performance status even after aggressive retreatment and almost one third of the patients were still alive 2 years after the diagnosis of a recurrent intracranial tumour. In univariate analysis, patients with adjuvant therapy (radiotherapy, chemotherapy or immunotherapy) after their first operation showed significantly longer survival spans after intracranial relapse than patients without specific therapy after their first operation. Patients with local therapy for their recurrence also survived longer than patients who received WBRT, chemotherapy or supportive therapy. However, these effects might be due to selection bias, as patients with systemic progression or low performance status tended to receive less aggressive therapy. Some authors report on a higher rate of intracranial relapse or leptomeningeal seeding in cases of peacemeal resection or the use of a Cavitron Ultrasonic Surgical Aspirator (CUSA)  and . We did not include the mode of resection into our statistical analysis as CUSA was used in one of our cases, only. However, the influence of the mode of resection on intracranial relapse should be further studied.
Our results on local retreatment of recurrent BM of MM are supported by reports on resurgeries of recurrent BM of other tumour entities. Arbit et al. report on prolonged survival in patients reoperated for recurrent brain metastases from non-small cell lung cancer  . Similarly, Bindal et al. could show prolonged survival and improved quality of life in their group of patients with a variety of different recurrent brain metastases  .
Although prognosis in patients with BM of MM remains poor, surgery can be a good palliative treatment in patients with single BM. Repeated local therapy may be performed in patients with intracranial tumour recurrence, lack of systemic progression and high performance status to further prolong survival with a good quality of life.
-  R.E. Sawaya, R.K. Bindal, F.F. Lang, D. Abi-Said. Metastatic brain tumors. A.H. Kaye, E.R.J. Laws (Eds.) Brain Tumors (Curchill Livingstone, Philadelphia, 2001) 999-1026
-  J.S. Barnholtz-Sloan, A.E. Sloan, F.G. Davis, F.D. Vigneau, P. Lai, R.E. Sawaya. Incidence proportions of brain metastases in patients diagnosed (1973 to 2001) in the Metropolitan Detroit Cancer Surveillance System. Journal of Clinical Oncology. 2004;22:2865-2872 Crossref
-  B.E. Harrison, J.L. Johnson, R.W. Clough, E.C. Halperin. Selection of patients with melanoma brain metastases for aggressive treatment. American Journal of Clinical Oncology. 2003;26:354-357
-  A. Houghton, D. Coit, W. Bloomer, A. Buzaid, D. Chu, B. Eisenburgh, et al. NCCN melanoma practice guidelines National Comprehensive Cancer Network. Oncology (Williston Park). 1998;12:153-177
-  M.H. Amer, M. Al-Sarraf, L.H. Baker, V.K. Vaitkevicius. Malignant melanoma and central nervous system metastases: incidence, diagnosis, treatment and survival. Cancer. 1978;42:660-668 Crossref
-  D.R. Budman, E. Camacho, R.E. Wittes. The current causes of death in patients with malignant melanoma. European Journal of Cancer. 1978;14:327-330 Crossref
-  D.E. Bullard, E.B. Cox, H.F. Seigler. Central nervous system metastases in malignant melanoma. Neurosurgery. 1981;8:26-30 Crossref
-  A.J. Patel, D. Suki, M.A. Hatiboglu, H. Abouassi, W. Shi, D.M. Wildrick, et al. Factors influencing the risk of local recurrence after resection of a single brain metastasis. Journal of Neurosurgery. 2010;113:181-189 Crossref
-  M.A. Davies, P. Liu, S. McIntyre, K.B. Kim, N. Papadopoulos, W.J. Hwu, et al. Prognostic factors for survival in melanoma patients with brain metastases. Cancer. 2011;117:1687-1696 Crossref
-  T.K. Eigentler, A. Figl, D. Krex, P. Mohr, C. Mauch, K. Rass, et al. Number of metastases, serum lactate dehydrogenase level, and type of treatment are prognostic factors in patients with brain metastases of malignant melanoma. Cancer. 2011;117:1697-1703 Crossref
-  K.M. Fife, M.H. Colman, G.N. Stevens, I.C. Firth, D. Moon, K.F. Shannon, et al. Determinants of outcome in melanoma patients with cerebral metastases. Journal of Clinical Oncology. 2004;22:1293-1300 Crossref
-  J.H. Sampson, J.H. Carter Jr., A.H. Friedman, H.F. Seigler. Demographics, prognosis, and therapy in 702 patients with brain metastases from malignant melanoma. Journal of Neurosurgery. 1998;88:11-20 Crossref
-  J.M. Skibber, S.J. Soong, L. Austin, C.M. Balch, R.E. Sawaya. Cranial irradiation after surgical excision of brain metastases in melanoma patients. Annals of Surgical Oncology. 1996;3:118-123 Crossref
-  M.S. Mitchell. Relapse in the central nervous system in melanoma patients successfully treated with biomodulators. Journal of Clinical Oncology. 1989;7:1701-1709
-  M. Salvati, A. Frati, A. D’Elia, L. Pescatori, M. Piccirilli, A. Pietrantonio, et al. Single brain metastases from melanoma: remarks on a series of 84 patients. Neurosurgical Review. 2012;35:211-218 Crossref
-  M. Wronski, E. Arbit. Surgical treatment of brain metastases from melanoma: a retrospective study of 91 patients. Journal of Neurosurgery. 2000;93:9-18 Crossref
-  A.C. Zacest, M. Besser, G. Stevens, J.F. Thompson, W.H. McCarthy, G. Culjak. Surgical management of cerebral metastases from melanoma: outcome in 147 patients treated at a single institution over two decades. Journal of Neurosurgery. 2002;96:552-558 Crossref
-  C.M. Balch, J.E. Gershenwald, S.J. Soong, J.F. Thompson, M.B. Atkins, D.R. Byrd, et al. Final version of 2009 AJCC melanoma staging and classification. Journal of Clinical Oncology. 2009;27:6199-6206 Crossref
-  L.E. Gaspar, C. Scott, K. Murray, W. Curran. Validation of the RTOG recursive partitioning analysis (RPA) classification for brain metastases. International Journal of Radiation Oncology, Biology, Physics. 2000;47:1001-1006 Crossref
-  M. Staudt, K. Lasithiotakis, U. Leiter, F. Meier, T. Eigentler, M. Bamberg, et al. Determinants of survival in patients with brain metastases from cutaneous melanoma. British Journal of Cancer. 2010;102:1213-1218 Crossref
-  J. Zakrzewski, L.N. Geraghty, A.E. Rose, P.J. Christos, M. Mazumdar, D. Polsky, et al. Clinical variables and primary tumor characteristics predictive of the development of melanoma brain metastases and post-brain metastases survival. Cancer. 2011;117:1711-1720 Crossref
-  R.R. Lonser, D.K. Song, J. Klapper, M. Hagan, S. Auh, P.B. Kerr, et al. Surgical management of melanoma brain metastases in patients treated with immunotherapy. Journal of Neurosurgery. 2011;115:30-36 Crossref
-  K. Margolin, M.S. Ernstoff, O. Hamid, D. Lawrence, D. McDermott, I. Puzanov, et al. Ipilimumab in patients with melanoma and brain metastases: an open-label, phase 2 trial. The Lancet Oncology. 2012;13:459-465 Crossref
-  D. Mathieu, D. Kondziolka, P.B. Cooper, J.C. Flickinger, A. Niranjan, S. Agarwala, et al. Gamma knife radiosurgery in the management of malignant melanoma brain metastases. Neurosurgery. 2007;60:471-481
-  Y. Mori, D. Kondziolka, J.C. Flickinger, J.M. Kirkwood, S. Agarwala, L.D. Lunsford. Stereotactic radiosurgery for cerebral metastatic melanoma: factors affecting local disease control and survival. International Journal of Radiation Oncology, Biology, Physics. 1998;42:581-589 Crossref
-  A.E. Radbill, J.F. Fiveash, E.T. Falkenberg, B.L. Guthrie, P.E. Young, S. Meleth, et al. Initial treatment of melanoma brain metastases using gamma knife radiosurgery: an evaluation of efficacy and toxicity. Cancer. 2004;101:825-833 Crossref
-  U. Selek, E.L. Chang, S.J. Hassenbusch 3rd, A.S. Shiu, F.F. Lang, P. Allen, et al. Stereotactic radiosurgical treatment in 103 patients for 153 cerebral melanoma metastases. International Journal of Radiation Oncology, Biology, Physics. 2004;59:1097-1106 Crossref
-  C. Yu, J.C. Chen, M.L. Apuzzo, S. O’Day, S.L. Giannotta, J.S. Weber, et al. Metastatic melanoma to the brain: prognostic factors after gamma knife radiosurgery. International Journal of Radiation Oncology, Biology, Physics. 2002;52:1277-1287 Crossref
-  N.A. Hagen, C. Cirrincione, H.T. Thaler, L.M. DeAngelis. The role of radiation therapy following resection of single brain metastasis from melanoma. Neurology. 1990;40:158-160
-  G.T. Gibney, P.A. Forsyth, V.K. Sondak. Melanoma in the brain: biology and therapeutic options. Melanoma Research. 2012;22:177-183 Crossref
-  S. Jang, M.B. Atkins. Which drug, and when, for patients with BRAF-mutant melanoma?. The Lancet Oncology. 2013;14:e60-e69 Crossref
-  G.V. Long, U. Trefzer, M.A. Davies, R.F. Kefford, P.A. Ascierto, P.B. Chapman, et al. Dabrafenib in patients with Val600Glu or Val600Lys BRAF-mutant melanoma metastatic to the brain (BREAK-MB): a multicentre, open-label, phase 2 trial. The Lancet Oncology. 2012;13:1087-1095 Crossref
-  J.H. Ahn, S.H. Lee, S. Kim, J. Joo, H. Yoo, S.H. Lee, et al. Risk for leptomeningeal seeding after resection for brain metastases: implication of tumor location with mode of resection. Journal of Neurosurgery. 2012;116:984-993 Crossref
-  E. Arbit, M. Wronski, M. Burt, J.H. Galicich. The treatment of patients with recurrent brain metastases. A retrospective analysis of 109 patients with nonsmall cell lung cancer. Cancer. 1995;76:765-773 Crossref
-  R.K. Bindal, R. Sawaya, M.E. Leavens, K.R. Hess, S.H. Taylor. Reoperation for recurrent metastatic brain tumors. Journal of Neurosurgery. 1995;83:600-604 Crossref
a Department of Neurosurgery, University Hospital Essen, Essen, Germany
b Department of Neurosurgery, Ospedale San Giovanni Battista Molinette, University of Torino, Torino, Italy
c Epartment of Dermatology, University Hospital Essen, Essen, Germany
© 2013 Elsevier B.V., All rights reserved.