Keywords metastatic melanoma - brain metastases - neurosurgery
Palavras-chave melanoma metastático - metástase cerebral - neurocirurgia
Introduction
Cutaneous melanoma is one of the most common tumors to metastasize (8%) to the brain;
is currently the third most common cause of cerebral metastasis, after lung (17%),
renal cell (10.5%), and breast (5.2%) cancers; and is the most common cancer in young
adults in Australia and in many other Western countries, with an estimated lifetime
risk of between 1:25 and 1:87.[1 ]
[2 ]
[3 ]
[4 ]
[5 ] Large clinical series show that brain metastasis (BM) are diagnosed in 10 to 40%
of melanoma patients during their lifetime, and autopsy data demonstrate that 49 to
73% of patients who die of disseminated metastatic melanoma have brain involvement.[4 ]
[6 ]
[7 ] Improvements in imaging quality and accessibility have also contributed to the increased
number of patients in whom metastasis is diagnosed.[2 ]
Melanoma has a propensity for multiorgan involvement, and central nervous system complications
are frequent. Metastatic spread to the brain is the most serious event in the course
of melanoma because it carries the worst prognosis of all visceral metastasis and
represents the major cause of death in patients with disseminated disease.[6 ] The prognosis of patients with disseminated melanoma is particularly poor if cerebral
metastasis is present.[5 ]
[8 ]
The recent management of cerebral melanoma metastasis mainly depends on the number
and size of the metastases and on extracranial extension of the disease. Neurosurgery
or stereotactic radiosurgery (STR) is usually offered to patients with single or few
metastatic lesions. However, surgical removal is a mainstay. Patients who are not
eligible for surgery or STR are usually offered whole-brain radiotherapy (WBRT), although
no survival advantage has been demonstrated to date.[1 ]
[2 ]
[6 ] Nevertheless, surgical resection in conjunction with WBRT is feasible in a subgroup
of these patients and may prolong their survival.[5 ]
[8 ] Generally accepted standards for the application of different treatment modalities
in patients with BM from melanoma do not exist so far. Oncologists had little to offer,
given that chemotherapy generally does not penetrate the blood-brain barrier.[9 ]
This retrospective study aimed to analyze prognostic factors, effects of treatment,
and survival outcome of patients with BM from melanoma at a single institution over
the last 22 years.
Clinical Material and Methods
Clinical Material and Methods
From January 1991 to April 2013, total 34 patients with histologically proven cerebral
metastasis from melanoma underwent surgical treatment in the Department of Neurosurgery
at the Santa Isabel Hospital. Most patients lived in Blumenau, with small numbers
living in other cities around. Clinical data on these patients were collected retrospectively
from their medical records, including operative and pathologic reports, and information
from the office files of their neurosurgeons. Detailed follow-up review was conducted
by telephone contact with patients our family members. The following information was
obtained: demographic, time interval between diagnoses of primary melanoma and BM,
localization and number of BM, Karnofsky performance status (KPS), presence of extracranial
disease, administered treatment, response to treatment, date of death or last follow-up,
and cause of death.
Diagnosis of BM from melanoma was based on computed tomography (CT) and magnetic resonance
imaging (MRI) scans. Usually, patients were offered surgery if they had one or few
surgically accessible BM and good KPS. STR was not available in our department. Patients,
who received WBRT, usually commencing within 6 weeks of surgery, were treated with
a total dose of 30 Gy with different fractionations regimens. Systemic chemotherapy
was administered at some time after craniotomy, mainly for progressive extracranial
disease.
Follow-up images were obtained by the neurosurgeon by using CT and MRI scans were
obtained in patients being considered for reoperation.
Statistical Analysis
Kaplan–Meier survival curves and median values were calculated by standard formulas
with commercially available software (EPINFO, 2006 e EPIDATA). A comparison of the
median survival times among treatment groups was performed using log-rank tests. Standard
tests of multivariate and univariate analysis were performed using the test with Bartlett's
Test for Inequality of Population Variances and parametric test ANOVA, and test for
two samples Mann–Whitney/Wilcoxon 13 Statistical significance was defined at the 5%
level.
Results
The demographic profile of the patients included in the study is summarized in [Table 1 ]. The median patient age was 50 years (range, 23–68 years) and 67.6% of the patients
were men. A history of melanoma was found in 16 patients (47%), with a median interval
of 12.2 months from diagnosis of primary melanoma. The remaining 18 patients (53%)
presented with a cerebral metastasis as the first evidence of melanoma.
Table 1
Demographic features in 34 patients who underwent craniotomy for cerebral metastasis
of melanoma
Variable
No of patients (%)
Age in years
23–29
8 (23.3)
30–49
8 (23.3)
50–68
18 (53.3)
Sex
M
23 (67.6)
F
11(32.4)
Time at initial diagnosis until cerebral metastases in months
<1[a ]
18 (53)
1–6
6 (17.6)
7–12
3 (8.9)
13–24
3 (8.9)
>24
4 (11.6)
a Melanoma was diagnosed following presentation with cerebral metastasis.
The clinical characteristics of all patients at the time their cerebral metastasis
was diagnosed are presented in [Table 2 ]. Most patients (70.6%) had a single cerebral metastasis. This was assessed using
CT scanning only over the period of 1991 through 1998 and by MRI beginning in 1998.
Of the remainder, 14.7% had two metastases and the same number had three metastases;
91% of the patients had supratentorial lesions. The locations of single cerebral lesions
were frontal (50%), parietal (5%), temporal (40%), and occipital (1%). Overall, more
than 20% of the patients in this series had evidence of extracranial metastasis at
the time their cerebral metastatic disease was diagnosed. The KPS at presentation
was ≥70 in 12 patients (35.2%).
Table 2
Characteristics of 34 patients with cerebral metastasis of melanoma
Variable
No. of patients (%)
Median survival (mo)
Karnofsky at presentation
p = 0.0940
0–60
22(64.7)
3
70–100
12(35.3)
35.2
No. of metastasis
p = 0.487
1
24 (70.6)
21
2
5 (14.7)
3.4
≥3
5 (14.7)
3.4
Location of metastasis
p = 0.089
Supratentorial
31 (91)
14.3
Infratentorial
3 (9)
15
Extracranial metastasis
p = 0.0528
yes
7 (20.6)
1.42
no
27 (79.4)
17.57
Extent of extracranial metastases
Skina
27 (79.4)
17.57
Lung
2 (6)
3.5
Liver, lung
1 (2.6)
0
Liver
2 (6)
1
Paravertebral, spinal cord
2 (6)
1
The treatment details for the cerebral metastases and patient outcome are given in
[Table 3 ]. A macroscopically complete tumor excision was achieved in 38.3% of patients. Subtotal
resection or biopsy sampling were performed in the remainder, based on an assessment
by the surgeon at the time of the procedure. Most patients (64.7%) had either improved
or unchanged neurologic status when assessed at the time of discharge from the hospital.
Table 3
Treatment of cerebral metastases in 34 patients with melanoma
Variable
No. of patients (%)
Survival (mo)
Treatment
p = 0.02
Surgery
13 (38.3)
1.9
Surgery/WBRT
4 (11.7)
11.2
Surgery/chemo
13 (38.3)
24.8
Surgery/WBRT/chemo
4 (11.7)
21.5
Extent of resection
p = 0.08
Macroscopically complete
13 (38.3)
15
Subtotal
4 (11.7)
46.7
Biopsy only
8 (23.5)
3.3
NAa
9 (26.5)
Neurologic status
p = 0.0002
Improved
13 (38.2)
34.4
Unproved
12 (35.3)
1.1
Unchanged
9 (26.5)
2
Intracranial recurrence
p = 0.5
Op site
0 (0.0)
0
Same lobe
4 (11.7)
4.75
Adjacent lobe
2 (5.9)
8
None
28 (82.4)
15.9
Outcome
p = 0.04
Alive w/melanoma
1 (2.9)
12
Death from disseminated melanoma
11 (32.4)
37.7
Death from intracerebral metastases
22 (64.7)
2.9
Abbreviations: NA, not applicable; WBRT, whole-brain radiotherapy.
Clinical or radiologic evidence of recurrent intracranial melanoma was found during
follow-up review in 17.6% of patients. Recurrence was assessed using CT scanning,
followed by MRI if further surgical treatment was contemplated. Radiosurgery was not
available for the patients in this series.
Progressive intracerebral melanoma was the direct cause of death in 64.7% of the entire
series. The remainder died of progressive extracranial disease.
The overall median survival time from diagnosis of cerebral metastasis was 14.4 months,
which is demonstrated in [Fig. 1 ]. On univariate analysis, the absence of extracranial metastasis was associated with
a significantly prolonged survival (p = 0.052), as was adjuvant therapy (p = 0.02). Other factors that did not affect survival included patient age, sex, the
time interval from primary melanoma to cerebral metastasis, and the location of the
cerebral lesion (infra- or supratentorial).
Fig. 1 Survival in months. *The overall median survival time from diagnosis was 14.4 months.
Although the median survival of patients with single lesions was higher (21 months),
there was no statistical significance (p = 0.48) ([Fig. 2 ]).
Fig. 2 (a) Kaplan-Meier curve for survival in patients with one or more lesions and (b)
patients with or without extracranial metastasis.
Discussion
Melanoma has the highest propensity to metastasize to the brain of all primary neoplasms
in adults, and this paper shows that the most prevalent age range is between 50 and
69 years (53.3%). This population is affected in their productive age, aggressive
therapy contributes impressively in some cases. Patients with melanoma tend to be
younger (∼50 years old) and often have excellent KPS scores on admission; however,
our study found that KPS less than 70 were most frequent (64.7%) and the patients
was oldest (50–69 years old).[10 ]
In this series, there was early diagnosis of BM, on average 12.2 months after diagnosis
of the primary disease, which typically occurs late in the course of melanoma, with
an interval from 2.2 to 3.8 years. In the study of Wronski et al, the median time
from diagnosis of the primary tumor to that of the BM was 14.4 months.[4 ]
[6 ]
[11 ]
[12 ]
[13 ]
In addition, Sampson et al investigated that patients with a single lesion and an
absence of extracranial metastasis who initially presented with BM have a better prognosis.
This article was evidenced that the most significant determinant of survival is the
absence of extracranial disease (17.34 menses), unlike those who had extracranial
disease (1.42). The same was asserted by Nieder et al that patients with extracranial
metastasis had a poor median survival (3.2 months).[10 ]
[11 ]
[14 ]
[15 ]
Nieder et al asserted that the most important prognostic factor was KPS.[10 ] Their study showed a poor survival (2 months) for patients with KPS <70 years. In
this study, all patients had BM initially and their survival was 3.28 months with
KPS <70 years. Both data Sampson and Nieder had the same from database.[10 ]
[11 ]
The number of brain lesions from melanoma is a factor prognostic used in the scores
of assessment of melanoma, like melanoma-specific graded prognostic assessment (GPA),
and this article confirms that patients with two to three or more than three BMs had
a poor prognosis with 3.4 months.[10 ]
[16 ]
[17 ]
[18 ]
In the study of Wronski and Arbit, the 1 -year mortality of patients who underwent
craniotomy was 60%, whereas our results show higher mortality (70.6%) for the overall
group, without discrimination of other variables.[4 ]
The survival of supratentorial and infratentorial was 14.3 and 15 months, respectively;
however, as other studies show, the location of cerebral metastasis was not significant
in survival.
The primary goal of surgery is gross-total resection of the tumor with minimal disruption
of, or injury to, the brain. In some cases, the additional goal of debulking and relief
of mass effect are considered. Account of the criterion of the extent of resection
during surgery is a subjective evaluation of the surgeon, and our team has four members,
because this is not statistically significant if resection is macroscopically complete
or subtotal and de survival (p = 0.08). Although the mortality of patients who underwent biopsy only was 3.3 months,
per Zacest et al showed that a macroscopically complete excision significantly affected
survival on univariate analysis (p < 0.05). Thus, the improvement in the technology and techniques of resection should
be craved in our institution.[2 ]
[5 ]
[19 ]
Combining the modalities of surgery, WBRT, and chemotherapy has improved in a statistically
significant survival (21.5 months) for patients who underwent these three therapies
(p = 0.02).[13 ]
[20 ]
[21 ]
[22 ]
[23 ]
[24 ]
Conclusion
The early diagnosis, with total resection of melanoma metastasis and association of
adjuvant therapies, showed a positive influence on survival. However, the number of
lesions and extracranial disease decreased the survival rates. We believe that further
studies will follow protocols to establish treatment.
