The treatment of metastasized melanoma (MM) has changed dramatically within the past decade, mainly due to the introduction of immune checkpoint inhibitors (ICI) and targeted treatment, leading to possible long-term remissions. This also includes the treatment of melanoma brain metastases (MBM) [1
], the most aggressive subtype of metastases, with poor prognosis [4
]. About 50% of all MM patients develop MBM. For MBM, radiotherapy (RT) plays a major role in disease control [6
]. Historically, whole-brain radiotherapy (WBRT) was considered standard of care, especially for patients suffering from a high number of symptomatic MBM, and is still a potential treatment option [8
]. However, treatment is progressively switching to stereotactic radiosurgery (SRS) as the preferred treatment, especially in patients with a limited number of metastases [10
]. Adjuvant WBRT has been shown to be ineffective in patients with melanoma. Except for an increased local tumor control in patients with 1–3 MBM, the general outcome is not favorable [12
]. Current European Society for Medical Oncology (ESMO) guidelines recommend avoiding WBRT in asymptomatic melanoma metastasis with regard to lack of efficacy and long-term toxicities [13
Several prospective studies have compared the rate of neurocognitive function (NCF) decline of patients treated by SRS alone compared to patients receiving either SRS and WBRT in combination [11
] or SRS vs. WBRT [15
]. Brown et al. [11
] demonstrated superiority of SRS alone in terms of cognitive deterioration at 3 months from initiating RT; the differences at 12 months were still significant, yet less conclusive. However, all studies focused on short- to medium-term cognitive differences. As such, data on long-term functional outcomes with respect to NCF at more than 12 months are missing.
Through the introduction of novel treatment options like ICI, potential long-term survival and long-term toxicities such as neurocognitive outcome should be considered when discussing treatment options such as WBRT [16
]. The aim of our study was to describe neurological function with focus on NCF in a series of MBM long-term survivors treated with WBRT.
In this study, we present a group of 8 patients with long-term survival after WBRT for MBM. When discussing the results, it should be noted that the patient sample is small, which represents a limitation of our study. Therefore, any conclusions based on these data need to be taken with care.
It is remarkable that in 6/8 patients, neurological outcome and especially neurocognitive function after WBRT allowed patients return to their previous work. Further, most patients reported no subjective impairment in neurocognitive function.
Simultaneously, we also identified 3/8 patients (ID 1, 2, 4) with reduced neurocognitive function in NAB screening after WBRT (Table 2
; Supplementary Table 1). Subjective memory deficits in patients 1 and 4 are mirrored in corresponding below average scores in the memory module (Table 2
). Both patients noted cognitive deterioration soon after WBRT, showing an impact on their everyday life. Thus, a relevant proportion of patients (38% of our cohort) still had a low-average to below-average neurocognitive function, and 2 patients (25%) had relevant impairment of their everyday life. It has to be noted that premorbid NCF was not available for the analysis. As such, other causes beside WBRT have to be considered and will be discussed below.
In general, up to two thirds of patients with brain metastases experience neurocognitive impairment within 2–6 months after WBRT, including concentration deficits but also decreased short- and long-term memory [11
However, in our patients, the cause of cognitive deterioration should not only be seen in potential radiogenic toxicity. As such, parietal lesion localization itself may certainly at least partially explain deficits in spatial skills (patient 1, 2; e.g., identifying and self-constructing different patterns in NCF testing), reduced memory performance of patterns (patient 1), and also reduced attention (patients 2, 4: marking “X” and “numbers” in 16 and 8 rows, respectively, in the NCF test battery). In addition, in patient 4, a partial visual field defect due to a metastatic lesion at the visual cortex might contribute to reduced visual ability to identify objects as well as levetiracetam intake reducing working memory and attention in some patients. Further, in comparison to the other patients, in patient 1, the WBRT treatment dose was increased and surgical resection of parietal metastasis was performed (Table 1
Decreased NCF may also be based on concurrent systemic anticancer treatments. As shown in Table 1
, 7 of 8 patients (88%) received ICI, for which neurological adverse events are rare, but may be severe [20
]. An influence of ICI on neurocognitive function is rare [21
] and to the best of our knowledge, has not been evaluated systematically so far. In our cohort, only 1 patient (patient 8) suffered from ICI-induced sensorimotor neuropathy as an autoimmune related adverse event, which required temporary systemic steroids and had completely resolved 6 months before NCF testing. Further, an influence of chemotherapy on NCF might be discussed as well, but may be especially relevant in elderly patients [22
]. It is rather unlikely that concurrent chemotherapy with temozolomide and dacarbazine (in patients 1 and 2, respectively) can explain a cognitive impairment but may of course be a contributing factor.
Temozolomide is even expected to prevent neurocognitive decline in patients with primary or metastatic CNS tumors [23
]. However, cognitive changes associated with cancer treatment may be diffuse and the topic of chemobrain is broad [21
Neurotoxic deficits usually involve the domains of attention and concentration, verbal and visual memory, and processing speed. Two of our patients had received targeted treatment with BRAF and MEK inhibitors in the past, which may rarely cause central neurotoxicity [21
] as well as paclitaxel in very rare cases.
Certainly we acknowledge that 75% of our patients held a university degree and would be expected to have a higher-than-average premorbid neurocognitive function. This might be a clinical hint toward potential deterioration after WBRT, as frequently seen. Also, it has to be noted that our cohort was highly selected and included potentially very fit patients; similar to toxicity seen in chemotherapy, more frail patients may suffer many more short- and long-term side effects from WBRT. Without any premorbid NAB status available, we cannot prove a deterioration of NCF over the course of treatment without a full return to normal. In patient 6, high intracranial tumor load as well as bifrontal and biparietal localization of metastases may explain the worse cognitive performance than expected with regard to profession as a teacher.
When deciding on an optimal treatment for each patient, potential deterioration of NCF always has to be considered. Short-term deterioration of NCF due to WBRT is well recognized and has been shown in randomized controlled trials [11
]. However, long-term effects in melanoma patients are widely unknown, as previously highly unlikely without an effective anticancer systemic treatment. Thus, factors predictive of long-term neurocognitive toxicity of WBRT are poorly studied, especially in melanoma [21
]. Several studies have evaluated potentially beneficial additions to WBRT to preserve neurocognition, mainly including the use of systemic memantine [27
]. A hippocampal-sparing technique in WBRT to avoid cognitive decline, especially in memory, should be preferred and is recommended [28
Patients studied in our case series mostly show a favorable outcome with good everyday functioning and return to activities. This is in line with Jiang et al. reporting on beneficial outcomes in MBM patients after WBRT, showing only limited neurocognitive side effects. However, neurocognitive function was not measured objectively [31
]. Favorable outcomes of children treated with cerebral RT have been described as well [32
These findings may encourage consideration of WBRT as a potential treatment option in selected cases, as the goal in melanoma treatment is progressively switched to long-term tumor control, even in patients with MBM.
Meanwhile, there is also a large body of evidence suggesting progressive use of SRS as the primary treatment option in combination with systemic treatment [33
]. This does not only include a limited number of 1–3 brain metastases; SRS can currently be used safely in up to 15 brain metastases [34
]. While radionecrosis may be a factor to consider in SRS [37
], the main advantage of SRS versus WBRT is reduced neurotoxicity, which has been demonstrated in several studies, especially confirming short-term neurocognitive decline induced by WBRT [35
]. Therefore, studies on long-term neurocognitive decline after WBRT are needed especially in tumors aiming for long-term tumor control.
Especially in patients with brain metastasis in whom local therapy has failed, or with neurological symptoms requiring steroids or leptomeningeal disease, an infrequent response to ipilimumab/nivolumab has been shown [38
]. This population can be treated by WBRT, even in the case of leptomeningeal disease or very extensive involvement.
We are aware of limitations of this analysis, including the very small number of patients, the partial bias due to localization of brain metastasis in neuropsychologically relevant areas, and the lack of premorbid NCF, precluding a general conclusion on long-term effects of WBRT for MBM. The strength of our study was the differentiated neurocognitive test battery applied to all long-term survivors, allowing a validated statement on the neurocognitive function of these patients.