Clinical Outcomes in Patients with Renal Cell Carcinoma Metastases to the Choroid Plexus

Abstract

Objective

Intraventricular metastatic brain tumors account for a small but challenging fraction of metastatic brain tumors (0.9–4.5%). Metastases from renal cell carcinoma (RCC) account for a large portion of these intraventricular tumors, and while patient outcomes are assumed to be poor, these have not been reported in a modern series with a multimodality treatment paradigm including radiation, resection and CSF diversion. Here we present the first case series of patients with intraventricular metastatic tumors from renal cell carcinoma.

Methods

This is a single institution retrospective review of patients with intraventricular RCC metastases treated between January 2003 and January 2019. Volumetric analysis was used to delineate tumor size, and the Kaplan-Meier method was used to evaluate survival data.

Results

Twenty-two intraventricular RCC metastases were identified in 19 patients with 61.3 patient-years of follow up. The median patient age was 64 years, and the median tumor volume was 2.2 cm³. Overall, 19 metastases were treated initially with radiation. Of these, 16 received stereotactic body radiotherapy (SBRT) and 3 metastases received whole brain radiotherapy. Three tumors were surgically excised and received adjuvant SBRT in the upfront setting. While five patients presented with obstructive hydrocephalus, none required cerebrospinal fluid (CSF) diversion. Following treatment, 5 metastases progressed, resulting in 1- and 3-year progression free survival (PFS) rates of 81.6% and 68%, respectively. The median overall survival was 2.8 years, with 1- and 5-year overall survival rates of 76.7% and 28.3%, respectively. Leptomeningeal carcinomatosis was not observed.

Conclusions

Despite relatively limited overall survival for this population with metastatic cancer, comparable to contemporary parenchymal brain metastasis cohorts, reasonable local CNS control was achieved in the majority of patients using a paradigm of focal irradiation and resection where indicated. Importantly, CSF diversion was not required even in patients presenting with hydrocephalus.

Introduction

Each year 338,000 new cases of renal cell carcinoma (RCC) are diagnosed, making it the ninth most common cancer worldwide. RCC often remains asymptomatic in its early stages and partly as a result is found to be metastatic by the time of diagnosis in approximately one-quarter of patients.^1,2 Two to four percent of RCC patients have brain metastases at the time of diagnosis, and 4–17% will develop brain metastases at some point during the course of their disease.^3–5 While metastatic disease to the ventricles is rare and accounts for only 0.9–4.5% of all brain metastases, metastatic RCC appears to have a unique propensity to spread to the ventricular system.^6–9 In fact, most metastases to the choroid plexus have been reported in patients with RCC.¹⁰ Indeed, RCC accounts for 36–64% of all metastatic intraventricular tumors, and some centers have reported that approximately one-third of all RCC brain metastases are found in the ventricular system.^11–13 These ventricular lesions are frequently symptomatic and morbid owing to their deep location and propensity to cause hydrocephalus.

Approaches to treatment include surgical resection via open craniotomy and irradiation including stereotactic body radiotherapy/radiosurgery (SBRT/SRS). Intraventricular lesions are challenging surgical targets given their deep location and as many access corridors involve or abut eloquent areas of the brain. Surgical resection of intraventricular metastases has been reported to result in local control rates of 79%, but this treatment has been associated with significant morbidity, with previous series describing major neurologic deficit in 38% of patients and mortality in 7% of patients treated with surgical resection.¹¹ SBRT/SRS presents an alternative option for treating these lesions, with relatively limited morbidity, with demonstrated effectiveness in a cohort of patients with intraventricular metastases including renal cell carcinoma.¹²

While there are a few retrospective case series in the literature analyzing surgical and stereotactic radiosurgical treatment of intraventricular metastases across histologies, there is no study isolating clinical outcomes in RCC patients.^10–12 Specifically, no modern series reports on the local control and long-term outcome of intraventricular RCC treated predominantly with SRS. Here, we retrospectively report the treatments and outcomes of a cohort of patients presenting with renal cell carcinoma metastatic to the choroid plexus.

Materials and Methods

Patient Population

The Memorial Sloan Kettering patient database was searched for patients with a diagnosis of renal cell carcinoma and at least one accessible magnetic resonance imaging (MRI) study of the brain dating from January 2003 through January 2019. Patients with a lesion contiguous with the choroid plexus, regardless of location, were included within the series. When resection was performed, the surgical pathology was reviewed to confirm the diagnosis of renal cell carcinoma. Two resected choroid plexus metastases were determined on pathological examination to arise from another primary malignancy and were removed from the series and were not included in the analysis. Treatment decisions regarding radiosurgery with or without surgical resection were rendered by a multidisciplinary team according to standards of care and in keeping with each patient’s presentation, functional status, and prognosis.

Data Collection

Available records of all patients satisfying the above criteria were reviewed, and data on demographics, presenting symptoms, tumor location, presence of hydrocephalus, extent of metastatic disease, treatment modalities, surgical complications, local recurrence, radiation necrosis, leptomeningeal disease, overall survival, and length of follow-up were collected. The determination of radiation necrosis was made by a neuroradiologist on the basis of available imaging, and leptomeningeal carcinomatosis by the treatment team. When date of death was unknown due to loss of follow up, the most recent communication confirming survival was recorded for censoring purposes.

Statistics

Quantitative variables were described using mean, median, standard deviation, and range. Categorical variables were assessed using frequency and percentage. Overall patient survival was determined from the date of treatment to death, with censoring using last follow-up for patients not known to have passed. For local failure, patients were censored at last intracranial imaging. The Kaplan-Meier method was used to evaluate survival data. Treatment group sizes did not permit statistically valid comparisons. All statistical analyses were completed using RStudio (RStudio, Inc.).

Results

Nineteen patients with 22 RCC metastatic lesions to the choroid plexus were treated at Memorial Sloan Kettering between January 2003 and January 2019 (Table 1). The median patient age was 64 ± 5.8 years (range 55–73 years). Sixty-eight percent (13/19) of patients were men. The ventricular metastases were discovered on surveillance imaging in 10 patients (53%) who were otherwise asymptomatic at the time of diagnosis. In those patients with symptoms, headache was most the common, and experienced by 5 patients. Four patients experienced confusion and one described preceding memory loss. The patient describing memory loss displayed notable vasogenic edema within the corpus callosum. Three patients had dizziness and gait instability, and one patient reported a visual deficit. At diagnosis, 5 patients had hydrocephalus or an entrapped ventricle. All 19 patients were evaluated with gadolinium-enhanced brain MRI. The diagnosis of metastatic intraventricular RCC was pathologically confirmed in the 6 patients who underwent surgery during the course of their treatment. Of the remaining 13 patients who did not undergo surgical sampling, 9 patients harbored an additional brain metastasis at the time of diagnosis, resulting in a high likelihood that the intraventricular pathology was concordant, in the setting of metastatic RCC with demonstrated competency to seed the brain. Tumors in 3 of the remaining patients exhibited hemorrhage, which is common with RCC and uncommon with meningioma and other rare intraventricular tumors such as choroid plexus tumors, with reported prevalence approximating 70% in RCC and 1% in meningioma, and the fourth tumor presented in the third ventricle (a location significantly less prevalent than the lateral ventricle in meningioma) and progressed rapidly despite irradiation.^14–16 In addition to the aforementioned radiographic features, these 4 cases with a solitary intraventricular tumor were furthermore considered to be far more likely to represent RCC than alternative pathologies given the rarity of the latter entities, with an estimated prevalence of intraventricular meningioma on the order of 1/100,000, and an estimated annual incidence of 1/2,000,000 for choroid plexus tumors.^17,18 In addition, none of these patients exhibited signs of a tumor predisposition syndrome or premorbid brain irradiation, risk factors for primary brain tumors, and all were treated by their multidisciplinary cancer teams for this diagnosis.

Table 1:

Patient demographics, tumor characteristics, and treatments

	Number (%)
Number of patients	19
Number of choroid plexus metastases	22
Male	13 (68.4%)
Median age, years (range)	64 (55–73)
Patients with an extracranial metastasis	15 (78.9%)
Patients with an additional intracranial metastasis	7 (36.8%)
Symptoms at Presentation
Asymptomatic	10 (52.6%)
Headache	5 (26.3%)
Confusion	4 (21.1%)
Memory Loss	1 (5.3%)
Hydrocephalus at presentation	5 (26.3%)
Median tumor volume, cm3 (range)	2.2 (0.1–31.7)
Initial treatment
SBRT	16
Surgery+SBRT	3
WBRT	3

Metastasis to the choroid plexus was diagnosed an average of 5.7 years following the initial diagnosis of renal cell carcinoma, ranging from synchronous with primary RCC presentation to a latency of 25 years. One patient presented with a symptomatic metastasis to the choroid plexus as the initial manifestation of RCC, with subsequent work-up leading to their RCC diagnosis. Fully four patients (21%) had central nervous system-only disease at the time their intraventricular metastases were identified, with the remaining 15 (79%) harboring systemically active disease. Seven patients (37%) presented with additional intracranial metastases in addition to the choroid plexus lesions focused upon in this series.

Nineteen of 22 choroid plexus metastases (86.4%), were located within the lateral ventricles (Figure 1), 11 within the right lateral ventricular system and 8 in the left. One metastasis developed within the roof of the third ventricle, and two were located in the choroid of the foramen of Luschka (one of these patients also had an additional synchronous lesion within a lateral ventricle). Notably, the patient presenting with the solitary metastasis to the foramen of Luschka subsequently developed a metastasis to the choroid plexus within the right lateral ventricle.

Figure 1.

Axial T1-weighted post-contrast MRI of the brain demonstrating an intraventricular metastasis to the right lateral ventricle.

Patient Karnofsky Performance Status (KPS) scale scores ranged from 80 to 100 with a median of 90. The Graded Prognostic Assessment (GPA) ranged from 1 to 4 with a median score of 3, correlating to a median expected survival of 11.3 months.¹⁹ Seventeen patients scored as class II on the Recursive Partitioning Assessment (RPA) and 2 patients class I.²⁰

Initial treatments at the time of diagnosis included surgical resection via craniotomy with adjuvant SBRT, SBRT alone, and whole brain radiation therapy (WBRT). Three patients with one intraventricular metastasis each underwent surgical resection followed by planned adjuvant SBRT within 6 weeks of their craniotomy. One patient underwent a subtotal resection while the other two patients had complete removal of their tumors. The patient who had a subtotal resection required further SBRT for progression 18 months after his initial adjuvant SBRT course. The other two patients had no evidence of recurrence during their follow-up period. None of these patients were treated with systemic therapy within one month of intracranial treatment, however one patient ultimately was treated with sunitinib for CNS progression and another with pembrolizumab for extracranial disease. The third patient treated with upfront neurosurgical resection was never prescribed adjuvant systemic treatment in the setting of good initial CNS and extracranial control, and deferred additional treatment upon systemic progression.

The remaining 19 metastases underwent definitive or palliative upfront radiation therapy. Sixteen metastases (72.7%) in 14 patients received SBRT alone as their primary treatment modality. This was initiated between 1 and 34 days after diagnosis. Three patients were treated with concurrent systemic therapy, reflecting their relatively limited efficacy data. One patient was treated with bevacizumab, another with axitinib, and the third with sunitinib. SBRT approaches included single fraction (SRS) treatments with a median dose of 21 Gy (range 18–21 Gy) as well as hypofractionated treatments (3 or 5 fractions) delivering a total of 27–30 Gy. The median tumor volume for metastases treated with SBRT was 1.5 cm³ (range 0.1–17.4 cm³). Of these 16 lesions that underwent SBRT, 4 developed radiation necrosis at a median of 1.1 years after treatment (range 0.5 to 1.7 years). For the metastatic tumors treated with SBRT, 3 lesions had a complete response, 3 lesions partial response, and 2 lesions were stable at 12 months by RANO-BM criteria.²¹ Five metastatic lesions had progression within 12 months, and 3 patients had insufficient follow-up for analysis. One patient was symptomatic from the radiation necrosis and this was managed with steroids. Delayed systemic therapy was ultimately used in 10 of the 11 remaining patients treated with CNS SBRT – 5 in the setting of mixed CNS and systemic progression and 5 for systemic salvage with stable brain metastases, with agents including sunitinib, pazopanib, sorafenib and temserolimus.

WBRT was performed for the other 3 metastases (13.6%) in 2 patients with initiation of therapy within 8 days of diagnosis in both patients. Both of these patients had systemically advanced disease at the time the choroid plexus metastases were identified. One had multifocal intracranial metastatic burden in both the infratentorial and supratentorial compartments, while the second had systemically advanced disease and was pursuing palliative measures. The latter patient was lost to follow-up immediately upon completion of WBRT, while the former passed 213 days following the diagnosis of choroid plexus metastases. Neither patient was treated with salvage systemic therapy.

Six of the 19 patients (31.6%) with seven metastases underwent operative treatment. These surgical tumors ranged in preoperative volume from 1.9 cm³ to 31.7 cm³, with a median volume of 11.6 cm³. Three patients received planned adjuvant SBRT following resection. The other three patients underwent surgical resection for recurrent disease confirmed on pathology following prior SRS and took place between 182 and 460 days following their radiation therapy. One patient initially treated with SBRT required two resections. The first resection took place 1.5 years following radiation, and the second resection took place 2.8 years following the first resection (without interval re-irradiation). This patient was one of only two patients requiring CSF diversion, and in this case, ventriculoperitoneal shunt was placed two months prior to the first resection. Patients who required surgery to treat choroid plexus metastases after failing upfront SBRT survived an average of 3.8 years following the diagnosis of choroid plexus metastases, with a range of 1.4 to 5.7 years.

Two patients who were treated with upfront surgical resection presented with obstructive hydrocephalus. Both patients underwent gross total resection and although one experienced transient post-operative aphasia, she ultimately returned to her neurologic baseline and did well without long term neurologic morbidity. Another patient treated with initial surgery presented with a symptomatic cerebellopontine angle tumor 5 years following treatment for localized renal cell carcinoma, and had no evidence of metastatic disease at that time. The lesion, which was associated with imbalance and headache, appeared consistent with a possible vestibular schwannoma and resection was performed for diagnostic and palliative indications.

Despite five patients presenting with hydrocephalus, none of these patients required cerebrospinal fluid (CSF) diversion. Notably, both patients who did ultimately require ventriculoperitoneal shunting did not initially present with hydrocephalus or an entrapped ventricle. Both were shunted following SBRT; however, one patient, previously discussed, failed to achieve tumor control with radiation and ultimately required both shunting and two surgical resections. The second patient demonstrated tumor contraction six months following radiation therapy, yet demonstrated ventricular enlargement concomitant with symptoms of communicating hydrocephalus.

There was a total of 61.3 patient-years of follow-up for this group. Five of the 22 metastases developed local progression in the follow-up period. The 1-year and 3-year actuarial progression free survival rates were 81.6% and 68%, respectively (Figure 2). Fourteen patients are known to have died during the follow-up period. Of the patients who died, the median time to death was 2.4 years following first treatment of their index choroid plexus metastasis. The median overall survival was 2.8 years. The 1-year and 5-year actuarial overall survival rates were 76.7% and 28.3%, respectively (Figure 3). No patient developed radiographic or cytological leptomeningeal disease despite the presence of cases with choroid-avid tumors. Eight patients developed additional cerebral metastases on MRI during the follow-up period. Due to the small sample sizes, a statistically relevant comparison of survival between treatment groups is not feasible.

Figure 2.

Kaplan-Meier curve for progression free survival at the index choroid plexus lesion. Actuarial progression free survival rates were 81.6% and 68% at 1 and 3 years, respectively.

Figure 3.

Kaplan-Meier curve for overall survival. Actuarial survival rates were 76.7% and 28.3% at 1 and 5 years, respectively.

Discussion

Renal cell carcinoma has a noted predilection for the choroid plexus within the ventricular system and CNS, making these lesions particularly daunting in neurosurgical oncologic practice. Understanding contemporary anticipated survival, and the lesions’ response to modern multimodality treatments is essential, particularly given the demonstrated relative longevity of these patients and not-infrequent manifestation of these as a patient’s only (and in one case herein, first) systemic site of involvement. Importantly, local surgical and radiotherapeutic strategies are mainstays of RCC brain metastasis treatment as modern systemic agents demonstrate limited CNS activity. Two nonrandomized phase II trials have evaluated the efficacy of immunotherapy for these lesions. GETUG-AFU 26 NIVOREN demonstrated a 12% response rate with PD-1 inhibitor nivolumab in previously-untreated metastases (and 0 for patients with multiple lesions or lesions larger than 1cm) with intracranial progression-free survival of 2.7 months; 72% subsequently were treated with focal therapy.²² CheckMate920, published in abstract form, evaluated combination nivolumab with CTLA4 inhibitor ipilimumab for untreated and asymptomatic brain metastases, followed by maintenance nivolumab, which resulted in a 29% response rate and progression-free survival (not restricted to CNS) of 9 months, and grade 3–4 toxicity of 21%.²³ Other modern systemic treatment strategies including sunitinib, and vascular endothelial growth factor and molecular target of rapamycin (mTOR) inhibitors, also have limited efficacy and can result in toxicity in these potentially hemorrhagic tumors.^2,24

Metastases within the ventricular system present unique challenges given their potential for ventricular obstruction, hydrocephalus and rapid neurologic decline. In this series, five patients exhibited hydrocephalus or a trapped ventricle on presentation. Moreover, these lesions are challenging surgically due to their deep location. Although surgery or CSF diversion may be attractive for acute restoration of CSF flow, in this series stereotactic radiosurgery was a safe and effective treatment strategy for carefully selected neurologically stable patients with asymptomatic and mild ventricular dilatation. Two patients with initially restricted CSF flow were treated surgically while the others received targeted radiotherapy and none required permanent CSF diversion.

Prior retrospective reviews have evaluated both surgical and stereotactic radiosurgical treatment of metastatic tumors to the choroid plexus. Hassaneen et al. evaluated the surgical management of 29 patients, 19 of who had a primary cancer occurring in the kidney.¹¹ The 3-year and 5-year survival rates for the cohort were 17% and 11%, respectively, lower than those seen in our more contemporary cohort. On multivariate analysis, a primary cancer diagnosis of RCC was noted to have a favorable survival benefit compared to other histologies. The 30-day mortality rate was 7%.

There are two retrospective reviews evaluating the efficacy of SRS for the treatment of intraventricular metastases. The first examined 16 patients with a mean tumor volume of 2.4 cm³ who received a mean treatment dose of 20.3 Gy.¹⁰ None of the patients experienced local failure with the metastasis stable in 8 patients, partially responsive in 6, and completely resolved in 1 patient. One patient had no follow-up imaging. The second study evaluated 25 patients with 30 metastases, of which 13 patients had RCC.¹² The median tumor volume was 0.75 cm³ and the median dose delivered was 20 Gy. With a median follow-up of 11.4 months, the actuarial 6-month and 1-year local control rates were 85.2% and 56.2%, respectively. Following SRS, the median overall survival time was 11.6 months with 6-month and 1-year actuarial rates of 87.1% and 46.7%, respectively.

While the above series looked at intraventricular metastases from a variety of primary cancers, the current study is the first to specifically analyze a cohort of patients with RCC with its unique radiobiological and survival attributes. Within this retrospective review, the majority of patients were treated primarily with stereotactic radiosurgery with a minority undergoing surgery at the outset. Patients generally did well with SBRT, with comparable progression free survival rates to the literature, with 3 patients failing focused irradiation and requiring salvage surgical resection later in their treatment course. Overall survival was somewhat longer than experiences reported in other series, which may reflect the aggressive multimodality treatment rendered at a cancer center, but this cohort remains small enough to disallow firm statistical conclusions. While leptomeningeal carcinomatosis has been reported to be an extremely rare manifestation in RCC, it was notable that this was seen in no cases with one or more choroid plexus metastases studied herein^25,26. Permanent CSF diversion was avoided in all but two cases.

Limitations of the current study include those inherent to a retrospective single institution experience. Additionally, the rarity of this entity makes it difficult to draw firm conclusions but also makes larger prospective studies not feasible.

Conclusions

This series addresses the outcome of patients with renal cell carcinoma metastases to the choroid plexus at a NCI-designated cancer center with ready access to both neurosurgical and radiosurgical treatment paradigms. Groups were too small to permit meaningful statistical comparison; however, surgical resection was associated with no long-term deficits and stereotactic radiosurgery, with few exceptions, provided lasting control even in cases with impaired CSF flow. Survival was similar to or longer than that previously described, and strikingly, leptomeningeal carcinomatosis was not observed in any patients despite cases with multiple choroidal deposits. Renal cell carcinoma when metastatic to the choroid plexus proved amenable to treatment within this cohort; although both surgery and SBRT were successfully utilized, treatment decisions must be made on a case-by-case basis with attention to patient-specific variables.

Abbreviations:

RCC

renal cell carcinoma

SBRT

stereotactic body radiotherapy

SRS

stereotactic radiosurgery

WBRT

whole brain radiation therapy

CSF

cerebrospinal fluid