Repeat stereotactic radiosurgery as salvage therapy for locally recurrent brain metastases previously treated with radiosurgery

Full access

OBJECTIVE

There are a variety of salvage options available for patients with brain metastases who experience local failure after stereotactic radiosurgery (SRS). These options include resection, whole-brain radiation therapy, laser thermoablation, and repeat SRS. There is little data on the safety and efficacy of repeat SRS following local failure of a prior radiosurgical procedure. This study evaluates the clinical outcomes and dosimetric characteristics of patients who experienced tumor recurrence and were subsequently treated with repeat SRS.

METHODS

Between 2002 and 2015, 32 patients were treated with repeat SRS for local recurrence of ≥ 1 brain metastasis following initial SRS treatment. The Kaplan-Meier method was used to estimate time-to-event outcomes including overall survival (OS), local failure, and radiation necrosis. Cox proportional hazards analysis was performed for predictor variables of interest for each outcome. Composite dose-volume histograms were constructed for each reirradiated lesion, and these were then used to develop a predictive dosimetric model for radiation necrosis.

RESULTS

Forty-six lesions in 32 patients were re-treated with a second course of SRS after local failure. A median dose of 20 Gy (range 14–22 Gy) was delivered to the tumor margin at the time of repeat SRS. Local control at 1 year was 79% (95% CI 67%–94%). Estimated 1-year OS was 70% (95% CI 55%–88%). Twelve patients had died at the most recent follow-up, with 8/12 patients experiencing neurological death (as described in Patchell et al.). Eleven of 46 (24%) lesions in 11 separate patients treated with repeat SRS were associated with symptomatic radiation necrosis. Freedom from radiation necrosis at 1 year was 71% (95% CI 57%–88%). Analysis of dosimetric data revealed that the volume of a lesion receiving 40 Gy (V40Gy) was the most predictive factor for the development of radiation necrosis (p = 0.003). The following V40Gy thresholds were associated with 10%, 20%, and 50% probabilities of radiation necrosis, respectively: 0.28 cm3 (95% CI 3%–28%), 0.76 cm3 (95% CI 9%–39%), 1.60 cm3 (95% CI 26%–74%).

CONCLUSIONS

Repeat SRS appears to be an effective salvage option for patients with brain metastases experiencing local failure following initial SRS treatment. This series demonstrates durable local control and, although rates of radiation necrosis are significant, repeat SRS may be indicated for select cases of local disease recurrence. Because the V40Gy is predictive of radiation necrosis, limiting this value during treatment planning may allow for a reduction in radiation necrosis rates.

ABBREVIATIONS CTCAE = Common Terminology Criteria for Adverse Events; DVH = dose-volume histogram; HR = hazard ratio; IQR = interquartile range; KPS = Karnofsky Performance Scale; OS = overall survival; SRS = stereotactic radiosurgery; WBRT = whole-brain radiation therapy.

Abstract

OBJECTIVE

There are a variety of salvage options available for patients with brain metastases who experience local failure after stereotactic radiosurgery (SRS). These options include resection, whole-brain radiation therapy, laser thermoablation, and repeat SRS. There is little data on the safety and efficacy of repeat SRS following local failure of a prior radiosurgical procedure. This study evaluates the clinical outcomes and dosimetric characteristics of patients who experienced tumor recurrence and were subsequently treated with repeat SRS.

METHODS

Between 2002 and 2015, 32 patients were treated with repeat SRS for local recurrence of ≥ 1 brain metastasis following initial SRS treatment. The Kaplan-Meier method was used to estimate time-to-event outcomes including overall survival (OS), local failure, and radiation necrosis. Cox proportional hazards analysis was performed for predictor variables of interest for each outcome. Composite dose-volume histograms were constructed for each reirradiated lesion, and these were then used to develop a predictive dosimetric model for radiation necrosis.

RESULTS

Forty-six lesions in 32 patients were re-treated with a second course of SRS after local failure. A median dose of 20 Gy (range 14–22 Gy) was delivered to the tumor margin at the time of repeat SRS. Local control at 1 year was 79% (95% CI 67%–94%). Estimated 1-year OS was 70% (95% CI 55%–88%). Twelve patients had died at the most recent follow-up, with 8/12 patients experiencing neurological death (as described in Patchell et al.). Eleven of 46 (24%) lesions in 11 separate patients treated with repeat SRS were associated with symptomatic radiation necrosis. Freedom from radiation necrosis at 1 year was 71% (95% CI 57%–88%). Analysis of dosimetric data revealed that the volume of a lesion receiving 40 Gy (V40Gy) was the most predictive factor for the development of radiation necrosis (p = 0.003). The following V40Gy thresholds were associated with 10%, 20%, and 50% probabilities of radiation necrosis, respectively: 0.28 cm3 (95% CI 3%–28%), 0.76 cm3 (95% CI 9%–39%), 1.60 cm3 (95% CI 26%–74%).

CONCLUSIONS

Repeat SRS appears to be an effective salvage option for patients with brain metastases experiencing local failure following initial SRS treatment. This series demonstrates durable local control and, although rates of radiation necrosis are significant, repeat SRS may be indicated for select cases of local disease recurrence. Because the V40Gy is predictive of radiation necrosis, limiting this value during treatment planning may allow for a reduction in radiation necrosis rates.

Although local failure after stereotactic radiosurgery (SRS) remains a relatively uncommon phenomenon, with reported rates of local control after 1 year > 90%, the recent improvements in survival seen in patients with brain metastases have made the management of local failures after radiosurgery a relevant clinical issue.1,4,11,14,22,26 Factors associated with local failure that need to be considered when treating these lesions include primary tumor histological type, radiation treatment dose, and intracranial and systemic disease burden.3,5,7,25 Other important factors to consider are the life expectancy of the patient and the possibility that imaging changes that appear consistent with treatment failure could at times represent radiation necrosis. Differentiating between radiation necrosis and tumor recurrence is an important consideration. Although the only definitive method of confirming recurrence is tissue biopsy, careful protocols of surveillance with perfusion-weighted MRI can distinguish between recurrence and necrosis,20 thus helping mitigate the risks associated with re-treatment.

Resection is used at some institutions as the salvage therapy of choice for patients who experience local treatment failure after SRS because it gives providers the ability to confirm progression by evaluating pathological specimens and has the potential to quickly palliate any symptoms present if the lesion is in a favorable location with regard to accessibility and proximity to eloquent brain structures. One drawback of surgery, however, is that it often requires further adjuvant therapy to decrease the risk of local failure.16,17 Another localized treatment option is laser thermoablation, an approach that has recently been used for surgically inaccessible lesions.19 In cases wherein treatment failure is more diffuse or there is evidence of leptomeningeal failure, whole-brain radiation therapy (WBRT) is commonly used.9 However, because WBRT increases the risk of acute worsening of performance status and chronic cognitive decline, its use for isolated local failure or for local failure with only limited distant brain metastases is often not ideal.6

In select patients who have experienced local treatment failure after SRS and in whom neither surgery nor WBRT is optimal, repeat SRS is often performed; however there is little data on the safety and efficacy of this technique. The concern with repeat SRS is the theoretically high risk of radiation necrosis, given the high cumulative dose and extreme hypofractionation of radiation delivered to the same volume. The current study represents the largest series with the longest follow-up in which outcomes are reported for patients receiving SRS as re-treatment following local failure after initial SRS. This study evaluates the clinical outcomes of repeat Gamma Knife (Elekta AB) SRS, including local control, overall survival (OS), and likelihood of neurological death, as well as factors predictive of radiation necrosis.

Methods

Data Acquisition

This is a single-institution retrospective review that was approved by our institutional review board. The Radiation Oncology Gamma Knife database at our institution was queried for all patients who received radiosurgical treatment more than once. The GammaPlan Treatment Planning System (Elekta AB) was subsequently used to identify lesions that were treated a second time after local failure of the initial radiosurgery. Local failure was defined either as a histologically proven recurrence or as serial increases in the size of an enhancing mass on axial slice MRI with corresponding increased perfusion on perfusion-weighted MRI. For small recurrences below the sensitivity of perfusion (< 1 cm3), tumors were followed with serial MRI, and a volumetric increase of ≥ 200% was required to determine local failure significant enough to warrant repeat SRS. Patients were selected for repeat SRS if they had either an unresectable local recurrence or multifocal (concurrent local and distant) recurrences that could be addressed in the same treatment session. Patients with small local recurrences were considered for this technique if they met the above criteria or if they had either worsening performance status or refused surgical treatment. Electronic medical records were reviewed to determine patient characteristics including age, sex, histological findings, extracranial disease status, Karnofsky Performance Scale (KPS) score, prior WBRT, and prior surgery. Outcomes including local control, toxicity, and radiation necrosis were also determined via the electronic medical records.

Radiosurgery Technique

Prior to SRS, each patient underwent high-resolution contrast-enhanced MRI studies of the brain. Treatment planning was performed using the GammaPlan system. The SRS was performed using either the Leksell Model B (used from 2002 to 2004), Model C (used from 2004 to 2009), or Perfexion (used from 2009 to 2015) Gamma Knife units (Elekta AB). Prescriptions for initial SRS treatments were determined based on guidelines published by Shaw et al.21 Doses chosen for the repeat SRS treatment were determined by physician discretion, but were generally also based on these guidelines. At the time of repeat SRS, doses were commonly chosen at the lower end of the dose range for an appropriate volume, particularly in those patients who had also received previous WBRT. In all patients in this series the SRS was delivered in a single fraction.

Response Assessment

Patients were followed with serial MRI after each SRS procedure. Posttreatment MRI was performed 6–8 weeks following SRS and subsequently every 3 months thereafter. Local failure was defined as either a histologically proven recurrence or serial increases in the size of the area of enhancement with corresponding increased perfusion on perfusion-weighted MRI. As stated above, small recurrences were followed with serial MRI, and a volumetric increase of ≥ 200% was required to determine local failure significant enough to warrant repeat SRS. For lesions > 1 cm3, local progression was determined by either biopsy confirmation or a 25% increase in the area of enhancement on axial slice MRI, with a corresponding increase in perfusion on perfusion-weighted MRI. For cases in which it was initially difficult to differentiate local failure from radiation necrosis or the development of granulation tissue, patients were treated conservatively with dexamethasone until recurrence was determined as detailed above. Cases were deemed radiation necrosis after either pathological confirmation or when serial imaging and conservative treatment with steroids ultimately led to stabilized imaging findings. The prescribed regimen of dexamethasone was 4 mg twice daily for 1 week, followed by a taper of 2 mg per week thereafter until the patient was taken off the medication.

Local failures were classified as either central or marginal failures. Central failures were defined as occurring within the prescribed treatment margin, whereas marginal failures were defined as occurring outside of the prescribed treatment margin. Figure 1 shows representative imaging for a patient treated with repeat SRS after a central failure. Neurological death was defined as previously reported by Patchell et al.17 Imaging results were also correlated with observed clinical deterioration as defined by the Common Terminology Criteria for Adverse Events (CTCAE), version 4.03. Patients diagnosed with radiation necrosis were treated conservatively with dexamethasone (duration and dose as detailed above) and the combination of vitamin E and pentoxifylline. In cases of severe radiation necrosis, patients underwent resection to palliate symptomatic deterioration.

FIG. 1.
FIG. 1.

A: Axial contrast-enhanced T1-weighted MRI sequence showing 50% isodose line from an SRS plan delivering 17 Gy to the margin of the tumor resection bed. B: Axial contrast-enhanced T1-weighted MRI sequence at time of a central local failure (24 months after initial SRS). The yellow line represents the new treatment volume, whereas the blue line represents prior treatment volume. C: Axial contrast-enhanced T1-weighted MRI sequence obtained 9 months after repeat SRS. D: Axial contrast-enhanced T1-weighted MRI sequence obtained 36 months after repeat SRS. Figure is available in color online only.

Statistical Analysis

Follow-up time and time-to-event outcomes were calculated from the time of SRS re-treatment to the time of most recent follow-up or to the event of interest. Time-to-event end points included time to local failure, time to death (i.e., OS), and time to development of radiation necrosis. All time-to-event data were summarized using the Kaplan-Meier estimator. Univariate Cox proportional hazards analysis was performed for each predictor variable and for all time-to-event outcomes.

Dosimetric Analysis

For each SRS treatment, relevant Digital Imaging and Communications in Medicine–Radiation Therapy (DICOM-RT) files were exported from GammaPlan and reconstructed using MIM Maestro v6.4 (MIM Software). Radiation therapy doses were then calculated using the Tissue Maximum Ratio 10-dose algorithm and composite plans were then formed, which allowed for the creation of an absolute cumulative dose-volume histogram (DVH) for each treated lesion. Absolute dose, rather than biologically equivalent dose, was used to maximize clinical applicability. Dose-volume relationships were analyzed at intervals of 0.1 Gy from total doses of 0–100 Gy. Serial logistic regression analyses were performed for each discrete dose (at 0.1-Gy intervals) received by a certain volume. A fitted logistic regression model was developed after selecting a discrete dose that was statistically predictive of radiation necrosis as well as clinically applicable.

Results

Patient Population

Between 2002 and 2015, 738 patients received upfront SRS for brain metastases, with 58 experiencing local failure. Of these 58 patients experiencing local failure, 32 were treated with a second course of Gamma Knife SRS to a total of 46 discrete lesions. Local failure in this cohort was confirmed by biopsy in 11 patients (34%); in the remainder failure was determined by serial increases in size on axial MRI in accordance with the aforementioned response criteria. The median patient age was 59 years (range 36–88 years). The median time to repeat SRS was 19 months (range 2–98 months). The median tumor volume at first and second SRS was 1.28 cm3 (range 0.01–22.57 cm3) and 0.98 cm3 (range 0.01–19.74 cm3), respectively. The mean tumor volume at repeat SRS was 2.23 cm3.

Six patients (19%) had multiple lesions treated with repeat SRS. The most common primary tumors originated in the lung, with 10 (31%) identified as adenocarcinoma, 6 (19%) as squamous cell carcinoma, and 1 (3%) as small cell carcinoma. Nine (28%) primary tumors were of breast origin, with 3 (9%) were identified as Her2/Neu-positive adenocarcinoma. Two (6%) patients presented with renal cell carcinoma and another 2 (6%) with metastatic melanoma. A single patient (3%) was treated for a brain metastasis from esthesioneuroblastoma. Twenty-one (66%) patients had stable extracranial disease at the time of repeat SRS, with the remaining 11 (34%) showing evidence of progressive disease. The median KPS score was 80 (range 60–100). The median dose delivered to the tumor margin at the time of first SRS was 20 Gy (range 12–24 Gy). The median dose delivered to the tumor margin at the time of repeat SRS was 20 Gy (range 14–22 Gy). Figure 2 plots the radiation doses chosen for each patient at initial and repeat SRS. The SRS was delivered in a single fraction to each lesion. Table 1 summarizes the patient characteristics in this study.

FIG. 2.
FIG. 2.

Radar plot of initial and re-treatment radiation doses (in Gy) for each lesion. Cases are numbered circumferentially and doses are plotted from the center of the figure extending outward radially. Brackets indicate multiple lesions treated in a single patient.

TABLE 1.

Characteristics in 32 patients and 46 individual lesions

CharacteristicsValue (%)
Patientsn = 32
  Mean age in yrs, range59, 36–88
  Median time in mos to repeat SRS, range19, 2–98
  Prior WBRT8 (25)
  Extracranial disease status
    Stable21 (66)
    Progressive11 (34)
  KPS score
    100–8012 (38)
    ≤7020 (62)
  Primary tumor type/location
    NSCLC16 (50)
    Breast9 (28)
    Melanoma2 (6)
    RCC2 (6)
    Esthesioneuroblastoma1 (3)
    Colorectal carcinoma1 (3)
    SCLC1 (3)
  No. of brain metastases receiving repeat SRS per patient
    126 (82)
    2–33 (9)
    ≥43 (9)
Lesionsn = 46
  Median time in mos to repeat SRS, range19, 2–98
  Median vol of lesion in cm3 at initial SRS, range1.28, 0.01–22.57
  Median vol of lesion in cm3 at repeat SRS, range0.98, 0.01–19.74
  Mean vol of lesion in cm3 at initial SRS, SD3.57, 5.26
  Mean vol of lesion in cm3 at repeat SRS, SD2.23, 3.65
  Median initial SRS treatment dose in Gy, range20, 12–24
  Median repeat SRS treatment dose in Gy, range20, 14–22
  Posterior fossa17 (37)
  1st SRS to a resection cavity9 (20)
  2nd SRS to a resection cavity14 (30)
  Failure location
    Central18 (39)
    Marginal28 (61)
  Location
    Cerebellum15 (33)
    Frontal14 (30)
    Parietal6 (13)
    Temporal6 (13)
    Pons2 (5)
    Motor strip1 (2)
    Parasagittal1 (2)
    Periventricular1 (2)
  Primary tumor type/location
    NSCLC21 (46)
    Breast18 (40)
    Melanoma2 (4)
    RCC2 (4)
    Esthesioneuroblastoma1 (2)
    Colorectal carcinoma1 (2)
    SCLC1 (2)
  Radiation necrosis following repeat SRS14 (30)

NSCLC = non–small cell lung carcinoma; RCC = renal cell carcinoma; SCLC = small cell lung carcinoma.

Overall Survival

The median length of follow-up after repeat SRS was 24 months (range 2–124 months). The OS at 1 year following re-treatment was 70% (95% CI 55%–88%). The median OS had not yet been reached at the time of analysis. The Kaplan-Meier plot for OS is shown in Fig. 3A.

FIG. 3.
FIG. 3.

Kaplan-Meier plots for OS (A), freedom from local failure (B), and freedom from radiation necrosis (C), all following re-treatment with SRS. Dashed lines represent 95% CIs. The number of cases reaching each time marker is noted at the bottom of each graph.

Cox univariate analysis revealed a statistically significant relationship between the lowest treatment dose per patient at the time of the second SRS and OS (hazard ratio [HR] 0.64, p = 0.001). Table 2 depicts Cox univariate analysis for factors affecting OS by patient. The most common cause of mortality was neurological death (67%).

TABLE 2.

Cox proportional hazards univariate analysis of factors affecting OS (by patient), local failure (by metastasis), and radiation necrosis (by metastasis)

VariableHR (95% CI)p Value
OS
  Age1.02 (0.97–1.07)0.398
  >1 metastasis treated w/repeat SRS0.97 (0.63–1.50)0.907
  Time from initial SRS to repeat SRS1.00 (0.97–1.07)0.845
  ≥1 lesion in posterior fossa0.93 (0.29–2.94)0.900
  1st SRS to a resection cavity0.24 (0.03–1.85)0.169
  2nd SRS to a resection cavity0.64 (0.19–2.14)0.467
  Prior WBRT2.04 (0.64–6.46)0.226
  Radiation necrosis in ≥1 lesion0.87 (0.27–2.76)0.810
  Symptomatic radiation necrosis1.70 (0.19–15.43)0.636
  Vitamin E or pentoxifylline given0.76 (0.23–2.57)0.659
  Lowest dose at 2nd SRS0.64 (0.49–0.84)0.001
  KPS Score ≥800.34 (0.10–1.08)0.068
  Largest tumor size at 1st SRS1.01 (0.91–1.12)0.858
  Total tumor burden at 1st SRS1.01 (0.91–1.12)0.843
  Largest tumor size at 2nd SRS1.08 (0.97–1.21)0.178
  Total tumor burden at 2nd SRS1.08 (0.97–1.20)0.184
  Progressive systemic disease0.71 (0.19–2.71)0.620
Local failure
  Time from initial SRS to repeat SRS1.01 (0.98–1.04)0.358
  Tumor vol at 1st SRS1.04 (0.93–1.15)0.506
  Tumor vol at 2nd SRS0.71 (0.32–1.55)0.389
  Treatment dose at 1st SRS0.94 (0.74–1.19)0.616
  Treatment dose at 2nd SRS0.80 (0.56–1.15)0.225
  Total combined SRS dose0.91 (0.76–1.09)0.288
  Location in posterior fossa0.55 (0.14–2.23)0.406
  2nd SRS to a resection cavity0.53 (0.11–2.60)0.435
  Prior WBRT2.97 (0.74–11.97)0.126
  Radiation necrosis1.35 (0.36–5.09)0.653
Radiation necrosis
  Time from initial SRS to repeat SRS1.00 (0.98–1.03)0.839
  Tumor vol at 1st SRS1.03 (0.93–1.13)0.526
  Tumor vol at 2nd SRS1.19 (1.07–1.32)0.002
  Treatment dose at 1st SRS0.90 (0.75–1.08)0.272
  Treatment dose at 2nd SRS0.64 (0.48–0.84)0.002
  Prior WBRT0.89 (0.29–2.68)0.831
  Location in posterior fossa0.93 (0.32–2.70)0.896
  1st SRS to a resection cavity0.68 (0.15–3.10)0.620
  2nd SRS to a resection cavity0.84 (0.26–2.69)0.768

Patterns of Failure

One-year local control for each lesion was 79% (95% CI 67%–94%). Figure 3B depicts a Kaplan-Meier plot for freedom from local failure. The median time to local failure after the second SRS was 95.6 months, with only 9 events reported. Of those 9, 5 (56%) occurred outside the margins of the treatment area and 4 (44%) occurred within the margins of the treatment area.

Radiation Necrosis and Adverse Events

In 14 of 46 cases (30%) the patient experienced radiation necrosis, with 11 of 46 (24%) showing evidence of symptomatic decline. In 3 (9%) patients radiation necrosis was confirmed by biopsy, with the remainder confirmed by imaging. Five patients (16%) experienced unsteadiness and dyscoordination as a result of cerebellar radiation necrosis; all of them showed improvement over time with corticosteroid treatment and rehabilitation. Three patients (9%) experienced motor weakness of the hand and face; 2 of them showed some improvement with therapy. Two patients (6%) experienced visual field changes; one had some improvement, although with residual blurriness, whereas the other still has a complete homonymous hemianopia after resection of significant radiation necrosis. One patient experienced transient word-finding difficulty, and another experienced expressive aphasia and continues to have some residual difficulty. A single patient had hemorrhage at the re-treatment site, which required surgical intervention. Aside from toxicity from radiation necrosis, 4 patients (13%) experienced Grade 1 toxicity according to the Radiation Therapy Oncology Group grading system; there were no other serious adverse events.

After re-treatment, the 1-year freedom from radiation necrosis was 71% (95% CI 57%–88%). The median time from repeat SRS to the development of radiation necrosis for the entire cohort was 44.4 months by Kaplan-Meier estimation. For those lesions in which radiation necrosis actually developed, the mean time to this event was 8.0 months (interquartile range [IQR] 3.8–19.7 months). The median time to radiation necrosis after initial SRS as estimated by the Kaplan-Meier method was 91.1 months. The median time to necrosis for tumors that actually developed radiation necrosis after the first SRS was 28.4 months (IQR 21.0–54.7). Figure 3C depicts a Kaplan-Meier plot for freedom from radiation necrosis. The volume of recurrent tumor at the time of repeat SRS was predictive of radiation necrosis on univariate analysis (HR 1.19, p = 0.002).

Of 46 lesions treated, 38 had the necessary dosimetric data for inclusion in dosimetric analysis. For these lesions, cumulative DVH data were analyzed and serial logistic regression analyses were performed at 0.1-Gy intervals, the results of which revealed the cumulative dose delivered to a volume of tissue (VD) for doses between 22 Gy and 63 Gy to be statistically significant predictors of the development of radiation necrosis. The most statistically significant relationship between the VD and radiation necrosis was at V40.8Gy (p = 0.003) (Fig. 4). To maximize clinical applicability, V40Gy was used to develop the fitted logistic regression model (Fig. 5). The following V40Gy thresholds were associated with 10%, 20%, and 50% probabilities of radiation necrosis, respectively: 0.28 cm3 (95% CI 3%–28%), 0.76 cm3 (95% CI 9%–39%), 1.60 cm3 (95% CI 26%–74%).

FIG. 4.
FIG. 4.

Logistic regression p values for radiation necrosis as predicted by dose to volume of tissue (DV), both for the entire range of significant values (upper), and also for the range wherein the most statistically significant relationships were identified (lower).

FIG. 5.
FIG. 5.

Fitted logistic regression model predicting the probability of radiation necrosis (PRN) by volume of brain receiving a cumulative dose of ≥ 40 Gy (V40Gy). The V40Gy values predictive of a probability of radiation necrosis of 10%, 20%, and 50% are 0.28 cm3, 0.76 cm3, and 1.60 cm3, respectively.

Further Salvage

Of the 8 patients (with a total of 9 lesions) who experienced local failure after repeat SRS, 3 went on to receive WBRT, 3 underwent resection, a single patient underwent laser thermoablation, and 1 patient opted for no further therapy and was placed in hospice care. Three patients experiencing local failure were dead at the time of data analysis, with 2 of these deaths being attributable to neurological causes.

Discussion

In the current study we investigated repeat SRS for brain metastases in patients who have experienced local treatment failure. This series demonstrates durable local control, successful salvage outcomes, and a possible dosimetric threshold for predicting radiation necrosis. Several salvage approaches currently exist for addressing local failure after radiosurgery, and the clinical scenario may ultimately predict which approach is optimal. Repeat SRS may be a reasonable option for cases wherein patients experience simultaneous local and distant brain recurrences, because SRS can be applied to treat both progressive and new brain metastases in the same treatment session while minimizing the total radiation exposure of surrounding tissue. Repeat SRS may also be helpful for lesions that are located in surgically inaccessible regions. Repeat SRS is probably a suboptimal approach in the treatment of larger lesions or in lesions that have experienced rapid local failure, due to the high risk of suboptimal local control or the development of radiation necrosis in those situations.

Several patients in the current series lived multiple years after repeat SRS for treatment of a local failure. This finding highlights the survivability following local failure if properly treated. Several recent analyses have been performed assessing the causes of death in patients with brain metastases. Whereas most patients with brain metastases ultimately die of systemic disease, those that have controlled extracranial disease are more likely to die of brain metastases.13 Patients with well-controlled extracranial disease are also more likely to experience local failures intracranially. This would explain the high rate of neurological death (67%) for patients who died in the current series. Durable salvage options for intracranial disease are thus quite important for patients who experience local failure after initial treatment with SRS.

In the present literature, there exist only a few series reporting outcomes following repeat radiosurgery for local failure in brain metastases. Jayachandran et al. showed a higher radiation necrosis rate in lesions treated with repeat SRS following local failure as compared with either a single application of SRS or SRS after failure of conventionally fractionated radiotherapy.10 Bhatnagar et al. reported good treatment response in both local and distant brain recurrences with repeat SRS. They also noted an association between increased re-treatment volume and the probability of neurological decline.2 Holt et al. reported acceptable rates of toxicity in patients treated with repeat SRS following resection of a prior SRS. Resection of the previously treated lesion may have influenced the toxicity rates by decreasing the volume of re-treated tumor.8 In our series and in those mentioned above, radiosurgery was delivered in a single fraction for each patient. A recent report by McTyre et al., however, describes good outcomes in 3 patients with brain metastases treated with hypofractionated SRS after having failed single-fraction radiosurgery; this strategy was chosen for large lesions deemed to be at higher risk for radiation necrosis.15 Given that hypofractionation is thought to decrease the risk of radiation necrosis, their findings represent an intriguing clinical option for patients with local recurrence after a prior SRS treatment.

In the present series, approximately 24% of re-treated lesions were associated with symptomatic radiation necrosis. Three patients (9%) who experienced necrosis had a CTCAE Grade 3 or higher toxicity. Five patients were left with residual deficits after the development of radiation necrosis, the severity of which was dependent on the grade of toxicity and the proximity of the lesion to eloquent brain. No patients died of toxicity. A single patient in the present series had a lesion that was re-treated within the brainstem. This particular patient was not deemed a surgical candidate, experienced improvement of a facial nerve palsy after repeat SRS, and lived for another 7 months without evidence of radiation necrosis. Although these results are encouraging, there is certainly a need for additional reported cases before re-treatment of lesions within the brainstem can be considered safe, because prior series have demonstrated higher rates of toxicity with SRS in this location.12

Volumetric analysis from the present series was able to determine that the volume that received a cumulative dose of 40 Gy over 2 SRS treatments was predictive of the development of radiation necrosis. This threshold represents a significant advance; the practice of repeating SRS after previously treated SRS failures has been communicated anecdotally at some centers and is scantily reported, with no previous evidence for dose or volume constraints. It is thought that the mechanism of damage to tumors as well as normal tissues with extreme hypofractionation is different than with conventionally fractionated radiation because the vasculature is affected more with hypofractionation.24 Several investigators have reported the use of repeat SRS for arteriovenous malformations in which there were acceptable rates of radiation necrosis.18,23 The use of this dosimetric data in the future will probably be 2-fold. First, it may assist in determining which patients are candidates for safe administration of repeat SRS treatments. Second, it may guide the process of determining the dose prescribed at re-treatment. Patients with lesions predicted to receive a cumulative V40Gy < 0.3 cm3 after repeat SRS have a 10% probability of radiation necrosis and are probably safe to re-treat with SRS. For lesions with a predicted V40Gy > 1.6 cm3 (50% likelihood of radiation necrosis), lowering the treatment dose or finding an alternative treatment modality would be advisable.

There are several limitations to the current series. As a retrospective series, its findings are generally limited to hypothesis generation. The potential use of the V40Gy as a dosimetric constraint to consider in re-treatment scenarios is intriguing, although further evaluation of the V40Gy and other potential dosimetric parameters as possible predictors of radiation necrosis is warranted. This is particularly true given the small sample size and number of radiation necrosis events in this series. Furthermore, all patients in this series received Gamma Knife SRS, so despite the theoretical generalizability of Gamma Knife SRS dosimetric constraints to LINAC-based SRS, this deserves further investigation. Despite these limitations, and in the absence of other constraints from which to guide the re-treatment of brain metastases with SRS, minimizing the V40Gy may be a useful strategy for clinicians who wish to decrease the probability of radiation necrosis and its associated toxicities.

Conclusions

Repeat Gamma Knife SRS appears to be an effective salvage option for brain metastases initially treated with SRS. This series demonstrates durable local control and successful salvage outcomes. Although this strategy does carry a relatively high risk for radiation necrosis, it may be indicated for select cases of disease recurrence. Furthermore, the risk for radiation necrosis may be minimized with careful patient selection and treatment planning. The volume receiving 40 Gy is the most significant predictor of the development of radiation necrosis; therefore it may be a useful consideration during the SRS planning process as a means of minimizing the risk to these patients.

Disclosures

Dr. Chan is on the advisory board of Novocure.

Author Contributions

Conception and design: all authors. Acquisition of data: McKay, McTyre, Alphonse-Sullivan, Ruiz, Munley, Qasem, Laxton, Tatter, Chan. Analysis and interpretation of data: McKay, Okoukoni, Lo, Xing, Laxton, Tatter, Watabe, Chan. Drafting the article: McKay, Chan. Critically revising the article: all authors. Reviewed submitted version of manuscript: all authors. Approved the final version of the manuscript on behalf of all authors: McKay. Statistical analysis: McTyre. Study supervision: Chan.

References

  • 1

    Baschnagel AMMeyer KDChen PYKrauss DJOlson REPieper DR: Tumor volume as a predictor of survival and local control in patients with brain metastases treated with Gamma Knife surgery. J Neurosurg 119:113911442013

  • 2

    Bhatnagar AHeron DEKondziolka DLunsford LDFlickinger JC: Analysis of repeat stereotactic radiosurgery for progressive primary and metastatic CNS tumors. Int J Radiat Oncol Biol Phys 53:5275322002

  • 3

    Black PJPage BRLucas JT JrHughes RTLaxton AWTatter SB: Factors that determine local control with Gamma Knife radiosurgery: the role of primary histology. J Radiosurg SBRT 3:2812862015

  • 4

    Cho KRLee MHKong DSSeol HJNam DHSun JM: Outcome of Gamma Knife radiosurgery for metastatic brain tumors derived from non-small cell lung cancer. J Neurooncol 125:3313382015

  • 5

    Ebner DRava PGorovets DCielo DHepel JT: Stereotactic radiosurgery for large brain metastases. J Clin Neurosci 22:165016542015

  • 6

    Greene-Schloesser DRobbins MEPeiffer AMShaw EGWheeler KTChan MD: Radiation-induced brain injury: a review. Front Oncol 2:732012

  • 7

    Greto DScoccianti SCompagnucci AArilli CCasati MFrancolini G: Gamma Knife radiosurgery in the management of single and multiple brain metastases. Clin Neurol Neurosurg 141:43472016

  • 8

    Holt DEGill BSClump DALeeman JEBurton SAAmankulor NM: Tumor bed radiosurgery following resection and prior stereotactic radiosurgery for locally persistent brain metastases. Front Oncol 5:842015

  • 9

    Huang AJHuang KEPage BRAyala-Peacock DNLucas JT JrLesser GJ: Risk factors for leptomeningeal carcinomatosis in patients with brain metastases who have previously undergone stereotactic radiosurgery. J Neurooncol 120:1631692014

  • 10

    Jayachandran PShultz DModlin LVon Eyben RGibbs ICChang S: Repeat stereotactic radiosurgery (SRS) for brain metastases locally recurrent following initial SRS. Int J Radiat Oncol Biol Phys 90:S3202014

  • 11

    Jensen CAChan MDMcCoy TPBourland JDdeGuzman AFEllis TL: Cavity-directed radiosurgery as adjuvant therapy after resection of a brain metastasis. J Neurosurg 114:158515912011

  • 12

    Kilburn JMEllis TLLovato JFUrbanic JJBourland JDMunley MT: Local control and toxicity outcomes in brainstem metastases treated with single fraction radiosurgery: is there a volume threshold for toxicity?. J Neurooncol 117:1671742014. (Erratum in J Neurooncol 120: 223 2014)

  • 13

    Kocher MSoffietti RAbacioglu UVillà SFauchon FBaumert BG: Adjuvant whole-brain radiotherapy versus observation after radiosurgery or surgical resection of one to three cerebral metastases: results of the EORTC 22952-26001 study. J Clin Oncol 29:1341412011

  • 14

    Lucas JT JrColmer HG IVWhite LFitzgerald NIsom SBourland JD: Competing risk analysis of neurologic versus nonneurologic death in patients undergoing radiosurgical salvage after whole-brain radiation therapy failure: who actually dies of their brain metastases?. Int J Radiat Oncol Biol Phys 92:100810152015

  • 15

    McTyre EHelis CAFarris MWilkins LSloan DHinson WH: Emerging indications for fractionated Gamma Knife radiosurgery. Neurosurgery [epub ahead of print]2016

  • 16

    Mu FLucas JT JrWatts JMJohnson AJDaniel Bourland JLaxton AW: Tumor resection with carmustine wafer placement as salvage therapy after local failure of radiosurgery for brain metastasis. J Clin Neurosci 22:5615652015

  • 17

    Patchell RATibbs PARegine WFDempsey RJMohiuddin MKryscio RJ: Postoperative radiotherapy in the treatment of single metastases to the brain: a randomized trial. JAMA 280:148514891998

  • 18

    Pollock BEKondziolka DLunsford LDBissonette DFlickinger JC: Repeat stereotactic radiosurgery of arteriovenous malformations: factors associated with incomplete obliteration. Neurosurgery 38:3183241996

  • 19

    Rao MSHargreaves ELKhan AJHaffty BGDanish SF: Magnetic resonance-guided laser ablation improves local control for postradiosurgery recurrence and/or radiation necrosis. Neurosurgery 74:6586672014

  • 20

    Reddy KWesterly DChen C: MRI patterns of T1 enhancing radiation necrosis versus tumour recurrence in high-grade gliomas. J Med Imaging Radiat Oncol 57:3493552013

  • 21

    Shaw EScott CSouhami LDinapoli RKline RLoeffler J: Single dose radiosurgical treatment of recurrent previously irradiated primary brain tumors and brain metastases: final report of RTOG protocol 90-05. Int J Radiat Oncol Biol Phys 47:2912982000

  • 22

    Sperduto PWWang MRobins HISchell MCWerner-Wasik MKomaki R: A phase 3 trial of whole brain radiation therapy and stereotactic radiosurgery alone versus WBRT and SRS with temozolomide or erlotinib for non-small cell lung cancer and 1 to 3 brain metastases: Radiation Therapy Oncology Group 0320. Int J Radiat Oncol Biol Phys 85:131213182013

  • 23

    Stahl JMChi YYFriedman WA: Repeat radiosurgery for intracranial arteriovenous malformations. Neurosurgery 70:1501542012

  • 24

    Truman JPGarcia-Barros MKaag MHambardzumyan DStancevic BChan M: Endothelial membrane remodeling is obligate for anti-angiogenic radiosensitization during tumor radiosurgery. PLoS One 5:2010

  • 25

    Won YKLee JYKang YNJang JSKang JHJung SL: Stereotactic radiosurgery for brain metastasis in non-small cell lung cancer. Radiat Oncol J 33:2072162015

  • 26

    Yamamoto MSerizawa TShuto TAkabane AHiguchi YKawagishi J: Stereotactic radiosurgery for patients with multiple brain metastases (JLGK0901): a multi-institutional prospective observational study. Lancet Oncol 15:3873952014

If the inline PDF is not rendering correctly, you can download the PDF file here.

Article Information

Correspondence Will McKay, 1 Medical Center Blvd., Winston-Salem, NC 27157. email: wmckay@wakehealth.edu.

INCLUDE WHEN CITING Published online August 5, 2016; DOI: 10.3171/2016.5.JNS153051.

Disclosures Dr. Chan is on the advisory board of Novocure.

© AANS, except where prohibited by US copyright law.

Headings

Figures

  • View in gallery

    A: Axial contrast-enhanced T1-weighted MRI sequence showing 50% isodose line from an SRS plan delivering 17 Gy to the margin of the tumor resection bed. B: Axial contrast-enhanced T1-weighted MRI sequence at time of a central local failure (24 months after initial SRS). The yellow line represents the new treatment volume, whereas the blue line represents prior treatment volume. C: Axial contrast-enhanced T1-weighted MRI sequence obtained 9 months after repeat SRS. D: Axial contrast-enhanced T1-weighted MRI sequence obtained 36 months after repeat SRS. Figure is available in color online only.

  • View in gallery

    Radar plot of initial and re-treatment radiation doses (in Gy) for each lesion. Cases are numbered circumferentially and doses are plotted from the center of the figure extending outward radially. Brackets indicate multiple lesions treated in a single patient.

  • View in gallery

    Kaplan-Meier plots for OS (A), freedom from local failure (B), and freedom from radiation necrosis (C), all following re-treatment with SRS. Dashed lines represent 95% CIs. The number of cases reaching each time marker is noted at the bottom of each graph.

  • View in gallery

    Logistic regression p values for radiation necrosis as predicted by dose to volume of tissue (DV), both for the entire range of significant values (upper), and also for the range wherein the most statistically significant relationships were identified (lower).

  • View in gallery

    Fitted logistic regression model predicting the probability of radiation necrosis (PRN) by volume of brain receiving a cumulative dose of ≥ 40 Gy (V40Gy). The V40Gy values predictive of a probability of radiation necrosis of 10%, 20%, and 50% are 0.28 cm3, 0.76 cm3, and 1.60 cm3, respectively.

References

1

Baschnagel AMMeyer KDChen PYKrauss DJOlson REPieper DR: Tumor volume as a predictor of survival and local control in patients with brain metastases treated with Gamma Knife surgery. J Neurosurg 119:113911442013

2

Bhatnagar AHeron DEKondziolka DLunsford LDFlickinger JC: Analysis of repeat stereotactic radiosurgery for progressive primary and metastatic CNS tumors. Int J Radiat Oncol Biol Phys 53:5275322002

3

Black PJPage BRLucas JT JrHughes RTLaxton AWTatter SB: Factors that determine local control with Gamma Knife radiosurgery: the role of primary histology. J Radiosurg SBRT 3:2812862015

4

Cho KRLee MHKong DSSeol HJNam DHSun JM: Outcome of Gamma Knife radiosurgery for metastatic brain tumors derived from non-small cell lung cancer. J Neurooncol 125:3313382015

5

Ebner DRava PGorovets DCielo DHepel JT: Stereotactic radiosurgery for large brain metastases. J Clin Neurosci 22:165016542015

6

Greene-Schloesser DRobbins MEPeiffer AMShaw EGWheeler KTChan MD: Radiation-induced brain injury: a review. Front Oncol 2:732012

7

Greto DScoccianti SCompagnucci AArilli CCasati MFrancolini G: Gamma Knife radiosurgery in the management of single and multiple brain metastases. Clin Neurol Neurosurg 141:43472016

8

Holt DEGill BSClump DALeeman JEBurton SAAmankulor NM: Tumor bed radiosurgery following resection and prior stereotactic radiosurgery for locally persistent brain metastases. Front Oncol 5:842015

9

Huang AJHuang KEPage BRAyala-Peacock DNLucas JT JrLesser GJ: Risk factors for leptomeningeal carcinomatosis in patients with brain metastases who have previously undergone stereotactic radiosurgery. J Neurooncol 120:1631692014

10

Jayachandran PShultz DModlin LVon Eyben RGibbs ICChang S: Repeat stereotactic radiosurgery (SRS) for brain metastases locally recurrent following initial SRS. Int J Radiat Oncol Biol Phys 90:S3202014

11

Jensen CAChan MDMcCoy TPBourland JDdeGuzman AFEllis TL: Cavity-directed radiosurgery as adjuvant therapy after resection of a brain metastasis. J Neurosurg 114:158515912011

12

Kilburn JMEllis TLLovato JFUrbanic JJBourland JDMunley MT: Local control and toxicity outcomes in brainstem metastases treated with single fraction radiosurgery: is there a volume threshold for toxicity?. J Neurooncol 117:1671742014. (Erratum in J Neurooncol 120: 223 2014)

13

Kocher MSoffietti RAbacioglu UVillà SFauchon FBaumert BG: Adjuvant whole-brain radiotherapy versus observation after radiosurgery or surgical resection of one to three cerebral metastases: results of the EORTC 22952-26001 study. J Clin Oncol 29:1341412011

14

Lucas JT JrColmer HG IVWhite LFitzgerald NIsom SBourland JD: Competing risk analysis of neurologic versus nonneurologic death in patients undergoing radiosurgical salvage after whole-brain radiation therapy failure: who actually dies of their brain metastases?. Int J Radiat Oncol Biol Phys 92:100810152015

15

McTyre EHelis CAFarris MWilkins LSloan DHinson WH: Emerging indications for fractionated Gamma Knife radiosurgery. Neurosurgery [epub ahead of print]2016

16

Mu FLucas JT JrWatts JMJohnson AJDaniel Bourland JLaxton AW: Tumor resection with carmustine wafer placement as salvage therapy after local failure of radiosurgery for brain metastasis. J Clin Neurosci 22:5615652015

17

Patchell RATibbs PARegine WFDempsey RJMohiuddin MKryscio RJ: Postoperative radiotherapy in the treatment of single metastases to the brain: a randomized trial. JAMA 280:148514891998

18

Pollock BEKondziolka DLunsford LDBissonette DFlickinger JC: Repeat stereotactic radiosurgery of arteriovenous malformations: factors associated with incomplete obliteration. Neurosurgery 38:3183241996

19

Rao MSHargreaves ELKhan AJHaffty BGDanish SF: Magnetic resonance-guided laser ablation improves local control for postradiosurgery recurrence and/or radiation necrosis. Neurosurgery 74:6586672014

20

Reddy KWesterly DChen C: MRI patterns of T1 enhancing radiation necrosis versus tumour recurrence in high-grade gliomas. J Med Imaging Radiat Oncol 57:3493552013

21

Shaw EScott CSouhami LDinapoli RKline RLoeffler J: Single dose radiosurgical treatment of recurrent previously irradiated primary brain tumors and brain metastases: final report of RTOG protocol 90-05. Int J Radiat Oncol Biol Phys 47:2912982000

22

Sperduto PWWang MRobins HISchell MCWerner-Wasik MKomaki R: A phase 3 trial of whole brain radiation therapy and stereotactic radiosurgery alone versus WBRT and SRS with temozolomide or erlotinib for non-small cell lung cancer and 1 to 3 brain metastases: Radiation Therapy Oncology Group 0320. Int J Radiat Oncol Biol Phys 85:131213182013

23

Stahl JMChi YYFriedman WA: Repeat radiosurgery for intracranial arteriovenous malformations. Neurosurgery 70:1501542012

24

Truman JPGarcia-Barros MKaag MHambardzumyan DStancevic BChan M: Endothelial membrane remodeling is obligate for anti-angiogenic radiosensitization during tumor radiosurgery. PLoS One 5:2010

25

Won YKLee JYKang YNJang JSKang JHJung SL: Stereotactic radiosurgery for brain metastasis in non-small cell lung cancer. Radiat Oncol J 33:2072162015

26

Yamamoto MSerizawa TShuto TAkabane AHiguchi YKawagishi J: Stereotactic radiosurgery for patients with multiple brain metastases (JLGK0901): a multi-institutional prospective observational study. Lancet Oncol 15:3873952014

TrendMD

Cited By

Metrics

Metrics

All Time Past Year Past 30 Days
Abstract Views 0 0 0
Full Text Views 389 389 68
PDF Downloads 321 321 38
EPUB Downloads 0 0 0

PubMed

Google Scholar