Resection Of Brain Metastases Research Articles

Case report: Long-term intracranial effect of zimberelimab monotherapy following surgical resection of high PD-L1-expressing brain metastases from NSCLC

Non-small cell lung cancer (NSCLC) accounted for the majority of lung cancer cases worldwide. Brain metastases (BM) frequently complicate NSCLC and portend a dismal prognosis. To control neurological symptoms, surgical resection is commonly followed by brain radiotherapy (RT). However, RT is often complicated by neurotoxicity. For patients with tumors that harbor positive driver genes, tyrosine kinase inhibitors are considered the standard of care. Nevertheless, treatment options for those without driver gene mutations are still debated. Programmed death receptor 1 (PD-1)/ligand 1 (PD-L1) inhibition has emerged as a novel therapeutic strategy for NSCLC patients with PD-L1-positive tumors, as well as for those with asymptomatic BM. However, the effect of anti-PD-1 antibodies on active BM within such specific populations is undetermined. Herein we present a case of a 65-year-old patient with NSCLC and high PD-L1-expressing BM. The patient underwent surgical resection of BM followed by first-line monotherapy with 31 cycles of zimberelimab, a novel anti-PD-1 antibody, and has already achieved 24 months of progression-free survival and intracranial recurrence-free survival. To our knowledge, this is the first report regarding the intracranial effect of zimberelimab on BM from primary lung cancer. This case report might facilitate an understanding of the intracranial effects of different anti-PD-1 antibodies for such populations.

Abstract PO5-20-03: Phase I/II study of stereotactic radiosurgery with concurrent olaparib followed by adjuvant durvalumab and physician’s choice systemic therapy in subjects with breast cancer brain metastases

Abstract BACKGROUND: Despite progress in the treatment of brain metastasis for HER2+ breast cancer, outcomes for patients with HER2-negative breast cancer brain metastases remain poor. Current standard of care consists of local therapies, including surgery and radiotherapy, followed by systemic therapy. Preclinical studies show inhibitors of poly(ADP-ribose) polymerase (PARP) are effective in combination with radiation therapy as a DNA damage response inhibitor. Triple-negative breast cancer (TNBC) has higher rates of homologous recombination deficiency compared to other breast cancer subtypes, and together with HER2-negative, BRCA-mutated breast cancer would be particularly sensitive to PARP inhibition. PARP inhibition has also demonstrated promising efficacy combined with immunotherapy in patients with germline BRCA-mutant and metastatic TNBC in clinical trials (MEDIOLA, TOPACIO). In addition, immunotherapy with stereotactic radiosurgery (SRS) is associated with favorable intracranial control and survival in patients with brain metastases. We hypothesize that this biologically-driven combination will enhance local control of SRS-treated brain metastases through synergy with PARP inhibition, while controlling micrometastatic disease in the brain and extracranial sites via potentiation of the immune response. METHODS: We are conducting a multi-institution, Phase I/II trial of SRS plus olaparib, followed by durvalumab (with physician’s choice systemic therapy), for patients with TNBC (any BRCA status) or HER2-negative with BRCA-mutated (germline or somatic) breast cancer brain metastases [NCT04711824]. A total of 41 patients are planned for enrollment at 8 sites. The primary objectives are to evaluate safety and tolerability (Phase I) and determine intracranial disease control at 6 months (Phase II) of this treatment combination. Secondary objectives include determining clinical activity via intracranial and global progression-free survival, overall survival, and intracranial and extracranial response rate. Exploratory objectives will assess potential biomarkers of treatment response, including changes in circulating tumor cells and DNA in blood and cerebrospinal fluid, germline and tumor mutations in DNA repair pathway genes, and PD-L1 expression, as well as quality of life and patient-reported outcomes. A surgical sub-study (n=5) will evaluate olaparib concentration/distribution in resected brain metastases. As of June 2023, cohort 2 of phase I has been completed without dose-limiting toxicity. Citation Format: Colette Shen, Yara Abdou, Linda Chen, Xianming Tan, Gaorav Gupta, Mina Lobbous, Filipa Lynce, Erica Stringer-Reasor, Carey Anders. Phase I/II study of stereotactic radiosurgery with concurrent olaparib followed by adjuvant durvalumab and physician’s choice systemic therapy in subjects with breast cancer brain metastases [abstract]. In: Proceedings of the 2023 San Antonio Breast Cancer Symposium; 2023 Dec 5-9; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2024;84(9 Suppl):Abstract nr PO5-20-03.

Abstract PS11-02: Clinical risk factors of Central Nervous System (CNS)-related death in patients with HER2-positive metastatic breast cancer

Abstract Background: Central nervous system (CNS) relapse represents a challenge in the management of patients with HER2-positive (+) metastatic breast cancer. Data regarding the impact of brain metastases (BrM) and CNS involvement on mortality are sparse. In this study we sought to determine the proportion of HER2+ MBC patients in which CNS disease is the cause of death and identify risk factors associated with CNS-related mortality. Methods: We reviewed medical records of 294 consecutive patients with HER2+ MBC and diagnosis of CNS disease including parenchymal BrM, leptomeningeal disease (LMD) or dural metastases (DM), treated at Memorial Sloan Kettering Cancer Center between August 2010 and April 2022. HER2-positivity was assessed at the first diagnosis of metastatic disease (any site). Clinicopathologic characteristics including disease burden at presentation, timing of CNS disease and extracranial metastases (ECM), neurologic complications, and local and systemic treatments were collected. CNS-related death was defined as any death caused by BrM or LMD or DM progression or CNS treatment-associated complications. CNS-related death was estimated using cumulative incidence in the competing risks setting (with death due to other causes as a competing event) and overall survival (OS) was estimated using Kaplan Meier methodology. Risk factors associated with CNS-related death were assessed using sub-distribution hazards regression modeling. Treatments given after CNS disease diagnosis were treated as time-varying variables. Tests were two-sided with statistical significance < 0.05. Results: After excluding 19 patients (11 for discordant HER2 status, 5 for concomitant second solid tumor, and 3 for missing data), 275 patients were included in these analyses (258 patients with parenchymal BrM +/- LMD, 8 patients with DM only, 2 patients with DM concomitantly with LMD, and 6 patients with LMD only). Overall, 63/275 (23%) presented with CNS as first and only site of metastasis, 210 (76%) patients developed CNS disease synchronously with or following ECM, and 2 had unknown timing. 125/275 (45%) had de novo MBC, and nearly all patients were treated with CNS local therapies (92% ≥1 radiation treatment and 28% underwent ≥1 BrM resection). The median number of lines of systemic therapy after CNS disease diagnosis was 1 (range: 0-14); 105/275 (38%) patients received a HER2 tyrosine kinase inhibitor (lapatinib: 21%, neratinib: 4.4%, and tucatinib: 12%). The median follow-up was 3.6 years for survivors (range: 0.22 years-12 years). 193/275 (70%) patients died, of whom 105 (54%) died of CNS-related cause. The 3-year OS rate was 40% (95% CI=34%-46%), which varied for patients with BrM only at the diagnosis of metastatic disease [56% (95% CI=43%-69%] compared to patients with ECM +/- BrM [35% (95% CI=28%-42%] (p=0.05). The cumulative incidence of CNS-related death at 3 years was 33% (95% CI=27%-39%). Upon univariable modeling, LMD and CNS radiation treatment (RT) were associated with CNS-related death, and these associations remained statistically significant in a multivariable model [LMD: HR=2.49 (95% CI=1.62-3.83), p< 0.0001, and RT: HR=2.91 (95% CI=1.27-6.64), p=0.01]. Conclusions: Greater than half of patients with HER2+ with CNS involvement suffered of CNS-related death, with greatest risk among patients with LMD. We hypothesize that the association between radiotherapy and CNS-related death may reflect upfront response to systemic cancer-directed treatments of those in patients with DM only or limited CNS disease extension. CNS-only relapse at metastatic presentation portended improved survival than intra- plus extracranial progression, supporting an approach of aggressive local therapy for selected patients. Multimodality approaches using HER2-directed agents effective against parenchymal BrM, LMD along with extracranial metastases, in combination with local CNS-directed therapies need to be optimized. Citation Format: Emanuela Ferraro, Rabih Bou Nassif, Anne Reiner, Samantha Brown, Umberto Tosi, Katherine Panageas, Chau Dang, Andrew D Seidman, Nelson Moss. Clinical risk factors of Central Nervous System (CNS)-related death in patients with HER2-positive metastatic breast cancer [abstract]. In: Proceedings of the 2023 San Antonio Breast Cancer Symposium; 2023 Dec 5-9; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2024;84(9 Suppl):Abstract nr PS11-02.

Abstract PO2-20-10: Exceptional long-term disease control on pembrolizumab monotherapy in a patient with TMB-high PMS2-mutated MMR-deficient triple negative breast cancer brain metastasis

Abstract Introduction The brain is among the most common sites of breast cancer (BC) spread. Prognosis of BC brain metastasis (BCBM) differs by receptor subtype, with poorest median overall survival (OS) for triple-negative (TN) BCBM at under 5 months. Though treatment guidelines have evolved in recent years, most strategies have focused on local therapies including surgery and stereotactic radiotherapy (SRT), rather than on systemic agents. The 2020 FDA approval of pembrolizumab (pembro) for patients (pts) with metastatic high-tumor mutational burden (TMB-h, ≥ 10 mutations/megabase) solid tumors, agnostic of primary site, combined with clinical trials in melanoma reporting efficacy of immune checkpoint inhibitors (ICI) against BM, suggest a potential use of ICI for pts with TMB-h BCBM. Results of the TAPUR study supported modest therapeutic benefit of pembro for 28 pts with metastatic BC selected for TMB-h, with median OS of 30.6 weeks, but did not include pts with active BCBM. TMB-h is uncommon in BC, and further genomic signatures remain to be elucidated to identify pts with BCBM responsive to ICI. We present a remarkable case of a pt with Lynch Syndrome (LS) due to a germline PMS2 mutation, who developed TMB-h, mismatch repair (MMR)-deficient TN BCBM, and has maintained stable disease for over 6 years on pembro monotherapy. Case description A 62-y.o. woman underwent screening mammography that identified a right breast mass. Pt has a family history of a mother with uterine and pancreatic cancer, and father with bladder cancer. Biopsy confirmed invasive ductal carcinoma (IDC), grade 1 (G1), ER/PR-positive, HER2 score 0. She underwent partial mastectomy, axillary dissection, radiation, and received adjuvant anastrozole. Three years later, while still on anastrozole, she presented with a 6 cm right breast mass. Mastectomy disclosed G3, ER/PR-negative, HER2-positive (3+, Her2:CEP17 ratio 5.6) IDC. She received adjuvant chemotherapy (AC-T) and 1 year of trastuzumab. Seven years later, pt presented with headaches and word finding difficulties. MRI identified a 3.6 cm frontal mass, which biopsy revealed as TN BCBM. Systemic imaging ruled out distant metastases. Pt proceeded with SRT and exemestane (despite receptor staining ER- in case of a possible false negative result). Despite radiotherapy, pt had BCBM progression within the same year with the largest mass diameter grown to 4.8 cm. Resection (followed by SRT) identified the dominant histology as G3 TNBC. Following transition of care to our institution, germline testing revealed a pathogenic mutation in PMS2 (c943C >T). Tumor sequencing of the resected brain metastasis showed TMB of 48 and MMR deficiency. Therefore, systemic therapy with pembro monotherapy was initiated and maintained with good tolerability. Two years later, concern for progression from rising tumor markers led to addition of concurrent carboplatin/gemcitabine for 5 cycles. Imaging and tumor markers stabilized, and pt returned to pembro monotherapy. Today, 3 years later, and over 6.5 years after initial TN BCBM diagnosis, pt continues to have stable disease on pembro with an excellent performance status (ECOG 0) and no evidence of intracranial or extracranial progression. Discussion We document an exceptional case of prolonged benefit to pembro in a pt with TMB-h PMS2-mutated MMR-deficient TN BCBM. This case highlights two major pragmatic points to clinical practice; (i) the significance of germline testing in all pts with metastatic BC, and (ii) the impact of molecular testing of tumor tissue to identify rare alterations which may have dramatic effects on pt outcomes. Altogether, this case is a remarkable addition to growing evidence that TMB could be a biomarker for predicting response to ICI therapy and provides an example of ICI efficacy in the setting of BCBM. Further research on additional genomic, clinical, and pathologic features associated with ICI responsiveness is warranted for greater precision in BCBM management. Citation Format: Jaeyoon Cha, Albert Grinshpun, Lauren Buckley, Brian Alexander, Sara Tolaney, Nancy Lin. Exceptional long-term disease control on pembrolizumab monotherapy in a patient with TMB-high PMS2-mutated MMR-deficient triple negative breast cancer brain metastasis [abstract]. In: Proceedings of the 2023 San Antonio Breast Cancer Symposium; 2023 Dec 5-9; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2024;84(9 Suppl):Abstract nr PO2-20-10.

Abstract PO5-16-01: Clinicogenomic analysis reveals genomic associations of brain metastatic tropism in breast cancer

Abstract Introduction: Recent progress in imaging and therapeutics has led to improvements in the treatment of advanced breast cancer. However, the incidence of brain metastasis (BM) in breast cancer is increasing and continues to be associated with poor prognosis and significant morbidity. Herein, we describe genomic enrichments associated with brain tropism from a large institutional clinico-genomic cohort. Methods: Patients who underwent tumor and matched normal sequencing utilizing hybridization-capture based targeted exonic sequencing (MSK-IMPACT) from April 2014 to March 2023 were included in this analysis; this cohort included 6,377 samples from 5,135 patients. We analyzed genomic data to inform the full spectrum of somatic and germline mutations, copy number changes and structural variants enriched in samples in resected BM specimens (n = 179), compared to primary tumors (n = 2980), stratified by receptor status. Enrichment analysis was conducted using a Fisher exact test for each respective receptor status. Similar analyses were conducted between extracranial metastatic (ECM) sites (n = 3186) and primary tumors, to differentiate the unique genomic features driving the emergence of BM as compared to general metastatic propensity. Multiple hypothesis testing was performed using Benjamini-Hochberg. Further genomic analysis, including allele specific copy number changes and mutational signatures will be reported at the 2023 SABCS meeting. Results: Oncogenic alterations in several genes were found to be enriched in BM samples compared to primary tumor, with an odds ratio exceeding that of the corresponding ECM comparison. For example, in the HR+/HER2- subset (4329 samples from 3476 patients), pathogenic variants in the following genes were enriched in BM:Primary analysis: NF1 (OR 6.4 [2.9 – 13.7], q = 0.001, 15.5% of BM), MYC (OR 3.9 [2.0 – 7.4], q = 0.001, 24.4% of BM), AGO2 (5.2 [2.4 – 10.9], q= 0.002, 15.5% of BM), CDKN2A (OR 3.9 [1.6 – 9.4], q = 0.06, 10.3% of BM), ERBB2 (OR 3.6 [1.5 – 8.6], q = 0.06, 10.3% of BM), RB1 (OR 4.2 [1.5 – 11.4], q = 0.06, 8.6% of BM), TP53 (OR 2.0 [1.1 – 3.5], q = 0.07, 40% of BM). In contrast, ESR1 variants were enriched in BM:Primary (OR 14.0 [6.7 – 28.2], but at the same level as ECM:Primary (OR 12.3 [8.7 – 17.7]). When considering HER2+ tumors (919 samples from 739 patients), only NKX.3 was enriched in the BM:Primary comparison (OR 11.5 [3.2 – 42.3], q = 0.004, 11.6% of BM). Although no TNBC samples met significance thresholds after adjustment for multiple hypothesis testing, CDH1, NTRK1, EGFR, and TP53 emerged as putatively significant. Conclusions: In a large cohort of genomically-profiled breast cancer samples, we find recurrent involvement of genes involved in cell cycle regulation (RB1, CDKN2A, TP53); these have been found to be enriched in brain metastases from other cancer types and are implicated in resistance to common lineage-directed therapies. Our approach also uncovered several targets for which therapies are actively in development (ERBB2, MYC). Further study of genomic evolution of breast cancer in the context of brain metastasis and therapeutic resistance will uncover conserved site-specific genomic changes and will identify novel, rational therapeutic targets. Citation Format: Anton Safonov, Deborah Smith, Emanuela Ferraro, Junchao Shen, Ishaani Khatri, Rahul Kumar, Julia An, Mark Robson, Sarat Chandarlapaty, Adrienne Boire, Nikolaus Schultz, Nelson Moss, Luke Pike, Pedram Razavi. Clinicogenomic analysis reveals genomic associations of brain metastatic tropism in breast cancer [abstract]. In: Proceedings of the 2023 San Antonio Breast Cancer Symposium; 2023 Dec 5-9; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2024;84(9 Suppl):Abstract nr PO5-16-01.

Fractionated stereotactic radiotherapy of intracranial postoperative cavities after resection of brain metastases – Clinical outcome and prognostic factors

Background and PurposeAfter surgical resection of brain metastases (BM), radiotherapy (RT) is indicated. Postoperative stereotactic radiosurgery (SRS) reduces the risk of local progression and neurocognitive decline compared to whole brain radiotherapy (WBRT). Aside from the optimal dose and fractionation, little is known about the combination of systemic therapy and postoperative fractionated stereotactic radiotherapy (fSRT), especially regarding tumour control and toxicity. MethodsIn this study, 105 patients receiving postoperative fSRT with 35 Gy in 7 fractions performed with Cyberknife were retrospectively reviewed. Overall survival (OS), local control (LC) and total intracranial brain control (TIBC) were analysed via Kaplan-Meier method. Cox proportional hazards models were used to identify prognostic factors. ResultsMedian follow-up was 20.8 months. One-year TIBC was 61.6% and one-year LC was 98.6%. Median OS was 28.7 (95%-CI: 16.9–40.5) months. In total, local progression (median time not reached) occurred in 2.0% and in 20.4% radiation-induced contrast enhancements (RICE) of the cavity (after median of 14.3 months) were diagnosed. Absence of extracranial metastases was identified as an independent prognostic factor for superior OS (p = <0.001) in multivariate analyses, while a higher Karnofsky performance score (KPS) was predictive for longer OS in univariate analysis (p = 0.041). Leptomeningeal disease (LMD) developed in 13% of patients. ConclusionFSRT after surgical resection of BM is an effective and safe treatment approach with excellent local control and acceptable toxicity. Further prospective randomized trials are needed to establish standardized therapeutic guidelines.

377 CDKN2A Alterations Are Associated With Increased Postoperative Seizure Risk After Resection of Brain Metastases

INTRODUCTION: Seizures are common and significantly disabling for patients with brain metastases (BMs). While surgical resection can provide seizure control, a subset of patients with BMs may continue to suffer seizures postoperatively. Molecular BM characteristics may influence which patients are at-risk for postoperative seizures. METHODS: All patients underwent BM resection at one institution with available clinical and sequencing data on &gt; 500 oncogenes. Clinical seizures were documented preoperatively and postoperatively. Random forest machine learning classification was used to determine candidate genomic alterations associated with postoperative seizures, and clinical and top genomic variables were correlated with postoperative seizures using Cox proportional hazards models. RESULTS: There were 112 patients with BMs who underwent 114 surgeries and had at least one month of postoperative follow-up. Seizures occurred preoperatively in 26 (22.8%) patients and postoperatively in 25 (21.9%). Engel classification (EC) at six months achieved for those with preoperative seizures was EC I in 13 (50%); II in 6 (23.1%); III in 5 (19.2%), and IV in 2 (7.7%). In those with postoperative seizures, only 8 (32.0%) had seizures preoperatively. Preoperative seizures were not a significant predictor of postoperative seizures (hazards ratio (HR) = 1.84; 95% confidence interval (CI) = 0.79 – 4.37; p = 0.156). On random forest classification and multivariate Cox analysis controlling for factors including recurrence, extent of resection, and number of BMs, CDKN2A alterations were associated with postoperative seizures (HR = 3.22; 95% CI = 1.27–8.16; p = 0.014). Melanoma BMs were associated with higher risk of postoperative seizures compared with all other malignancies (HR = 5.23; 95% CI = 1.37–19.98; p = 0.016). CONCLUSIONS: CDKN2A alterations and melanoma primary malignancy are associated with increased postoperative seizure risk following resection of BMs. These results may help guide postoperative seizure prophylaxis in patients undergoing resection of BMs.

385 Genomic Alterations Associated With Postoperative Nodular Leptomeningeal Disease After Resection of Brain Metastases

INTRODUCTION: The relationship between brain metastasis resection and risk of nodular leptomeningeal disease (nLMD) is unclear. Prior studies have not explored genomic alterations associated with postoperative nLMD. METHODS: A retrospective, single-center study was conducted with patients who underwent resection of a BM with available clinical and genomic data available. Postoperative nLMD was the primary endpoint of interest. Targeted next-generation sequencing of over 500 oncogenes using a CLIA certified assay was performed in brain metastases. Cox proportional hazards analyses were performed to identify clinical features and genomic alterations associated with nLMD. RESULTS: One hundred and one patients composed the cohort with tumors originating from multiple cancer types. There were 15 cases of nLMD (14.9% of the cohort) with a median time from surgery to nLMD diagnosis of 8.2 months. Two supervised machine learning algorithms consistently identified CDKN2A/B co-deletion and ERBB2 amplification as the top predictors associated with postoperative nLMD across all cancer types. In a multivariate Cox proportional hazard analysis including clinical factors and genomic alterations observed in the cohort, tumor volume (HR 8.8, 95% CI 1.1-69, p = 0.04), CDKN2A/B co-deletion (HR 5.3, 95% CI 1.7-16.9, p = 0.004), and ERBB2 amplification (HR 3.9, 95% CI 1.1-14.4, p = 0.04) were associated with decreased time to postoperative nLMD. CONCLUSIONS: In addition to increased resected tumor volume, ERBB2 amplification and CDKN2A/B deletion were independently associated with an increased risk of postoperative nLMD across multiple cancer types. Additional work is needed to determine if targeted therapy decreases this risk in the postoperative setting.

Human Epidermal Growth Factor Receptor 2-Low Breast Cancer Brain Metastases: An Opportunity for Targeted Systemic Therapies in a High-Need Patient Population.

Trastuzumab deruxtecan is a new treatment option for patients with advanced human epidermal growth factor receptor 2 (HER2)-low breast cancer (BC). Although HER2-low status has been characterized in early and advanced BC, it has yet to be fully characterized in brain metastases (BrM). Patients who underwent surgery for BC BrM at Sunnybrook Health Sciences Centre and for whom HER2 status was available on resected BrM were studied. Estrogen receptor, progesterone receptor, and HER2 status were assessed on the basis of ASCO/College of American Pathologists (CAP) guidelines. HER2-zero was defined as immunohistochemistry (IHC) 0; HER2-low was defined as IHC 1+ or IHC 2+ with fluorescence in situ hybridization (FISH)-negative status. HER2-positive (HER2+) was defined as IHC 3+ or IHC 2+ with positive FISH. Clinicopathologic features were recorded. We also assessed the prognostic association between extent of HER2 expression and (1) brain-specific progression-free survival (bsPFS), as well as (2) overall survival (OS). In this retrospective cohort of 102 patients with resected BC BrM, 53% (n = 54) were HER2+, 29.4% (n = 30) were HER2-low, and 17.6% (n = 18) had HER2-zero status. Among BrM that were triple-negative on the basis of ASCO/CAP guidelines, 63.6% (n = 14/22) were reclassified as being HER2-low. Sixty percent (n = 15/25) of BrM that were hormone receptor-positive/HER2-negative (HR+/HER2-) were reclassified as being HER2-low. In total, 51 patients had matched primary breast and BrM tissue available; results of HER2 status when categorized as HER2-zero, HER2-low, and HER2+ were concordant in 82.3% (n = 42/51) of cases (Cohen's kappa, 0.58; P = .07). There was no significant association between HER2-zero, HER2-low, and HER2+ status in BrM and either bsPFS or OS. Among patients with surgically resected BrM, a high proportion of those with metastatic triple-negative BC and HR+/HER2- disease have HER2-low BrM with potential to benefit from HER2-targeted therapy.

Predictive role of intracranial PD-L1 expression in a real-world cohort of NSCLC patients treated with immune checkpoint inhibition following brain metastasis resection

BackgroundEmerging evidence suggests that treatment of NSCLC brain metastases with immune checkpoint inhibitors (ICIs) is associated with response rates similar to those of extracranial disease. Programmed death-ligand 1 (PD-L1) tumor proportion score (TPS) serves as a predictive biomarker for ICI response. However, the predictive value of brain metastasis-specific (intracranial) PD-L1 TPS is not established. We investigated the role of intra- and extracranial PD-L1 TPS in NSCLC patients treated with ICI following brain metastasis resection.MethodsClinical data from NSCLC patients treated with ICI following brain metastasis resection (n = 64) were analyzed. PD-L1 TPS of brain metastases (n = 64) and available matched extracranial tumor tissue (n = 44) were assessed via immunohistochemistry. Statistical analyses included cut point estimation via maximally selected rank statistics, Kaplan–Meier estimates, and multivariable Cox regression analysis for intracranial progression-free survival (icPFS), extracranial progression-free survival (ecPFS), and overall survival (OS).ResultsPD-L1 expression was found in 54.7% of brain metastases and 68.2% of extracranial tumor tissues, with a median intra- and extracranial PD-L1 TPS of 7.5% (0 – 50%, IQR) and 15.0% (0 – 80%, IQR), respectively. In matched tissue samples, extracranial PD-L1 TPS was significantly higher than intracranial PD-L1 TPS (p = 0.013). Optimal cut points for intracranial and extracranial PD-L1 TPS varied according to outcome parameter assessed. Notably, patients with a high intracranial PD-L1 TPS (> 40%) exhibited significantly longer icPFS as compared to patients with a low intracranial PD-L1 TPS (≤ 40%). The cut point of 40% for intracranial PD-L1 TPS was independently associated with OS, icPFS and ecPFS in multivariable analyses.ConclusionOur study highlights the potential role of intracranial PD-L1 TPS in NSCLC, which could be used to predict ICI response in cases where extracranial tissue is not available for PD-L1 assessment as well as to specifically predict intracranial response.

HER2 expression and genOmic characterization of rESected brain metastases from colorectal cancer: the HEROES study

BackgroundLittle is known about prognostic factors of brain metastases (BM) from colorectal cancer (CRC). HER2 amplification/overexpression (HER2+) was previously described; its impact on prognosis remains uncertain.MethodsIn the translational study HEROES, extensive molecular analysis was performed on primary CRC (prCRC) and their matched resected BM by means of NGS comprehensive genomic profiling and HER2 status as assessed by immunohistochemical/ in situ hybridization. Count of tumour-infiltrating lymphocytes (TILs) was also performed. Primary objective: to describe the molecular landscape of paired BM/prCRC. Secondary objectives: to search for new prognostic biomarkers of outcome after BM resection: intracranial-only Progression-Free Survival (BM-iPFS), Progression-Free Survival (BM-PFS), and Overall Survival (BM-OS).ResultsOut of 22 patients having paired samples of prCRC and BM, HER2+ was found on 4 (18%) BM, 3 (75%) of which also HER2+ in matched prCRC. Lower tumour mutation burden (HR 3.08; 95%CI 1.06–8.93; p = 0.0386) and HER2-negative BM (HER2neg) (HR 7.75;95%CI 1.97–30.40; p = 0.0033) were associated with longer BM-iPFS; HER2neg BM (HR 3.44; 95%CI 1.03–11.53; p = 0.0449) and KRASmut BM (HR 0.31; 95%CI 0.12–0.80; p = 0.0153) conferred longer BM-PFS. Longer BM-OS was found in pts with TILs-enriched (≥1.6/HPF) BM (HR 0.11; 95%CI0.01–0.91; p = 0.0403).ConclusionsThis study shows HER2+ enrichment in both BM and their prCRC. TILs-enriched BM conferred better BM-OS.

Genomic alterations associated with postoperative nodular leptomeningeal disease after resection of brain metastases.

The relationship between brain metastasis resection and risk of nodular leptomeningeal disease (nLMD) is unclear. This study examined genomic alterations found in brain metastases with the aim of identifying alterations associated with postoperative nLMD in the context of clinical and treatment factors. A retrospective, single-center study was conducted on patients who underwent resection of brain metastases between 2014 and 2022 and had clinical and genomic data available. Postoperative nLMD was the primary endpoint of interest. Targeted next-generation sequencing of > 500 oncogenes was performed in brain metastases. Cox proportional hazards analyses were performed to identify clinical features and genomic alterations associated with nLMD. The cohort comprised 101 patients with tumors originating from multiple cancer types. There were 15 patients with nLMD (14.9% of the cohort) with a median time from surgery to nLMD diagnosis of 8.2 months. Two supervised machine learning algorithms consistently identified CDKN2A/B codeletion and ERBB2 amplification as the top predictors associated with postoperative nLMD across all cancer types. In a multivariate Cox proportional hazards analysis including clinical factors and genomic alterations observed in the cohort, tumor volume (× 10 cm3; HR 1.2, 95% CI 1.01-1.5; p = 0.04), CDKN2A/B codeletion (HR 5.3, 95% CI 1.7-16.9; p = 0.004), and ERBB2 amplification (HR 3.9, 95% CI 1.1-14.4; p = 0.04) were associated with a decreased time to postoperative nLMD. In addition to increased resected tumor volume, ERBB2 amplification and CDKN2A/B deletion were independently associated with an increased risk of postoperative nLMD across multiple cancer types. Additional work is needed to determine if targeted therapy decreases this risk in the postoperative setting.

Multicentric Assessment of Safety and Efficacy of Combinatorial Adjuvant Brain Metastasis Treatment by Intraoperative Radiation Therapy and Immunotherapy

PurposeFollowing surgical resection of brain metastases (BMs), intraoperative radiation therapy (IORT) provides a promising alternative to adjuvant external beam radiation therapy (EBRT) by enabling superior organ at risk preservation, reduction of in-hospital times and timely admission to subsequent systemic treatments, which increasingly comprise novel targeted immunotherapeutic approaches. We sought to assess safety and efficacy of IORT in combination with immune checkpoint inhibitors (ICIs) and other targeted therapies (TTs). MethodsIn a multicentric approach incorporating individual patient data from six international IORT centers, all patients with BMs undergoing IORT were retrospectively assessed for combinatorial treatment with ICIs/TTs and evaluated for toxicity and cumulative rates, including wound dehiscence, radiation necrosis (RN), leptomeningeal spread (LMS), local control (LC), distant brain progression (DBP) and estimated overall survival (OS). ResultsA total of 103 lesions with a median diameter of 34 mm receiving IORT combined with immunomodulatory systemic treatment or other TTs were included. The median follow up was 13.2 (1.2-102.4) months and the median IORT dose was 25 (18-30) Gy prescribed to the applicator surface. There was one grade 3 adverse event related to IORT recorded (2.2%). A 4.9% cumulative RN rate was observed. The 1-year LCR was 98.0% and the 1-year DBP-free rate 60.0%. Median time to DBP was 5.5 (1.0-18.5) months in the subgroup of patients experiencing DBP and the cumulative LMS rate was 4.9%. The median estimated OS was 26 (1.2-not reached) months with a 1-year survival rate of 74.0%. Early initiation of IT/TT was associated with a non-significant trend towards improved DBP rate and OS. ConclusionThe combination of ICIs/TT with IORT for resected BMs does not seem to increase toxicity, while yielding encouraging local control outcomes in the difficult-to-treat subgroup of larger BMs. Time gaps between surgery and systemic treatment could be shortened or avoided. The definitive role of IORT in local control after BM resection will be defined in a prospective trial.

Prognostic impact of the number and total tumor burden of secondary cerebral lesions in patients with resected brain metastases of non-small cell lung cancers.

Systemic therapeutic advancements have improved the prognosis of cancer patients, leading to surgery more frequently being carried out for patients with multiple brain metastases (BM). The underlying evidence for the strategy is currently lacking. This study aimed to evaluate the prognostic significance of the number of BM and total tumor burden (TTB) on the overall survival (OS) of patients with resected BM of non-small cell lung cancer (NSCLC) in a modern series. In this monocentric retrospective series, patients who underwent resection of BM of NSCLC between 2015 and 2021 were included. Demographic, clinical, and histological parameters were collected, and formal radiological volumetric analyses were performed. Prognostic biomarkers for cerebral progression-free survival (C-PFS) and OS were analyzed with univariate and multivariate Cox proportional hazards analysis. One hundred eighty-four patients were included in the study. Among these, 108 patients (58.7%) presented with a single brain metastasis, 36 patients (19.6%) with 2 BM, 22 patients (11.9%) with 3 BM, and 18 patients (9.8%) with more than 3 BM (maximum 15 BM). The mean ± SD (range) preoperative tumor burden was 23.1 ± 25.3 (1.1-145.3) cm3. The mean residual tumor burden after surgery was 0.3 ± 0.8 (0.0-6.3) cm3. By the time of the analysis, 128 patients (69.6%) had died. The median follow-up duration was 49.0 months (95% CI 39.6-63.6). The median OS was 19.2 months (95% CI 13.2-24.0), and the survival rates at 6 months, 1 year, and 2 years were 76% (95% CI 69%-82%), 61% (95% CI 53%-67%), and 43% (95% CI 35%-50%), respectively. The median C-PFS was 8.4 months (95% CI 7.2-12.0). In the Cox multivariate regression model, younger age (< 65 years), single brain metastasis, adjuvant brain radiation therapy, adjuvant use of targeted therapy, and TTB < 7 cm3 were all independent predictors of longer OS. In this era of modern systemic treatments for cancer, the number of BM and total cerebral tumor burden remain significant prognostic factors of OS. However, resection should be considered as an option even in those patients with multiple BM in order to enhance patient clinical status, enable further local and systemic treatment delivery, and improve their survival and quality of life.

Clinical Application of the Association between Genetic Alteration and Intraoperative Fluorescence Activity of 5-Aminolevulinic Acid during the Resection of Brain Metastasis of Lung Adenocarcinoma.

The primary objective of this study was to investigate the association of certain genetic alterations and intraoperative fluorescent activity of 5-aminolevulinic acid (ALA) in brain metastasis (BM) of lung adenocarcinoma. A retrospective cohort study was conducted among 72 patients who underwent surgical resection of BM of lung adenocarcinoma at our institute for five years. Cancer cell infiltration was estimated by the intraoperative fluorescent activity of 5-ALA, and genetic alterations were analyzed by next-generation sequencing (NGS). The sensitivity and specificity for detecting cancer cell infiltration using 5-ALA were 87.5% and 96.4%, respectively. Genes associated with cell cycle regulation (p = 0.003) and cell proliferation (p = 0.044) were significantly associated with positive fluorescence activity of 5-ALA in the adjacent brain tissue. Genetic alterations in cell cycle regulation and cell proliferation were also associated with shorter recurrence-free survival (p = 0.013 and p = 0.042, respectively) and overall survival (p = 0.026 and p = 0.042, respectively) in the multivariate analysis. The results suggest that genetic alterations in cell cycle regulation and cell proliferation are associated with positive fluorescence activity of 5-ALA in the adjacent infiltrative brain tissue and influence the clinical outcome of BM of lung adenocarcinoma.

RADT-09. CLINICAL OUTCOMES OF FRACTIONATED STEREOTACTIC RADIOTHERAPY FOR INTACT AND POSTOPERATIVE RADIORESISTANT RENAL CELL CARCINOMA AND MELANOMA BRAIN METASTASES

Abstract PURPOSE Fractionated Stereotactic Radiotherapy (FSRT) may be an optimal strategy for radioresistant renal cell carcinoma (RCC) and melanoma brain metastases (BM). We reviewed our institutional experience of FSRT to intact BM and postoperative cavities for patients with melanoma and RCC. METHODS We retrospectively identified consecutive patients with melanoma and renal cell carcinoma with unresected or resected BM treated with FSRT at our institution from 1/2018 to 1/2021. Time to event outcomes were estimated using the Kaplan-Meier method with log-rank tests and multivariable Cox regression analyses. RESULTS We identified 44 patients with a total of 64 BM: 31 intact (48.4%) and 33 resected BMs (51.6%). Thirty-six (81.8%) were melanoma and eight (18.2%) were RCC. The majority of intact BM (n = 22/31; 71%) and cavities (n = 21/33; 64%) were treated to 30 Gy (24-30) in 5 fractions and 25 Gy (24-35), respectively. Median PTV volume for intact BM was 5.04 cc (1.2-19.8) and for resected BM was 16.7 cc (4.8-89.1). Median interval from planning MRI to FSRT was 10 days (3-28). Median time from surgery to FSRT was 37 days (9-90). At 18.6 months median follow-up, local and intracranial distant failures were observed in 4 and 20 patients, respectively. Our cohort had a local recurrence-free survival (LRFS) of 83% and overall survival (OS) of 57% at 12 months; no difference between primary FSRT versus surgery followed by FSRT (p = 0.93 and p = 0.52, respectively). On MVA, higher FSRT dose was associated with better local control (p = 0.01). Radiation Necrosis (RN) was observed in eight treatment areas (12.5%). Neither resection status (p = 0.99) nor PTV volume (p = 0.61) were associated with RN. CONCLUSION FSRT is associated with low rates of toxicity and clinically significant disease control for both intact and resected melanoma and RCC BM, with higher FSRT dose being correlated with improved control.

RADT-05. SINGLE VERSUS MULTI-FRACTION STEREOTACTIC RADIOSURGERY FOR LARGE BRAIN METASTASES: RATES OF LOCAL PROGRESSION AND NECROSIS

Abstract PURPOSE For large brain metastases, it is unknown whether single fraction stereotactic radiosurgery (sSRS) or multi-fraction stereotactic radiosurgery (fSRS) is optimal. We aimed to compare rates of local progression and necrosis between sSRS and fSRS for large brain metastases. METHODS Patients with intact or resected brain metastases with a target volume ≥ 4.2 cc (corresponding to a diameter ≥ 2 cm) treated with Gamma Knife (GK) or linear accelerator-based SRS from 2014-2022 were included in this retrospective review. RESULTS A total of 343 large brain metastases in 276 patients were included; 171 lesions were treated with sSRS (96% with GK) and 172 lesions with fSRS (100% with linear accelerator). Median lesion volume was 19.8 cc and 8.4 cc for fSRS and sSRS, respectively. Resection was performed prior to SRS in 228 (66%), with the remaining 115 metastases treated intact. There was no difference in rate of local progression between sSRS versus fSRS, or between resected versus intact targets (p &amp;gt;0.8 for both). Cumulative incidence of local progression (with death as a competing risk) at 1 and 3 years for patients receiving sSRS was 24% and 30%, respectively, compared to 22% and 34% for fSRS. FSRS was associated with significantly higher rates of necrosis than sSRS (HR 5.58, 2.38-13.08, p &amp;lt; 0.001). This association remained significant after controlling for target volume. Cumulative incidence of necrosis (with death as a competing risk) at 1 and 3 years for patients receiving sSRS was 1% and 3%, respectively, compared to 10% and 18% for fSRS. Necrosis rates were significantly higher with fSRS for both intact and resected targets (P &amp;lt; 0.02 for both). CONCLUSION While local control for large resected or intact brain metastases appears similar after sSRS or fSRS, this retrospective study shows higher rates of necrosis after fSRS. Prospective validation is needed.

RADT-13. A BIOLOGIC EQUIVALENT DOSE COMPARISON BETWEEN RADIOSURGERY AND COLLAGEN TILE BRACHYTHERAPY FOR POST-OPERATIVE TREATMENT OF A BRAIN METASTASIS RESECTION CAVITY

Abstract INTRODUCTION Recurrence of a previously irradiated brain metastasis represents a challenging clinical scenario. Repeat radiosurgery results in sub-optimal local control (LC). Two independently published prospective cohort studies demonstrated durable LC in this setting with collagen tile brachytherapy (CTBT), a novel Cesium-131 carrier. Here we compare biologically equivalent dosimetric parameters of CTBT to radiosurgery to understand dose metrics that may underlie durable LC from CTBT. METHODS A patient was selected from the multi-institutional prospective CTBT registry (NCT04427384) with a brain metastasis resection cavity near brainstem. Planning was performed for radiosurgery (30Gy in 5 fractions, BED of 48Gy) and CTBT (Cs-131, activity of 3.5U). To directly compare dosimetric parameters between radiosurgery and CTBT, biologic equivalent doses (BED) were calculated for brainstem(α/β of 3, BED3) and oncologically at-risk tissue(α/β of 10, BED10). TG101 brainstem dose-constraints were referenced, and respective BED were calculated. RESULTS The resection cavity was 4mm away from brainstem and contained 7 CTBT tiles (28 Cs131 seeds). Gradient index (GI) (24Gy BED10 volume/48Gy BED10 volume) was 3.8 for radiosurgery and 2.2 for CTBT. Conformality index (CI) (48Gy BED10 volume/CTV volume) was 1.2 for radiosurgery and 2.6 for CTBT. BED10 to 90% of the cavity wall (2mm rim around the resection cavity) was 49.6Gy for radiosurgery and 92.6Gy for CTBT. BED3 dose limit goals (max dose 31Gy in 5 fractions, BED3 95.1Gy and threshold dose 23Gy in 5 fractions, BED3 of 58.3Gy to 0.5cc) were met for both radiosurgery and CTBT. CONCLUSION BED comparisons demonstrate that CTBT leads to higher CI, lower GI, and higher dose to the cavity wall than radiosurgery while meeting BED3 dose limit goals to an adjacent organ at risk. These dose characteristics might underlie durable LC seen in the recurrent brain metastasis setting, though larger scale analyses are being performed with with clinical outcomes on NCT04427384.

RADT-04. DOSE PAINTING WITH THE GAMMA KNIFE LIGHTNING OPTIMIZER TO ESCALATE RADIATION DOSE TO RESIDUAL TUMOR WITHIN BRAIN METASTASIS RESECTION CAVITIES

Abstract BACKGROUND The novel Gamma Knife (GK) Lightning optimizer allows simultaneous automated optimization of multiple intracranial radiotherapy targets. We used this optimizer to create a dose-painting treatment plan where an escalated dose was prescribed to residual disease within a brain metastasis resection cavity. We validated this method by measuring dose-delivery accuracy using radiochromic film in an anthropomorphic phantom. METHODS A 15.7cc irregular contour was drawn to represent a brain metastasis resection cavity, a 2mm expansion contour created, and a 1.6cc contour drawn representing a nodule of residual disease. In GammaPlan v11.3.1, three targets were created with overlapping dose matrices. Targets were prescribed 3Gy (2mm expansion), 4Gy (resection cavity), and 5Gy (residual disease) in one fraction. On the GK Icon system, an occipital mold was made, an unexposed film placed in the phantom’s cranium, cone-beam coregistration conducted, and treatment delivered. The film was scanned and calibrated for absolute dosimetry. To evaluate dose-painting accuracy, global gamma index (GI) analyses were performed at various dose and distance tolerances, excluding points below 30% of the maximum dose. RESULTS An 18-minute treatment plan with 40 shots was created and delivered. Prescription isodose lines were 3Gy at 55% (2mm expansion, mean 4.8Gy), 4Gy at 69% (resection cavity, mean 5.2Gy), and 5Gy at 75% (residual disease, mean 6.1Gy). All target volumes had ≥ 99% prescription dose coverage and the maximum dose was 6.9Gy. Paddick conformality indices were 0.79 (2mm expansion), 0.74 (resection cavity), and 0.15 (residual disease). Gamma index pass rate/mean GI/median GI were 73%/0.68/0.55 at 1%/1mm tolerance, 85%/0.58/0.48 at 2%/1mm tolerance, and 97%/0.34/0.28 at 2%/2mm tolerance. An average of 17,883 points were evaluated per analysis. CONCLUSIONS We successfully performed GK dose painting with the Lightning optimizer, verifying dose delivery within clinically acceptable tolerances. Future work is needed to determine optimal dose levels for use in clinical practice.

EPCO-01. THE GENOMIC LANDSCAPE OF BRAIN METASTASIS

Abstract BACKGROUND The prognosis for patients with brain metastasis (BM) is devastating, while the underlying biology of BM development remains poorly understood. Identification of genomic biomarkers can offer promising opportunities for prediction of BM development and guidance in personalized treatment. In this study we evaluate the genomic alterations present in a large cohort of resected BM. METHODS Upon retrospective review, we identified 868 patients with diagnosis of solid primary cancers who had undergone a standard of care craniotomy at Memorial Sloan Kettering Cancer Center (MSKCC). BM samples were profiled by MSK-IMPACT, a next-generation sequencing (NGS) assay designed to detect a wide range of genetic alterations in 341-505 cancer genes. Genomic alterations were filtered for driver variants using OncoKB. Disease sites included cancers of head and neck, breast, lung, lower and upper gastrointestinal tract, kidney, ovary, uterus, prostate, skin, soft tissues, and others. RESULTS More than half (57%; 493/868) of patients were female and the median age was 63 (range 22-93). Foreseeably, the most common histology was non-small cell lung cancer (NSCLC; 38%), followed by invasive carcinoma of the breast and melanoma (16% and 13%, respectively). The most frequently encountered alterations were inTP53 (60%), CDKN2A (25%), TERT (23%), KRAS (18%), and PTEN (12%) across all BM samples. The median fraction of genome altered was 0.41 [range: 0-0.99] and the median tumor mutational burden was 6.6 muts/Mb [range: 0-395]. Additionally, TP53 was the most frequently altered gene after stratifying by primary histology in most cancer types, except for melanoma, renal, and thyroid, which were enriched for alterations in TERT, CDKN2A, and TERT, respectively. CONCLUSION The landscape of genomic alterations present in BM is distinct and varies by primary histologic diagnosis. Ongoing analyses of matched primary-BM pairs and clinicogenomic correlation will identify factors predisposing patients to BM development and progression.