Association of Machine Learning-Based Assessment of Tumor-Infiltrating Lymphocytes on Standard Histologic Images With Outcomes of Immunotherapy in Patients With NSCLC

Importance

Currently, predictive biomarkers for response to immune checkpoint inhibitor (ICI) therapy in lung cancer are limited. Identifying such biomarkers would be useful to refine patient selection and guide precision therapy.

Objective

To develop a machine-learning (ML)-based tumor-infiltrating lymphocytes (TILs) scoring approach, and to evaluate TIL association with clinical outcomes in patients with advanced non-small cell lung cancer (NSCLC).

Design, Setting, and Participants

This multicenter retrospective discovery-validation cohort study included 685 ICI-treated patients with NSCLC with median follow-up of 38.1 and 43.3 months for the discovery (n = 446) and validation (n = 239) cohorts, respectively. Patients were treated between February 2014 and September 2021. We developed an ML automated method to count tumor, stroma, and TIL cells in whole-slide hematoxylin-eosin-stained images of NSCLC tumors. Tumor mutational burden (TMB) and programmed death ligand-1 (PD-L1) expression were assessed separately, and clinical response to ICI therapy was determined by medical record review. Data analysis was performed from June 2021 to April 2022.

Exposures

All patients received anti-PD-(L)1 monotherapy.

Main Outcomes and Measures

Objective response rate (ORR), progression-free survival (PFS), and overall survival (OS) were determined by blinded medical record review. The area under curve (AUC) of TIL levels, TMB, and PD-L1 in predicting ICI response were calculated using ORR.

Results

Overall, there were 248 (56%) women in the discovery cohort and 97 (41%) in the validation cohort. In a multivariable analysis, high TIL level (≥250 cells/mm2) was independently associated with ICI response in both the discovery (PFS: HR, 0.71; P = .006; OS: HR, 0.74; P = .03) and validation (PFS: HR = 0.80; P = .01; OS: HR = 0.75; P = .001) cohorts. Survival benefit was seen in both first- and subsequent-line ICI treatments in patients with NSCLC. In the discovery cohort, the combined models of TILs/PD-L1 or TMB/PD-L1 had additional specificity in differentiating ICI responders compared with PD-L1 alone. In the PD-L1 negative (<1%) subgroup, TIL levels had superior classification accuracy for ICI response (AUC = 0.77) compared with TMB (AUC = 0.65).

Conclusions and Relevance

In these cohorts, TIL levels were robustly and independently associated with response to ICI treatment. Patient TIL assessment is relatively easily incorporated into the workflow of pathology laboratories at minimal additional cost, and may enhance precision therapy.

Incidence of Germline Variants in Familial Bladder Cancer and Among Patients With Cancer Predisposition Syndromes

BACKGROUND

The familial aggregation of bladder cancers has been observed, but the incidence and association of familial bladder cancer with germline pathogenic and likely pathogenic (P/LP) variants is unknown.

PATIENTS AND METHODS

A retrospective analysis was conducted of patients with bladder cancer treated at the Dana-Farber Cancer Institute to identify those with a first-degree relative with bladder cancer. A second cohort of patients referred to DFCI for suspicion of a cancer predisposition syndrome was analyzed for candidate P/LP germline variants. Descriptive statistics were generated.

RESULTS

Among 885 patients with bladder cancer, 38 patients (4.3%) had a family history of bladder cancer in a first-degree relative. No significant association of age of diagnosis was observed between patients with and without a first-degree family history of bladder cancer (P = .3). In the second cohort, 27 of 80 (34%) patients with bladder cancer evaluated for cancer predisposition syndromes harbored a P/LP germline variant. P/LP variants were identified most commonly in the following genes: BRCA1 (n = 5), MSH2 (n = 5), MLH1 (n = 4), ATM (n = 3), and CHEK2 (n = 2). Of the 27 patients with identified germline P/LP variants, 20 (74%) had a family history of a tumor component syndrome in a first- or second-degree relative and 3 were subsequently diagnosed with another genetically-linked associated cancer.

CONCLUSION

Familial bladder cancer defined as bladder cancer in the proband and a first-degree relative, was present in 4.3% of patients with bladder cancer and was not associated with age of diagnosis. Additionally, among patients suspected to have a familial cancer syndrome, one-third harbored a germline P/LP variant. Further study of germline variants in patients with familial bladder cancer including somatic testing for loss of heterozygosity may provide insights regarding disease pathogenesis and inform therapy.

Reciprocal effects of mTOR inhibitors on pro-survival proteins dictate therapeutic responses in tuberous sclerosis complex

mTORC1 is aberrantly activated in cancer and in the genetic tumor syndrome tuberous sclerosis complex (TSC), which is caused by loss-of-function mutations in the TSC complex, a negative regulator of mTORC1. Clinically approved mTORC1 inhibitors, such as rapamycin, elicit a cytostatic effect that fails to eliminate tumors and is rapidly reversible. We sought to determine the effects of mTORC1 on the core regulators of intrinsic apoptosis. In TSC2-deficient cells and tumors, we find that mTORC1 inhibitors shift cellular dependence from MCL-1 to BCL-2 and BCL-XL for survival, thereby altering susceptibility to BH3 mimetics that target specific pro-survival BCL-2 proteins. The BCL-2/BCL-XL inhibitor ABT-263 synergizes with rapamycin to induce apoptosis in TSC-deficient cells and in a mouse tumor model of TSC, resulting in a more complete and durable response. These data expose a therapeutic vulnerability in regulation of the apoptotic machinery downstream of mTORC1 that promotes a cytotoxic response to rapamycin.

Ancestry-driven recalibration of tumor mutational burden and disparate clinical outcomes in response to immune checkpoint inhibitors

The immune checkpoint inhibitor (ICI) pembrolizumab is US FDA approved for treatment of solid tumors with high tumor mutational burden (TMB-high; ≥10 variants/Mb). However, the extent to which TMB-high generalizes as an accurate biomarker in diverse patient populations is largely unknown. Using two clinical cohorts, we investigated the interplay between genetic ancestry, TMB, and tumor-only versus tumor-normal paired sequencing in solid tumors. TMB estimates from tumor-only panels substantially overclassified individuals into the clinically important TMB-high group due to germline contamination, and this bias was particularly pronounced in patients with Asian/African ancestry. Among patients with non-small cell lung cancer treated with ICIs, those misclassified as TMB-high from tumor-only panels did not associate with improved outcomes. TMB-high was significantly associated with improved outcomes only in European ancestries and merits validation in non-European ancestry populations. Ancestry-aware tumor-only TMB calibration and ancestry-diverse biomarker studies are critical to ensure that existing disparities are not exacerbated in precision medicine.

Neoadjuvant atezolizumab for resectable non-small cell lung cancer: an open-label, single-arm phase II trial

In an ongoing, open-label, single-arm phase II study ( NCT02927301 ), 181 patients with untreated, resectable, stage IB-IIIB non-small cell lung cancer received two doses of neoadjuvant atezolizumab monotherapy. The primary end point was major pathological response (MPR; ≤10% viable malignant cells) in resected tumors without EGFR or ALK alterations. Of the 143 patients in the primary end point analysis, the MPR was 20% (95% confidence interval, 14-28%). With a minimum duration of follow-up of 3 years, the 3-year survival rate of 80% was encouraging. The most common adverse events during the neoadjuvant phase were fatigue (39%, 71 of 181) and procedural pain (29%, 53 of 181), along with expected immune-related toxicities; there were no unexpected safety signals. In exploratory analyses, MPR was predicted using the pre-treatment peripheral blood immunophenotype based on 14 immune cell subsets. Immune cell subsets predictive of MPR in the peripheral blood were also identified in the tumor microenvironment and were associated with MPR. This study of neoadjuvant atezolizumab in a large cohort of patients with resectable non-small cell lung cancer was safe and met its primary end point of MPR ≥ 15%. Data from this single-arm, non-randomized trial suggest that profiles of innate immune cells in pre-treatment peripheral blood may predict pathological response after neoadjuvant atezolizumab, but additional studies are needed to determine whether these profiles can inform patient selection and new therapeutic approaches.

Mosaicism in Tumor Suppressor Gene Syndromes: Prevalence, Diagnostic Strategies, and Transmission Risk

A mosaic state arises when pathogenic variants are acquired in certain cell lineages during postzygotic development, and mosaic individuals may present with a generalized or localized phenotype. Here, we review the current state of knowledge regarding mosaicism for eight common tumor suppressor genes-NF1, NF2, TSC1, TSC2, PTEN, VHL, RB1, and TP53-and their related genetic syndromes/entities. We compare and discuss approaches for comprehensive diagnostic genetic testing, the spectrum of variant allele frequency, and disease severity. We also review affected individuals who have no mutation identified after conventional genetic analysis, as well as genotype-phenotype correlations and transmission risk for each tumor suppressor gene in full heterozygous and mosaic patients. This review provides new insight into similarities as well as marked differences regarding the appreciation of mosaicism in these tumor suppressor syndromes.

Midkine expression by stem-like tumor cells drives persistence to mTOR inhibition and an immune-suppressive microenvironment

mTORC1 is hyperactive in multiple cancer types1,2. Here, we performed integrative analysis of single cell transcriptomic profiling, paired T cell receptor (TCR) sequencing, and spatial transcriptomic profiling on Tuberous Sclerosis Complex (TSC) associated tumors with mTORC1 hyperactivity, and identified a stem-like tumor cell state (SLS) linked to T cell dysfunction via tumor-modulated immunosuppressive macrophages. Rapamycin and its derivatives (rapalogs) are the primary treatments for TSC tumors, and the stem-like tumor cells showed rapamycin resistance in vitro, reminiscent of the cytostatic effects of these drugs in patients. The pro-angiogenic factor midkine (MDK) was highly expressed by the SLS population, and associated with enrichment of endothelial cells in SLS-dominant samples. Inhibition of MDK showed synergistic benefit with rapamycin in reducing the growth of TSC cell lines in vitro and in vivo. In aggregate, this study suggests an autocrine rapamycin resistance mechanism and a paracrine tumor survival mechanism via immune suppression adopted by the stem-like state tumor cells with mTORC1 hyperactivity.

miRNAs and isomiRs: Serum-Based Biomarkers for the Development of Intellectual Disability and Autism Spectrum Disorder in Tuberous Sclerosis Complex

Tuberous sclerosis complex (TSC) is a rare multi-system genetic disorder characterized by a high incidence of epilepsy and neuropsychiatric manifestations known as tuberous-sclerosis-associated neuropsychiatric disorders (TANDs), including autism spectrum disorder (ASD) and intellectual disability (ID). MicroRNAs (miRNAs) are small regulatory non-coding RNAs that regulate the expression of more than 60% of all protein-coding genes in humans and have been reported to be dysregulated in several diseases, including TSC. In the current study, RNA sequencing analysis was performed to define the miRNA and isoform (isomiR) expression patterns in serum. A Receiver Operating Characteristic (ROC) curve analysis was used to identify circulating molecular biomarkers, miRNAs, and isomiRs, able to discriminate the development of neuropsychiatric comorbidity, either ASD, ID, or ASD + ID, in patients with TSC. Part of our bioinformatics predictions was verified with RT-qPCR performed on RNA isolated from patients’ serum. Our results support the notion that circulating miRNAs and isomiRs have the potential to aid standard clinical testing in the early risk assessment of ASD and ID development in TSC patients.

nab-Sirolimus for patients with advanced malignant PEComa with or without prior mTOR inhibitors: Biomarker results from AMPECT and an expanded access program

11574

Background: Malignant perivascular epithelioid cell tumor (PEComa) is a rare and aggressive sarcoma. nab-Sirolimus is an albumin-bound intravenous (IV) mTOR inhibitor (mTORi) approved for the treatment of adult patients with locally advanced unresectable or metastatic malignant PEComa. The AMPECT trial (NCT02494570) was the first prospective study in advanced malignant PEComa. In exploratory biomarker analyses, TSC1 or TSC2 alterations were associated with response. We report data from the final analysis of AMPECT patients, who were naïve to mTORi, and in patients with malignant PEComa with prior mTORi exposure treated with nab-sirolimus in an expanded access program (EAP) (NCT03817515). Methods: In AMPECT, patients with malignant PEComa naïve to mTORi received nab-sirolimus (100 mg/m2 IV days 1 and 8 of every 21-day cycle) until progression or unacceptable toxicity. The primary endpoint was ORR by independent radiology review. Other endpoints included duration of response (DOR) and disease control rate (DCR), defined as complete response (CR), partial response (PR), or stable disease (SD) at ≥12 weeks. In the EAP, patients with malignant PEComa and prior mTORi exposure received the same dose of nab-sirolimus as in AMPECT. Responses and DCR were evaluated post hoc via electronic medical record review. Genetic profiling, including TSC1 or TSC2 status, was assessed in both protocols, but no specific mutation criteria were required for enrollment. Results: Data include 47 total efficacy-evaluable patients, 31 in AMPECT and 16 with malignant PEComa and prior mTORi exposure treated in the EAP from July 2019–July 2021. Prior mTORi on the EAP included sirolimus, everolimus, temsirolimus, or sapanisertib; 12 patients had exposure to 1 prior mTORi and 4 to ≥2 prior mTORi, and 50% had had progressive disease as best response on mTORi. In AMPECT, ORR was 39% (12/31 patients), and DCR was 71%. Median DOR was not reached after 3 years of follow-up. On the EAP, 4/16 patients (25%) achieved PR (DOR range: 1.3+–25.2+ months, 3 ongoing), and 8/16 (50%) had SD as best response; the DCR was 63% (10/16). Of patients with known TSC1 or TSC2 inactivating alterations in the combined datasets (n = 23), 57% had a response (AMPECT, 64%; EAP, 44%). There were no Grade 4 or 5 treatment-related adverse events on either protocol Conclusions: nab-Sirolimus provided durable responses in mTORi-naïve patients with malignant PEComa and clinical benefit in an expanded access protocol for patients with malignant PEComa with prior mTORi therapy. Although AMPECT and the EAP cannot be directly compared, response rates showed similar trends regardless of prior mTORi exposure and in patients with TSC1 or TSC2 alterations. Based on the emerging biomarker results, a tissue-agnostic study in patients with TSC1 and TSC2 alterations has been initiated (NCT05103358). Clinical trial information: NCT02494570, NCT03817515.

LCMC LEADER neoadjuvant screening trial: LCMC4 evaluation of actionable drivers in early-stage lung cancers

TPS8596

Background: Comprehensive genomic profiling (CGP) has transformed the care of patients with advanced non-small cell lung cancer (NSCLC), giving many patients access to precision targeted treatment and immunotherapy with remarkable improvements in outcomes. Studies show that patients with lung cancers with oncogenic drivers are the least likely group to benefit from checkpoint inhibitors and are better served by enrollment in studies of targeted therapies. Early-stage NSCLC is now poised to benefit from these precision approaches with the regulatory approval of the first tyrosine kinase inhibitors and checkpoint inhibitors for the adjuvant treatment of resected NSCLC, each requiring testing for precision biomarkers. Neoadjuvant precision therapy for NSCLC has the potential to further improve treatment outcomes. Methods: The LCMC4 Evaluation of Actionable Drivers in EaRly Stage Lung Cancer (LEADER) Neoadjuvant Screening Trial (NCT04712877) is a collaborative diagnostic study developed by the Lung Cancer Mutation Consortium (LCMC), supported by the Thoracic Surgery Oncology Group and the Lung Cancer Research Foundation. The primary objective is to determine the proportion of patients with stage IA2-III lung cancers who possess actionable oncogenic drivers, defined as 1 of 11 actionable genomic alterations: mutations in EGFR, BRAFV600E, MET exon 14, KRAS G12C, and HER2, rearrangements in ALK, RET, NTRK, and ROS1, and amplification of MET and HER2. The study will also assess the feasibility of CGP to detect actionable oncogenic drivers in patients with suspected early-stage lung cancers scheduled to undergo biopsies to establish the diagnosis of lung cancer. The protocol will enroll 1000 patients with operable stage IA2-III (TNM 8th edition) lung cancer who will undergo CGP utilizing the Foundation Medicine 324 gene assay as well as paired liquid biopsy analysis. Results will enable selection of neoadjuvant therapy and enrollment onto independent therapeutic trials with genomically matched neoadjuvant treatment, standard therapies, or other trials if no driver is detected. The approach will be considered feasible if >35% of non-squamous NSCLCs have 1 of the 11 actionable alterations. Tumor mutational burden and PD-L1 IHC will be assessed. Plasma specimens collected pre- and post neoadjuvant treatment and post-surgery will be used for research to study the ability of circulating tumor DNA to assess neoadjuvant treatment response and minimal residual disease. 26 academic sites in the US plan to enroll patients. Clinical trial information: NCT04712877.

Germline variants across self-reported racial populations with urothelial carcinoma (UC)

4569

Background: Prior studies of the UC germline landscape centered around White patients with minimal representation of other racial populations. Herein, we examine the frequency of germline pathogenic and likely pathogenic (P/LP) variants in 2,582 patients with UC from various racial populations. Methods: 2,582 patients with UC underwent germline testing of 1 to 126 genes using massively parallel sequencing with customized capture bait-sets to analyze exonic regions, flanking intronic regions, and copy number alterations (CNAs). P/LP variants including single nucleotide variants, indels and CNAs were reported. Fisher’s Exact test and multivariable logistic regression were used after accounting for the number of genes tested, age at diagnosis, site of disease (upper versus lower tract), gender, family history of UC, and personal history of any cancer. Results: Among the 2,582 patients with UC, median age at diagnosis was 63 years (range, 4-90) and 1158 (44.8%) were female. There were 58 Asians (2.2%), 110 Blacks (4.3%), and 2,414 Whites (93.5%). Overall, 1,639/2,582 (63.5%) patients had a personal history of another cancer and 284/2,393 (11.9%) had history of UC in a family member. 465 P/LP variants were identified in 18% of patients, among whom 286 (11.1%) harbored ≥1 clinically actionable variants. P/LP in cancer-associated genes were most frequently reported in MSH2 (72/2,512, 2.9%), monoallelic MUTYH (45/2,136, 2.1%), BRCA2 (44/2,299, 1.9%) and MSH6 (47/2,511, 1.9%). Patients with upper tract UC had significantly more P/LP (72/247, 29.1%) compared to lower tract UC (332/2,076, 16%, p= 1.3x10-5). Age at diagnosis, gender, personal history of other primary cancers, or family history of UC were not significantly associated with the prevalence of P/LP variants. There were no significant differences ( p= 0.33) in P/LP variants across Asians (11/58, 19.0%), Blacks (14/110, 12.7%), and Whites (440/2,414, 18.2%) although a trend towards lower P/LP in Blacks is notable. Compared to Whites, Blacks and Asians harbored significantly more variants of unknown significance (VUS, Whites vs Blacks: 241/2414 vs 25/110, p= 0.0015; Whites vs Asians: 241/2414 vs 17/58, p= 4e-7). Asians with UC harbored significantly more P/LP variants in ATM (2/50, 4%) compared to Whites (30/2122, 1.4%, OR = 1.1 [95% CI, 1.0-1.2]) and Asian controls from the gnomAD Database. There were no significant differences across racial populations for other highly altered genes ( BRCA1/2, CHEK2, FH, MSH2/6, MUTYH) or for actionable variants. Conclusions: Germline P/LP variants were identified in 18% of patients with UC and were enriched in upper tract tumors. Although no significant differences in P/LP prevalence were noted among patients of different racial populations, a trend towards lower P/LP in Blacks and a higher rate of VUS in Asians and Blacks suggest that ongoing analysis by genetic ancestry may provide richer admixture data and insights.

Dual CDKN2A/MTAP loss compared to CDKN2A loss alone and response to immune-checkpoint inhibitors (ICI) in advanced solid tumors

2622

Background: We previously showed that CDKN2A genomic alterations (GAs) are associated with resistance to ICI (Adib E, Clinical Cancer Research, 2021). The majority of such GAs are homozygous deletions, which commonly (̃50-80%) include MTAP, located 100kb telomeric of CDKN2A. MTAP loss leads to 5′-deoxy-5′-methylthioadenosine (MTA) accumulation and immunosuppressive effects in tumors. We examined combined CDKN2A/MTAP deletion vs. CDKN2A deletion/mutation alone as predictors of poor ICI response. Methods: We curated clinical data for cancer patients (pts) treated with ICI at Dana-Farber Cancer Institute through 6/2021, who had targeted panel sequencing. Inclusion criteria were: ICI in metastatic setting, ≥2 cycles, no concurrent systemic therapy, cancer type with > 50 pts treated. CDKN2A/ MTAP GAs were defined as a deep deletion affecting both genes; CDKN2A only GAs included both homozygous deletions and truncating mutations. Hazard ratios (HR) for overall survival (OS) and time-to-treatment failure (TTF) were derived using multivariable Cox regression, adjusted for prior lines of therapy, treatment type (single vs. combination ICI), tumor mutational burden and ECOG PS. We also used a machine learning approach to quantify the density of tumor-infiltrating lymphocytes (TILs) in digital whole-slide H&E images of 144 melanoma pts with available genomic data. Results: 921 pts with 6 cancer types were studied: non-small cell lung cancer (NSCLC, n = 366), melanoma (mel, n = 228), urothelial carcinoma (UC, n = 120), esophagogastric carcinoma (EGC,n = 90), head and neck squamous cell carcinoma (HNSCC, n = 58), and renal cell carcinoma (RCC, n = 59). UC pts with MTAP/ CDKN2A GAs had shorter OS and TTF than pts without GA in either gene (OS HR = 1.9[1.1-3.4], p = 0.005; TTF HR = 1.8[1.0-3.1], p = 0.0016) after adjusting for covariates. Similar results were seen for melanoma (OS HR = 2.5[1.4-2.6],p = 0.00065; TTF HR = 1.9[1.1-3.2],p = 0.018). There was no significant difference between pts with CDKN2A GA only and those without GA in either gene for OS or TTF in either UC or melanoma. CDKN2A/MTAP status was not associated with significantly shorter survival for NSCLC and EGC; while the analysis was confounded by HPV events for HNSCC, and underpowered for RCC. ML-based analysis of digital slides for melanoma, showed that tumors with CDKN2A GAs only (n = 42) had similar median density of TILs compared to tumors without GAs in either gene (n = 84; 920 vs. 943 TILs/mm2; p = 0.42). In contrast, tumors with co-occurring CDKN2A/ MTAP GAs had lower TIL density (529 TIL/mm2, n = 17 vs. 925 TIL/mm2, n = 126 (pooled); p = 0.018, Wilcoxon rank sum). Conclusions: In this study, we showed that co-occurrence of MTAP/CDKN2A GAs, but not CDKN2A GA only, was associated with worse outcomes in pts with UC and melanoma treated with ICI. Lower TIL density was also seen in melanoma tissue samples with combined MTAP/CDKN2A GA.

Distinct oncogenic signatures in malignant PEComa and leiomyosarcoma identified by integrative RNA-seq and H3K27ac ChIP-seq analysis

11552

Background: Malignant perivascular epithelioid cell tumor (PEComa) and leiomyosarcoma (LMS) are two sarcomas with overlapping morphologic and immunophenotypic features which can make their diagnostic distinction challenging. We aimed to characterize the transcriptional and epigenetic landscape of PEComa and LMS to identify distinguishing features. Methods: We performed whole transcriptome RNA-sequencing on 19 PEComas and compared their gene expression profile to 259 sarcomas from The Cancer Genome Atlas (TCGA) including 104 LMS. ChIP-sequencing for H3K27ac, a histone modification associated with activation of nearby genes/open chromatin, was conducted on 9 malignant PEComas and 12 LMS and were compared with publicly available data from 4 other sarcoma subtypes (chordoma; osteosarcoma; undifferentiated pleomorphic sarcoma; rhabdomyosarcoma; n = 29 tumors). Results: Genome-wide epigenetic and transcriptional analyses revealed overlapping patterns between PEComa and LMS, which were distinct from other sarcomas. However, we also identified a set of highly expressed and epigenetically distinct transcripts which may represent diagnostic?biomarkers: e.g., DAPL1, MLANA, SULT1C2, GPR143, and CHI3L1 for PEComa; and MYOCD, WDFC2, DES, MYH11, and CNN1 for LMS; each of which showed >17x fold higher expression for each tumor entity by DESeq2 (FDR<0.0001). Gene Set Enrichment Analyses (GSEA) demonstrated enrichment in the KEGG Lysosome pathway for PEComa (FDR=0.11), whereas myogenesis and smooth muscle contraction pathways were enriched in LMS (FDR=0.09). Integrative transcriptomic and epigenetic analyses revealed a unique set of master core transcription factors for each tumor type including among others MYOCD for LMS; MITF for PEComa, which require further functional investigation. Twelve selected genes including new as well as known and standard diagnostic markers (e.g., DAPL1, MLANA, GPR143, PNL2, CHI3L1, DES, MYH11, ER, CD68, PU.1, pS6 and CNN1) were validated by immunohistochemistry (IHC) in multiple sections from PEComa and LMS (n = 26). The combination of three melanocytic markers (HMB45, MLANA, PNL2) and pS6 can distinguish LMS from PEComas (**** p<0.0001). IHC for CD68 and PU.1 macrophage markers did not show any difference regarding the degree of immune infiltration in PEComa vs. LMS. Conclusions: Our studies revealed novel epigenetic signatures translating into lysosomal and melanocytic proteins for PEComa and myogenic proteins for LMS, which may serve as useful diagnostic biomarkers in the distinction of these two sarcoma subtypes.

TSC1-mutant bladder cancer expression signature in relation to nuclear localization of TFE3 and potential for targetable dependency

e16532

Background: Mutation and inactivation of the tumor suppressor gene TSC1 is a recurrent (6-10%) event in bladder cancer, but whether it functions as a driver event for tumor development has been uncertain. Methods: We performed differential gene expression and pathway analyses using RNA-seq data from the curated TCGA TSC1 mutant BLCA (n = 26) and TSC1 wild-type BLCA (n = 382) cohort and compared to an internal cohort of putative TSC1/TSC2-driven tumors (n = 63). Mechanistic studies, as well as RNA-seq and H3K27ac ChIP-seq analyses, were conducted in 2 TSC1 mutant/WT BLCA cell lines. Results: Comparison of The Cancer Genome Atlas (TCGA) TSC1-mutant bladder cancers ( TSC1mBLCA) with TCGA TSC1 wildtype tumors ( TSC1WTBLCA) identified a conserved TSC-associated expression signature, similar to ones seen in syndromic TSC tumors. GSEA and DESeq2 analyses implicated both mTORC1 hyperactivation, as well as activation of lysosomal pathways in TSC1mBLCA. We validated our findings by IHC analysis of a separate cohort of TSC1mBLCA (n = 5), compared to TSC1WTBLCA (n = 5). In addition, we found that TFE3, a transcriptional regulator of lysosomal gene expression, was relatively highly expressed in BLCA (compared to other MiT-TFE genes) and was localized to the nucleus in TSC1mBLCA but not in TSC1WTBLCA. Mechanistic studies of two TSC1mBLCA cell lines and their respective TSC1 addback derivatives, recapitulated the phenotype found in human tumors and demonstrated that TFE3 was both post-translationally modified and predominantly nuclear in TSC1-null cell lines compared to TSC1 addbacks. RNA-seq and H3K27ac ChIP-Seq analyses showed that TSC1mBLCA cell lines retained elements of the TSC-associated expression signature that was seen in TSC1mBLCA tumors, confirming differential activation of TFE3 in response to TSC1 loss. Nuclear localization of TFE3 in TSC1mBLCA cell lines was only partially reversed by rapamycin treatment and was unaffected by treatment of Torin1. SiRNA mediated knockdown of TFE3 significantly decreased cell growth and viability in TSC1mBLCA cell lines and did not result in compensatory upregulation of TFEB and MITF. Conclusions: Our findings indicate that TSC1 mutant bladder tumors retain elements of a conserved transcriptional signature that is characterized by nuclear localization and activation of TFE3. Aberrant TFE3 activation likely contributes to TSC1mBLCA development and may therefore be amenable to targeted therapy.

Artificial intelligence in digital pathology approach identifies the predictive impact of tertiary lymphoid structures with immune-checkpoints therapy in NSCLC

9065

Background: The presence of Tertiary Lymphoid Structures (TLS) in multiple cancer types has been recognized as a potential predictive biomarker for response to immune-checkpoint blockade. However, there is no standardized method to quantify their presence. In this context, Artificial Intelligence (AI)-based assessment of histology images may well contribute to improve reproducibility, accuracy and speed of TLS quantification. Methods: We developed an automated workflow for quantification of TLS on digitized H&E slides through A) pixel-level classification of tissue using supervised artificial neural networks model, B) object-level cell classification of candidate TLS regions, C) merging the two approaches for curation and validation of TLS versus non-TLS regions. 433 advanced stage non-small cell lung cancer (NSCLC) patients treated with first or subsequent line of anti-PD-(L)1 single agent at DFCI were included in this study. Results: TLS were detected in 37% (n = 161) of the patients H&E slides, with the highest score of 4.7 TLS per mm2 (interquartile range: Q1 = 0, Q2 = 0, Q3 = 0.03 TLS/mm2). TLS density (per mm2) was significantly higher in surgically resected (n = 246; TLSPOS= 49%) compared to bioptic samples (n = 187; TLSPOS= 21%). No association was observed between TLS and tumor mutational burden (TMB) or PD-L1 protein expression as continuous variables. Among clinically actionable mutations, EGFR (all subtypes) mutated patients (n = 38) had a significantly lower number of TLS compared to patients without EGFR mutations. Patients with ≥ 0.01 TLS/mm2 had a significantly higher objective response rate (32% vs 22%, p = 0.03), a significantly longer median progression-free survival (PFS, 4.8 vs 2.7 months, HR: 0.73, 95% CI: 0.59-0.90, p = 0.004), and a significantly improved median overall survival (OS, 16.5 vs 12.5 months, HR: 0.72, 95% CI: 0.57-0.92, p = 0.008). In multivariable analysis, after adjusting for PD-L1 (≥ vs < 50%), TMB (≥ vs < 10 mu/Mb), sex, age, ECOG score, smoking and line of treatment, TLS/mm2 (≥ vs < 0.01) levels were found to be an independent positive predictive factor for both PFS (HR:0.69, 95% CI: 0.54-0.88, p = 0.003) and OS (HR: 0.70, 95% CI: 0.52-0.93, p = 0.01). Conclusions: These findings suggest that TLS status is an independent predictor of immunotherapy effectiveness in NSCLC, with predictive value similar to that of PD-L1 expression and TMB. This novel AI system has potential for automated identification and quantification of the TLS on digital histopathological slides, and could be utilized in a standard pathology workflow with relative ease. These findings are currently being validated in other solid tumors and cohorts.

Digital quantification of lymphocytic infiltration on routine H&E images and immunotherapy response in non–small cell lung cancer

9066

Background: Current biomarker(s) for immuno-oncology (IO) therapy response prediction in lung cancer are limited. Additional predictive biomarkers are useful to help refine patient selection and guide precision therapy. Methods: Biopsy and surgical specimens stained with hematoxylin-eosin (H&E) were subjected to whole-slide scanning for 446 advanced stage non-small cell lung cancer (NSCLC) treated with single agent immune check point inhibitors (ICI). A machine learning model was trained on H&E images for classification of tumor infiltrating lymphocytes (TILs), tumor cells, and stromal cells in specific tissue types. Results: TIL levels were found to be highly variable, with a range of 12 to 4270 cells/mm2, and median of 319 (Q1 = 159, Q3 = 681). TIL levels were assessed on tissue samples from multiple organs which had shown primary or metastatic NSCLC, and were similar across all specimen sites except the liver, for which median TIL levels were significantly lower, at 90 cells/mm2. There was no correlation between tumor mutational burden (TMB) and TIL levels, while high TIL levels were correlated with high PD-L1 (≥ 50%) expression. Patients who experienced a partial/complete response to ICI therapy had a trend to higher median TILs compared to those who had progressive/stable disease (350 versus 310 cells/mm2, P = 0.09). In a multivariable analysis after controlling for covariates (incl. sex, age, cigarette smoking, ECOG, PD-L1, TMB & treatment line), a higher TIL level (≥ 250 cells/mm2) was an independent predictor of IO response for both progression-free survival (PFS; HRadj 0.70; 95% CI, 0.55 - 0.89; P = 0.003) and overall survival (HRadj 0.73; 95% CI, 0.56 - 0.95; P = 0.02). In a ROC analysis considering single biomarkers, PD-L1 had the highest AUC (0.68, P < 0.001), while TIL (AUC = 0.53, P = 0.08) and TMB (AUC = 0.55, P = 0.05) had similar AUC values for classifying responders from non-responders based on objective response rate. Using weighted linear regression approach to combine the biomarkers, paired PD-L1/TMB had the greatest AUC (0.70, P < 0.001) compared to PD-L1 single assay. In the PD-L1 negative (< 1%, N = 50) subgroup, TIL levels had superior predictive performance for classification of IO responders (AUC = 0.77, P = 0.02) compared to TMB (AUC = 0.57, P = 0.3), such that patients with a high TIL level (≥ 250 cells/mm2) had an improved PFS (median PFS: 2.7 vs 2.2 months; HR = 0.48; 95% CI, 0.26 - 0.87; P = 0.02). Conclusions: Digital TIL quantification with use of machine learning is feasible. TIL levels appear to be a robust and independent biomarker of likelihood of response to IO treatment in NSCLC, especially in the PD-L1 negative subgroup. The findings of this study are under validation in additional lung cancer cohorts.

Ultrasensitive profiling of UV-induced mutations identifies thousands of subclinical facial tumors in tuberous sclerosis complex

BackgroundTuberous sclerosis complex (TSC) is a neurogenetic syndrome due to loss-of-function mutations in TSC2 or TSC1, characterized by tumors at multiple body sites, including facial angiofibroma (FAF). Here, an ultrasensitive assessment of the extent and range of UV-induced mutations in TSC facial skin was performed.MethodsA multiplex high-sensitivity PCR assay (MHPA) was developed, enabling mutation detection at extremely low (<0.1%) variant allele frequencies (VAFs).ResultsMHPA assays were developed for both TSC2 and TP53, and applied to 81 samples, including 66 skin biopsies. UV-induced second-hit mutation causing inactivation of TSC2 was pervasive in TSC facial skin with an average of 4.8 mutations per 2-mm biopsy at median VAF 0.08%, generating more than 150,000 incipient facial tumors (subclinical "micro-FAFs") in the average TSC subject. The MHPA analysis also led to the identification of a refined UV-related indel signature and a recurrent complex mutation pattern, consisting of both a single-nucleotide or dinucleotide variant and a 1- to 9-nucleotide deletion, in cis.ConclusionTSC facial skin can be viewed as harboring a patchwork of clonal fibroblast proliferations (micro-FAFs) with indolent growth, a small proportion of which develop into clinically observable FAF. Our observations also expand the spectrum of UV-related mutation signatures.FundingThis work was supported by the TSC Alliance; the Engles Family Fund for Research in TSC and LAM; and the NIH, National Heart, Lung, and Blood Institute (U01HL131022-04 and Intramural Research Program).

Variation in targetable genomic alterations in non-small cell lung cancer by genetic ancestry, sex, smoking history, and histology

Background

Genomic alterations in 8 genes are now the targets of FDA-approved therapeutics in non-small cell lung cancer (NSCLC), but their distribution according to genetic ancestry, sex, histology, and smoking is not well established.

Methods

Using multi-institutional genetic testing data from GENIE, we characterize the distribution of targetable genomic alterations in 8 genes among 8675 patients with NSCLC (discovery cohort: DFCI, N = 3115; validation cohort: Duke, Memorial Sloan Kettering Cancer Center, Vanderbilt, N = 5560). For the discovery cohort, we impute genetic ancestry from tumor-only sequencing and identify differences in the frequency of targetable alterations across ancestral groups, smoking pack-years, and histologic subtypes.

Results

We identified variation in the prevalence of KRAS^G12C, sensitizing EGFR mutations, MET alterations, ALK, and ROS1 fusions according to the number of smoking pack-years. A novel method for computing continental (African, Asian, European) and Ashkenazi Jewish ancestries from panel sequencing enables quantitative analysis of the correlation between ancestry and mutation rates. This analysis identifies a correlation between Asian ancestry and EGFR mutations and an anti-correlation between Asian ancestry and KRAS^G12C mutation. It uncovers 2.7-fold enrichment for MET exon 14 skipping mutations and amplifications in patients of Ashkenazi Jewish ancestry. Among never/light smokers, targetable alterations in LUAD are significantly enriched in those with Asian (80%) versus African (49%) and European (55%) ancestry. Finally, we show that 5% of patients with squamous cell carcinoma (LUSC) and 17% of patients with large cell carcinoma (LCLC) harbor targetable alterations.

Conclusions

Among patients with NSCLC, there was significant variability in the prevalence of targetable genomic alterations according to genetic ancestry, histology, and smoking. Patients with LUSC and LCLC have 5% rates of targetable alterations supporting consideration for sequencing in those subtypes.

Supplementary Information

The online version contains supplementary material available at 10.1186/s13073-022-01041-x.

Evolution of electroencephalogram in infants with tuberous sclerosis complex and neurodevelopmental outcome: a prospective cohort study

AIM

To describe the evolution of electroencephalogram (EEG) characteristics in infants with tuberous sclerosis complex (TSC) and the relationship with neurodevelopmental outcome at 24 months.

METHOD

Eighty-three infants were enrolled in the EPISTOP trial and underwent serial EEG follow-up until the age of 24 months (males n=45, females n=37, median age at enrolment 28d, interquartile range 14-54d). Maturation of the EEG background and epileptiform discharges were compared between the TSC1 and TSC2 variants and between preventive and conventional groups respectively.

RESULTS

Children with TSC2 more frequently had a slower posterior dominant rhythm (PDR) at 24 months (51% vs 11%, p=0.002), a higher number of epileptiform foci (median=8 vs 4, p=0.003), and a lower fraction of EEGs without epileptiform discharges (18% vs 61%, p=0.001) at follow-up. A slower PDR at 24 months was significantly associated with lower cognitive (median=70 vs 80, p=0.028) and motor developmental quotients (median=70 vs 79, p=0.008). A higher fraction of EEGs without epileptiform discharges was associated with a lower probability of autism spectrum disorder symptoms (odds ratio=0.092, 95% confidence interval=0.009-0.912, p=0.042) and higher cognitive (p=0.004), language (p=0.002), and motor (p=0.001) developmental quotients at 24 months.

INTERPRETATION

TSC2 is associated with more abnormal EEG characteristics compared to TSC1, which are predictive for neurodevelopmental outcome.

FOXP3+ T-cell infiltration is associated with improved outcomes in metastatic urothelial carcinoma (mUC) treated with immune-checkpoint inhibitors (ICI)

549

Background: Given that ICIs benefit a minority of mUC patients and are associated with significant costs, biomarkers are necessary to optimally utilize them in the clinic. Although FOXP3+ T-cells have been associated with an immune-cold environment in many cancers, studies in urothelial carcinoma have shown an opposite trend. Methods: Formalin-fixed paraffin-embedded slides from tumor specimens were collected for patients with mUC treated with ICI at Dana-Farber Cancer Institute. A novel multiplex immunofluorescence (mIF) panel, ImmunoProfile, was performed for PD-L1, PD-1, FOXP3, CD8 and DAPI, then scanned by a Vectra Polaris platform. Regions of interest were defined and used for quantitative analysis using PerkinElmer/Akoya. Intratumoral (IT) and tumor-stroma interface (TSI) density (cells/mm2) of each cell type was calculated. Clinical data was collected through chart review, and associations between cell density and response were assessed. Wilcoxon Rank-Sum test between responders (CR/PR) and non-responders (SD/PD) was used to generate p-values, followed by Benjamini-Hochberg correction. Receiver-operating curve (ROC) and area-under-curve (AUC) calculations were performed to determine the optimal cutoff (OC) differentiating responders from non-responders. Cox proportional hazards models were used to estimate OS and PFS, accounting for type of therapy (single vs. combination ICI), baseline neutrophil-to-lymphocyte ratio (NLR), PD-L1 CPS, prior therapy, non-urothelial component, ECOG-PS and liver metastases (mets). Results: Of 35 patients assessed by ImmunoProfile, 32 were evaluable for response. Most patients (88%) were male and the median age at ICI start was 73 years. Median number of prior lines of therapy was 1 (range 0 – 3), and the majority (72%) was treated with single-agent ICI. Eight patients (25%) had CR/PR, eight had SD and 16 had PD as best response. Of all IF stains assessed, IT-TSI FOXP3 was the strongest predictor of objective response (q-value = 0.006), followed by IT-CD8 (q = 0.014). ROC analysis yielded an AUC of 0.812 (0.656 – 0.969) and the optimal cutpoint was set at 75 IT-TSI-FOXP3 cells/mm2. ORR was 46% in FOXP3-high ( > 75/mm2) vs 14% in FOXP3-low mUCs. A combined model using IT-TSI-FOXP3 and clinical covariates (NLR, ECOG-PS, line of therapy, histology, and liver mets) had an AUC of 0.929. PFS was significantly longer in the FOXP3-high group (7.9 [5.7 – NR] months) compared to the FOXP3-low group (2.3 [2.1 – 6.1] months) on multivariable analysis (p = 0.007). OS also showed the same trend (p = 0.1). Conclusions: In this pilot study of ImmunoProfile, a novel mIF panel, higher FOXP3+ infiltration in tumors was associated with better outcomes and a composite clinico-IHC panel exhibited robust prognostic impact in mUC pts treated with ICI. Further study of this mIF panel is warranted to implement it in routine practice.

Biomarker-Based Phase II Study of Sapanisertib (TAK-228): An mTORC1/2 Inhibitor in Patients With Refractory Metastatic Renal Cell Carcinoma

PURPOSE

Sapanisertib is a kinase inhibitor that inhibits both mammalian target of rapamycin complex 1 (mTORC1) and mTORC2. In this multicenter, single-arm phase II trial, we evaluated the efficacy of sapanisertib in patients with treatment-refractory metastatic renal cell carcinoma (mRCC; NCT03097328).

METHODS

Patients with mRCC of any histology progressing through standard therapy (including prior mTOR inhibitors) had baseline biopsy and received sapanisertib 30 mg by mouth once weekly until unacceptable toxicity or disease progression. The primary end point was objective response rate by RECIST 1.1. Tissue biomarkers of mTOR pathway activation were explored.

RESULTS

We enrolled 38 patients with mRCC (clear cell = 28; variant histology = 10) between August 2017 and November 2019. Twenty-four (63%) had received ≥ 3 prior lines of therapy; 17 (45%) had received prior rapalog therapy. The median follow-up was 10.4 (range 1-27.4) months. Objective response rate was two of 38 (5.3%; 90% CI, 1 to 15.6); the median progression-free survival (PFS) was 2.5 months (95% CI, 1.8 to 3.7). Twelve patients (32%) developed treatment-related grade 3 adverse events, with no grade 4 or 5 toxicities. Alterations in the mTOR pathway genes were seen in 5 of 29 evaluable patients (MTOR n = 1, PTEN n = 3, and TSC1 n = 1) with no association with response or PFS. Diminished or loss of PTEN expression by immunohistochemistry was seen in 8 of 21 patients and trended toward shorter PFS compared with intact PTEN (median 1.9 v 3.7 months; hazard ratio 2.5; 95% CI, 0.9 to 6.7; P = .055).

CONCLUSION

Sapanisertib had minimal activity in treatment-refractory mRCC independent of mTOR pathway alterations. Additional therapeutic strategies are needed for patients with refractory mRCC.

Association of Early MRI Characteristics With Subsequent Epilepsy and Neurodevelopmental Outcomes in Children With Tuberous Sclerosis Complex

BACKGROUND AND OBJECTIVE

Multiple factors have been found to contribute to the high risk of epilepsy in infants with Tuberous Sclerosis Complex (TSC), including evolution of EEG abnormalities, TSC gene mutation and MRI characteristics. The aim of the present prospective multi-center study was to: 1) identify early MRI biomarkers of epilepsy in infants with TSC aged < 6 months and before seizure onset, and 2) associate these MRI biomarkers with neurodevelopmental outcomes at 2 years of age. The study was part of the EPISTOP project.

METHODS

We evaluated brain MRIs performed in infants with TSC younger than 6 months of age. We used harmonized MRI-protocols across centers and children were monitored closely with neuropsychological evaluation, and serial video EEG. MRI characteristics defined as tubers, radial migration lines, white matter abnormalities, cysts, calcifications, subependymal nodules (SEN) and subependymal giant cell astrocytoma (SEGA) were visually evaluated and lesions were detected semi-automatically. Lesion to brain volume ratios were calculated and associated with epilepsy and neurodevelopmental outcomes at two years.

RESULTS

Lesions were assessed on MRIs from 77 TSC infants, 62 MRIs were sufficient for volume analysis. The presence of tubers and higher tuber-brain ratios were associated with the development of clinical seizures, independently of TSC gene mutation and preventive treatment. Furthermore, higher tuber-brain ratios were associated with lower cognitive and motor development quotients at two years, independently of TSC gene mutation and presence of epilepsy.

DISCUSSION

In infants with TSC, there is a significant association between characteristic TSC lesions detected on early brain MRI and development of clinical seizures, as well as neurodevelopmental outcomes in the first two years of life. According to our results, early brain MRI findings may guide clinical care for young children with TSC.

CLASSIFICATION OF EVIDENCE

This study provides Class I evidence that in infants with TSC, there is a significant association between characteristic TSC lesions on early brain MRI and the development of clinical seizures and neurodevelopmental outcomes in the first two years of life.

Down-regulation of the brain-specific cell-adhesion molecule contactin-3 in tuberous sclerosis complex during the early postnatal period

Background

The genetic disorder tuberous sclerosis complex (TSC) is frequently accompanied by the development of neuropsychiatric disorders, including autism spectrum disorder and intellectual disability, with varying degrees of impairment. These co-morbidities in TSC have been linked to the structural brain abnormalities, such as cortical tubers, and recurrent epileptic seizures (in 70–80% cases). Previous transcriptomic analysis of cortical tubers revealed dysregulation of genes involved in cell adhesion in the brain, which may be associated with the neurodevelopmental deficits in TSC. In this study we aimed to investigate the expression of one of these genes – cell-adhesion molecule contactin-3.

Methods

Reverse transcription quantitative polymerase chain reaction for the contactin-3 gene (CNTN3) was performed in resected cortical tubers from TSC patients with drug-resistant epilepsy (n = 35, age range: 1–48 years) and compared to autopsy-derived cortical control tissue (n = 27, age range: 0–44 years), as well as by western blot analysis of contactin-3 (n = 7 vs n = 7, age range: 0–3 years for both TSC and controls) and immunohistochemistry (n = 5 TSC vs n = 4 controls). The expression of contactin-3 was further analyzed in fetal and postnatal control tissue by western blotting and in-situ hybridization, as well as in the SH-SY5Y neuroblastoma cell line differentiation model in vitro.

Results

CNTN3 gene expression was lower in cortical tubers from patients across a wide range of ages (fold change = − 0.5, p < 0.001) as compared to controls. Contactin-3 protein expression was lower in the age range of 0–3 years old (fold change = − 3.8, p < 0.001) as compared to the age-matched controls. In control brain tissue, contactin-3 gene and protein expression could be detected during fetal development, peaked around birth and during infancy and declined in the adult brain. CNTN3 expression was induced in the differentiated SH-SY5Y neuroblastoma cells in vitro (fold change = 6.2, p < 0.01).

Conclusions

Our data show a lower expression of contactin-3 in cortical tubers of TSC patients during early postnatal period as compared to controls, which may affect normal brain development and might contribute to neuropsychiatric co-morbidities observed in patients with TSC.

Supplementary Information

The online version contains supplementary material available at 10.1186/s11689-022-09416-2.

Abstract P138: nab-Sirolimus improves mTOR pathway suppression and antitumor activity versus oral mTOR inhibitors in PTEN null bladder cancer (UMUC3) and TSC2 null liver cancer (SNU398) xenografts

Background: TSC1, TSC2, and PTEN genes are tumor suppressors in the mTOR pathway and can be inactivated or deleted across many cancers (Kwiatkowski, Clin Cancer Res 2016). The mTORC1 pathway is frequently activated in cancer and causes phosphorylation of downstream targets S6K (activation) and 4EBP1 (inactivation). Nanomolar concentrations of mTOR inhibitors (mTORi) sirolimus and everolimus can effectively inhibit S6K but not 4EBP1 and may lead to therapeutic resistance (Kang, Science 2013). In a registrational phase 2 trial (AMPECT) with malignant PEComa, nab-sirolimus (ABI-009) had a response rate of 64% (9/14) in patients with TSC1 or TSC2 mutations. In an expanded access program (NCT03817515), of 8 patients with TSC1 or TSC2 mutations (2 had prior mTORi) treated with nab-sirolimus, 5 had partial responses (all mTORi naïve). This study investigated mTOR pathway inhibition, tumor drug levels, and antitumor activity of nab-sirolimus vs equal doses of oral mTORi in PTEN-null and TSC2-null xenograft models. Methods: Athymic mice bearing subcutaneous PTEN-null UMUC3 bladder cancer xenografts were treated with either saline or equal weekly doses (15 mg/kg) of nab-sirolimus (IV, 7.5 mg/kg, 2x/wk), and sirolimus or everolimus (PO, 3 mg/kg/day, 5 days/wk). Tumors were harvested and analyzed for tumor drug levels (LC-MS/MS) and pS6 inhibition by immunohistochemistry (IHC). The same treatment conditions were repeated in a subsequent experiment with TSC2-null SNU-398 hepatocellular carcinoma xenografts, which further analyzed pS6K, pS6, and p4EBP1 via western blot (WB). Results: In UMUC3 xenografts, compared with oral mTORi, IV nab-sirolimus resulted in significantly higher drug exposure (AUC 7d) in the tumor (P&lt;0.0001) and greater pS6 inhibition as measured by IHC (P=0.0001 vs sirolimus, P=0.0034 vs everolimus). Correspondingly, nab-sirolimus resulted in significantly greater tumor growth inhibition (TGI) than sirolimus (69.6% vs 24.3%, P&lt;0.0001) and everolimus (36.2%, P=0.0023), and prolonged animal survival vs both oral mTORi (P&lt;0.05 log-rank). Based on WB in SNU-398 xenografts, IV nab-sirolimus consistently inhibited mTOR targets pS6K, pS6, and p4EBP1, whereas oral sirolimus only partially decreased pS6K and pS6 and did not appear to reduce p4EBP1 levels. Correspondingly, nab-sirolimus resulted in significantly greater TGI than sirolimus (67.8% vs 36.2%, P&lt;0.05) and prolonged animal survival (P&lt;0.05 log-rank). Conclusions: The relatively low tumor concentrations achieved with oral mTORi may limit their effectiveness as anticancer therapies. IV nab-sirolimus at equal dose showed significantly higher tumor accumulation and inhibition of pS6 in a PTEN-null bladder cancer xenograft and increased inhibition of mTOR targets pS6K, pS6, and p4EBP1 in a TSC2-null hepatocellular carcinoma xenograft. This was accompanied with significantly greater antitumor activity, suggesting that nab-sirolimus may have a more optimal pharmacologic profile than the oral mTORi. Clinical studies in cancers harboring these alterations are planned.

Citation Format: Shihe Hou, Heng Du, Anita N. Schmid, David J. Kwiatkowski, Neil P. Desai. nab-Sirolimus improves mTOR pathway suppression and antitumor activity versus oral mTOR inhibitors in PTEN null bladder cancer (UMUC3) and TSC2 null liver cancer (SNU398) xenografts [abstract]. In: Proceedings of the AACR-NCI-EORTC Virtual International Conference on Molecular Targets and Cancer Therapeutics; 2021 Oct 7-10. Philadelphia (PA): AACR; Mol Cancer Ther 2021;20(12 Suppl):Abstract nr P138.