The clinical success of immune-checkpoint inhibitors (ICI) in both resected and metastatic melanoma has confirmed the validity of therapeutic strategies that boost the immune system to counteract cancer. However, half of patients with metastatic disease treated with even the most aggressive regimen do not derive durable clinical benefit. Thus, there is a critical need for predictive biomarkers that can identify individuals who are unlikely to benefit with high accuracy so that these patients may be spared the toxicity of treatment without the likely benefit of response. Ideally, such an assay would have a fast turnaround time and minimal invasiveness. Here, we utilize a novel platform that combines mass spectrometry with an artificial intelligence-based data processing engine to interrogate the blood glycoproteome in melanoma patients before receiving ICI therapy. We identify 143 biomarkers that demonstrate a difference in expression between the patients who died within six months of starting ICI treatment and those who remained progression-free for three years. We then develop a glycoproteomic classifier that predicts benefit of immunotherapy (HR=2.7; p=0.026) and achieves a significant separation of patients in an independent cohort (HR=5.6; p=0.027). To understand how circulating glycoproteins may affect efficacy of treatment, we analyze the differences in glycosylation structure and discover a fucosylation signature in patients with shorter overall survival (OS). We then develop a fucosylation-based model that effectively stratifies patients (HR=3.5; p=0.0066). Together, our data demonstrate the utility of plasma glycoproteomics for biomarker discovery and prediction of ICI benefit in patients with metastatic melanoma and suggest that protein fucosylation may be a determinant of anti-tumor immunity.

Background: Protein glycosylation is the most common and complex form of post-translational protein modification. Glycosylation profoundly affects protein structure, conformation, and function. The elucidation of the potential role of differential protein glycosylation as biomarkers has been limited by the technical complexity of generating and interpreting this information. We have recently established a novel, powerful platform that combines liquid chromatography-mass spectrometry with a proprietary artificial-intelligence-based data processing engine that allows, for the first time, highly scalable interrogation of the glycoproteome. Here we report the performance of this platform to predict likely benefit from immune-checkpoint inhibitor (ICI) therapy in advanced non-small cell lung cancer (NSCLC). Methods: Our platform was utilized to assess 532 glycopeptide (GP) and peptide signatures representing 75 serum proteins in pretreatment blood samples from a cohort of 123 individuals (54 females, 69 males, age range 30 to 88 years). Inclusion criteria were a diagnosis of unresectable stage 3 or 4 NSCLC, treatment with pembrolizumab monotherapy (26 patients), or treatment with combination pembrolizumab-chemotherapy (97 patients). Overall survival (OS) data were available for all patients. Results: An ensemble multivariable-model-based glycoproteomic classifier consisting of 7 GP and non-glycosylated peptide biomarker features selected from a generalized additive model for OS was developed using ≈2/3rds of the full cohort (n=88) and validated in the remainder of patients (n=35). The classifier yielded similar statistical significance in Cox regression analysis for separating patients who are likely to benefit from ICI therapy from those who are not, to accurately predict likely ICI benefit with a sensitivity of >95% while performing at a specificity of 33% to predict those who are unlikely to benefit. Results were further analyzed in patients with either non-squamous or squamous NSCLC with first-line therapy (n=98). The classifier yielded a hazard ratio (HR) for prediction of likely ICI benefit of 3.6 with median OS of 13.9 vs. 4.2 months, and of 3.5 with median OS of 13.5 vs. 4.5 months in the entire cohort and the first-line treated patients, respectively. Conclusions: The glycoproteomic classifier described here predicts with high sensitivity which patients are likely to benefit from ICI therapy. In addition to potentially reducing the use of ICIs in a safe manner in patients who would be unnecessarily subjected to possible adverse drug reactions, our classifier simultaneously has the potential of reducing the burden of health care expenditures. Our results indicate that glycoproteomics holds a strong promise as a predictor for ICI treatment benefit which appears to significantly outperform other currently pursued biomarker approaches. Citation Format: Klaus Lindpaintner, Chad Pickering, Alan Mitchell, Gege Xu, Xin Cong, Daniel Serie. A peripheral blood-based glycoproteomic predictor of checkpoint inhibitor treatment benefit in advanced non-small cell lung cancer. [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2023; Part 1 (Regular and Invited Abstracts); 2023 Apr 14-19; Orlando, FL. Philadelphia (PA): AACR; Cancer Res 2023;83(7_Suppl):Abstract nr 5314.

Background Immune checkpoint inhibitors (ICIs) have revolutionized melanoma treatment, necessitating predictive biomarkers to identify patients likely to benefit. To that end, this study leverages a novel platform that combines liquid chromatography/mass spectrometry with a proprietary artificial-intelligence-based data processing engine, allowing for highly scalable and reproducible interrogation of glycoproteins with site-and glycan-specificity, capable of identifying blood-based predictive biomarkers using pre-treatment plasma samples from metastatic melanoma (MM) patients. Methods We interrogated 521 glycopeptide (GP) and 75 peptide biomarkers in a discovery cohort of pre-treatment plasma samples obtained from 202 patients with metastatic melanoma (MM) treated with anti-PD-1 monotherapy (pembrolizumab or nivolumab (57%), or anti-CTLA-4 (ipilimumab) with/without nivolumab (43%) (table 1). In addition to using age- and sex-adjusted regression to identify differentially abundant biomarkers where overall survival (OS) from ICI therapy start was the primary endpoint, patients were divided into those having early treatment failures (death within 6-months), intermediate controls (progression of death l between 6-months and 3-years), and sustained controls (progression-free for at least 3 years). Next, the discovery cohort was divided into a training, test, and validation set to develop and assess a repeated cross-validated LASSO-regularized Cox-based glycoproteomic classifier. To externally validate the classifier, an independent cohort of 27 MM patients were tested (table 2). Lastly, given the link between fucosylation and MM, engineered fucosylation-features were used in a second classifier. Results We identified 143 markers that significantly distinguished patients with early treatment failure from those with sustained controls (figure 1). A 14-marker classifier achieved a high degree of separation (table 3-detailed performance metrics) between those likely to benefit (i.e. those predicted to achieve long-term clinical benefit) and unlikely to benefit (Cox proportional hazard ratio/H.R. = 2.7, p-value = 0.026) (figure 2) while also yielding comparable performance in an independent cohort (H.R = 5.6, p-value = 0.027) (table 3). The secondary fucosylated-based classifier was also able to distinguish patients with and without long-term benefit (H.R = 3.5, p-value = 0.0066) (figure 4). Conclusions Using glycoproteomic profiling, our classifier predicted which MM patients treated with ICIs had nearly a 3-fold greater likelihood of durable benefit, with the finding validated in an independent cohort. Our results also suggest circulating glycoprotein fucosylation may be an important determinant of anti-tumor immunity. These data demonstrate the utility of plasma glycoproteomics for biomarker discovery and prediction of ICI benefit in patients with MM. Future directions include prospective confirmatory testing. Acknowledgements The authors thank James Richard Hartness, Jr. and Kim Vigal for their alliance management efforts and critical inputs for this abstract. References Shum B, Larkin J, Turajlic S. Predictive biomarkers for response to immune checkpoint inhibition. Semin Cancer Biol. 2022 Feb;79:4–17. doi: 10.1016/j.semcancer.2021.03.036. Epub 2021 Apr 2. PMID: 33819567. Dhar C, Ramachandran P, Xu G, Pickering C, Caval T, Rice R, Zhou B, Srinivasan A, Hundal I, Cheng R, Aiyetan P. Diagnosing and staging epithelial ovarian cancer by serum glycoproteomic profiling. medRxiv 2023.03.20.23287422 [Preprint]. March 20, 2023 [cited 2023 Jun 26]. Available from: https://doi.org/10.1101/2023.03.20.23287422 Agrawal P, Fontanals-Cirera B, Sokolova E, Jacob S, Vaiana CA, Argibay D, Davalos V, McDermott M, Nayak S, Darvishian F, Castillo M, Ueberheide B, Osman I, Fenyö D, Mahal LK, Hernando E. A Systems Biology Approach Identifies FUT8 as a Driver of Melanoma Metastasis. Cancer Cell. 2017 Jun 12;31(6):804–819.e7. doi: 10.1016/j.ccell.2017.05.007. PMID: 28609658; PMCID: PMC5649440. Ethics Approval Plasma samples were collected under MGH IRB protocols 12–488 & 11–181and Central Adelaide Local Health Network Human Research Ethics Committee protocol HREC/16/RAH/95. Written informed consent was obtained from all patients prior to inclusion in the study.

69 Background: Colorectal cancer (CRC) remains a leading cancer despite current screening modalities. Precancerous lesions, or Advanced Adenomas (AA), commonly precede invasive cancer development by years. Newer technologies use circulating tumor DNA and/or proteins for CRC detection but have not been able to effectively detect AA. Aberrant protein glycosylation is associated with (pre-)malignant lesions. To detect glycoproteome profiles associated with the occurrence of AA, we studied serum glycoproteins in AA/CRC. Methods: A novel platform combining liquid-chromatography/mass-spectrometry (LC-MS) and artificial-intelligence (AI)-powered data processing allowing high resolution, high throughput glycoproteomic profiling was used to identify glycoprotein biomarkers in peripheral blood. Samples were sourced from biorepositories and included patients diagnosed with CRC, AA, ulcerative colitis (UC) and controls. The samples were split into a training (50%) and a hold-out testing set (50%) for the development of a machine learning (ML)-based multivariable predictive model. Statistical analysis was performed on normalized data to identify biomarkers differentiating AAs and different stages of CRC from controls. Results: We studied 563 patient samples: 196 controls (mean age 51.7; 52% female); 32 AA (mean age 68.6; 53% female); 247 CRC (mean age 65.6; 50% female) and 88 UC (mean age 44.1; 47% female). There were 250 differentially abundant (FDR < 0.05) glycopeptides/peptides when comparing CRC and AA samples with healthy and UC controls. A subset was assessed, generating a six (6) biomarker ML classification model. This model was applied to the hold-out test and achieved an overall sensitivity of 91.4% and specificity of 91.8% for predicting AA/CRC versus healthy/UC with an area under the receiver operating characteristic of 0.962. AA and CRC separately were predicted with a sensitivity of 84.4% and 92.8%, respectively, relative to healthy/UC with sensitivities for CRC stage 1/2 and stage 3/4 being 91.2% and 93.2%, respectively). Conclusions: Glycoproteomic serum profiles accurately detect precancerous AA in addition to CRC and offer a new approach to effective CRC screening. We will have completed an interim analysis of a large prospective observational study at the time of the meeting. Clinical trial information: NCT05445570 . [Table: see text]

e21148 Background: Protein glycosylation is the most abundant and complex form of post-translational protein modification. Glycosylation profoundly affects protein structure, conformation, and function. The elucidation of the potential role of differential protein glycosylation as biomarkers has so far been limited by the technical complexity of generating and interpreting this information. We have recently established a novel, powerful platform that combines liquid chromatography/mass spectrometry with a proprietary artificial-intelligence-based data processing engine that allows, for the first time, highly scalable interrogation of the glycoproteome. Methods: Using this platform, we interrogated 694 glycopeptide (GP) and non-glycosylated peptide transitions derived from 74 serum proteins in pre-treatment peripheral blood samples from a cohort of 316 individuals with non-small-cell lung cancer (NSCLC) (128 females, 187 males, 1 with unknown sex, median age 66 years, age range 31-89 years, stage 0-4 N’s: 1 / 99 / 80 / 84 / 49, 3 missing) and a comparison cohort of 194 healthy control samples (102 females, 92 males, median age 52 years, age range 30-63 years). Age- and sex-adjusted differential expression analysis for 596 normalized biomarkers were performed to evaluate statistically significant differential abundances using an FDR-adjusted q-value of 0.05 as a cutoff. Repeated five-fold cross-validated LASSO-regularized logistic regression was performed to create a multivariable classifier that predicts whether a serum sample belongs to the healthy or NSCLC cohort. Results: We identified 432 biomarkers with significant abundance differences at FDR ≤ 0.05 between samples with NSCLC and healthy controls. Using 70% of the complete cohort (balanced by case/control membership, NSCLC stage, sex, and age quartile) as a training set, we selected a total of 375 glycopeptide and non-glycosylated peptide biomarker features that remained differentially expressed at FDR-adjusted q-value ≤ 0.05 as input into a LASSO-regularized multivariable classifier. This resulting in a 19-biomarker model exhibiting an accuracy of 94.8% (96.9% sensitivity, 91.2% specificity) and AUC of 0.989. This classifier was validated in an independent test set comprising the remaining 30% of subjects, yielding an accuracy of 94.5% (95.5% sensitivity, 93.0% specificity) and AUC of 0.975. Sensitivity in the test set was 100% / 96% / 99% / 96% / 94% / 10%, in stages 0-4 and missing, respectively. Conclusions: Our results indicate that glycoproteomic biomarkers can be leveraged as a strong liquid biopsy-based screening tool for patients at high risk of NSCLC, as an alternative to imaging modalities.

Background Protein glycosylation is the most abundant and complex form of post-translational protein modification. Glycosylation profoundly affects protein structure, conformation, and function. The elucidation of the potential role of differential protein glycosylation as biomarkers has been limited by the technical complexity of generating and interpreting this information. We have recently established a novel, powerful platform that combines liquid chromatography-mass spectrometry with a proprietary artificial-intelligence-based data processing engine that allows, for the first time, highly scalable interrogation of the glycoproteome. Here we report the performance of this platform to predict likely benefit from immune-checkpoint inhibitor (ICI) therapy in advanced non-small cell lung cancer (NSCLC). Methods Our platform was utilized to assess 532 glycopeptide (GP) and peptide signatures representing 75 serum proteins in pretreatment blood samples from a cohort of 125 individuals (54 females, 71 males, age range 60 to 75 years). Inclusion criteria were as follows: a diagnosis of unresectable stage 3 or 4 NSCLC, treatment with pembrolizumab monotherapy (27 patients), or treatment with combination pembrolizumab-chemotherapy (98 patients). Overall survival (OS) data were available for all patients. Samples and de-identified clinical data were obtained from Tempus Labs (Chicago, IL). Results A multivariable-model-based classifier for OS was created utilizing 70% of the cohort as a training set and seven glycopeptide and non-glycosylated peptide biomarker features selected from a generalized additive model. The classifier yielded a hazard ratio (HR) for prediction of likely ICI benefit of 3.96 at p < 0.0001. Additionally, the classifier was validated using a test set comprised of the withheld 30% of patients, yielding a HR of 3.86 at p< 0.01 which separated patients likely benefiting from ICI therapy from those likely not benefiting from ICI therapy (median OS of 23.2 vs. 5.9 months, respectively, based on classifier score above/below cutoff). Conclusions The glycoproteomic classifier described here predicts with high sensitivity which patients are likely to benefit from ICI therapy. In addition to potentially reducing the use of ICIs in a safe manner in patients who would be unnecessarily subjected to possible adverse drug reactions, our classifier simultaneously has the potential of reducing the burden of health care expenditures. Our results indicate that glycoproteomics holds a strong promise as a predictor for ICI treatment benefit which appears to significantly outperform other currently pursued biomarker approaches. Ethics Approval The study was conducted under IRB approval obtained by Tempus Labs, with all patients involved providing informed consent for the use of theit blood samples for biomarker research.

e15529 Background: Excluding skin cancers, colorectal cancer is the third most common cancer diagnosed in both men and women in the United States. Colorectal cancer (CRC) affects men and women of all racial and ethnic groups and is most often found in people who are 50 years old or older. To aid diagnosis and improve screening for CRC, this study focuses on identifying glycoprotein biomarkers using blood serum. Methods: Novel methods including liquid-chromatography/mass-spectrometry (LC-MS) with in-house peak integration software PB-Net were used to identify glycoprotein biomarkers by analyzing blood serum. Samples were sourced from different biorepositories including 245 CRC, 38 adenoma and 196 healthy controls. The data were split into 75% training and 25% hold-out test set for multivariable predictions. Statistical analysis was performed on normalized data to identify potential biomarkers differentiating adenoma and different stages of CRC samples from the healthy controls. Results: There were 419 significantly differentially expressed glycopeptides/peptides from comparisons between CRC and adenoma samples against the healthy control samples with an FDR < 0.05. A subset of these biomarkers were assessed, generating a 21-biomarker multivariable classifier model. We observed a test set AUC of 0.926, and the sensitivity for all stages of CRC was 90% (87% early stage, 92% late stage). Notably, sensitivity for adenomas was 79%, a large improvement upon the state of the art in adenoma diagnosis. Conclusions: Identification of these key glycopeptides/peptides in blood serum could prove to be a promising non-invasive diagnostic tool that can help improve screening and aid in early detection of advanced adenomas and CRC.

Introduction: While immune checkpoint inhibitor (ICI) therapy has added a powerful new arsenal of highly effective drugs for the subset of cancer patients who respond to these agents, their use remains burdened by the fact that we lack reliable biomarkers to identify likely responders, to avoid the adverse event incidence and cost of treating likely non-responders. Likewise, we currently have no tools which would help identify the optimal choice of agents; current prescribing practice is not guided by any objective criteria. We have recently demonstrated that interrogating the serum glycoproteome, using a proprietary platform that couples artificial intelligence to targeted liquid chromatography-mass spectrometry yields highly informative biomarkers for a range of use cases, including prediction of response to ICI treatment. We recently demonstrated this for metastatic malignant melanoma (MM). In the current study, we examined if we could also predict preferential response to individual ICIs. Experimental Procedures: We carried out glycoproteomic analysis of pretreatment blood samples in advanced MM patients treated with pembrolizumab (P; n=24) or nivolumab-ipilimumab (N; n=11). Individual glycopeptide (GP)signatures derived from 67 serum proteins were analyzed and correlated with treatment, and progression-free survival (PFS). Summary of New Data: Two response groups were defined based on PFS: early failures (EF; PFS event within 6 months) and sustained control (SC; no events for ≥ 12 months). Differential relative abundances for 498 serum GPs were calculated between SC and EF patients to determine GPs more abundant in SC vs. EF by treatment group. A score was developed for each treatment group based on the 20 GPs within each treatment group identified as most statistically significant (one-sided Wilcoxon test). For any patient, the score is the proportion of GPs with relative abundance exceeding their median abundance. A low score is associated with high risk for EF. When examined in all patients in the cohort (regardless of treatment), both scores isolated EF from SC. Only 2 glycopeptides overlapped between the treatment group scores, suggesting that the information is indeed drug specific. Algorithmic assignment was performed by choosing the treatment with the highest treatment-specific score (e.g., if N- score > P-score, then assign to N). PFS was superior for cases where the assigned treatment matched the treatment received. Log-rank p-values comparing PFS by assigned treatment within P- and N -treated cases were 0.009 and 0.0004, respectively. Conclusions: Our results show that a proprietary serum glycoproteomic analytical approach can guide individualized treatment assignment to the most likely successful agent among different ICIs. This could importantly improve the clinical use of immuno-therapy. Citation Format: Klaus Lindpaintner, Gege Xu, Rachel Rice, Alan Mitchell, Dennie Frederick, Genevieve Boland, Daniel Seie. Glycoproteomics-based liquid biopsy informs optimal checkpoint-inhibitor drug choice [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2022; 2022 Apr 8-13. Philadelphia (PA): AACR; Cancer Res 2022;82(12_Suppl):Abstract nr 1270.

e12545 Background: Breast cancer is the most common cancer among women worldwide.Traditional methods of cancer detection such as tissue biopsy are invasive, costly, time consuming and not amenable for repetition. As a result, minimally invasive liquid biopsies, especially blood-based biomarkers show potential value for breast cancer risk prediction and early detection. In this study, we investigated the use of serum glycoproteins circulating in blood to identify a panel of potential prognostic markers that may aid in predicting breast cancer in women. Methods: We applied a novel platform for characterizing blood glycoproteomic biomarkers, combining liquid-chromatography/mass spectrometry (LC-MS) with artificial intelligence/neural networks (AI-NN) to analyze serum samples from 279 breast cancer patients (median age 56 years, with stage 0-4 N’s: 1 / 83 / 114 / 56 / 25) and 102 healthy control samples (median age 52 years). A panel of 596 serum glycosylated and non-glycosylated peptides, representing 71 serum proteins, were analyzed. Age-adjusted differential expression analysis for 596 normalized biomarkers were performed to evaluate statistically significant differential abundances using an FDR q-value of 0.05 as a cutoff. Using the top differentially expressed markers as input, a LASSO penalized logistic regression model with 5-fold repeated cross validation was applied to identify the top biomarkers contributing to the separation between healthy controls and breast cancer patients. Results: We identified 243 out of 596 markers that were differentially expressed (FDR <<0.05) between breast cancer samples and healthy controls. Out of those, 11 markers were obtained as the top predictors in classifying breast cancer patients and healthy controls. The classification algorithm yielded an accuracy of 94% (95.9% sensitivity, 88.7% specificity) and an AUC of 0.983 on the training set. This classifier was validated on an independent test set with 30% of the subjects, yielding an accuracy of 93% (96.4% sensitivity, 83.9% specificity) and an AUC of 0.974. Test sensitivity was high across stages, at 96% / 90% / 95% / 90% in stages 1-4, respectively. Conclusions: Based on the results, we conclude that circulating glycoproteins in serum may be useful in screening applications in breast cancer, and strongly demonstrates the utility of glycoprotein profiles as a powerful non-invasive diagnostic tool.