PROTEIN N-GLYCOSYLATION TRAITS CAN ACCURATELY DISTINGUISH PANCREATIC CANCER CASES FROM A HETEROGENEOUS GROUP OF CONTROLS

UEG Poster

Standard Poster

Log in to access this content.

Free for all myUEG account holders. Your access level is set automatically based on your occupation. Medical professionals get full access to all content. If you are a non-medical user, you can only access UEG Week content from congresses you attended.

Log In Create a free account

Not sure what you can access? Learn more about account types.

Introduction

New methods are urgently needed to detect pancreatic ductal adenocarcinoma (PDAC) at an earlier stage to improve outcomes.^{1, 2} We have previously reported six protein N-glycosylation traits (NGTs) and further grouped these into three biologically distinct ones to discriminate PDAC patients from healthy controls with an area under the curve (AUC) of 0.81-0.88.³ A subsequent longitudinal study revealed glycosylation differences up to 50 months prior to diagnosis, indicating potential for early detection.⁴ However, external validation is needed to determine accuracy of this biomarker panel. Moreover, it is unknown whether this panel can differentiate PDAC in a diverse group that resembles clinical practice better. Therefore, this study aims to validate the biomarker panel in a heterogeneous cohort.

Aims & Methods

Protein N-glycans in prospectively collected plasma samples from Mayo Clinic were analyzed by mass spectrometry at the Leiden University Medical Center using a semi-automated preparation protocol.⁵ Glycosylation profiles were determined and compared between PDAC cases and controls. The control group consisted of imaging-confirmed pancreatic disease-negative controls and individuals with benign pancreatic diseases, including (non)-mucinous cysts and pancreatitis. Logistic regression was performed to calculate odds ratios (ORs) for the six previously reported protein NGTs. Additionally, AUC, sensitivity and specificity were calculated to determine the accuracy of the three biologically distinct NGTs (antennarity (CA4), sialylation (A3F0L), and fucosylation (CFa)) in discriminating PDAC cases from controls.

Results

We included 232 individuals, of whom 42 (18.1%) were diagnosed with PDAC, and 190 (81.9%) served as controls. Among controls, 59 (31.1%) were pancreatic disease-negative and 131 (68.9%) had benign pancreatic diseases. Median age of PDAC cases and controls was 71.7 (IQR=14.5) and 65.5 years (IQR=17.6), respectively. Among PDAC cases, 19 (45.2%) were female, and 106 (52.2%) among controls. The BMI in PDAC cases and controls was equal (26.8 (IQR=7.0) and 26.6 (IQR=7.6), respectively). Table 1 shows ORs for the six NGTs and PDAC. The AUC for the three biologically distinct traits was 0.77 (95% CI: 0.69-0.86), with a sensitivity of 0.81 (95% CI: 0.69-0.93) and specificity of 0.70 (95% CI: 0.64-0.76). When comparing PDAC cases to pancreatic disease-negative controls, the AUC was 0.83 (95% CI: 0.74-0.92), with a sensitivity of 0.79 (95% CI: 0.64-0.90) and specificity of 0.80 (95% CI: 0.69-0.90).

Table 1: The odds ratios for six glycosylation traits distinguishing pancreatic ductal adenocarcinoma cases from a heterogeneous cohort, including pancreatic disease-negative and benign pancreatic diseases controls. A subanalysis was performed for disease-negative controls. The Bonferroni correction was used to account for multiple testing.
Abbreviations: PDAC= pancreatic ductal adenocarcinoma, CI= confidence interval.
	PDAC compared to Heterogeneous cohort		PDAC compared to pancreatic disease-negative cohort
Variables	Odds ratio (95% CI)	P-value	Odds ratio (95% CI)	P-value
CA2	0.90 (0.85-0.95)	<0.001	0.86 (0.79-0.93)	0.001
CA4	1.59 (1.30-1.97)	<0.001	2.08 (1.49-3.10)	<0.001
CFa	1.76 (1.28-2.47)	0.004	3.40 (1.86-7.15)	0.002
A3F0L	0.91 (0.84-0.97)	0.043	0.82 (0.72-0.91)	0.005
A3F0E	1.17 (1.07-1.29)	0.004	1.37 (1.18-1.64)	0.001
A3FE	1.90 (1.39-2.66)	0.001	2.58 (1.69-4.23)	<0.001

Conclusion

Distinctive NGTs from plasma protein differentiate PDAC from non-PDAC in a diverse group, consisting of cystic pancreatic diseases and pancreatitis. This panel holds potential for future early detection methods.

References

1. Klatte DCF, Wallace MB, Löhr M, Bruno MJ, van Leerdam ME. Hereditary pancreatic cancer. Best Pract Res Clin Gastroenterol. 2022;58-59:101783.
2. Goggins M, Overbeek KA, Brand R, Syngal S, Del Chiaro M, Bartsch DK, et al. Management of patients with increased risk for familial pancreatic cancer: updated recommendations from the International Cancer of the Pancreas Screening (CAPS) Consortium. Gut. 2020;69(1):7-17.
3. Vreeker GCM, Hanna-Sawires RG, Mohammed Y, Bladergroen MR, Nicolardi S, Dotz V, et al. Serum N-Glycome analysis reveals pancreatic cancer disease signatures. Cancer Med. 2020;9(22):8519-29.
4. Levink IJM, Klatte DCF, Hanna-Sawires RG, Vreeker GCM, Ibrahim IS, van der Burgt YEM, et al. Longitudinal changes of serum protein N-Glycan levels for earlier detection of pancreatic cancer in high-risk individuals. Pancreatology. 2022;22(4):497-506.
5. Vreeker GCM, Nicolardi S, Bladergroen MR, van der Plas CJ, Mesker WE, Tollenaar R, et al. Automated Plasma Glycomics with Linkage-Specific Sialic Acid Esterification and Ultrahigh Resolution MS. Anal Chem. 2018;90(20):11955-61.