Lead-Time Trajectory of CA19-9 as an Anchor Marker for Pancreatic Cancer Early Detection

Fahrmann, Johannes F.; Schmidt, C. Max; Mao, Xiangying; Irajizad, Ehsan; Loftus, Maureen; Zhang, Jinming; Patel, Nikul; Vykoukal, Jody; Dennison, Jennifer B.; Long, James P.; Anh, Kim; Zhang, Jianjun; Chabot, John A.; Kluger, Michael D.; Kastrinos, Fay; Brais, Lauren K.; Babić, Ana; Jajoo, Kunal; Lee, Linda; Clancy, Thomas E.; Ng, Kimmie; Bullock, Andrea J.; Genkinger, Jeanine M.; Yip-Schneider, Michele; Maitra, Anirban; Wolpin, Brian M.; Hanash, Samir M.

doi:10.1053/j.gastro.2020.11.052

Cited by 104 publications

(106 citation statements)

References 52 publications

Supporting

Mentioning

102

Contrasting

Unclassified

Order By: Relevance

“…There is an additional opportunity to tailor the nature of clinical investigations for different sections of the high risk population. For example, a patient predicted to get cancer within the next 3 to 6 months would immediately be assigned to detailed testing with imaging and blood tests, such as CA19-9, cell free DNA for mutations or changes in methylation patterns (Widschwendter et al 2017; Liu et al 2020; Fahrmann et al 2021). On the other hand, someone projected to get cancer in the two to three year time frame might be assigned to a serial surveillance program, with periodic reassessment of risk.…”

Section: Discussionmentioning

confidence: 99%

Pancreatic cancer risk predicted from disease trajectories using deep learning

Yuan

et al. 2021

Preprint

Self Cite

View full text Add to dashboard Cite

Pancreatic cancer is an aggressive disease that typically presents late with poor patient outcomes. There is a pronounced medical need for early detection of pancreatic cancer, which can be facilitated by identifying high-risk populations. Here we apply artificial intelligence (AI) methods to a large corpus of more than 6 million patient records spanning 40 years with 24,000 pancreatic cancer cases in the Danish National Patient Registry. In contrast to existing methods that do not use temporal information, we explicitly train machine learning models on the time sequence of diseases in patient clinical histories. In addition, the models predict the risk of cancer occurrence in time intervals of 3 to 60 months duration after risk assessment. For cancer occurrence within 12 months, the performance of the best model trained on full trajectories (AUROC=0.91) substantially exceeds that of a model without time information (AUROC=0.81). For the best model, lower performance (AUROC=0.86) results when disease events within a 3 month window before cancer diagnosis are excluded from training, reflecting the decreasing information value of earlier disease events. These results raise the state-of-the-art level of performance of cancer risk prediction on real-world data sets and provide support for the design of real-world population-wide clinical screening trials, in which high risk patients are assigned to serial imaging and measurement of blood-based markers to facilitate earlier cancer detection. AI on real-world clinical records has the potential to shift focus from treatment of late- to early-stage cancer, benefiting patients by improving lifespan and quality of life.

show abstract

Section: Discussionmentioning

confidence: 99%

Pancreatic cancer risk predicted from disease trajectories using deep learning

Yuan

et al. 2021

Preprint

Self Cite

View full text Add to dashboard Cite

show abstract

“…More specifically, 86% and 51% of surgeons would offer upfront resection in patients with CA19-9 >1000U/ml or PS ≥2, respectively. CA19-9 is currently the only available biomarker in clinical practice for PDAC and despite mediocre sensitivity and specificity, high values are suggestive of aggressive tumour biology ( 28 ) and have been associated with early disease recurrence ( 29 ). The study of biological behaviour in PDAC gradually introduced further criteria for resectability beyond anatomic considerations ( 11 , 30 ) aiming at the improvement of recurrence rates and overall survival.…”

Section: Discussionmentioning

confidence: 99%

Surgical Management of Non-Metastatic Pancreatic Cancer in the United Kingdom: Results of a Nationwide Survey on Current Practice

et al. 2021

View full text Add to dashboard Cite

BackgroundIt is presently unclear what clinical pathways are followed for patients with non-metastatic PDAC in specialised centres for pancreatic surgery across the United Kingdom (UK).MethodsBetween August 2019 and August 2020 an electronic survey was conducted aiming at a national cohort of pancreatic surgeons in the UK. Participants replied to a list of standardised questions and clinical vignettes, and data were collected and analysed focusing on management preferences, resectability criteria, and contraindications to surgery.ResultsWithin the study period, 65 pancreatic surgeons from 27 specialist centres in the UK (96%) completed the survey. Multidisciplinary team meetings are utilised universally for the management of patients with PDAC, however, different staging systems for resectability classification are being applied. In borderline resectable PDAC, most surgeons were keen to proceed with surgical exploration post NAT, but differences were noted in preferred chemotherapy regimens. Surgeons from standard volume institutions performed fewer vein resections annually and were more likely to deem patients with locally advanced PDAC as unresectable. Intra-institutional variability in patient management was also present and ranging between 20-80%.ConclusionsSignificant variability in the surgical management of non-metastatic PDAC was identified both on inter- and intra-institutional level.

show abstract

“…Carbohydrate antigen (CA) 19-9, an epitope of sialylated Lewis antigen, appears elevated in PC[ 10 ], and there have been numerous studies on its usefulness for screening, diagnostic, prognostic, and predictive purposes and for assessing resectability[ 11 ]. It has been documented as an early detector of PC[ 12 ], and it is the only biomarker approved for monitoring its progression and the therapeutic response[ 13 ]. However, some shortcomings have been reported, including poor sensitivity (69%-98%) and specificity (46%-98%) for PC diagnosis[ 14 ].…”

Section: Carbohydrate Antigen 19-9 Other Carbohydrate Markers and Carcinoembryonic Antigenmentioning

confidence: 99%

Liquid biopsy approach to pancreatic cancer

Perales

Torres

Jiménez-Luna

et al. 2021

WJGO

View full text Add to dashboard Cite

Pancreatic cancer (PC) continues to pose a major clinical challenge. There has been little improvement in patient survival over the past few decades, and it is projected to become the second leading cause of cancer mortality by 2030. The dismal 5-year survival rate of less than 10% after the diagnosis is attributable to the lack of early symptoms, the absence of specific biomarkers for an early diagnosis, and the inadequacy of available chemotherapies. Most patients are diagnosed when the disease has already metastasized and cannot be treated. Cancer interception is vital, actively intervening in the malignization process before the development of a full-blown advanced tumor. An early diagnosis of PC has a dramatic impact on the survival of patients, and improved techniques are urgently needed to detect and evaluate this disease at an early stage. It is difficult to obtain tissue biopsies from the pancreas due to its anatomical position; however, liquid biopsies are readily available and can provide useful information for the diagnosis, prognosis, stratification, and follow-up of patients with PC and for the design of individually tailored treatments. The aim of this review was to provide an update of the latest advances in knowledge on the application of carbohydrates, proteins, cell-free nucleic acids, circulating tumor cells, metabolome compounds, exosomes, and platelets in blood as potential biomarkers for PC, focusing on their clinical relevance and potential for improving patient outcomes.

show abstract

Lead-Time Trajectory of CA19-9 as an Anchor Marker for Pancreatic Cancer Early Detection

Cited by 104 publications

References 52 publications

Pancreatic cancer risk predicted from disease trajectories using deep learning

Pancreatic cancer risk predicted from disease trajectories using deep learning

Surgical Management of Non-Metastatic Pancreatic Cancer in the United Kingdom: Results of a Nationwide Survey on Current Practice

Liquid biopsy approach to pancreatic cancer

Contact Info

Product

Resources

About