67 Appraisal of Follow-up Protocols and Outcomes of High Risk Pancreatic Cancer Surveillance Programs: A Systematic Review and Meta-Analysis

Abstract

INTRODUCTION: Individuals with familial or genetic predisposition are at high risk of pancreatic cancer (PC) and undergo screening. After baseline imaging, these high-risk individuals (HRIs) undergo regular follow-up through cross-sectional imaging, endoscopic ultrasound (EUS), or a combination of these modalities. The follow-up of HRIs is largely based on guidelines, and the evidence and consensus for optimal imaging modality and screening intervals are scarce. The current study is the first systematic review appraising various follow-up strategies among HRIs screened for PC for the detection of new lesions and progression of preexisting abnormalities. METHODS: Prospective studies on PC surveillance were analyzed to determine the imaging intervals and predominant modality-Cross sectional imaging: CT or MRI, or Combined EUS and cross sectional imaging. The outcomes of interest included incidence of new lesions where baseline was normal and progression of screen-prevalent lesions, defined as pre-existing abnormalities at baseline imaging. Follow-up was quantified as per 1000 patient-years (pt-yrs). A random effect model was used to calculate cumulative effect measures. The systematic review was performed in adherence with PRISMA guidelines and a preregistered protocol (Figure 1). RESULTS: 9 uncontrolled studies with 1438 subjects and 6274 pt-yrs of follow-up were analyzed; protocols included cross-sectional imaging (n = 4,4709 pt-yrs), or combined-EUS and cross sectional imaging (n = 5,1565 pt-yrs). HRIs with normal baseline were screened every 12 months, pooled radiologic incidence of 24.74 per 1000 pt-yrs [8.44–48.60, P < 0.01] and diagnostic rate of advanced pathology of 1.06 per 1000 pt-yrs [0–3.99, P < 0.08] (Figure 2). HRIs with screen-prevalent lesions were followed every 3–6 months and longer when stable, with cumulative progression rate of 19.47 per 1000 pt-yrs [1.05–52.98, P < 0.01], and that of advanced pathology-1.54 per 1000 pt-yrs [0–7.55, P < 0.01] (Figure 3). Subgroup analysis showed no significant difference in radiologic incidence or progression or in the histologic diagnosis of advanced pathologies between cross-sectional and combined follow-up protocols. CONCLUSION: Surveillance protocols exhibited significantly higher radiologic than histologic diagnostic rates. Though heterogeneity existed between studies, there was no significant difference in the diagnostic rates of incidence or progression between protocols employing predominately cross-sectional vs combined protocols of cross-sectional imaging and EUS.