Pancreatic Cancer Surgical Resection Margins: Molecular Assessment by Mass Spectrometry Imaging

Abstract

Background: Surgical resection with microscopically negative margins remains the main curative option for pancreatic cancer; however, in practice intraoperative delineation of resection margins is challenging. Ambient mass spectrometry imaging has emerged as a powerful technique for chemical imaging and real-time diagnosis of tissue samples. We applied an approach combining desorption electrospray ionization mass spectrometry imaging (DESI-MSI) with the least absolute shrinkage and selection operator (Lasso) statistical method to diagnose pancreatic tissue sections and prospectively evaluate surgical resection margins from pancreatic cancer surgery.

Methods and findings: Our methodology was developed and tested using 63 banked pancreatic cancer samples and 65 samples (tumor and specimen margins) collected prospectively during 32 pancreatectomies from February 27, 2013, to January 16, 2015. In total, mass spectra for 254,235 individual pixels were evaluated. When cross-validation was employed in the training set of samples, 98.1% agreement with histopathology was obtained. Using an independent set of samples, 98.6% agreement was achieved. We used a statistical approach to evaluate 177,727 mass spectra from samples with complex, mixed histology, achieving an agreement of 81%. The developed method showed agreement with frozen section evaluation of specimen margins in 24 of 32 surgical cases prospectively evaluated. In the remaining eight patients, margins were found to be positive by DESI-MSI/Lasso, but negative by frozen section analysis. The median overall survival after resection was only 10 mo for these eight patients as opposed to 26 mo for patients with negative margins by both techniques. This observation suggests that our method (as opposed to the standard method to date) was able to detect tumor involvement at the margin in patients who developed early recurrence. Nonetheless, a larger cohort of samples is needed to validate the findings described in this study. Careful evaluation of the long-term benefits to patients of the use of DESI-MSI for surgical margin evaluation is also needed to determine its value in clinical practice.

Conclusions: Our findings provide evidence that the molecular information obtained by DESI-MSI/Lasso from pancreatic tissue samples has the potential to transform the evaluation of surgical specimens. With further development, we believe the described methodology could be routinely used for intraoperative surgical margin assessment of pancreatic cancer.

Fig 1. Flowchart summarizing the study performed.

Fig 2. Representative negative ion mode DESI-MSI mass spectra obtained from pancreatic tissue.

(A) Normal pancreatic glandular tissue from sample PC775 (note 10× zoom applied from m/z 700 to m/z 900 to assist visualization); (B) pancreatic cancer tissue from sample PC13702; and (C) normal pancreatic stromal tissue from sample PC0423.

Fig 3. Selected 2-D negative ion mode DESI-MSI ion images obtained from pancreatic tissue samples.

(A) Ion images of m/z 281.3 (oleic acid), m/z 303.3 (arachidonic acid), m/z 788.5 (PS(36:1)), m/z 861.6 (PI(36:2)), and m/z 885.6 (PI(38:4)) for samples PC7817, which is composed of 90% normal pancreatic glands and 10% normal stromal tissue, sample PC13702, which is composed of 90% pancreatic cancer and 10% normal stromal tissue, and sample PC0423, which is entirely composed of normal stroma. (B) Optical images of the same tissue sections subjected to H&E stain and evaluated by histopathology.

Fig 4. Selected DESI-MSI ion image and Lasso prediction results obtained for training samples PC5756 and PC699 and validation samples PC14836 and PC13702.

(A) Negative ion mode DESI-MSI ion images of m/z 303.2, m/z 885.7, m/z 281.3, and m/z 281.3 are shown for samples PC5756, PC699 (pair of normal pancreas and pancreatic cancer), PC14836, and PC13702, respectively. (B) Lasso prediction results are shown for each sample, with pixels predicted as pancreatic cancer shown in red, normal pancreatic glands in green, and normal stroma in blue. (C) Optical images of the H&E-stained tissue sections are shown, with the regions diagnosed by the pathologist delineated using the same color representation.

Fig 5. Statistical predictions in comparison with histopathologic diagnosis for samples of mixed histology.

Results are shown for samples PC13336, PC12809, PC14851, PC0411, and PC14132 using (A) the Lasso classification system built with the training set of samples and (B) the customized training set approach. Pixels predicted as pancreatic cancer are shown in red, normal pancreatic glands in green, and normal pancreatic stroma in blue. (C) Optical images of the H&E-stained tissue sections are shown, with the regions diagnosed by the pathologist delineated using the same color representation.

Fig 6. Lasso prediction results for cancer and margin tissues obtained prospectively from surgery in pancreatic cancer patients.

Lasso prediction results are shown for the cancer tissue as well as the surgical margins (uncinate and/or neck) of surgical cases PCP4 (negative margins), PCP9 (negative margin), PCP21 (negative margin), and PCP14 (positive margin). Pixels predicted as pancreatic cancer are shown in red, normal pancreatic glands in green, and normal stroma in blue.

Fig 7. Kaplan-Meier curves comparing overall survival after surgical resection between patients with negative surgical margins by both DESI-MSI/Lasso and frozen section analysis and patients with positive margins by DESI-MSI/Lasso but negative margins by frozen section analysis.

The blue line indicates patients with negative surgical margins by both DESI-MSI/Lasso and frozen section analysis (n = 23, median 26 mo); the green line indicates patients with positive margins by DESI-MSI/Lasso but negative margins by frozen section analysis (n = 8, median 10 mo). Log-rank test, p = 0.209. One patient in the prospective arm of the study had positive margins by both techniques and was excluded from this comparison.