Accurately Assess Lymph Node Status

USCAP 2020 Abstract

The Utility of Acetone and Automated Compressive Filtration (ACF) in Lymph Node (LN) Procurement: Interobserver Variability

Elias Makhoul DO, Brian K Cox MD, Bonnie L. Balzer MD PhD Elena Chang MD, David Frishberg MD, Maha Guindi MD, Richard Mertens MD PhD, Mary Wong MD, Kevin M. Waters MD PhD


Lymph node (LN) recovery from gross examination and the correct analysis of LN status are necessary for accurate staging after colorectal cancer resection. The use of acetone and Automated Compressive Filtration (ACF) of pericolic adipose tissue (PAT) has been identified as a novel technique to yield more LNs. We aim to investigate the effect on interobserver variability in identifying lymph nodes with this novel technique.


We utilized a case-control study design consisting of 40 total PAT slides from archival colon resection cases for colorectal carcinoma. The control cohort (N=20) was comprised of representative sections of manually compressed PAT that were submitted after the initial LN search yielded fewer than 12 LNs. The comparison cohort consisted of representative sections of PAT (N=20) that underwent ACF. All 40 slides were blinded to preparation technique and randomized. Eight pathologists counted the number of LNs on each slide. LNs were defined as lymphoid tissue with at least one of the following structures: a capsule, LN sinus, or hilar vessels. To investigate whether morphology of large grossly positive-appearing LN would alter due to ACF, a large LN from PAT underwent ACF and morphology was assessed. Fleiss' kappa was run to determine agreement.


There was moderate overall interobserver agreement in the assessment of LN in the control cohort, κ=0.694 (95% CI=0.652, 0.735). PAT that underwent ACF also showed moderate overall agreement, κ=0.639 (95% CI=0.599, 0.678). Kappas for the control cohort were consistently moderate while the ACF cohort showed greater agreement when lymph node counts were 0-1. Both the case and control cohorts showed less agreement with an increasing number of lymph nodes. No alteration of morphology or increased fragmentation was noted in the ACF slides. Also, a grossly positive large LN (1.6 x 1.0 x 0.8 cm) that underwent ACF did not fragment and maintained normal histology when compared to FFPE H&E stained LN.


There was essentially no significant difference in interobserver agreement between lymph node counts in manually versus automated compressed PAT with each group displaying moderate agreement. The enlarged LN that underwent ACF did not fragment or have altered morphology. We found no evidence that ACF negatively impacts LN count in colon resections specimens.

to AdiPress
Book Now

Want to see AdiPress in action?

Thank you! Your submission has been received!
Oops! Something went wrong while submitting the form.
CTA Image