Melton and Genta from Caris Research Institute and UT-Southwestern Medical Center in Dallas have published an excellent review of these uncommon lesions in the December 2010 American Journal of Surgical Pathology that is worthy of note for the practicing pathologist, in spite of their relative rarity.
First, I emphasize that these are rare but I think if you sign out a lot of GI biopsies, especially from outpatient endoscopy centers, you will (or have) come across these lesions in practice. Caris Life Sciences is a GI pathology subspecialty referral lab that receives specimens from such endoscopy centers from across the U.S. This is one of the strengths of the paper--the huge denominator of GEJ biopsies for comparison. They diagnosed a cardiac polyp in 330 patients out of 120,817 patients with distal or junctional esophageal biopsies--0.27% of all biopsies.
Second, usually a nodule or polyp is noted by endoscopy and may be noted by the endoscopist as a "sentinel fold." In this series, a nodule or polyp was noted during endoscopy in 60% of patients diagnosed with cardiac polyps.
Characteristic histologic features: prominent (polypoid) foveolar hyperplasia, edematous stroma with mild or no chronic or active inflammation, mucosal erosion/ulceration in about 50%.
Uncommonly associated with gastric hyperplastic polyps (10/241) by slightly more than three times more common compared to patients without cardiac polyps.
Inversely associated with: active esophagitis/GERD, Barrett mucosa, fundic gland polyp, Helicobacter gastritis.
I note this study because it most likely reflects experience in routine general practice (assuming your local GI group endoscopy center doesn't send their specimens to Caris : ) ) as opposed to the usual descriptive studies from academic centers that reflect referral bias and it provides a simple description with excellent images of an uncommon but diagnostic entity.