Peptic Ulcer Disease and H. pylori: NSAID Ulcers, Refractory Disease, and Perforation

0:00Welcome to Board Pearls. This is episode two of two of the Peptic Ulcer Disease and H. pylori chapter, in the Stomach and Small Bowel Disorders module. This episode is the ulcers that aren't from H. pylori: the NSAID and aspirin ulcers and how you prevent them, the refractory and idiopathic ulcers where you have to figure out what was missed, and the perforation and penetration emergencies that split by whether the leak is free or contained.

0:26NSAIDs damage the stomach and duodenum two ways at once, and the one that matters for prevention is the systemic one. The COX-1 enzyme in the gut lining makes the prostaglandins that drive mucus, bicarbonate, and blood flow, the three defenses that let the surface neutralize acid, and when an NSAID blocks COX-1 all three fail together, so the mucosa can't buffer acid at its own surface and erosions appear. Because that mechanism is systemic, a suppository or an injection causes ulcers just as readily as a pill; the local contact injury from a swallowed tablet is layered on top but isn't the main event. Aspirin does both of those and adds a permanent block of platelet function, which is why even a low daily aspirin keeps a bleeding risk for the life of each platelet.

1:12The risk factors are the real teaching point, because they don't add, they multiply. The list runs through age over sixty-five, a prior ulcer especially a prior bleed or perforation, high-dose or more-than-one NSAID, the first few months of therapy before the mucosa adapts, and then concurrent steroids, anticoagulants, other antiplatelets including aspirin and clopidogrel, SSRIs, and coexisting H. pylori. Each roughly doubles the risk on its own, and stacking them multiplies rather than sums, so a young person on two weeks of ibuprofen sits at baseline while an older patient with a prior bleeding ulcer on a full-dose NSAID plus an antiplatelet plus a steroid is in another world entirely. The tiers are simple: high risk is a prior complicated ulcer or more than two factors, moderate is one or two, and low is none.

2:08The H. pylori interaction deserves its own note because the two injure the lining by different routes, the bacterium through inflammation and the NSAID through prostaglandin depletion, so together the bleeding-ulcer risk is markedly higher than either alone. The intervention follows the mechanism: eradicating H. pylori before starting chronic NSAIDs lowers the ulcer rate over the following year, and that protection is on top of whatever an acid blocker adds, which is exactly why long-term NSAID and aspirin use are now on the test-and-treat list.

2:43Co-prescribing an acid blocker is what actually prevents most of these ulcers, and the threshold follows the tier. High-risk patients shouldn't take a non-selective NSAID without a proton pump inhibitor, continued as long as the NSAID. Moderate-risk patients on long-term NSAIDs get one. Low-risk patients without another reason for acid suppression generally don't. Misoprostol is mechanistically the cleanest, since it replaces exactly the prostaglandin the NSAID suppressed, but its diarrhea is intolerable for most and it can't be used in women who might become pregnant, so the proton pump inhibitor wins in practice. Standard-dose histamine blockers are inadequate and are the wrong answer for NSAID prevention.

3:28The COX-2 selective drugs change the math when chronic anti-inflammatory therapy is truly needed. Celecoxib spares COX-1 and so spares the stomach's prostaglandin defense, dropping the GI complication rate, but the trade is cardiovascular risk, so it's not a blanket replacement; it earns its place in the patient who has bled from an NSAID ulcer and still needs an anti-inflammatory, where the answer is a COX-2 drug plus an acid blocker together. And there's a trap: a patient also on aspirin loses much of that advantage, because aspirin itself blocks COX-1 and restores most of the risk celecoxib was meant to remove, so celecoxib plus aspirin with any additional risk factor brings the acid blocker back. For someone on long-term aspirin alone, use the lowest effective dose and add an acid blocker when a second risk factor raises the bleeding probability.

4:22The dual-antiplatelet patient after a stent is a recurring case: aspirin plus a second antiplatelet generates real upper-GI bleeding risk that climbs with prior ulcers, older age, anticoagulation, or NSAIDs on top, and an acid blocker cuts that bleeding risk substantially without compromising the cardiac protection. The theoretical interaction between proton pump inhibitors and clopidogrel hasn't produced real cardiovascular harm in prospective trials, so the right answer is to give the acid blocker alongside the dual antiplatelet therapy, and the same logic gets an anticoagulated patient with a prior ulcer bleed safely back on their anticoagulant. In fact, after a high-risk ulcer bleed in someone who needs to stay anticoagulated, the plan has three parts: resume the anticoagulant after about a week or two once bleeding is stable, continue long-term acid suppression, and retest and eradicate H. pylori, which together drive recurrence low.

5:19A non-healing ulcer at the follow-up scope is a different problem, because the usual explanations have been addressed and the job is to find what was missed, and there's an order to it. The first thing to suspect is a falsely negative H. pylori test, because the patient is almost always still on an acid blocker, which suppresses the bacteria across all the enzyme-based tests; histology on a body biopsy is the most resilient but even that is better off the drug, so the right move is to retest off the acid blocker for one to two weeks and off antibiotics or bismuth for four, bridging symptoms with a histamine blocker if needed. A negative test done the day after the scope on a patient who's been on a proton pump inhibitor for weeks isn't really a negative test, it's a deferred one. The second thing to find is hidden NSAID or aspirin use, because patients don't report over-the-counter ibuprofen, the aspirin they take on their own, or the naproxen buried in a cold remedy, so you ask about each by name, and many "idiopathic" ulcers become NSAID ulcers on the second pass. The third is non-adherence, both to the acid blocker, which loses much of its effect taken with food instead of before a meal, and to the eradication regimen, which is easy to miss doses on, so refill records and a frank conversation matter.

6:36Once H. pylori, NSAIDs, and adherence are excluded, the differential widens to the things that imitate ordinary ulcer disease, and cancer is the one you cannot miss. Gastric adenocarcinoma can present as a chronic non-healing gastric ulcer that fails acid suppression, and most missed cancers are sampling error rather than truly absent disease, which drives a firm rule: every gastric ulcer gets a follow-up scope at eight to twelve weeks to confirm healing, and every non-healing one gets repeat biopsies, many of them, from both the rim and the base, because the first set may have missed the lesion. Duodenal ulcers don't need that routine follow-up because primary cancer at a duodenal ulcer is rare, and that contrast is itself testable. Lymphoma sits next to adenocarcinoma, so biopsies go for the appropriate staining when it's in the picture. Crohn's of the stomach or duodenum makes ulcers that look peptic but fail acid suppression, with clues like multiple aphthous lesions, ileal disease on imaging, granulomas, or a perianal fistula, and the answer there is the inflammatory bowel workup, not more acid suppression. Vasculitis can make gastric ulcers as part of a systemic disease, with the rim biopsy showing it. CMV ulcers belong to the immunocompromised, with the owl-eye inclusions on deep biopsy. Eosinophilic gastritis makes ulcers with heavy mucosal eosinophils on biopsy. And drugs besides NSAIDs, potassium chloride, bisphosphonates, some transplant drugs, cocaine, cause ulcers too, so a careful medication review matters.

8:09Zollinger-Ellison syndrome is the highest-yield rare cause, and the recognition pattern matters more than the workup: multiple ulcers, ulcers in unusual spots like the distal duodenum or jejunum, ulcers that recur despite proven eradication and no NSAIDs, an ulcer with severe esophagitis, an ulcer with secretory diarrhea, or a personal or family history of MEN1. The mechanism is gastrin oversecretion from a neuroendocrine tumor overstimulating the acid-making cells. The screen is a fasting gastrin, and it has two pitfalls that are themselves the test point. The first is the acid blocker, which raises gastrin by feedback, so a true diagnostic level needs the drug held for at least a week, bridged with a histamine blocker if tolerable. The second is low stomach acid: autoimmune atrophic gastritis destroys the acid-making cells, and the resulting low acid drives a compensatory gastrin rise that mimics the syndrome. Gastric pH tells them apart, because true gastrinoma drives high acid and a fasting pH under two, while atrophic gastritis produces low acid and a pH above four. So a very high fasting gastrin with a pH under two confirms it; an intermediate gastrin needs a secretin stimulation test, where secretin paradoxically pushes the gastrin up in gastrinoma. Imaging is a DOTATATE PET with endoscopic ultrasound for small tumors, and you screen for MEN1. The full workup is in chapter seven.

9:37Only after all of that is excluded, a correctly done negative H. pylori retest off the drug, no hidden NSAIDs, repeat biopsies clearing malignancy, Crohn's and the rest ruled out, and the gastrinoma workup negative, do you call an ulcer idiopathic. Truly idiopathic ulcers exist, recur more and complicate more than H. pylori ulcers, and get long-term acid suppression with a low threshold to re-investigate.

10:03Perforation is the surgical emergency, and imaging answers the question that drives management: is the leak free into the abdomen or contained against an adjacent organ. A free perforation is the ulcer eroding through the whole wall so contents spill into the peritoneum, giving immediate chemical peritonitis that becomes bacterial within hours, and it presents as sudden, severe, generalized pain the patient can time to the minute, quickly developing rigidity, rebound, and guarding with signs of sepsis. Free air under the diaphragm on an upright chest film confirms it in most cases, but that film misses a meaningful share, so CT is the modern test of choice, since it finds the perforation, localizes the leak, distinguishes contained from free, and rules out mimics. Time to surgery drives mortality, so resuscitation runs in parallel with calling the surgeon: fluids, broad-spectrum antibiotics covering gut and oral flora such as piperacillin-tazobactam or ceftriaxone plus metronidazole, nasogastric decompression, and an intravenous acid blocker, and delay beyond a day sharply raises mortality, especially in the elderly and the septic.

11:14The operation is the Graham omental patch, and the board expects the name and the concept: a piece of living, vascularized omentum is laid over the hole and secured, sealing it with healthy tissue rather than trying to sew together an inflamed, friable ulcer rim that would tear out under tension. Open and laparoscopic approaches are both standard. Definitive ulcer surgery like vagotomy isn't added at that operation anymore, because acid suppression and eradication handle the underlying disease afterward and piling a bigger operation onto a septic patient adds risk without benefit. Postoperatively it's a high-dose acid blocker, H. pylori testing and eradication once stable, stopping NSAIDs, and a follow-up scope to confirm healing.

11:58Contained perforation and penetration run differently, because an adjacent structure has already sealed the leak and the patient isn't in free peritonitis. Penetration is the ulcer boring into a neighboring organ instead of the peritoneum, classically a posterior duodenal ulcer into the pancreas, and the tell is that the pain changes character, from intermittent and meal-related to constant and radiating to the back, with the lipase rising from local pancreatic inflammation so the whole thing mimics acute pancreatitis. What breaks the tie is the absence of free air on CT alongside the inflammatory change next to the duodenal ulcer. Management of penetration is medical, high-dose acid suppression, eradication, stopping NSAIDs, and observation, with surgery reserved for medical failure, abscess, a fistula, or instability.

12:51Endoscopic closure is now an option for selected non-free perforations, especially small ones caught during a therapeutic scope, using clips for small defects and larger over-the-scope clips or suturing for bigger ones, but only in a stable patient with no diffuse peritonitis, ideally recognized at the moment it happens, by an experienced operator with surgery immediately available. And surgical consultation accompanies any perforation whatever the plan, because a patient who deteriorates during endoscopic or medical management has to be escalated fast. A small group of stable patients with a sealed, contained leak and no peritonitis or shock can be managed without surgery on bowel rest, decompression, fluids, an acid-blocker infusion, and antibiotics, and many carefully selected patients heal, but the threshold to convert to a Graham patch is low: any peritonitis, any hemodynamic deterioration, or imaging that progresses.

13:45So the non-H.-pylori side of ulcer disease runs on three ideas. NSAID and aspirin ulcers run on multiplicative risk and acid-blocker co-prescription, with the celecoxib-plus-aspirin trap and the dual-antiplatelet patient as the recurring tests. Refractory and idiopathic ulcers run on a sequential exclusion, the falsely negative H. pylori test on an acid blocker first, hidden NSAIDs second, repeat biopsy of any non-healing gastric ulcer third, then the gastrinoma workup, Crohn's, CMV, and the rest. And perforation runs on the free-versus-contained distinction, free perforation to a Graham omental patch and pancreatic penetration to medical management.

14:31Chapter seven picks up the gastric story from the other side: chronic gastritis separating into autoimmune, H. pylori, and Menetrier, the gastrinoma returning in full, and the gastric tumors from adenocarcinoma through MALT lymphoma to the GIST.

14:47For the full chapter, the practice vignettes, and the topic-tagged question bank, head to board pearls dot com. You'll find the rest of the series on Apple Podcasts, Spotify, or wherever you listen to podcasts. That brings us to the end of episode two of two of chapter six, and I'll see you in the next one.