My goals in this article are to give you my current recommendations on how you should risk stratify and treat patients with peptic ulcer bleeding (PUB) and other types of non-variceal upper GI hemorrhage (NVUGIH), based upon recent studies performed by me and my research group, the CURE Hemostasis Research Group.
For more than four decades, risk stratification of PUBs for re-bleeding and treatment recommendations have been based on stigmata of recent hemorrhage (SRH). This is in spite of limitations such as difficulty of standardization and lack of agreement in classifying PUBs. Even among international experts, the agreement in classifying SRH in PUBs is only fair to poor.
My first major message is that what actually determines whether PUBs re-bleed is not the stigmata of recent hemorrhage but rather the arterial blood flow underneath the SRH. If there is still a patent artery after endoscopic hemostasis, there is a much higher risk of PUB re-bleeding. Now we can easily and reliably detect arterial blood flow with a Doppler endoscopic probe (DEP) in lesions and use this as a guide along with the visual stigmata of recent hemorrhage to risk stratify patients and improve endoscopic hemostasis. So, use both your eyes (for SRH) and ears (for DEP blood flow detection).
My second message is to separate oozing bleeding PUBs (Forrest IB) from PUBs with spurting or pulsatile arterial bleeding (Forrest IA) and other major stigmata of recent hemorrhage including non-bleeding visible vessels, or NBVV – FIIA, and adherent clots, or FIIB. Compared to other major stigmata of recent hemorrhage oozers FIB have a much lower risk of re-bleeding and DEP detection rate of arterial blood flow both before (46.7 percent vs. 87.4 percent) and after standard endoscopic hemostasis (0 percent vs 27.4 percent). Evidence is from a large CURE prospective cohort study of PUBs with different SRH, with underlying blood flow studied by endoscopic probe (GIE 2016;83:129). In prospectively followed subgroups, oozing PUBs, when compared to spurting PUBs, had a significantly lower rate of arterial flow underneath the stigmata of recent hemorrhage (46.7 percent vs. 100 percent), a lower rate of residual blood flow after standard endoscopic treatment (0 percent vs. 35.7 percent) and a lower re-bleeding rate (0 percent vs. 28.6 percent). This was in spite of different medical management treatments after successful endoscopic hemostasis (oral proton pump inhibitors in oozers versus high dose IV proton pump inhibitors in spurters). Also, the oozers were much easier to treat endoscopically than the spurters.
Other confirmatory evidence comes from a secondary data analysis that I participated in that included PUBs in a large international randomized controlled trial (RCT) of high dose esomeprazole versus placebo to reduce rebleeding. The re-bleeding rate of oozing PUBs treated for 72 hours with IV placebo after successful endoscopic hemostasis was only 4.9 percent (8/163), whereas re-bleeding with other major SRH combined (FIA, FIIA or FIIB) was significantly higher at 14.2 percent (32/225). Furthermore, when we compared re-bleeding rates of patients up to 72 hours according to treatments, major SRH PUBs – spurters, nonbleeding visible vessel or adherent clot – had a significant reduction in re-bleeding rate with high dose proton pump inhibitor infusion (6.3 percent vs. placebo 14.2 percent). However, those with oozing PUBs did not (proton pump inhibitor 5.4 percent vs. placebo 4.9 percent).
So the take home message to clinicians is to separate oozers from spurters and other major stigmata of recent hemorrhage, endoscopically treat them and consider oral proton pump inhibitor treatment. In contrast, treat spurters, NBVV and adherent clot PUBs with high dose IV proton pump inhibitor infusions after successful endoscopic hemostasis.
My third major recommendation is based upon a recent CURE RCT of standard (visually guided) versus DEP-assisted hemostasis of NVUGIB patients. 148 patients with severe UGI bleeding were randomized when SRH were found- Forrest I A (spurting), FIB (oozing), FIIA (NBVV), FIIB (adherent clot) or FIIC (flat spot) in NVUGIB. 84.5 percent of patients (125) had peptic ulcers, 12.8 percent had Dieulafoy’s lesions (19) and 2.7 percent had Mallory Weiss tears (4).
All patients with SRH received high dose IV PPI infusion after endoscopic hemostasis, except spots had BID oral proton pump inhibitors. After 72 hours, all PUB patients had oral BID PPI. At baseline, risk factors for re-bleeding were similar but 30-day outcomes in the standard treatment group compared to the Doppler endoscopic probe group were significantly different: including rates of re-bleeding (20/76 vs. 8/72), major complications (4/76 vs. 0/72), surgery (4/76 vs. 0/72), and more RBC’s transfused after randomization (1.09 vs. 0.56 units).
Our conclusions from this randomized controlled trial were that Doppler endoscopic probe determination of blood flow underneath SRH improved both risk stratification for re-bleeding and completion of definitive hemostasis. Based upon strong scientific evidence, I now recommend that DEP detection of blood flow be included in assessment and treatment of patients with severe non-variceal upper GI hemorrhage. It’s easy to learn how to use, effective in improving outcomes, safe and inexpensive.
Dr. Jensen is consults for Boston Scientific and Vascular Technology Inc. The research studies cited and Dr. Jensen were funded by a VA Clinical Merit Review Grant (CLIN-013-07F) & NIH-NIDDK P30 CURE DDRC 41301 (Human