• Helicobacter pylori infection in steroid users has not been well evaluated in the past.
WHAT THIS STUDY ADDS
• Prednisolone 60 mg daily for 3 months caused few endoscopic ulcers (1.6%) and no ulcer complications in 60 nephrotic patients, which was comparative with the controls, with an odds ratio of 0.5.
• Corticosteroid therapy did not to increase PUD in nephrotic syndrome patients. Further larger studies are needed to clarify the role of corticosteroid in PUD.
Whether corticosteroids induce peptic ulcer disease (PUD) remains uncertain. The study evaluated and compared the occurrence of PUD between nephrotic patients receiving oral prednisolone therapy and nephritic patients without steroid therapy.
The prospective case control study compared 60 nephrotic syndrome patients who received 60 mg daily prednisolone therapy for 3 months with 30 age-and sex-matched nephritic patients without steroid therapy. Each patient underwent endoscopic examination and tissue and blood sampling before and after the study. Examined parameters included Helicobacter pylori (H. pylori) infection, and gastric and serum prostaglandin (PG) E2 and thromboxane (TX) B2 concentrations. The primary endpoint was the occurrence of endoscopic peptic ulcers between the two groups, while the secondary end point was the occurrence of ulcer complications.
The two groups were comparable in sex, age, smoking habits, alcohol drinking, past history of PUD, H. pylori infection rate and serum creatinine. There were no differences in the occurrence of endoscopic peptic ulcers (1.6% vs. 3.3%) and ulcer complications (0% vs. 0%), pre-therapy gastric PGE2, and pre- and post-therapy gastric TXB2, serum PGE2 and serum TXB2 between the two groups. However, there was significantly lower post-therapy gastric PGE2 concentrations in the prednisolone group.
Three-month therapy with 60 mg daily prednisolone caused few endoscopic ulcers (1.6%) and no ulcer complications in nephrotic patients. Corticosteroids therapy did not increase PUD in nephrotic syndrome patients [odds ratio 0.492 with 95% confidence interval (CI) 0.03, 8.142, P= 0.620]. Further larger studies are needed to clarify the role of corticosteroids in PUD.
Whether corticosteroids use causes gastric mucosal injury and peptic ulcer disease (PUD) remains controversial . It is suggested that the mechanisms responsible for peptic ulcer formation induced by corticosteroids include enhanced gastrin and parietal cell hyperplasia with increased acid secretion, diminished gastric mucus synthesis, and suppressed arachidonic acid metabolism and prostaglandin (PG) synthesis [1, 2]. Initial meta-analysis of 26 placebo-controlled, double-blind studies showed no association between corticosteroids use and PUD . However, subsequent meta-analysis of 71 clinical trials revealed that systemic corticosteroids or adrenocorticotrophic hormones increased the risk of PUD . Another study showed that non-steroidal anti-inflammatory drug (NSAID) use and smoking, but not Helicobacter pylori (H. pylori) infection, were risk factors for PUD in autoimmune disease receiving long-term corticosteroids therapy .
Recently, a prospective paired study showed that short-term large-dose use of steroids alone did not induce gastroduodenal mucosal injury in systemic lupus erythematosus patients receiving pulse methylprednisolone therapy . Animal studies revealed that non-ulcergenic doses of dexamethasone delayed rat gastric ulcer healing via inhibition of cyclo-oxygenase (COX)-2 expression and PGE2 synthesis without inhibition of COX-1 expression . Another animal study showed that dexamethasone damaged rat stomach only when the function of COX-1 was inhibited .
There has been no prospective study investigating whether long-term use of medium-dose corticosteroids causes PUD. Patients who suffer from nephrotic syndrome due to minimal change disease (MCD) and focal segmental glomerulosclerosis (FSGS) usually receive steroid therapy for their renal disease . This prospective case control study aimed to evaluate and compare the occurrence of PUD between nephrotic patients receiving 60 mg daily oral prednisolone therapy for 3 months and nephritic patients without steroid therapy.
Patients diagnosed with nephrotic syndrome with renal histology of MCD or FSGS who received oral prednisolone 60 mg day−1 for 3 months were enrolled. Nephrotic syndrome was defined as heavy proteinuria >3 g day−1 with serum albumin <2.5 g ml−1, hypercholesterolaemia and peripheral oedema . MCD and FSGS were diagnosed by renal biopsy based on their histologic characteristics . Patients were excluded if they (i) were receiving acid-reducing agents or mucosal protective agents (i.e. proton-pump inhibitor, histamine2-receptor antagonists or misoprostol) or NSAIDs, including selective COX-2 inhibitors, aspirin or selective serotonin re-uptake inhibitors, (ii) had current PUD or history of surgery for PUD complications, (iii) had bleeding tendency (platelets <80 000/mm3, prolonged prothrombin time, partial thromboplastin time) or (iv) had poor general health. The Institutional Review Board of Taipei Veterans General Hospital (VGHIRB No: 94-08-01) approved this study, which was conducted in accordance with the Declaration of Helsinki guidelines. All patients provided written informed consent.
Procedures and measurement
Within 1 week before oral prednisolone therapy, the patients' medical history and medications were reviewed. Habitual consumption of cigarette tobacco (≥half pack daily), alcohol (≥80 g weekly), tea (≥three cups daily), or coffee (≥three cups daily), and past history of PUD were recorded . Each subject underwent video oesophago-gastroduodenoscopy (EGD) (Olympus GIF-Q260 gastrointestinal videoscope, Aizo Olympus, Fukushima, Japan). An ulcer was defined as mucosal break ≥3 mm with significant depth [11, 12].
H. pylori infection was diagnosed based on histologic observation showing H. pylori in stomach specimens and a positive rapid urease test (CLO test, Ballard medical products, Draper, Utah, USA) [1, 11]. Gastric juice was also collected for pH assessment. Antral gastric mucosa was collected to measure PGE2 and thromboxane (TX) B2 concentrations. Serum was collected before EGD to measure serum albumin, creatinine, PGE2, and TXB2 concentrations. Blood COX-2 activity was measured as lipopolysaccharide-induced PGE2 and blood COX-1 activity was measured as serum TXB2[5, 13]. LPS-induced PGE2 in whole blood, a marker of monocyte COX-2 activity, and serum-generated TXB2 in whole blood, a marker of platelet COX-1 activity, were measured as previously reported [5, 13]. PGE2 and TXB2 were measured using PGE2 and TXB2 ELISA kits, respectively (Quantikine, R&D systems Inc, Minneapolis, MN, USA).
Gastric tissues were homogenized with buffer for 30 s, and then centrifuged at 24 148 g for 15 min at 4°C. Supernatants were assayed in duplicate using commercially available PGE2 and TXB2 ELISA kits. A protein assay kit determined protein concentrations in each sample, while mucosal PGE2 and TXB2 concentrations were expressed as pg mg−1 protein .
After taking prednisolone, patients were followed-up with monthly outpatients visits and haemoglobin checks as well as stool occult blood tests for 3 months. Compliance with these regimens was assessed by counting the number of pills that were returned. Rescue medication (antacids with an acid-binding capacity <16 mmol HCl per tablet) was provided, and patients took no more than four tablets per day only when they had ulcer symptoms (e.g. epigastric pain, dyspepsia). Patients reported to a gastroenterologist if they had persistent ulcer symptoms for 1 week and reported to the emergency room if they had evidence of GI bleeding or ulcer complications (i.e. melaena, haematemesis or sudden-onset severe epigastric pain).
Urgent EGD was undergone to document any gastro-duodenal ulcers with or without complications. Gastric juice pH, PGE2 and TXB2 concentrations, as well as serum albumin, creatinine, PGE2, and TXB2 were measured if available. Patients without persistent ulcer symptoms or evidence of ulcer complications received follow-up EGD within 2 days after 3 months of prednisolone therapy to detect any gastroduodenal ulcers. These measurements were taken again during follow-up EGD.
This study also enrolled 30 age- and sex-matched nephritic patients who did not receive glucocorticoid therapy as controls for comparison. They had received baseline EGD with tissue and blood sampling and had similar exclusion criteria. They received urgent EGD to document any gastroduodenal ulcers with or without complications if patients developed symptoms before the end of the study period. Patients without ulcer symptoms or complications also received follow-up EGD at the end the study. Tissue and blood sampling during follow-up EGD were also performed.
The primary endpoint was the occurrence of endoscopic peptic ulcers (symptomatic ulcers during the 3-month therapy and asymptomatic ulcers at the end of therapy) between nephrotic patients receiving prednisolone therapy and nephritic patients without such therapy. The secondary endpoint was the occurrence of ulcer complications.
Statistical analysis of endoscopic ulcers and their complications was conducted on the intention-to-treat population. Data were expressed as means ± SD and results were compared between groups depending on the type of data analyzed, using chi-square test, Fisher's exact test, Student's t-test or non-parametric Mann–Whitney U-test, when appropriate. Paired t-test or non-parametric Wilcoxon signed-rank test was used to compare continuous data for the same patient before and after prednisolone therapy. Logistic regression was used to analyze the odds ratio of PUD in prednisolone users. All statistical analyses were performed using spss for Windows version 14.0 (SPSS Inc, Chicago, IL, USA). All P values were two-tailed and a value <0.05 was considered statistically significant.
From August 2007 to March 2009, 65 nephrotic patients who received oral prednisolone therapy for 3 months were initially enrolled. After screening, one subject was excluded due to PUD identified by EGD, two for aspirin use as cardiovascular medication, one for taking NSAIDs for osteoarthritis and one with poor general health. Sixty nephrotic patients were finally included in the intention-to-treat analysis. During the follow-up period, two (3%) of 60 subjects dropped out (one for non-compliance to medication without gastrointestinal symptoms and another for changing prednisolone therapy to another regimen due to poor response after 2 months (Figure 1).
Those in the prednisolone group had an average age of 50 ± 20 years (range 24–83 years). Thirty-three (55%) were male and 27 (45%) were female; three (5%) were smokers, one (1.3%) had a drinking habit, six (9%) used tea, and one (1.3%) used coffee. Twenty (33%) patients had H. pylori infection and two (3%) had a history of PUD. During the 3-month prednisolone therapy period, one case (1.6%) developed symptomatic gastric ulcer with positive H. pylori status after 4 weeks of therapy. None developed ulcer complications and none had endoscopic ulcers at the end of 3 months therapy. Comparison of pre- and post-therapy parameters in the 58 patients revealed no significant difference in terms of gastric pH, serum TXB2 (COX-1 activity), serum PGE2 (COX-2 activity) and gastric TXB2. However, there was significantly lower gastric PGE2 (1505 ± 795 vs. 939 ± 730 ng mg−1 protein, P < 0.001) in the prednisolone treated group post-therapy (Table 1).
Table 1. Comparison of the demographic and laboratory data between nephrotic patients with prednisolone therapy and nephritic patients without prednisolone therapy
Nephrotic patients with prednisolone therapy n= 60
Nephritic patients without prednisolone therapy n= 30
In the control group, the average age was 50 ± 20 years (range 26–80 years). Seventeen (56%) patients were male, thirteen (44%) were female, two (6.7%) had smoked, one (3.3%) had a drinking problem, two (6.7%) used tea and none used coffee. Eleven (37%) had an H. pylori infection and one (3.3%) had a past history of PUD.
During the 3-month follow-up period, none developed symptomatic ulcers or complications. One (3.3%) with positive H. pylori infection had asymptomatic duodenal ulcer at the end of the follow-up period. Comparison of pre- and post-follow-up parameters in the 30 patients revealed no significant difference in gastric pH, serum TXB2 (COX-1 activity), serum PGE2 (COX-2 activity), gastric TXB2 and gastric PGE2 (Table 1).
Data comparison between the prednisolone therapy and control groups
The two groups were comparable in terms of baseline demographic and risk factors of peptic ulcer, including sex, age, smoking and alcohol drinking habits, tea and coffee consumption, past history of PUD, H. pylori infection rate and serum creatinine (Table 1). There were no differences in the occurrence of endoscopic peptic ulcers (1.6% vs. 3.3%) and ulcer complications (0% vs. 0%), pre-therapy gastric PGE2, and pre- and post-therapy gastric pH level, gastric TXB2, serum PGE2 and serum TXB2 (Table 1). However, the prednisolone group had significantly lower serum albumin and lower post-therapy gastric PGE2 (Table 1).
Analyses of protective or risk factors for PUD among the 90 subjects
There was no significant difference for PUD in sex distribution, smoking and alcohol drinking habits, use of tea and coffee, past history of PUD, H. pylori infection and prednisolone use (all P > 0.05) although two cases with PUD (one in the study group and the other in the control group) had H. pylori infection (P= 0.116). The odds ratio of PUD in prednisolone users was 0.492 (95% confidence interval 0.03, 8.142, P= 0.620).
This study demonstrated that 3-month medium-dose corticosteroids (60 mg prednisolone day–1) therapy decreased gastric PGE2 formation but caused few endoscopic ulcers (1.6%) and related complications (0%) in nephrotic patients, which was comparable with the occurrence of endoscopic ulcers (3.3%) and complications (0%) in nephritic patients without prednisolone therapy.
It is well recognized that NSAIDs, including aspirin, are ulcerogenic agents. The PLUTO study showed 20.4% endoscopic ulcers in 6 months among NSAIDs users with high gastrointestinal risks, while Scheiman et al. and Cullen et al.'s showed 16.5% endoscopic ulcers in 6 months among NSAIDs users with general gastrointestinal risks [14, 15]. Yeomans et al.'s study showed 5.4% endoscopic ulcers in 26 weeks in H. pylori-negative aspirin users (75–325 mg day−1) with negative baseline gastroduodenal ulcers . Laine et al.'s study showed 7.3% endoscopic ulcers in 12 weeks in aspirin (81 mg day−1) users with negative baseline gastroduodenal ulcers . However, whether corticosteroids induce peptic ulcer formation remains uncertain. Our study appears to be the first prospective, matched, case-controlled study with a negative baseline for gastroduodenal ulcers to show that 3 months of 60 mg daily prednisolone does not increase PUD in nephrotic syndrome patients when compared with nephritic patient controls.
The laboratory data here showed that prednisolone therapy did not cause significant change in serum TXB2 (COX-1 activity), serum PGE2 (COX-2 activity), and gastric TXB2, but significantly decreased gastric PGE2 when compared with the pre-therapy data of the prednisolone group and the post-follow-up data of the control group. These findings suggest that prednisolone use, although it diminishes more COX-2 activity rather than COX-1 activity, causes no more endoscopic ulcers when compared with the controls. These results are consistent with previous animal studies showing that corticosteroids do not decrease gastric mucosal COX-1 function  and inhibition of both COX-1 and COX-2 functions is required for gastric mucosa injury in rats [7, 18].
H. pylori infection is a risk factor for PUD in aspirin and NSAIDs users . However, there are no studies showing that H. pylori infection is an aggressive factor for PUD in steroid users. A previous study showed that H. pylori infection was not a risk factor for PUD in autoimmune disease patients receiving low-dose oral steroid therapy . Another study also showed that H. pylori infection did not increase gastric mucosal injury in SLE patients receiving pulse methylprednisolone therapy . In the present study, only one of the 60 nephrotic patients developed PUD. Though the subject had H. pylori infection, there were too few ulcer patients to analyze whether H. pylori infection played an important role for PUD in nephrotic patients receiving prednisolone therapy.
Though patients with chronic renal failure and haemodialysis have a higher prevalence of PUD , there are no reports demonstrating that nephrotic patients or nephritic patients have a higher PUD rate. The current study cannot conclude that 3-month medium-dose corticosteroid therapy does not increase PUD. There were several limitations in this study. First, only patients with relatively low gastrointestinal risks were enrolled due to ethical reasons. Patients who had current PUD (one subject during screening), who took aspirin or NSAIDs (three subjects during screening) or who had history of surgery for PUD complications (zero in this study) were excluded. Second, the study was limited to 60 nephrotic patients who needed steroid therapy nephritic patients without steroid therapy were compared as control group. Further studies enrolling more subjects including healthy subjects, elderly patients, or other disease patients who need steroid therapy are needed to clarify whether steroid therapy is ulcerogenic or not for the human upper gastrointestinal tract. Third, there was only one ulcer subject in the study group and, therefore, it was impossible to evaluate the role of H. pylori infection for PUD in steroid users in this study.
In conclusion, a 3-month medium-dose corticosteroid therapy (60 mg prednisolone day–1) caused few endoscopic ulcers (1.6%) and no related complications in nephrotic patients, which was comparable with the occurrence of endoscopic ulcers (3.3%) and complications (0%) in nephritic patients without prednisolone therapy. Corticosteroid therapies seem not increase PUD in nephrotic syndrome patients. Further larger studies are needed to clarify the role of corticosteroids in PUD.
There are no competing interests to declare.
This work was supported by a grant from the Taipei Veteran General Hospital (V98C1-051) and a grant from Pfizer (ID#7984/IIR2007-05) but was independent of them.
We thank Miss PC Lee (Department of Medicine, Taipei Veterans General Hospital) for her help in the statistical consultation, as well as the support of the Core Laboratory, Department of Medical Research and Education, Taipei Veteran General Hospital.