Proton Pump Inhibitors, revealing new side effects

1
Proton Pump Inhibitors, revealing new side effects

• Andres Marin, MD
• September 2007

2
Use of proton pump inhibitors and the risk of
community-acquired pneumonia

• Archives of Internal Medicine
• May 2007. 167 950-955.

3
Background

• Annually, 20-40 of the general population
  experience dyspepsia or GERD
• PPIs and H2RAs are among the most frequently
  prescribed drugs
• Between 1995 and 2000, 300 increase in PPI use
  (Denmark)
• These medicines effectively reduce gastric acid
  secretion and raise gastric pH

4
Background Cont.

• Intragastric acidity constitutes a major
  nonspecific defense mechanism of stomach ingested
  pathogens
• Normal gastric environment has pH below 4
• Effective tx. With PPIs and H2RAs maintain
  gastirc pH above 4 for at least 18 hours
• Theory that acid suppressive therapy may lead to
  insufficient elimination, or colonization of
  ingested pathogens

5
Background Cont

• Use of PPIs have previously been associated with
  lower GI infections (Salmonella, Campylobacter,
  C-diff.)
• Nested case-control study in 2004 by Laheij et
  al. reported association between gastric-acid
  suppressive therapy and increased risk of CAP
  (JAMA 20042921955-60)
• Current study tested this association in larger
  population

6
Methods

• Population-based case control study using linked
  data from 4 registries
• A) Patient Registry County of Fumen, Denmark
• B) Danish Civil Registry
• C) Odense University Pharmacoepidemiological
  Database
• D) Department of Clinical Microbiology at Odense
  University Hospital

7
Methods Cont.

• Cases (n7642) All ages defined by first
  admission with CAP (ICD-10 codes) 2000-2004 .
• Controls (n34,176) Randomly extracted and
  frequency matched by age and sex and assigned a
  random index date
• Cases could also be extracted as controls prior
  to CAP (index date)
• Exclusion Criteria Malignancy, hospital
  discharge within 7 days of index date

8
Methods Cont.

• Exposure Definition Considered exposed if had
  redeemed a Rx for PPI within 90 days of index
  date (current use)
• If filled Rx over 90 days before index date, then
  considered past users
• Data Analysis Unconditional Logistic Regression
  calculating crude and adjusted ORs

9
Results

• 7642 cases identified, and 34176 controls
• 10.7 of cases and 4.6 of controls were current
  PPI users
• Cases generally more burdened by chronic
  diseases than were controls

10
Results Cont.

• Adjusted OR associating current PPI use with CAP
  was 1.5 (95CI, 1.3-1.7)
• No significant association was found for past
  PPI use
• No significant association was found for current
  or past H2RAs use
• Attributable proportion, ie, fraction of CAP
  potentially caused by PPIs was 4

11
Results Cont.
Big Drop in OR when adjusting for Confounders
No Dose Response
12
Results Cont.

• Stratum Specific Results
• Overall, little variation for subgroup analysis.
• PPI users younger than 40 had Adj. OR2.3 (95
  CI, 1.3-4.0)
• Subjects with no previous hospitalization had
  Adj. OR1.8 (95 CI, 1.0-3.2)
• Subjects who had not received abx. during 90 days
  prior to index date had Adj. OR1.6 (CI, 1.4-1.8)
• PPI users with cirrhosis also had above average
  ORs (OR, 4.6 95 CI, 1.3-17.2)

13
Results
14
Results
Association between current use of PPIs and CAP,
according to the timing of first PPI prescription
15
Discussion

• Their results show that current use of PPI was
  moderately associated with CAP. Overall, it was
  a large, truly population based observational
  study with many strengths.BUT WE MUST DIG DEEPER

16
Discussion

• Potential Problems
• Selection Bias Only included hospitalized
  patients. Also, cannot account for PPI
  non-adherence
• Protopathic Bias Symptoms of CAP could have been
  misinterpreted as reflux leading to PPI
  prescription
• Reflux is associated with some airway symptoms
  (ie. cough) that could lead to a dx of CAP
• ETOH and smoking, known risk factors for both
  reflux and CAP, were not directly adjusted for
• Possible additional residual confounding by
  unmeasured co-morbid conditions/general frailty

17
Discussion

• Support theory

• Problems supporting theory

• Large population based study
• Young PPI users greater OR
• Greater OR with no history of antibiotics
• Greater OR with 1st hospitalization
• Strong temporal relationship
• No increase in OR with airborne pathogens

• OBSERVATIONAL STUDY
• Used Frequency matched controls
• Likely residual confounders
• No dose-response relationship found
• No cumulative dose response
• No response with H2RAs

18
Discussion Reevaluating the mechanistic
explanation

• Results may support association, but acid
  suppression as etiological explanation does not
  add up since no association with H2RAs
• Other possible theories include
• PPIs inhibit immune function
• People with reflux may aspirate gastric contents
  -gt dx of CAP
• Others? Feel free to propose theories

19
Future Studies

• RCT may not be practical. Could analyze pooled
  data from RCTs of PPIs to look at differences in
  rates of CAP (power issues)
• Can repeat analyze data looking as associations
  with other common medications to see if
  association is specific for PPIs
• Repeat using time series analysis
• Repeat using Nested Case-Control Analysis
• Look at patients with pernicious anemia or
  surgical vagotomy to see if complete acid
  suppression increases CAP risk

20
Bottom line

• Though etiology is unclear, there is likely a
  moderate association between PPI use and CAP.
• In general, not a problem since risk for
  developing CAP is low
• However, I would be careful in those who are at
  risk of PNA, and in whom PNA is a major source of
  mortality (asthma, COPD, immunocompromised,
  children and elderly)

21
Adults admitted to SFGH FMIS and GMIS
Courtesy of Drs. Todd May and Katie Murphy
22
Thank You Dr.THom
23
References

• Gulmez et al. Use of Proton Pump Inhibitors and
  the Risk of Community-Aquired Pneumonia A
  Population-Based Case-Control Study. Archives of
  Internal Medicine. May, 14, 2007 167 (9)
  950-955
• Laheij RJ, Sturkenboom MC, Hassing RJ, Dieleman
  J, Stricker BH, Jansen JB. Risk of community
  acquired pneumonia and use of gastric
  acid-suppressive drugs. JAMA. 2004 292
  1955-1960
• Mayhar E., Pharm, D. Ali, S. Pharmacoepidemiology
  I A Review of Pharmacoepidemiologic Study
  Designs. Pharmacotherapy. 2004 24(8) 964-969
• Mayhar E., Pharm, D. Pharmacoepidemiology II The
  Nested Case-Control Study-A Novel Approach in
  Pharmacoepidemiologic Research. Pharmacotherapy.
  2004 24 (9) 1105-1109