CRITICALLY APPRAISED TOPIC

Janelle Krasovich 10/5/01

Brown NJ, et al. Recurrent Angiotensin-Converting Enzyme Inhibitor-Associated Angioedema. JAMA1997; 278(3) 232-3.

Clinical Question: Is a patient at increased risk for recurrent angioedema with possible worse morbidity/mortality after a first episode of ACEi induced angioedema and are these patients safe with an ARB?

Background: We know from the HOPE trial that ACEi (ramipril) has been shown to reduce mortality and cardiovascular morbidity in adults at high risk for cardiovascular events. We also have learned that ACEi help prevent progression of proteinuria in diabetic patients. As all therapies have risks involved, some patients on ACEi experience side effects such as cough or angioedema and physicians are faced with the decision to continue treatment or change to an alternate therapeutic plan that may or may not carry the same risk reductions and benefits as ACEi therapy. Incidence of ACEi induced angioedema has been reported between 0.1-1%, but there has been little evidence to determine recurrence rate if patient continues on ACEi or recurrence if placed on ARB, and are these occurrences of increased morbidity/mortality.

Study Design: Retrospective Cohort. The setting is a Tennessee Medicaid Program. Enrollees aged 15 years or older who used an ACEi and had a first documented episode of angioedema between 1986 and 1992 were followed up for recurrent episodes through June 1993. The cohort consisted of 82 patients who had a first confirmed episode of angioedema during exposure to ACEi from 1/1/86-12/31/92.

Inclusion Criteria: First documented diagnosis of angioedema was derived from coded diagnosis in admits, D/C, and outpt visits. The diagnosis was then confirmed by medical record review. Each patient’s computerized record was then searched for additional physician, other outpt, or hospital paid claim with the coded diagnosis of angioneurotic edema.

Exclusion Criteria: Diagnosis codes including urticaria and other specified causes, or medical review by trained nurse or physician which didn’t support the diagnosis code of angioneurotic edema.

Results: The results of this study (due to its design) are reported in episodes per 100 patient years.

• From 51,752 patient years on ACEi (1986-1992) - angioedema occurred at a rate of 0.16 episodes/100 patient years

• Of these 82 patients, 30% were hospitalized
• During follow up from 1986-1993 in patients who continued on ACEi therapy – 16 total episodes of recurrent angioedema occurred in 13 patients
• 2 episodes occurred in follow-up in which there was no exposure to ACEi therapy as defined by number of days the drug supply should last, but one patient may actually have been taking the ACEi and was still calculated in the non-exposure group.
• Of the group not reexposed to ACEi, angioedema occurred at a rate of 1.8 episodes per 100 patient years
• Of the group exposed to ACEi, angioedema occurred at a rate of 18.7 episodes per 100 patient years
• Therefore the rate of recurrent angioedema in continued ACEi users was 10 times the rate seen during nonuse (P<0.001)
• Of the 11ACEi users with recurrent angioedema, 5 of these patients were hospitalized during there first occurrence, 3 required ICU care. Of the 3 ICU patients, 2 of these patients were admitted to the ICU during there second admission and required intubation
• No deaths occurred in any patient during the study follow up

Critical Review

Based on EBM guidelines found in JAMA 1994; 271(20): 1615-19.

Are the results valid?

• No clearly identified comparison group was identified. The study group was often compared to a larger group of patients of which they were a subset of, not a comparison group with similar features except the one of interest.
• No other known prognositic factors in the patient group was identified or adjusted for. The only demographic factors known about the study group includes age and economic status. We cannot assess other factors such as confounding diagnosis, other drug therapy, dosage response, etc that may have affected the results.
• Selection of participants were not affected by recall or interview bias, also we may be missing patients that really had angioedema but were coded as alternate diagnosis, not all patients used the same definition of angioedema when coding the diagnosis
• Follow up included a sufficient time course and temporal relationship between therapy and harmful outcome were well documented

What are the results?

• The study design limits the statistical analysis, the outcomes can only be reported as episodes per patient years. The patients are included in the study from different starting points and therefore an accurate determination of relative risk or an odds ratio (the correct statistical analysis inferred in articles reviewing harm) cannot be made
• Confidence intervals are very complicated and difficult to attain when data is presented as episodes per patient years
• However, the rate of recurrence of angioedema in patients with continued exposure to ACEi certainly showed significant statistical significance when compared to the “control” group, and the rate may actually be greater than 10 times given the patient from the “control” group who was likely exposed to ACEi

Will the results help me in my patient care?

• The patients included in this study (age >15 and enrolled in Tennessee Medicaid program) are similar to my patient poplulation , but we do not know anything else about the population group.
• Treatments are certainly similar, I have many patients in my clinic who are currently taking ACEi. Given this study occurred in 1986-1993, I likely have many more patients in my clinic population taking ACEi, and am possibly more likely to find episodes of harmful events than during this time period and am likely more cautious about discontinuing ACEi therapy than physicians during that time period given all of the current evidence of the benefit of ACEi therapy in improving morbitity/mortality.
• Number needed to harm cannot be calculated given design of the study, its results and inablility to calculate absolute risk increase, however I do feel that given these results I would not be likely to continue an ACEi under these circumstances given there are other alternative therapies
• Many interesting questions arise that are not answered by this study such as are the second occurances worse in morbidity, are these patients rehospitalized because of knowledge of previous hospitalization or did these patients actually meet specific admission criteria, what is the number needed to harm given the known benefit of ACEi? Are ARB’s an acceptable second line therapy or do these people have the same increased rate of recurrance?

There is little data to support or refute the use of ARB’s in patients who have experienced angioedema from an ACEi. Another article (EraW, et al. Angioneurotic Edema Attributed to the Use of Losartan. Arch Intern Med 1998; 158: 2063-5.) reviews 13 case reports of angioedema attributed to the use of losartan in Netherlands from 1995-7. They conclude that the onset of angioedema is associated with the use of losartan and suggest that physicians strongly consider the use of ARB in patients with known history of angioedema.