Released by FDA: 7/19/00.  Posted by FDA:  7/24/00

Mr. Douglas N. Dobak 
Quality Liaison Leader 
AstraZeneca L.P. 
725 Chesterbrook Blvd. 
Wayne, PA 19087

RE:    NDA#19-810 
          Prilosec (omeprazole) Delayed-Release Capsules 
          MACMIS ID #9086

Dear Mr. Dobak:

This letter concerns AstraZeneca L.P.’s (AstraZenecas) dissemination of promotional labeling and advertising for Prilosec (omeprazole) Delayed-Release Tablets.  The Division of Drug Marketing, Advertising, and Communications (DDMAC) has reviewed Prilosec promotional materials (sales aids #162159, #162105, #160452, reprint carrier #161856, and advertisement #O12OPEOP), as part of its monitoring program and has concluded that AstraZeneca is disseminating materials that contain misleading promotional claims in violation of the Federal Food, Drug, and Cosmetic Act and implementing regulations.  A description of our objections follows.

Misleading Efficacy Claims based on Intragastric pH Levels

Prilosec promotional materials use intragastric pH data from healthy volunteers to suggest that specific levels of intragastric pH acidity correlate with clinical efficacy in the treatment of erosive esophagitis (EE) patients.  However, the relationship between specific intragastric pH levels in healthy volunteers and the clinical efficacy of Prilosec in EE patients has not been demonstrated.  The use of nonclinical data to suggest a clinical benefit where none exists is misleading.  Prilosec promotional materials also use intragastric pH level data to suggest that the clinical effect of Prilosec is superior to that of lansoprazole.  The use of nonclinical data to suggest superiority to another drug is misleading.  For example, Sales Aid #160452 presents a graphical comparison of the lengths of time that Prilosec and lansoprazole each cause intragastric acidity levels in healthy volunteers to exceed a pH value > 4.  The graph shows that 20 mg and 40 mg doses of Prilosec result in pH levels > 4 for 53% and 78% of the day respectively, while lansoprazole 30mg and 60 mg doses result in pH levels >4 for only 46% and 70% of the day respectively.  The implication that, at both doses, Prilosec’s clinical efficacy depends on its ability to maintain an intragastric pH level >4 for a certain percentage of time, is misleading.  Similarly, the implication that the pattern and extent of Prilosec’s daily intragastric acid suppression translates into an improved clinical benefit for Prilosec over Iansoprazole in the EE patient, is misleading.

The small, bifurcated statement at the bottom of the page that “Intragastric pH levels are not indicative of efficacy of healing” is not sufficient to correct the overwhelming message that Prilosec’s intragastric pH level correlates with clinical benefit.  The tiny footnote that P=NS, is also not sufficient to correct the overall misleading message that Prilosec is superior to lansoprazole in clinical effect due to superior acid suppression in healthy volunteers.

Misleading Dose-Related Duration of Effect Claims

Prilosec’s promotional materials misleadingly suggest that increasing the dose of Prilosec will result in an increased clinical efficacy in the treatment of erosive esophagitis or duodenal ulcers due to the increased duration of Prilosec’s antisecretory effect at higher doses.  For example, Sales Aid #160452 states that “as you increase the dose of PRILOSEC, you can increase acid inhibition.”  An accompanying graph shows that the dose of Prilosec 40 mg suppresses acid (pH>3) for a duration of 22 hours a day while lower doses of Prilosec suppress acid for a shorter duration of time throughout the day. Although the Sales Aid is directed at the erosive esophagitis indication, the graph measures pH levels in duodenal ulcer patients.  The graph and accompanying text in Sales Aid #160452, implies that both erosive esophagitis and duodenal ulcer patients will benefit from higher doses of Prilosec due to a correlation between increased clinical effect and increased duration of antisecretory effect.

There is, however, no substantial evidence to prove that a dose-related increase in duration of antisecretory action of Prilosec results in an increased clinical benefit in the treatment of EE or duodenal ulcers.  In fact, the CLINICAL STUDIES section of the approved product labeling for Prilosec states that [for erosive esophagitis] "the 40 mg dose was not superior to the 20 mg dose of Priosec in the percentage healing rate.”  Similarly, the CLINICAL STUDIES section of the approved product labeling for Prilosec states that [for duodenal ulcers] “At 2 and 4 weeks... 40mg was not superior to 20 mg of PRILOSEC, and at 8 weeks there was no significant difference.. ."  Thus, the suggestion of improved clinical effect, based on the Prilosec’s duration of antisecretory effect in the promotional materials where no such effect has been demonstrated, is misleading.

Promotion of Unapproved Dosage Regimens

Prilosec promotional materials suggest dosage regimens for the treatment of erosive esophagitis and duodenal ulcers that are not supported by substantial evidence and are inconsistent with approved product labeling for Prilosec.  Specifically, these materials promote a dose of 40 mg/day for both conditions for a duration of up to 12 years. Prilosec, however, is indicated for the short-term treatment of both conditions at an approved dose of 20mg/day for a period of 4 to 8 weeks.

Misleading Comparative Presentations of EE Healing Rates

Reprint Carrier #161856 makes misleading graphical and textual comparisons between the EE healing rates of Prilosec and other Proton Pump Inhibitors (PPls), based on the Castell study.  The upper graph in Reprint Carrier #161856 is entitled “Healing Rates in Patients with Erosive Esophagitis.”  The graph is misleading because it does not prominently disclose the placebo effect from the Castell Study at 4 and 8 weeks, nor does it prominently disclose the lack of statistical significance of some of the numbers shown.

The accompanying statement that “No PPI is proven to work better.. . for healing all grades of erosive esophagitis in clinical trials” is also not accurate, since the Castell study (and other cited references) did not make head-to-head comparisons between the healing rates of Prilosec and all other PP Is, delineated by each grade of erosion severity.  In addition, the reprint carrier presents healing rates for EE that are inconsistent with the approved product labeling.

Misrepresentative Healing Rates in Severe EE

Sales Aid #162159 compares a combined EE healing rate of 89% for Prilosec 20 mg in patients with grades 3 and 4 erosive esophagitis, to a combined EE healing rate of 85% for lansoprazole 30 mg.  The accompanying text states: “No PPI is proven to work better... in severe erosive esophagitis in clinical trials.”  The graph, however, misrepresents the extent of Prilosec experience in. severe EE patients because it does not present the background incidence rate of patients studied that had the most severe erosions of grade 4.  In the actual study, less than 7% of the Prilosec patients (and less than 9% of the lansoprazole patients) had an erosion severity of grade 4, and less than 27% of the Prilosec (and approximately 30% of the lansoprazole patients) had an EE erosion severity rate of grade 3.  Thus, the healing rate of 89% for Prilosec and its comparison to healing rate of 85% for lansoprazole, for erosion severity of grades 3 and 4, without additional background incidence, suggests that the drug has been studied in a larger population of grade 4 erosion patients than it actually has been studied.  The Sales Aid misrepresents the extent of experience with this drug in healing erosions in grade 4 patients.

Misleading Presentation of Clinical Data re Heartburn Relief

Reprint Carrier #161856, (Castell et al), selectively presents data from the Castell study.  The front flap of the reprint carrier states that “Prilosec 20mg and lansoprazole 30mg provided comparable decreases in heartburn in patients with EE.  There were only minor and inconsistent differences in heartburn symptom assessments.”  However, on page 1753 of the actual reprint, the article states, “Patient diaries revealed significant differences between active treatment groups in the relief of day and night heartburn (Table 2).  Patients receiving lansoprazole 30 mg reported significantly less day and night heartburn during the first day and the first week of treatment than did patients receiving omeprazole 20mg (Table 2).  Similar results were observed when diary entries from the intent-to-treat population were evaluated.”  The Castell reprint summary also stated that lansoprazole provided superior symptomatic relief early in the treatment and was more effective than omeprazole 20 mg with respect to alleviating nighttime heartburn throughout the 8-wk course of therapy.  Thus, the claims of comparable heartburn relief between omeprazole and lansoprazole are not supported by the referenced study.

Expanded Indication

Advertisement #0120PEOP promotes Prilosec for the relief of heartburn without adequately describing Prilosec’s approved indication.  Prilosec is indicated for the treatment of heartburn and other symptoms associated with gastroesophageal reflux disease (GERD).  Heartburn related to GERD or acid reflux disease is heartburn that occurs two (2) or more days a week and persists despite diet or treatment.  Prilosec is not indicated for the occasional relief of heartburn in the absence of GERD.

Lack of Fair Balance

The promotional materials are lacking in fair balance because the risk information is not presented in a manner that is reasonably comparable to the presentation of promotional claims for PRILOSEC.  Promotional materials must present information relating to contraindications, warnings, precautions and adverse effects with a prominence and readability reasonably comparable with the presentation of information relating to the efficacy of the drug.

For example, Sales Aid #162105 states, in large-size colorful header font, that Prilosec has an “Excellent safety record...ZERO cases of ECL cell dysplasia or carcinoids in continuous, open label studies of up to 12 years.. .With PRILOSEC you can increase the dose to 40mg without increasing adverse events.”  In much smaller type at the bottom of the page, however, the balancing statements that “Gastroduodenal carcinoids have been reported in patients with Zollinger-Ellison syndrome on long-term treatment with omeprazole.  This finding is believed to be a manifestation of the underlying condition, which is known to be associated with such tumors... PRILOSEC should be used only for the conditions, dosage, and duration specified in the Prescribing Information.”  This important safety information is not presented with reasonably comparable readability to the efficacy claims presented and lacks fair balance.

In addition, some of the Prilosec materials lack important risk information.  Sales Aid #162159, for example, states a Prilosec claim for the indication of H. pylon-associated duodenal ulcer disease in combination with clarithromycin and amoxicillin.  The Sales Aid, however, fails to include any important risk information that accompanies the use of Prilosec in combination with clanithromycin and amoxicillin.  Specifically, the approved product labeling for Prilosec states that clarithromycin is contraindicated in patients with a known hypersensitivity to clarithromycin, erythromycin, or any of the macrolide antibiotics, that clarithromycin is contraindicated in patients receiving cisapride, or pimozide who have pie-existing cardiac abnormalities or electrolyte disturbances, and that clarithromycin should not be used in pregnant women except in circumstances where no alternative therapy is appropriate. If resistance to clarithromycin is demonstrated or susceptibility testing is not possible, alternative antimicrobial therapy should be instituted.  The Prilosec labeling also states that amoxicillin is contraindicated in patients with a history of allergic reaction to any of the penicillins. 

Reprint Carrier #161856 also fails to provide fair balance.  The carrier extensively details the efficacy results of the Castell study on each page, while limiting the safety information related to the study to one statement detailing the three most common adverse events of the study: headache, and diarrhea and abdominal pain.  However, in the study, nausea was also a common adverse event, that was not mentioned in the carrier, nor does the carrier mention the severe adverse reactions to omeprazole 20 mg that were experienced by the patients in the study.  In the Castell study, four omeprazole 20 mg patients developed severe events that were possibly or probably treatment—related, including rhabdomyolysis, uticarial wheals, severe headache, and severe thromobocytopenic fever.

Requested Action

In order to address these objections, DDMAC requests that AstraZeneca:

1. Immediately ceases further use of these and other materials and practices with the same or similar messages.
2. Provide DDMAC, in writing, with AstraZeneca’s intent to comply with the above.  This response should include a list of all violative promotional materials and AstraZeneca’s methods for discontinuing their use.

AstraZeneca’s response should be received no later than July 28, 2000.  If you have any questions, you should direct them to the undersigned in writing or by facsimile at (301) 594-6759 or at the Food and Drug Administration, Division of Drug Marketing, Advertising, and Communications, HFD-42, Rm. 17B-20, 5600 Fishers Lane, Rockville, MD 20857.  DDMAC reminds AstraZeneca that only written communications are considered official.

In all future correspondence regarding this particular matter, please refer to MACMIS ID #9086 in addition to the NDA number.

Sincerely,

Patricia Kuker Staub, R.Ph, J.D. 
Regulatory Review Officer 
Division of Drug Marketing, 
   Advertising and Communications