Comparative Tolerability of Intravenous Azithromycin, Clarithromycin and Erythromycin in Healthy Volunteers: Results of a Double-Blind, Double-Dummy, Four-Way Crossover Study

Torsten Zimmermann, Heinrich Laufen, Klaus-Dieter Riedel, Pfizer Research and Development, Illertissen, Germany; Glenda Treadway, Pfizer Pharmaceutical Group, Pfizer Inc., New York, USA; Alexander Wildfeuer, Pfizer Research and Development, Illertissen, Germany

Clin Drug Invest. 2001;21(8) 

In This Article

Participants and Methods

A total of 12 healthy males were enrolled. All participants were required to have a bodyweight of between 55 and 100kg and to be within 20% of the mean value according to height and frame using Metropolitan Life Insurance tables. Screening included a detailed history, physical examination and standard clinical laboratory tests (haematology, clinical chemistry and urinalysis). Participants with known hypersensitivity to antibacterials or to any of the formulation additives, with a history of any drug allergy, with pre-existing or recent clinically significant disease, or with a history of a significant organ disorder, were excluded from the study.

The study was approved by the Swabia Independent Ethics Committee, Bavaria, Germany, and each participant gave written informed consent.

The following drugs were employed in the study: azithromycin dihydrate 500mg in citrate buffer (Zithromax®, Pfizer); 1 clarithromycin lactobionic acid 500mg (Klaricid®, Abbott); erythromycin lactobionic acid 500mg (Erythrocin®, Abbott); and placebo [0.9% (isotonic) saline solution, Pharmacia]. All intravenous solutions were reconstituted as directed in the detailed descriptions of the manufacturers.

This randomised, double-blind, double-dummy, placebo-controlled, four-way crossover study was performed at the Phase I unit of Pfizer R & D, Illertissen, Germany. Throughout each 3-day study period, the participants remained within the Phase I unit and were provided with standard meals at set times (7am, midday and 6pm), with the provision of additional standard-content snacks at 10am and 2pm. No alcohol was allowed from 24 hours before the first infusion until 48 hours after the last dose of each study period. Caffeinated drinks were restricted to three cups per day. The use of prescription drugs and over-the-counter medication was not permitted from 2 weeks before drug administration until completion of each study period.

In the first study period, participants were randomised according to a pre-prepared, computer generated randomisation table to receive for 3 consecutive days azithromycin (500mg in 250ml isotonic saline) once daily, clarithromycin (500mg in 250ml isotonic saline) twice daily, erythromycin (500mg in 500ml isotonic saline) three times daily, or placebo (500ml isotonic saline) three times daily. To ensure blinding of different active medications, despite varying infusion volumes, each participant received additional placebo (250ml or 500ml for the blinding of the 500ml or 250ml infusions, respectively) in the other arm, so that a total of 750ml was infused on each occasion. The infusions, which lasted 1 hour, were administered into a vein in each arm in the morning (8am), afternoon (3pm) and evening (10pm) for 3 consecutive days. Infusions of active drug and placebo were randomly assigned to the right and left arm veins. In the event of an adverse GI event not allowing the infusion to be completed within 1 hour, the infusion could be prolonged to a maximum of 2 hours to avoid the discontinuation of further infusions.

At the end of each infusion, solutions from the infusion tubes were collected and analysed using high-performance liquid chromatography to verify that the drug administered was consistent with the randomisation schedule.

Intravenous catheters (Adsyte® 1.3 x 44mm, Becton, Dickinson, Franklin Lakes, NJ, USA) remained in place until termination of the administration of study drug at 11pm on each study day; the catheters were then removed to minimise over-night discomfort and inconvenience. Between the infusions during the day, the catheters were kept closed by an obturator (Luer Lok®,Becton, Dickinson, Franklin Lakes, NJ, USA). When a severe local reaction required premature removal of a catheter, the participant was given no further doses that day. If possible, the identical infusion site was used throughout each 3-day drug administration period. In case of a limited local reaction at the infusion site on the previous day, an alternative site on the same arm was used and, if possible, the identical vein was punctured.

After a washout period of at least 4 weeks, participants received, in a double-blind manner, the second medication assigned to them in the randomisation table. Thereafter, the whole procedure was repeated twice more, so that all participants had received all active medications and placebo by the end of the fourth study period.

Participants were under full medical supervision throughout the infusion of each study drug. All adverse events were recorded and evaluated with respect to severity. The site and vein used for the infusion were inspected by the investigators every 15 minutes during the infusion and at 15, 30 and 60 minutes after the infusion. If the infusion

resulted in irritation, these inspections were continued at 1-hour intervals until the visible signs of irritation disappeared. Any signs and symptoms of inflammation (pain, swelling, tenderness and erythema) observed by the investigators were recorded using a four-point scale (0 = absent, 1 = mild, 2 = moderate, 3 = severe). A severe event resulted in the infusion being discontinued. Irritation was defined as a mild to moderate response confined to the injection site, and the presence of no more than two of the signs and symptoms of inflammation. Inflammation was defined as the presence of all four signs and symptoms of at least moderate intensity, but confined to the injection site. Phlebitis involved the extension of erythema and pain along the proximal course of the vein.

The intensity of GI events was evaluated by the investigators at the same intervals using a four-point scale (0 = absent, 1 = mild, 2 = moderate, 3 = severe). In the event of a severe GI event, infusion was discontinued.

For the evaluation of infusion site and GI tolerability by the study participants, each individual completed a visual analogue scale (VAS) before and at the end of each infusion, and 1 hour thereafter. The 11-point VAS (0 = no complaints, 10 = intolerable) was used to quantify injection site complaints (irritation, pain, tenderness, swelling, redness, warmth, discoloration, itching, rash and hardness). The same 11-point VAS was used to rate GI events (abdominal pain, nausea, 'butterflies', urgency to defaecate, flatulence/passing gas, regurgitation, hunger, heartburn, abdominal cramping and burping/belching).

At the end of each 3-day drug administration period, a physical examination was conducted, an ECG performed, and blood and urine samples obtained for routine haematology, clinical chemistry and urinalysis.

The sample size was determined on the basis of practicality rather than using biometric considerations. No power calculations were performed, since there were insufficient data available about variability of the primary parameters.

However, it was assumed that a strong (i.e. clinically relevant) variation in tolerability could be found within this sample size. Data from each volunteer receiving the four different regimens in turn (i.e. intrapatient comparison) also supported the chances of finding variation in the tolerability of each drug.

The primary parameters of interest were the incidence of venous irritation and GI intolerance. The time profiles of infusion site or GI adverse events were summarised by areas under the time-score curves (AUSs), with AUSs being estimated by linear trapezoidal integration of the time-score curve from time 0 to 2 hours (i.e. AUS0-2) after the start of study drug infusion. In addition, overall venous irritation and GI intolerance were calculated as the sum of all respective VAS scores: for example, overall venous irritation was evaluated as the sum of investigators' ratings for pain, tenderness, erythema and swelling at corresponding times. Placebo infusions that were administered for the purpose of blinding the infusion volumes were excluded from the evaluation.

Two sets of analyses were carried out. The first compared events after the first infusion on the first day, and the second involved data obtained from the last infusion of each day. Both analyses were a non-parametric analysis of variance (ANOVA).[15] All ANOVA calculations were performed with SAS, Version 6.10 (Cary, NC, USA). The primary comparison of interest was azithromycin versus erythromycin, clarithromycin or placebo; inter-group comparisons were carried out by sequential tests, and the levels of significance adjusted according to the Bonferroni procedure.[16] Individual tests for trends over time in irritation parameters were performed for the nine medication infusions, and for the median values of each medication group, applying Mann's trend test.[15]

Adverse events and discontinuations were summarised by descriptive statistics, and frequencies compared between medication groups by the 2 -test with Bonferroni adjustment for multiple comparisons.[16] The evaluation of data from dropouts from the medication schedules was carried out according to the intention-to-treat principle.

A summary comparison of all ratings, excluding all placebo infusions, was performed. In this comparison, rating sums were calculated over medication administration time for active medication only, and were divided by the number of daily active drug administrations. A relative tolerability index was defined by setting the maximum sum value at 100% and relating other values to the maximum.

1 Use of trade names is for product identification only and does not imply endorsement.


Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.