Utility of Renal Shift Tables to Assess the Magnitude and Chronicity of Antibiotic-Associated Changes in Renal Function in Clinical Trials: Results from Assessment of Telavancin for Treatment of Hospital-Acquired Pneumonia (ATTAIN) trials for Telavancin and Vancomycin

Background: Renal shift tables track categorical changes in renal function over time. The present analysis uses renal shift tables to enhance understanding of the comparable risk and chronicity of changes in renal function in patients receiving antibiotics, using data from the ATTAIN trials. These trials compared telavancin (TLV) and vancomycin (VAN) treatment for patients with Gram-positive hospital-acquired and ventilator-associated bacterial pneumonia.

Methods: Estimates of renal function (using standard Cockcroft-Gault creatinine clearance [CrCl]) from the overall ATTAIN safety population (n=1503) were used to develop shift tables comparing TLV to VAN at baseline vs. day 4, day 7, end of treatment (EOT), and worst observed CrCl through EOT. These tables evaluated the magnitude and timing of changes observed in renal function. The CrCl estimate values were grouped as <30, 30-50, >50-80, and >80 ml/min. A categorical change, based on the CrCl difference between the specified time points, was classified as Grade 1 (1 group shift), Grade 2 (2 group shifts), or Grade 3 (3 group shifts) and statistically compared.

Results: Negative shifts in CrCl categories (baseline vs. worst observed CrCl) during the study were observed in similar proportions of patients: TLV 25.1% and VAN 21.2% (difference 3.9%, 95% confidence interval [CI] -0.51%, 8.3%). For TLV, the greatest negative shifts were noted between baseline and day 7, while for VAN the greatest shifts occurred between day 7 and EOT. However, at EOT no significant differences were observed in negative shifts between TLV and VAN: TLV 18.6%, VAN 15.2% (difference 3.4%, 95% CI -0.6%, 7.3%). Percentage of patients with Grade 2/3 shifts for TLV were 5.6% (38/684) and for VAN were 4.7% (33/698) at EOT. For both treatments, Grade 3 shifts were observed in ≤1% of subjects at day 4, day 7, and EOT. Across all time points analyzed, 67% of TLV and 63% of VAN treated subjects experienced no shifts in renal CrCl. Regardless of treatment received, approximately 25% of patients at each time point analyzed had Grade 1 shifts (both in the positive and negative direction).

Conclusion: Renal shift table analyses can provide a useful assessment of the timing and magnitude of antibiotic-associated renal function changes.