Animal & Veterinary
PMF 005-637 - Tiger Muskellunge - Effectiveness Letter
PMF 005-637 A-0000
November 29, 1999
Leslie E. Holland-Bartels, Ph.D.
Upper Mississippi Science Center
P.O. Box 818
La Crosse, Wisconsin 54602-0818
Dear Dr. Holland-Bartels:
We refer to your submission dated July 29, 1998, in which you requested the establishment of a new public master file (PMF) for chloramine-T. You also requested our review of two studies conducted to demonstrate the effectiveness of chloramine-T for the prevention of mortality associated with bacterial gill disease in tiger muskellunge.
CVM letter dated August 20, 1998, stated that PMF number 005-637 had been assigned per your request.
We have completed our review of the submitted studies and offer the following comments.
1. We concur that the pivotal study (Report #9093 – 1997) in this submission demonstrates the effectiveness of chloramine-T at a concentration of 20 ppm for the prevention of mortality associated with bacterial gill disease in tiger muskellunge.
2. The draft labeling included in the submission proposes a dose range of 5 to 20 mg/L. In order to include this dose range on the product labeling, effectiveness would need to be demonstrated at the low end of the dose range (in this case 5 ppm), while data would need to be submitted that demonstrates the safety of administration at the high end of the range (20 ppm). Accordingly, in order to include a dosage level lower than 20 ppm in the labeling for this indication in tiger muskellunge, additional efficacy data would need to be submitted that demonstrate effectiveness at the lower dose.
1. On page 4 (second full paragraph), the mean cumulative mortality among treated and unmedicated controls should be 2.7% and 62.8%, respectively, based on the data presented in Table 3 on page 10.
2. In Table 2 on page 9, you should use “Source” rather than “Control” in the table under Measured Treatment Concentrations. Using Control implies that water was sampled from each control tank and analyzed for chloramine-T concentrations, when you analyzed only the source water for chloramine-T. In footnote 2, you refer to Appendix G for the false positives in the source water. However, Appendix M contains the results from the water samples, while Appendix G contains the calculations of tank water volumes and amount of chloramine-T to achieve 10 and 20 mg/L concentrations. The footnote and table should be corrected.
3. In Table 3 on page 10, the information in the “Number of Unaccounted Fish” should be included in the other columns. Missing fish should be presumed dead and included in the “Total Mortality” column. The excess fish are appropriately included in the “Number of Survivors” column for tank 11. Also, an additional column should be included for “Percent Mortality.”
Pivotal Study, Report #9093 – 1997
1. In Section 9.6 (Drug Administration), there is no description of how the treated water was disposed of.
2. In Section 11.3 (Statistical Methodology), you state that mortality was analyzed using Bartlett’s test for homogeneity of variance (Appendix K) and t-test with unequal variances (Appendix P). For both procedures, one of the assumptions is that the data are normally distributed. Based on the information presented in each appendix, you analyzed total counts of dead fish and not rates of mortality. You cannot assume counts of dead fish are normally distributed. The data should have been analyzed as mortality rates per tank, which are binomially distributed.
We transformed the mortality rates using an arc sine transformation (sin‑1Ö(proportion)) and used the MIXED procedure in SAS. We also analyzed the mortality rates using the generalized linear models (GENMOD procedure), using a binomial error distribution and a logit link function. Both models included treatment as a fixed effect. The least-squares means and back-transformed mortality rates (percentages) from the MIXED and GENMOD procedures are presented in the following table.
Table 1. The least-squares means and back-transformed mortality rates (percentages) from the MIXED and GENMOD procedures
Treatment MIXED LSMean MIXED Percent GENMOD LSMean GENMOD Percent 0 ppm 0.9187 63.2% 0.5251 62.8% 20 ppm 0.1638 2.7% -3.5973 2.7% p-value 0.0006 -- <.0001 --
3. Table 2 below shows the mortality that occurred from the initiation of medication through the end of the trial. These results should appear in the Freedom of Information Summary.
Table 2. Percent mortality occurring from the initiation of medication through the end of the trial
|No treatment||62.8 (1131/1800)|
|Chloramine-T 20 ppm||2.7 (48/1800)|
“Supportive” Study (no study number)
We consider this study which was conducted in 1996 under INAD 9093 to be non pivotal. Inaccurate dosing on the first three treatment days combined with the lack of assays to show how much drug was actually in the treated tanks cast doubt on the reliability of the data generated during the trial. Results of this study should not be summarized in the Freedom of Information Summary.
Future correspondence regarding this submission to the PMF should be identified by the date of the submission and our file number, PMF 5637 A 0000, and be addressed to the Document Control Unit, HFV-199. Please include only one request per submission, clearly stating the request in the first paragraph of the submission.
If you have any questions or comments regarding this correspondence, please telephone Dr. Joan Gotthardt, Leader, Aquaculture Drugs Group at 301-827-7571.
Steven D. Vaughn, D.V.M.
Director, Division of Therapeutic
Drugs for Food Animals
Office of New Animal Drug Evaluation
Center for Veterinary Medicine