Two pivotal efficacy studies were conducted at an Elanco Animal Health laboratory in St-Aubin, Switzerland, in compliance with Swiss guidelines covering efficacy studies with ectoparasiticides in cats and dogs, VICH GL9 Good Clinical Practice Guidelines, EMEA/CVMP/EWP/005/2000-Rev.2, and European Parliament Directive 2001/82/EC [14,15,16].
Both studies were performed in compliance with GCP quality standards and were blinded, randomized, parallel-group and negative controlled.
Study 1 was a dose confirmation study with the objective of confirming the efficacy of lotilaner tablets for cats as close as possible to the minimum recommended dose rate of 6 mg/kg, against adult fleas (Ctenocephalides felis), in experimentally infested cats. Sixteen cats were randomized to two groups: one treatment group and one negative control group of eight cats each. Efficacy was evaluated 24 (± 1) h after treatment on day 0 (i.e. on day 1, 48 h after flea infestation) and 24 (± 1) h after each subsequent flea infestation (i.e. on days 8, 15, 22, 29 and 36).
The objective of Study 2 was to evaluate the speed of kill of lotilaner dosed orally once, on day 0, at the minimum recommended dose rate of 6 mg/kg against experimental infestations with adult fleas (C. felis) in cats. Thirty-two cats were randomized to four groups of eight animals each: two treatment groups (Groups 2 and 4) and two negative control groups (Groups 1 and 3). Efficacy was measured 8 and 12 h after treatment and 8 and 12 h after each subsequent weekly flea infestation, until the end of the study on day 35 (see Table 1 for details on cat groups and time points).
A pre-dose infestation test was performed during the acclimatization phase in order to evaluate the flea retention rate measured at 24 (± 1) h post-infestation. This retention rate was used as a criterion for inclusion of the cats into the study and for randomization.
Body weight was recorded three times, twice during the acclimatization phase and then once at the end of the study. Full physical examinations were performed at the beginning of acclimatization and at the end of the study. General health observations were performed daily and additional clinical observations occurred at approximately 1 h, 6 h and 8 h post-treatment.
In order to obtain eight cats suitable for inclusion in each study group, 60 healthy European domestic short hair cats of both sexes (24 cats in the first study and 36 cats in the second study) from the St-Aubin cat colony, with a minimum age of 12 months, with body weight that allowed for administration of lotilaner at a minimum dose of 6 mg/kg and as close as possible to this target dose rate, were selected for acclimatization. These cats had to be clinically healthy and with no pre-existing skin lesions or known high allergic reactions to insect bites. No other ectoparasiticide treatment had been provided to them for at least 8 weeks (6 months for isoxazolines) before the first infestation in these studies.
An infestation test (“pre-dose infestation test”) was performed during the acclimatization phase (14 days before start of the study) in order to evaluate the flea retention rate measured at 24 (± 1) h post-infestation. Cats having a flea retention rate of at least approximately 50% in the acclimatization phase were considered eligible for inclusion in the study. Any cats that developed important skin lesions or had an allergic reaction to insect bites after the pre-infestation test of the acclimatization phase and cats considered difficult to manipulate were excluded from the study. Following the acclimatization phase, 48 cats were randomized in the studies (16 cats in Study 1 and 32 in Study 2). After discussion between the investigator and the sponsor, three cats with flea retention rates slightly lower than 50% (47-49%) were included in Study 1. Similarly, one cat with a flea retention rate of 42% was included in Study 2. These deviations were minor and did not affect the study results. Any cats excluded from the study were returned to the research site cat colony.
Cats were housed in rooms, each containing two pens, with up to 8 cats per room and 4 cats per pen. In both studies, cats were housed in individual cages from infestations until the flea counts were completed. On some days during acclimatization, cats were also individually housed to be trained to consume their feeding portion in 30 min. Additional individual housing was applied on some acclimatization days for approximately 30 min or 1 h in Study 1 and 2, respectively, to train cats to wear an Elizabethan collar.
Commercially available cat food was provided once daily. To motivate cats to eat their whole food portion within 30 min before treatment, each cat received only half of their normal daily food portion the day prior to treatment (i.e. day -1). Cats were not force-fed. Fresh drinking water from the communal water plant was available ad libitum. Temperature and humidity were adequately controlled for the duration of the study and a few noted excursions were short and minor without any impact on the study outcome. During the study, cat toys and cat scratchers were available for animal welfare reasons.
Randomisation and blinding
Randomization was carried out using SAS/STAT® software (ver. 9.2; SAS Institute Inc., Cary, NC, USA) per the site Standard Operating Procedure (SOP). Cats were ranked ordered from highest to lowest flea pre-dose counts and randomly allocated within blocks to treatment groups. Block sizes of two cats were used in Study 1 and of four cats were used in Study 2. When two or more cats had the same number of live fleas, the tie was broken randomly. Cats within each treatment group were randomly allocated to study rooms with each room containing an equal representation of treated and control cats of the same counting time point as applicable for the study. Blinding was accomplished by separation of function. In Study 1, study personnel responsible for general health observations, clinical observations, preparation of fleas, infestation with fleas, physical examination, safety data review, animal weighing, and/or animal maintenance were blinded throughout the study. The sponsor’s representative, monitor, quality assurance personnel, statistician, and dispenser and assistant were unblinded. In Study 2, blinding was guaranteed by a similar setting but the safety data review was performed by the unblinded sponsor’s representative.
Flea infestation and counting
Except for the pre-dose infestation test, in the intervals between flea counting and next infestation, cats were combed for 8-16 min in order to remove any fleas remaining from the previous infestation.
A mixed flea strain of Swiss and Danish origin was used in both studies. This strain was confirmed as a C. felis strain based on morphological examination and confirmation that the morphological features were identical to those described by Zentko & Richman .
Flea infestations were performed per the site SOP: 100 (± 5) unfed adult C. felis fleas (both sexes and aged 17-21 days in Study 1 and 17-22 days in Study 2, were released on the back of the cats, from the neck to the central part of the spine. For each infestation period (from infestation until flea count was completed), cats wore an individual Elizabethan collar. The Elizabethan collar was removed prior to start combing the cats. Cats were not sedated for flea infestation.
For flea counts, cats were first combed for 10 min. If during the last 2 min, a minimum of three live fleas were found, the animal was combed for two additional min, for a maximum of 16 min combing time. During each flea count, all fleas were removed, and live fleas killed in 70% ethanol solution and/or in a freezer (for at least 24 h); all fleas were then disposed of in compliance with local requirements. No cat needed to be sedated for the flea counts.
In Study 1, fleas were classified as either live or dead. In Study 2, the additional category “moribund” was added. Fleas were considered live if they could maintain an upright posture and if they could actively move through hair removed from the animal during the combing procedure. A moribund flea was a flea that was laterally recumbent, could not normally move through hair or “right” itself when placed on a flat surface, but still had leg movement or twitching. A dead flea was a completely immobile flea.
Based on day -2 body weights, all cats in the treated groups received a single oral administration of lotilaner on day 0 at a minimum dose rate of 6 mg/kg and as close as possible to this target dose rate; the treatment was administered 30 ± 5 min following feeding. Two tablet strengths (12 mg and 48 mg) were available. The required number of tablets was placed over the tongue of the cats, in the back of the oral cavity followed by administration of 1-5 ml of water via a syringe to facilitate swallowing. This was followed by a check of each cat’s mouth to ensure that the full dose had been swallowed. Mock dosing was performed for cats in the negative control groups where the treatment administrator opened and massaged the cat’s mouth followed by administration of 1-5 ml of water via a syringe. In both studies, no animal receiving lotilaner spat out the tablet(s) or vomited within 30 to 45 min following dosing.
Study assessments and statistical analyses
All statistical analyses were performed using SAS® (ver. 9.2.2, SAS Institute Inc., Cary, NC, USA). The experimental unit was the individual animal. Descriptive statistics (arithmetic and geometric mean, minimum, maximum, standard deviation, coefficient of variation and median) were calculated with respect to study groups separately for flea counts and body weights.
Efficacy was defined as the ability of lotilaner to reduce flea infestations on the cats at the specified time points (24 h in Study 1 and 8 h and 12 h in Study 2) after treatment following the infestation on day -2 and after each post-treatment infestation. Efficacy calculation was based on live flea counts; in Study 2, two different calculations were performed, in which moribund fleas were considered as either live or dead. Efficacy was determined based on the percent reduction in live adult flea counts in the treated group compared with the negative control (untreated) group for the same time point.
In Study 1, efficacy analyses were performed using the intent-to-treat data set, which included all randomized animals, i.e. animals that received lotilaner and the untreated animals. In Study 2, some cats could not be infested on some days for animal welfare reasons, due to reactions to flea bites, leading to smaller (but still sufficient) group sizes. On each day, data were present only for those animals which had been infested. Therefore, only these animals were included in the statistical analysis for efficacy (per protocol analysis).
Flea infestation was considered adequate at each time point after lotilaner administration if the arithmetic mean of the flea retention rate of the control animals was at least approximately 50%. There were separate calculations for each time point. Arithmetic and geometric means of the flea counts were calculated. Efficacy using geometric means was performed for information purposes only. Efficacy was calculated as follows:
where C is the arithmetic/geometric mean number of live fleas in the negative control group and T is the arithmetic/geometric mean number of live fleas in the treated group.
In order to avoid taking the log of zero, geometric means were calculated after adding 1 to all flea counts, and 1 was subtracted from the resulting geometric mean, if any of the flea counts were zero. Log-transformed flea counts were compared between groups in an analysis of variance (ANOVA). In Study 2, as the assumption of normal distribution of log-transformed flea counts was not valid, non-parametric methods were additionally applied to compare the treated and negative control groups (Mann-Whitney U-test).
Lotilaner was considered effective for the control of C. felis at a given time point if infestation was adequate, if there was a statistically significant difference between flea counts of the treated group and the untreated group (2-sided P < 0.05), and if the treated group had a calculated efficacy of ≥ 95% (live fleas) using the arithmetic means.
Adverse events were coded according to VeDDRA SOC and PT (Veterinary Dictionary for Drug Regulatory Activities, System Organ Class, Preferred Term). For each SOC and PT, the number of animals showing the respective sign was counted for untreated and treated animals, and the number of adverse events with the respective sign was calculated. Some of the adverse events that occurred during group housing could not be attributed to a specific cat but only to the set of four cats in the same pen. These adverse events were summarized separately, by pen rather than by cat. Full physical examinations were performed at the beginning of acclimatization and at the end of each study.
In Study 1, body weights were recorded on days -16, -9, -2 and 37. In Study 2, body weights were recorded on days -15, -9, -2 and 37 for Groups 1 and 2; and on days -16, -10, -3 and 36 for Groups 3 and 4. Body weight change was calculated as the change in weight from baseline (weight closest to dosing was used as baseline weight) to weight at the end of the study and was analysed in the first study using an analysis of covariance containing the model effects “Treatment Group” and “Baseline”. In Study 2, this was analysed using an analysis of variance (ANOVA) with the fixed effect of “Treatment Group” and the random effects of “Block” and “Room”.
French translation of the Abstract is available in Additional file 1.