December 2025 Journal Review

Hirotaka Tomiyasu, DVM, PhD

Sridhar M Veluvolu, Robert B Rebhun, Jaeyoung Kim, Luke A Wittenburg

8 Oct 2025; 2(1):24 https://link.springer.com/article/10.1186/s44356-025-00038-z#change-history

Doxorubicin (DOX) is used for many canine malignancies, and dosing of DOX is based on patient body surface area (BSA) as it is for many other chemotherapeutic drugs. However, BSA-based dosing cannot account for several key factors including drug distribution, metabolism and excretion, and newer dosing schemes are needed to better predict drug exposure and limit patient toxicity. In human medicine, pharmacokinetic (PK)-guided dosing has been shown to reduce inter-individual variability in the degree of bone marrow suppression and to improve objective response rates. A previous study developed a limited sampling (LS) model to predict DOX PK in canine patients, and it was shown to accurately predict area-under-the drug concentration-time curve (AUC), which was correlated with neutrophil count at the seven-day nadir after one dose of DOX. The two main objectives of this study were: - to compare the within- and between-patient variability in DOX exposure (AUC predicted by LS model) after three consecutive doses of DOX. - to validate the ability of the LS model to predict nadir neutrophil count and to investigate potential correlations between DOX exposure and gastrointestinal (GI) adverse events (AEs).

Prospective, single-center study in client-owned dogs.

This study included any canine patients with a cytologic or histologic diagnosis of cancer susceptible to DOX with favorable performance status. The inclusion criteria also included body weight of at least 15 kg, but one dog with body weight of 7.5 kg was eventually included. Exclusion criteria included an absolute neutrophil count ≤ 2,000 /µl, platelet count ≤ 100,000 /µl, bilirubin greater than or equal to the upper limit of reference, serum creatinine greater than or equal to the upper limit of reference, any patient with known or suspected cardiac systolic dysfunction or any other significant underlying cardiac disease, and any breed predisposed to MDR-1 gene mutations.

All dogs initially received a standard dose of DOX (30 mg/m2 or 1 mg/kg) by a 20-minute intravenous infusion. Blood samples were obtained at 5-, 45-, and 60-min post-infusion, and the serum concentrations of DOX were measured by liquid chromatography tandem-mass spectrometry assay to predict AUC based on LS model. Dose-normalized predicted AUC was also calculated by dividing the predicted AUC by mg of DOX. Total DOX dose (mg) received, BSA (m2) of the patient at each dose, and mg/m2 per dose were recorded. CBCs were performed before each dose of DOX and seven days following each dose. Hematocrit, total white blood cell count, total neutrophil count, and platelet count were collected. Surviving fractions were calculated by dividing the measured number of respective cells at seven days by the measured number of cells at baseline. A previously reported quality-of-life form was provided to owners to fill out to assess GI AEs after each dose of doxorubicin. The average coefficient of variation (CV) for AUC between treatment cycles was calculated for each patient. The degree of intrapatient and interpatient variability (%CV) was compared using a two-tailed Welch’s t-test following determination of normal distribution of data by the Shapiro-Wilk test. Least squares linear regression analysis was performed on predicted AUC against measured hematologic parameters. Spearman correlation was performed on predicted AUC versus GI toxicity scores. Correlation and linear regression analysis was additionally performed only for the subset of dogs whose measured nadir neutrophil counts were lower than baseline. For all analyses, p-values less than 0.05 were considered significant.

Ten dogs were included in this study, and two small dogs were prescribed a starting DOX dose of 1 mg/kg. Three dogs only received one dose of DOX, and data from these dogs were only included in the analysis of interpatient variability in AUC of the first dose. Three neutropenic events were captured at seven days, two of which occurred during the first cycle of DOX and one during the second cycle in a patient that also experienced neutropenia during the first cycle. One patient who experienced grade 4 neutropenia after dose one received a 20% dose reduction for subsequent doses. There was nearly a 10-fold variation between patients in measured and dose-normalized DOX concentrations at collected time points. No correlation was found between total mg dose of DOX and predicted AUC using the developed LS model, although two out of three neutropenic events occurred at the highest predicted AUCs. The interpatient CVs for AUC (22.5%) appeared to be larger than intrapatient CVs (10.7%) before dose-normalization, although no significant difference was found. After dose normalization, a significant difference was noted; the average dose-normalized intrapatient and interpatient CVs were 4.7% and 25.4%, respectively. No significant relationship was found between predicted AUC and absolute lymphocyte count, surviving fraction of lymphocytes, absolute platelet count, surviving fraction of platelets, absolute neutrophil count, or surviving fraction of neutrophils. Since the dogs whose day 7 neutrophil count was higher than baseline likely had an unobserved nadir, two data points with a surviving fraction of neutrophils of >1 were then removed. After removal, a significant relationship was found between AUC and surviving fraction of neutrophils for only those patients with day 7 neutrophil counts lower than baseline. The previously developed LS model could predict the absolute nadir neutrophil count by utilizing the baseline neutrophil count and predicted AUC. In the present study, this model could predict absolute nadir neutrophil count for the data points with a surviving fraction of neutrophils < 1 with a mean prediction error of 3.3% and mean absolute prediction error of 41%. Regarding GI AEs, a significant correlation for higher AE scores and predicted AUC was observed for decreased appetite and increased nausea, but not for vomiting or diarrhea. Non-significant trends were noted for average GI AE score and total GI AE score with predicted AUC.

- Interpatient variability in predicted AUC across three cycles is significantly higher than intrapatient variability. - Increased predicted AUC was significantly associated with decreased surviving fraction of neutrophils in patients who had a decreased absolute neutrophil count at seven days compared to baseline. - This study validated a previously established LS model of DOX AUC to predict absolute neutrophil count at the nadir in patients who had a decreased absolute neutrophil count at seven days compared to baseline. - Increased predicted AUC was correlated with increased owner reporting of reduced appetite and nausea.

- Small number of included dogs. - Inherent owner bias in scoring subjective GI signs such as appetite and nausea. - No specific DOX administration dose (1 mg/kg or 30 mg/m2) or schedule (q2 weeks, q3 weeks, or multi-agent) was required for patient enrollment. - The occurrence of neutropenia may have been underestimated, as CBCs were performed only 7 days after DOX administration.

Although this is a preliminary study performed in a single center, the findings suggest the need to establish PK-guided dosing of DOX and other chemotherapeutic drugs to optimize the treatment outcomes.