• Open Access

A Pilot Phase II Study of the Efficacy and Biosafety of Doxorubicin Chemotherapy in Tumor-Bearing Equidae




The efficacy and biosafety of a previously established tolerable dosage of doxorubicin have not been established in horses.


To provide preliminary evidence of the efficacy of doxorubicin in tumor-bearing horses, explore drug pharmacokinetics profile, and estimate period of risk of exposure to drug residues.


Twelve horses with 37 tumors.


Treatment protocol included 6 treatments at 3-week intervals. Eight horses were uniformly treated at a dosage of 70 mg/m2 and 4 horses received 4 of 6 treatment cycles at 70 mg/m2. Clinical signs, tumor responses, and toxicoses were evaluated. Drug residue concentrations were quantitated in 3 horses receiving of 65, 70, and 75 mg/m2 by high-performance liquid chromatography with ultraviolet detection (plasma, feces) and liquid chromatography and tandem mass spectrometry (urine).


Thirty tumors, including lymphomas, carcinomas, sarcoids, and melanoma, were evaluated for efficacy. The overall response rate was 47% (95% CI, 28–65%). Doxorubicin was not found to be effective against melanomas. Lymphomas and carcinomas were most responsive. Pooled serum Cmax and half-life of doxorubicin were 121.3 ng/mL and 12.9 hours, respectively. There were no detectable residues in fecal samples up to 3 weeks after treatment and in plasma and urine after 2 and 3 days, respectively.

Conclusion and Clinical Relevance

This study provides preliminary evidence that single-agent doxorubicin at a dosage of 70 mg/m2 has a broad spectrum of activity. The risk of exposure to drug residues in plasma and feces was low. Direct contact with urine-contaminated wastes should be avoided for 2 days after treatment.


area under the plasma concentration–time curve for the first 24 hours


area under the curve from time 0 to infinity


drug clearance


maximum plasma concentration


dose-limiting toxicosis


lower limits of detection


maximum tolerated dose


terminal half-life


volume of distribution at steady state

Doxorubicin is one of the most clinically effective and widely used antineoplastic drugs in humans, dogs, and cats.[1-3] However, its therapeutic potential has not been explored extensively in horses. Dose-limiting toxicoses (DLT) of doxorubicin and target organs have been determined in a Phase I trial conducted in 17 horses with 34 localized or multicentric advanced tumors.1 DLT included reversible drug hypersensitivity and neutropenia. Other minor toxicoses included reversible hair loss and skin reactions including multifocal dermatitis, neutrophilic necrotizing, and ulcerative epidermitis and patchy crusting and scaling of the skin in areas of alopecia. Cardiac, gastrointestinal, and renal toxicoses common in other species were not observed. Cardiac toxicosis was monitored by sequential echocardiograms, serum CK, SDH, AST activities, and cTnI concentrations. Gastrointestinal toxicosis was monitored by weight loss, anorexia, diarrhea, gastrointestinal (GI) bleeding assessed by occult blood test, and BUN : Cr ratio. Renal toxicity was evaluated by serum concentrations of creatinine and BUN, and urinalysis.

For cytotoxic chemotherapy, the therapeutic margin is narrow and it is important to treat close to the MTD to evaluate efficacy. Maximum tolerated dose and safe dosage of doxorubicin have been established for use in tumor-bearing horses in a previous dose-ranging study. However, the wide range of doxorubicin dosage (40–85 mg/m2) and dosage escalation protocol did not allow evaluation of efficacy at the determined recommended dosage.1

The large quantity (1,127–2,900 mg)1 of doxorubicin administered to horses for treatment prompted consideration of health risks and environmental hazards associated with drug residues in body fluids or excreta. Guidelines on doxorubicin use and protection in horses1 have been based on protocols for handling cytotoxic drugs and related wastes in small animal medicine.[4] However, drug and metabolites residue concentrations in body fluids and feces depend on their pattern of distribution, metabolism, and elimination, which differs among species. Because the pattern of distribution and elimination of doxorubicin in dogs[5] and humans[6, 7] are different, it seems inappropriate to rely on protocols used in dogs and humans for handling body fluids and wastes in horses.

The hypothesis of this study was that doxorubicin is effective in horses with tumors, and an appropriate use protocol could minimize a potential biohazard risk. The goals of this pilot study were to explore antitumor activity in horses treated with an established safe dosage and provide preliminary information on pharmacokinetic variables and drug residues to assess potential occupational and environmental hazards.

Materials and Methods


Horses with advanced or multicentric tumors referred to the Equine Medical Oncology Clinic at the Veterinary Medical Teaching Hospital, UC Davis were enrolled in this study over a period of 58 months. Four of 17 horses from a previous Phase I study were1 enrolled in 14 months and the remainder horses in this study were enrolled over 22 months. Horses eligible for this study had histologically confirmed tumors. Horses with multiple lesions had biopsies of representative lesions, the other nontarget lesions were diagnosed on cytologic or clinical evaluation. Horses entered in the study had measurable disease, good general health, and no prior history of local or systemic treatment with corticosteroids or other antineoplastic agents. Inclusion criteria were a life expectancy ≥3 months, normal renal and hepatic function, and hematologic and cardiac parameters within normal limits. Echocardiogram parameters evaluated included left ventricular internal dimension at end-diastole, left ventricular internal dimension at end-systole, fractional shortening, interventricular septal wall thickness at end-diastole, interventricular septal wall thickness at end-systole, and aortic dimension. Horses were evaluated clinically and staged (tumor measurements, evaluation of regional nodes, and thoracic radiography and abdominal ultrasound examination as needed). Staging of cutaneous tumors was done by use of a modified tumor/node/metastasis classification system[8] in which the T category indicated maximum diameter of the tumor, N category indicated the extent of regional nodal involvement, and the M indicated the presence of distant metastases. Staging of horses with lymphoma was done by use of the World Health Organization system for clinical staging for lymphoma in domestic animals.[9] Protocol approval was obtained from the UCD Institutional Animal Care and Use Committee. Signed informed consent forms were obtained from owners.

Study Design

The study was planned with a Fleming 2-stage design. It allowed testing for efficacy or lack thereof at predefined thresholds.[10] In this study, the drug would be considered ineffective for a tumor type if the response (completer response [CR] + partial response [PR]) rate was ≤10%. In contrast, the drug would be considered effective if the tumor response rate was ≥50%. The design was optimized to minimize the sample size in the first stage of the study.[11] In horses with multiple tumors of the same (melanomas or sarcoids) or different histologic types, each tumor (N ≤ 5 per horse) was counted as an individual tumor. The study required the evaluation of 10 tumors of each type with the possibility of early termination after evaluation of 6 tumors if response rates were above or below the predetermined response rate thresholds. In the first stage of the study, 6 tumors of each type were evaluated. For each histologic type, the study was stopped for lack of efficacy if there were ≤1 (1 or 0) response in the first 6 tumors. If there were ≥3 responses, the study was stopped for sufficient activity against a particular tumor type. If there were 2 responses, 4 more tumors were added in a second stage for a total sample of 10 tumors. If ≥4 responses were observed in 10 tumors of the same type, the study was stopped for sufficient efficacy. If 3 responses were observed in 10 tumors of the same type, the drug was considered active, but not sufficiently active (response rate <50%) according to the predefined thresholds. This design provided 98% probability of correctly concluding that the drug is insufficiently active (α = 0.021) when its true response rate was ≤10%, and 82% probability of correctly concluding that the drug has a sufficient level of activity when its true response rate was ≥50%. Positive conclusions would provide a strong rationale for further testing of accurate estimates of efficacy.

Treatment Procedures

Twelve horses were included in this study. The dosage was calculated on an mg/m2 basis.1 Doxorubicin2 was drawn into 50 cc luer-lock syringes. The drug was administered by slow IV push through the injection port (Y-port) of the solution set of a jugular IV line at a slow, steady rate over a period of approximately 1 hour. As doxorubicin was instilled, normal saline solution (0.9% sodium chloride USP) was administered by IV drip at a rate of approximately 400–450 drops/min to ensure continued dilution (approximately 0.2 mg/mL at a rate of 5 mg/min) of the drug. Eight horses received 6 treatments at 3-week intervals at a dosage of 70 mg/m2. Four horses selected from a previous study were included in this study.[2] They received 4 of 6 treatments at a dosage of 70 mg/m2 and 2 doses at 65 or 75 mg/m2. These horses were included in this study because they received doses in the therapeutic range (65–75 mg/m2) and were needed for the evaluation of drug residues in that dosage range. In addition, they allowed for screening of therapeutic efficacy against a balanced distribution of tumor types because the majority of recently recruited horses had lymphomas.

Handling and administration of doxorubicin2, as well as disposal of related waste, were done according to guidelines on chemotherapy use in veterinary oncology as previously described.[2] To minimize hypersensitivity reactions, prophylactic medical treatment with diphenhydramine hydrochloride (1 mg/kg, IM) and flunixin meglumine (1 mg/kg, IV) was carried out 30 min before infusion.1 Diphenhydramine hydrochloride (0.5 mg/kg IM) and flunixin meglumine (0.5 mg mg/kg IV) were repeated every 4 hours as needed.

Tumor Response Evaluation

Tumor measurements were recorded before initiation of treatment and at 3-week intervals, and assessed by response evaluation criteria in solid tumors.[12] Tumor volume was calculated as (height × width × depth) × (π/6). Tumor measurements were made by use of calipers or ultrasound examination during the course of treatment and at 1 month, 3 months, and every 3 months thereafter or until relapse. A CR was defined as total reduction in measured tumor volume, and a PR was defined as ≥50% reduction in tumor volume that was maintained for a minimum of 4 weeks. Stable disease was defined as <50% reduction in tumor volume or <25% increase in tumor volume, and progressive disease was defined as ≥25% increase in tumor volume. Tumor response included CR and PR. In horses with multiple lesions excluding lymphomas, up to 5 lesions were identified as independent target lesions based on accessibility and ability to reproduce measurements; a minimum diameter of 10 mm was required. Response was measured clinically during follow-up or until progressive disease or study withdrawal. The duration of CR was defined as the time from achievement of response until last disease-free follow-up or relapse. The duration of PR or stable disease was defined as the time from completion of treatment to subsequent progression of disease.

Drug Residue Sampling

Blood, urine, and fecal samples for pharmacokinetic analysis were obtained from 3 horses receiving dosages of 65, 70, and 75 mg/m2. Samples were analyzed in 2 horses evaluated for 2 consecutive cycles (65 and 70 mg/m2; 70 and 75 mg/m2) and in 1 horse for 1 cycle (70 mg/m2). An indwelling catheter was inserted in the jugular vein, and blood samples were collected into heparinized tubes before and at the end of infusion, and at 5, 10, 15, 20, 30, and 60 minutes and 2, 3, 4, 8, 12, and 24 hours, and 21 days after the end of the doxorubicin infusion. Samples also were collected at additional time points when possible. Samples were centrifuged, and plasma was harvested and frozen at −80°C until analysis. Urine and fecal samples were scheduled to be collected in the same 3 horses on their first chemotherapy cycle (65, 70, and 75 mg/m2) before and at the end of drug administration and approximately 4, 24, 48, and 72 hours, and 21 days after administration. Urine samples were collected by bladder catheterization and fecal samples were obtained by rectal palpation during the 24 hours the horse was hospitalized and 21 days after administration before the next cycle. The 48 and 72 hour samples were obtained by collection of voided urine and collection of fresh floor fecal samples in the stall or by rectal collection. Samples were refrigerated and then stored at −80°C until processed. Coded samples were sent overnight to and analyzed at the Pharmacology Core Facility, Colorado State University.

Measurements of Doxorubicin and Doxorubicinol Concentrations

Doxorubicin hydrochloride,3 doxorubicinol,4 and daunorubicin hydrochloride5 (as internal standard) stock solutions were prepared in methanol. Biological sample preparation was achieved by methanol protein precipitation (plasma) and liquid-liquid extraction (urine and feces).

Doxorubicin and doxorubicinol concentrations were measured in plasma and fecal samples by high-performance liquid chromatography (HPLC)–ultraviolet (UV) assay. The HPLC system consisted of 2 liquid chromatography pumps6 operated in binary mode and an autosampler.7 Separation was attained through a HPLC column8 with a guard cartridge.9 The mobile phase consisted of an aqueous component of 15 mM sodium phosphate, pH 4.0, and an organic component of acetonitrile. Drug concentration was quantitated by fluorescence5 with an excitation wavelength of 475 nm and an emission wavelength of 580 nm. Standard plasma pharmacokinetic variables were assessed and included maximum plasma concentration (Cmax, ng/mL), area under the plasma concentration–time curve for the first 24 hours (AUC0→24hr, ng/mL/min), area under the curve from time 0 to infinity (AUC0→∞, ng/mL/min), terminal half-life (T1/2, hours), volume of distribution at steady state (Vdss, L/kg), and drug clearance (CL, mL/min/kg).

Doxorubicin and doxorubicinol were measured in urine by liquid chromatography and tandem mass spectrometry as previously described[13] and optimized for equine biological samples. Positive ion electrospray ionization mass spectra were obtained with a mass spectrometer10 with a turbo ionspray source interfaced to a binary pump system.11 Separation was attained through a HPLC column12 with a guard cartridge13 and maintained at 40°C. The mobile phase consisted of an aqueous component of 0.1% formic acid and an organic component of acetonitrile.

The methods were validated for linearity, accuracy (recovery), precision, limit of detection (LOD), and limit of quantitation (LOQ) of the analytes as previously described.[13]

Treatment Toxicity

Although treatment toxicosis was not a primary end point, hematologic and nonhematologic adverse events were graded as previously described.1 Cardiac evaluation included a physical examination, auscultation, and echocardiogram (ECHO) performed at our institution was done before treatment and scheduled before cycle 5. On the day of treatment, each horse had a thorough history and physical examination, complete blood count (CBC), and serum biochemical profile as needed to determine any clinically adverse effects of drug treatment. Between treatments, blood tests were done 10–12 days after each treatment unless the horse experienced grade 2 or 3 hematologic toxicity at the previous cycle. There was no dose reduction in horses experiencing grade 2 reactions, but CBC was done on a weekly basis after administration. There was a dose reduction in horses experiencing grade 3 reactions and CBC was done on a weekly basis after administration.

Statistical Analysis

Optimal sample sizes for stage I and II of the 2-stage design as well as type I error (α) and type 2 error (β) were calculated by use of a dedicated software.14 Serum doxorubicin pharmacokinetic parameters were calculated by noncompartmental analysis.[14] Estimates of response duration were computed by use of the product-limit method.[15] Computations were performed by statistical software.15



Twelve horses with 37 measurable tumors were included in this study. Mean and median ages were 13.5 and 13 years, respectively (range, 5–22 years), and the male : female ratio was 1. There were 6 geldings and 6 mares. Breeds included Quarter Horse (n = 4), Thoroughbred (4), Arabian (2), American Paint Horse (1), and Appaloosa (1). Weights ranged between 443 and 600 kg (mean, 509; SD, 70). Histologic tumor types included lymphoma (N = 6), carcinoma (6), melanoma (13), and sarcoid (12). Four horses had 2 different tumor types. There were 2 T-cell-rich B-cell lymphomas (stage IIIa, and IIa), 2 T-cell and histiocyte-rich B-cell lymphomas (stage IIa and IVb), and 2 T-cell lymphoma (both stage IVb). There were 1 anaplastic carcinoma (stage T4N2M1, and T3N2M0) and 5 squamous cell carcinomas (3 stage T3N0M0, 2 stage T3N2M0) in 4 horses. There was 1 SC anaplastic melanoma (stage T4N2M1) in a nongraying QH. A graying Arabian horse had 2 noncutaneous melanomas affecting the parotid (T3N0M0) and mammary gland (T3N1M0) and also had multiple verrucous sarcoids (6 lesions, T2N0M0). A graying TB horse had dermal melanomatosis (4 largest measurable lesions) affecting the tail base and penis (stage T3N0M0) and also had T-cell and histiocyte-rich B-cell lymphoma. There were 2 horses each with a fibroblastic (stage T3N0M0) and a target verrucous (T2N0M0) sarcoid. One horse had multiple verrucous sarcoids (N = 8 lesions, 4 target lesions T2N0M0).


Seventy-two chemotherapy cycles were administered to 12 horses. All horses received 6 treatment cycles. No death caused by treatment-related toxicity occurred. Most horses received a dosage of 70 mg/m2. In the 4 horses pooled from a previous study, dose modifications were attributable to grade 2 hypersensitivity reactions that precluded dose escalation, but allowed repeat of the same dose. Two horses were treated with 65 mg/m2 of doxorubicin on their first 2 cycles and the following 4 cycles at a dosage of 70 mg/m2. One horse received the first 4 cycles at a dosage of 70 and 75 mg/m2 on the 5th cycle, and 80 mg/m2 on the 6th cycle. One horse received the first 2 cycles at a dosage of 70 mg/m2, 75 mg/m2 on the 3rd cycle, 70 mg/m2 on the 4th and 5th cycles, and 75 mg/m2 on the 6th cycle. In the 8 horses scheduled to receive 6 cycles at 70 mg/m2, 2 horses had dosage reduction to 65 mg/m2 at cycle 3 and 4 for grade 3 drug hypersensitivity. In these horses, the dose was re-escalated with no protocol changes to 70 mg/m2 at the following cycle without grade 3 reactions. Total doses ranged from 2,171 to 2,906 mg (mean, 2,603 mg; median, 2,576 mg). On mg/m2 basis, cumulative doses ranged from 410 to 435 mg/m2 (mean, 419.5 mg/m2; median, 420 mg/m2). On an mg/kg basis, doses ranged from 0.84 to 0.96 (mean and median, 0.9 mg/kg). Diphenhydramine hydrochloride (0.5 mg/kg IM) and flunixin meglumine (0.5 mg mg/kg IV) were administered after doxorubicin infusion once after 5 treatments, twice (q4h) after 56 treatments, and 3 times (q4h) after 11 treatments.


There were no hematologic toxicosis >grade 2 neutropenia seen after any cycles (Fig 1). The most common nonhematologic toxicoses were grade 1 (N = 47), grade 2 in 6 horses (1 at cycle 2, 3 at cycle 4, 2 at cycle 5 and 6) and grade 3 in 5 horses (1 at cycle 3, 3 at cycle 4, and 2 at cycle 6) adverse drug reactions. Only 7 horses had ECHOs rechecked before cycle 5. In 8 horses, hepatic and renal toxicoses were evaluated by serum biochemistry analysis, and urinalysis was done before cycle 4 only. Although other nonhematologic toxicoses were not monitored consistently, the tests obtained did not identify any drug-related changes.

Figure 1.

Graph of absolute neutrophil count/μL (mean-SD) in horses receiving 6 treatment cycles at 70 mg/m2. Time 0 is baseline before the first infusion, arrows represent infusion times.

Tumor Responses

Thirty target tumors were evaluated for efficacy. Because several horses were affected by multiple and different tumor types, tumor response rates were calculated based on response of (1) individual target tumors and (2) tumor types. Except for lymphomas, all lesions of the same histologic type in a patient were counted as up to 5 target tumors and response measured as described previously.[16] Mean and median follow-up durations of horses after treatment completion were 19 and 11 months (range, 4–57 months), respectively.

Overall tumor response rate was 50% (95% CI, 16–84%). The response rate of lymphomas in 6 horses was 100%, with estimated mean response duration of 30 months (SE, 12 months). The 2 T-cell lymphomas had CR lasting 9 and 29 months, but both horses had tumor progression. The 4 B-cell lymphomas also had CR, 3 still disease-free 11, 18, and 57 months after treatment, whereas the fourth horse with extensive disease (Stage IVb) had a PR lasting 4 months before being euthanized because of disease progression. No additional lymphomas were accrued for stage II of the study because of sufficient activity.

The response rate of 6 carcinomas in 4 horses was 100% with an estimated mean response duration of 17 months (SE, 5 months). Four squamous cell carcinomas in 2 horses affecting eyelid, genitalia, perirectal area, and cheek had CR. One horse with 2 carcinomas on its rectum, penis, and sheath was disease-free 28 months after treatment completion. The other horse had a tumor involving the eyelid and perirectal area extending to the vulva. This horse had salvage surgery after local recurrence of the perirectal lesion 6 months after chemotherapy and was disease-free 12 months post surgery. Two horses had PR. Duration of PR was 6 months in 1 horse with a carcinoma involving the floor of the mouth metastatic to regional lymph nodes. Tumor progression was associated with widespread metastasis. The other horse with a vulvar carcinoma had surgical resection of residual disease 2 months after chemotherapy completion. The horse was euthanized 13 months later because of metastatic spread, but no local recurrence. No additional carcinomas were accrued for stage II of the study because of sufficient activity.

One PR was observed in the first 6 melanomas. This was a fast-growing anaplastic melanoma that had a PR lasting 5 months. One horse with noncutaneous melanomas affecting the parotid area gland and dermal melanomatosis (5 target lesions) affecting the penis had stable disease lasting 11 months. As a result, 4 additional melanomas were added in a second stage. Four target lesions were evaluated on the same horse that had noncutaneous melanomas affecting the mammary gland and dermal melanomatosis (4 target lesions) affecting the perineal area and tail base. The treated horse had stable disease lasting 9 months. As a result, only 1 response in 10 evaluable lesions was observed, which does not warrant further evaluation of doxorubicin for equine melanomas.

One PR was observed in 1 of 5 target verrucous sarcoids and 1 fibroblastic sarcoid in 2 horses. One horse had 4 verrucous sarcoids that remained stable in size for 14 months. The other horse with cutaneous verrucous and eyelid fibroblastic sarcoid had PR lasting 12 months; this horse also had a T-cell-rich B-cell lymphoma that was still in remission. The eyelid lesion was retreated with intratumoral chemotherapy with cisplatin as previously described.[8] The horse was disease-free at the last follow-up 12 months after retreatment of the eyelid and 24 months after treatment of the lymphoma. As a result, 4 more sarcoids were required in stage II. The next horse entered had a fibroblastic and verrucous sarcoid. The fibroblastic sarcoid (stage T3N0M0) was fast growing, affecting the flank, and had CR lasting 57 months. This horse also had a T-cell-rich B-cell lymphoma that was still in remission. Mean response duration was 27 months (SE, 17 months). Because of slow accrual for sarcoids and the number of responses, 3 of 8 target lesions (11 total), the study was stopped because the drug was shown to be active, but the expected response rate, likely between 10 and 50%, must be determined using a larger sample size.

Drug Residues in Biosamples

The lower LOD (LLOD) for both doxorubicin and doxorubicinol was 2.5 ng/mL in urine and 5 ng/mL in feces. The LLOD was 1 ng/mL for doxorubicin and 10 ng/mL for doxorubicinol in plasma. The lower LOQ (LLOQ) was 25 ng/mL in urine, 2.5 ng/mL in plasma, and 10 ng/mL in feces. Accuracy of the standard curves was within 15% for urine, 10% for plasma, and 15% for feces.

Comparison of values for pharmacokinetic variables in 3 horses that received 65, 70, or 75 mg/m2 in this study indicated similar plasma concentration–time curves (Fig 2). When all samples were pooled, serum Cmax and half-life of doxorubicin were 121.3 ng/mL (SD, 46) and 11.7 hours, respectively. Plasma pharmacokinetic parameters assessed are given in Table 1. Mean Cmax were 108.6 and 140.3 ng/mL after the first and second cycles, respectively. Mean AUC0→24hr were 4,486 and 6,487 ng/mL∙min after the first and second cycles, respectively. Mean AUC0→∞ were 7,683 and 11,212 ng/mL∙min after the first and second cycles, respectively. Doxorubicinol was detected only in 1 horse that received 70 mg/m2 in the first cycle and 75 mg/m2 in the second cycle. Cmax were 73.7 and 46 ng/mL after the first and second cycles, respectively. AUC0→24hr were 10,463 and 2,734 ng/mL∙min after the first and second cycles, respectively.

Table 1. Pharmacokinetic variables assessed after first and second administration of doxorubicin (§ and represent samples from horses treated at 2 consecutives cycles).
Analysis TimeCmax (ng/mL)AUC0→24hr (ng/mL/min)AUC0→∞ (ng/mL/min)Terminal T1/2 (hr)Vss (L/kg)CL (mL/min/kg)
After 1st treatment
After 2nd treatment
Figure 2.

Plasma doxorubicin concentration–time relationship in tumor-bearing horses treated with IV administration of doxorubicin at dosages of 65 mg/m2 (first 2 cycles in the same horse), 70 mg/m2 (first cycle in 2 different horses), and 75 mg/m2 (third cycle in 1 horse). Data are presented as mean ± SD.

Because urine color reflected high concentrations of drug and may have represented a health hazard during collection, urine doxorubicin concentration at the end of infusion was measured only in 1 horse. Urine doxorubicin and doxorubicinol concentrations in this sample were 2,430 and 1,630 ng/mL, respectively. After a dosage of 65 mg/m2, urine doxorubicin concentrations 4 hours post administration and on day 1 were 36.4 ng/mL and <LLOD, respectively. At the same dosage, urine doxorubicinol concentrations were 65.1 and 43.8 ng/mL, respectively. After a dosage of 70 mg/m2, urine concentrations of doxorubicin and doxorubicinol 4 hours post administration were 109 and 61.1 ng/mL, respectively. At the same dosage, urine concentrations of doxorubicin and doxorubicinol on day 1 were LLOD and 9.82 ng/mL, respectively, whereas urine concentrations of both analytes were LLOD on day 3. After a dosage of 75 mg/m2, urine doxorubicin concentrations 4 hours post administration and on day 1 and 3 were 230, 28.7, and 25.7 ng/mL, respectively. At the same dosage, urine doxorubicinol concentrations were 827, 135, and 120 ng/mL, respectively. Urine samples collected on day 10 (1 horse) and day 22 and 28 (2 horses, 2 cycles) after treatment had no measurable doxorubicin residues.

Detectable concentrations of doxorubicin or doxorubicinol were not found in the fecal samples at any collection time point after treatment.


The main aim of phase II cancer clinical trials is to evaluate the antitumor effect of a treatment, screening out agents that are insufficiently active and selecting active agents for future studies. Because of the proven efficacy of doxorubicin in veterinary[17, 18] and human medicine[19], a 2-stage Fleming Phase II design was used to minimize expected sample sizes by stopping accrual for lack of efficacy, but also for sufficient activity.[10] The number of patients in the first screening stage was minimized according to Bayesian decision theory.[11] The study design allowed for simultaneously testing the hypothesis that the response rate was less than or equal to the specified ineffective treatment response rate set at 10% and the alternative hypothesis that the response rate was greater than the effective response rate set at 50%. This approach saves drug development time, brings useful treatments into clinical practice more quickly, and minimizes the number of patients used.

In this study, doxorubicin was found to be very active against lymphomas and carcinomas. The true response rate for these tumors is likely >50%. Doxorubicin appeared active against fast-growing fibroblastic sarcoids. However, the drug was less active on slow-growing verrucous sarcoids and resulted in growth control rather than tumor reduction. Stable disease in 4 verrucous sarcoids for 14 months was far greater than what would be expected if the horse had not been treated. As a result, the overall response rate for sarcoid tumors is likely to be between 10 and 50%. Further accrual of patients with these tumor types will be necessary to better estimate doxorubicin efficacy. Doxorubicin was not found active against melanomas and additional trials are not warranted.

The dose of doxorubicin as single-agent chemotherapy used in this study ranged from 2,171 to 2,906 mg (mean, 2,603 mg; median, 2,576 mg) based on a dosage of 70 mg/m2. On an mg/m2 basis, cumulative dosage ranged from 410 to 435 mg/m2. On mg/kg basis, dosages ranged from 0.84 to 0.96. This dose was comparable to that used in humans, both as determined on an mg/m2 and mg/kg basis.[7, 20] Although the sample size in this study did not allow accurate determination of the true efficacy of single-agent doxorubicin, the response rate confirms the efficacy of doxorubicin against lymphomas, carcinomas, and sarcomas in horses.[21] On the basis of the observed efficacy for horses with gross disease, doxorubicin is likely to have superior results in an adjuvant setting.

Adverse drug reactions (grade 3) were a dose-limiting toxicity after 6 cycles. Mild bone marrow suppression was common at 10–12 days after each treatment and a self-limiting adverse effect of chemotherapy with doxorubicin. Double nadir was not observed. Grade 3 drug reactions were seen after 6 cycles (<10% of cycles) and resulted in dosage reduction to 65 mg/m2. After a cycle at 65 mg/m2 without dose-limiting toxicity, dosage at 70 mg/m2 was resumed without grade 3 drug reactions. These results confirm the validity of the study design used in a previous phase I study to determine a safe dosage.1

The low Cmax and AUC and large clearance and volume of distribution calculated from plasma concentrations in 3 horses in this study indicated extensive drug distribution into tissues. Plasma drug concentrations are affected by the rate at which drug is administered, the volume in which it distributes, and its clearance. When compared with dogs receiving 30 mg/m2 administered IV over 20 minutes,[5] Cmax and AUC were approximately 10 times lower in the 3 horses in this study receiving 60–75 mg/m2. Conversely, Vss was approximately 3 times higher and CL more than 10 times higher in horses. Because the study was not designed to determine drug pharmacokinetics in horses, but rather provide safety guidelines, the pharmacokinetic parameters and failure to account for the total administered drug must be interpreted with caution. When compared with dogs and humans, the striking differences in pharmacokinetic variables may reflect a poor correlation between drug pharmacokinetics and drug dosages expressed in mg/m2 (Table 1). Finally, the preliminary pharmacokinetic findings may have reflected an inadequate formula of surface area in horses, resulting in an inaccurate dosage on mg/m2 basis.

The lack of doxorubicin-induced cardiotoxicity in this study and in a previous Phase I study may be a reflection of the low plasma concentration of doxorubicin and metabolites in horses. Doxorubicin-induced cardiotoxicity has been shown in humans to be dependent on Cmax.[22] The low Cmax of doxorubicin in the 3 horses evaluated relative to dogs[5] and humans[6, 7] may explain the lack of cardiotoxicity.

The anticancer activity of doxorubicin has been reported in humans to be dependent on the AUC.[22] Two (2 lymphomas) of the 3 horses that were evaluable for tumor response and had pharmacokinetics data had CR. These 2 horses had lower Cmax and AUC relative to dogs[5] and humans,[6, 7] but substantially more extensive tissue drug distribution into tissues potentially including tumor tissues. These findings may provide preliminary evidence that drug distribution may contribute more to the anticancer effects than the AUC and warrant additional studies.

The increased use of chemotherapeutics and large amounts of drugs used in equine medicine prompted consideration of health risks and environmental hazards associated with residues of these drugs in patient excretions. The present studies indicated that residues of doxorubicin and its metabolite, doxorubicinol, in serum, urine, and feces of treated horses were substantially lower than those in treated dogs or humans. There were no detectable residues in plasma 48 hours after administration. As a result, the risk of occupational hazards for veterinarians and laboratory personnel by collecting and handling blood samples ≥2 days after administration is minimal. There were no detectable residues in fecal samples at any time point after administration. This also has been reported in humans[23] despite the fact that biliary excretion is an important route of plasma clearance.[24] This could have resulted from extensive metabolism of the drug and metabolites by the intestinal flora in horses, poor recovery from the feces, or efficient reabsorption from the intestine during a process of enterohepatic circulation. As a result, the risk of exposure to owner and personnel handling feces with no direct contact appears to be negligible.

In this study, urine doxorubicin concentrations in horses receiving dosages up to 75 mg/m2 were below the range reported in dogs receiving 30 mg/m2. Residues in urine were not detectable in 2 horses (at dosages of 60 and 70 mg/m2) 2 days after treatment, whereas 1 horse (75 mg/m2) had detectable residues up to 3 days after treatment. Although the concentrations of doxorubicin and doxorubicinol in this horse were low and below urine concentrations in dogs receiving 30 mg/m2,[25] the large urine output may have resulted in a potential contamination risk. However, this risk is minimized by the rapid degradation of doxorubicin in urine when exposed to daylight (complete degradation in 4 hours) and artificial light (50% degradation in 3.2 hours).[26] In addition, owners and staff are not likely to come into close contact with a horse's excreta after treatment. As a result, a 2-day period of risk to owner, personnel, and public for doxorubicin and metabolites in equine urine appears reasonable. At any time during that period, daily excreta or soiled bedding should be kept on site for an additional 24 hours before disposal to allow drug degradation to nondetectable concentrations.

In summary, systemic chemotherapy with doxorubicin offers a potentially highly effective, affordable, and practical treatment option for horses with advanced cancers. Horse owners are aware of progress made in cancer treatment and are seeking treatment options for their horses when diagnosed with advanced or metastatic cancers. This study establishes a step toward developing effective chemotherapy regimens in equine oncology. The drug residue findings in this study provide information on which guidelines can be developed for safe handling and avoiding exposure risk for veterinary personnel, public, and the environment.


The authors thank Catherine Glines for her help with coordination of the clinical trial. Supported in part by the Cunningham and Doyle Charitable Trust Fund and the UC Davis Center for Equine Health.

Conflict of Interest: Authors disclose no conflict of interest.


  1. 1

    Théon AP, Pusterla N, Magdesian KG, et al. Phase I dose escalation of doxorubicin chemotherapy in tumor-bearing equidae. Accepted for publication, Journal of Veterinary Internal Medicine

  2. 2

    Adriamycin, Bedford Laboratories, Bedford, OH

  3. 3

    Doxorubicin hydrochloride, Sigma-Aldrich, St Louis, MO

  4. 4

    Doxorubicinol, Ventas Labs, Branford, CT

  5. 5

    Daunorubicin hydrochloride, Calbiochem- EMD, Billerica, MA

  6. 6

    Shimadzu LC-20AD Prominence, Shimadzu Corporation, Kyoto, Japan

  7. 7

    Shimadzu SIL-20AC Prominence autosampler, Shimadzu Corporation

  8. 8

    C18 HPLC Phenomenex C18 column, Phenomenex, Torrance, CA

  9. 9

    SecurityGuard C18 cartridge, Phenomenex

  10. 10

    Shimadzu RF-10AXL fluorescence detector, Shimadzu Corporation

  11. 11

    API 3200 triple quadrupole mass spectrometer, Applied Biosystems, Inc, Foster City, CA

  12. 12

    HPLC Agilent 1200 Series Binary Pump SL HPLC system, Agilent Technologies, Santa Clara, CA

  13. 13

    Waters Sunfire C8 column, Waters Corporation, Milford, MA

  14. 14

    Software developed by Dr A Ivanova, University of North Carolina at Chapel Hill, Department of Biostatistics, Chapel Hill, NC

  15. 15

    IBM SPSS Statistics, Somers, NY