Toxicity and Efficacy of Cisplatin and Doxorubicin Combination Chemotherapy for the Treatment of Canine Osteosarcoma

Introduction

Appendicular osteosarcoma (OSA) is the most common primary bone tumor in dogs.^1–5 Radiographically detectable pulmonary metastatic disease is present in <10% of affected dogs at the time of diagnosis; but with surgery alone, >90% of these dogs are dead from metastatic disease within 150 days of diagnosis.^2,4 The current standard of care for OSA involves surgical amputation or limb-salvage procedures followed by cisplatin, carboplatin, or doxorubicin chemotherapy.^1–3,6–10 Reported median survival times of dogs treated with surgery and any of the aforementioned chemotherapeutic agents range from 262 to 413 days. More than 80% of affected dogs are dead from metastases within 2 years of diagnosis.^{1–3,6,9–13}

Combination chemotherapy theoretically is more effective in the management of neoplasia than is single-agent chemotherapy.^10,11 Utilizing drugs with different mechanisms of action can maximize tumor cell death and minimize the development of resistant cell populations. Previously studied combinations of cisplatin and doxorubicin for the treatment of canine OSA have varied from alternating the drugs every 3 weeks to administering both drugs within a 24-hour period.^9,12,13 When doxorubicin and cisplatin were administered every 3 weeks in alternating doses, median survival times were similar to those for single-agent therapy.^{1–3,6–9,14,15} When combined into a protocol where both drugs were administered on the same day for a total of three cycles, the 1-, 2-, and 3-year survival rates were 48%, 28.3%, and 18%, respectively.¹² Another report described administering four cycles of a similar protocol with a disease-free interval of 15 months and a median survival time of 18 months.¹³ A carboplatin and doxorubicin combination administered within a 24-hour period every 3 weeks has also been evaluated for the treatment of canine OSA, with a median survival time of 235 days for 24 dogs.¹⁶ The purpose of the current study was to retrospectively evaluate the toxicity and efficacy of limb amputation, combined with four cycles of cisplatin and doxorubicin given within 24 hours of each other for the treatment of canine OSA.

Materials and Methods

Entry criteria for this retrospective study included a histological diagnosis of appendicular OSA and no evidence of pulmonary metastatic disease on thoracic radiographs at the time of diagnosis. All dogs were free of concurrent disease on the basis of physical examination and a complete blood count (CBC), serum biochemical profile, urinalysis, and thoracic radiographs. All dogs underwent surgical amputation of the affected limb and combination chemotherapy consisting of cisplatin and doxorubicin administered on consecutive days every 21 days, with the intent of administering four treatment cycles. Cisplatin^a (50 mg/m² intravenously [IV]) was administered over 20 minutes, in conjunction with a previously described 6-hour saline-induced diuresis.¹⁷ Butorphanol^b (0.4 mg/kg intra-muscularly [IM]) was administered as an antiemetic 20 minutes before the cisplatin. Twenty-four hours after cisplatin administration, doxorubicin^c (15 mg/m² IV) was given as a slow bolus through an indwelling catheter. Hemograms were monitored immediately before and at 7 to 10 days after chemotherapy administration. Blood urea nitrogen (BUN) and creatinine concentrations and urine specific gravity were measured immediately before cisplatin administration.

Toxicities were categorized as gastrointestinal, myelosuppressive, or both. Toxicities were classified as mild, moderate, or severe [see Table]. Dose reductions or discontinuation of either drug occurred at the discretion of the clinician in charge of the case. Thoracic radiographs were performed at the time of the last treatment and every 2 to 3 months after completion of chemotherapy.

Disease-free interval was defined as the time from surgery to documentation of metastatic disease. Survival time was defined as the time from surgery to death or euthanasia because of tumor or treatment. The effect of postulated independent variables on measured outcomes was assessed using a Kaplan-Meier product limit survival analysis.^d Dependent variables included survival time in days and disease-free interval. Independent variables included tumor site, sex, chemotherapy timing (e.g., <14 days, >14 days, or starting before surgery), toxicity (severe versus not severe), protocol alteration, and serum alkaline phosphatase activity (above reference range was considered high; within the reference range of 22 to 116 U/L was considered normal). Survival times and disease-free intervals were significantly different among groups when the log-rank statistic yielded a P<0.05. Median survival time and disease-free interval were calculated for each independent variable.^d

Results

Thirty-five dogs were entered into the study. Age at diagnosis ranged from 2 to 13 years, with a median of 8 years. There were 19 spayed females, 14 castrated males, and two intact males. The breeds included rottweiler (n=7), greyhound (n=6), Labrador retriever (n=6), Doberman pinscher (n=3), golden retriever (n=2), Great Pyrenees (n=2), and one each of Alaskan malamute, Saint Bernard, Old English sheepdog, German shorthaired pointer, Irish setter, German shepherd dog, boxer, and mixed-breed dog. Tumor locations included the radius (n=10), femur (n=9), tibia (n=8), humerus (n=5), ulna (n=2), and scapula (n=1). All dogs had preoperative alkaline phosphatase values determined. Ten values were above the reference range for the laboratory utilized. Information regarding prior nonsteroidal anti-inflammatory drug (NSAID) or steroid administration was unavailable.

Chemotherapy treatment was initiated between days 1 and 42 after surgery, with a median time of 6 days. Two dogs had one cycle of chemotherapy administered before limb amputation. Sixteen dogs completed four treatment cycles of both drugs. In two dogs, the dose of doxorubicin was decreased to 12 mg/m² and 11.25 mg/m²; in one dog, cisplatin was decreased to 40 mg/m² because of toxicity. Cisplatin was discontinued in four dogs because of elevations in BUN and creatinine, concurrent with a low urine specific gravity. In one of these four dogs, the BUN and creatinine subsequently normalized. The BUN and creatinine did not deteriorate in the other three dogs, and renal values were not followed after completion of doxorubicin chemotherapy. Doxorubicin was discontinued in two dogs, because cardiomyopathy developed in one dog (the boxer), and anorexia developed in the other.

Treatment was discontinued in one dog after three treatment cycles because of metastatic disease. Treatment was discontinued in four dogs after two treatment cycles because of toxicity (n=3) or metastatic disease (n=1). Eight dogs only had one treatment cycle. Five of these had documented toxicity, which resulted in discontinuance of therapy. Three of these eight dogs died within 1 week of the first treatment. One dog was overdosed with chemotherapy and died despite supportive care. The other two dogs died at home with no veterinary intervention, and necropsies were not performed.

Out of a total of 126 treatment cycles, excluding the dog that was overdosed, gastrointestinal toxicity was mild after 15 cycles, moderate after six cycles, and severe after six cycles. Neutropenia was mild after one cycle, moderate after one cycle, and severe after two cycles. Only one episode of mild thrombocytopenia was documented.

Ten dogs were alive at the last follow-up; 22 dogs had died; and three were lost to follow-up. Of the 32 dogs with follow-up, 14 had no evidence of disease, 17 had suspected metastases, and disease status was unknown in one dog at the time of death. The sites of suspected metastases included the lungs (n=6), bones (n=5), soft tissues (n=1), lungs and bone (n=3), and lungs and soft tissue (n=1). The site of metastasis was not documented in one dog. Five animals were necropsied. Four had metastatic OSA confirmed, and splenic hemangiosarcoma was found in one dog. The dog with hemangiosarcoma was censored in the analysis of disease-free interval and survival.

The median survival time for all dogs enrolled in the study was 300 days [Figure 1]. Ten of the dogs survived 1 year. Median survival time for this group of dogs was 510 days. Median disease-free interval for all dogs enrolled in the study was 240 days. Prior to surgery, 10 dogs had elevated alkaline phosphatase values. Median disease-free intervals for dogs with normal and elevated alkaline phosphatase were not significantly different (P=0.714) at 188 and 330 days, respectively [Figure 2]. Median survival times for dogs with normal and elevated alkaline phosphatase were not significantly different (P=0.347) at 195 and 330 days [Figure 3].

All 35 dogs were included in the data set analyzed to determine the effects of tumor site, sex, chemotherapy timing, toxicity, protocol alteration, and serum alkaline phosphatase activity on survival time and disease-free interval. None of the parameters were significantly associated with survival. Of the listed parameters, only increased toxicity was significantly and negatively associated with disease-free interval.

Discussion

Studies of canine appendicular OSA treated with surgical amputation followed by platinum-based chemotherapy have reported <50% survival beyond 1 year.^1,3,7,15 Although those findings were not confirmed by the current study, the protocol described in this study has been associated with the longest disease-free interval and overall survival time reported to date for dogs with OSA.¹³ A larger study evaluating the efficacy of three cycles of a cisplatin-doxorubicin protocol reported similar survival times as those seen with single-agent protocols.^12,18 Significant toxicity was reported after the first combination chemotherapy cycle when it was administered 2 days after amputation.^12,18

The chemotherapeutic protocol used in this study was associated with substantial toxicity. Only 16 dogs completed four cycles of therapy. Six (17%) dogs were switched to a single-agent protocol because of toxicity, and two (5.7%) dogs died from presumed chemotherapy toxicity. While the cause of death was unknown in two of the dogs that died at home, myelosuppression and sepsis may have played a role. Eight (7.8%) of 102 dogs treated with a similar protocol in a prior report died or were euthanized because of severe myelosuppression.¹² A third dog in the current study was overdosed with chemotherapy and died. Four (11%) dogs had cisplatin discontinued because of renal toxicity. None of the dogs in this study had evidence of renal insufficiency before chemotherapy was started. In one study, dogs treated with the same diuresis protocol and 70 mg/m² of cisplatin had a 6.6% incidence of nephrotoxicity.¹⁹ In another study, the same dosage used with a 4-hour diuresis protocol had a 7.8% incidence of nephrotoxicity.²⁰ No nephrotoxicity was reported with earlier cisplatin and doxorubicin combination protocols.^12,13,18

Aside from renal toxicity, the side effects reported in this study mirrored those reported from a larger number of dogs.^12,21 When combining chemotherapeutic drugs, it is important to avoid overlapping toxicities. As single agents, both cisplatin and doxorubicin are associated with gastrointestinal toxicity, myelosuppression, and nephrotoxicity; however, doxorubicin is only associated with renal failure in cats.^21–23 Combination chemotherapy could cause a cumulative effect, resulting in more profound toxicities than seen with either drug alone. The use of NSAIDs in combination with cisplatin has been reported to increase nephrotoxicity.²⁴ Because this was a retrospective study, it was not known whether any of the dogs that developed nephrotoxicity were also given NSAIDs.

The concept of summation dose intensity has recently been reviewed in the veterinary literature.^16,18,22 Briefly, summation dose intensity allows for comparisons between chemotherapy protocols by mathematically manipulating doses into fractional dose intensities.²⁵ For example, when cisplatin is given as a single agent at 70 mg/m² every 3 weeks, it has a dose intensity of 23.3 mg/m² per week.²⁶ When administered at 50 mg/m² every 3 weeks (as in this study), it has a dose intensity of 16.6 mg/m² per week. As used in the protocol reported here, cisplatin had a fractional dose intensity of 16.6 divided by 23.3 or 0.71. Doxorubicin given at 30 mg/m² every 2 weeks has a dose intensity of 15 mg/m² per week. The dose of 15 mg/m² every 3 weeks, which was used in the current study, has a dose intensity of 5 mg/m² per week. Thus, the fractional dose intensity of doxorubicin in the study reported here was 0.33. Assuming that there is a linear relationship between dosage and response, the combined fractional dose intensity for the two drugs was 1.04, compared to a summation dose intensity of 1.00 for either drug alone. Based on these calculations, the combination protocol described here did not have a higher summation dose intensity over single-agent protocols. This fact may explain why survival times in the current study were not better than those with single-agent therapy. Alternatively, survival may have been affected by the fact that only 16 dogs completed all four treatment cycles.

Unlike in earlier studies, dogs with an elevated alkaline phosphatase did not have shorter disease-free intervals and survival times when compared to dogs with normal alkaline phosphatase.^27–29 Possibly some of the dogs had received prior NSAID or steroid therapy that induced alkaline phosphatase elevations, but this information was not available. Similarly, the specific alkaline phosphatase isoenzyme responsible for the elevations was not analyzed.

The previously reported prolonged survival time of dogs treated with four cycles of cisplatin and doxorubicin administered within a 24-hour period strongly supported further evaluation of this protocol for the management of canine appendicular OSA.¹³ The median survival time of 300 days reported here did not approach the previously reported median survival time of 540 days.¹³ While the current study contained a subset of dogs with prolonged survival, the median survival time was inferior to that reported with single-agent protocols.^2,3,6–8,26 This poor survival time may have reflected the fact that only 16 of 35 dogs completed four cycles of the protocol.

As reported in this study, the toxicity associated with the chemotherapeutic protocol was significant. The reason for the increased toxicity compared to the prior reports was unclear.¹⁵ Given the fact that the current study was not conducted in a standardized, prospective fashion, it is possible that factors such as drug interactions with concurrent pain medications or clinician variability in tolerance for the induced side effects may have impacted the grading of the toxicities for the drug combination protocol.

Conclusion

The combination chemotherapy protocol reported in this study was associated with significant toxicity and disease-free interval, and overall survival times did not exceed those reported for single-agent chemotherapy protocols. A prospective, randomized, clinical trial comparing single-agent cisplatin to the combination protocol described in this paper would resolve the discrepancies in toxicities, disease-free intervals, and survival times between this series of dogs and those reported previously by Chun et al. Until such a study is reported, the authors encourage frank client communications regarding the concern for toxicity versus the potential benefit of using this chemotherapy protocol for dogs with appendicular OSA.