Evaluation of dexamethasone as a chemoprotectant for CCNU-induced bone marrow suppression in dogs*
This information was presented in part at the Veterinary Cancer Society 27th Annual Conference, Ft Lauderdale, FL, USA, November 2007, and the American College of Veterinary Internal Medicine 26th Annual Conference, San Antonio TX, USA, June 2008.
Abstract
In mice and people, administering corticosteroids before chemotherapy can reduce the severity of myelosuppression without reducing antitumour effects. This study investigated whether pretreatment with dexamethasone would reduce the incidence of grade 4 neutropenia in dogs receiving CCNU. Twenty-five dogs received dexamethasone [0.1 mg kg−1 per os (PO) every 12 h] for 5 days and on the sixth day received CCNU (90 mg m−2 PO). Historical dogs (n = 67) received CCNU alone (90 mg m−2 PO). Forty-five percent of historical dogs had grade 4 neutropenia, while 64% of dogs pretreated with dexamethasone had grade 4 neutropenia (P = 0.16). Dexamethasone plasma levels were quantified by enzyme-linked immunosorbent assay in three healthy dogs. Peak plasma concentrations after a single oral 0.1-mg kg−1 dose were <80 ng mL−1, the minimum level associated with chemoprotective effects of dexamethasone in people. Pretreatment with dexamethasone did not reduce the incidence of grade 4 neutropenia in dogs receiving CCNU.
Introduction
Myelosuppression, specifically neutropenia, is an important dose-limiting toxicity of many chemotherapy agents. The destruction of cycling progenitor cells within the bone marrow results in a reduction in peripheral blood cell counts until previously quiescent progenitor cells and/or stem cells become reactivated and can replenish the depleted circulating pools.1 Chemotherapy-induced neutropenia is associated with increased risk of infection. In dogs, the risk of infection might be greatest when absolute neutrophil counts decrease to <500 cells μL−1;2 however, controlled studies to verify this are unavailable. Hospitalisation, chemotherapy treatment delays, chemotherapy dosage reductions, increased medical costs and an overall reduced quality of life are additional consequences of chemotherapy-induced myelosuppression.3
Several strategies have been developed to enhance recovery of haematopoietic cells, including postchemotherapy administration of haematopoietic growth factors, autologous bone marrow transplants and peripheral blood stem cell transplants.4–6 Such treatments are associated with considerable expense, have limited efficacy and availability for veterinary patients and do not eliminate chemotherapy-induced genetic damage to cycling progenitor cells. An alternative approach would be to protect haematopoietic cells before chemotherapy is given, thereby decreasing or preventing chemotherapy-associated myelosuppression.
In mice and people, the administration of dexamethasone before myelosuppressive chemotherapy reduces the associated haematologic toxicities without reducing antitumour effects. Pretreating mice for five consecutive days with dexamethasone (12 mg m−2 subcutaneous every 24 h) reduced fatal haematologic toxicity from a single dose of carboplatin (600 mg m−2 intraperitoneal) from 80 to 10%.1 In a pilot study, people receiving carboplatin-based chemotherapy that were pretreated for 5 days with dexamethasone (8 mg PO every 12 h) showed improvement in the average granulocyte nadir and more rapid recovery times to average granulocyte counts >1500 μL−1 and platelet counts to >100 000 μL−1.7 Recently, a phase 1–2 trial in people with nonsmall cell lung cancer showed that patients pretreated for 5 days with dexamethasone (8 or 16 mg PO every 12 h) before combination chemotherapy had significantly improved average granulocyte counts and platelet nadirs, and recovery times of both lines, when compared with patients that did not receive pretreatment with dexamethasone.8
The precise mechanisms by which pretreatment with dexamethasone reduce chemotherapy-associated haematologic toxicity are unknown. In preclinical studies, marrow colony forming units of granulocytic/monocytic lineage (CFU-GM) derived from mice pretreated with corticosteroids had increased survival after in vitro exposure to cisplatin, 5-fluorouracil and methotrexate compared with CFU-GM derived from mice treated with chemotherapy alone.1,9 Pretreatment with corticosteroids has also been shown to decrease the fraction of splenic and marrow CFU-GM that are in S phase, thereby potentially reducing the number of haematopoietic progenitors at risk for chemotherapy-induced damage.1 The protective effects of corticosteroids are dose and schedule dependent. In people, single-dose peak plasma levels of dexamethasone >80 ng mL−1 are required.7 Administration of dexamethasone after chemotherapy has not been shown to be beneficial.9,10
CCNU is a noncell cycle specific alkylating chemotherapeutic agent in the nitrosourea subclass that is effective in the treatment of canine lymphoma, mast cell tumour and histiocytic sarcoma.11–17 CCNU is capable of causing severe haematopoietic toxicity in dogs. At a dosage of 90 mg m−2, the incidence of grade 4 (<500 cells μL−1) neutropenia is reported to be up to 31%.18,19 Overall, the incidence of febrile neutropenia in dogs treated with CCNU at this dosage might be as high as 29%, and nearly, all episodes of fever are observed when neutrophil counts are <500 cells μL−1.12
The purpose of this study was to evaluate the chemoprotective effects of dexamethasone in tumour-bearing dogs treated with CCNU. The null hypothesis was that pretreatment with dexamethasone would not reduce the frequency of severe neutropenia in dogs prospectively given CCNU compared with an historical group of dogs treated with CCNU without dexamethasone pretreatment. Accepting the alternative hypothesis could lead to use of dexamethasone as a chemoprotective agent in dogs treated with CCNU.
Materials and methods
Study design
The frequency of grade 4 neutropenia (<500 cells μL−1)20 in the historical group of dogs was 45%. This study was designed to test the hypothesis that pretreatment with dexamethasone would reduce the frequency of grade 4 neutropenia by 50% (one-sided α error = 0.05 and β error = 0.80). A minimum of 74 dogs in the prospective group was required to fulfil this criteria.21
Animals
This project was approved by, and adhered to the guidelines of, the Institutional Animal Care and Use Committee of Cornell University. Client-owned dogs with spontaneous cancers were prospectively enrolled. To be eligible, dogs were required to have histologically or cytologically confirmed neoplasia for which CCNU would be a reasonable treatment. Dogs also needed to have adequate neutrophil (>3000 cells μL−1) and platelet counts (>100 000 μL−1), normal hepatic function and a life expectancy greater than 30 days without treatment. All owners provided informed consent before enrolment.
Historical control group (CCNU alone)
A search of the database of the Cornell University Hospital for animals identified 67 dogs treated with a single dose of CCNU (CeeNu; Bristol Laboratories, Princeton, NJ, USA) (90 mg m−2 PO) between March 2002 and September 2006. After treatment with CCNU, dogs received prophylactic antibiotics (trimethoprim–sulfadiazine (Tribrissen; Interfarm, Hauppauge, NY, USA), 15 mg kg−1 per os (PO) every 12 h for 14 days). Complete blood counts (CBCs) were obtained 7 days after administration of CCNU. Signalment, disease, previous chemotherapy, date of treatment with CCNU, neutrophil nadir, frequency of febrile neutropenia and outcome were recorded.
Prospective group (Dex + CCNU)
Pretreatment evaluation included physical examination, tumour measurements (when applicable), CBC, chemistry profile, urinalysis, and when appropriate, thoracic radiographs, abdominal ultrasonography and/or bone marrow cytology. Eligible dogs then received dexamethasone (Boehringer Ingelheim, Columbus, OH, USA) (0.1 mg kg−1 PO every 12 h) for five consecutive days and on the sixth day received a single dose of CCNU (90 mg m−2 PO). This dosage of dexamethasone was based on the dosage found to be effective in people (8 mg) and was calculated using allometric scaling from an average-sized person weighing 75 kg to an average-sized dog weighing 32 kg.22 The dose of CCNU was delivered to the nearest 5 mg using standard 40- and 10-mg capsules and reformulated 5-mg capsules. This group of dogs also was treated with prophylactic antibiotics, using the same regimen as dogs in the CCNU alone group. A CBC was obtained 7 days after administration of CCNU. Signalment, disease, previous chemotherapy, date of treatment with CCNU, neutrophil nadir, frequency of febrile neutropenia and outcome were recorded as for dogs in the CCNU alone group.
Measurement of plasma concentrations of dexamethasone
Three healthy mixed-breed dogs (age range 3–7 years; body weight range 36.9–41.4 kg) were used to measure plasma dexamethasone levels. The dogs were fasted for at least 12 h before collection of blood samples. Peripheral sampling catheters were placed in the lateral saphenous vein and a blood sample was obtained (time = 0). The dogs were administered a single oral dose of dexamethasone (0.1 mg kg−1 PO). Three millilitres of blood were collected from the sampling catheters into heparinised tubes at the following time points: 15, 30, 40, 50, 60, 70, 80, 90, 100, 110, 120, 180, 240, 360, 480 and 720 min after dexamethasone was given. Blood tubes immediately were transferred to ice. Samples were centrifuged (10 min, 3500 g) and plasma was separated and stored at −70 °C until pharmacokinetic analyses were performed. Plasma concentrations of dexamethasone were measured using an enzyme-linked immunosorbent assay (Enhanced Dexamethasone Kit; Neogen Corp., Lexington, KY, USA) as previously described and in accordance with the manufacturer’s instructions.23 All time points for each dog were measured on the same sample plate and all samples were measured in duplicate, with positive and negative controls. A standard curve was generated using serial dilutions of a 10-mM stock solution of dexamethasone (BioVision, Mountain View, CA, USA). A 1:10 dilution was necessary to quantify plasma dexamethasone concentrations up to 10 ng mL−1, ensuring that measured concentrations were within the linear portion of the standard curve. Dexamethasone concentrations in the samples were measured with a microplate reader using dual wavelength settings (Safire Multi-Detection Monochrometer Microplate Reader; TECAN, Männedorf, Switzerland) (W1 at 650 nm and W2 at 490 nm).
Data analysis and statistical methods
To verify baseline comparability between dogs prospectively treated with dexamethasone before CCNU and the historical group of dogs treated with CCNU alone, dogs in each treatment group were compared with respect to age (continuous), body weight (continuous), disease (categorical) and prior chemotherapy (categorical). Since lymphoma was the tumour type most likely to infiltrate the bone marrow severely enough to potentially adversely affect the ability to recover from cytotoxic chemotherapy, the proportion of dogs with lymphoma versus other neoplasias was compared. The primary outcome analysed was the frequency of grade 4 neutropenia between the treatment groups. The frequency of febrile neutropenia was a secondary outcome analysed. Comparisons of categorical variables between dogs in each treatment group were made using the chi-square test of independence or the Fisher’s exact test when cell values were less than 5. Continuous variables were compared by the Wilcoxon rank sum test (for non-Gaussian data) or the Student’s t test (for Gaussian data). All statistical analyses were performed by use of a software package (SPSS, version 10, Chicago, IL, USA). For statistical analysis, two-sided values of P ≤ 0.05 were considered significant.
Results
Population characteristics
Of the 67 dogs in the CCNU alone group, 34 were diagnosed with multicentric lymphoma, 22 with mast cell tumours, 9 with disseminated histiocytic sarcoma and 1 each with acute lymphoblastic leukaemia and metastatic oral melanoma. Twenty-five dogs were enrolled in the Dex + CCNU group between September 2006 and July 2007. Fifteen were diagnosed with multicentric lymphoma and 10 with mast cell tumours. Dogs in the CCNU alone group were comparable to dogs in the Dex + CCNU group in regards to age (P = 0.45), weight (P = 0.79) and disease (P = 0.49). Characteristics of the dogs in this study are presented in Table 1.
| Treatment group | |||
|---|---|---|---|
| Dex + CCNU (n = 25) | CCNU alone (n = 67) | P | |
| Age (years) | |||
| Median | 9.3 | 8.6 | 0.45 |
| Range | 3.9–13.6 | 2.6–15.1 | |
| Weight (kg) | |||
| Median | 29.5 | 29.8 | 0.79 |
| Range | 8.8–58 | 3.3–62.8 | |
| Tumour type, n (%) | |||
| Lymphoma | 15 (60) | 34 (51) | 0.49 |
| Other tumour | 10 (40) | 33 (49) | |
| Previous chemotherapy, n (%) | |||
| Yes | 14 (56) | 58 (87) | 0.003 |
| No | 11 (44) | 9 (13) | |
Prior chemotherapy
Dogs in the CCNU alone group were significantly (P = 0.003) more likely to have received other chemotherapy before CCNU. Fifty-eight of the 67 dogs (87%) in the CCNU along group had received chemotherapy before CCNU (median number of drugs, 3; range, 1–10 drugs). Eighteen dogs received vinblastine (median number of doses, 2; range, 1–4) and one dog received an investigational monoclonal antibody (four doses). Two dogs received four doses of vinblastine followed by daily hydroxyurea, one dog received one dose each of l-asparaginase and vincristine, one dog received cisplatin (one dose) and carboplatin (two doses) and one dog received vinblastine (three doses) and cyclophosphamide (one dose). One dog received one dose each of vincristine, cyclophosphamide and doxorubicin. Thirty-one dogs received l-asparaginase, cyclophosphamide, doxorubicin and vincristine (L-CHOP); 22 dogs received 10 weeks24 and 9 dogs received 4 weeks of L-CHOP. One dog received 4 weeks of L-CHOP and nine cycles of actinomycin-D, melphalan, cytosine arabinoside and dexamethasone. One dog received 35 weeks of L-CHOP where two doses of doxorubicin were substituted with mitoxantrone. The median time from the last chemotherapy dose to CCNU dose was 14 days (range, 5–533 days) for dogs in this group.
Fourteen of the 25 dogs (56%) in the Dex + CCNU group received previous chemotherapy (median number of drugs, 5; range, 1–10 drugs). Two dogs received vinblastine (one received one dose and one received two doses). One dog received one dose of doxorubicin and 6 months of daily chlorambucil. One dog received 3 weeks of treatment with CHOP. Nine dogs received L-CHOP; three received between 10 and 12 weeks and six received 4 weeks of L-CHOP. One dog received 2 years of L-CHOP with methotrexate,25 three cycles of mustargen, vincristine, procarbazine and prednisone, one cycle of mustargen, vinblastine, procarbazine and prednisone and one dose of mitoxantrone. The median time from the last chemotherapy dose to CCNU dose was 17 days (range, 7–20 days) for dogs in this group. Prior chemotherapy in both treatment groups is summarised in Table 2.
| Treatment group | No. of dogs | No. of prior drugs | |||||
|---|---|---|---|---|---|---|---|
| 0 | 1 | 2 | 3 | 4 | >4 | ||
| Dex + CCNU | 25 | 11 | 2 | 1 | 1 | 9 | 1 |
| CCNU alone | 67 | 9 | 19 | 5 | 1 | 31 | 2 |
CCNU-induced neutropenia
The median neutrophil nadir for dogs in the CCNU alone group was 573 cells μL−1 (range, 0–8600 cells μL−1). The median neutrophil nadir for dogs in the Dex + CCNU group was 282 cells μL−1 (range, 0–22 400 cells μL−1). The frequency of grade 4 neutropenia was not different between the two treatment groups; 30 of 67 (45%, 95% confidence interval [CI]: 33–57) of dogs treated with CCNU alone had grade 4 neutropenia compared with 16 of 25 (64%, 95% CI: 43–81) of dogs in the Dex + CCNU group (P = 0.16). Four dogs (6%) in the CCNU alone group had fevers compared with five dogs (20%) in the Dex + CCNU groups (P = 0.06). All episodes of fever were documented during the time of grade 4 neutropenia and all dogs recovered with supportive treatment (intravenous fluid and antibiotics). CCNU-induced neutrophil toxicosis for both treatment groups is summarised in Table 3.
| Treatment group | No. of dogs | Neutropenia grade, n (%) a | ||||
|---|---|---|---|---|---|---|
| 0 | 1 | 2 | 3 | 4 | ||
| Dex + CCNU | 25 | 3 (12) | 0 (0) | 1 (4) | 5 (20) | 16 (64) |
| CCNU alone | 67 | 6 (9) | 14 (21) | 8 (12) | 9 (13) | 30 (45) |
- a From reference 20, haematologic toxicity grading criteria: grade 0 (≥3000 neutrophils μL−1), grade 1 (1500–2999 neutrophils μL−1), grade 2 (1000–1499 neutrophils μL−1), grade 3 (500–999 neutrophils μL−1) and grade 4 (<500 neutrophils μL−1).
Follow-up CBC information 14 days postadministration of CCNU was available in 11 of 30 dogs with grade 4 neutropenia in the CCNU group and 8 of 16 dogs with grade 4 neutropenia in the Dex + CCNU group. Neutrophil counts were normal in 8 of 11 dogs with grade 4 neutropenia in the CCNU alone group and 4 of 8 dogs with grade 4 neutropenia in the Dex + CCNU group by 14 days postadministration of CCNU. Persistent neutropenia beyond day 14 was observed in three dogs in the CCNU alone group (all grade 1) and four dogs in the Dex + CCNU alone group (three grade 1 and one grade 3). All three persistently neutropenic dogs in the CCNU alone group went on to receive chemotherapy on day 14 post-CCNU, precluding the ability to record the precise time to recovery to a normal neutrophil count. The median time to a normal neutrophil count for the four persistently neutropenic dogs in the Dex + CCNU was 24 days postnadir (range, 10–21 days). While the targeted sample size was not reached, the incidence of grade 4 neutropenia identified in the Dex + CCNU group was high enough to justify early termination of this portion of the study.
Dexamethasone plasma concentrations
The mean peak plasma level of dexamethasone in three healthy dogs after receiving a single oral dose of dexamethasone (0.1 mg kg−1) was 27 ng mL−1 (actual peak plasma levels = 24, 28 and 29 ng mL−1). Time to peak plasma concentration of dexamethasone was 15 min in two dogs and 100 min in the third dog.
Dose escalation of dexamethasone
Based on the pharmacokinetic results, the dosage of dexamethasone was increased from 0.1 to 0.5 mg kg−1 PO every 12 h for five consecutive days. Eight tumour-bearing dogs received the increased dosage of dexamethasone and the median neutrophil nadir for these dogs posttreatment with CCNU (90 mg m−2 PO) was 400 cells μL−1 (range, 0–2400 cells μL−1). Six of the eight dogs (75%, 95% CI: 41–93) had grade 4 neutropenia and three (38%) had fevers. The three dogs with fever recovered with supportive care, but because of the high rate of treatment-related complications, the trial was terminated.
Discussion
Few studies in veterinary medicine have addressed methods to reduce or prevent chemotherapy-associated haematologic toxicity. Although corticosteroids are often incorporated into treatment protocols for various canine cancers, there is limited information regarding their effects on haematologic toxicity, pharmacokinetics or antitumour activity of chemotherapeutic agents. This study was designed to investigate the use of dexamethasone as a haematoprotective agent for tumour-bearing dogs receiving treatment with CCNU. The results from the current study indicate pretreating dogs with dexamethasone does not reduce the haematologic toxicity associated with CCNU.
The haematoprotective effects of dexamethasone in people are attained when the peak plasma levels following a single dose of dexamethasone exceed 80 ng mL−1.7 Previous pharmacokinetic studies indicate that this peak plasma level is achieved when people are administered a single oral dose of 8 mg of dexamethasone.26–28 To the authors’ knowledge, there is no published information regarding pharmacokinetic parameters when dexamethasone is given orally to dogs. Therefore, we used allometric scaling calculations to determine that a peak plasma level of 80 ng mL−1 of dexamethasone could be obtained when dogs received a single oral dose of 0.1 mg kg−1 of dexamethasone.22 In attempt to verify that this dosing strategy was accurate, we subsequently measured the peak plasma levels of dexamethasone in healthy dogs following administration of a single oral dosage of 0.1 mg kg−1 of dexamethasone. Peak plasma levels of dexamethasone achieved were 24–29 ng mL−1, which was substantially lower than the targeted peak plasma level of 80 ng mL−1. Failure to achieve this targeted peak plasma level is therefore a possible explanation for negative results reported here. Although allometric scaling is a useful tool for predicting interspecies pharmacokinetic parameters, it is well known that many drugs show no allometric correlation and that there are varying degrees of success with different methods for different drugs.29,30 Given that our dosing strategy might not have been accurate, we increased the dosage of dexamethasone to 0.5 mg kg−1 PO every 12 h in eight tumour-bearing dogs. Pharmacokinetic data from two of these dogs demonstrated this increased dosage resulted in peak plasma levels of dexamethasone exceeding 80 ng mL−1 (data not shown). However, 75% of dogs treated with this increased dexamethasone dosage before CCNU still developed grade 4 neutropenia and 38% had fevers. This increased incidence of toxicity lead to termination of enrolment of dogs in this study before reaching the targeted sample size as well as a re-evaluation of the study design.
Bone marrow CFU-GM from mice pretreated with corticosteroids shows increased resistance to chemotherapeutic agents in vitro as evidenced by a reduction in percent inhibition of growth of cells following exposure to the antineoplastic drug being evaluated. The resistance appears to be related to decreases in the number of CFU-GM in S phase as evidenced by a reduction in sensitivity to high-specific activity 3H-thymidine. In mice, the postchemotherapy resistance of haematopoietic progenitor cells following corticosteroid exposure is seen with either cell-cycle-specific drugs or platinum agents.9,31 Therefore, it is possible that our negative results are a function of investigating CCNU as our chemotherapeutic drug of choice rather than using an anticancer agent already established in preclinical models. Of interest, nonproliferating cells are most susceptible to cell kill from nitrosourea agents in vitro, whereas in vivo, the normally quiescent haematopoietic stem cells appear to be less sensitive than either rapidly proliferating marrow cells or tumour cells. Nitrosoureas cause arrest in the G2 phase of the cell cycle, further supporting the concept that these agents cause cell death more effectively in nondividing cells than in dividing cells.32 If dexamethasone is capable of causing cell cycle arrest of haematopoietic stem and/or progenitor cells, and CCNU is more effective against nondividing cells then treating dogs with dexamethasone before administration of CCNU could therefore enhance the damage to haematopoietic cells, rather than protect them.
Further support for the possibility of increased CCNU-induced damage to haematopoietic progenitor cells following pretreatment with dexamethasone is seen when examining the time to recovery to a normal neutrophil count for dogs in both treatment groups. In people, treatment with dexamethasone before administration of chemotherapy resulted in a shortening in the duration of time required for neutrophil counts to return to baseline values.7 Follow-up CBC information was available for the majority of dogs in both treatment groups in this study; however, CBCs were not obtained at standardised intervals beyond 1 week posttreatment, so we are unable to accurately report time to recovery to a normal neutrophil counts for all patients. Despite not having complete data for all dogs, it appears as though some dogs treated with dexamethasone before receiving CCNU showed prolonged durations of neutropenia. No dog in either treatment group experienced prolonged grade 4 neutropenia, so patients did not appear to have a protracted duration of time of increased risk for sepsis. However, depending on tumour type and chemotherapy protocol, prolonged neutropenia might result in treatment delays and negatively impact remission rates, durations and overall survival.
The two treatment groups were similar with respect to age, body weight and underlying neoplasia. For purposes of the study, all nonlymphoma neoplasms were categorised as ‘other neoplasia’, and treatment groups were compared as lymphoma versus other. This was performed because lymphoma is the most likely tumour to have bone marrow involvement. Many dogs in both groups had neoplastic infiltrates in bone marrow detected during initial staging, but tests were not always performed immediately before the CCNU dose. However, it is unlikely that myelophthesis had an impact on the results reported here.
A significantly larger percentage of dogs in the CCNU alone group had received previous chemotherapy compared with the Dex + CCNU group (P = 0.003). This comparison was examined as we were concerned that dogs that were more heavily pretreated with chemotherapy could have potentially accrued more cumulative damage to their bone marrow progenitor cells and therefore could have been more susceptible to CCNU-induced myelosuppression, less responsive to the protective effects of dexamethasone or both. Given that the more heavily pretreated group had a numerically lower and statistically similar frequency of severe neutropenia, it is unlikely that this had a substantial impact on outcome. Additionally, the entry criteria for the pilot study conducted in people permitted prior treatment with chemotherapy with no apparent adverse effect on outcome.7
In summary, the administration of dexamethasone before CCNU did not result in a decrease in frequency of grade 4 neutropenia when compared with a historical group of dogs treated with CCNU alone. Additional investigations incorporating the administration of dexamethasone before cell-cycle-specific chemotherapeutic drugs could still be worthwhile. Further characterisation of the pharmacokinetic parameters of oral dexamethasone in both healthy and tumour-bearing dogs would also be valuable for understanding the potential haematoprotective mechanisms of corticosteroids.
