Lymphomas are among the most common cancers in dogs, having an annual incidence rate of approximately 24 cases/100,000 dogs at risk.1 The majority of lymphomas in dogs have a multicentric distribution characterized by generalized lymphadenomegaly, with other organs such as the liver, spleen, and bone marrow also commonly affected.2 Combination chemotherapy with CHOP, with or without L-asparaginase (ie, L-CHOP), represents the current standard of care for dogs with nodal lymphomas, affording complete cancer remission rates of approximately 70% to 90% and median overall survival times of approximately 9 to 14 months.3–5 However, the prognosis for individual dogs treated with chemotherapy is reportedly highly variable, affected by factors such as tumor immunophenotype, histopathologic grade, stage, and substage as well as several other cancer- or treatment-related factors.2
Among treatment-related factors that may affect the prognosis for dogs with nodal lymphomas, treatment with prednisone or other glucocorticoids prior to initiating cytotoxic chemotherapy repeatedly has been reported to have a deleterious effect on remission rate and survival time.6–8 The pathogenesis of this clinical phenomenon is not fully understood; however, it has been hypothesized that corticosteroid treatment upregulates expression of the cell membrane–associated drug efflux pump P-glycoprotein, conferring multidrug resistance.9,10 Corticosteroids increase cellular expression of P-glycoprotein in vitro,11,12 and higher tumoral P-glycoprotein expression at the time of chemotherapy induction is associated with decreased durations of remission and overall survival times in dogs with nodal lymphomas.13,14
Although prior treatment with prednisone is reportedly strongly correlated with poor outcome in dogs with chemotherapy-treated lymphomas,6–8 the effect of omitting prednisone from the CHOP chemotherapy protocol on treatment outcome has undergone limited investigation. Hypothetically, we suggest that the omission of a drug that induces multidrug resistance from a chemotherapy protocol for treatment of lymphomas could be advantageous and might improve the survival of dogs treated in such fashion. Zandvliet et al9 recently reported that omission of prednisolone from a CHOP-based chemotherapy protocol did not have a significant effect on remission rate or progression-free survival time in dogs with nodal lymphomas. These authors, however, did not enroll their target sample size; as such, the trial may have been underpowered to detect significant differences in clinically important outcome variables between study groups. Furthermore, to our knowledge, that study represents the only report to date describing the effect of prednisone omission from CHOP chemotherapy, and the results have not been reproduced by others.
Therefore, the primary objective of the study reported here was to determine the effect of prednisone omission from a CHOP-based chemotherapy protocol on the median progression-free survival time of dogs with histopathologically confirmed peripheral nodal lymphomas. We hypothesized that the median progression-free survival time for dogs treated with a CHOP-based protocol omitting prednisone would be at least 33% longer than the median progression-free survival time for dogs receiving a traditional CHOP chemotherapy protocol. Secondary objectives were to determine whether prednisone omission significantly affected remission rate, overall survival time, and the rate of serious treatment-related adverse events in study dogs.
Materials and Methods
Study protocol
The trial was reported in accordance with CONSORT 201015,16 recommendations (Supplemental Table S1, available at http://avmajournals.avma.org/doi/suppl/10.2460/javma.249.9.1067).
Study population
A nonblinded, parallel-group, randomized, controlled trial was conducted at the Purdue University Veterinary Teaching Hospital. All dogs enrolled in the study were client-owned pet animals examined for untreated peripheral nodal lymphomas. Dogs were enrolled between September 2008 and October 2013. The study protocol was approved by the Purdue University Animal Care and Use Committee, and written informed consent was obtained from all dog owners prior to study enrollment. Trial enrollment was voluntary. A small discount to the cost of chemotherapy was offered to the owners of dogs participating in this study, although these owners were responsible for the majority of study-related medical expenses. Dogs were included in the study if they met the following criteria: cytologic or histopathologic confirmation of the diagnosis of primary peripheral nodal lymphoma, expected survival time of > 4 weeks with treatment, and written informed owner consent. Dogs were excluded from the study if they met any of the following criteria: a diagnosis of primary extranodal lymphoma (as defined by physical examination and diagnostic imaging findings), platelet count < 100,000/μL, neutrophil count < 2,500/μL, history or clinical signs of cardiac disease that would reasonably preclude doxorubicin treatment (such as ventricular arrhythmia, cardiomyopathy, or congestive heart failure), clinically important evidence of hepatic dysfunction (defined as serum alanine aminotransferase activity ≥ 4 × the upper limit of the reference range [3 to 69 U/L], hyperbilirubinemia, or serum biochemical evidence of hepatic synthetic failure, including hypoglycemia, hypoalbuminemia, hypocholesterolemia, or low BUN concentration), prior chemotherapy or radiotherapy for the lymphoma, and treatment with any systemic glucocorticoid within 30 days of initiation of chemotherapy. Approximately 27 months after initiating the trial, we decided that dogs with a high breed-associated risk for carrying the ABCB1-1Δ gene mutation17 would also be excluded from the study.
Study interventions and timing
At the time of study enrollment, all dogs underwent incisional wedge biopsy or surgical extirpation of a peripheral lymph node to provide adequate tissue for histologic examination. In dogs in which surgical biopsies had been performed previously by the primary care veterinarian, unstained slides made from formalin-fixed, paraffin-embedded tumor tissues were obtained for review. All lymphomas were histopathologically subtyped according to World Health Organization criteria (Appendix 1).18 Histopathologic confirmation of a diagnosis of lymphoma was made by reviewing H&E-stained tissue sections as well as via immunohistochemical detection of CD79a or CD3 expression by the tumor cells, as previously described.19 A single board-certified pathologist (JAR) reviewed all biopsy specimens and reported tumor grade, mitotic index (defined as number of mitotic figures observed per 400X field, for a mean of 10 counted fields), and World Health Organization sub-type18 for each tumor. Tumors with a mitotic index of ≤ 5 mitoses/400X field were considered low grade, those with a mitotic index of 6 to 10 mitoses/400X field were considered intermediate grade, and those with a mitotic index > 10 mitoses/400X field were considered high grade.
For staging purposes, all dogs underwent standardized diagnostic testing at the time of study enrollment, including CBC, serum biochemical analysis, thoracic and abdominal radiography (including right and left lateral thoracic, right lateral abdominal, and dorsoventral abdominal and thoracic views), abdominal ultrasonography, bone marrow aspiration with cytologic examination, and an ECG. Cancer stage and substage for all dogs were determined according to World Health Organization criteria (Appendix 2).20 A diagnosis of hepatic or splenic infiltration was made on the basis of results of abdominal radiographic and ultrasonographic examinations; fine-needle aspiration of these organs was not performed. Bone marrow infiltration by the lymphoma was diagnosed on the basis of the presence of malignant lymphoblasts observed on microscopic evaluation of a bone marrow aspirate. Restaging was not performed in any study dog during its initial course of treatment.
Dogs were randomly assigned to treatment with a combination L-CHOP chemotherapy protocol previously described by Garrett et al3 or a protocol designated as L-CHO that was identical except for the exclusion of prednisone (Table 1). A CBC was performed in all dogs immediately prior to each chemotherapy treatment. If the segmented neutrophil count was < 2,500/μL before doxorubicin or cyclophosphamide treatment or < 1,500/μL before vincristine treatment, chemotherapy was withheld and a CBC was repeated in 3 to 5 days. If necessary, this process was repeated until the neutrophil count was above the acceptable threshold, at which time chemotherapy was administered. Doxorubicin was administered at a dose of 30 mg/m2 in dogs weighing > 15 kg (33 lb) and at a dose of 1 mg/kg (0.45 mg/lb) in dogs weighing ≤ 15 kg. At the time of study initiation, we had planned that all patients would receive cyclophosphamide IV. However, during the study, the cost of parenteral cyclophosphamide at the Purdue University Veterinary Teaching Hospital increased by approximately 1,700%. As such, to permit more affordable treatment of study dogs, and because the bioavailability of cyclophosphamide's active cytotoxic metabolite has been reported as relatively equivalent when administered IV versus by mouth,21 dogs were allowed to receive cyclophosphamide either orally or IV during the later stages of the study, from approximately March 2012 through July 2014.
Protocol3 for treatment of dogs (n = 40) enrolled in a single-center, nonblinded, parallel-group, randomized, controlled trial evaluating the effect of prednisone omission from a multidrug chemotherapy protocol on treatment outcome in dogs with peripheral nodal lymphomas.
| Week | ||||||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Treatment | 1 | 2 | 3 | 4 | 6 | 7 | 8 | 9 | 11 | 13 | 15 | 17 | 19 | 21 | 23 | 25 |
| l-asparaginase (400 U/kg, SC) | x | |||||||||||||||
| Vincristine (0.7 mg/m2, IV) | x | x | x | x | x | x | x | x | ||||||||
| Cyclophosphamide (250 mg/m2, IV or PO) | x | x | x | x | ||||||||||||
| Doxorubicin (30 mg/m2, IV, or 1 mg/kg, IV) | x | x | x | x | ||||||||||||
| Prednisone* | ||||||||||||||||
Dogs assigned to the L-CHOP treatment group received 2 mg/kg (0.9 mg/lb), PO, once a day for 7 days; 1.5 mg/kg (0.68 mg/lb), PO, once a day for 7 days; 1 mg/kg (0.45 mg/lb), PO, once a day for 7 days; and 0.5 mg/kg (0.23 mg/lb), PO, once a day for 7 days, after which prednisone administration was discontinued. Dogs assigned to the L-CHO treatment group did not receive prednisone.
x = Patient received specified treatment.
For the sake of convenience, some patients received a portion of their chemotherapy treatments with their referring veterinarian. In these cases, medical records related to treatment and follow-up, including chemotherapy dosing records (ie, dates of administration, drug and dose administered, and route of administration), lymph node measurements, results of CBCs, treatment-related adverse events, and dates of tumor progression or death from any cause, were obtained from the referring veterinarian. Standardized data capture forms were not provided to the referring veterinarian, and as such, data were abstracted from referring veterinarian medical records retrospectively.
Rescue chemotherapy was offered to the owners of all study dogs at the time of first evidence of cancer progression following L-CHOP or L-CHO, although rescue treatments were not standardized.
Outcome measures
The primary outcome of interest was progression-free survival time, defined as the time elapsed from the date of chemotherapy initiation to the date that cancer progression (ie, progressive disease) was first detected or to the date of death from any cause, whichever came first. The date of chemotherapy initiation and the date of study enrollment were the same for all dogs enrolled in the study. Secondary outcomes of interest included overall response rate, the rate of serious treatment-related adverse events in each group, and overall survival time, defined as the time elapsed from the date of chemotherapy initiation to the date of death from any cause. The target sample size for the trial was 80 dogs (40 dogs/group). This was determined on the basis of a calculation to ascertain the sample size necessary to detect a 33% improvement in median progression-free survival time3 (375 days vs 282 days) with 80% power and an α value (error rate) of 0.05.
Overall response to chemotherapy was determined with caliper-based measurement of peripheral lymph nodes and was defined according to criteria established by Vail et al.22 Briefly, complete remission was defined as absence of measurable tumor burden, partial remission was defined as ≥ 30% reduction in the sum of the longest diameters of measurable tumor lesions, progressive disease was defined as ≥ 20% increase in the sum of the longest diameters of measurable tumor lesions or the appearance of new lesions, and stable disease was defined as measurable tumor burden that was neither partial remission nor progressive disease. Caliper measurements of peripheral lymph nodes were obtained from all dogs at the time of study enrollment. Overall response to chemotherapy was based on repeated caliper measurements obtained at the time of each chemotherapy treatment. Reassessment of peripheral lymph node size was then scheduled to occur once monthly thereafter for dogs that completed the initial chemotherapy protocol. Overall response was not assessed by a single evaluator blinded to the dog's group assignment, but rather by the veterinarian in charge of that dog's care at the time of initial evaluation and study enrollment. In cases in which a dog's best overall response to chemotherapy was complete remission, progressive disease was confirmed by means of cytologic or histopathologic examination of a lymph node specimen whenever possible.
The occurrence of treatment-related adverse events was recorded for each patient, and serious treatment-related adverse events were defined as those ≥ grade 3 of 5 according to the Veterinary Cooperative Oncology Group's common terminology criteria for adverse events.23 Adverse events were graded by reviewing clinicopathologic data at the time of each scheduled recheck together with owner interviews at these times. The owners' answers were recorded on a questionnaire form (Supplemental Appendix S1, available at http://avmajournals.avma.org/doi/suppl/10.2460/javma.249.9.1067) distributed as routine practice to all owners of animals with cancer at our hospital. Data regarding adverse events in dogs receiving chemotherapy treatments from their referring veterinarians were abstracted retrospectively from medical records. Standardized data capture forms and attempts to more objectively grade adverse events according to owners' recollections (eg, request the owner to grade severity of an adverse event according to a visual analogue scale) were not employed. Chemotherapy dose reductions of 20% were planned for dogs reportedly experiencing ≥ grade 3 adverse events. However, the general management of treatment-related adverse events in study dogs was not standardized. The need for and method of treatment for adverse events were at the discretion of the dogs' attending clinician at the time an adverse event occurred. Dogs with neutropenia and fever (regardless of grade) typically received IV fluid therapy and treatment with broad-spectrum antimicrobials. Dogs with grade 3 or 4 neutropenia, but no fever or other serious signs of illness, typically received oral antimicrobial prophylaxis, although the type and dose of antimicrobial varied with clinician preference. Management of gastrointestinal adverse events generally included dietary adjustments and the use of antiemetic (eg, metoclopramide, ondansetron, and maropitant) or antidiarrheal (eg, metronidazole) medications. Whether gastrointestinal adverse events were managed in an inpatient or outpatient setting was a matter of clinician preference and was dependent on individual clinician assessments of the severity of clinical signs associated with the adverse event in each dog. Dogs experiencing urinary tract adverse events attributable to cyclophosphamide treatment received chlorambucil at a dose of 1.4 mg/kg (0.64 mg/lb) PO in place of cyclophosphamide at the time of all subsequent doses in their treatment protocols. Treatment delays were defined as delays in a dog's treatment that deviated from the scheduled treatment protocol by ≥ 5 days. Whether adverse events of grade 3 or higher, dose reductions, or treatment delays occurred was recorded for each patient. If a dog experienced more than 1 adverse event of grade 3 or higher, dose reduction, or treatment delay, only the first such event was included in the statistical analysis. Statistical analysis of repeated events in these categories was not performed.
Group assignment and blinding
Individual dogs were randomly assigned to treatment with L-CHOP or L-CHO. A simple randomization scheme was generated by coin toss; block or stratified randomization schemes were not employed. The randomization scheme was generated by one of the authors (MOC) by means of 80 consecutive coin tosses prior to the enrollment of any dogs. The randomization scheme called for dogs to be allocated individually into L-CHOP and L-CHO groups. Dogs were enrolled consecutively in the study by a rotating team of board-certified veterinary oncologists and oncology residents working within the hospital during the study period.
The personnel enrolling dogs in the study were not blinded to the randomization scheme or to the dogs' group assignments, but group assignment was concealed from the dog owners until after written consent for study participation was obtained. Following this, owners were not blinded to their dogs' treatment. A placebo was not offered in place of prednisone in the L-CHO group. The same personnel involved in dog enrollment were also responsible for assessing overall response and adverse events in study dogs and were thus not blinded to the dogs' treatment group assignment when making these assessments.
Statistical analysis
Descriptive statistics, including patient-, tumor-, and treatment-related variables of interest, were recorded for each group. Patient-related variables for each dog included sex, neuter status, age, and weight at study enrollment. Tumor-related variables included World Health Organization stage and substage, immunophenotype, World Health Organization histopathologic subtype, grade, mitotic index, presence of anemia, and presence of bone marrow involvement at the time of diagnosis. Treatment-related variables included whether chemotherapy dose reductions or dose delays occurred, whether cyclophosphamide was administered by mouth or IV, and whether the dog received some portion of its chemotherapy from the referring veterinarian.
Proportionate differences in patient-, tumor-, and treatment-related variables were compared between the 2 groups with either the χ2 or Fisher exact test for categorical variables and 2-sample t test for continuous variables. Normality of distribution of continuous variables was assessed by means of the Shapiro-Wilk test prior to statistical analysis. Progression-free and overall survival times were estimated for each group with the Kaplan-Meier method. Data regarding progression-free survival and overall survival times were analyzed on an intention-to-treat basis (ie, on the basis of the patients' original group assignment) because 3 study dogs were unable to receive the full treatment to which they were originally randomized as a result of adverse events. Differences in survival times were assessed with the log-rank test. Hazard ratios with corresponding 95% confidence intervals for cancer progression and death were calculated with Cox proportional hazards modeling. Statistical significance for all tests and final models was set at P ≤ 0.05. Statistical calculations were performed with a commercial software package.a
Results
Study population
Forty dogs were enrolled in the study. The flow of dogs through the study is presented (Figure 1). Three dogs were screened for eligibility but excluded from trial enrollment because of cardiac disease (1 patient with atrial fibrillation and 1 patient with episodes of ventricular tachycardia) or the presence of the ABCB1-1Δ mutation (1 dog with a homozygous mutant configuration, as determined by PCR-based testing of cheek swab samplesb). Several other dogs examined during the study period were presumably eligible for this study, but their owners declined to pursue eligibility screening or enroll these dogs for various reasons that were not documented. The number of such dogs examined at the hospital during the study period was not recorded.
Study flow diagram showing eligibility screening and progression of patients enrolled in a single-center, nonblinded, parallel-group, randomized, controlled trial evaluating the effect of prednisone omission from a multidrug chemotherapy protocol (L-CHOP) on treatment outcome in dogs with peripheral nodal lymphomas.
Citation: Journal of the American Veterinary Medical Association 249, 9; 10.2460/javma.249.9.1067
Eighteen dogs were randomized to chemotherapy with L-CHOP and 22 dogs were randomized to chemotherapy with L-CHO. Patient- and tumor-related variables for the 2 groups were summarized (Table 2; Supplemental Table S2, available at http://avmajournals.avma.org/doi/suppl/10.2460/javma.249.9.1067). Twenty different dog breeds were included in the study population, with mixed-breed dog (n = 13), Golden Retriever (6), and Labrador Retriever (3) being most frequently represented. Consistent with previous reports,18,24 the most commonly represented histopathologic subtype was diffuse large B-cell lymphoma. Two dogs had B-cell lymphomas characterized by diffuse effacement of nodal architecture by a population of neoplastic cells in which the nuclear size was too small for the subtype to be considered diffuse large B-cell lymphoma according to a strict interpretation of World Health Organization guidelines. These guidelines stipulate that the size of the nuclei of the neoplastic cells in diffuse large B-cell lymphoma must be 2X to 2.5X the diameter of an erythrocye.25 These 2 tumors were therefore classified as diffuse intermediate-size B-cell lymphoma.24
Summary of patient- and tumor-related variables for the dogs in Table 1.
| Variable | L-CHOP (n = 18) | L-CHO (n = 22) |
|---|---|---|
| Patient variables | ||
| Sex | ||
| Castrated male | 12 | 5 |
| Sexually intact male | 0 | 3 |
| Spayed female | 5 | 14 |
| Sexually intact female | 1 | 0 |
| Age (y) | 7.5 (4.2–11.2) | 7.0 (2.6–15.1) |
| Weight (kg) | 34.9 (7.2–62.5) | 33.6 (12.3–58.3) |
| Tumor variables | ||
| World Health Organization stage | ||
| 3 | 5 | 3 |
| 4 | 6 | 11 |
| 5 | 7 | 8 |
| World Health Organization substage | ||
| a | 15 | 18 |
| b | 3 | 4 |
| Immunophenotype | ||
| B cell | 17 | 16 |
| T cell | 1 | 6 |
| World Health Organization subtype | ||
| DLBCL | 16 | 14 |
| DIBCL | 1 | 1 |
| FL | 0 | 1 |
| PTCL-NOS | 0 | 3 |
| TZL | 1 | 1 |
| T-LBL | 0 | 2 |
| Tumor grade | ||
| Low | 6 | 8 |
| Intermediate | 6 | 11 |
| High | 6 | 3 |
| Mitotic index | 7.3 (0.3–12.6) | 5.4 (0.5–12.3) |
| Anemic at diagnosis | ||
| No | 15 | 19 |
| Yes | 3 | 3 |
| Bone marrow involved | ||
| No | 14 | 13 |
| Yes | 3 | 7 |
| Nondiagnostic sample | 1 | 2 |
Age, weight, and mitotic index values are reported as median (range). DIBCL = Diffuse intermediate-size B-cell lymphoma.
DLBCL = Diffuse large B-cell lymphoma. FL = Follicular lymphoma. PTCL-NOS = Peripheral T-cell lymphoma, not otherwise specified. T-LBL = T-lymphoblastic lymphoma. TZL = T-zone lymphoma.
Small Animals & Exotic
Fifteen dogs received a portion of their chemotherapy administered by their referring veterinarians; 4 of these dogs received L-CHOP, and 11 received L-CHO. Although dogs receiving L-CHO were more likely to receive some portion of their chemotherapy treatments with their primary care veterinarians than dogs receiving L-CHOP, this difference was not significant (P = 0.104). Thirty dogs received cyclophosphamide only via the IV route, 6 dogs received cyclophosphamide only via the oral route, and 4 dogs received cyclophosphamide via both routes. Fourteen dogs receiving only IV cyclophosphamide were treated with L-CHOP, and 16 dogs receiving only IV cyclophosphamide were treated with L-CHO; there was no significant (P = 0.873) difference in the proportion of dogs receiving only IV cyclophosphamide between the 2 groups.
Study participant flow
Twenty-one dogs completed the full course of chemotherapy treatment to which they were randomized, and 19 did not (6 treated with L-CHOP and 13 treated with L-CHO). The reason for failure to complete the assigned course of treatment was progressive disease in 16 cases and treatment-related adverse event in 3 cases (Figure 1). The median duration of follow-up for all dogs (n = 40) was 341.5 days (range, 41 to 1,545 days), whereas for dogs still alive at the end of the study (n = 6), the median duration of follow-up was 464.5 days (range, 278 to 1,545).
The adherence of pet owners to the prescribed recheck schedule following completion of L-CHOP or L-CHO (ie, physical examination once monthly) was poor, with only 5 of 21 dogs that completed their treatment protocol examined at the hospital monthly until the time that progressive disease was documented. The remaining dogs were examined at our hospital or by their referring veterinarian sporadically, according to the convenience of their owners, between the time that their initial treatment protocol was completed and the time that progressive disease was diagnosed.
Recruitment of dogs for the study was slow, with only half of the target sample size of 80 dogs recruited after 5 years. In view of this difficulty in case recruitment, with the expectation that a similar length of time would be required to enroll the second half of the target sample size, a decision was made to terminate the study early.
Efficacy of treatment
Overall response to chemotherapy was classified as complete remission in 28 of 40 (70%) dogs and partial remission in 12 (30%) dogs. In dogs receiving L-CHOP, overall response was complete remission in 13 of 18 (72%) dogs and partial remission in 5 (28%) dogs. In dogs receiving L-CHO, overall response was complete remission in 15 of 22 (68%) dogs and partial remission in 7 (32%) dogs. There was no significant (P = 0.784) difference in the complete remission rate between the 2 groups.
Of the 28 dogs experiencing complete remission of their lymphoma, 22 eventually experienced cancer progression (9 treated with L-CHOP and 13 treated with L-CHO). In 21 of these dogs, progressive disease was confirmed by cytologic or histopathologic examination of lymph node specimens. The owner of the remaining dog declined to return to the hospital at the time of progressive disease, and the date of cancer progression for this dog was recorded as the date that recurrence of lymphadenomegaly was diagnosed by the referring veterinarian. Of the 12 dogs experiencing partial remission, 11 eventually experienced cancer progression (4 treated with L-CHOP and 7 treated with L-CHO). In 10 of these dogs, progressive disease was confirmed with lymph node measurements taken at a visit to the Purdue University Veterinary Teaching Hospital. The owner of the remaining dog declined to return to the Purdue University Veterinary Teaching Hospital, and the date of progressive disease in this dog was recorded as the date that a subjective worsening of the dog's lymphadenomegaly during the course of chemotherapy was reported by the referring veterinarian.
At the time of data analysis, 6 of 40 dogs were still alive, and 34 dogs had died. Of the dogs alive at the time of data censoring, 2 dogs had measurable disease and 4 had lymphoma in complete remission. The former 2 dogs were censored with respect to overall survival time, whereas the latter 4 dogs were censored with respect to progression-free and overall survival times. Twenty-seven of the 40 dogs died or were euthanatized because of progressive lymphoma. Three dogs died of neoplasia other than lymphoma (1 each of cholangio-cellular carcinoma, high-grade mast cell tumor, and disseminated histiocytic sarcoma). Two dogs were euthanatized because of hemoabdomen 551 and 715 days after chemotherapy induction. In both dogs, hemoabdomen was detected by means of ultrasonographic imaging and cytologic examination of an abdominocentesis fluid sample. One of these dogs had clinically apparent drug-resistant lymphoma at the time of death, and the hemoabdomen present in this dog was presumed to be lymphoma-related, but postmortem examination was declined by the dog's owner. The other dog was receiving rescue chemotherapy for lymphoma relapse, but had no physical evidence of lymphoma at the time of death. It did, however, have multiple cavitated hepatic and splenic masses that were identified with ultrasonography. Cytologic examination of a fine-needle aspirate of one of the hepatic masses revealed moderate hepatocellular hyperplasia with no evidence of lymphoma. This dog was presumed to have nonlymphoid cancer of the liver, spleen, or both as the cause for the hemoabdomen, but postmortem examination was declined by the dog's owner. One dog was euthanatized because of acute renal failure 341 days after chemotherapy induction; there was no clinical, clinicopathologic, or diagnostic imaging–based evidence of lymphoma in this dog at the time of euthanasia, and postmortem examination was declined by the owner. One dog died spontaneously while asleep at home 659 days after chemotherapy induction, with its lymphoma still in apparent complete remission. The owner declined postmortem examination; therefore, a cause of death could not be ascertained.
The estimated median progression-free survival time for dogs receiving L-CHO was 142.5 days, and that for dogs receiving L-CHOP was 292 days; L-CHO was associated with an increased hazard for cancer progression, compared with L-CHOP (hazard ratio, 1.79; 95% confidence interval, 0.85 to 3.75), but this result was not significant. The estimated median overall survival time for dogs receiving L-CHO was 255 days, and that for dogs receiving L-CHOP was 387 days; L-CHO was associated with an increased hazard for death, compared with L-CHOP (hazard ratio, 1.53; 95% confidence interval, 0.74 to 3.17), but this result was not significant. There was no significant difference in estimated median progression-free (P = 0.117) or overall (P = 0.330) survival time between the 2 groups (Figure 2).
Kaplan-Meier curves of progression-free (A) and overall (B) survival times for dogs (n = 40) receiving combination chemotherapy for peripheral nodal lymphomas. Solid lines represent dogs receiving an L-CHOP chemotherapy protocol, and dashed lines represent dogs receiving an L-CHO chemotherapy protocol.
Citation: Journal of the American Veterinary Medical Association 249, 9; 10.2460/javma.249.9.1067
Adverse events
Adverse events recorded for all 40 dogs enrolled in the study were summarized (Table 3). Consistent with previous reports regarding CHOP-based chemotherapy in dogs,3–5 most recorded adverse events were not serious (ie, grade 1 or 2)23 and were self-limiting. Two dogs receiving L-CHO experienced clinical signs (hematuria, stranguria, and pollakiuria) attributed to sterile hemorrhagic cystitis induced by cyclophosphamide treatment. Sterile hemorrhagic cystitis was diagnosed on the basis of the presence of compatible abnormalities on urinalysis and a lack of bacterial growth following microbial culture of urine samples from these 2 dogs. These dogs received 3 doses and 1 dose of chlorambucil in place of cyclophosphamide during their initial courses of treatment. One dog receiving L-CHOP experienced grade 2 ileus after receiving its first vincristine treatment. The severity of this adverse event was not acceptable to the dog's owner, who refused additional treatments with vincristine. This dog therefore received vinblastine (2 mg/m2, IV) in place of all subsequent vincristine doses (n = 7) in its initial course of treatment. Dogs receiving L-CHO were more likely to experience grade 3 or 4 adverse events (P = 0.160), chemotherapy dose reductions (P = 0.073), and treatment delays (P = 0.106) than were dogs receiving L-CHOP, although none of these differences were significant.
Summary of adverse events in dogs treated with an L-CHOP or L-CHO chemotherapy protocol for peripheral nodal lymphomas.
| Adverse event | Grade | L-CHOP | L-CHO |
|---|---|---|---|
| Neutropenia | 1 | 16 | 35 |
| 2 | 4 | 17 | |
| 3 | 4 | 10 | |
| 4 | 2 | 3 | |
| Thrombocytopenia | 1 | 15 | 15 |
| 2 | 2 | 0 | |
| 3 | 1 | 1 | |
| 4 | 0 | 0 | |
| Anemia | 1 | 5 | 0 |
| 2 | 0 | 2 | |
| 3 | 0 | 0 | |
| 4 | 0 | 0 | |
| Anorexia | 1 | 8 | 5 |
| 2 | 4 | 4 | |
| 3 | 0 | 1 | |
| 4 | 0 | 0 | |
| Vomiting | 1 | 7 | 4 |
| 2 | 0 | 3 | |
| 3 | 0 | 2 | |
| 4 | 0 | 0 | |
| Ileus | 1 | 0 | 0 |
| 2 | 1 | 0 | |
| 3 | 0 | 0 | |
| 4 | 0 | 0 | |
| Cystitis | 1 | 0 | 0 |
| 2 | 0 | 1 | |
| 3 | 0 | 1 | |
| 4 | 0 | 0 | |
| Lethargy | 1 | 11 | 5 |
| 2 | 2 | 1 | |
| 3 | 0 | 0 | |
| 4 | 0 | 0 | |
| Alopecia | 1 | 0 | 0 |
| 2 | 0 | 1 | |
| 3 | 0 | 0 | |
| 4 | 0 | 0 |
The values reported in the table reflect the number of adverse events occurring out of a total of 236 chemotherapy treatments administered to dogs in the L-CHOP group and 264 treatments administered to dogs in the L-CHO group. Grade was assigned on the basis of the Veterinary Cooperative Oncology Group common terminology criteria for adverse events.23 Grades for anemia were assigned on the basis of this document's parameters for PCV.
Discussion
Results of the present single-center, nonblinded, parallel-group, randomized, controlled trial were consistent with those previously reported by Zandvliet et al.9 In the present trial, the exclusion of prednisone from a CHOP-based chemotherapy protocol did not have a significant effect on overall response to chemotherapy, progression-free survival time, or overall survival time for dogs with primary nodal lymphomas. Therefore, it would appear that our hypothesis (ie, that prednisone omission from a CHOP-based chemotherapy protocol would significantly improve progression-free survival time) should be rejected on the basis of these results. This interpretation should be made with caution, however, as the possibility of a type II error cannot be excluded. The small sample size of this study, which was insufficient to detect a significant difference in progression-free survival time between treatment groups on the basis of a priori statistical power calculations, may explain our negative results. Furthermore, it is also possible that the opposite of our hypothesis is true: prednisone omission may have had a harmful effect. In the present trial, dogs from which prednisone was withheld had an approximately 80% greater risk for cancer progression (hazard ratio, 1.79; 95% confidence interval, 0.85 to 3.75) and 50% greater risk for death (hazard ratio, 1.53; 95% confidence interval, 0.74 to 3.17), compared with dogs receiving prednisone. Whereas the present study was not designed, nor likely sufficiently powered, to test the hypothesis that prednisone omission from L-CHOP is harmful, we cannot exclude such a possibility on the basis of our results.
Several other shortcomings in the design and execution of this clinical trial unfortunately limited the internal and external validity of our findings. First, the lack of concealment of the treatment allocation schedule could have contributed to selection bias in how dogs were enrolled in one treatment group or the other. Failure to blind personnel performing response assessments and the failure to offer a placebo treatment in place of prednisone in the L-CHO group may also have contributed to information bias in how overall response rate, progression-free survival time, overall survival time, and adverse events were assessed. Second, because progression-free survival time was the primary outcome of interest in this study, it was critical that all dogs be re-assessed for tumor progression at regularly scheduled intervals so that progression-free survival time would be calculated in an unbiased fashion. This did not occur. Because of poor owner compliance, only 5 of 21 dogs that completed the full course of L-CHOP or L-CHO chemotherapy were examined at the hospital in accordance with the monthly postchemotherapy recheck schedule. Furthermore, the owners of 2 dogs enrolled in the study did not permit objective assessment of disease progression by means of lymph node measurement or biopsy. Therefore, there was likely considerable information bias in how progressive disease and progression-free survival time were assessed in a large proportion of the dogs in this study. It is important to note that these biases, rather than a true effect of treatment, may have contributed to the increased hazards for progression and death in dogs treated with L-CHO. Finally, the criteria for inclusion of dogs in this study were likely too broad for a meaningful assessment of how omission of prednisone from a CHOP-based chemotherapy protocol affected the primary outcome of interest, progression-free survival time. Dogs were randomized to treatment groups without stratifying for factors that might influence progression-free survival time, such as tumor stage, substage, histologic subtype, or immunophenotype. This might have created an imbalance in the distribution of these prognostic factors between the 2 groups, creating a biased impression of how the treatment intervention affected progression-free survival time. Whereas a stratified randomization scheme could have been employed to account for these known prognostic factors, this would have increased the number of dogs required to show a significant difference in progression-free survival time between groups. A more practical solution might therefore have been to restrict entry in this trial to dogs with specific types and stages of lymphoma (such as diffuse large B-cell lymphoma, substage a) to afford greater patient homogeneity in the 2 treatment groups. This may have allowed a more meaningful interpretation of how the exclusion of prednisone from the L-CHOP protocol affects progression-free survival time in a clearly defined subset of dogs with lymphoma.
Although progression-free survival time was the primary end point assessed, a secondary objective of this study was to determine whether the omission of prednisone affected the rate of adverse events detected in treated dogs. Similar to our observations on treatment efficacy, we did not observe a significant difference in the tolerability of chemotherapy between the 2 treatment groups. Dogs receiving L-CHO experienced a greater number of adverse events overall, particularly neutropenia. These dogs were also more likely to experience serious, potentially life-threatening adverse events.23 However, there was no significant (P = 0.160) difference in the proportion of dogs experiencing grade 3 and 4 adverse events between the 2 groups. In addition, when considering the apparently greater number of adverse events in the L-CHO group, other factors that may have skewed these results should be considered. First, systematic and rigorous methods were not in place for recording and grading adverse events or attributing them specifically to treatment in this study. Information regarding adverse events was determined by review of CBC results and questionnaire forms that were provided as standard practice to the owners of dogs receiving chemotherapy at our hospital during the study period. Specialized forms for capturing and grading adverse events were not used, and the grading and attribution of adverse events in this study, particularly nonhematologic adverse events, were therefore subject to information bias. Bias was also introduced into the assessment of adverse events because some dogs received a portion of their chemotherapy from their referring veterinarians. Neither standard data capture forms nor specific instructions for recording the severity of adverse events or their likely attributions were provided to the referring veterinarians caring for dogs in this study. Grading and attributing adverse events in these dogs were performed by review of medical records provided by the referring veterinarians, a practice that likely also introduced information bias into the study results. Thus, we are unable to conclude from these results whether the exclusion of prednisone truly affected the tolerability of chemotherapy in these dogs.
Whereas several flaws in the execution of this study, chief among them including lack of blinding and failure to provide a placebo control, could and should have been corrected through better study design, other limitations were imposed by the circumstances under which the trial was conducted. A lack of funding for this study, in particular, made it challenging to accrue data in a timely and rigorous manner. For instance, this clinical trial was terminated early prior to enrollment of the target sample of 80 dogs because of slow case accrual. Poor recruitment of cases in this trial was likely influenced by several factors related to lack of funding support, such as competition from funded clinical trials and less expensive standard treatment options for dogs with lymphoma at the Purdue University Veterinary Teaching Hospital. Generally poor economic conditions in the geographic area from which referrals to the hospital were made during the study period likely also contributed to slow case accrual. Better financial support might have provided greater incentive for dog owners to consistently travel to the hospital in accordance with the study schedule rather than elect to have some scheduled treatments or recheck examinations performed by their referring veterinarian. This in turn might have afforded greater compliance with the study protocol and allowed a more meaningful determination of whether prednisone exclusion from L-CHOP chemotherapy has a significant effect on progression-free survival time. The effect that a lack of funding support would have on the validity of this study clearly was not fully considered at the time of its conception and should be a primary consideration in the design of any future studies attempting to address similar clinical questions with adequate statistical power.
Despite the many limitations imposed by flaws in the design and execution of this study, our results suggested that omission of prednisone from a CHOP-based chemotherapy protocol does not lead to improved progression-free survival time in dogs with peripheral nodal lymphomas. Whereas the question of whether omitting prednisone improves progression-free survival time was compelling to us on the basis of past observation of poorer outcomes in lymphoma-bearing dogs pretreated with glucocorticoids,6–8 we suggest that perhaps a more relevant question to everyday practice is whether L-CHO chemotherapy is no less effective, in terms of overall response rate or progression-free survival time, than L-CHOP. Answering such a question is important, as dogs with lymphoma are typically older and may therefore have comorbid conditions (eg, diabetes mellitus and hyperadrenocorticism) or require treatment with medications (eg, NSAIDs for osteoarthritis) that would contraindicate the use of glucocorticoid treatment. Because the present study was designed to show the superiority of one treatment versus another, the question of whether L-CHO is noninferior to L-CHOP cannot be addressed from our results. An adequately powered, properly blinded, placebo-controlled, randomized noninferiority trial employing more rigorous data collection practices than the present study would be a logical next step in this line of investigation.
Acknowledgments
No third-party funding or support was received in connection with this study or the writing or publication of the manuscript. The authors declare that there were no conflicts of interest.
ABBREVIATIONS
| CHOP | Cyclophosphamide, doxorubicin, vincristine, and prednisone |
| CONSORT | Consolidated Standards of Reporting Trials |
| L-CHO | L-asparaginase, cyclophosphamide, doxorubicin, and vincristine |
| L-CHOP | L-asparaginase, cyclophosphamide, doxorubicin, vincristine, and prednisone |
Footnotes
Stata, version 13.1, StataCorp LP, College Station, Tex.
Multidrug resistance (MDR1) gene test, Veterinary Clinical Pharmacology Lab, College of Veterinary Medicine, Washington State University, Pullman, Wash.
References
1. Ito D, Frantz AM, Modiano JF. Canine lymphoma as a comparative model for human non-Hodgkin lymphoma: recent progress and applications. Vet Immunol Immunopathol 2014; 159: 192–201.
2. Vail DM, Pinkerton ME, Young KM. Canine lymphoma and lymphoid leukemia. In: Withrow SJ, Vail DM, Page RL, eds. Small animal clinical oncology. 5th ed. St Louis: Elsevier-Saunders, 2013;608–638.
3. Garrett LD, Thamm DH, Chun R, et al. Evaluation of a 6-month protocol with no maintenance for dogs with lymphoma. J Vet Intern Med 2002; 16: 704–709.
4. Hosoya K, Kisseberth WC, Lord LK, et al. Comparison of COAP and UW-19 protocols for dogs with multicentric lymphoma. J Vet Intern Med 2007; 21: 1355–1363.
5. Sorenmo K, Overly B, Krick E, et al. Outcome and toxicity associated with a dose-intensified, maintenance-free CHOP-based chemotherapy protocol in canine lymphoma: 130 cases. Vet Comp Oncol 2010; 8: 196–208.
6. Price GS, Page RL, Fischer BM, et al. Efficacy and toxicity of doxorubicin/cyclophosphamide maintenance therapy in dogs with multicentric lymphosarcoma. J Vet Intern Med 1991; 5: 259–262.
7. Dobson JM, Gorman NT. Canine multicentric lymphoma 2: comparison of response to two chemotherapeutic protocols. J Small Anim Pract 1994; 35: 9–15.
8. Marconato L, Stefanello D, Valenti P, et al. Predictors of long-term survival in dogs with high-grade multicentric lymphoma. J Am Vet Med Assoc 2011; 238: 480–485.
9. Zandvliet M, Rutteman GR, Teske E. Prednisolone inclusion in a first-line multidrug cytostatic protocol for the treatment of canine lymphoma does not affect therapy results. Vet J 2013; 197: 656–661.
10. Bergman PJ. Mechanisms of anticancer drug resistance. Vet Clin North Am Small Anim Pract 2003; 33: 651–667.
11. Schuetz JD, Silverman JA, Thottassery JV, et al. Divergent regulation of the class II P-glycoprotein gene in primary cultures of hepatocytes versus H35 hepatoma by glucocorticoids. Cell Growth Differ 1995; 6: 1321–1332.
12. Altuvia S, Stein WD, Goldenberg S, et al. Targeted disruption of the mouse mdr1b gene reveals that steroid hormones enhance mdr gene expression. J Biol Chem 1993; 268: 27127–27132.
13. Bergman PJ, Ogilvie GK, Powers BE. Monoclonal antibody C219 immunohistochemistry against P-glycoprotein: sequential analysis and predictive ability in dogs with lymphoma. J Vet Intern Med 1996; 10: 354–359.
14. Lee JJ, Hughes CS, Fine RL, et al. P-glycoprotein expression in canine lymphoma: a relevant, intermediate model of multidrug resistance. Cancer 1996; 77: 1892–1898.
15. Schulz KF, Altman DG, Moher D, et al. CONSORT 2010 statement: updated guidelines for reporting parallel group randomized trials. Ann Intern Med 2010; 152: 726–732.
16. Moher D, Hopewell S, Schulz KF, et al. CONSORT 2010 explanation and elaboration: updated guidelines for reporting parallel group randomised trials (Erratum published in BMJ 2011;343:d6131). BMJ 2010; 340:c869.
17. Mealey KL, Meurs KM. Breed distribution of the ABCB1-1Δ (multidrug sensitivity) polymorphism among dogs undergoing ABCB1 genotyping. J Am Vet Med Assoc 2008; 233: 921–924.
18. Valli VE, San Myint M, Barthel A, et al. Classification of malignant lymphomas according to the World Health Organization Criteria. Vet Pathol 2011; 48: 198–211.
19. Vara JA, Miller MA, Valli VE. Immunohistochemical detection of multiple myeloma 1/interferon regulatory factor 4 (MUM1/IRF-4) in canine plasmacytoma: comparison with CD79a and CD20. Vet Pathol 2007; 44: 875–884.
20. Owen LN. TNM classification of tumors in domestic animals. Geneva: World Health Organization, 1980.
21. Warry E, Hansen RJ, Gustafson DL, et al. Pharmacokinetics of cyclophosphamide after oral and intravenous administration to dogs with lymphoma. J Vet Intern Med 2011; 25: 903–908.
22. Vail DM, Michels GM, Khanna C, et al. Response evaluation criteria for peripheral nodal lymphoma in dogs (v1.0)—a Veterinary Cooperative Oncology Group (VCOG) consensus document. Vet Comp Oncol 2010; 8: 28–37.
23. Veterinary Cooperative Oncology Group—common terminology criteria for adverse events (VCOG-CTCAE) following chemotherapy or biological antineoplastic therapy in dogs and cats v1.1 [published online ahead of print Jul 20, 2011]. Vet Comp Oncol doi: 10.1111/j.1476-5829.2011.00283.x.
24. Valli VE, Kass PH, San Myint M, et al. Canine lymphomas: association of classification type, disease stage, tumor subtype, mitotic rate, and treatment with survival. Vet Pathol 2013; 50: 738–748.
25. Valli VE. B-cell neoplasms. In: Veterinary comparative hematopathology. Ames, Iowa: Blackwell Publishing, 2007;119–274.
Appendix 1
Histopathologic classification of canine lymphomas according to World Health Organization criteria.
| B-cell neoplasms |
| Precursor B-cell neoplasms |
| Precursor B lymphoblastic leukemia/lymphoma |
| Mature B-cell neoplasms |
| B-cell chronic lymphocytic leukemia/prolymphocytic leukemia/small lymphocytic lymphoma |
| B-cell prolymphocytic leukemia |
| Lymphoplasmacytic lymphoma |
| Splenic marginal zone B-cell lymphoma |
| Plasma cell myeloma/plasmacytoma |
| Extranodal marginal zone B-cell lymphoma of mucosa-associated lymphoid tissue |
| Nodal marginal zone lymphoma |
| Follicular lymphoma |
| Mantle cell lymphoma |
| Diffuse large B-cell lymphoma |
| Mediastinal large B-cell lymphoma |
| Burkitt-like lymphoma/Burkitt-like cell leukemia |
| Primary effusion lymphoma |
| T-cell and putative natural killer cell neoplasms |
| Precursor T-cell neoplasms |
| Precursor T lymphoblastic lymphoma/leukemia |
| Mature (peripheral) T-cell and natural killer cell neoplasms |
| T-cell prolymphocytic leukemia |
| Mature (peripheral) nodal T-cell (T-zone) lymphoma |
| Large granular lymphocyte leukemia |
| Aggressive natural killer cell leukemia |
| Peripheral T-cell lymphomas, not otherwise specified |
| Adult T-cell lymphoma/leukemia |
| Intestinal T-cell lymphoma (± enteropathy associated) |
| Hepatosplenic γδT-cell lymphoma |
| Subcutaneous panniculitis-like T-cell lymphoma |
| Mycosis fungoides/Sézary syndrome |
| Anaplastic large-cell lymphoma, T- and null-cell primary cutaneous type |
| Angioimmunoblastic T-cell lymphoma |
| Angiocentric T-cell lymphoma |
(Adapted from Valli VE, San Myint M, Barthel A, et al. Classification of malignant lymphomas according to the World Health Organization Criteria. Vet Pathol 2011;48:198–211. Reprinted with permission.)
Appendix 2
World Health Organization staging system for canine lymphomas.
| Stage | Description |
|---|---|
| I | Cancer confined to a single lymph node |
| II | Involvement of many lymph nodes (± tonsils) in a regional area on 1 side of the diaphragm |
| III | Generalized lymph node involvement |
| IV | Any stage I–III with liver or spleen involvement |
| V | Any stage I–IV with blood, bone marrow, or any other organ involvement |
| Substage | A—No systemic signs of illness present |
| B—Systemic signs of illness present |
(Adapted from Owen LN. TNM classification of tumors in domestic animals. Geneva: World Health Organization, 1980. Reprinted with permission.)
