Lymphoma is the most common hematopoietic

Transcription

1 J Vet Intern Med 2007;21: Comparison of COAP and UW-19 Protocols for Dogs with Multicentric Lymphoma Kenji Hosoya, William C. Kisseberth, Linda K. Lord, Francisco J. Alvarez, Ana Lara-Garcia, Carrie E. Kosarek, Cheryl A. London, and C. Guillermo Couto Background: Various chemotherapy protocols for treating lymphoma in dogs have been published; however, comparison of protocols from different studies is difficult, especially when evaluating survival time and toxicoses. Hypothesis: The choice of COAP (C, cyclophosphamide; O, vincristine; A, cytosine arabinoside; P, prednisone) and a modified University of Wisconsin 19-week (UW-19) induction protocol has no influence on overall survival times in dogs with lymphoma. Animals: One hundred and one dogs with multicentric lymphoma. Methods: Retrospective study ( ). Dogs induced with either an 8-week COP-based protocol (C, cyclophosphamide; O, vincristine; and P, prednisone) with maintenance therapy (COAP group) or a 19-week CHOP (C, cyclophosphamide; H, doxorubicin; O, vincristine; and P, prednisone) based protocol (UW-19 group) were compared in terms of the duration of first remission, survival time, toxicoses, and cost. Results: There were 71 dogs in the COAP group and 30 dogs in the UW-19 group. Various protocols were used after the first relapse. The median duration of the first remission for the COAP and UW-19 groups were 94 days (range, days) and 174 days ( days), respectively (P,.01). The median survival times for dogs in the COAP and UW-19 groups were 309 days (6 620 days) and 275 days ( days), respectively (P 5.09). Dogs in the COAP group had a hazard ratio of 1.9 (95% CI ) for death relative to the UW-19 group (P 5.03), after controlling for the confounders (World Health Organization clinical stage, age, sex, use of doxorubicin during reinduction). The severity of neutropenia and gastrointestinal toxicoses were significantly higher in the UW-19 group than in the COAP group (P 5.01 and P,.01, respectively). Conclusion and Clinical Importance: Use of a long-term doxorubicin-containing sequential combination chemotherapy protocol is associated with a decreased risk of relapse and death relative to a non-doxorubicin-containing protocol. Key words: Cancer; Chemotherapy; CHOP; COP; Cost. Lymphoma is the most common hematopoietic neoplasm in dogs, occurring in 13 to 24 per 100,000 dogs. 1 The current standard of treatment for dogs with lymphoma is sequential combination chemotherapy, and the results of treatment with various chemotherapy protocols have been published L- asparaginase, cyclophosphamide, vincristine, prednisone, and doxorubicin are the most effective drugs, and some or all of these agents are incorporated into most induction chemotherapy protocols. During the past 3 decades, the standard of care has shifted from socalled COP-based protocols (C, cyclophosphamide; O, vincristine; and P, prednisone) to CHOP-based protocols (C, cyclophosphamide; H, doxorubicin; O, vincristine; and P, prednisone). It is generally accepted that CHOP-based protocols are associated with longer From the Department of Veterinary Clinical Sciences and Veterinary Teaching Hospital (Hosoya, Kisseberth, Alvarez, Lara- Garcia, Kosarek, London, Couto); and Department of Veterinary Preventive Medicine (Lord), College of Veterinary Medicine, The Ohio State University, Columbus, OH. Dr Kosarek is presently affiliated with the Department of Small Animal Medicine and Surgery, College of Veterinary Medicine, University of Georgia, Athens, GA. Presented in part at the 26th Annual Conference of the Veterinary Cancer Society, Pine Mountain, GA, October 19 22, Reprint requests: William C. Kisseberth, DVM, PhD, DACVIM (Oncology), Department of Veterinary Clinical Sciences and Veterinary Teaching Hospital, College of Veterinary Medicine, The Ohio State University, Columbus, OH 43210; kisseberth.2@ osu.edu. Submitted November 28, 2006; Revised March 14, 2007, May 2, 2007, June 2, 2007; June 29, Copyright E 2007 by the American College of Veterinary Internal Medicine /07/ /$3.00/0 median duration of remission than COP-based protocols. Median duration of remission that range from 3.3 to 6.0 months have been reported with COP-based protocols, whereas the median duration of remission that range from 5.0 to 10.9 months have been reported with CHOP-based protocols. 2 15,17 23 The role of maintenance therapy after the initial induction period also has been the subject of debate and investigation. The addition of maintenance therapy immediately after induction does not improve the duration of first remission or overall survival for dogs induced with different doxorubicin-based combination chemotherapy protocols. 11,13 15,18 The effect of maintenance or no maintenance in COP-based protocols has not been formally evaluated, although anecdotal experience suggests that, without some form of maintenance therapy, dogs treated with COP-based protocols often relapse soon after induction. Although the duration of remission is an objective measurement and can be used to compare the initial efficacy of induction protocols, it is less clear whether a longer duration of first remission correlates with longer overall survival in dogs with multicentric lymphoma. Survival time, as opposed to the duration of first remission, is influenced by many factors, including the reinduction or rescue protocols used and the owner s willingness to continue treatment. Historically, longer survival times have been reported with CHOP-based protocols, ranging from 5.8 to 17.0 months, compared with 7.3 to 9.7 months with COP-based protocols. 2 7,11,13 15,17 However, these reports were published over a 30-year period, during which time the quality of animal care and the average owner s commitment to their pet s well being may have in-

2 1356 Hosoya et al creased. Furthermore, in most instances, comparisons between protocols have been made between studies done at different institutions, by different oncologists, and from different animal populations. Recently, a retrospective study by this group suggested that there was no difference in survival times between dogs with lymphoma induced with COP- versus CHOP-based protocols. 9 That study only included a small number of dogs who received CHOP-based protocol, and the CHOP-based protocol was selectively used for cases that were thought to be less responsive to COP-based chemotherapy, such as mediastinal form or gastrointestinal form. Two protocols (COAP [C, cyclophosphamide; O, vincristine; A, cytosine arabinoside; P, prednisone] protocol and a modified University of Wisconsin- Madison 19-week protocol [UW-19]) have been used since 2001 as the initial induction protocols in dogs with multicentric lymphoma treated at The Ohio State University Veterinary Teaching Hospital (OSU-VTH), where the choice of initial induction protocol largely was dependent on the admitting clinician s and owner s preferences. After choosing the induction protocol, dogs were managed by the same group of clinicians, thus potentially eliminating many of the biases inherent in comparing dogs treated with different protocols at different institutions. Several reinduction or rescue protocols were equally available at the time of relapse and also were chosen based on owner preference and the judgment of the clinician who was assigned to the care of the dog at the time of relapse. The purpose of the study was to compare the durations of first remission and survival in dogs with multicentric lymphoma induced with either COAP with maintenance therapy or UW-19 protocol without maintenance therapy. We hypothesized that UW-19 protocol would be associated with longer duration of first remission than COAP but that the choice of initial induction protocol would have no significant influence on overall survival time. Materials and Methods Dog Population and Selection Medical records of dogs with multicentric lymphoma treated at the OSU-VTH from January 1, 2001, to January 31, 2006, were retrospectively reviewed. Inclusion criteria were (1) histologic or cytologic diagnosis of lymphoma; (2) initiation of induction chemotherapy during the study period; (3) clinical manifestation of multicentric nodal involvement, that is, lymphomas that primarily involve extranodal sites were excluded; (4) absence of a previous history of chemotherapy other than corticosteroids. The diagnosis of lymphoma was made on cytology of the enlarged lymph nodes, histology of lymph-node biopsy, or both. Immunophenotype (B or T cell) was determined by immunohistochemistry by using antibodies to CD3 (T-cell marker a ) and CD 79a and BLA 36 (B-cell markers b ) flow cytometry by using multiple cellsurface markers, or clonality assay by using polymerase chain reaction All dogs were retrospectively staged by using the World Health Organization (WHO) clinical classification system for canine lymphoma, on the basis of the available records of some or all of the following: physical examination, CBC count, thoracic radiographs, abdominal ultrasonography, and bone marrow aspiration cytology. 27 Bone marrow aspiration was not routinely performed if CBC count changes indicated bone marrow involvement. Therefore, dogs were considered to have bone marrow involvement when (1) presence of neoplastic lymphocytes was demonstrated by bone marrow cytology or (2) thrombocytopenia (, platelets/l; reference range, platelets/l) with or without neutropenia (, cells/l; reference range, cells/l), and the presence of circulating neoplastic lymphoid cells were documented in the peripheral blood smear. Dogs also were classified as substage a (absence of or mild clinical illness) or substage b (moderate-to-severe lethargy or other systemic signs). Clinical, hematologic, or biochemical abnormalities without systemic illness, evidenced by ocular signs, cytopenias, hypercalcemia were not by themselves considered as criteria for substage b. Information extracted from the medical records included breed, age, sex, hormonal status (intact versus neutered), weight, WHO clinical stage (I V), substage (a or b), immunophenotype (B versus T), presence of extranodal involvement (eye/central nervous system [CNS]/testicle, bone marrow, other), and presence of hypercalcemia (.13.0 mg/dl; reference, mg/dl). Treatment Protocols Two chemotherapy protocols, COAP (C, cyclophosphamide; O, vincristine; A, cytosine arabinoside; P, prednisone) and a modified University of Wisconsin 19-week (UW-19) protocols, were used (Table 1). 9,15 For dogs on the COAP protocol, LMP (L, chlorambucil; M, methotrexate; P, prednisone) maintenance 9 was used at the conclusion of induction if the animal was in remission. For dogs on the UW-19 protocol, chemotherapy was discontinued after completion of the induction phase. In dogs who developed hemorrhagic cystitis, chlorambucil was substituted for cyclophosphamide. Reinduction or rescue protocols (Table 2) included vincristinebased protocols (COP, CLOP, COAP, LMP/vincristine, and LMP/ vincristine/l-asparaginase protocols), doxorubicin-based protocols (single agent doxorubicin, doxorubicin/cyclophosphamide [AC], CHOP, UW-19, and modifications of UW-19), CCNU (lomustine)- based protocols (single agent CCNU; CCNU/L-asparaginase; CCNU/vincristine; and CVM: cyclophosphamide, vincristine, methotrexate); DMAC (dexamethasone, melphalan, actinomycin D, cytosine arabinoside); and others (single agent L-asparaginase, cytosine arabinoside/mitoxantrone, and investigational agents). 28 Tumor Response Response was categorized as complete response (CR), evident as complete resolution of disease; partial remission (PR), evident as.50% but,100% reduction of lymph node size; stable disease (SD), evident as,50% reduction or,25% increase of lymph-node size; progressive disease (PD), evident as.25% increase of lymphnode size or development of a new extranodal lesion. Remission duration of at least 3 weeks was required for classification of CR or PR. Response rate (RR) was defined as the percentage of the dogs who achieved CR or PR with a given protocol among all dogs who received the same protocol. Duration of remission for a given protocol was defined as the time, in days, from the start of the protocol to development of progressive disease. Censored events for the duration of remission analysis included death unrelated to the disease and loss to follow-up with CR or PR. Survival time was defined as the time, in days, from the initiation of chemotherapy to the time of death. Censored events for the survival time analysis included alive at the time of analysis (October 1, 2006), death unrelated to the disease, loss to follow-up immediately after initiation of the first or second reinduction protocol, and loss to follow-up with CR or PR. Loss to follow-up while the disease was in SD or PD was considered as a disease-related death, and counted at the time of the last contact.

3 Protocols for Canine Lymphoma 1357 Table 1. Induction protocols for dogs with multicentric lymphoma. Protocol COAP Cytosine arabinoside (300 mg/ m 2 SC/IV drip) N Vincristine ( mg/m 2 IV) N N N N N N N N Cyclophosphamide (50 mg/m 2 PO EOD) N N N N N N N N Prednisone a N N N N N N N N LMP Chlorambucil (20 mg/m 2 PO) N N N N N N Methotrexate (2.5 5 mg/m 2 PO twice/wk) N N N N N N N N N N N Prednisone (20 mg/m 2 PO EOD) N N N N N N N N N N N UW-19 L-asparaginase (400 mg/kg IM) N Vincristine ( mg/m 2 IV) N N N N N N N N Cyclophosphamide (200 mg/ m 2 IV) N N N N Doxorubicin (30 mg/m 2 or 1 mg/kg IV) b N N N N Prednisone c N N N N a 40 mg/m 2 PO q24h for 7 days, then 20 mg/m 2 PO q48h thereafter. b 30 mg/m 2 for dogs.10 kg, 1 mg/kg for dogs,10 kg. c 2 mg/kg PO q24h for 7 days, 1.5 mg/kg PO q24h for 7 days, 1 mg/kg PO q24h for 7 days, then 0.5 mg/kg PO q24h for 7 days. Week Toxicosis Hematologic toxicoses were assessed by reviewing the CBC counts results and recorded at each treatment according to the Veterinary Co-operative Oncology Group-Common Terminology Criteria for Adverse Events v. 1.0 (VCOG-CTCAE), with minor modification, in that sepsis was classified as grade 5 neutropenia and was distinguished from grade 4 neutropenia without development of sepsis. 29 Gastro- Table 2. Reinduction protocols. Vincristine-based protocols COAP protocol COP protocol LMP/vincristine Doxorubicin-based protocols Single agent doxorubicin AC protocol CHOP protocol UW-19 protocol DMAC protocol CCNU-based protocols Single agent CCNU CCNU/vincristine protocol CCNU/vincristine/methotrexate (CVM) See Table 1 induction protocols COAP without cytosine arabinoside Add vincristine mg/m 2 IV q2 wk to LMP Doxorubicin 30 mg/m 2 IV q3 wk Day 1: Doxorubicin 30 mg/m 2 IV q3 wk Day 10: Cyclophosphamide 200 mg/m 2 PO Day 1: Doxorubicin 30 mg/m 2 IV q3 wk Day 8: Vincristine 0.75 mg/m 2 IV Day 10: Cyclophosphamide 200 mg/m 2 PO Day 15: Vincristine 0.75 mg/m 2 IV See Table 1 induction protocols Day 1: Actinomycin-D 0.75 mg/m 2 IV Cytosine arabinoside 300 mg/m 2 SC Dexamethasone 2.2 mg/kg SC Day 14: Melphalan 20 mg/m 2 PO Dexamethasone 2.2 mg/kg PO CCNU mg/m 2 PO q3 wk Day 1: CCNU mg/m 2 PO q3 wk Day 10: Vincristine mg/m 2 IV Add methotrexate mg/m 2 PO twice/wk in CCNU/vincristine protocol.

4 1358 Hosoya et al intestinal toxicoses were assessed by reviewing the descriptions of the clinical signs and therapy instituted in the medical record. Gastrointestinal toxicoses were graded after each dose of chemotherapy according to VCOG-CTCAE. Toxicosis was scored as 0 if the dog had abnormalities before treatments in the given parameter, unless it progressed to the next grade after the treatment. Toxicosis grades for a given protocol were expressed as the highest grade during the period from the first dose of the protocol to subsequent progression of the disease. Toxicoses recorded while dogs were in PD were not taken into consideration in the analysis, because it was difficult to assess whether abnormality related to therapy or disease. Cost of the Treatment For dogs who received both treatment and management of adverse effects entirely at the OSU-VTH, the bill during the entire remission period (including the period during and after reinductions until the ultimate progression of the disease) was extracted from the OSU-VTH accounting database. Dogs who were presented with hypercalcemia were excluded from the analysis because of the cost associated with initial hospitalization. Although there were some minor changes in hospital charges during the study period ( ; typically approximately a 5% increase per year), this factor was not taken into account in the analysis. Statistical Analysis Comparisons were made between categorical variables by using the chi-square test. The Fisher exact test was used for categorical variables when the expected value of a given cell in the comparison was less than 5. Comparisons between continuous variables were made by using the Student s t-test. The severity of toxicoses was compared by using the Mann-Whitney rank sum test and the frequency of the toxicoses $grade 3 was compared between data sets by using the chi-square test. The total cost during the entire remission period were plotted against the time and compared by the Student s t-test after excluding the outliers. The duration of remission and survival time curves were generated by using the Kaplan-Meier product limit method, and median duration of first remission and survival time were calculated. Comparisons between treatment groups were made by using the log-rank test. Cox proportional hazards models were used to compare duration of first remission and survival time between the 2 treatment groups after adjusting for potential confounders. Potential confounders that were evaluated in the models included age (in years), hormonal status (intact or neutered), sex, body weight, immunophenotype (B cell, T cell, or unknown), WHO clinical stage (stage I IV or V), substage (a or b), primary faculty clinician, presence of bone marrow involvement, presence of hypercalcemia, year of admission, use of corticosteroid before initiation of chemotherapy, and use of doxorubicin during reinduction period. Because the primary comparison focused on the difference between the 2 treatment groups, the variable for treatment group was forced into the initial model. Bivariate Cox proportional hazards regression model analyses were performed to screen potential confounding variables for subsequent inclusion in the multivariate analysis. Variables with P values #.25 in these bivariate analyses were included in the multivariate analysis. Variables were removed from the full multivariate model on the basis of results of the likelihood ratio test. Biologically meaningful interactions between the main effect variables in the model were tested for inclusion in a similar manner. Before the model building process, the proportional hazards assumption was tested on each variable by means of Schoenfeld residuals and graphical techniques to determine whether the log hazard ratio function was constant over time. 30 After the model building process, the assumption of proportional hazards was again tested for each variable included in the final model. Table 3. Distribution of WHO clinical stage, substage, immunophenotype, and hypercalcemia. COAP Group, no. (%) (n 5 71) In all analyses, a P value,.05 was considered statistically significant. Statistical analyses were performed by using Prism 4 c and Stata version 9.1. d Results UW-19 Group, no. (%) (n 5 30) P Value Staging methods CBC/ blood smear 71 (100) 30 (100) 1.00 Thoracic radiographs 37 (52) 24 (80).01 Abdominal ultrasound 34 (48) 22 (73).03 Bone marrow 4 (6) 6 (20).06 aspiration WHO stage.29 I 1 (1) 0 (0) II 2 (3) 0 (0) III 26 (37) 7 (23) IV 22 (31) 13 (43) V 20 (28) 10 (33) Substage.15 a 48 (68) 25 (83) b 23 (32) 5 (17) Extranodal involvement Bone marrow 8 (11) 1 (3).27 Eye/CNS/testicle 10 (14) 1 (3).17 Other sites 7 (10) 8 (27).11 Immunophenotype,.01 B cell 34 (39) 27 (84) T cell 13 (15) 1 (3) Non-B/non-T/ 42 (47) 4 (13) unknown Hypercalcemia 10 (14) 0 (0).03 COAP, cyclophosphamide, vincristine, cytosine arabinoside, prednisone; UW-19, University of Wisconsin 19-week induction protocol; CNS, central nervous system. Dogs One hundred and one dogs met the inclusion criteria: 71 dogs (70%) were induced with COAP protocol (COAP group), and 30 dogs (30%) with the UW-19 protocol (UW- 19 group). Breeds included were mixed breed (13), Golden Retriever (9), Labrador Retriever (6), Boxer (4), Rottweiler (4), Doberman Pinscher (4), and German Shepherd Dog (4) in the COAP group; and mixed breed (8), Golden Retriever (4), Shih Tzu (3), Labrador Retriever (2), and Rottweiler (2) in the UW-19 group. Age was years (mean 6 SD) and years in the COAP and UW-19 groups, respectively (P 5.47). Body weight was kg and kg in the COAP and UW-19 groups, respectively (P 5.86). There were 37 males (6 intact) and 34 females (1 intact) in the COAP group, and 13 males (2 intact) and 17 females (1 intact) in the UW-19 group (P 5.51). There was no significant difference in the proportions of neutered animals between groups (P ). The methods of staging and the distribution of the WHO clinical stage, substage, immunophenotype, and presence of hypercalcemia are summarized in Table 3.

5 Protocols for Canine Lymphoma 1359 Table 4. Distribution of reinduction protocols. No. Reinduction Attempt (%) Reinduction protocol 2nd 3rd 4th 5th 6th COAP group Vincristine based 32 (45) 2 (3) 0 (0) 0 (0) 0 (0) Doxorubicin based 4 (6) 4 (6) 8 (11) 4 (6) 2 (3) CCNU based 13 (18) 13 (18) 6 (8) 4 (6) 0 (0) DMAC 4 (6) 16 (23) 5 (7) 2 (3) 0 (0) Other 0 (0) 2 (3) 1 (1) 1 (1) 0 (0) Total 53 (75) 37 (52) 20 (28) 11 (15) 2 (3) UW-19 group Vincristine-based 7 (23) 1 (3) 2 (7) 1 (3) 0 (0) Doxorubicin-based 6 (20) 1 (3) 0 (0) 1 (3) 0 (0) CCNU-based 6 (20) 4 (13) 1 (3) 1(3) 0 (0) DMAC 4 (13) 7 (23) 1 (3) 0 (0) 0 (0) Other 0 (0) 1 (3) 2 (7) 1 (3) 1 (3) Total 23 (77) 15 (50) 6 (20) 4 (13) 1 (3) COAP, cyclophosphamide, vincristine, cytosine arabinoside, prednisone; CCNU, lomustine; DMAC, dexamethasone, melphalan, actinomycin D, cytosine arabinoside; UW-19, University of Wisconsin 19-week induction protocol. Thoracic radiographs and abdominal ultrasonography were more frequently performed in the dogs in the UW- 19 group than in the COAP group (P 5.01, and.03, respectively). There were no significant differences in WHO clinical stage (P 5.29), substage (P 5.15), or extranodal involvement (bone marrow, P 5.27; ocular/ CNS, P 5.17; other sites, P 5.11) between the COAP and UW-19 groups. Hypercalcemia was more common in dogs in the COAP group than in the UW-19 group (P 5.03). Similarly, there were significantly more dogs with T-cell lymphoma in the COAP group than in the UW-19 group (P,.01), although immunophenotype was less frequently known in the COAP group than the UW-19 group (P,.01). Six dogs received a short course of corticosteroid for their lymphoma before presentation (,2 weeks), and additional 2 dogs were chronically on corticosteroid for reasons other than lymphoma. Reinduction Protocols Eighteen dogs (25%) in the COAP group and 7 dogs (23%) in the UW-19 group did not receive reinduction chemotherapy. The percentage of dogs that underwent the 2nd, 3rd, 4th, 5th, and 6th induction chemotherapy were 75, 52, 28, 15, and 3% in COAP group, respectively, and 77, 50, 20, 13, and 3% in UW-19 group, respectively (Table 4). There was no difference in the proportion of dogs who received these reinduction chemotherapy protocols between the COAP and UW-19 groups (P 5.79). Duration of First Remission The initial response in the COAP group was CR in 43 dogs (61%), PR in 22 dogs (31%), SD in 2 dogs (3%), and PD in 4 dogs (6%), resulting in a RR of 92%. The initial response in the UW-19 group was CR in 23 dogs (77%) and PR in 7 dogs (23%), resulting in a RR of 100%. There was no significant difference in CR rate or RR between groups (P 5.17 and.18, respectively). However, the median duration of first remission was significantly shorter in the COAP group at 94 days (95% CI days) than the UW-19 group at 174 days (95% CI days) (P,.01) (Fig 1). The WHO clinical stage (I IV versus V) was identified as a significant confounder in the multivariate Cox proportional hazards model (Table 5). After adjusting for the WHO clinical stage, dogs in the COAP group were found to have a 2.6 times higher risk for relapse than dogs in the UW-19 group (95% CI , P,.01). Overall Survival Time The median survival times for dogs in the COAP and UW-19 groups were 309 days (95% CI days) and 275 days (95% CI days), respectively (P 5.09) (Fig 2). The 1- and 2-year survival rate for the Fig 1. Kaplan-Meier curves for the duration of first remission for dogs with multicentric lymphoma. Dogs were treated with COAP (n 5 71, solid line) or modified UW-19 protocol (n 5 30, dashed line).

6 1360 Hosoya et al Table 5. Final multivariate Cox proportional hazards models for the analysis of duration of first remission and survival time for comparison of COAP to UW-19 protocols for treatment of canine lymphoma. Model for duration of first remission Variable Treatment (COAP versus referent group UW-19) WHO clinical stage (stage V versus referent group I IV) Hazard Ratio (95% CI) P Value , Note: Other potential confounders that did not enter the final model were age (in years), hormonal status (intact or neutered), sex, body weight, immunophenotype (B cell, T cell, or unknown), substage (a or b), primary faculty clinician, presence of bone marrow involvement, presence of hypercalcemia, year of admission, use of corticosteroid before initiation of chemotherapy, and use of doxorubicin during reinduction period. Fig 2. Kaplan-Meier curves for the survival times for dogs with multicentric lymphoma. Dogs were induced with COAP (n 5 71, solid line) or modified UW-19 protocol (n 5 30, dashed line). Various protocols were used after the first relapse. Model for survival time Variable Hazard Ratio (95% CI) P Value Treatment (COAP versus referent group UW-19) Sex (male versus referent group ,.01 female) WHO clinical stage (stage V versus referent group I IV) Age (in years) Use of doxorubicin during reinduction Note: Other potential confounders that did not enter the final model were hormonal status (intact or neutered), body weight, immunophenotype (B cell, T cell, or unknown), substage (a or b), primary faculty clinician, presence of bone marrow involvement, presence of hypercalcemia, year of admission, and use of corticosteroid before initiation of chemotherapy. COAP, cyclophosphamide, vincristine, cytosine arabinoside, prednisone; UW-19, University of Wisconsin 19-week induction protocol; WHO, World Health Organization. COAP group was 32 and 0%, respectively. The 1- and 2- year survival rate for the UW-19 group was 45 and 17%, respectively. The results were similar when only dogs with B-cell lymphoma were compared (data not shown). When T- and B-cell lymphomas within the COAP group (n 5 11 and 24, respectively) were compared, there was no significant difference in survival time between immunophenotypes (B cell, 321 days; T cell, 378 days; P 5.58). A comparison of survival times in dogs with B- and T-cell lymphomas was not performed in UW-19 group, because there was only 1 dog with T-cell lymphoma in the group. The WHO clinical stage (I IV favorable to V), sex (female favorable to male), age (younger dogs favorable to older dogs), and use of doxorubicin during reinduction (dogs received doxorubicin at any time during reinduction phase favorable to dogs who did not) were identified as significant confounders in the multivariate Cox proportional hazards model (Table 5). After adjusting for these confounders, dogs in the COAP group were found to have a 1.9 times higher risk for death compared with the dogs in the UW-19 group (95% CI , P 5.03). Toxicoses The mean neutropenia and gastrointestinal toxicosis grades were significantly higher in the UW-19 group (P 5.01 and P,.01, respectively) (Table 6). Hematologic toxicoses $grade 3 occurred in 7 dogs (10%) in the COAP group and in 7 dogs (23%) in the UW-19 group. Gastrointestinal toxicoses $grade 3 occurred in 9 dogs (13%) in the COAP group and in 7 dogs (23%) in the UW-19 group. Other toxicoses included acute tumor lysis syndrome (1 dog) in the COAP group, and hemorrhagic cystitis (2 dogs) and acute tumor lysis syndrome (1 dog) in the UW-19 group. The differences in frequency of $grade 3 neutropenia (P 5.08), thrombocytopenia (P 5.15), and $grade 3 gastrointestinal toxicoses (P 5.23) between the 2 groups were not statistically significant. Cost Analysis Thirty-four dogs in the COAP group and 22 dogs in the UW-19 group received all treatment and management of toxicoses at the OSU-VTH from the initiation of the induction protocol until the final disease progression or the loss to follow-up. Two dogs had exceptionally long total durations of remission (794 and 1102 days) in the UW-19 group and were excluded from the cost comparison. For the remaining dogs, the mean (6SD) duration of entire remission (or follow-up time) for the COAP and UW-19 groups were days and days, respectively (P 5.30). The mean (6SD) total cost during these period for the

7 Protocols for Canine Lymphoma 1361 Table 6. Distribution of toxicosis grade during the initial induction (and maintenance in COAP group) period. Hematologic toxicoses COAP Group, no. (%) (n 5 71) COAP and UW-19 groups were $3,515 6 $366 and $3,581 6 $432, respectively (P 5.91). Discussion UW-19 Group, no. (%) (n 5 30) Neutropenia Grade 1 19 (27) 9 (30) Grade 2 2 (3) 4 (13) Grade 3 2 (3) 3 (10) Grade 4 3 (4) 3 (10) Grade 5 0 (0) 0 (0) Thrombocytopenia Grade 1 3 (4) 0 (0) Grade 2 1 (1) 2 (7) Grade 3 1 (1) 1 (3) Grade 4 0 (0) 0 (0) Anemia Grade 1 30 (42) 16 (53) Grade 2 7 (10) 5 (17) Grade 3 1 (1) 0 (0) Grade 4 0 (0) 0 (0) Gastrointestinal toxicoses Grade 1 16 (23) 8 (27) Grade 2 10 (14) 9 (30) Grade 3 8 (11) 5 (17) Grade 4 1 (1) 2 (7) COAP, cyclophosphamide, vincristine, cytosine arabinoside, prednisone; UW-19, University of Wisconsin 19-week induction protocol. This study compared 2 groups of dogs with multicentric lymphoma induced with either COAP and LMP maintenance chemotherapy or the UW-19 protocol without maintenance for induction of remission at a single institution over the same time period. Use of the UW-19 protocol as the initial induction protocol was associated with a longer duration of first remission and a significant survival advantage compared with dogs treated with the COAP protocol and LMP maintenance. However, the severity of neutropenia and gastrointestinal toxicoses with the UW-19 protocol were significantly higher than with the COAP protocol. The patient population was comparable in the 2 groups in this study; however, although the choice of initial induction protocol was determined in part by clinician preference, there are several potential animalselection biases. For example, owners with financial limitations or those who were unsure about pursuing chemotherapy for their pets may have tended to select the COAP protocol rather than the UW-19 protocol as an induction. Furthermore, COAP protocol is preferred by some clinicians in our practice for dogs with ocular, CNS, or both involvement, or with hypercalcemia. Also, during this time period, there was a clinical study in progress that required induction with UW-19 for enrollment and dogs with T-cell lymphomas or with hypercalcemia were more commonly treated with COAP. Because, historically, dogs with the T-cell phenotype have been reported to have a poor prognosis, this selection bias could have skewed the remission and survival data in favor of the UW-19 group; however, the duration of remission and the survival time curves were not different between phenotypes in the COAP group in animals where immunophenotyping was available. 22,31 34 Differences in staging methods may have influenced the WHO clinical stage distribution in the 2 groups; abdominal ultrasonography and thoracic radiographs were more commonly performed in dogs in the UW-19 group, also reflecting individual clinician preferences. However, this difference is unlikely to have had any impact on the remission and survival data, because dogs with lymphoma in stages III and IV historically have the same prognosis, and only 1 dog in COAP group was classified as stage I, without results of thoracic radiographs and abdominal ultrasonography. 17,35 Because many of stage V cases were classified as such because of extranodal involvement found in thoracic radiographs or abdominal ultrasonography, the Cox proportional hazards model was repeated by using only bone marrow involvement as the classification of stage V, and the results were similar (data not shown). Finally, evaluating dogs at only one practice, particularly a tertiary-referral practice, could mean that the types of cases included are not typical of dogs with lymphoma seen by a nonspecialty general practice. Duration of first remissions and survival times for the COAP and UW-19 protocols in this study generally were consistent with previously published studies that used the same protocols, although the duration of first remission and survival time for the UW-19 protocol were somewhat shorter than that reported for other versions of the Wisconsin protocol. 4,9,14,15 In another retrospective study where the UW-19 protocol was used, the first remission duration and overall survival time were 206 and 310 days, respectively, comparable with this study. 15 Duration of first remission was calculated from the beginning of chemotherapy induction to relapse. Although duration of remissions were longer with the UW-19 protocol, because it is considerably longer (19 weeks) than that of COAP protocol (8 weeks), the median duration of remission after completion of the induction protocol was similar (44 days in COAP protocol and 48 days in UW-19 protocol) in the 2 protocols. Comparison of survival time was complicated by the various reinduction protocols used, and, thus, care must be taken when interpreting these data. Because all dogs were treated during the same time period by the same group of clinicians and there were no differences in the proportion of dogs who underwent reinduction or in the type of reinduction protocols used, we believe this is a valid comparison. The median survival times were not significantly different between the COAP and UW-19 groups in the univariate analysis. In fact, the survival curves were nearly identical up to approximately 300 days, although

8 1362 Hosoya et al the 1- and 2-year survival rates were higher in the UW-19 group than the COAP group. Furthermore, when the data were adjusted for significant confounders (age, sex, use of doxorubicin during reinduction, and WHO clinical stage), the dogs in the COAP group had a significantly higher risk for death than the dogs in UW-19 group. Sex and WHO clinical stage have also been reported to be prognostic factors in other studies. 3,4,19,20,36 These findings might indicate that although the influence of the choice of initial induction protocol on survival is relatively small for the majority of dogs, there is a portion of dogs who have a long-term survival advantage by being treated with the UW-19 protocol for initial induction. Further studies are warranted to test this hypothesis. The frequency of $grade 3 hematologic or gastrointestinal toxicoses was not different between the 2 induction protocols, although the retrospective assessment of gastrointestinal toxicosis requires a careful interpretation. The frequency of toxicoses was expected to be higher with UW-19 protocol, but the lack of significance could be explained by the low frequency of such toxicosis in both protocols. Although the difference was not statistically significant between the 2 protocols in the current study, the actual frequency was similar to the previous studies (8.7% with COP- and 17% with CHOP-based protocols). 9,17 Although the frequency of $grade 3 toxicoses did not differ significantly, the mean grade of neutropenia and gastrointestinal toxicoses were significantly greater in UW-19 protocol than COAP protocol. This is expected, because the UW-19 protocol contains doxorubicin and the treatment protocol is considerably longer than the COAP protocol. Although the COAP protocol initially was associated with lower treatment costs and potentially lower toxicosis management cost, most dogs treated with COAP received subsequent reinduction protocols sooner than dogs treated with UW-19, therefore, the total cost became similar in the long term. In fact, dogs in both groups received very similar treatment overall, differing only in the order and timing of treatments. As with any retrospective study for treatment of lymphoma in dogs, biases may exist in this study because of uncontrolled factors, such as factors influencing protocol selection, owner s commitment to treat and financial considerations, animal selection based on the presentation, and inconsistently used reinduction protocols. Therefore, care must be taken not to overstate the findings of the current study. It also is possible that other confounders exist that were not controlled for in our modeling. Nevertheless, our models suggest that dogs treated with COAP were at approximately 2 times higher risk for relapse and death than dogs induced with the UW-19 protocol. Future, randomized prospective studies are warranted to further assess the influence of induction protocol and maintenance chemotherapy on survival in dogs with lymphoma. Footnotes a CD3 T-cell marker, DAKO Corp, Carpinteria, CA b CD 79a and BLA 36 B-cell markers, DAKO Corp, Carpinteria, CA c Prism 4, GraphPad, Inc, San Diego, CA d Stata version 9.1 (StataCorp, College Station, TX References 1. Dorn CR, Taylor DO, Schneider R. The epidemiology of canine leukemia and lymphoma. Bibl Haematol 1970; MacEwen EG, Brown NO, Patnaik AK, et al. Cyclic combination chemotherapy of canine lymphosarcoma. J Am Vet Med Assoc 1981;178: MacEwen EG, Hayes AA, Matus RE, et al. Evaluation of some prognostic factors for advanced multicentric lymphosarcoma in the dog: 147 cases ( ). J Am Vet Med Assoc 1987;190: Keller ET, MacEwen EG, Rosenthal RC, et al. Evaluation of prognostic factors and sequential combination chemotherapy with doxorubicin for canine lymphoma. J Vet Intern Med 1993;7: Myers NC, 3rd, Moore As, Rand WM, et al. Evaluation of a multidrug chemotherapy protocol (ACOPA II) in dogs with lymphoma. J Vet Intern Med 1997;11: Valerius KD, Ogilvie GK, Mallinckrodt CH, et al. Doxorubicin alone or in combination with asparaginase, followed by cyclophosphamide, vincristine, and prednisone for treatment of multicentric lymphoma in dogs: 121 cases ( ). J Am Vet Med Assoc 1997;210: Khanna C, Lund EM, Redic KA, et al. Randomized controlled trial of doxorubicin versus dactinomycin in a multiagent protocol for treatment of dogs with malignant lymphoma. J Am Vet Med Assoc 1998;213: Zemann BI, Moore AS, Rand WM, et al. A combination chemotherapy protocol (VELCAP-L) for dogs with lymphoma. J Vet Intern Med 1998;12: Baskin CR, Couto CG, Wittum TE. Factors influencing first remission and survival in 145 dogs with lymphoma: A retrospective study. J Am Anim Hosp Assoc 2000;36: Boyce KL, Kitchell BE. Treatment of canine lymphoma with COPLA/LVP. J Am Anim Hosp Assoc 2000;36: Chun R, Garrett LD, Vail DM. Evaluation of a high-dose chemotherapy protocol with no maintenance therapy for dogs with lymphoma. J Vet Intern Med 2000;14: Dobson JM, Blackwood LB, McInnes EF, et al. Prognostic variables in canine multicentric lymphosarcoma. J Small Anim Pract 2001;42: Moore AS, Cotter SM, Rand WM, et al. Evaluation of a discontinuous treatment protocol (VELCAP-S) for canine lymphoma. J Vet Intern Med 2001;15: Garrett LD, Thamm DH, Chun R, et al. Evaluation of a 6 month chemotherapy protocol with no maintenance therapy for dogs with lymphoma. J Vet Intern Med 2002;16: MacDonald VS, Thamm DH, Kurzman ID, et al. Does L-asparaginase influence efficacy or toxicosis when added to a standard CHOP protocol for dogs with lymphoma? J Vet Intern Med 2005;19: Frimberger AE, Moore AS, Rassnick KM, et al. A combination chemotherapy protocol with dose intensification and autologous bone marrow transplant (VELCAP-HDC) for canine lymphoma. J Vet Intern Med 2006;20: Siedlecki CT, Kass PH, Jakubiak MJ, et al. Evaluation of an actinomycin-d-containing combination chemotherapy protocol with extended maintenance therapy for canine lymphoma. Can Vet J 2006;47:52 59.

9 Protocols for Canine Lymphoma Simon D, Nolte I, Eberle N, et al. Treatment of dogs with lymphoma using a 12-week, maintenance-free combination chemotherapy protocol. J Vet Intern Med 2006;20: Cotter SM, Goldstein MA. Treatment of lymphoma and leukemia with cyclophosphamide, vincristine, and prednisone, I: Treatment of dog. J Am Anim Hosp Assoc 1983;19: Carter RF, Withrow SJ, et al. Chemotherapy of canine lymphoma with histopathological correlation: Doxorubicin alone compared to COP as first treatment regimen. J Am Anim Hosp Assoc 1987;23: Jeffreys AB, Knapp DW, Carlton WW, Thomas RM, et al. Influence of asparaginase on a combination chemotherapy protocol for canine multicentric lymphoma. J Am Anim Hosp Assoc 2005;41: Greenlee PG, Filipa DA, Quimby FW, et al. Lymphoma in dogs: A morphologic, immunologic, and clinical study. Cancer 1990;66: Stone MS, Goldstein MA, Cotter SM. Comparison of two protocols for induction of remission in dogs with lymphoma. J Am Anim Hosp Assoc 1991;27: Aquino SM, Hamor RE, Valli VE, et al. Progression of an orbital T-cell rich B-cell lymphoma to a B-cell lymphoma in a dog. Vet Pathol 2000;37: Gibson D, Aubert I, Woods JP, et al. Flow cytometric immunophenotype of canine lymph node aspirates. J Vet Intern Med 2004;18: Burnett RC, Vernau W, Modiano JF, et al. Diagnosis of canine lymphoid neoplasia using clonal rearrangements of antigen receptor genes. Vet Pathol 2003;40: Owen LN. TNM classification of tumours in domestic animals, 1st ed Geneva; 1980: Alvarez FJ, Kisseberth WC, Gallant SL, Couto CG. Dexamethasone, melphalan, actinomycin D, cytosine arabinoside (DMAC) protocol for dogs with relapsed lymphoma. Journal of Veterinary Internal Medicine 2006;20: Veterinary co-operative oncology group common terminology criteria for adverse events (VCOG-CTCAE) following chemotherapy or biological antineoplastic therapy in dogs and cats v1.0. Vet Comp Oncol 2004;2: Cleves MA, Gould WW, Gutierrez RG. An Introduction to Survival Analysis Using Stata.Revised ed. College Station, TX: Stata Press; 2004: Vail DM, Kisseberth WC, Obradovich JE, et al. Assessment of potential doubling time (T pot ), argyrophilic nucleolar organizer regions (AgNOR), and proliferating cell nuclear antigen (PCNA) as predictors of therapy response in canine non-hogikin s lymphoma. Exp Hematol 1996;24: Ruslander DA, Gebhard DH, Tompkins MG, et al. Immunophenotypic characterization of canine lymphoproliferative disorders. In Vivo 1997;11: Appelbaum FR, Sale GE, Storb R, et al. Phenotyping of canine lymphoma with monoclonal antibodies directed at cell surface antigens: classification, morpholoty, clinical presentation and response to chemotherapy. Hematol Oncol 1984;2: Teske E, van Heerde P, Rutteman GR, et al. Prognostic factors for treatment of malignant lymphoma in dogs. J Am Vet Med Assoc 1994;205: Vail DM, MacEwen EG, Young KM. Canine lymphoma and lymphoid leukemias. In: Withrow SJ, MacEwen EG, eds. Small Animal Clinical Oncology. 3rd ed. Philadelphia, PA: WB Saunders; 2000: Jagielski D, Lechowski R, Hoffmann-Jagielska M, et al. A retrospective study of the incidence and prognostic factors of multicentric lymphoma in dogs ( ). J Vet Med A Physiol Pathol Clin Med 2002;49: