The study was performed at the Veterinary Oncology and Hematology Center in Norwalk, CT, The Center for Specialized Veterinary Care in Westbury, NY and the Katonah-Bedford Veterinary Center in Bedford Hills, NY. This study was presented at the 25th Veterinary Cancer Society conference, Huntington Beach, CA, 2005.
Background: Dogs with multicentric lymphoma are treated with various cyclophosphamide, doxorubicin, vincristine, and prednisone (CHOP)-based chemotherapy protocols with variable success.
Objectives: To describe the progression-free survival (PFS) time and overall survival time (OST) of dogs with T-cell lymphoma or hypercalcemic lymphoma treated with l-asparaginase and mechlorethamine, vincristine, prednisone, procarbazine (MOPP).
Animals: Fifty dogs with T-cell lymphoma, hypercalcemic lymphoma, or both treated at 3 referral veterinary hospitals.
Methods: Retrospective study. Case were selected based on histologic or cytologic diagnosis of lymphoma; presence of the T-cell phenotype, presence of hypercalcemia or both; and absence of previous chemotherapy. The T-cell phenotype was determined by flow cytometry, immunocytochemistry, immunohistochemistry, or polymerase chain reaction of antigen receptor rearrangement.
Results: The overall response rate was 98% (78% complete response, 20% partial response). The median PFS for the entire study population was 189 days with 25% PFS at 939 days. The median OST for the entire study population was 270 days with 25% surviving 939 days. Twenty percent of the dogs required hospitalization for treatment related complications.
Conclusions and clinical importance: l-Asp/MOPP chemotherapy might result in longer PFS and OST for dogs with multicentric T-cell lymphoma, dogs with hypercalcemic lymphoma or both, than achieved with CHOP.
polymerase chain reaction of antigen receptor rearrangement
progression free survival
Veterinary Comparative Oncology Group
World Health Organization
Lymphoma in dogs is most commonly treated with a combination chemotherapy protocol consisting of cyclophosphamide, doxorubicin, vincristine, and prednisone (CHOP). The response rate of this protocol approaches 90% with median remission and survival times in the range of 6–12 months.1 The T-cell phenotype is a poor prognostic indicator.2–12 Dogs with T-cell lymphoma have significantly decreased response rates, as well as significantly shorter remission and survival times, when compared with dogs with the more common B-cell phenotype.2–4,6,7,11,12 The median first remission duration for dogs with T-cell lymphoma is 1.7–5 months, compared with a median first remission duration for dogs with B-cell lymphoma of 5–11 months when CHOP-based protocols are used.2,4,11 Furthermore, the median survival times for dogs with T-cell lymphoma (4–5.3 months) are less than half of those for dogs with B-cell lymphoma (11–13 months) when CHOP-based protocols are used.2,4,11 The use of a high-dose chemotherapy CHOP protocol, as well as the inclusion of half-body radiation therapy within a CHOP protocol, do not improve response and survival times of dogs with T-cell lymphoma.11,12 The presence of hypercalcemia, which is associated with the T-cell phenotype, is a negative prognostic indicator for dogs with multicentric lymphoma.3,4,13–15
The mechlorethamine, vincristine, prednisone, procarbazine (MOPP) chemotherapy protocol (mechlorethamine, vincristine, procarbazine, and prednisone) is an effective protocol for dogs with lymphoma that failed to respond or became refractory to previous chemotherapy.16 MOPP was originally used in people as a first line chemotherapy protocol for both Hodgkin's and non-Hodgkin's lymphoma.17,18 The protocol induces a complete remission in up to 85% of the humans treated.18,19 The purpose of this retrospective study was to evaluate the use of a non–CHOP-based chemotherapy protocol, l-asparaginase (l-asp)/MOPP, as first line therapy for the treatment of T-cell lymphoma, hypercalcemic lymphoma, or both in dogs.
Material and Methods
This study was retrospective in nature and included 50 dogs with multicentric T-cell lymphoma, hypercalcemic lymphoma, or both, treated at 3 facilities (Veterinary Oncology and Hematology Center, The Center for Specialized Veterinary Care, and the Katonah-Bedford Veterinary Center) between January 2003 and January 2008. Inclusion criteria were (1) histologic or cytologic diagnosis of multicentric lymphoma; (2) presence of the T-cell phenotype, presence of hypercalcemia or both; and (3) absence of previous chemotherapy. Dogs with lymphoma of the gastrointestinal tract or skin were excluded.
Diagnosis and Staging
The diagnosis of lymphoma was based on cytologic or histologic evaluation of lymph node or extranodal lesion (liver, spleen, effusion, bone marrow). Clinical staging for all dogs was based on the World Health Organization (WHO) 5-stage criteria for canine lymphoma.20 In addition, dogs were subcategorized into either substage a (self-supporting) or substage b (non–self-supporting).13 Pretreatment evaluation included CBC, chemistry panel, urinalysis, thoracic radiographs, abdominal ultrasound, bone marrow evaluation, ionized calcium when indicated, and immunophenotype. Immunophenotype was determined by flow cytometry, immunocytochemistry, immunohistochemistry, and polymerase chain reaction of antigen receptor rearrangement (PARR). For flow cytometry, immunocytochemistry, and immunohistochemistry, the presence of the CD3 antigen was considered diagnostic for T-cell lymphoma.14 Unfortunately, complete staging was not performed in all dogs due to owner preference or financial constraints. Therefore, it is possible that some dogs with a higher clinical stage were classified as having a lower clinical stage.
Fifty dogs were treated with the l-aspa/MOPP chemotherapy protocol (Table 1). l-Asp (400IU/kg, maximum of 10,000 U) was given intramuscularly 1 week before the initiation of the MOPP protocol. Procarbazine was delivered to the nearest 10 mg. Protocol alterations or dosage adjustments were made at the discretion of the veterinarian overseeing the dog at that visit.
Table 1. MOPP chemotherapy protocol for treatment of multicentric T-cell lymphoma in dogs.
A CBC and a physical examination were performed before each administration of chemotherapy. Evidence of gastrointestinal toxicosis was obtained by owner description of clinical signs. Hematologic and gastrointestinal toxicoses were graded based on the Veterinary Comparative Oncology Group (VCOG) consensus document.21 Grades III or IV toxicosis was considered as being of moderate to severe severity.
Assessment of Response
Complete response (CR) was elimination of all detectable disease. CR was then subdivided into complete response-1 (CR-1) and complete response-2 (CR-2). CR-1 refers to dogs that achieved a CR and maintained it throughout the entire 28 day MOPP cycle. CR-2 refers to dogs that achieved a CR, relapsed between days 7 and 28, and subsequently responded to the next MOPP cycle.16 Partial response (PR) was considered a >50% reduction but <100% reduction in measurable disease for 1 chemotherapy cycle, stable disease (SD) was considered <50% reduction or no change in size of all measurable disease for 1 chemotherapy cycle and progressive disease (PD) was considered a >25% increase in measurable disease or appearance of new neoplastic lesions. In addition, a dog was also considered to have PD if hypercalcemia recurred despite no change in gross disease in a dog that was previously in CR/PR/SD with normal plasma ionized calcium concentration.
Response rate was defined as the number of dogs that achieved CR or PR compared with the total number of dogs treated. Overall survival time (OST) was defined as the time from initiation of treatment until cancer-related death. Progression-free survival (PFS) was defined as the time from initiation of treatment until evidence of PD (either recurrence of gross disease or return of hypercalcemia) that was unresponsive to MOPP. Dogs were censored from OST or PFS if they were alive at the time of study closure, lost to follow-up or died due to causes unrelated to neoplasia or its treatment.
The PFS and OST curves were generated by using the Kaplan-Meier product limit method and the median PFS and OSTs were calculated. For each of the dogs in the study, the following prognostic factors were evaluated by the Gehan-Breslow test for their influence on PFS and OST: sex; stage; substage; hypercalcemia; cranial mediastinal mass; bone marrow involvement; low-grade lymphoma; hospitalization; protocol alteration; treatment delay; rescue chemotherapy; anemia at the time of diagnosis; prior steroid use; response to therapy (CR [CR-1 versus CR-2] versus PR versus SD); hospitalization and grade III or IV neutropenia. Gehan-Breslow was chosen because it weighs early data points in a survival analysis more heavily than later ones, as the early data points tend to be the most reliable. All statistical analyses were performed by use of Systat 12.e
Dog Characteristics, Diagnosis, and Clinical Staging
Fifty dogs were included in this study. Median age was 7 years (range 1.5–14 years). There were 20 breeds: Boxer (n = 12, 24%), Golden Retriever (n = 7, 14%), mixed breed (n = 6, 12%), and Labrador Retriever (n = 4, 8%). The median weight at the time of diagnosis was 30 kg (range, 5–55 kg). Twenty-eight (56%) of the dogs were male, with 4 (8%) being male intact and 24 (48%) being male neutered. Twenty-two (44%) of the dogs were female, with 2 (4%) being unspayed females and 20 (40%) being spayed females.
Thirty-three (66%) of the dogs were diagnosed by cytologic examination as having lymphoma. Twenty (40%) of the dogs diagnosed by cytology had a microscopic description consistent with lymphoblastic lymphoma, 8 (16%) had intermediate cell lymphoma, and in 5 (10%) dogs the microscopic description was not recorded. However, in 3 of the 5 dogs with unrecorded microscopic descriptions, hypercalcemia was present. Seventeen (34%) dogs were diagnosed with lymphoma, by histopathology. Of these 17 dogs, 11 (22%) had high-grade, 2 (4%) had intermediate grade, and 5 (10%) had low grade. T-cell lymphoma was diagnosed by flow cytometry in 27 (54%) dogs, immunohistochemistry in 12 (24%) dogs, immunocytochemistry in 3 (6%) dogs, and polymerase chain reaction of antigen receptor rearrangement in 1 (2%) dog. Seven (14%) of the dogs were hypercalcemic without an immunophenotype analysis performed.
One dog (2%) was classified as stage I, 1 dog (2%) was classified as stage II, 14 dogs (28%) were classified as stage III, 13 dogs (26%) were classified as stage IV, and 21 dogs (42%) were classified as stage V. Sixteen of the dogs (32%) were substage B, while 34 of the dogs (68%) were substage A. Fifteen dogs (30%) had bone marrow biopsy that was consistent with lymphomatous infiltration and 18 dogs (36%) had no evidence of lymphomatous infiltration. Bone marrow was not collected for evaluation in 17 dogs (34%). Fifteen dogs (30%) had a cranial mediastinal mass, 30 dogs (60%) had no cranial mediastinal mass, and chest radiographs were not done in 5 dogs (10%). Twenty-seven of the dogs (54%) were hypercalcemic at the time of diagnosis, 6 dogs (12%) had steroids administered before inclusion in the study, 7 dogs (14%) were anemic at presentation, and 8 dogs (16%) experienced at least 1 episode of grade III or IV neutropenia.
Response to Treatment, PFS, and OST
Fifty dogs received 267.5 cycles of MOPP chemotherapy. The median number of cycles was 5 (range, 1–12.5). CR was achieved in 39 of the 50 dogs (78%). Twenty-eight dogs (56%) had a CR-1 and 11 dogs (22%) had a CR-2. Ten of the 50 dogs (20%) achieved a PR for an overall response rate of 98%. One dog (2%) had SD. Complete response (including CR1 versus CR2), PR or SD was not significantly related to either PFS or OST (Table 2). The median PFS for the entire study population was 189 days (95% confidence interval [CI], 99–278 days) with 25% PFS at 939 days (Fig 1). The median OST for the entire study population was 270 days (95% CI, 206–333 days) with 25% survival at 939 days (Fig 2). Twenty-one dogs (42%) received rescue chemotherapy. The most commonly used protocols were a CHOP-based protocol and single agent lomustine.f Three dogs died due to causes other than lymphoma while responding to MOPP or to a rescue chemotherapy protocol at 166, 468, and 933 days. Eleven dogs were alive and responding to MOPP at a median of 202 days. Four dogs were lost to follow-up at 71, 151, 309, and 379 days. None of the prognostic variables evaluated had a statistically significant impact either on PFS or OST (Table 2).
Table 2. Univariate analysis of potential prognostic factors for progression free survival (PFS) and survival (OST) in dogs with lymphoma.
CR, complete response; PR, partial response; SD, stable disease; n, number of cases; NR, not reached; OST, overall survival time.
Yes (n = 27)
No (n = 23)
Cranial mediastinal mass
Yes (n = 15)
No (n = 30)
A (n = 34)
B (n = 16)
Bone marrow involvement
Yes (n = 15)
No (n = 18)
CR (n = 39)
PR (n = 10)
SD (n = 1)
Complete response (subcategories)
CR-1 (n = 28)
CR-2 (n = 11)
Yes (n = 10)
No (n = 40)
Yes (n = 35)
No (n = 15)
Yes (n =19)
No (n = 31)
Yes (n = 21)
No (n = 29)
Yes (n = 7)
No (n = 43)
Grade III/IV neutropenia
Yes (n = 8)
No (n = 42)
Seventy-six percent (38/50) of the dogs had at least 1 episode of hematologic toxicosis. Thirty-eight percent (19/50) experienced at least 1 episode of neutropenia and 60% (30/50) experienced at least 1 episode of thrombocytopenia. However, the majority of the hematologic toxicoses were mild and clinically insignificant. Of the 29 episodes of neutropenia, 38% (11/29) were grade III (500–999/μL) or IV (< 500/μL), and of the 110 episodes of thrombocytopenia 18% (20/110) were grade III (25,000–49,000/μL) or IV (<25,000/μL). Three dogs in the study were hospitalized because of fever and signs consistent with sepsis. Two of the dogs had grade III neutropenia and 1 dog had grade IV neutropenia. All dogs recovered with supportive care consisting of IV fluids and antibiotics. Two dogs were hospitalized due to anemia (hematocrits 16.3 and 20.3%) necessitating blood transfusion. One of these dogs that had received 7 cycles of MOPP and was in clinical remission had concurrent grade IV thrombocytopenia and subsequently died. A bone marrow evaluation performed 2 days before death did not detect lymphoma. This dog was the only one to die because of adverse effects of chemotherapy.
Gastrointestinal toxicosis was a common finding in this study, with 78% (39/50) of the dogs having at least 1 episode. The majority of the episodes of gastrointestinal toxicosis were mild and self-limiting. Of the 47 episodes of vomiting only 4 (8.5%) were grade III or IV. Furthermore, of the 46 episodes of diarrhea, only 1 (2%) was grade III or IV. Finally, of the 31 episodes of anorexia, only 1 (3.2%) was grade III or IV. There were 4 incidents of hospitalization due to gastrointestinal toxicosis. All resolved with IV administration of fluids and supportive care. In addition, 1 dog was hospitalized for IV administration of fluids and supportive care for 2 days because of an episode of pancreatitis occurring after administration of mustargen, vincristine, prednisone, and procarbazine. Another dog was hospitalized for IV administration of fluids due to severe polyuria and polydypsia resulting in dehydration. The extreme polyuria and polydypsia resolved with a dramatic reduction in dose of prednisone. In total, there were 11 episodes of hospitalization that occurred in 10 of the 50 dogs (20%).
There were 26 treatment delays in 19 of the 50 dogs treated (38%). Fourteen of the treatment delays were due to neutropenia, 5 due to gastrointestinal toxicosis, 2 due to neutropenia and gastrointestinal toxicosis, and 1 each due to polyuria and polydypsia (prednisone intoxication), perivascular slough secondary to either mustargen or vincristine, cervical abscess, staphylococcus aureus skin infection, lymph node abscess, and anemia. Treatment delay did not have a statistically significant effect on either the median PFS or median OST.
Thirty-five of the 50 dogs (70%) had a total of 59 protocol alterations. The protocol alterations were separated into 2 categories: single—alterations that were not continued throughout the remainder of the protocol; and chronic—alterations that were continued throughout the remainder of the protocol. There were 36 incidents of single alterations and 23 incidents of chronic alterations. Of the 36 single alterations, the most commonly performed changes included 25 alterations in which vincristine was not given with mechlorethamine and 5 alterations in which chlorambucilg was substituted for procarbazine due to unavailability of procarbazine. Of the 23 chronic alterations, the most commonly performed changes included 11 alterations where the mustargen and vincristine were moved from days 7 to 14, 4 alterations in which the 14-day administration of the prednisone and procarbazine was moved to another 14-day period within the 28-day cycle, 3 dosage reductions (25% dose reduction of vincristine; 15% dose reduction of both mechlorethamine and vincristine; 10% dose reduction of BCNUh which was substituted for mechlorethamine in 1 dog for 6 of the 12 MOPP chemotherapy cycles due to unavailability of mechlorethamine) and 2 deletions of vincristine from the protocol. The presence of a protocol alteration did not result in a statistically significant difference in either median PFS or median OST. The majority of single alterations and the majority of chronic alterations were due to hematologic toxicosis.
The goal of this study was to evaluate the use of a non–CHOP-based chemotherapy protocol (l-asp/MOPP) as first line therapy for the treatment of T-cell lymphoma, hypercalcemic lymphoma or both in dogs. When the median PFS and OST from the current study are compared with the median remission and survival times from the historical controls it appears that use of MOPP chemotherapy might have resulted in a benefit in both PFS and OST. There are limitations to using historical controls when making comparisons between PFS times and OSTs. One means of comparing PFS and OST to historical controls is with the use of CIs.12 In 3 previous studies evaluating the use of CHOP-based protocols, dogs with T-cell lymphoma had median durations of remission of 52, 96, and 150 days.2,4 The 95% CI for PFS in the current study (99–278 days) exceeds the median remission duration in 2 of the 3 previous studies. The 3rd study had a median remission duration of 150 days, which falls within the 95% CI in the current study, but is still lower than the median PFS of 189 days.11 In that previous study, however, a high dose CHOP chemotherapy was used, which might have increased the median remission duration in comparison with other CHOP-based protocols. Unfortunately, there was a significant increase in frequency of adverse effects of chemotherapy as compared with CHOP protocols using conventional chemotherapy doses.11 In the high dose chemotherapy study, 27% of the dogs died due to chemotherapy-related adverse events, as compared with only 2% in the current study.11 This suggests that the PFS for dogs with T-cell lymphoma treated with MOPP exceeds that of dogs with T-cell lymphoma treated with standard CHOP-based protocols. In 3 previous studies evaluating CHOP-based protocols the median survival times were 120, 153, and 159 days for dogs with T-cell lymphoma.2,4,11 All 3 of these median survival times are outside of the 95% CI range (206–333 days) identified in this study for OST, suggesting that MOPP chemotherapy increases survival time of dogs with T-cell lymphoma when compared with dogs with T-cell lymphoma that receive CHOP protocols. An important fact to note in the current study was the long tail in PFS and OST identified on final analysis, with 25% of the dogs having PFS and OST at 2.6 years. This compares favorably to dogs (regardless of immunophenotype) with multicentric lymphoma that receive CHOP-based protocols, in which 25% were alive approximately 2 years after treatment.22
Two prognostic factors, the presence of hypercalcemia and the presence of a cranial mediastinal mass, have been associated with the T-cell phenotype and with a poor prognosis, with survival times of 3.7 and 3 months, respectively. 2–4,7,13–15,23, 24 In the current study, the presence of a cranial mediastinal mass was not associated with a decrease in either PFS or OST. The median PFS of dogs with a cranial mediastinal mass was not reached at 260 days and the median OST was 387 days. Furthermore, the median PFS and median OST for dogs with hypercalcemia in this study were 189 and 270 days respectively and were not significantly different from the median PFS and median OST of dogs without hypercalcemia. These results suggest that within our population of dogs with T-cell lymphoma, the presence of a cranial mediastinal mass, hypercalcemia, or both does not result in shorter PFS or OST when MOPP chemotherapy is used.
Along with the T-cell phenotype, substage has been shown to be a strong negative prognostic factor for dogs with lymphoma2,3,7,12,24 Median remission times and survival times for dogs presenting with substage B lymphoma range from 37 to 165 days 44 to 137 days, respectively.2,12,26 However, in the current study substage was not predictive of PFS or OST. The median PFS and OST for dogs presenting substage B were 173 and 252 days respectively. This was not statistically different than dogs presenting in substage A.
One possible reason for the success of l-asp/MOPP for the treatment of multicentric canine T-cell lymphoma is because it is an alkylating agent-rich protocol. MOPP contains mechlorethamine and procarbazine, 2 alkylating agents. A previous study which evaluated chromosome aberrations in lymphoma in dogs found the T-cell lymphomas had a considerably higher number of chromosomal aberrations when compared with B-cell lymphomas.25 This genetic instability could allow T-cell lymphomas to develop methods to evade the cytotoxic effects of chemotherapy agents more easily than their B-cell counterparts.26 Cross resistance is uncommon with alkylating agents, and alkylating agents are not affected by p-glycoprotein (MDR phenotype).27,28 Therefore, if the lymphoma develops resistance to one of the alkylating agents in the MOPP protocol due to inherent genetic instability, the remaining alkylating agent might still be active, maintaining the cytotoxicity of the protocol. Previous studies have identified that the inclusion of multiple alkylating agents within chemotherapy protocols to treat canine multicentric lymphoma results in the T-cell phenotype no longer being a negative prognostic factor.29,30
Mild toxicosis was fairly common with MOPP with 76% of the dogs experiencing at least 1 episode of hematologic toxicosis and 78% of the dogs experiencing at least 1 episode of gastrointestinal toxicosis. The most clinically relevant toxicosis was neutropenia, which was considered moderate to severe in 38% of the cases in which it was documented. Only 3 out of the 50 dogs (6%) required hospitalization for sepsis, which is similar to previous reports using MOPP where 4% of the dogs were hospitalized from sepsis.16 Furthermore, neutropenia was the reason for the majority of treatment delays and protocol alterations in this study. The overwhelming majority of GI toxicoses were mild and resolved with supportive care.
The occurrence of treatment delays secondary to neutropenia was reduced by a variety of protocol alterations. An important point is that both treatment delay and protocol alteration did not statistically influence survival time in this study. A possible explanation for the lack of influence on PFS and OST is that many of the protocol alterations did not affect dose intensity.31
There are some limitations to this study. There was not a contemporaneous control population of dogs with T-cell lymphoma receiving a CHOP-based protocol to compare with the study population. It was decided not to use a control population as it was felt that the MOPP chemotherapy protocol resulted in better responses and longer PFS and OST and it would be unethical to use a protocol that was unsuccessful historically. Another limitation to this study was that not all dogs had a lymph node biopsy. Thirty-four percent of the dogs had a lymph node biopsy. It is possible that some of the dogs that were diagnosed as lymphoma via cytologic evaluation would have been diagnosed as having a low-grade lymphoma via histopathology. However, none of the dogs diagnosed via cytology had microscopic descriptions consistent with a small-cell or low-grade lymphoma. There were 5 dogs in this study diagnosed as having a low-grade T-cell lymphoma via histopathology. One of these dogs was hypercalcemic and another dog had an anterior mediastinal mass and was in substage B, conditions believed not associated with true low-grade/indolent T-cell lymphomas.32 A 3rd dog had a biopsy diagnosis of peripheral nodal T-cell lymphoma of intermediate size and low-grade, which may behave differently than the T-zone indolent lymphoma reported in dogs to have prolonged survival times.33 The 2 remaining dogs had a histopathologic diagnosis consistent with low-grade lymphoma. However, 4 out of the 5 dogs with low-grade T-cell lymphoma were censored from PFS and OST analysis, and the final dog had a PFS time of 172 days and an OST of 215 days. Only 81% of the dogs in this study had confirmation of the presence of the CD3 antigen or T cell receptor. Fourteen percent of the dogs were presumed to have T-cell lymphoma based on the presence of hypercalcemia despite not having an immunophenotypic analysis performed. The presence of hypercalcemia in dogs with multicentric lymphoma is consistent with the T-cell phenotype.3,4,7,14 Even if some of these dogs were hypercalcemic and were not of a T-cell phenotype, although this would be extremely rare as suggested by the lack of hypercalcemic B-cell lymphomas reported in the literature, the presence of hypercalcemia in canine lymphoma is a negative prognostic factor and MOPP chemotherapy was able to improve PFS and OST over those reported with CHOP protocols previously.3,4,13–15
Dr VE Valli, DVM, DACVP; VDx Veterinary Pathology Services.