All work was performed at the Matthew J. Ryan Veterinary Hospital of the University of Pennsylvania, 3900 Delancey St, Philadelphia PA 19104.
Prospective Clinical Trial to Compare Vincristine and Vinblastine in a COP-Based Protocol for Lymphoma in Cats
Version of Record online: 17 NOV 2012
Copyright © 2012 by the American College of Veterinary Internal Medicine
Journal of Veterinary Internal Medicine
Volume 27, Issue 1, pages 134–140, January/February 2013
How to Cite
Krick, E.L., Cohen, R.B., Gregor, T.P., Salah (Griessmayr), P.C. and Sorenmo, K.U. (2013), Prospective Clinical Trial to Compare Vincristine and Vinblastine in a COP-Based Protocol for Lymphoma in Cats. Journal of Veterinary Internal Medicine, 27: 134–140. doi: 10.1111/jvim.12006
This research was presented as an oral abstract presentation at the 30th annual Veterinary Cancer Society Conference, San Diego, CA, October 2010.
- Issue online: 11 JAN 2013
- Version of Record online: 17 NOV 2012
- Manuscript Accepted: 25 SEP 2012
- Manuscript Revised: 20 AUG 2012
- Manuscript Received: 21 MAY 2012
- Clinical Oncology Research Fund of the Veterinary Hospital of the University of Pennsylvania
- a University of Pennsylvania School of Veterinary Medicine Department of Clinical Studies Research Grant
- Vinca alkaloid
Current standard chemotherapy protocols for lymphoma in cats carry risks of gastrointestinal toxicity, which can decrease quality of life and complicate response assessment. Protocols with less gastrointestinal toxicity may improve treatment tolerance.
The study purpose was to compare response rate, outcome, and toxicity between cats that received vincristine or vinblastine as part of combination chemotherapy for lymphoma. We hypothesized that vinblastine would have similar efficacy, but less gastrointestinal toxicity, compared with vincristine.
Forty client-owned cats with confirmed diagnosis of lymphoma.
Cats were randomized to 1 of 2 treatment arms and received weekly COP-based chemotherapy for 6 months or until disease progression. Response rate, progression-free survival (PFS), lymphoma-specific survival (LSS), and incidence and severity of gastrointestinal and hematologic toxicity were compared between arms. Arm cross-over occurred if specific gastrointestinal toxicity criteria were noted.
Cats in both arms had similar response rates, PFS, and LSS (48 versus 64 days, P = .87; 139 versus 136 days, P = .96). Cats that received vincristine were significantly more likely to switch arms based on gastrointestinal toxicity than cats that received vinblastine (44.4 versus 10.5%, P = .02). Lower baseline weight was significantly negatively associated with PFS and LSS (P = .01, P = .003, respectively). Baseline anemia was significantly negatively associated with LSS (P = .04).
Conclusions and Clinical Importance
Results suggest that vinblastine is a reasonable alternative to vincristine in the treatment of some cats with lymphoma. Baseline body weight remains a significant prognostic factor for cats with lymphoma.
cyclophosphamide, doxorubicin, vincristine, prednisone
cyclophosphamide, vincristine, prednisone
Lymphoma is one of the most commonly diagnosed and treated cancers in cats. The most commonly reported chemotherapy protocols used to treat large cell lymphoma in cats consist of a combination of vincristine, cyclophosphamide, and prednisone (COP-based), or these drugs and doxorubicin (CHOP-based). Both COP- and CHOP-based protocols are effective, with varying response rates and survival times. Complete response rates reported for COP protocols range from 33 to 75%, with similar complete response rates of 38–80% for CHOP protocols.[2-8] Remission duration and survival times with CHOP protocols also are similar to COP protocols (5–21 and 2–10 months versus 3–8 and 2–9 months, respectively).[2-8]
The most commonly reported adverse effects of these chemotherapy protocols in cats are gastrointestinal signs and myelosuppression.[2, 3, 5, 6, 8, 9] Vincristine was the drug most commonly associated with chemotherapy toxicity in 3 studies that reported on specific drug-associated toxicity.[3, 8, 9] Vincristine-related gastrointestinal toxicity occurred in 11 of 38 cats during a COP induction protocol. In addition, vincristine was associated with 47% of the episodes of neutropenia and 56% of the episodes of gastrointestinal toxicity in 23 cats receiving a CHOP-based protocol. In another study, the dose of vincristine was reduced in 20 of 61 cats receiving a CHOP-based protocol. Vincristine was the most commonly dose reduced drug, and 19 of these dose reductions were because of gastrointestinal toxicity.
The majority of cats with lymphoma are ill at the time of diagnosis, often with gastrointestinal signs, as this is the most common anatomic location of lymphoma in cats.[1, 10] Low body weight, body condition score, and muscle mass have been reported in cats with lymphoma, and low body condition score at diagnosis is a negative prognostic factor. The gastrointestinal tract is also commonly affected by chemotherapy, and cats that are already experiencing gastrointestinal signs may have poor tolerance of chemotherapy. If these cats experience continued weight loss or decreased appetite, it can be difficult to know if those signs represent lymphoma progression or intolerance of chemotherapy. Weight loss during the 1st month of chemotherapy recently has been associated with shorter survival in cats with large cell lymphoma.
Given the poor body condition of many cats with lymphoma, the prognostic impact of weight loss during treatment, and the relatively high risk of gastrointestinal toxicity with vincristine, it is reasonable to consider alternate drugs to decrease this risk. Vinblastine is another vinca alkaloid that is associated with less gastrointestinal and neurotoxicity but more myelosuppression in humans. Several case reports describe the use of vinblastine in cats, but side-by-side or randomized comparison of vincristine and vinblastine has not been performed.[14-17]
The purpose of our study was to compare the response rate, progression-free survival (PFS), lymphoma-specific survival (LSS), and gastrointestinal and hematopoietic toxicity between cats that received vincristine as compared with vinblastine as part of a COP-based chemotherapy protocol. We hypothesized that cats in both arms would have similar response rates and outcome and that cats that received vinblastine would have less gastrointestinal toxicity compared with cats that received vincristine.
Client-owned cats with a cytological or histopathological diagnosis of intermediate to large or small cell LSA in any anatomic location with a life expectancy > 1 month with treatment were eligible for enrollment in this prospective, randomized, single-blinded study. Cats with intermediate or large cell LSA that had received prior chemotherapy (except cats that received L-asparaginase and prednisone ≤1 week before enrollment) were excluded. Prior treatment with prednisone and chlorambucil was allowed for cats with small cell lymphoma, provided that those cats had progressed on that protocol and had not received additional chemotherapy. Cats were excluded if they had undergone surgical resection of all lymphoma lesions, had received radiation therapy or both, or if radiation therapy was planned in addition to chemotherapy. This protocol was approved by the University of Pennsylvania's Institutional Animal Care and Use Committee, and written informed consent from owners was required for enrollment.
Complete staging (including physical examination, complete blood count, serum biochemistry panel, serum cobalamin and folate concentrations, urinalysis, thoracic radiographs, and abdominal ultrasound examination, plus or minus bone marrow cytology) was intended to be performed on all cats. Cats were stratified by lymphoma cell type (intermediate to large or small cell) then were randomized to 1 of 2 treatment arms (ArmVCR=vincristine, ArmVBL=vinblastine) in blocks of 12 using a computerized random numbers generator program. Owners were blinded to their cats' treatment arm.
All cats received a standardized weekly sequential combination chemotherapy protocol for the 6-month study period (Table 1). At the conclusion of the study period, owners were unblinded as to which drug their cat had received and had the option of continuing the chemotherapy protocol for a 10- to 12-month maintenance period. Owners of cats that had progressive lymphoma during the study period were offered the option of continuing treatment with a rescue chemotherapy protocol. The choice of rescue chemotherapy protocol was at the discretion of the attending clinician.
|L-asparaginase 400 IU/kg SC||•|
Vincristine 0.5 mg/m2 IV (ArmVCR)
Vinblastine 1.5 mg/m2 IV (ArmVBL)
|Cyclophosphamide 25 mg PO for 2 doses||•|
|Methotrexate 2.5 mg PO||•|
|Prednisone 1–2 mg/kg/day PO||•||•||•||•||•|
Cats that were found to be cobalamin-deficient were supplemented with 250 μg of cobalamin SC once a week for 4 weeks, once every 2 weeks for 8 weeks, then monthly as needed based on a previously published protocol. Cobalamin and folate concentrations were planned to be re-evaluated after 3 months of supplementation.
After each chemotherapy treatment, owners received a daily log on which to record the frequency and severity of any toxicity experienced by their cat. This log was reviewed at the subsequent visit to the oncology service. Toxicity that was reported ≤7 days postdrug administration was attributed to that drug. Toxicity was graded according to criteria established by the Veterinary Comparative Oncology Group (VCOG). All cats were prescribed antiemetics (metoclopramide or ondansetron, depending on drug cost) for owners to administer at home according to standard instructions to dose cats on a daily basis. Standard dosages were used for both medications. Cats that experienced clinically relevant gastrointestinal toxicity within 7 days after receiving a vinca alkaloid without signs of lymphoma progression were crossed over to the opposite treatment arm. Gastrointestinal toxicity that warranted treatment arm cross-over included >grade 1 vomiting/diarrhea toxicity, >grade 2 anorexia, or both, including decreased appetite for >2 days resulting in substantial weight loss (≥10% of body weight), anorexia for >1 day, vomiting >3 times in 1 day or 1–2 times per day for >3 days, and diarrhea for >2 days. Additional criteria for cross-over were determined by the attending clinician and could include severe adverse effects not listed above determined to be caused by drug administration, such as constipation. Because abdominal ultrasound examination or thoracic radiographs are sometimes necessary to determine if such clinical signs are because of drug toxicity or progression of lymphoma, these tests were performed at the discretion of the attending clinician. Cats that experienced myelosuppression without gastrointestinal toxicity were intended to remain in their original treatment arm, but drug dose decreases were allowed. Dose reductions of 10–15% of the original dose were intended for patients with asymptomatic neutropenia depending on the toxicity grade. If neutropenia persisted despite a total dose reduction of >30%, cats could be crossed over to the opposite treatment arm.
Response rate at 6 weeks and 6 months of treatment (based on restaging tests appropriate to the original location of the cat's lymphoma), PFS, LSS, and the number of cats in each arm that crossed over to the other arm because of gastrointestinal toxicity were the primary study endpoints. Cats were monitored after the 6-month study endpoint for PFS and LSS data. Cats that did not respond to the protocol or relapsed during the 6-month study period were considered off-study when progressive disease was noted or when their chemotherapy protocol was changed because of lack of response. Response to treatment was be defined as complete response (100% resolution of clinically detectable disease), partial response (≥50 but ≤100% reduction in tumor burden), stable disease (decrease of ≤50% or increase of ≤25% in tumor burden with no development of new lesions), and progressive disease (≥25% increase in tumor burden or development of new lesions). Slightly different criteria were used in cats for which the anatomical location of lymphoma precluded percent measurements of tumor burden changes. In these patients, if partial regression of the lymphoma lesions was noted, and lymphoma-associated clinical signs improved, they were placed in the partial response category. If no regression or progression of lymphoma lesions was noted, and clinical signs were unchanged, they were placed in the stable disease category.
It was estimated that a difference of 30% in PFS between the 2 treatment arms would be considered a clinically relevant difference in efficacy between vinblastine and vincristine. With power = 0.8 and α of 0.05, a minimum of 36 cats would be required in each arm to detect such a difference. If this difference were detected, the study hypothesis that the 2 drugs have equal efficacy would be rejected. Based on prior literature, the probability of gastrointestinal toxicity in the control group (ArmVCR) was estimated to be 0.4,[3, 8, 9] and a 75% reduction in that risk (probability of 0.1 in the experimental group [ArmVBL]) was considered clinically relevant. Therefore, a minimum of 38 cats would be required in each arm to detect a difference in the proportion of cats that changed treatment groups because of gastrointestinal toxicity with power = 0.8 and α of 0.05.
Chi-square and Fisher's exact test were used to determine if the treatment arms differed by the number of responders (partial or complete) at the 1 and 6 month time points and the number of patients that crossed over treatment arms because of gastrointestinal and hematologic toxicity.
Progression-free survival (PFS) was defined as the time from treatment initiation to tumor progression (physical examination findings, diagnostic tests, or both consistent with progressive lymphoma or change in chemotherapy protocol because of lack of response) or death from any cause. Lymphoma-specific survival (LSS) was defined as time from treatment initiation to death because of lymphoma. Cats were censored from PFS analysis if they were alive with no evidence of relapse or lost to follow-up before relapse. Cats were censored from LSS analysis if they were alive, lost to follow-up, or died of causes other than lymphoma. Kaplan-Meier survival curves were used to generate median PFS and LSS times, and the log rank test was used to compare the 2 treatment arms. A second analysis using Kaplan-Meier product limit method and log rank test was used to compare cats according to which vinca alkaloid(s) they received in the protocol. The 3 arms for that analysis were the following: cats in ArmVCR that did not cross over to ArmVBL; cats in ArmVBL that did not cross over to ArmVCR; and all cats in ArmVCR or ArmVBL that did cross over to the opposite treatment arm. Kaplan-Meier product limit method and log rank test were used to compare PFS and LSS between cats with baseline body weight above versus below the median baseline weight, low versus normal baseline serum cobalamin concentration, and low versus normal baseline hematocrit. P-values <.05 were considered statistically significant. All analyses were performed using commercial software.1 2
Forty cats were enrolled in the study, and signalment, staging, and tumor information are listed in Table 2. Most cats had large cell LSA, and the gastrointestinal (GI) tract was the most common location. There were 18 cats assigned to ArmVCR (vincristine), and 22 cats were assigned to ArmVBL (vinblastine). One cat (in ArmVBL) had progressive disease at week 2 of the protocol and was taken off the study and changed to another chemotherapy protocol before receiving a vinca alkaloid; the remaining 39 cats received at least 1 dose of their assigned vinca alkaloid. The 2 treatment arms did not differ significantly according to baseline body weight, anatomic location of lymphoma (GI versus non-GI), baseline serum cobalamin concentration (low versus normal), cell type, or baseline anemia. In addition, the incidence of baseline clinical signs of diarrhea, vomiting, decreased appetite, and weight loss did not significantly differ between the 2 treatment arms. Ten cats were cobalamin-deficient at baseline (5 in ArmVCR and 5 in ArmVBL). Half (5/10) of the cats that were cobalamin-deficient had GI lymphoma, whereas 29% of the cats with GI lymphoma had cobalamin deficiency. The lymphoma locations of the remaining cobalamin-deficient cats were peripheral lymph nodes (n = 2), nasopharyngeal LSA (n = 1), mesenteric lymph nodes (n = 1), and liver and spleen (n = 1). The small intestine of the cat with liver and spleen lymphoma was thickened, but no evidence of GI lymphoma was noted in any other cat based on abdominal ultrasound examination. The majority (n = 8) of the cats that were cobalamin-deficient had large cell lymphoma, as had the majority of cats in the study population. None of the cats with small cell lymphoma were cobalamin-deficient.
|VCR (n = 18)||VBL (n = 22)|
|Age (years; mean ± SD)||11.35 ± 3.76||12.57 ± 4.01|
|Baseline weight (kg; mean ± sd)||4.36 ± 0.83||3.94 ± 1.3|
|LSA cell type|
|Abdominal non-GI (liver, spleen, kidney, etc.)||6||33.3||4||18.2|
|Thoracic (mediastinum, pleural effusion)||0||0||1||4.5|
|Peripheral LNs (+/- systemic disease)||2||11.1||5||22.7|
|Extranodal (nasal, extraocular, etc.)||2||11.1||2||9.1|
|Bone marrow, peripheral lymphoblasts only, or both||0||0||0||0|
|Other (mixed locations/body cavities)||1||5.6||0||0|
|Anemic at baseline|
|Hypocobalaminemia at baseline|
|Clinical signs at baseline|
Cats in ArmVCR were significantly more likely to be crossed over to ArmVBL because of gastrointestinal toxicity (44.4 versus 10.5%, relative risk [RR] 4.9, 95% CI, 1.2–20.2, P = .02). Of the cats that changed treatment groups because of gastrointestinal toxicity (ArmVCR n = 8, ArmVBL n = 2), 4 cats in ArmVCR and 1 cat in ArmVBL had GI lymphoma. None of the cats in ArmVCR that crossed over to ArmVBL because of GI toxicity had further dose reductions because of GI toxicity. One cat in ArmVBL that crossed over to ArmVCR had 1 dose reduction because of GI toxicity.
Although dose reductions because of neutropenia occurred after both vincristine and vinblastine, the number of cats that had such dose reductions was not statistically different between the 2 drugs (10.5 versus 26%, P = .44). After treatment with vincristine, 2 cats became neutropenic. There were 2 episodes each of grade 1 and grade 2 neutropenia. After vinblastine, 6 cats became neutropenic. There were 2 episodes of grade 1, 5 episodes of grade 2, 2 episodes of grade 3, and 3 episodes of grade 4 neutropenia. No cat died of chemotherapy toxicity, and no incidents of neutropenic sepsis occurred.
Response and Survival
The complete response rates at 6 weeks and 6 months into treatment were similar between treatment arms (ArmVCR, 11 and 11%; ArmVBL, 18 and 18%). The overall response rates at 6 weeks and 6 months into treatment (complete and partial response rates combined) also were similar between treatment arms (ArmVCR, 56 and 17%; ArmVBL, 50 and 18%; P = .76, .37).
At the time of analysis, 3 cats were alive, 2 cats were lost to follow-up, and 4 died of lymphoma-unrelated or unknown causes (undiagnosed pulmonary disease despite response to chemotherapy, congestive heart failure, died at home with no lymphoma progression, and uncontrolled diabetes mellitus, electrolyte abnormalities, and anemia despite response to chemotherapy). Of the cats that were alive, 2 had no progression of lymphoma, and 1 cat had relapsed once (after completing the study protocol), but was currently in remission after receiving rescue chemotherapy. No significant difference was found in PFS (ArmVCR, median [95% CI] 48 [24, 72] versus ArmVBL, 64 [26, 102] days; P = .87) or LSS (ArmVCR, 139 [77, 201] versus ArmVBL, 136 [1, 271] days; P = .96) between treatment arms (Fig 1). When the cats were grouped according to whether they crossed over to the other treatment arm (ArmVCR [n = 10], ArmVBL [n = 19], and crossed over between arms [n = 11]), PFS (37 [14, 60] and 39 [2, 76] versus 183 [84, 282] days) and LSS (139 [16, 262] and 125 [115, 134] versus 205 [59, 351] days) were longer in cats that crossed over treatment arms, but not significantly.
Median PFS and LSS did not differ significantly according to anatomic location of lymphoma (GI versus non-GI) and whether or not cats were cobalamin-deficient at the start of treatment. Only 1 cat with low baseline cobalamin concentration was available for cobalamin reassessment after 3 months of supplementation, and that cat's serum cobalamin concentration was normal. Baseline body weight was significantly associated with PFS and LSS. Cats with baseline weight of <4.2 kg had median (95% CI) PFS and LSS of 37 (31, 43) and 98 (32, 164) days, whereas cats with baseline weight of ≥4.2 kg had median PFS and LSS of 112 (43, 181) and 284 (129, 439) days (P = .011 and .003, respectively; Fig 2). Cats that were anemic at baseline had similar PFS compared to cats that were not (48 [6, 90] versus 70 [0, 150] days; P = .4); but the LSS between cats that were or were not anemic at baseline was significantly different (125 [104, 146] versus 277 [174, 380] days; P = .04).
The results of this prospective randomized clinical trial indicate that vinblastine and vincristine have similar efficacy when used as part of a COP-based chemotherapy protocol, but that vinblastine is associated with less GI toxicity compared with vincristine. This study is the 2nd published randomized study reporting on chemotherapy for cats with lymphoma, and as such provides a contemporaneous comparison of 2 different chemotherapy protocols in a population of cats with lymphoma. Protocols, response rates and outcome with chemotherapy for cats with lymphoma have not changed substantially over the past 2 decades, and our study is the first to describe the use of vinblastine for lymphoma in a cohort of cats.
Body weight was the only variable significantly associated with both PFS and LSS in this study. Body condition score, which previously has been associated with survival in cats with lymphoma, was not assessed in the current study. Given the internal location of lymphoma in most cats, it is reasonable for body weight to be a marker of severity of disease. In fact, weight loss before treatment has been reported as a negative prognostic factor for cats with lymphoma. Body weight also may be a marker for response to treatment and tolerance of chemotherapy during treatment, because weight loss early during treatment is associated with shorter survival time in cats with large cell lymphoma. In addition, the majority of owners of cats with lymphoma that were treated with chemotherapy and who completed a survey inquiring about the experience noted that appetite is an indicator of their cats' quality of life. The use of chemotherapy drugs that are less likely to cause substantial GI toxicity, such as vinblastine, may help decrease the risk of chemotherapy-associated weight loss during treatment, particularly in cats that are responding to treatment. Anemia at baseline was significantly negatively associated with LSS, which has been inconsistently reported previously for cats with lymphoma.[4, 8, 9, 23]
Fewer cats were enrolled in this study than originally planned based on our power analysis. After 40 cats were enrolled and treated, interim analysis showed that vinblastine carried a significantly lower risk of GI toxicity compared with vincristine. Additional enrollment was discontinued because the null hypothesis that GI toxicity risk between the 2 drugs was equal was found to be incorrect. It is possible that with a larger number of cats, a difference in response rate, PFS, or LSS could have been determined, but given the similarity in outcome between treatment arms 1 and 2 that does not seem likely. A larger difference was detected when outcome was compared among cats that stayed in ArmVCR or ArmVBL and cats that crossed over to the opposite treatment arm, and it is possible that a larger sample size could have better informed this potentially clinically important difference.
The prevalence of cobalamin deficiency at diagnosis was lower than what has been reported for cats with small cell lymphoma. Unfortunately, only 1 cat that received cobalamin supplementation was available for follow-up serum cobalamin sampling, so it remains unknown if persistent cobalamin deficiency could be associated with poor response to treatment. Several of the cats that were cobalamin-deficient did not have evidence of GI lymphoma based on clinical signs or abdominal ultrasound findings, and this is similar to what was reported in a study describing the prevalence and impact of cobalamin deficiency in dogs with lymphoma. Measuring serum methylmalonic acid may have been a more accurate method to assess clinically relevant cobalamin deficiency in the current study population.[26-28]
Clinicians were not blinded, which could have allowed some bias in toxicity and response assessment, but the response and toxicity endpoints were designed to be as objective as possible to minimize bias. It can be more challenging to quantify GI versus hematologic toxicity, as the variables are more subjective. Therefore, we asked each client to fill out a standardized daily diary to decrease the effect of recall bias and to ensure that questions regarding toxicity were asked in the same way to all clients. We also prospectively set toxicity endpoints based on VCOG criteria so that treatment arm cross-over was standardized. Relatively low grade toxicity cut-offs were used to dictate treatment arm cross-over because those levels of toxicity were considered by the authors to have potential to substantially impact quality of life and owners' willingness to continue treatment. Although all clients were given the same medication instructions, we were not able to assess compliance with medication administration. However, the randomization scheme should have resulted in an equal distribution of owners that did and did not comply with the medication instructions.
The dose of vinblastine used in this study is lower than what has been previously reported in cats and dogs. However, myelosuppression (with mild or no GI toxicity) was noted in several cats, indicating that this dose still has the potential to cause toxicity. In addition, the outcome of cats in the 2 treatment arms was similar, indicating that using this dose of vinblastine did not negatively impact remission duration or survival.
Accurate assessment of response to treatment remains a challenge for cats with lymphoma, because the majority of affected cats have internal disease and locations that make the determination of percentage change in tumor burden difficult. Because the majority of cats with lymphoma are ill at diagnosis, it seems reasonable to include clinical sign improvement or progression as part of response assessment in these patients. In fact, several studies describing small cell lymphoma base response solely on changes in clinical signs.[24, 29, 30]
In conclusion, the results reported here suggest that vinblastine is a reasonable alternative to vincristine in cats receiving COP-based chemotherapy for lymphoma. Response rates and outcome were not significantly different for cats that received vinblastine versus vincristine, and vinblastine was associated with a lower risk of GI toxicity versus vincristine. In addition, baseline body weight and anemia remain prognostic factors for cats with lymphoma.
This study was supported by the Clinical Oncology Research Fund of the Veterinary Hospital of the University of Pennsylvania, and a University of Pennsylvania School of Veterinary Medicine Department of Clinical Studies Research Grant.Conflict of Interest: Authors disclose no conflict of interest.
- 1Feline lymphoma and leukemia. In: Withrow SJ, Vail DM, eds. Withrow and MacEwen's Small Animal Clinical Oncology, 4th ed. St. Louis MO: Saunders Elsevier; 2007:733–756..
- 13Antimicrotubule agents. In: Chabner BA, Longo DL, eds. Cancer Chemotherapy and Biotherapy, 3rd ed. Philadelphia PA: Lippincott Williams and Wilkins; 2001:329–373., .
- 19Veterinary Co-operative Oncology Group. Common terminology criteria for adverse events (VCOG-CTCAE) following chemotherapy or biological antineoplastic therapy in dogs and cats v1.0. A consensus statement from the Veterinary Co-operative Oncology Group. Vet Comp. Oncol 2004;2(4):194–213.
- 29Feline gastrointestinal lymphoma: 67 cases (1988–1996). Eur J Comp Gastroenterol 1999;4:5–11., , , et al.