Outcome and toxicity associated with a dose-intensified, maintenance-free CHOP-based chemotherapy protocol in canine lymphoma: 130 cases

Original Article DOI: 10.1111/j.1476-5829.2010.00222.x

Outcome and toxicity associated witha dose-intensified, maintenance-freeCHOP-based chemotherapy protocolin canine lymphoma: 130 cases

K. Sorenmo, B. Overley∗, E. Krick, T. Ferrara†, A. LaBlanc and F. Shofer‡

School of Veterinary Medicine of the University of Pennsylvania, Philadelphia, PA, USA

AbstractA dose-intensified/dose-dense chemotherapy protocol for canine lymphoma was designed and

implemented at the Veterinary Hospital of the University of Pennsylvania. In this study, we describe

the clinical characteristics, prognostic factors, efficacy and toxicity in 130 dogs treated with this

protocol. The majority of the dogs had advanced stage disease (63.1% stage V) and sub-stage b

(58.5%). The median time to progression (TTP) and lymphoma-specific survival were 219 and

323 days, respectively. These results are similar to previous less dose-intense protocols. Sub-stage

was a significant negative prognostic factor for survival. The incidence of toxicity was high; 53.9 and

45% of the dogs needed dose reductions and treatment delays, respectively. Dogs that required dose

reductions and treatment delays had significantly longer TTP and lymphoma-specific survival times.

These results suggest that dose density is important, but likely relative, and needs to be adjusted

according to the individual patient’s toxicity for optimal outcome.

Keywordschemotherapy, dogs,lymphoma, summationdose density

Introduction

The history of lymphoma therapy in dogs spans

more than three decades: from the early reports

on the sequential basic combination proto-

cols (COP)-based (cyclophosphamide, vincristine,

prednisone) protocols of the late 1970s and

early 1980s, to the more complex CHOP-based

(COP + doxorubicin) protocols with a prolonged

maintenance phase of the 1980s and 1990s, to

the shorter, maintenance-free protocols that have

become the standard in the more recent years.1 – 20

Early in the history of lymphoma therapy, it was

∗Present address: CARES (Center for Animal Referral andEmergency Services), Langhorne, PA, USA

†Present address: Med Infectious Diseases Section, Universityof Pennsylvania Health System, Philadelphia, PA, USA

‡Present address: Department of Emergency Medicine,University of North Carolina, Chapel Hill, NC, USA

recognized that a subset of patients would go into

durable remissions (i.e. be cured); however, despite

a high initial response rate (higher than 80% in most

protocols), the majority of dogs would relapse and

succumb to their lymphoma. Only 20–25% of the

cases would be alive for 2 years or longer.18 The high

response rates to relatively low dose-intense proto-

cols indicate that canine lymphomas are uniquely

chemotherapy sensitive tumours, and also suggest

that further improvements, both in terms of remis-

sion rates and survival times, may be made through

dose intensification. Results from human oncology

trials, however, have been mixed; some studies have

documented clinically significant improvements

in outcome in patients treated with such dose-

intense/dose-dense protocols, whereas others have

resulted in only marginal or no improvements in

remission and survival durations.21 – 26 The incon-

sistent results from dose intensification may in part

be due to the presence of cancer stem cells, a distinct

Correspondence address:Dr Karin SorenmoDepartment of ClinicalStudiesSchool of VeterinaryMedicineUniversity of Pennsylvania3900 Delancey Street,PhiladelphiaPA 19104, USAe-mail: [email protected]

196 © 2010 Blackwell Publishing Ltd

A dose-intense/dense protocol for canine lymphoma 197

population of cells within the tumour responsible

for tumour repopulation. These cancer stem cells

have been found to be more resistant to conven-

tional cytotoxic therapies than non-stem cell cancer

cells.27 Therefore, dose intensification by itself may

not be enough; specific targeting and eradication

of these cancer stem cells through novel therapies

may be necessary to cure more patients.27 – 29

Nevertheless, in theory, dose intensification

makes sense, both from a chemotherapeutic phar-

macokinetic, and a tumour biology point of view:

response to chemotherapy is characterized by a

steep dose–response curve; a modest increase in

dose (i.e. increased dose intensity) should the-

oretically translate into major improvements in

response rates. Most tumours consist of a het-

erogeneous tumour cell population; therefore, a

combination of several effective drugs is needed

to eradicate all sub-clones and affect a cure. The

concept of dose intensity/density incorporates both

of these principles into protocol development and

advocates for a combination of the most effective

agents to be given concurrently and/or sequentially

over a condensed period of time, thus maximizing

cell kill and preventing tumour cell re-population

between treatments. Applying these strategies in

protocol design should in theory improve efficacy

and translate into higher cure rates.

Few dose-intense protocols have been evalu-

ated in canine lymphoma; most of the short,

maintenance-free protocols are just shorter ver-

sions of the longer protocols of the 1990s. However,

two studies evaluating dose-intense protocols with-

out stem cell rescue have been published in canine

lymphoma.8,9 Both of the studies were associated

with a significant increase in toxicity, but did not

document dramatic improvements in progression-

free survival or survival. A dose-intense protocol for

canine lymphoma was developed and implemented

by the oncology service of the University of Penn-

sylvania in 2001. In this study, we report on the

toxicity and efficacy of this dose-intense protocol.

Methods

This work was performed at the M J Ryan Veterinary

Hospital of the University of Pennsylvania,

Philadelphia, Pennsylvania.

Protocol design and dose-intensity calculation

Six chemotherapeutic drugs (L’asparaginase, vin-

cristine, cyclophosphamide, doxorubicin, metho-

trexate and prednisone) were incorporated in the

protocol. The protocol was composed of three

cycles of induction chemotherapy followed by a

consolidation phase consisting of three cycles of

doxorubicin (Table 1). The combination of con-

comitant vincristine (0.5 mg m−2) and cyclophos-

phamide (50 mg m−2 orally for 4 days) in weeks

2 and 3 of the induction provided a higher dose

intensity/density compared with the combination

weekly sequential protocols even though the doses

were reduced from the standard full dose for both

vincristine (0.7 mg m−2) and cyclophosphamide

(75 mg m−2 for 4 days, total 300 mg m−2). This

was followed by a consolidation phase of three

additional back to back cycles of doxorubicin

given every 2 weeks. The summation dose intensity

(SDI) of the protocol was calculated according to

modified version of the relative efficacy method

described by Dr Frei et al.: the estimated effective-

ness [combination of complete response (CR) and

partial response (PR), CR + PR] of each individual

drug when given at the full dose as a single agent

was established based on previous publications or

empirically based on the authors’ experience.30 The

fractional dose intensity (FDI) is derived by the

fraction or percentage of full dose delivered mul-

tiplied by the relative efficacies of the individual

agents. The sum of the individual FDI is the SDI

for the whole protocol. The SDI per week was

calculated by dividing the SDI by the number of

weeks over which the entire protocol was given

(Tables 2–4).30 Single agent activity for many of

the drugs used in this protocol has not been well

established in canine lymphoma, with the excep-

tion of doxorubicin.10,18 – 20,31,32 Most of the other

agents used in this protocol have been used in

combinations with other drugs, and there are

no publications on single agent activity of these

drugs in canine lymphoma. However, publications

on COP-based (cyclophosphomide, vincristine and

prednisone) chemotherapy reflect similar response

rates and duration to that of doxorubicin sin-

gle agent chemotherapy, with remission rates of

70–75% and remission duration of 3–6 months,

© 2010 Blackwell Publishing Ltd, Veterinary and Comparative Oncology, 8, 3, 196–208

198 K. Sorenmo et al.

Table 1. Protocol

Weeks 1 2 3 4 6 7 8 9 11 12 13 14 16 18 20

L’asparaginase • •Vincristine • • • • • •Cyclophosphamide • • • • • •Doxorubicin • • • • • •Methotrexate •Prednisone

→L’asparaginase: 400 IU kg−1 subcutaneously; vincristine: 0.5 mg m−2 intravenously; cyclophosphamide: 50 mg m−2 orally indays 1–4 (total dose 200 mg m−2); doxorubicin: 30 mg m−2 intravenously, 1 mg kg−1 for dogs less than 15 kg; methrotrexate:0.6 mg kg−1 intravenously.Prednisone tapering over 4 weeks: week 1: 2 mg kg−1 daily; week 2: 1 mg kg−1 daily; week 3: 0.5 mg kg−1 daily; week 4:0.5 mg kg−1 daily. Then discontinued.

Table 2. Summation dose intensity (SDI) calculation

Estimated responsefull dose (CR + PR)

Fraction offull dosea

Relativeefficacyb

Fractional doseintensity (FDI)c

No. ofweekly tx

Total FDI,20 weeks

L’asparaginase 0.5 1.0 1.1 1.1 2 2.2

Vincristine 0.5 0.71 1.1 0.78 6 4.68

Cyclophosphamide 0.5 0.67d 1.1 0.74 6 4.44

Doxorubicin 0.7 1.0 1.6 1.6 6 9.6

Methotrexate 0.2 1.0 0.44 0.44 1 0.44

Prednisone 0.3 1.0 0.6 0.6 2e 1.2

SDI, 20 weeks 22.56

SDI/week 1.13

aThe fraction of full dose was calculated using the dose used in combination divided by the dose used when given as singleagent, e.g. for vincristine 0.5/0.7 = 0.71, for cyclophosphamide 50 mg × 4 (200 mg m−2)/75 mg × 4(300 mg m−2)∗ = 0.67.bThe relative efficacy was calculated by dividing the response rate (CR + PR) of the individual agent by the mean efficacy of allagents in the study, the reference standard: (0.5 + 0.5 + 0.5 + 0.7 + 0.2 + 0.3)/6 = 0.45. For example, the relative efficacy ofvincristine and cyclophosphamide was 0.5/0.45 = 1.1 and the relative efficacy of doxorubicin was 0.7/0.45 = 1.6.cThe FDI is the product of the fraction of full dose multiplied by the relative efficacy (i.e. for vincristine 0.71 × 1.1 = 0.78).dBased on 300 mg m−2 full dose.ePrednisone was given at tapering dose over 4 weeks, only the first 2 weeks when dose was 2 mg kg−1 and 1 mg kg−1, respectively,was included and added to the total SDI calculation.

which is similar to the response rates of 59–85%

and response duration of 4.3–6.8 months reported

with doxorubicin alone.18 – 20 For the purpose of

this study, doxorubicin was therefore estimated

to have the highest efficacy of 0.7 (70% response

rate).10,18 – 20,31,32 The other drugs were assigned

a lower efficacy than doxorubicin: L’asparaginase,

vincristine and cyclophosphamide were each esti-

mated to have an efficacy of 0.5 (50%), methotrex-

ate 0.2 (20%) and prednisone 0.3 (30%), respec-

tively. According to these assigned efficacy val-

ues, the FDI of doxorubicin was 1.6, whereas

the FDI for the combination of vincristine and

cyclophosphamide was 1.52 (0.78 + 0.74), respec-

tively. These results are consistent with the previous

reports reflecting similar activity between these

two protocols in canine lymphoma and suggest

that the efficacy values assigned to vincristine

(0.5) and cyclophosphamide (0.5) were appropri-

ate (Table 2). The SDI per week for the protocol

reported here was 1.13. The dose intensity of two

other protocols, The University of Wisconsin Madi-

son 25-week (UWM-25) protocol and a high-dose

chemotherapy protocol, with no maintenance was

calculated for comparison.8,11 The SDI per week

for the UMW-25 protocol was 0.79 and the for the

high-dose chemotherapy protocol was 0.84 when

using the formula stated above.

Clinical data

The medical records of the MJ Ryan Veterinary

Hospital of the University of Pennsylvania were



Table 3. SDI per week UWM-25 week protocol11

CR + PRFraction offull dose Relative efficacy FDI

No. oftreatments Total FDI, 25 weeks

L’asparaginase 0.5 1.0 1 1 1 1

Vincristine 0.5 1.0 1 1 8 8

Cyclophosphamide 0.5 0.83a 1 0.83 4 3.32

Doxorubicin 0.7 1.0 1.4 1.4 4 5.6

Prednisone 0.3 1.0 0.6 0.6 3 1.8

Total SDI 19.72

SDI per week 0.79

aBased on 300 mg m−2 full dose.

Table 4. SDI per week: a high-dose chemotherapy protocol with no maintenance for dogs with lymphoma8

CR + PRFraction offull dose Relative efficacy FDI

No. oftreatments Total FDI

L’asparaginase 0.5 1.0 1 1 1 1

Vincristine 0.5 1.0 1 1 8 8

Cyclophosphamide 0.5 0.83a 1 0.83 4 3.32

Doxorubicin 0.7 1.25 1.4 1.75 4 7

Prednisone 0.3 1.0 0.6 0.6 3 1.8

Total SDI 21.12

SDI/week (25) 0.84

aBased on 300 mg m−2 full dose.

searched for dogs with lymphoma treated with this

particular protocol. All dogs that were identified

through this search and started on this particular

protocol were included in this analysis. Lymphoma

was diagnosed by fine needle aspirate and cyto-

logical evaluation in all dogs. Complete staging,

including blood work (complete blood count),

serum chemistry panel, urinalysis, three-view tho-

racic radiographs, abdominal ultrasound and bone

marrow aspirate, was requested on all cases. In

addition, cardiac examinations including cardiac

echocardiography were performed before the first

and after the fourth doxorubicin treatment. Cases

that had changes in their cardiac auscultation before

the planned re-exam after four doses underwent

additional cardiac examinations as needed. Dox-

orubicin was discontinued in dogs with cardiac

changes suspicious for doxorubicin-induced dam-

age, e.g. clinically significant changes in shortening

fraction or the development of frequent ventricular

arrhythmias or persistent supra-ventricular tachy-

cardia. The protocol was modified in these cases,

and doxorubicin was substituted with a combina-

tion of vincristine and cyclophosphamide (similar

to weeks 2 and 3). All cases that completed the

protocol underwent restaging to ensure complete

remission before discontinuing chemotherapy. The

dogs were requested to return for clinical exam-

inations once monthly to assess remission status

after completion of the chemotherapy protocol.

Relapses were confirmed by fine needle aspirates

and cytological evaluation in addition to clini-

cal examinations. The specific implementation of

the protocol (i.e. when to perform dose reduction

and when to delay treatment) was decided by the

treating clinician. However, our general guidelines

included a treatment delay from 3 to 7 days if

the neutrophil count was less than 2000 cells μL−1,

and dose reductions ranging from 10 to 25% in

cases with toxicity. The degree of dose reduction

depended on the severity of toxicity; cases with

asymptomatic neutropenia would generally receive

a 10–15% dose reduction, whereas cases with severe

toxicity requiring hospitalization would receive a

20–25% dose reduction. Dogs that required dose

reductions continued subsequent treatments at the

lower dose and were rarely escalated back to the

initial dose. The goal of the dose reductions was

to maintain the dose density (i.e. to find the dose

that permitted treatment to be administered weekly



where the protocol called for it) rather than stay

with the current dose (in cases with asymptomatic

neutropenia) and treat at more extended intervals

to allow for bone marrow recovery. The dogs were

routinely discharged with prophylactic oral anti-

emetics (metoclopramide 0.2 mg kg−1 every 8 h or

odansetron 0.3–0.5 mg kg−1 every 24 h depend-

ing on the clinicians’ discretion) to be given after

doxorubicin administration. Anti-emetics were dis-

pensed as needed for the other drugs if the dogs

had prior nausea and vomiting when receiving that

particular treatment.

Information regarding signalment, tumour stage

and sub-stage, immunophenotype, treatment (dose

reductions, treatment delay, completion of proto-

col), response to treatment [response rate, time to

progression (TTP), response to rescue, lymphoma-

specific survival and cause of death] and toxic-

ity (acute and chronic, specifically doxorubicin-

induced cardio-toxicity) were collected from the

medical records and through follow-up phone calls

with owners whose animals’ status was not clear

from the medical records. Response to treatment

was categorized as a CR if the peripheral lymph

nodes palpated within normal range ± supportive

cytology. PR was defined as 50% or more reduction

in size, but clinically still enlarged ± cytology to

confirm residual disease. Stable disease was defined

as less than 50% reduction or less than 20% in-

crease in peripheral lymph nodes. Progressive dis-

ease (PD) was defined as more than 20% increase in

peripheral lymph node size or enlargement of previ-

ously normal lymph nodes. The following variables

were evaluated for prognostic significance for both

TTP and lymphoma-specific survival: body weight

(more or less than 15 kg), stage, sub-stage, presence

of multicentric disease, hypercalcemia and toxicity

(resulting in treatment delay or dose reductions).

Statistical analysis

Continuous data were expressed as means ±standard deviation (SD) and categorical data as

frequencies and percentages. TTP and lymphoma-

specific survival times were calculated from the

date of diagnosis and treatment initiation to the

date of death, PD or relapse. Dogs that were

still alive or died because of other causes than

lymphoma or lymphoma therapy were censored

at the last date they were reported to be alive or

when they died. The Kaplan–Meier product limit

method was used to estimate the proportion of dogs

that were alive/had died because of lymphoma or

had not progressed/had relapsed from lymphoma.

Differences in outcome (TTP and lymphoma-

specific survival) according to prognostic variables

(body weight, more or less than 15 kg, stage, sub-

stage, immunophenotype presence of multicentric

disease, hypercalcemia, dose reduction or treatment

delay) were assessed by the log rank test. Statistical

significance was defined as P < 0.05. All analyses

were performed using SAS® statistical software

Version 9.2 (SAS Institute, Cary, NC, USA).

Results

A total of 130 dogs with lymphoma were included

in this analysis. Information regarding signalment,

stage, sub-stage, immunophenotype and clinical

presentation have been presented in Table 5.

Complete staging data were not available in all

of the dogs. Two of the three dogs with stage I

or stage II disease did not have bone marrow

aspirates performed. Six of the 12 dogs with stage III

disease did not have complete staging, specifically,

thoracic radiographs, abdominal ultrasounds and

bone marrow aspirates were not performed in

two dogs, and bone marrow aspirates were not

performed in the other four dogs. Five of the

33 dogs with stage IV disease did not have bone

marrow aspirates performed. Seventy-one dogs had

information regarding immunophenotyping; of

these 52 dogs were enrolled in a prospective analysis

of various immunological parameters including

lymphoma immunophenotype by flow cytometry,

and the other 19 dogs had immunophenotyping

as part of their regular staging. Forty-three dogs

(60.6%) had B-cell lymphoma, 23 dogs (32.4%)

had T-cell lymphoma and 5 dogs (7%) stained for

both T- and B-cell markers (B/T).

Response to treatment and outcome

In all, 108 dogs (83.1%) had a CR to the protocol,

whereas 15 dogs (11.5%) had a PR, 5 dogs (3.8%)

had stable disease and 2 dogs (1.5%) had PD. Fifty-

six dogs (43.1%) did not complete the protocol,



Table 5. Signalment and stage of disease in 130 dogs withlymphoma

Parameter

Age (years), mean ± SD 7.4 ± 2.7

Body weight (kg), mean ± SD 30.7 ± 13.1

Sex N (%)

MC 58 (44.6)

FS 51 (39.2)

M 17 (13.1)

F 4 (3.1)

Breed

Mix breeds 25 (19.2)

Golden Retrievers 19 (14.6)

Labrador Retrievers 8 (6.1)

Rottweilers 7 (5.4)

Cocker Spaniels 7 (5.4)

Beagles 5 (3.9)

German Shepherds 5 (3.9)

Boxers 4 (3.1)

Other pure breeds 50 (0.77–2.3)

Stage of disease

I 1 (0.77)

II 2 (1.5)

III 12 (9.2)

IV 33 (25.4)

V 82 (63.1)

Sub-stage

a 54 (41.5)

b 76 (58.5)

Immunophenotype

B-cell 43 (60.6)

T-cell 23 (32.4)

B/T-cell 5 (7.0)

Hypercalcemia

Yes 21 (16.2)

No 109 (83.8)

Multicentric disease

Yes 119 (91.5)

No 11 (8.5)

whereas 71 (54.6%) completed the protocol. Three

dogs were lost to follow-up before completion of

protocol. It is unknown whether these dogs com-

pleted the protocol elsewhere. Four of the dogs

that did not complete the protocol were switched

over to a longer, less dose-intensified protocol. The

remaining dogs either failed (n = 22) to achieve

complete remission or relapsed before the protocol

was completed (n = 27). Eighty-nine dogs received

rescue chemotherapy when they failed to respond

to the initial protocol or relapsed after complet-

ing the original protocol; 54 of these dogs were

re-induced with drugs included in the original pro-

tocol given at a weekly schedule similar to weeks 1,

2 and 3 in the induction protocol, specifically vin-

cristine, cyclophosphamide, L’asparaginase, pred-

nisone (L’COP). Forty-one (75.9%) of the dogs

had a CR to this protocol. Other rescue proto-

cols included CCNU, MOPP and a single dose of

L’asparaginase or prednisone alone. At the time of

data analysis, 101 of the dogs had died because

of tumour or treatment-related causes, 10 died

because of other causes, 7 were lost to follow-up and

12 were still alive. The median TTP and lymphoma-

specific survival for all dogs was 219 days (95%

CI: 189–272) and 323 days (95% CI: 243–436),

respectively (Fig. 1).

Toxicity

Of the treated dogs, 69 (53.9%) had dose reductions,

of which 36 dogs had one reduction, 24 had two

reductions, 7 had three reductions and 2 had

four reductions. Fifty-eight (45%) of the dogs had

treatment delays, of which 39 dogs had one delay, 8

had two delays, 7 had three delays, 2 had four delays

and 2 had five delays. Fifty-four percent of dogs

that weighed less than 15 kg had treatment delays

and 58% had dose reductions. Eighty-four percent

of the delays/reductions occurred in the early

induction after a treatment with vincristine and

cyclophosphamide. Fifteen dogs were hospitalized

and treated for sepsis or severe gastrointestinal

toxicity; in addition, another four dogs presented

to the emergency service because of the owners’

concerns over toxicity, but were triaged away

and not admitted. One dog died approximately

7 weeks into induction (1 week after the second

L’asparaginase). The cause of the rapid decline

in clinical status and death was unclear, and at

the time of death the dog appeared to be in

remission, presented with vague clinical signs, was

not neutropenic, but had coagulopathy, fatigue

and severe depression, which was unresponsive to

supportive care. Two dogs died because of cardiac

disease, one of which had cardiac changes consistent

with doxorubicin-induced cardiomyopathy. This

particular dog, a Great Dane, received a modified

protocol and only four total doses of doxorubicin

when changes in the shortening fraction were

noted on routine cardiac echocardiogram before

the scheduled fifth dose of doxorubicin. The second



0 365 730 1095 1460 1825 2190 2555 2920

Days

0.0

0.2

0.4

0.6

0.8

1.0

Pro

port

ion

Sur

vivi

ng/T

TP

TTP 219 daysLymphoma-specific Survival

323 days

Figure 1. TTP and lymphoma-specific survival of all dogs treated with a short, dose-intensified CHOP-based protocol.

dog, a Bull Terrier, which died because of cardiac

disease received all six doses of doxorubicin, had

been treated for ventricular arrhythmias and did

not have dilated cardiomyopathy when he died. It

is unclear whether this particular dog’s heart disease

was related to the previous doxorubicin treatments.

The survival times of these two dogs were 164

and 445 days, respectively. In addition to these

two dogs, four other dogs (a Bulldog, a German

Shepherd, a Collie and a Golden Retriever) had

protocol modifications because of clinical concerns

of early doxorubicin cardio-toxicity. These dogs

received vincristine/cyclophosphamide instead of

doxorubicin in the consolidation phase to complete

the 20-week protocol. None of these four dogs

developed dilated cardiomyopathy later on.

Prognostic factors

Dogs that required dose reduction had significantly

longer TTP than dogs that did not need dose

reductions, with a median TTP of 234 versus

178 days, P = 0.04 (Table 6). There was no

statistically significant difference in TTP between

dogs that needed treatment delays compared

with those that did not, but when evaluating

the combined effect of dose-reduction and/or

treatment delay, the effect on TTP duration

became statistically significant, 242 versus 177 days,

P = 0.03 (Fig. 2). Stage, sub-stage, body weight,

presence of hypercalcemia or multicentric disease

did not significantly impact TTP in this study

(Table 6). Five of the 21 dogs with hypercalcemia

had immunophenotyping performed, and all five

dogs had T-cell lymphoma. The other 16 dogs

were not immunophenotyped. The TTP and

lymphoma-specific survival times in dogs with

B-cell lymphoma, B/T or T-cell lymphoma were

not significantly different, P = 0.19 and P = 0.17,

respectively (Table 6). Dogs with B-cell and B/T

immunostaining had similar outcomes and were

therefore combined as one group in a subsequent

analysis: B-cell + B/T versus T-cell, resulting in a

median TTP and lymphoma-specific survival times

of 247 and 399 days, respectively, in the B + B/T

group, versus 192 and 184 days, respectively, in

the T-cell group, P = 0.069 (TTP) and P = 0.07

(lymphoma-specific survival) (power 0.83). Sub-

stage, however, had a statistically significant impact

on lymphoma-specific survival; dogs with sub-

stage ‘a’ had a median survival of 450 days

versus 242 days in dogs that had sub-stage ‘b’,

P = 0.03 (Fig. 3). Dogs that required either dose

reductions or treatment delays had significantly

longer lymphoma-specific survival than dogs that

did not require dose reductions and/or treatment

delays.



0 365 730 1095 1460 1825 2190 2555 2920

Days

0.0

0.2

0.4

0.6

0.8

1.0

Pro

port

ion

with

TT

P No 177Yes 242

Tx delay and/orDose reduction

TTP (days)

No 177Yes 242


TTP (days)

0 365 730 1095 1460 1825 2190 2555 2920

Days

0.0

0.2

0.4

0.6

0.8

1.0

Pro

port

ion

Sur

vivi

ng (

LSA

spe

cific

)

No 174Yes 385


MST (days)

No 174Yes 385


MST (days)

P=0.002

P=0.03

Figure 2. TTP and lymphoma-specific survival and in dogs with dose reduction and/or treatment delay (yes) versus dogsthat did not require dose reduction and/or treatment delay (no).

Discussion

TTP and lymphoma-specific survival associated

with this protocol were similar to previously

published protocols. However, these results must

be interpreted with caution since the majority of

dogs in our study had advanced stage lymphoma

(63% stage V), were sick (58.5% sub-stage b) at

initial presentation and 32% had T-cell lymphoma.

Most previous reports have included dogs with

predominantly sub-stage ‘a’ disease, fewer with

stage V disease and fewer dogs with T-cell

lymphoma.4 – 11,13 A more recent publication,

however, included a population of dogs where the

majority had sub-stage ‘b’ lymphomas, similar to

our study.12 Despite the fact that assigning sub-

stage is somewhat subjective, sub-stage has been

one of the most consistent negative prognostic

factors in the canine lymphoma literature, as

it was in our current study. In addition to

sub-stage, immunophenotype is a recognized

prognostic factor in canine lymphoma. Results

from immunophenotyping were available for only

a subset of the dogs (n = 71) treated on this



Table 6. Prognostic variables and time to progression (TTP)/lymphoma-specific survival

Variable TTP median (95% CI) P-valueLymphoma-specific survivaldays, median (95% CI) P-value

Stage of disease 0.34 0.19

Stages I–IV 238 (211–322) 470 (260–594)

Stage V 195 (169–247) 286 (212–399)

Sub-stage 0.057 0.03

a 288 (219–408) 450 (383–588)

b 192 (146–225) 242 (193–313)

Multicentric LSA 0.67 0.77

Yes 221 (182–272) 373 (244–442)

No 205 (179–NC) 242 (65–492)

Body weight 0.56 0.81

≤15 kg 192 (82–225) 270 (127–594)

>15 kg 234 (189–288) 323 (243–436)

Hypercalcemia 0.95 0.67

Yes 179 (131–322) 244 (203–646)

No 225 (195–283) 372 (243–445)

Immunophenotype 0.19 0.17

B-cell 225 (180–408) 399 (194–555)

T-cell 192 (83–272) 184 (127–445)

B/T 416 (174–597) 425 (235–NC)

Dose reduced 0.04 0.004

Yes 234 (195–300) 383 (290–527)

No 178 (103–238) 203 (138–434)

Treatment delay 0.13 0.027

Yes 272 (197–394) 425 (313–560)

No 195 (137–242) 226 (168–399)

Dose reduced/treatment delay 0.03 0.002

Yes 242 (197–322) 385 (301–535)

No 177 (103–225) 174 (129–399)

Completed protocol <0.0001 <0.0001

Yes 322 (272–432) 574 (456–646)

No 96 (80–121) 138 (126–184)

particular protocol. Despite the fact that dogs with

B-cell or B/T-cell lymphoma appeared to have a

longer median survival (399 days) than dogs with

T-cell lymphoma (184 days), this difference did

not reach statistical significance at 0.05. It is quite

possible that a difference might have been found if

additional cases had been available for comparison,

even though the power of this comparison was

0.83 (83% probability of appropriately rejecting

the null-hypothesis). Alternatively, it is also

possible that this high dose-intensity protocol

was particularly effective for T-cell lymphoma,

thus eliminating immunophenotype as a negative

prognostic factor in this study. This latter theory

may be supported by the fact that hypercalcemia,

which is often associated with T-cell lymphomas,

was not associated with a poorer prognosis in dogs

treated with this protocol.

The incidence of toxicity was significant in

dogs treated with this protocol. Dose reduction

and treatment delays were common, and the rate

of hospitalization was also higher than typically

reported in veterinary literature. Because of the high

incidence of toxicity, especially secondary to the

combination of vincristine and cyclophosphamide

in weeks 2 and 3 of the protocol, the majority

of the dogs had to have dose reduction and

treatment delays relatively early in the protocol.

Subsequent treatments were administered at the

reduced dosages. Several dogs required more than

one dose reduction, reflecting that the approach

to dose reductions were performed in a gradual

stepwise fashion in order to keep the dose as

close to what the individual patient could tolerate.

Nevertheless, these protocol changes resulted in a

significant decrease in the SDI when compared



0 365 730 1095 1460 1825 2190 2555 2920

Days

0.0

0.2

0.4

0.6

0.8

1.0

Pro

port

ion

Sur

vivi

ng ly

mph

oma-

spec

ific

b 242a 450

SubstageMST

(days)

ðoðoðoðoðoðo

0 365 730 1095 1460 1825 2190 2555 2920

Days

0.0

0.2

0.4

0.6

0.8

1.0

Pro

port

ion

TT

P

b 192a 288

SubstageTTP

(days)

P=0.03

P=0.057

Figure 3. TTP and lymphoma-specific survival in dogs with sub-stage ‘a’ versus sub-stage ‘b’.

with the original protocol. However, contrary

to what might have been expected, these dose

reductions and delays were not detrimental to

the outcome; rather, the opposite occurred; dogs

that experienced toxicity and needed reductions

and delays fared better both in terms of TTP and

lymphoma-specific survival than those that did not

require dose modifications or treatment delays.

These results are similar to the findings in another

recent study, which also reported that dogs that

needed dose reductions had a better outcome.16

What implications do these findings have for the

concept of dose density? Does this mean that dose

intensification is a failed strategy? Not necessarily, in

fact, these results may confirm that dose intensity

is indeed important, but that it is relative, and

must be individualized and tailored to what a

particular dog or patient can tolerate. The dogs that

experienced toxicity and required modifications

in dose might have been receiving closer to his

or her maximally tolerated dose and therefore a

more appropriate, and thus a more effective dose

of chemotherapy. Chemotherapeutic drug doses

are typically calculated by converting body weight

in kilograms to the body surface area (BSA) of

the patient. This dosing strategy is based on the

assumption that BSA is a better estimate of cardiac

output and metabolic rate and, therefore, a more

accurate way of dosing these drugs. This method

may be particularly flawed in dogs as the variation



in size and body conformation can vary quite

dramatically depending on the breed of dog.33,34,

This study, however, did not find a significant

improvement in outcome in smaller dogs (<15 kg)

when compared with larger dogs (>15 kg), and the

incidence of dose reductions/treatment delays were

similar between the smaller and larger dogs. This

may be due to the fact that clinicians often dose

smaller dogs more conservatively than larger dogs,

especially when calculating the dose of doxorubicin.

This practice may result in decreased toxicity but

may also negate the previously reported outcome

advantage in smaller dogs. Many of the drugs used

in this protocol are metabolized by the liver, which

may add other somewhat unpredictable variables

because of organ dysfunction from concurrent

disease processes or individual variations in

pharmacogenetics into the equation and change

the pharmacokinetics of the drugs. Nevertheless,

calculating drugs based on BSA may provide a

reasonable starting point; however, the results

from our study suggest that this dose should be

re-assessed and increased or decreased according to

how the patient tolerated it in the earlier cycles of

treatment. In fact, adjusting the initial BSA-based

calculated dose to a toxicity adjusted dose according

to the patients’ response has been advocated by

other authors.35 Dose reductions are commonly

performed in the practice of veterinary oncology,

but our results suggest that perhaps we should

be open to gradually increasing the doses over

the initial BSA-based dose calculation in patients

that have no toxicity at standard doses. Such dose

escalations or de-escalations according to response

(i.e. toxicity) may result in improved outcome

by allowing for a more effective, dose-intensified

protocol that is tailored to the individual patient’s

maximal tolerance.

We used the relative efficacy methods to cal-

culate the SDI for this protocol. There is limited

information regarding the single dose efficacy of

the drugs used in this protocol, other than doxoru-

bicin. The effectiveness was therefore assigned to

each drug according to the results from COP and

according to the authors’ experience. This method

may have resulted in some drugs being overesti-

mated and other drugs being underestimated for

their contribution to the SDI, which is a weakness

to this approach. Doxorubicin was considered to be

the most valuable single agent in the protocol and

assigned the highest relative efficacy value in the for-

mula, and we believe most other oncologists would

agree with this assignment. Similarly, methotrex-

ate was given the lowest value in our formula

(Tables 2–4), and we also believe most oncologists

would agree with us that methotrexate is the weak-

est drug in this protocol. Some might even question

why it is included at all. In fact, methotrexate was

included in the third cycle instead of L’asparaginase

in order to spare the limited resources of avail-

able L’asparaginase during the parts of the period

this protocol was used. Most likely methotrexate

detracts from the total dose intensity in this pro-

tocol. The relative efficacies of the various drugs

were assigned fixed value throughout the protocol.

This is likely not likely the case in a real patient

where the effectiveness of the drugs is likely to

change and diminish over time after exposure.30

However, it would be impossible to accurately cal-

culate the dose intensity of a chemotherapy protocol

and incorporate the individual patient variation in

drug metabolism as well as the dynamic changes

in drug resistance that may occur over time in the

tumour cell population. Introducing an estimate

for drug resistance and decreased efficacy into the

formula would very likely have introduced more

error; hence, it is safest to assume that our esti-

mated SDI is most likely the maximum effect given

what we know. And even though the formula may

be flawed, it still provides us a metric by which we

can compare protocols and use as a guide when

designing other dose-intense protocols.

The results from this study confirm that canine

lymphoma is a heterogeneous disease with a wide

range in outcome. Approximately 20% of the dogs

did not relapse and did extremely well long term on

this protocol, which is similar to some of the previ-

ous studies. However, a significant portion (43%)

of the dogs in this study did not complete the pro-

tocol, because of either failure to achieve complete

progression-free survival or early relapse. Because

of referral filter bias, our institution appears to be

seeing more dogs with advanced lymphoma that are

sick from their cancer. Many of these dogs are clearly

not treated effectively with the current first-line

agents. It is unlikely that significant improvements



can be achieved through further dose intensifica-

tions without increasing the incidence of serious

toxicity or death. Eleven percent of the dogs in this

study were hospitalized because of serious acute tox-

icity, which is a higher incidence than what many

veterinarians and owners might find acceptable.

Newer agents or treatment strategies are needed to

improve the outcome in this particular population

of dogs.

Acknowledgments

This study was supported in part by the oncology

clinical research fund of the School of Veterinary

Medicine of the University of Pennsylvania and by

the AnCan Foundation.

References

1. Squire RA and Bush M. The therapy of canine and

feline Lymphosarcoma. Bibliotheca Haematologica

1983; 39: 189–197.

2. Page RL, Macy DW, Ogilvie GK, Rosner GL,

Dewhirst MW, Thrall DE, Withrow SJ,

McEntee MC, Cline JM and Heidner GL. Phase III

evaluation of doxorubicin and whole-body

hyperthermia in dogs with lymphoma. International

Journal of Hyperthermia 1992; 8: 187–197.

3. Hahn KA, Richardson RC, Teclaw RF, Cline JM,

Carlton WW, DeNicola DB and Bonney PL. Is

maintenance chemotherapy appropriate for the

management of canine lymphoma? Journal of

Veterinary Internal Medicine 1992; 6: 3–10.

4. Keller ET, MacEwen EG, Rosenthal RC, Helfand SC

and Fox LE. Evaluation of prognostic factors and

sequential combination chemotherapy with

doxorubicin for canine lymphoma. Journal of


5. Valerius KD, Ogilvie GK, Mallinckrodt CH and

Getzy DM. Doxorubicin alone or in combination

with asparaginase, followed by cyclophosphamide,

vincristine, and prednisone for treatment of

multicentric lymphoma in dogs. Journal of the

American Veterinary Medical Association 1997; 210:

512–516.

6. Zemann BI, Moore AS, Rand WM, Mason G,

Ruslander DM, Frimberger AE, Wood CA,

L’Heureux DA, Gliatto J and Cotter SM. A

combination chemotherapy protocol (VELCAP-L)

for dogs with lymphoma. Journal of Veterinary

Internal Medicine 1998; 12: 465–470.

7. Khanna C, Lund EM, Redic KA, Hayden DW,

Bell FW, Goulland EL and Klausner JS.

Randomized controlled clinical trial of doxorubicin

versus dactinomycin in a multiagent protocol for

treatment of dogs with malignant lymphoma.

Journal of the American Veterinary Medical

Association 1998; 213: 985–990.

8. Chun R, Garrett LD and Vail DM. Evaluation of a

high-dose chemotherapy protocol with no

maintenance therapy for dogs with lymphoma.

Journal of Veterinary Internal Medicine 2000; 14:

120–124.

9. Moore AS, Cotter SM, Rand WM, Wood CA,

Williams LE, London CA, Frimberger AE and

L’Heureux DA. Evaluation of a discontinuous

treatment protocol (VELCAP-S) for canine

lymphoma. Journal of Veterinary Internal Medicine

2001; 15: 348–354.

10. Mutsaers AJ, Glickman NW, DeNicola DB,

Widmer WR, Bonney PL, Hahn KA and

Knapp DW. Evaluation of treatment with

doxorubicin and piroxicam or doxorubicin alone

for multicentric lymphoma in dogs. Journal of The

American Veterinary Medical Association 2002; 220:

1813–1817.

11. Garrett LD, Thamm DH, Chun R, Dudley R and

Vail DM. Evaluation of a 6-month chemotherapy

protocol with no maintenance therapy for dogs

with lymphoma. Journal of Veterinary Internal

Medicine 2002; 16: 704–709.

12. Simon D, Nolte I, Eberle N, Abbrederis N,

Killich M and Hirschberger J. Treatment of dogs

with lymphoma using a 12-week, maintenance-free

combination chemotherapy protocol. Journal of

Veterinary Internal Medicine 2006; 20:

948–954.

13. Frimberger A, Moore A, Rassnick K, Cotter S,

O’Sullivan J and Quesenberry P. A combination

chemotherapy protocol with dose intensification

and autologous bone marrow transplant

(VELCAP-HDC) for canine lymphoma. Journal of


14. Hosoya K, Kisseberth WC, Lord LK, Alvarez FJ,

Lara-Garcia A, Kosarek CE, London CA and

Couto CG. Comparison of COAP and UW-19

protocols for dogs with multicentric lymphoma.


1355–1363.

15. Rassnick KM, McEntee MC, Erb HN, Burke BP,

Balkman CE, Flory AB, Kiselow MA, Autio K and

Gieger TL. Comparison of 3 protocols for treatment

after induction of progression free survival in dogs

with lymphoma. Journal of Veterinary Internal

Medicine 2007; 21: 1364–1373.



16. Vaughan A, Johnson JL and Williams LE. Impact of

chemotherapeutic dose intensity and hematologic

toxicity on first progression free survival duration

in dogs with lymphoma treated with a

chemoradiotherapy protocol. Journal of Veterinary

Internal Medicine 2007; 21: 1332–1339.

17. Simon D, Moreno SN, Hirschberger J, Moritz A,

Kohn B, Neumann S, Jurina K, Scharvogel S,

Schwedes C, Reinacher M, Beyerbach M and

Nolte I. Efficacy of a continuous multiagent

chemotherapy protocol versus a short-term single

agent protocol in dogs with lymphoma. Journal of

the American Veterinary Medical Association 2008;

232: 879–885.

18. Vail DM and Young KM. Hematopoietic tumors,

chapter 31. In: Withrow and MacEwen’s Small

Animal Oncology, 4th edn., SJ Withrow and

DM Vail, eds., St. Louis, MO, Saunders Elsevier,

2007: 699–784.

19. Carter RF, Harris CK and Withrow SJ.

Chemotherapy of canine lymphoma with

histopathological correlation: doxorubicin alone

compared to COP as first treatment regimen.

Journal of the American Animal Hospital Association

1983; 23: 587–596.

20. Cotter SM and Goldstein MA. Treatment of

lymphoma and leukemia with cyclophosphamide,

vincristine, and prednisone I treatment of the dog.

Journal of the American Animal Hospital Association

1983; 19: 159–165.

21. Itoh K and Fukuda H. Randomized phase II study

of biweekly CHOP and dose-escalated CHOP with

prophylactic use of lenograstim (glycosylated

G-CSF) in aggressive non-Hodgkin’s lymphoma:

Japan Clinical oncology group study 9505. Annals of

Oncology 2002; 13: 1347–1355.

22. Balzarotti M, Spina M and Sarina B. Intensified

CHOP regimen in aggressive lymphomas: maximal

dose intensity and dose density of doxorubicin and

cyclophosphamide. Annals of Oncology 2002; 13:

1341–1346.

23. Kaiser U, Uebelacker I, Abel U, Birkmann J,

Trumper L and Schmalenberg H. Randomized

study to evaluate the use of high-dose therapy as

part of primary treatment for ‘‘aggressive’’

lymphoma. Journal of Clinical Oncology 2002; 20:

4413–4419.

24. Nademee A, Schmidt GM and O’Donnell MR. High

dose chemotherapy followed by autologous bone

marrow transplantation as consolidation therapy

during first complete progression free survival in

adult patients with poor-risk aggressive lymphoma:

a pilot study. Blood 1992; 80: 1130–1134.

25. Fisher RI, Gaynor ER and Dahlberg S. Comparison

of a standard regimen (CHOP) with three intensive

chemotherapy regimens for advanced

Non-Hodgkin’s lymphoma. New England Journal of

Medicine 1993; 328: 1002–1006.

26. Kluin-Nelemans H, Zagonel V and

Anastasopoulou A. Standard chemotherapy with or

without high-dose chemotherapy for aggressive

Non-Hodgkin’s lymphoma: randomized phase III

EORTC study. Journal of the National Cancer

Institute 2001; 93: 22–30.

27. Tu LC, Foltz G, Lin E, Hood L and Tian Q.

Targeting stem cells-clinical implications for cancer

therapy. Current Stem Cell Research & Therapy

2009; 4: 147–153.

28. Winquist RJ, Boucher DM, Wood M and Furey BF.

Targeting cancer stem cells for more effective

therapies: taking out cancer’s locomotive engine.

Biochemical Pharmacology 2009; 78: 326–334.

29. Tung DC and Chao KS. Targeting hedgehog in

cancer stem cells: how a paradigm shift can improve

treatment response. Future Oncology 2007; 3:

569–574.

30. Frei E III, Elias A, Wheeler C, Richardson P and

Hryniuk W. The relationship between high-dose

treatment and combination chemotherapy: the

concept of summation dose intensity. Clinical

Cancer Research 1998; 4: 2027–2037.

31. Simon D, Moreno SN, Hirschberger J, Moritz A,

Kohn B, Neuman S, Jurina K, Scharvogel S,

Schwedes C, Reinacher M, Beyerbach M and

Nolte I. Efficacy of a continuous, multiagent

chemotherapeutic protocol versus a short-term

single-agent protocol in dogs with lymphoma.

Journal of the American Veterinary Medical

Association 2008; 232: 879–885.

32. Postorino NC, Susaneck SJ and Withrow SJ. Single

agent therapy with Adriamycin for canine

lymphosarcoma. Journal of the American Animal

Hospital Association 1989; 25: 221–225.

33. Arrington KA, Legendre AM and Tabeling GS.

Comparison of body surface area-based and

weight-based dosage protocols for doxorubicin

administration in dogs. American Journal of

Veterinary Research 1994; 55: 1587–1592.

34. Frazier DL and Price GS. Use of body surface area

to calculate chemotherapeutic drug dose in dogs: II.

Limitations imposed by pharmacokinetic factors.


272–278.

35. Gurney H. Dose calculation of anticancer drugs: a

review of the current practice and introduction of

an alternative. Journal of Clinical Oncology 1996; 14:

2590–2611.


Outcome and toxicity associated with a dose-intensified, maintenance-free CHOP-based chemotherapy protocol in canine lymphoma: 130 cases

Documents