Adverse events

A total of 283 episodes of adverse events were noted for all 64 dogs in this study, with 192 ones from the 25-week group and 91 ones from the 15-week group.

Hematologic AEs comprised 32.2% of total episodes, while gastrointestinal AEs made up 67.8% of all. 5 dogs didn’t experience any toxicity, with 3 from the 25-week group and 2 from the 15-week group. Instead, 9 dogs experienced at least one grade 4 toxicity, with 5 dogs from the 25-week group and 4 dogs from the 15-week group. The detailed distribution of frequency and severity of adverse events was summarized in Table 9.

Among the AEs of hematologic in nature, 4 (57.1%) dogs in the 25-week group and 10 (45.5%) dogs in the 15-week group experienced neutropenia during treatment.

Grade 1 neutropenia was the most frequently presented in both groups, accounting for 51% of neutropenia episodes in the 25-week group and 40% of that in the 15-week group. Grade 4 neutropenia was noted in 4 dogs from both groups, being 9.5% and 18.2% in each population, and among them only 1 dog from the 15-week group experienced grade 4 neutropenia for twice. Febrile neutropenia with grade 1 fever (39.9

) was noted in 1 dog from the 15-week group on the third day after administration of 1^st vincristine. The blood examination on the same day revealed grade 2 anemia (PCV=24.9%), grade 1 neutropenia (neutrophil=1581/µL) with decreased segmentation and mild toxic reaction, and grade 2 thrombocytopenia (platelet=84000/µL, by manual calculation). This dog recovered after administration of oral antibiotics (i.e. Augmentin).

There was neither significance in the occurrence of neutropenia, nor occurrence of grade 4 neutropenia between two groups. (P=0.869 and 0.320, respectively). For

thrombocytopenia, only 5 (21.4%) dogs in the 25-week group and 5 (22.7%) dogs in the 15-week group experienced the toxicity. Grade 1 thrombocytopenia predominated for both groups. Toxicity other than grade 1 was only noted in a minority of 23.5% of episodes from the 25-week group being grade 2. No significance was noted in the occurrence of thrombocytopenia between two groups (P=0.905). Generally, the frequency and severity of neutropenia were more profound than those of thrombocytopenia in both groups.

For AEs of gastrointestinal in nature, more than half of dogs in both groups experienced anorexia and vomiting. Grade 1 and grade 2 toxicity were in majority, with grade 2 anorexia and grade 1 vomiting being the most frequently presented situations in both two groups. Grade 4 toxicity was only noted once in 1 dog of the 25-week group with anorexia. No significance was noted in the occurrence of anorexia, vomiting and diarrhea between two groups (P=0.299, 0.749 and 0.487, respectively).

Two dogs from the 25-week group and 1 dog from 15-week group developed sterile hemorrhagic cystitis after first, second and third administration of cyclophosphamide, respectively. Cyclophosphamide was subsequently substituted by chlorambucil in the following cycles for these dogs.

One dog from the 15-week group developed dilated cardiomyopathy speculated to be resulted from cumulative cardiotoxicity of doxorubicin. This was a 5-year old Corgi.

Panting and exercise intolerance had been noted for 1 month after last dose of doxorubicin injection, with the cumulative dose of doxorubicin being 180 mg/m². Deterioration of respiratory signs and presence of pleural effusion and ascites were noted at second month after last dose. Echocardiography revealed dilated

Three days after discharge, this dog eventually died at home (one week after diagnosis of cardiomyopathy). Re-evaluation of echocardiography was thus unavailable.

Overall 20 dose reductions were observed in 13 (31%) dogs of the 25-week group and 15 dose reductions in 9 (40.9%) dogs from the 15-week group. The most common cause for dose reduction was neutropenia, accounting for 16 (80%) of episodes from the 25-week group and 8 (53.3%) episodes from the 15-week group. Other miscellaneous causes included gastrointestinal toxicity (n=6), elevated liver enzyme (n=3). There was no significant difference in the rate of dogs requiring dose reduction between two groups (P=0.326).

A total 62 episodes of dose delay from 29 (69%) dogs in the 25-week group and 30 episodes in 15 (68.2%) dogs from the 15-week group were noted. Neutropenia still remained the major cause for dose delay, accounting for 43 (69.4%) times of delay in the 25-week group and 19 (63.3%) times of delay in the 15-week group.

Gastrointestinal toxicity resulted in 7 (11.3%) times of delay from the 25-week group and 4 (13.3%) times of delay from the 15-week group. Elevated liver indexes were associated with 3 (4.8%) episodes and 5 (16.7%) episodes from the 25 and 15-week group respectively. Other miscellaneous causes were such as leukocytosis, foreign body digestion before visit or owner’s request. There was no significant difference in the rate of dogs requiring dose delay between two groups (P=0.385).

Other than the dose reductions meant for ATLS prevention at treatment initiation, 11 (26.2%) dogs from the 25-week group and 6 (27.3%) dogs from the 15-week group required neither dose reduction nor delay during treatment.

4.5 Prognostic factors

4.5.1 Prognostic factors analysis for CR

The following factors were analyzed to see if they played a role in attaining CR or not: age (< or ≥ median age), body weight (< or ≥ median body weight), breed (Golden retriever or other breeds), sex (spayed female, intact female, castrated male or intact male), stage (< 5 or 5), substage (a or b), immunophenotype (B-cell, T-cell or null), administration of steroid before chemotherapy, presence of thoracic involvement in imaging, characteristics of thoracic imaging (infiltration or swollen lymph node), presence of laboratory abnormalities present at diagnosis (including anemia, thrombocytopenia, lymphocytosis and leukocytosis), baseline globulin (< or ≥ median value of globulin), baseline neutrophil (< or ≥ median value of neutrophil), presence and severity of AEs during treatment (including neutropenia, thrombocytopenia, anorexia, vomiting and diarrhea), need for dose delay, number of dose delay (< or ≥ 2 delays), need for dose reduction, number of dose reductions (< or ≥ 2 reductions), the use of L-asparaginase in protocol, dose reduction for ATLS prevention at treatment initiation, and the use of mitoxantrone in protocol. Hypercalcemia is not evaluated because only one dog has hypercalcemia among the whole population.

In univariate analysis of the above factors, body weight, breed and thoracic involvement in imaging were showed result in statistically significance in attaining CR or not during treatment. All dogs with body weight lower than median value (i.e. 16.15 kg) achieved CR; however, only 75% of dogs with body weight higher than median value attained CR (P=0.002). For dogs who were not Golden retrievers, the rate of CR

dogs with an abnormality in thoracic imaging (P=0.018). Interestingly, no significance was noted between different characteristics of thoracic imaging (85.7% rate of CR for infiltration versus 73.7% for swelling of lymph nodes, P=0.417).

Relatively higher rate of CR was shown in dogs presented with neutropenia during the treatment (94.1% versus 80.0%) and dogs encountering more than 2 times of dose delay (96.4% versus 80.6%), but both were not significant enough (P=0.088 and 0.057, relatively). The result of analysis was summarized in Table 10-12.

No statistical significance was found between dogs ever receiving mitoxantrone during the treatment and those who did not (P=0.517). Particularly in 25-week group, either the use of CHOP or CMOP resulted no significant difference in achieving CR (P=0.780).

4.5.2 Prognostic factors analysis for TTP

Factors mentioned previously in univariate analysis for attaining CR or not, along with response to treatment, were analyzed for TTP in all patients as well. Body weight, breed and response to treatment were factors shown to have significant influence on TTP (Table 13-15). Dogs with body weight higher than median value had significantly shorter median TTP than those whose body weight lower than median value (231 days versus 209 days, P=0.010). Golden retrievers had median TTP of 126 days which was significantly shorter than that of 244 days in other breeds (P=0.024). In dogs able to attain CR during the treatment, median TTP was 244 days and significantly longer than 56 days from those failed to achieve CR (P=0.000). Dogs with substage a also had relatively longer median TTP than those with substage b, but significance was not reached (273 days versus 209 days, P=0.069, Figure 6). The Kaplan-Meier curves for the above factors with significance were shown in Figure 3 to 5.

In multivariate analysis for factors with significance (P<0.05) in TTP, only response to treatment remained statistically significant. Hazards ratio for dogs without CR and those who achieving CR was 3.560 (95% confidence interval, 1.433-8.842, P=0.006). Both body weight and breed were not significant (P=0.176 and 0.999,

respectively) in multivariate analysis. Results of multivariate analysis with hazards ratio (HR) and 95% confidence interval (95% CI) of factors for TTP were summarized in Table 16.

No significance was found between dogs ever receiving mitoxantrone during the treatment and those who didn’t (242 days versus 231 days, P=0.209). In the 25-week group, the median TTP between subgroup of CHOP and CMOP was revealed no statistically significant (242 days versus 273 days, P=0.274).

4.5.3 Prognostic factors analysis for OST

Factors mentioned previously in univariate analysis for TTP were analyzed for OST in all patients (Table 17-19). Body weight, breed, presence of neutropenia during treatment and response to treatment were significantly associated with median OST.

With similar findings as in TTP, dogs with body weight higher than median value and Golden retrievers both had significantly shorter median OST, with 307 days versus 555 days (P=0.002) for the former subgroup and 245 days versus 496 days for the latter subgroup (P=0.007). Response to treatment remained significant for median OST as well, with 392 days in dogs attaining CR and 189 days in dogs without CR. In the aspect of adverse events from treatment, dogs ever experiencing neutropenia had significantly higher median OST (514 days versus 309 days, P=0.029). The

However, no factors with significance (P ≤ 0.05) from univariate analysis remained significant in multivariate analysis for median OST. Results of multivariate analysis with hazards ratio (HR) and 95% confidence interval (95% CI) of factors for OST were summarized in Table 20.

For dogs ever receiving mitoxantrone during the treatment and those who didn’t, median OST was not statistically significant between two subgroups (313 days versus 379 days, P=0.136). In 25-week group, the median OST for subgroup of CHOP and CMOP was 313 days and 379 days, respectively (P=0.176), without significance found.

4.5.4 Prognostic factors analysis for large-sized population and non-GR population

According to the results above, we further recognized the dogs with body weight over 20 kg from the other relatively small-sized dogs. In this subgroup of large-sized population, the rate of complete remission, TTP and OST were compared between Golden retrievers and other large-sized non-GR dogs. Detailed distribution for body weight was listed in Table 21. Although the rate of CR, median TTP and median OST were all numerically lower in GR than in large-breed non-GR dogs, statistical significances were not observed for either endpoint between two groups (Table 22).

Again, we eliminated Golden retrievers and recognized the subgroup of non-GR population. In this subgroup, the comparisons were focused between dogs < 20 kg and dogs ≥ 20 kg. No significant difference was noted in median TTP and OST (P=0.171 and 0.100, respectively) between two groups. However, the rate of complete remission was significantly higher in dogs < 20 kg than dogs ≥ 20 kg (100% vs. 81.8%, P=0.049) (Table 22).