Portions of this material were presented in abstract form to the 28th Annual Conference of the Veterinary Cancer Society, Seattle, WA, October 2008 and the 29th Annual Conference of the Veterinary Cancer Society, Austin, TX, October 2009.
Corresponding author: B. J. Biller, DVM, PhD, Department of Clinical Sciences, Animal Cancer Center, Colorado State University, Fort Collins, CO 80523; e-mail: firstname.lastname@example.org.
Background: Low-dose, continuous (metronomic) chemotherapy improves tumor control by inhibiting tumor angiogenesis and suppressing regulatory T cells (Treg) in mice and humans. The effects of metronomic chemotherapy on Treg and tumor angiogenesis in dogs has not been investigated previously.
Objective: To determine whether metronomic cyclophosphamide (CYC) therapy decreases Treg or exhibits antiangiogenic activity or both in dogs with soft tissue sarcoma (STS). We hypothesized that Treg numbers would be increased in dogs with STS and that continuous dosing of CYC would decrease Treg in a dose-dependent manner, as well as exhibit antiangiogenic activity.
Animals: Eleven client-owned dogs with grade I or II STS. Twenty-one healthy dogs were used as controls.
Methods: Prospective, open, clinical trial. Dogs with STS were enrolled in 2 dose cohorts and administered CYC at 12.5 or 15 mg/m2 PO once daily for 28 days. Whole blood and tumor biopsy specimens were obtained on days 0, 14, and 28 to assess changes in T lymphocyte subsets by flow cytometry and tumor microvessel density (MVD), respectively.
Results: Administration of CYC at 12.5 mg/m2/d significantly decreased the number of Treg from days 0 to 28, but there was no change in the percentage of Treg or tumor MVD. In dogs that received CYC at 15.0 mg/m2/d, both the number and percent of Treg as well as tumor MVD were significantly decreased over 28 days.
Conclusions: CYC administered at 15 mg/m2/d should be used in further studies examining the antitumor properties of low-dose CYC in dogs.
Metronomic chemotherapy is defined as the continuous administration of chemotherapy drugs at doses that are significantly lower than conventional maximally tolerated dose (MTD) therapy.1 In contrast to the direct cytotoxic effects of MTD chemotherapy, metronomic drug delivery inhibits tumor angiogenesis, a process critical for tumor development and progression.2 There are several proposed mechanisms for these effects, including targeting of the drug-sensitive endothelial cell compartment of tumors, inhibiting mobilization of proangiogenic circulating endothelial cells (CECs) from the bone marrow, and induction of the endogenous angiogenesis inhibitor thrombospondin 1.3–5
In addition to antiangiogenic properties, there is accumulating evidence that metronomic chemotherapy modulates the immune system of tumor-bearing patients through inhibitory effects on regulatory T cells (Treg). Treg, a subset of CD4+ lymphocytes that normally function to prevent autoimmunity, inhibits antitumor immune responses.6–8 Treg are frequently present in high numbers in human cancer patients and increased numbers of Treg may be associated with a less favorable outcome for certain malignancies.9–12 Treg are also increased in the peripheral blood of dogs with various types of cancer; however, the prognostic significance of this finding is unclear.13–17 In mice and humans with advanced cancer, the administration of metronomic doses of cyclophosphamide (CYC) selectively decreases circulating Treg numbers and inhibits their function.18–20 The effects of low-dose CYC on canine Treg have not been reported previously.
Despite its frequent use in animals, metronomic chemotherapy has not been systematically evaluated and dosing schedules and drug combinations have been selected empirically.21,22 There exist several challenges in designing effective metronomic chemotherapy protocols. First, most protocols tend to utilize a combination of drugs, each with differing mechanisms of action. The efficacy and therefore validity of using these drugs can only be determined once single drug effects are known; such experiments have not yet been performed. Secondly, dose optimization has been difficult to establish because low-dose protocols are not defined by dose-limiting toxicities in the same manner as MTD chemotherapy. In addition, therapies that target tumor vasculature or a patient's immune system are unlikely to demonstrate a clinical benefit as quickly as high-dose, cytotoxic chemotherapy. Therefore, standard criteria for monitoring antitumor response could be inadequate for assessing efficacy of antiangiogenic or immune-modulating therapies.23 Finally, clinically applicable biomarkers of tumor response have not been identified, limiting the use of metronomic protocols as adjunctive therapies.
To address some of these challenges, we performed a prospective clinical trial to evaluate the effects of low-dose CYC on Treg numbers and percentages and tumor microvessel density (MVD) as potential markers of treatment response in a canine cancer model. Our specific goal was to determine the dosage of CYC required to decrease circulating Treg and inhibit tumor angiogenesis in dogs with soft tissue sarcoma (STS).
Materials and Methods
Dogs presenting to the Colorado State University Veterinary Teaching Hospital with histologically confirmed grade 1 or 2 STS were eligible for inclusion in this prospective clinical trial. This population was selected with the rationale that optimal treatment for low to intermediate grade STS was unlikely to be adversely affected by enrollment in a 4-week long study before pursuing definitive treatment. Dogs were required to have a measurable and biopsy-accessible tumor, and no evidence of metastatic disease or significant biochemical or hematological abnormalities. Dogs were also required to have a Veterinary Co-Operative Oncology Group performance status of 0 or 1 (0, normal activity; 1, restricted activity [decreased activity from predisease status]; 2, compromised [ambulatory only for vital activities, consistently defecates and urinates in acceptable areas]; 3, disabled [dog needs to be force-fed, is unable to confine urination and defecation to acceptable areas]; 4, dead). Previous therapy was allowable with a 3-week washout period from chemotherapy or radiation therapy and a 3-day washout from any nonsteroidal anti-inflammatory drug (NSAID). Information recorded included age, breed, sex, body weight, CBC results, and tumor size, grade, and location. This study was approved by the Institutional Animal Care and Use Committee at Colorado State University, and signed informed consent was obtained from all owners before study entry.
Treatment Protocols and Sample Collection
For analysis of Treg and effector T cells in normal dogs, blood samples were collected from 21 healthy, age-matched dogs. Dogs with STS were enrolled in 2 cohorts. The 1st group was treated with CYC at a target dosage of 12.5 mg/m2/d administered PO at home by the owners for 28 days; the 2nd group was treated with oral CYC at target dosage of 15.0 mg/m2/d. The initial CYC dosage was selected based on previously published clinical studies examining metronomic chemotherapy for canine cancer. In these studies, dosages of CYC ranged from 10 to 25 mg/m2/d.21,22 Additionally, our group performed a pilot study in which CYC was administered at 10 mg/m2/d to dogs with various tumor types to assess the effect of low-dose CYC on Treg numbers in peripheral blood. No significant alterations in percentages or absolute numbers of Treg were observed when CYC was given at 10 mg/m2/d (unpublished data).
To improve accuracy of drug administration in the present study, CYC was compounded into 2.5 and 5 mg capsules. A CBC, whole blood for collection of peripheral blood mononuclear cells (PBMCs), and tumor biopsy specimens were obtained at days 0, 14, and 28 after beginning CYC therapy. Tumor biopsies were performed using local anesthesia and a needle core biopsy instrument in an aseptically prepared site on the tumor. The longest diameter of the tumor was monitored at each time point and tumor response was assessed by Response Evaluation Criteria in Solid Tumours (RECIST).24 Clients completed a dog quality of life questionnaire at each visit.
PBMC were obtained from blood samples collected in EDTA tubes after lysis of red blood cells. PBMC were added at a concentration of approximately 1 × 106 cells per well in 96-well round bottom plates and then immunostained for surface expression of CD4 and CD8 with FITC-conjugated anticanine CD4 mAB (clone YKIX302.9)a and Alexa 647-conjugated anti-canine CD8 mAB (clone YCATE55.9)a following the method described previously.25 Immunostaining for FoxP3 expression was conducted as previously described.13 Briefly, after washing to remove unbound surface antibodies, intracellular detection of FoxP3 was performed with a cross-reactive, PE-conjugated murine FoxP3 antibody (clone FJK-16 seconds).b A directly conjugated rat IgG2A antibody was used as the isotype control.
Flow Cytometric Analysis
Flow cytometry was performed with a CyAn ADP flow cytometerc and Summit software for data analysis. Analysis gates were set on the live lymphocyte population based on typical forward and side scatter characteristics.26 The percentage of Treg was calculated by determining the percentage of CD4+FoxP3+ cells within the CD4+ T-cell population. The percentages of CD4+ and CD8+ T cells were also determined. Absolute numbers of Treg, CD4+, and CD8+ T cells in peripheral blood were calculated based on the total lymphocyte count determined from a CBC run on an automated cell counter.
Biopsy specimens were embedded in Tissue-Tek OCT compound,d snap frozen in liquid nitrogen, and stored at −80°C until processing. Tissues were cyrosectioned to a thickness of 4 μm and adhered to positively charged glass slides. Immunohistochemistry was performed with a mouse anti-human CD146 monoclonal antibody (clone P1H12),e that cross-reacts with dog endothelium and has been validated as a measure of tumor of angiogenesis.27,28 The slides were fixed in acetone for 4 minutes, allowed to air dry, and blocked with 5% donkey serum.f The tissues were then incubated with a 1 : 100 dilution of the primary antibody followed by a 1 : 400 dilution of biotin-conjugated donkey anti-mouse IgG.e The slides were then incubated with strepavidin-HRP,g followed by AEC substrate,g then hematoxylin counterstain. Normal canine liver and spleen were used as positive control tissues. Positive control samples were processed as outlined above and negative control samples were incubated with 1 × PBS rather than the primary antibody.
To quantify tumor MVD, 3–5 photomicrographs for each time point were obtained at × 20 magnification with a Zeiss AxioVision microscope and AxioVision Software v4.6.h Computerized determination of the percentage of CD146+ vessels positive within each × 20 field was performed and averaged for each time point. The number of microvessels also included immunopositive endothelial cells or endothelial cell clusters regardless of the presence or absence of a lumen.29,30 The final MVD value for each time point was calculated as the mean (± SD) percentage of microvessels for the 3–5 20 × fields.
Data regarding age of healthy and STS-bearing dogs were compared by an unpaired, 2-tailed t-test. Data regarding age, weight, and tumor size of the 2 CYC dose cohorts of dogs enrolled in the study were also compared by an unpaired, 2-tailed t-test. Data regarding sex of normal and STS-bearing dogs as well as data regarding the sex of the dogs in the 2 dose cohorts were compared by Fisher's exact test. Changes in mean percentages and absolute numbers of Treg and MVD were compared by paired, 1-tailed t-tests. Hematologic parameters and mean percentages and absolute numbers of CD4+ and CD8+ T cells were compared by paired, 2-tailed t-tests. Statistical calculations were performed by a commercial software program.1 A P-value of <.05 was considered statistically significant for all analyses.
Twelve dogs were enrolled in and 11 dogs completed this clinical trial. The 1 dog that did not complete the study had a large STS on the lateral lumbar spine, which was causing the dog significant discomfort at the time of enrollment. Tramadol was administered to the dog while on study but the owner elected euthanasia before the day 14 time point because of a continued poor quality of life. The pretreatment samples obtained from this dog were included in comparisons of dogs with STS to the healthy control dogs but excluded from other analyses. Of the dogs that completed the study, the first 5 received a mean CYC dose of 12.4 (± 1.0) mg/m2/d (range 11.2–13.9 mg/m2/d; target dose 12.5 mg/m2/d) and the next 6 received a mean CYC dose of 16.0 (± 1.2) mg/m2/d (range 14.0–17.0 mg/m2/d; target dose 15.0 mg/m2/d). Five dogs were castrated males and 6 were spayed females with an average age of 9.9 years (range 6.9–12.8 years). There was no significant difference in age, sex, or weight between the 2 groups; similarly there was no significant difference in age or sex between the healthy dogs and those with STS (data not shown). Tumor biopsies were performed for all dogs before enrollment; 8 dogs had grade 1 STS and 3 had grade 2 STS; 72.7% (8/11) of these tumors occurred at or distal to the elbow or stifle.
Average tumor maximal diameter was 9.3 cm with a range of 1.9–30 cm; there was no significant difference in tumor size between the 2 dose cohorts. For all dogs in the study, maximal tumor diameters remained stable during the treatment period, according to RECIST criteria. Additionally, no adverse events were reported for any of the dogs over the 28-day study period.
Circulating Treg Are Significantly Increased in Dogs with STS
Whole blood was collected from control dogs as well as dogs with STS and evaluated by flow cytometry to determine the percentage of CD4+ and CD8+ T cells and Treg in peripheral blood as described in “Materials and methods.” Both the percentage and absolute number of Treg in peripheral blood were significantly increased in the dogs with STS before therapy as compared with healthy dogs (Fig 1). The mean percentage of Treg for dogs with STS was 7.6 (± 3.0)% as compared with 4.8 (± 1.3)% for the normal dogs (P < .001). The mean absolute number of Treg in peripheral blood was also increased in dogs with STS (107.8, SD ± 58.2 cells/μL) as compared with the normal dog population (35.4, SD ± 19.6 cells/μL; P < .001). The total number and percentage of CD4+ and CD8+ cells did not differ significantly between the normal and diseased populations (data not shown).
Low-Dose CYC Decreases Treg Numbers in Dogs with STS
To determine the effects of metronomic CYC therapy on numbers of Treg in circulation, blood samples were collected on days 0, 14, and 28 after beginning oral CYC, and changes in the mean absolute number and percentage of Treg were analyzed at each time point for the 2 groups of dogs. For dogs with STS receiving CYC at a target dosage of 12.5 mg/m2/d, there was no significant change in the mean percentage of Treg over the 28-day period (Fig 2A). A significant decrease in the mean absolute number of Treg was seen, however, from days 0 to 28 (P= .001; Fig 2B). The mean absolute number of pretreatment Treg was 78.9 (± 43.1) cells/μL, which decreased to 54.1 (± 43.7) cells/μL after 28 days of CYC therapy. No significant change in the absolute numbers of Treg was observed between days 0 and 14 or days 14 and 28.
Dogs receiving CYC at a target dosage of 15.0 mg/m2/d had significant decreases in both the absolute number and percentage of Treg in the peripheral blood during the 28-day study period (Fig 3). For this group of dogs, a significant decrease in the mean percentage and mean absolute number of Treg occurred between days 14 and 28 (7.4 ± 3.1 versus 5.8 ± 3.5%; P= .035 and 121.1 ± 64.1 versus 55.7 ± 27.9 cells/μL; P= .049).
Low-Dose CYC Mediates Selective Decreases in Treg
To determine the effects of low-dose CYC on leukocyte numbers, a CBC was performed at each study time point for all dogs. There were no significant changes in the total numbers of neutrophils, lymphocytes, monocytes, or platelets in dogs with STS receiving CYC at 12.5 or 15.0 mg/m2/d. To determine the effect of CYC on T-lymphocyte subsets, the CD4+ and CD8+ T cell populations were also evaluated. Despite the decrease of Treg at the higher dosage of CYC, there were no changes in the absolute numbers or percentages of either CD4+ or CD8+ lymphocytes during the course of the clinical trial, suggesting that the effects of CYC were selective for the Treg subset of T-lymphocytes (Table 1).
Table 1. Alterations in CD4+ and CD8+ lymphocyte subpopulations in dogs with STS treated with metronomic CYC for 28 days.
12.5 mg/m2 Cohort (Mean ± SD)
15.0 mg/m2 Cohort (Mean ± SD)
STS, soft tissue sarcoma; CYC, cyclophosphamide.
P < .05.
45.7 (± 13.1)
44.6 (± 7.9)
33.8 (± 6.4)
51.3 (± 16.7)
694 (± 492)
500 (± 274)
463 (± 169)
611 (± 326)
26.9 (± 4.0)
25.6 (± 5.7)
20.5 (± 7.1)
17.6 (± 8.6)
396 (± 231)
318 (± 222)
270 (± 100)
204 (± 119)
Low-Dose CYC Decreases Tumor MVD in Canine STS
Tumor MVD was evaluated at each time point to assess whether low-dose CYC exhibits antiangiogenic properties in canine STS. This was assessed by measuring the tumor MVD in serial tumor biopsies for each dog. At the time of MVD analysis, it was determined that several samples consisted of only hemorrhage, or necrotic material or both with no identifiable tumor present and these samples were excluded from MVD analysis. This insufficient sampling resulted in the small number of dogs from each dose cohort that could be evaluated for changes in MVD over the study period. There were no significant changes in the mean tumor MVD between the three study time points in the tumors of dogs treated with 12.5 mg/m2/d CYC (Fig 4A). However, in the dogs receiving CYC at 15.0 mg/m2/d, the mean tumor MVD significantly decreased from a pretreatment mean of 2.6 (± 0.9)% to a mean of 1.3 (± 1.0)% on day 14 (P= .015). This decrease persisted through day 28 (1.8 ± 0.8%) and was also significantly decreased from the MVD before therapy (P= .004; Fig 4B). The percent MVD was not significantly different between days 14 and 28 in the 15.0 mg/m2/d dose cohort.
Several important findings emerged from this investigation. First, when compared with a population of healthy dogs, the blood of dogs with STS contained significantly increased numbers of Treg. When dogs with STS received oral CYC at a dosage of 15 mg/m2/d, Treg numbers and percentages declined over the 28-day study period to those of healthy dogs. In addition, daily administration of CYC at 15 mg/m2 was associated with a significant decrease in the density of blood vessels within tumor tissues, suggesting that metronomic delivery of CYC at this dosage may have both immunomodulatory and antiangiogenic effects. Because of the small number of dogs evaluated in this study, these results should be considered preliminary.
Previous work has established that Treg are increased in humans with a wide range of malignancies, including melanoma, carcinoma of the ovaries, pancreas, breast and liver, and a number of hematologic malignancies; however, increases in Treg in humans with STS have not been described previously.11,31 Recent reports suggest that Treg are also increased in the blood of dogs with various cancers including melanoma, lymphoma, and osteosarcoma.13–16,32 Although the results of the present study support the notion that circulating Treg are increased in dogs with cancer, conflicting reports exist in the veterinary literature, likely reflecting differences in experimental approach and the small numbers of dogs evaluated to date.33
A key finding of this study is that the administration of CYC at 15 mg/m2/d over the 28-day study period selectively decreases absolute numbers and percent of Treg in the peripheral blood of dogs with STS. Dosages of CYC used in previously published studies ranged from 10 to 25 mg/m2/d.21,22 In two of these studies, metronomic chemotherapy appeared to have antineoplastic effects but the mechanism of action of the drugs used was not investigated.21,22 Our data suggest that decreased numbers of circulating Treg may be one potential mechanism of action of metronomic CYC therapy. Recent work demonstrates decreased levels of intracellular ATP in murine and human Treg as compared with other CD4+ T cells; lower levels of ATP and subsequent decreased glutathione concentrations contribute to low-dose CYC Treg depletion.34 Whether this is the case in Treg in dogs is not known.
Serial changes in tumor MVD were used in this study to assess the antiangiogeneic properties of metronomic CYC therapy. We found that MVD was significantly decreased over the 28-day study period when CYC was administered PO at 15.0 mg/m2/d to dogs with STS. This finding is consistent with previous work in other species that has established the antiangiogenic properties of metronomic dosing of CYC.5,35,36 Our findings should be interpreted with caution due the small number of biopsy samples available for analysis and the knowledge that tumor MVD might not be the most accurate assessment of response to antiangiogenic therapy.37 Inherent in MVD assessment is tumor heterogeneity that could yield in nonrepresentative biopsy samples. In this study nearly half of the biopsies consisted of hemorrhage and necrosis rather than tumor tissue. Additionally, there is some thought that MVD might be more reflective of the metabolic activity of the tumor rather than its angiogenic dependence.37 Combining evaluation of tumor MVD with other markers of angiogenesis such as vascular endothelial growth factor concentrations or enumeration of CECs could provide greater understanding of the antiangiogenic properties of metronomic CYC in dogs with cancer.
Interestingly, significant decreases in both Treg numbers and tumor MVD occurred in dogs with STS treated with CYC at 15.0 mg/m2/d. One explanation for this is that alterations in tumor immunity and angiogenesis occur at a similar dose of CYC. This finding is in line with previous investigations in murine tumor models in which metronomic delivery of drugs such as CYC and paclitaxel was associated with both immunomodulatory and antiangiogenic effects.38 In a previously published study examining the impact of metronomic chemotherapy, dogs with incompletely excised STS received 10 mg/m2/d of CYC.21 Although our preliminary work indicated that decreases in Treg numbers do not occur at this dosage, it is possible that the addition of piroxicam in the Elmslie study had a synergistic effect on Treg depletion. Our data also suggest that decreases in Treg might occur at dosages <15.0 mg/m2/d because of the decrease in the absolute numbers of Treg observed in dogs receiving 12.5 mg/m2/d CYC; larger numbers of dogs will be required to optimize metronomic dosing of CYC. In addition to determining the dosage required to decrease Treg, the duration of CYC administration should also be examined as long-term therapy has been found to induce myeloid-derived suppressor cells and promote tumor growth in murine models.39 Initial tumor burden could also be a factor as Treg depletion might promote tumor growth in more advanced tumors.39,40
Surface staining for CD4 and detection of the intracellular transcription factor FoxP3 were used in this study to identify Treg in dogs by flow cytometry similar to the method use in earlier studies.13–15 In addition to these markers, CD25, the α chain of the IL-2 receptor, is also commonly used to identify Treg in other species. Recently both an anti-human CD25 antibody and a canine anti-CD25 antibody have been found to be useful in the identification of canine Treg.33,41 These antibodies were not available for use in the dogs of the present study; however, future studies should include the use of one of these CD25 antibodies to facilitate identification of CD4+CD25+FoxP3+ Treg.
No adverse events were identified for any of the dogs that completed this study. This result is in contrast to previous studies examining metronomic chemotherapy protocols for canine malignancies as both have reported up to 23% of dogs developed grade 1 or 2 gastrointestinal toxicity and 10–22% of dogs developed sterile hemorrhagic cystitis (SHC) when CYC is administered.21,22 Both of these studies included the use of a NSAID so it is possible that the gastrointestinal signs were secondary to the NSAID rather than the low-dose CYC. A limitation of the current study is that routine monitoring of urine was not performed to assess the presence of microscopic hematuria; however, clinical signs consistent with the development of SHC were not reported in any of the dogs. This could be because of the short duration of CYC administration as dogs receiving CYC therapy at the MTD have a reported incidence of up to 10%.42
In conclusion, we found that dogs with STS have significantly more Treg in circulation than healthy dogs. Additionally, CYC administered at 15.0 mg/m2/d selectively decreased the number and percentage of Treg in the peripheral blood and appears to have antiangiogenic properties in dogs with STS. Because of the small number of dogs in this study, these results must be considered preliminary but support further studies to better characterize the antitumor effects and optimal dosing of low-dose CYC in canine cancer therapy. Additional studies will require longer treatment of a larger number of dogs to gain valuable information regarding potential biomarkers of clinical response to metronomic CYC therapy.
a Serotec, Raleigh, NC
b Mouse FoxP3 staining kit, eBioscience, San Diego, CA
c Cyan ADP flow cytometer, Dako-Cytomation, Fort Collins, CO
d Ted Pella Inc, Redding, CA
e Millipore, Billerica, MA
f Jackson ImmunoResearch Laboratories Inc, West Grove, PA
g Vector Laboratories, Burlingame, CA
h Zeiss, Thornwood, NY
The authors thank Ms Dahlia Rice, Ms Linda Strange, Dr Amanda Guth, and Mr Jesse Harper for their excellent technical assistance as well as Dr Jens Eickhoff for his assistance with statistical analysis.
This study was supported by Grant #D08CA-054 from the Morris Animal Foundation.