This study was conducted at the Comparative Oncology Research Laboratory and Cancer Care Clinic, Department of Veterinary Clinical Medicine, University of Illinois at Urbana-Champaign, Urbana, IL. Findings of this study were presented in part at the 27th Annual Veterinary Cancer Society Conference, Fort Lauderdale, FL, 2007.
Corresponding author: Timothy M. Fan, DVM, PhD, Department of Veterinary Clinical Medicine, University of Illinois at Urbana-Champaign, Urbana, IL 61802; e-mail: email@example.com.
Background: Various bone resorption markers in humans are useful for supporting the diagnosis of malignant skeletal pathology, with certain bone resorption markers appearing to be more discriminatory for detecting cancer-induced osteolysis than others. Canine osteosarcoma (OSA) is characterized by focal bone destruction, but a systematic investigation for determining which bone resorption marker best supports the diagnosis of OSA in dogs has not been reported.
Hypothesis: Dogs with OSA will have increased concentrations of bone resorption markers compared with healthy dogs and dogs with orthopedic disorders. Differences will exist among various bone resorption markers for their ability to support the diagnosis of malignant osteolysis in dogs with OSA.
Animals: Single time point, cross-sectional, cohort study including dogs with OSA (n = 20) or orthopedic disorders (n = 20) and healthy dogs (n = 22).
Methods: Basal concentrations of urine and serum N-telopeptide (NTx), urine and serum C-telopeptide (CTx), and urine deoxypyridinoline (DPD) were compared among all 3 groups.
Results: Compared with healthy dogs and dogs with orthopedic disorders, urine NTx, serum NTx, and serum CTx concentrations were significantly increased in dogs with OSA. For urine NTx and serum NTx, the calculated lower and upper 95% confidence limits in dogs with OSA did not overlap with dogs diagnosed with orthopedic disorders or healthy dogs.
Conclusions and clinical importance: Of the markers evaluated in this study, urine NTx and serum NTx appear to be the most discriminatory resorption markers supporting the diagnosis of focal malignant osteolysis in dogs with OSA.
Homeostatic bone remodeling is a dynamic and continual process whereby skeletal mass and health are maintained by the coordinated balance of bone formation and bone resorption. The cellular components responsible for skeletal equilibrium are osteoclasts that function to degrade aged or diseased bone and osteoblasts that deposit osteoid, which subsequently mineralizes to new bone. Under normal physiologic conditions, osteoclast and osteoblast activities are tightly coupled with one another and form a functional element known as the basic multicellular unit.1 Pathologic bone disorders such as osteoporosis and skeletal neoplasms perturb finely balanced bone turnover, often favoring excessive bone resorption.2 Because mineralized bone is composed of approximately 90% type I collagen, pathologic skeletal resorption resulting from dysregulated osteoclast activity leads to proteolysis and release of type I collagen major cross-links such as deoxypyridinoline (DPD). In addition, specific collagen type I epitopes, including amino-(N-telopeptide [NTx]) and carboxy-(C-telopeptide [CTx]) terminal peptides, which hold collagen major cross-links together, also are released during bone degradation. Collectively, these cleavage products of type I collagen resulting from osteoclast activities are liberated into the bloodstream and ultimately are eliminated by the kidneys. Measurement of these bone resorption markers serves as a noninvasive method for assessing dynamic changes in bone metabolism.2–6
In humans and animals with pathologic skeletal conditions, circulating concentrations of bone resorption markers are increased.7–14 Common markers of bone resorption measured in humans with pathologic osteolytic diseases include urine NTx, urine CTx, urine DPD, serum NTx, and serum CTx.6,8,15 Although useful for supporting the diagnosis of dysregulated bone metabolism, some bone resorption markers also can be used to assess treatment responses. Bone resorption markers reflect bone metabolic changes in real time, and they have been shown in cancer patients to predict progression of skeletal metastases before radiographic changes are evident.16 In humans with skeletal metastases treated with antineoplastic or aminobisphosphonate therapies, the magnitude of reduction in bone resorption markers correlates with treatment effectiveness for reducing bone pain and tumor burden.2,10,17,18
Despite the widely documented diagnostic and therapeutic monitoring use of several urine and serum bone resorption markers in humans with skeletal metastases, comparative studies in companion animals with malignant osteolysis remain limited and incomplete.9,12,19 Although the utility of a solitary bone resorption marker, urine NTx, has been described in dogs with appendicular osteosarcoma (OSA),9,12,19 the contemporaneous examination of multiple bone resorption markers in dogs to determine differences in their ability for discriminating dogs with OSA from healthy dogs and those with orthopedic disorders has not previously been conducted. Therefore, the purpose of this study was to compare 5 bone resorption markers in 3 populations of age- and weight-matched dogs considered healthy or diagnosed with OSA or orthopedic disorders.
Materials and Methods
Twenty dogs with OSA were evaluated between November 2004 and June 2007. All dogs had a confirmed diagnosis of OSA by either histopathology or alkaline phosphatase positive staining concurrent with cytologic criteria of malignancy.20,21 All patients were clinically staged, which included a CBC, serum biochemical profile, urinalysis, primary tumor and thoracic radiographs, and abdominal ultrasound examination. Additional diagnostic tests such as urine culture or lymph node aspirates were performed at the discretion of the attending clinician. The population of dogs with OSA consisted of 6 male neutered dogs, 1 intact male dog, and 13 female spayed dogs. The mean age and weight were 7.8 years (range, 4–12 years) and 43.8 kg (range, 24.8–64.5 kg), respectively (Table 1).
Twenty-two healthy, geriatric dogs owned by faculty, staff, and students of the Veterinary Teaching Hospital served as normal controls. All dogs weighed >25 kg and were >6 years of age. Control dogs were considered to be in good health based on history, physical examination, CBCm and serum biochemical profile results. None of the dogs in the control population had any history of osteoarthritis or previously documented orthopedic disease (Table 1).
Twenty dogs with radiographically confirmed orthopedic disorders with lameness or osteoarthritis requiring interventional surgery were prospectively evaluated between June 2005 and January 2007. All dogs with orthopedic disorders were otherwise in good health as determined by physical examinations, limb and thoracic radiographs, CBC, serum biochemistry profile, and urinalysis. Dogs with orthopedic disorders included in this study were >25 kg in weight and >6 years of age (Table 1).
Basal Urine NTx, CTx, or DPD Measurement
On initial presentation, urine was collected from all dogs in the morning by either cystocentesis or voiding. Urine samples were immediately centrifuged at 4°C at 1500 rpm for 10 minutes, and the supernatant was collected and stored at −20°C in 2 mL polypropylene cryovials until analysis was performed. Urine NTx, CTx, and DPD concentrations were measured with commercial ELISAa,b,c test kits, all previously validated for use in the dog.7,22,23 Urine NTx was expressed as normalized nanomolar bone collagen equivalents (BCE) per millimole concentration of urine creatinine. Urine CTx was corrected for creatinine and expressed as micrograms of CTx per millimole urine creatinine. Urine DPD was expressed as nanomoles urine DPD per millimole of urine creatinine.
Basal Serum NTx and CTx Measurement
Upon initial presentation, blood samples were collected by jugular venipuncture from all dogs. Whole blood samples were centrifuged for 10 minutes at 2,000 rpm, and serum was separated and stored at −20°C in 2 mL polypropylene cryovials until analysis. Serum NTx and CTx concentrations were measured by commercially available immunoassays,d,e previously validated for use in the dog.23,24
Among the 3 groups of dogs, differences in each individual bone resorption marker were analyzed using 1-way ANOVA and posthoc comparisons were made with a Tukey-Kramer multiple comparisons test. For each bone resorption marker, 95% confidence intervals were calculated for all 3 groups of dogs. All values were reported as mean ± standard deviation and range. Statistical analysis was performed by commercial computer software.f Significance was defined as P < .05.
Basal Urine NTx Concentrations
In dogs with OSA, basal urine NTx concentrations averaged 427.2 ± 215.1 (range, 191.7–986.3) nM BCE/mM creatinine. In dogs with orthopedic disease and control dogs, basal urine NTx concentrations were significantly lower than in OSA dogs, at 207.8 ± 84.7 (range, 95.5–426.7) and 164.6 ± 73.3 (range, 29.2–289.8) nM BCE/mM creatinine, respectively (Fig 1A, P < .0001). Basal concentrations of urine NTx were not statistically different between dogs with orthopedic disease and control dogs (P > .05). The lower and upper 95% confidence limits for OSA dogs (326.5 and 527.9 nM BCE/mM creatinine, respectively), did not overlap results from dogs with orthopedic disease (168.2 and 247.5 nM BCE/mM creatinine, respectively) or those of control dogs (132.1 and 197.1 nM BCE/mM creatinine, respectively).
Basal Urine CTx Concentrations
In dogs with OSA, baseline urine CTx concentrations were 2.9 ± 1.3 (range, 1.1–6.4) μg/mM creatinine. In dogs with orthopedic disorders and control dogs, basal urine CTx concentrations were 3.3 ± 1.4 (range, 1.5–6.2) μg/mM creatinine and 2.4 ± 1.5 (range, 0.6–7.7) μg/mM creatinine, respectively. No statistical difference in basal urine CTx concentrations was identified among the 3 groups of dogs (Fig 1B, P= .11).
Basal Urine DPD Concentrations
In dogs with OSA, baseline urine DPD concentrations were 4.8 ± 1.3 (range, 2.0–6.6) nM DPD/mM creatinine. In dogs with orthopedic disorders and control dogs, basal urine DPD concentrations were 4.5 ± 2.0 (range, 2.3–9.8) and 3.1 ± 0.9 (range, 2.1–5.1) nM DPD/mM creatinine, respectively. Normal control dogs had significantly lower concentrations of urine DPD than did dogs with OSA and those with orthopedic disorders (Fig 1C, P < .01). No statistical difference in urine DPD concentrations was identified between dogs with OSA and dogs with orthopedic disorders, P > .05.
Basal Serum NTx Concentrations
In dogs with OSA, basal serum NTx concentrations were 61.4 ± 30.6 (range, 21.2–117.5) nM BCE, whereas in dogs with orthopedic disorders and normal dogs, serum NTx concentrations were 28.3 ± 13.0 (range, 12.2–61.1) nM BCE and 26.2 ± 13.8 (range, 8.3–64.4) nM BCE, respectively. Basal serum NTx concentrations in dogs with OSA were statistically higher than those in dogs with orthopedic disorders and in control dogs (Fig 1D, P < .001). No statistical difference in serum NTx concentrations was identified between dogs with orthopedic disorders and control dogs, (P > .05). The lower and upper 95% confidence limits for OSA dogs (47.0 and 75.7 nM BCE, respectively) did not overlap with those of dogs with orthopedic disease (22.2 and 34.4 nM BCE, respectively) or those of control dogs (20.1 and 32.3 nM BCE, respectively).
Basal Serum CTx Concentrations
In dogs with OSA, basal serum CTx concentrations were 454.7 ± 189.4 (range, 203.7–808.8) pg/mL. In dogs with orthopedic disorders and normal dogs, serum CTx concentrations were 359.6 ± 40.8 (range, 282.4–434.7) pg/mL and 324.7 ± 47.3 (range, 270.2–448.8) pg/mL, respectively. Basal serum CTx concentrations in dogs with OSA were statistically higher than those in dogs with orthopedic disorders and control dogs (Fig 1E, P < .01). No statistical difference in serum CTx concentrations was identified between dogs with orthopedic disorders and control dogs (P > .05). Although basal serum CTx concentrations were significantly increased in dogs with OSA, the lower and upper 95% confidence limits for OSA dogs (366.1 and 543.4 pg/mL, respectively) did overlap those of dogs with orthopedic disease (340.5 and 378.7 pg/mL, respectively).
Although bone resorption markers universally quantify the breakdown products of type I collagen, our study findings suggest that their discriminatory capacity for supporting the diagnosis of focal malignant osteolysis in dogs with OSA is not equivalent. Of the 5 resorption markers studied, only urine NTx, serum NTx, and serum CTx were increased in dogs with OSA, but not in age- and weight-matched healthy control dogs or in dogs with confirmed orthopedic lameness. In addition, only urine NTx and serum NTx concentrations derived from dogs with OSA generated nonoverlapping 95% confidence intervals compared with dogs with orthopedic disorders and healthy dogs. Based on these findings, urine NTx and serum NTx appear to be the most discriminatory bone resorption markers for supporting the diagnosis of focal malignant osteolysis associated with appendicular OSA.
The documented increase in urine NTx, serum NTx, and serum CTx concentrations in dogs with OSA most likely reflects ongoing focal malignant osteolysis in these clinical patients, a pathologic process that is presumed to be absent in normal dogs and in dogs with osteoarthritic diseases. The findings from this study are in agreement with those of a previous investigation that demonstrated increases of a solitary bone resorption marker, urine NTx, in dogs with OSA in comparison with normal, healthy control dogs.12 However, the current study provides additional information, because not only urine NTx but also serum NTx and serum CTx were determined to support the diagnosis of focal malignant osteolysis. Furthermore, basal concentrations of these 3 specific bone resorption markers were different between dogs with OSA and dogs with orthopedic lameness, a finding that has not been reported previously.
Interestingly, not every bone resorption marker evaluated in this study was capable of differentiating focal malignant osteolysis from either orthopedic disease or homeostatic bone turnover. No difference in basal urine CTx excretion could be identified among the 3 groups of dogs in this study, an unexpected finding given that CTx is a specific breakdown product of collagen type I, the major protein constituent of bone. Another unanticipated result was the similar urine DPD excretion identified in dogs with OSA and in those with orthopedic disorders. Given that chronic osteoarthritis results in articular cartilage breakdown, it would be predicted that orthopedically lame dogs may have increases in collagen type II metabolic products, but not DPD, which is a type I collagen major cross-link.
The identification of noninvasive, supportive diagnostic tests that aid in discriminating between the 2 most common causes of lameness (focal malignant osteolysis and osteoarthritis) in large breed, geriatric dogs may be of clinical importance to veterinarians and pet owners alike. Based upon calculated lower and upper 95% confidence limits, our findings suggest that quantifying basal concentrations of either urine NTx or serum NTx may help to differentiate malignant (OSA) from benign (osteoarthritis) causes of lameness in dogs. Already in humans, the practical assessment of urine NTx in patients being treated for osteoporosis has been facilitated through the development of disposable, handheld, devicesg for routine clinical use. With available technology, it may be possible to quickly and economically evaluate basal urine NTx concentrations in dogs with undetermined causes of lameness.
Although this study provides new information regarding the potential utility of 5 different bone resorption markers in dogs, several limitations should be highlighted. First, the sample sizes for the 3 groups were relatively small, but attempts were made to limit variations among groups, including age (skeletal maturity) and weight (skeletal mass). In addition, the orthopedic disorders that comprised one of our groups were restricted primarily to nonerosive arthropathies, usually cranial cruciate ligament rupture, which is not osteolytic in nature, not generally a primary differential diagnosis for focal malignant osteolytic processes, and highly variable in bone metabolism and clinical progression. As such, it is not surprising that for the most part, bone resorptive markers were not increased in this experimental group. Second, although some bone resorption markers were increased in dogs diagnosed with OSA, this patient population had advanced disease as demonstrated by radiographically evident osteolysis. It remains undetermined if bone resorption marker concentrations would provide similar differential information in dogs with smaller tumor burdens or lesser degrees of focal malignant osteolysis. Third, this investigation evaluated only osteolytic markers and did not assess counter-regulatory osteoblastic activities, both homeostatic and reparative, likely to be operative in the 3 groups of dogs studied. Although the prognostic value of bone alkaline phosphatase (bALP) activities in dogs with OSA has been described previously,25,26 no study has characterized additional bone formation markers such as osteocalcin (OC), procollagen I N-peptide (PINP), and procollagen I C-peptide (PICP). Given that bALP, OC, PINP, and PICP are proteins produced by osteoblasts, their concentrations may be dysregulated in dogs with OSA, as they are in human cancer patients.13,27–30 Fourth, it remains undetermined why discrepancies exist between serum and urine CTx concentrations in the 3 groups of dogs evaluated. Specifically, serum CTx was clearly increased in dogs with OSA, but its urine equivalent failed to differentiate dogs with OSA from normal controls and dogs with orthopedic disorders. Intuitively, bone resorption markers in the serum and urine should provide similar information regarding skeletal health, but this did not appear to be the case for all of the bone resorption markers evaluated in this study.
Because the clinical need to provide effective and durable pain relief for dogs suffering from malignant osteolysis and associated cancer pain increases, it will be necessary to validate and study surrogate biochemical markers of skeletal health. Findings from this study demonstrate that considerable variability exists among bone resorption markers for supporting the diagnosis of focal malignant osteolysis. Ultimately, information derived from similar studies may be useful not only for aiding the diagnosis of skeletal malignancy but also for assessing the effectiveness of novel palliative treatments.
The authors would like to thank Nancy George, Jenny Rose, Rebecca Moss, Carrie Bubb, and Kim Knap for their technical assistance in collection of blood and urine samples for patients included in this study.
aOsteomark NTx urine, Ostex International Inc, Seattle, WA