Advances in the detection, characterization, and monitoring of cancer in pets

Content sponsored by Blue Buffalo.

Radiograph of canine tibial osteosarcoma / AlexandruPh / stock.adobe.com

Historically, physicians have identified cancer in humans and animals through the appearance of an obvious mass/lump or the development of clinical signs that typically occur when a tumor is advanced (weight loss, vomiting, coughing, etc). Over the past few decades, technologies have improved substantially to identify cancer in earlier stages, dramatically improving the likelihood of cure. In humans, routine mammograms and colonoscopies screen for breast cancer and colon cancer, respectively. For patients with a family history of cancer, routine genetic tests determine whether specific genes (eg, BRCA1) carry mutations that enhance the likelihood the patient will develop cancer, supporting preemptive treatment strategies in some instances. Although such approaches have transformed the practice of human oncology, veterinary oncology is only beginning to implement them. This review will discuss standard and advanced imaging methodologies used in veterinary medicine to identify cancer as well as more recent developments in the realm of genetic testing that have been made available over the past few years.

Diagnostic Imaging

Routine diagnostic imaging, including radiography and ultrasonography, has represented the foundation of patient evaluation in veterinary medicine for more than 40 years because of ease of use, ready application in general practice, and relatively low cost. However, such modalities have inherent limitations for finding cancer, particularly in its earliest stages. For example, standard thoracic radiography will not detect lesions in the lungs until they are at least 7 to 9 mm in size. Similar issues exist with ultrasonography. Size of the patient and gas in the intestinal tract, among other factors, can impair the effectiveness of this approach to detect small lesions. Additionally, because ultrasound is performed in real time, static images may not capture the totality of exam results that typically require 30 to 60 minutes of time.

Increasingly, physicians are employing advanced imaging techniques in veterinary medicine, including CT and MRI. These have a far greater sensitivity for lesion detection (1 to 2 mm in size) and provide a mechanism for more accurate lesion assessment using contrast agents (eg, iodine, gadolinium). Barriers to more widespread use of these modalities include cost to install and maintain, the need for general anesthesia, and cost to pet owners, especially as a screening tool. More recently, advances in CT technologies have permitted scans to be performed under sedation, reducing risk and lowering costs, thereby enhancing use and application. However, access remains a challenge because CT machines are usually located at specialty practices and are most often used for patients who have an active illness rather than as a screening mechanism.

Within the context of cancer detection, the use of positron emission tomography (PET) can find small lesions in locations that are difficult to image using standard approaches. PET works as a functional imaging technique that employs radioactive substances known as radiotracers to visualize and measure changes in specific metabolic processes within cells. Whole-body images after tracer administration show sites of radiotracer accumulation. The most common radiotracer in use is 18F-fluorodeoxyglucose (18F-FDG), a radiolabeled glucose molecule. Because tumor cells often exhibit enhanced glucose metabolism, the 18F-FDG is preferentially taken up by tumor cells in the body, thereby providing a mechanism to identify small lesions within the context of a whole-body scan (i.e., very small tumors will concentrate the 18F-FDG making them easy to identify). Several other tracer agents have capacities to detect specific tumor types through their preferential expression of specific proteins, and these have helped to enhance accuracy of the approach. A CT scanner is often linked to the PET platform, permitting dual imaging in 1 session. This facilitates more precise anatomic localization of lesions found on the PET scan. Although the use of PET-CT has expanded in human oncology, several barriers hinder its widespread use in veterinary medicine. These barriers include cost and proximity to a cyclotron, the particle accelerator device necessary for generation of the radioisotopes necessary to make the radiotracers. Finally, several states have substantial regulations regarding the use of radioactivity in dogs, necessitating containment post imaging until sufficient decay of the radioactive isotope has occurred. Together, these barriers have constrained the use of PET-CT to a relatively limited number of academic institutions, thereby limiting its implementation in veterinary medicine.

Molecular and Genetic Approaches

Liquid biopsy

Although imaging approaches have undergone considerable advances, perhaps the most rapid progress in cancer detection and monitoring has been in the realm of what is termed liquid or blood biopsy. This is a minimally invasive tool that relies on the fact that dead and dying cells release DNA and RNA, among other components, into the bloodstream and that other body fluids (eg, urine) are detectable using newer genetic methods. Most efforts have focused on identifying circulating tumor DNA (ctDNA) from the broader pool of overall cell-free DNA (cfDNA) that results from general cellular turnover. Analysis of ctDNA uses a technique called ultralow-pass whole genome sequencing (ULPWGS). This technique calculates the total fraction of ctDNA in the cfDNA (known as tumor fraction). If this level is high enough (typically greater than 10%), it can be used for more advanced assays, such as in-depth sequencing of whole exomes (the DNA that codes for proteins). However, the current limit of detection for ULPWGS is a tumor fraction of 3% to 5%. As such, it is challenging to use this approach for early detection of cancer when the tumor fraction is quite low.

In human oncology, physicians generally use FDA-approved liquid biopsy diagnostics as an adjunct to standard diagnostic methodologies rather than as a screening tool specifically. These diagnostics include the Guardant360 CDx and FoundationOne Liquid CDx. The tests rely on an approach called exome capture. In this setting, a defined list of genes known to be involved with specific human cancers (eg, melanoma, lung cancer) is interrogated for mutations found in these cancers, using ctDNA level in the bloodstream as the template. As noted, previously, challenges remain with respect to sensitivity for early detection of cancer and identifying early recurrence of cancer post treatment. Additionally, the exome capture panel methodology does not work well for tumors in which specific gene mutations are uncommon, which is the case in most sarcomas.

Sensitivity of detection can be improved using a droplet digital polymerase chain reaction (ddPCR), but this method requires preexisting knowledge of mutations likely to be present in the tumor. In settings such as canine transitional cell carcinoma of the bladder, ddPCR works well. Indeed, the CADET test for canine bladder cancer leverages the fact that a mutation in the BRAF gene, which results in an amino acid switch from valine to glutamic acid at position 595, occurs in more than 80% of dogs, facilitating the application of ddPCR. In the setting of canine oncology, commercial canine blood-based biopsy tests aid in the diagnosis of cancer in dogs. One of these tests, OncoK9 from PetDx, incorporates ctDNA/tumor fraction detection followed by ULPWGS with a limited analysis of specific genes (eg, TP53). However, as with current human liquid biopsy platforms, it has a relatively low sensitivity for identifying cancers earlier in the course of disease (20% for detecting cancers less than 5 cm in size). Additionally, although some tumor types such as lymphoma and hemangiosarcoma are relatively easy to find using liquid biopsy, substantial variation in tumor fraction associated with other cancers, such as in mast cell tumors, even when disease burden is high, emphasizes current challenges with effective implementation of this tool as a reliable method for early detection.

Another blood-based test available in the veterinary space involves the detection of circulating nucleosomes, small fragments of chromosomes composed of DNA wrapped around a histone core made of 4 duplicate histone proteins forming an octamer. The test, termed the Nu.Q assay, is made by Volition Veterinary and uses an enzyme linked immunosorbent assay that contains a capture antibody directed at histone 3.1 and a nucleosome-specific detection antibody. Like the PetDx test, the Nu.Q assay has high specificity for certain canine cancers but a lower overall sensitivity (less than 50%) across all cancer types. Specifically, it performs better for the detection of lymphoma, hemangiosarcoma, histiocytic sarcoma, and malignant melanoma but is less likely to detect soft tissue sarcomas, osteosarcoma, and mast cell tumors.

Given the current limitations regarding the sensitivity of blood-based diagnostics for cancer, efforts are underway to expand the capacities of liquid biopsy by employing a multiomics approach integrating information from circulating DNA, methylation patterns in the ctDNA, circulating RNA, and other sources. Building upon this, studies are ongoing in human patients at risk of lung cancer to use such approaches as a screening tool and compare the sensitivity with CT scan–based screens.

Genetic testing

Another approach to characterizing patient cancers to confirm tumor origin/type and assist with treatment recommendations involves direct genetic analysis of the tumors after removal. If tumor specimens are collected and frozen, they are subject to whole genome sequencing, in which the entire coding and noncoding DNA is sequenced, or whole exome sequencing, in which the entire coding DNA (representing about 1% of the total DNA) is sequenced. Both processes identify mutations in the tumor DNA that may change the expression or function of a particular gene ultimately driving tumor cell growth. Additionally, knowledge regarding these mutations can help identify new therapeutic strategies. With respect to commercially available mutation tests for canine cancers, platforms are currently available through FidoCure and Vidium. Both platforms use a curated exome capture panel to analyze 50 to 60 specific genes commonly showing mutation across a variety of canine cancers. FidoCure also pairs mutations found in specific cancers with small molecule inhibitors to target pathways disrupted by the changes. Unfortunately, these assays are only available for dogs, with none yet developed for cats or other species.

Although characterizing the mutation landscape of cancer is important, understanding what germline genetic characteristics may contribute to a higher likelihood of contracting certain types of cancer is equally valuable. As previously mentioned, such genetic tests are widely used in patients with known familial histories of cancer (eg, the test for BRAC1 mutation). The canine genome has recently undergone a more thorough characterization and annotation, and efforts are underway to better define the normal variants across breeds. These can be used as a baseline to begin identifying specific DNA changes linked to cancers that occur with high incidence in certain breeds, such as T-cell lymphoma in boxers, transitional cell carcinoma in West Highland white terriers, and hemangiosarcoma in German shepherds. A better understanding of these heritable factors can then be used to create genetic tests for breed-specific mutation screening, with the goal of eliminating those carrying the cancer risk from the breeding pool.

Future directions

Technology-driven progress has catalyzed broad innovation in the realm of cancer detection and monitoring. Several new tools marry digital pathology advances with multiomics analysis, in which information about protein and gene expression in tumor cells and the surrounding microenvironment can be obtained from formalin-fixed paraffin-embedded biopsy specimens on a single-cell level. Such data can be analyzed using machine-learning algorithms to better define the exact pathways that are dysregulated in tumor cells and, more importantly, help design more effective treatment interventions. Although such innovations are only broadly available for human tumor specimens, capacities are growing to perform similar studies using canine tumors. Therefore, veterinary medicine will likely gain the necessary toolbox for undertaking similar studies, at least within the companion animal setting, over the next decade.