Received: 31 January 2022 Revised: 5 July 2022 Accepted: 7 July 2022

DOI: 10.1111/vco.12848

REVIEW

Blood biomarkers for canine cancer, from human to veterinary

oncology

Philippe Colombe1,2 | Jérémy Béguin2,3 | Ghita Benchekroun2,4 |

Delphine Le Roux1,5

1
Ecole Nationale Vétérinaire d'Alfort, BioPôle
Alfort, Maisons-Alfort, France Abstract
2
Ecole Nationale Vétérinaire d'Alfort, CHUVA, In recent decades, interest in circulating tumour biomarkers is increasing both in
Service de Médecine Interne, Maisons-Alfort,
human and veterinary oncology. An ideal tumour biomarker would allow early diag-
France
3
Anses, INRAE, Ecole Nationale Vétérinaire nosis of neoplasia, identify it specifically, accurately, establish a prognosis and predict
d'Alfort, UMR VIROLOGIE, Laboratoire de its behaviour, especially regarding different therapeutic solutions. It would also allow
Santé Animale, Maisons-Alfort, France
4
to monitor its evolution over time and all this in a non-invasive and inexpensive way.
Ecole nationale Vétérinaire d'Alfort, Univ
Paris Est Créteil, INSERM, IMRB, Maisons- Actually, no biomarkers meeting all of these criteria have been identified in veterinary
Alfort, France
medicine, particularly due to a lack of specificity of the main protein tumour bio-
5
Anses, INRAE, Ecole Nationale Vétérinaire
d'Alfort, UMR BIPAR, Laboratoire de Santé
markers studied to date. However, great hope is currently placed in biomarkers
Animale, Maisons-Alfort, France grouped under the name of liquid biopsy, which could prove to be effective tools for

Correspondence
common clinical use in the near future. This review gives an update on blood cancer
Delphine Le Roux, Ecole Nationale Vétérinaire biomarkers studied in dogs, such as ions, proteins, nucleic acids and also circulating
d'Alfort, 7 avenue du Général de Gaulle,
94700 Maisons-Alfort, France.
cells, of which some might become more prominent in the coming years to help
Email: [email protected] improve the management of animal care.

KEYWORDS
biomarker, cancer, liquid biopsy, veterinary

1 | I N T RO DU CT I O N sampling procedure is minimally invasive and the measurement of one

(or more) component is often quick. In contrast, tumour biopsies are
The use of biomarkers represents a major challenge in the field of more invasive and often take longer to analyse and interpret. More-
oncology in both human and veterinary medicine. The growing knowl- over, in veterinary medicine, tumour biopsies, by requiring sedation or
edge of biomarkers over the last two decades has enabled better general anaesthesia, increase the owner's expenses as well as expo-
management of patients with tumours. Ideally a tumour biomarker sure of the animal to additional risks such as dissemination of tumour
would be easy to measure, should have perfect sensitivity and speci- cells, enlargement of the surgical site, and possible post-anaesthetic
ficity, with the aim of perfectly differentiating individuals suffering complications. Furthermore, unlike the majority of tumour biopsies,
from a neoplastic process from healthy individuals, and at the earliest blood sampling can be repeated at close intervals, which allows moni-
possible stages. It should also be able to identify a subject affected by toring of parameters over time and therefore assessment of potential
any tumour process, allow monitoring of the evolution of cancer over changes in cancer behaviour, with or without treatment. There are
time and forecasting of relapses or recurrences.1 many types of blood biomarkers: ions, proteins, carbohydrates, lipids,
The vascular compartment is easily accessible, so blood is a sam- nucleic acids and also cells. However, none of them is yet commonly
ple of choice to identify and measure tumour biomarkers. Indeed, the used for diagnostic purposes in veterinary medicine. Instead of

This is an open access article under the terms of the Creative Commons Attribution License, which permits use, distribution and reproduction in any medium,
provided the original work is properly cited.
© 2022 The Authors. Veterinary and Comparative Oncology published by John Wiley & Sons Ltd.

Vet Comp Oncol. 2022;20:767–777. wileyonlinelibrary.com/journal/vco 767

768 COLOMBE ET AL.

focusing on one single category of biomarker or all biomarkers in one further studied in other tumour types or with bigger dog cohorts. This
single form of cancer, this review provides an overview of the scien- could be attributed to the fact that dietary intake may influence Zn2+
tific advances concerning selected blood parameters of various type, serum levels, which could be hard to monitor in animals. Yet, a lot of
that have shown potential interest as biomarkers for different canine studies in human oncology have shown a significant decrease in
cancers, inspired from extensive studies in human oncology, the latter serum Zn2+ concentration in many cancers and some also suggest the
being excluded from this manuscript as reviewed elsewhere.2 potential benefit of Zn2+ supplementation during therapy.10–14 This
should justify and motivate more research on Zn2+ blood levels in the
veterinary oncology field since it is now easily measurable in serum
2 | IONS AS BIOMARKERS with a colorimetric assay.12

2.1 | The copper ion and copper isotopes

3 | PROTEINS AS BIOMARKERS
Copper ion (Cu2+) is an essential micronutrient, with involvement in
many fundamental physiological processes. In humans, copper has 3.1 | The carcino-embryonic antigen
also been involved in the processes of tumorigenesis, angiogenesis,
tumour migration and metastasis development3–6 but only a few ani- The carcino-embryogenic antigen (CEA) is a membrane glycoprotein,
mal studies have focused on this topic. However, one study demon- mainly synthesised in certain portions of the digestive tract and plays
strated a two-fold increase in Cu2+ concentrations in bitches with a role in cell adhesion. CEA blood concentration is very low under
7
mammary tumours of various types compared with healthy bitches. physiological conditions but some tumour cells can produce this pro-
Unfortunately, the small size of this study and the lack of comparison tein in large quantities and express it over the entire membrane sur-
of nutrient intake, precludes for the moment, the possibility to use face when losing their polarity. As these cells no longer rest on a basal
serum Cu2+ as a diagnostic biomarker for mammary tumours in dogs. lamina, CEA can be found in high concentrations in the blood of
New mass spectrometry methods such as multi-collector induc- human patients with certain tumours.15 In healthy dogs, serum CEA
tively coupled plasma-mass spectometry (MC-ICP-MS), allowed the concentrations reference values were initially set using radio-immuno
measurement of copper isotopes 65Cu/63Cu ratio (δCu) in dogs' blood. assays, ranging from 0.12 to 0.23 ng/ml,16,17 and interestingly, bitches
Blood δCu measured in dogs with neoplastic or non-neoplastic dis- with mammary tumours presented CEA serum concentration above
ease was significantly lower compared with healthy dogs. However, the reference values. At the 0.23 ng/ml cut-off concentration, the
δCu levels between neoplastic and non-neoplastic dogs were not sig- sensibility and specificity of this assay was of 60% et 95% respectively
nificantly different. In addition, following a chemotherapy protocol for for tumour diagnosis in dogs.17 Other studies evaluated the difference
lymphoma, an increase of the δCu was seen in five out of six dogs in in CEA expression between female dogs with mammary tumours and
8
clinical remission. Interestingly, to measure these copper ions, whole healthy dogs using ELISA kits available for human CEA. Two studies
blood can be frozen at
20 C and processed once defrosted, to purify confirmed a significant increase in serum CEA concentration in
copper elements before performing MC-ICP-MS, which is therefore a bitches with mammary tumours compared with healthy bitches18,19
technical advantage to investigate larger cohorts. Indeed, to show a whereas a previous one showed no correlation of the CEA serum
possible interest of the δCu in monitoring the response to chemother- levels with the staging of the tumour.20 Interestingly, Senhorello et al.,
apy treatment in dogs, these larger studies are now required, with a observed a significant increase in CEA levels in bitches with tumours
focus on specific neoplastic conditions to identify those that may pre- greater than 3 cm in size compared with tumours less than 3 cm, and
sent significant isotope variation compared with other non-neoplastic a significant increase in CEA was noticed for grade III tumours com-
conditions. pared with grade I and II tumours. Based on the Receiver Operating
Characteristic (ROC) curve performed using 1.08 ng/ml as the cut-off
value, they established the sensitivity and specificity of serum CEA,
2.2 | The zinc ion measured by ELISA, to detect the presence of mammary tumour at
82.14% and 95.24%, respectively. Sensitivity for this threshold value
Only two published studies have focused on measuring Zinc ion was improved to 100% for tumours larger than 3 cm and metastasised
(Zn2+) serum concentration in neoplastic and healthy dogs. A signifi- tumours but down to 70% for tumours smaller than 3 cm.19 Taken
cant decrease in Zn2+ serum concentrations was observed in 50 dogs together, these studies confirmed that CEA is a relatively sensitive
diagnosed with IIIa or IVa lymphoma (classification according to the and specific diagnostic biomarker for canine mammary tumours, espe-
World Health Organisation's) compared with 50 healthy dogs. This cially in more advanced stages, which, however, might not allow CEA
decrease was also seen in dogs suffering from osteosarcoma but the to be a biomarker for early diagnostic of breast tumour. The use of
difference with the healthy dogs was not significant.9 A significant human ELISA kits, commercially available to measure CEA in dog
2+
decrease in serum Zn concentration was also observed in bitches blood, made these studies possible but more recently, canine specific
with mammary tumours compared with healthy bitches.7 Since these kits are also available, which is a great advantage to acknowledge CEA
two studies from the past decades, Zn2+ serum levels have not been as an easily accessible canine cancer biomarker. These more recent
COLOMBE ET AL. 769

studies are in agreement with what had been shown previously.21,22 major proteins in foetal circulation, however, its expression is tran-
Using canine specific CEA ELISA kits, lower CEA levels were observed scriptionally repressed at birth. AFP has been shown to have an
15 days after surgery on 41 dogs but this difference was not signifi- immunosuppressive activity and also pro-angiogenic properties that
cant.21 Interestingly, with human CEA ELISA kits, the reduced CEA promote neovascularisation in foetal and tumour tissues.38–40 Consid-
serum levels were significantly different 15 and 45 days post-masectomy ering that AFP is mainly produced by liver, veterinary studies have
as compared with before surgery in 11 female dogs.19 In terms of prog- focused on this protein as a biomarker for canine liver cancer.
nostic biomarkers, they also show that the higher the initial serum CEA At birth, serum AFP concentration in dogs is high:
levels in bitches, the shorter the survival time is,21 but longer-term stud- 14080 ± 5944 μg/ml, then a strong decrease during growth is
ies involving a larger number of bitches are required to assert this. observed until reaching 0.014 to 0.069 μg/ml in adult dogs.41 In vitro,
AFP can be produced by canine hepatocellular carcinoma cells,
suggesting that this production could be linked to the neoplastic pro-
3.2 | The carbohydrate antigen 15-3 cess in dogs.42 This observation had been confirmed by Marchesi
et al. 2007, in which, no significant differences were noted in the
The Carbohydrate Antigen 15–3 (CA15-3) is transmembrane glycopro- serum AFP concentrations between healthy dogs and those with
tein belonging to the mucin family produced by the MUC-1 gene. The non-neoplastic diseases but dogs with lymphomas and mastocytomas
products of the MUC-1 gene are involved in carcinogenesis as they par- showed an increase in serum AFP concentrations. However, AFP
ticipate to immunosuppression,23 promote the proliferation and survival concentrations in serum of dogs with sarcomas and carcinomas
of tumour cells,24 and also contribute to their dissemination.25 CA15-3 were not significantly different from healthy dogs.34 A high AFP
concentration is increased in blood when tissues are altered, however, serum concentration cut-off value could indicate a high probability of
this is not specific to a particular neoplastic disease and can be found in neoplastic process.43,44 but to date, no studies have proposed a
26–31 32 33
different human cancers as well as in hepatitis and arthritis. cut-off value in dogs defining a sensitivity and specificity value for
In veterinary oncology, several studies have found a significant this biomarker. Moreover, no studies have been conducted on
increase in serum CA15-3 concentration in female dogs suffering from hepatocellular carcinomas in dogs to establish a correlation between
canine mammary tumours compared with healthy ones.17,18,20–22,34–36 serum AFP concentration and histological characteristics, tumour
The sensitivity and specificity of CA15-3 to detect mammary neoplastic size, or tumour stage. Despite a significant decrease in serum AFP
damage at a threshold value of 7 IU/ml, was 100% and 95% respec- concentrations observed following surgical removal of affected
17
tively. However, a more recent study exhibited a sensitivity of liver lobes,43,44 the available evidence is not strong enough yet to
CA15-3 to 51.8% in the malignant group but with a specificity close to consider AFP serum concentrations as a good follow-up biomarker of
previously described (93.9%). More interestingly, combining CA15-3 to individuals with hepatocellular carcinoma.
CEA for instance, increased sensitivity to 64.2%, but combining these
two biomarkers reduced specificity down to 81.7%.22 In addition, a sig-
nificant positive correlation between increasing serum CA15-3 concen- 3.4 | Lactade dehydrogenase
trations and advancing stages of tumour has been observed.20
However, the preliminary study from 2007 by Marchesi et al., showed Rapid cancer cell proliferation and high metabolic demands lead to an
no significant difference in serum CA15-3 concentrations between increase in lactate dehydrogenase (LDH), which catalyses the revers-
dogs with different types of neoplastic processes and healthy dogs.34 ible transformation of pyruvate into lactate and is well established in
This lack of significance could be explained, as in humans, by the fact human medicine as a prognosis and follow-up biomarker in many can-
that not all tumour types are associated with an elevated CA15-3 pro- cers.45 It has been shown in veterinary medicine that LDH is signifi-
duction. Finally, as for CEA, low serum CA15-3 concentrations before cantly increased in blood of dogs with malignancies compared with
surgery are associated with longer survival times, while higher concen- healthy dogs or dogs with non-tumour disease and the highest LDH
trations are associated with shorter survival times.21 Thus, the value of concentration is observed among dogs with lymphoma.46 This enzyme
CA15-3 may provide prognostic information in bitches with mammary is easily measurable in freshly collected serum using commercially
tumours. Now, studies showing the independence of this parameter available kits and reference values in dog serum are set from 45 to
from those used in practice for prognostic purposes are required to rec- 233 U/L.20 Interestingly, a significant increase of serum LDH in dogs
ommend serum CA15-3 in prognostic procedures. Combining CA15-3 with malignant mammary tumours has been measured compared with
with other biomarkers, as suggested above,22,37 could also be a lead to healthy dogs and a positive correlation was shown between LDH
follow for canine cancer prognosis. serum concentrations and tumour stage.20 Therefore, serum LDH
concentration shows an interesting diagnostic value for canine
tumours, despite an earlier study showing that dogs affected with
3.3 | The alpha-fetoprotein malignant lymphoma kept LDH levels in the reference values.47 In
cats, a study shows a poor accuracy for serum LDH to discriminate
The alpha-fetoprotein (AFP) is synthesised in large quantities mainly oral lymphoma and inflammatory bowel disease (IBD)48 and LDH
by the liver, during embryonic development. It is indeed one of the levels were also high in serum of dogs infected with canine
770 COLOMBE ET AL.

parvovirus.49 However, interestingly, an increased LDH activity of this marker in the follow-up of animals with GCT or SCT, including
(> 280 U/L) anticipated clinical stages of lymphomas in dogs and was its ability to detect metastatic processes after surgical excision.
also often seen at completion of chemotherapy and at 1 month after
chemotherapy, in dogs with recurrence during the successive 45 days,
with a sensitivity of 77.8% and 96.2% of specificity.50 Thereby, rather 3.6 | The thymidine kinase 1
than a diagnostic biomarker, LDH could be an interesting prognostic
tool to predict recurrence and a follow-up biomarker in veterinary The thymidine kinase 1 (TK1) is one isoform of the thymidine kinase
medicine, but further studies are needed to assess this hypothesis. (TK) which phosphorylates thymidine, an essential nucleotide for DNA
replication. TK1 is located in the cytosol where its quantity and activ-
ity are low when the cell is quiescent and increase during cell divi-
3.5 | The anti-mullerian hormone sion.58 Many studies in dogs have focused on TK1 as a biomarker in
oncology. In most of these studies, it has been shown that the activity
The anti-mullerian hormone (AMH) is a glycoprotein secreted by Ser- of TK1, measured by radioenzymatic assays or ELISA, is increased par-
toli cells in males and by the granulosa cells of small growing follicles ticularly in the blood of dogs suffering from lymphomas and myeloid
in females. A decrease of its secretion has been observed after leukaemia compared with healthy dogs.47,59–66 The diagnostic perfor-
51
puberty in men, bulls and stallions. Thus, this hormone is mainly mances of TK1 activity, outlined by some of these authors, to differ-
studied and used in relation to its potential as a biomarker in granu- entiate dogs with haematopoietic neoplastic processes from healthy
losa cell tumours (GCT) and Sertoli cell tumours (SCT) in humans.52 dogs is shown in Table 1. Thus, the activity of TK1, significantly
Given the importance of reproduction in veterinary breeding, blood increased in haematopoietic tumours, may be interesting in terms of
AMH concentrations in the case of ovarian pathologies were first diagnostic performances mainly for these specific tumours. In terms
investigated in cows and mares. It was shown in both animals that an of prognostic interest, it has been shown that the activity of TK1
increase in the serum AMH concentration above the threshold limit increased with tumour stage.60,65 In addition, an increase in TK1 activ-
diagnostics the presence of GCT, with the cut-off set at 0.36 ng/ml ity above 30 U/L is significantly associated with a decrease in overall
for cows and 4 ng/ml for mares, with sensitivity respectively evalu- survival time, in dogs with lymphomas60 and more recent studies
53,54
ated at 100% and 98%. AMH serum concentrations were signifi- showed that a combination of TK1 with the inflammation marker C
cantly higher in female dogs with GCT compared with healthy bitches Reactive Protein (CRP) gave higher diagnosis accuracy and prognostic
as well as bitches with other ovarian neoplasia or non-neoplastic ovar- values than either of both alone, in dogs with haematological malig-
ian disease such as follicular or luteal cysts. The cut-off value was set nancies.65,67 Finally, several studies have investigated the interest of
at 0.99 ng/ml, which gives to AMH a sensitivity of 100% and specific- TK1 as a pharmacodynamic biomarker in dogs with haematopoietic
ity at 94.44% to detect GCT.55 These values suggest accurate diag- tumours (lymphomas and leukaemias) during treatment. All these
nostic performance for AMH, but further studies would be required in studies consistently showed that dogs undergoing chemotherapy had
bitches, including a larger number of animals, to confirm these results. a significant decrease in TK1 activity, with partial or complete
Interestingly, in male dogs, several articles have also shown a three- response, and up to physiological values.60,61,64,66,68,69 According to
fold or more increase in serum AMH concentrations in SCT dogs com- these same studies, these values remained significantly higher in cases
56,57
pared with healthy dogs. These results strengthened AMH as a of relapse or no response to treatment. Thus, TK1 shows good values
potential biomarker for SCT in male dogs as well. However, the diag- in terms of diagnostic, prognostic and follow-up performance in
nostic interest of this marker is limited by the fact that any pathology canine haematopoietic tumours, and therefore seems to be a good
concerning the ovaries or testicles is mostly treated by surgical exci- candidate as a biomarker in this field, despite a certain lack of specific-
sion in veterinary medicine. Therefore, a precise diagnostic test to ity of this enzyme for other neoplastic pathologies. Interestingly, the
detect a GCT or SCT before surgical removal is rarely necessary. Fur- stability of TK1 enzyme activity is not affected by storage at
20 C
ther studies are now needed to establish the potential clinical utility for up to 3 month60 and has also been correlated with ELISA values in

TABLE 1 Diagnostic performances of TK1 in dogs with haematopoietic cancers (lymphomas and leukaemias) compared with healthy dogsa

Study Dogs with neoplasia (n) Healthy dogs (n) Used technology Sensitivity Specificity
Nakamura et al., 199760 20 13 Radioenzyme assay 100% 100%
61
von Euler et al., 2004 52 21 Radioenzyme assay 92% 98.1%
Thamm et al., 201264 14 24 ELISA 80% n.d.
65
Sharif et al., 2012 34 35 Radiochemical assay 94% 68%
Jagarlamudi et al., 201567 43 42 ELISA 79% 97%

Abbreviation: n.d., non-determined.

a
Calculated by authors from data in the articles.
COLOMBE ET AL. 771

dogs with malignant lymphomas,47 therefore TK1 measurement could investigating the variation in the amount of cfDNA in dogs with neo-
easily become a routine procedure in veterinary oncology. plastic processes have shown a significant increase in the amount of
cfDNA in patients compared with healthy dogs as shown in
Table 2.78–80 Interestingly, Tagawa et al., show that the amount of
4 | N U C LE I C A C I D B I O M A R K E R S plasma cfDNA is correlated with clinical stage, with a significant
increase in plasma cfDNA in metastatic versus non-metastatic
4.1 | Circulating deoxyribonucleic acids dogs.78 In 2013, a first study carried out the complete sequencing of
the tumour genome of four dogs with mammary tumours. Based on
Increased concentrations of cell-free circulating Deoxyribonucleic the genetic rearrangements observed within the tumours, the
Acids (cell-free DNA or cfDNA) have been observed in blood of authors were able to detect in the four dogs, circulating DNA with
human cancer patients as compared with healthy individuals. How- these same rearrangements, that is, tumour circulating DNA.81 The
ever, this increase has been shown later not to be specific of tumours specific BRAF gene mutation was also investigated in dogs with
since cfCDNA is also increased in other human conditions such as urothelial carcinomas since it had been shown that nearly 80% of
pregnancy or after transplantation and in some inflammatory condi- these tumours in dogs have a mutation in this gene.82 The BRAF
70
tions. Whereas cfDNA is the total pool of circulating cell-free DNA, V595E mutation was found in almost 73% of the plasma of dogs
circulating-tumour DNA (ctDNA) is tumour-specific, released from cir- with tumour and the concentration increased with disease progres-
culating tumour cells (CTCs) and from cells within the tumour. Several sion. Thus, the authors suggest that the amount of ctDNA with BRAF
studies in humans suggest that cells within the tumour are the main mutations may represent a useful non-invasive diagnostic biomarker
source of ctDNA compared with CTCs since ctDNA has been in canine urothelial carcinoma.83 Interestingly the BRAF V595E
71,72
detected even in the absence of CTCs in the blood. The ctDNA mutation can also be monitored in urine samples using the new
forms a very small part of the total circulating cfDNA, so it is particu- droplet digital Polymerase Chain Reaction (dPCR) molecular tech-
larly complex to detect. Somatic mutations (point mutations, inser- nique which allowed to identify the mutation in 83% of dogs with
tions, rearrangements), variability in the number of copies of a gene, urothelial carcinoma (19 positive out of 23 patients) and 100% of
size of the fragment, aneuploidy and degree of methylation are used dogs with protastic carcinoma (3 dogs only) when it was completely
to identify ctDNA among cfDNA, using many different methods.73,74 absent in healthy controls.84
Furthermore, it was shown that, thanks to different properties of the Another specific element of ctDNA is the hypomethylation of the
ctDNA, such as its methylation profile, it was possible to determine Long Interspersed Nuclear Element-1 (LINE-1) sequences which are
the tissue of origin of the identified ctDNA.75–77 Thus, a combination associated with oncogenesis and metastases formations.85 In a study
of somatic and epigenetic changes could detect a tumour process and involving healthy dogs, dogs with malignant or benign cancers, the
determine its location. index of the relative level of methylation of the LINE-1 sequences of
A limited number of studies have been carried out on cfDNA in the cfDNA was significantly reduced in dogs with benign and malig-
veterinary oncology rather than on ctDNA. Most of the studies nant mammary tumours (0.29 ± 0.061 and 0.39 ± 0.066, respectively)

TABLE 2 Studies comparing cfDNA quantities in neoplastic and healthy dogs

Tumour group Healthy group

Type and number of tumour Number of [cfDNA]b in

Study bearing dogs b
[cfDNA] in blood healthy dogs blood p-valuea
Letendre and Goggs, Sarcomas n = 15 8.1 μg/mlc (5.1– n = 15 4.7 μg/mlc (0.6– p = .003
201781 15.8) 8.5)
Beffagna et al., Mammary tumours n = 44 Short fragments: n = 15 Short fragments: Short fragments:
201782 128.5 ng/ml 28.8 ng/ml p = .018
(56.8–200.2) (6.5–51.1) Long fragments:
Long fragments: Long fragments: p = .009
87.5 ng/ml 18.2 ng/ml
(33–142) (3.9–32.6)
Tagawa et al., 201980 Various tumoursd n = 50 6290 ng/ml (51– n = 11 481 ng/ml p = .011
115 000) (205–802)
a
In all studies, p < .05 is considered significant.
b
[cfDNA]: concentration of cell-free DNA.
c
Median values and when non-specified mean values.
d
Lymphoma & leukaemia (n = 11), hemangiocarcinoma (n = 8), urothelial carcinoma (n = 5), mammary carcinoma (n = 4), orar squamous cell carcinoma
(n = 3) apocrine gland anal sac carcinoma (n = 3) hepatic tumour (n = 3), soft tissue carcinoma (n = 2), malignant melanoma (n = 2), gastrointestinal stroma
tumour (n = 2), mast cell tumour (n = 1), osteosarcoma (n = 1), ceruminous adenocarcinoma (n = 1), salivary adenocarcinoma (n = 1), apocrine
adenocarcinoma (n = 1), pheocromocytoma (n = 1), seminoma (n = 1).
772 COLOMBE ET AL.

compared with healthy dogs (0.92 ± 0.067). However, this decrease thus attesting to the independent nature of these biomarkers.93,94
was less significant for other tumour types studied in this article.86 Interestingly, as in human medicine where a panel of several miRNA
cfDNA and ctDNA appear to be of interest as well as prognostic showed very high values of diagnosis accuracy to detect people with
and follow-up biomarkers in dogs with neoplastic processes. Indeed, early stage of non-small-cell-lung-tumours compared with either
dogs with high cfDNA concentrations had a significantly shorter dura- miRNA alone,95 a combination of miRNA-214 and miRNA-126 has
tion of remission/lower overall survival rate than dogs with low showed better diagnosis accuracy to detect different types of epithe-
cfDNA concentration with a threshold set at 25 ng/ml when the ref- lial and non-epithelial tumours than both independently.93,94 Several
87
erence value for cfDNA in normal canine plasma is 1–15 ng/ml. additional studies have looked at other mi-RNAs, particularly with
Looking more specifically at ctDNA, two other studies came to the regard to specific tumour conditions. Serum miRNA-122 concentra-
same conclusions.83,88 tions have been shown to be increased in dogs with neoplastic liver
Despite the low number of studies in dogs, cfDNA and ctDNA as damage compared with healthy dogs.96 RNA deep-sequencing
diagnostic, prognostic and monitoring biomarkers in patients suffering (RNASeq) and dPCR techniques identified differential expression of
from neoplastic processes should be considered. The advantages of 65 miRNAs in the blood of healthy female dogs and of dogs with
measuring cfDNA and/or ctDNA are numerous: it can be detected in mammary carcinomas, with two of them significantly over-expressed
blood as well as in non-invasive samples such as urine, appears reli- in the plasma of dogs with tumours: miRNA-19b and miRNA-125a. In
able in detecting a tumour process in a sensitive and specific manner, addition, the plasma concentration of miRNA-18a was significantly
and allows the entire tumour process of an individual to be properly increased in dogs with Stage III versus Stage I–II tumours. A significant
characterised. However, studies involving a larger number of dogs are increase in plasma miRNA-18a concentration was also observed when
needed to establish reliable results and to be integrated into clinical lymphatic invasion was noticed on histological analysis, and could
practice in veterinary oncology. therefore be a predictive biomarker of metastasis.97
Regarding the prognostic performance of blood miRNAs in dogs,
an increase in serum miRNA-125a concentration showed a correla-
4.2 | Circulating ribonucleic acids tion, although not significant, with a decrease in survival time in
bitches with mammary carcinoma.97 Moreover, combined miRNA-126
As for DNA, RNA can also be found in the bloodstream. The main cir- and -214 concentrations did not show a significant impact on survival
culating RNAs are microRNAs (miRNAs). Highly conserved in evolu- times of dogs affected by various type of neoplasia.93 A study focus-
tion, they act as post-transcriptional regulators, based on the principle ing on the prognostic impact of miRNA-126 and -214 concentrations
of partial or complete correspondence with the complementary in dogs with osteosarcoma treated by amputation and adjuvant che-
mRNA. To date, several thousand miRNAs have been identified in motherapy, revealed that an increase in serum miRNA-214 concentra-
humans.89 Several studies have been conducted on miRNAs as a bio- tion was significantly associated with poorer duration of remission
marker in dogs, but only a few have focused on their expression in the and shorter overall survival time. Higher miRNA-126 concentrations
blood of dogs with neoplastic processes. In vitro, the miRNA profiles were on the contrary, significantly associated with longer durations of
contained in exosomes released from normal canine mammary epithe- remission and overall survival. It should be noted that the values of
lial cells obtained from dogs without mammary tumours were com- these markers were independent of different histopathological char-
pared with exosomes released by tumour breast cells, derived from acteristics (mitotic index, histological subtype), but an association was
canine mammary carcinoma. A total of 145 miRNAs were differen- noted with tumour grade.98 For dogs with multicentric lymphomas,
90
tially expressed by tumour cells compared with healthy cells. How- elevated plasma miRNA-21, -31, -45, -150, -155 and -222 concentra-
ever, two other studies in dogs concluded that there was no tions were significantly correlated with lower progression-free sur-
correlation between the miRNA profiles recorded within the tumour vival rates, and elevated plasma miRNA-181c and -222
and in the blood of the same patient.91,92 Thus, the increases or concentrations significantly associated with a lower median overall
decreases in blood concentrations recorded for some miRNAs, may survival time but in the B-cell lymphoma group only.92 Thus, several
not be due solely to differential expression of the miRNAs by the circulating miRNAs appear to have shown potential as prognostic bio-
tumour, but may be the result of a general response of the body to markers in dogs with different tumour types, although few studies
the presence of the tumour. Nevertheless, some studies focused on have been performed on this subject.
specific miRNA, such as miRNA-126 and -214, which are known to Interestingly, serum miRNA-126 and -214 concentrations were
play a role in angiogenesis and tumorigenesis.93 Two studies showed decreased in dogs after surgical removal of a hemangiosarcoma,
a significant increase in blood expression of both biomarkers in dogs attesting the dynamic nature of these two miRNAs.94 Some other
with hemangiosarcoma, compared with dogs with benign processes miRNAs also showed significant variation in their serum concentra-
and healthy dogs. It should also be noted that the values of these two tions in dogs with lymphomas during relapses (increased miRNA-125b
markers did not show a strong correlation with other clinical parame- but miRNA-30b, -34a and -182 decreased).92 These observations
ters (haematocrit, platelet count, prothrombin time and activated par- could trigger an interest to use these miRNAs for monitoring tumour
tial thromboplastin time, blood urea nitrogen, creatinine, alanine processes in dogs and early detection of relapse. However, it is impor-
aminotransferase, aspartate aminotransferase, alkaline phosphatase), tant to underline that a lack of specificity of miRNAs in neoplastic
COLOMBE ET AL. 773

condition could be a main limitation for these as diagnostic bio- EpCAM+/CK+/DAPI+/CD45- CTCs have also been identified in
markers. Indeed, for example, there were no significant differences in the blood of bitches with metastatic mammary carcinoma using the
plasma concentrations of miRNA-122 between dogs with hepatic FDA approved method for human samples. This new technique
neoplasia, hepatic fibrosis or inflammatory liver disease.96 Moreover, showed a sensitivity of 44.4% for the detection of metastasized mam-
some miRNA detected in the blood such as miR-144 or miR-32, are mary carcinoma, and revealed that the presence of CTCs in the blood
not able to discriminate dogs with metastasis from dogs without,91 has a negative prognostic impact.106 Recently, flow cytometry was
adding one more limitation to their value as prognostic biomarkers. also used to detect CTCs in the blood of three dogs with osteosar-
In conclusion, some specific circulating miRNAs have been shown coma with labelling of Collagen I and Osteocalcin. Interestingly, CTC
in humans and dogs to be of interest for diagnosis, prognosis or pre- frequencies in blood were greatly reduced following amputation in all
diction of responses to treatment and for monitoring individuals with three dogs, were variable during chemotherapy and an increase has
neoplastic processes. For all these reasons, miRNAs are presented as been seen in all three dogs within the 4 weeks prior to apparent
the future major biomarkers in the field of oncology by many authors metastasis or death.107 This first study showed a dynamic evolution
and more studies on these in veterinary oncology can be expected in of CTCs during chemotherapy treatment in dogs, suggesting a poten-
the future. tial follow-up biomarker role for CTCs in dogs with osteosarcoma dur-
ing treatment.
In conclusion, CTCs have shown their potential for diagnosis of
5 | C I R C U L A T I N G CE L L S A S B I O M A R K E R S tumour processes despite limited and variable sensitivity. They have
also proven to be interesting in terms of prognosis, monitoring and
5.1 | Circulating tumour cells predictive tools. However, given that flow cytometry and CTCs
research is a relatively recent field in veterinary oncology, it is likely
Circulating Tumour Cells (CTCs) are responsible for the formation of that the increasing number of studies will lead in the near future to a
metastases. CTCs originate from a solid tumour from which they have wider use of these cellular biomarkers.
lost their ability to adhere, allowing them to migrate into the blood. As
these cells are very rare, about 1 CTC among 106 to 107 leukocytes,
the tools used for their detection must be highly sensitive and tech- 5.2 | Myeloid derived suppressor cells
niques for enrichment and isolation of these cells, are necessary.99,100
They are based on the recognition of CTCs' cell surface markers, on Myeloid derived suppressor cells (MDSCs) form a group of heteroge-
their physical or biological characteristics (density, size, invasiveness), neous cells mainly present in the bone marrow where they represent
or on molecular techniques such as RT-PCR for the detection of the immature stages of granulocytes and macrophages. They have a
CTCs-specific RNA. However, only one technique for identifying, iso- strong immunomodulatory activity by mainly inhibiting the prolifera-
lating and enumerating CTCs was approved in 2004 by the Food and tion of T lymphocytes, reducing their cytotoxic power and inducing
Drug Agency (FDA) which includes an enrichment step followed by a their apoptosis. They also reduce the proliferation and activity of B
more specific identification, both using specific antibodies and high lymphocytes, as well as natural killer (NK) cells and dendritic cells.
resolution imaging systems.100 The cells considered as CTCs express Acute or chronic inflammatory phenomena favour the accumulation
the Epithelial cell adhesion molecule (EpCAM), Cytokeratins (CK), and of MDSCs within the sites of inflammation, prevent their differentia-
are positive for 40 ,6-diamidino-2-phenylindole (DAPI) but do not tion and induce their activation.108 It has been shown that many fac-
express the Cluster of Differentiation 45 (CD45). Their phenotype is tors allow the recruitment and expansion of MDSCs into the tumour
therefore EpCAM+/CK+/DAPI+/CD45–.101 Due to the fact that this environment from the bone marrow and secondary lymphoid organs.
is the only FDA-approved technology, the majority of studies in Due to their immunosuppressive role and their angiogenic capacities,
human oncology have been conducted using this system, but in dogs, human MDSCs are key players in tumour escape and metastasis for-
studies have been published using RT-PCR as it is possible to detect mation.109 A significant increase in the proportion of MDSCs among
the presence of mRNAs specifically expressed in CTCs in the peripheral blood mononuclear cells (%MDSCs) has been shown in a
102
blood. Out of 16 human breast cancer CTCs markers mRNAs, six wide range of tumour types compared with the %MDSCs in healthy
were overexpressed in the blood of bitches with mammary carcinoma human subjects.110 Thus, due to the large number of tumour types
103,104
but none were in the blood of healthy dogs. Moreover, prod- with a significant increase in %MDSCs, this endpoint cannot represent
ucts of the CLDN7, CRYAB, ATP8B1 and EGFR genes can differentiate a diagnostic biomarker of a particular tumour but may indicate the
with varying degrees of accuracy between bitches with metastatic possibility of a tumour involvement. MDSCs were first identified in
breast carcinoma and bitches with no evidence of metastasis. In par- dog blood by flow cytometry as immature myeloid cells expressing
ticular, mRNAs from the CRYAB and ATP8B1 genes could be bio- CD11b, but lacking the expression of CD14 and the major histocom-
markers for CTCs in female dogs with mammary carcinoma, since they patibilty complex class II (MHCII).111 The immunosuppressive charac-
both exhibit respectively the best sensitivity and specificity values ter of these cells was confirmed by co-culture experiments in the
(35% and 100% for CRYAB and 32.5% and 90% for AT8B1) among all presence of autologous or heterologous T lymphocytes, whose activ-
the other genes tested.105 ity was reduced in the presence of MDSCs. Blood of dogs with
774 COLOMBE ET AL.

advanced neoplastic processes showed a significant increase in % are also accurate to follow in dogs, and conversely dogs could also be
MDSCs (36.04%) compared with healthy dogs (10.24%) and early- a good model to identify or better characterise them, leading to a true
111
stage tumours (9.40%). These results were confirmed by other personalised medicine, for both humans and dogs.
studies on canine MDSCs in the context of sarcoma, carcinomas and
oral melanomas112 and mammary tumours.113,114 It was proposed AUTHOR CONTRIBU TIONS
later, that canine MDSCs could be distinguished, as in humans and Philippe Colombe: Acquisition of data, analysis, and interpretation of
mice, in two different phenotypes: monocyte-derived MDSCs (M- data, drafting and revising the manuscript. Jérémy Béguin: revising
MDSCs) and polymorphonuclear-MDSCs (PMN-MDSCs) by flow the manuscript. Ghita Benchekroun: revising the manuscript. Del-
cytometry, microscopy, and co-culture experiments. Both cell popula- phine Le Roux: conception and design, analysis and interpretation of
tions are CD5
/CD21
/MHCII
and CD11b+ and they differ from data, drafting and revising the manuscript.
each other by the expression of CD14 and CADO48A, a canine spe-
cific neutrophil marker. M-MDSCs are CD14+/CADO48A
, whereas ACKNOWLEDG MENT
PMN-MDSCs are CD14
/CADO48A+. The percentage of PMN- The authors thanks Dr Darragh Duffy for critical reading of the
MDSCs among blood mononuclear cells (%PMN-MDSC) was manuscript.
increased in dogs with tumour processes compared with healthy
dogs, and this difference was significant in some tumour types.115 In FUNDING INF ORMATI ON
dogs with melanomas, a significant increase in %M-MDSC and % The authors received no funding for this work.
PMN-MDSC compared with healthy dogs was also observed.116
In humans, MDSCs have shown a moderate diagnostic capacity, with CONFLIC T OF INT ER E ST
modest sensitivities and specificities for the detection of tumour The authors have no conflict of interest to declare.
processes.117–119 However, the prognostic capacities of blood %
MDSCs have been widely demonstrated, as well as their ability to OR CID
predict the response to treatments based on their initial levels. Delphine Le Roux https://orcid.org/0000-0003-3252-9155
However, although their increase is more pronounced for advanced
neoplastic processes, their prognostic impact has not been clearly RE FE RE NCE S
established in dogs with tumours. Finally, their usefulness in terms of 1. Henry NL, Hayes DF. Cancer biomarkers. Mol Oncol. 2012;6(2):
follow-up during treatment could not yet be clearly established in both 140-146.
2. Febbo PG, Ladanyi M, Aldape KD, et al. NCCN task force report:
humans and dogs. Therefore, additional studies are necessary to con-
evaluating the clinical utility of tumor markers in oncology. J Natl
firm the results obtained, to study the prognostic impact of MDSCs and Compr Cancer Netw. 2011;9(Suppl 5):S1-S32. quiz S33.
their usefulness in the follow-up of animals under treatment. 3. Brady DC, Crowe MS, Turski ML, et al. Copper is required for onco-
genic BRAF signalling and tumorigenesis. Nature. 2014;509(7501):
492-496.
4. Pan Q, Kleer CG, van Golen KL, et al. Copper deficiency induced by
6 | C O N CL U S I O N tetrathiomolybdate suppresses tumor growth and angiogenesis. Can-
cer Res. 2002;62(17):4854-4859.
In this review, we focused on blood biomarkers studied in veterinary 5. MacDonald G, Nalvarte I, Smirnova T, et al. Memo is a copper-
dependent redox protein with an essential role in migration and
oncology, and that shows a certain degree of specificity and sensitiv-
metastasis. Sci Signal. 2014;7(329):ra56.
ity for neoplastic processes. Most of the biomarkers highlighted to 6. Karginova O, Weekley CM, Raoul A, et al. Inhibition of copper
date are protein-based, including CEA, CA15-3, AFP, LDH, AMH and transport induces apoptosis in triple-negative breast cancer cells and
TK1. They all show varying performances in terms of sensitivity and suppresses tumor angiogenesis. Mol Cancer Ther. 2019;18(5):
specificity to tumour, and each has shown some evidence of their 873-885.
7. Askar TK, Salmanoglu B, Salmanoglu R, Erkal N, Beskaya A. Changes
potential as diagnostic, prognostic, predictive and monitoring bio-
in the oxidative status and serum trace element levels in dogs with
markers. However, none of them has perfect specificity for distin- mammary tumours. Acta Vet (Beograd). 2009;59(4):405-411.
guishing individuals with tumour processes from healthy individuals or 8. Chamel G, Gourlan AT, Telouk P, et al. Retrospective evaluation of
those with non-neoplastic pathologies. The next generation of bio- blood copper stable isotopes ratio (65) Cu/(63) Cu as a biomarker of
cancer in dogs. Vet Comp Oncol. 2017;15(4):1323-1332.
markers from liquid biopsy, seems to be particularly encouraging such
9. Kazmierski KJ, Ogilvie GK, Fettman MJ, et al. Serum zinc, chromium,
as circulating miRNAs, ctDNA, cfDNA, CTCs and MDSCs. Some of and iron concentrations in dogs with lymphoma and osteosarcoma.
these biomarkers have shown high performances in terms of sensitiv- J Vet Intern Med. 2001;15(6):585-588.
ity and specificity for identifying individuals presenting a tumour con- 10. Cunzhi H, Jiexian J, Xianwen Z, Jingang G, Shumin Z, Lili D. Serum
and tissue levels of six trace elements and copper/zinc ratio in
dition. It also appears that they have the capacity to describe tumour
patients with cervical cancer and uterine myoma. Biol Trace Elem
characteristics and monitor their evolution. More specifically, they Res. 2003;94(2):113-122.
allow to predict precisely sensitivities of the tumours studied to the 11. Mao S, Huang S. Zinc and copper levels in bladder cancer: a system-
various treatment protocols, both in humans and dogs. According to atic review and meta-analysis. Biol Trace Elem Res. 2013;153(1–3):
5-10.
the One Health concept, most of the biomarkers studied in humans
COLOMBE ET AL. 775

12. Kumar R, Razab S, Prabhu K, Ray S, Prakash B. Serum butyrylcholi- 31. Dolscheid-Pommerich RC, Keyver-Paik M, Hecking T, et al. Clinical
nesterase and zinc in breast cancer. J Cancer Res Ther. 2017;13(2): performance of LOCI-based tumor marker assays for tumor markers
367-370. CA 15-3, CA 125, CEA, CA 19-9 and AFP in gynecological cancers.
13. Xie Y, Wang J, Zhao X, et al. Higher serum zinc levels may reduce Tumour Biol. 2017;39(10):1010428317730246.
the risk of cervical cancer in Asian women: a meta-analysis. J Int 32. Yeganeh-Amirkande S, Assmar M, Mansour-Ghanaei F, Mozafari-
Med Res. 2018;46(12):4898-4906. Noor A. The frequency of CA15-3, CA125, CA19-9 in patients with
14. Wang Y, Sun Z, Li A, Zhang Y. Association between serum zinc levels hepatitis B and C. Zahedan J Res Med Sci. 2015;17(5):e959.
and lung cancer: a meta-analysis of observational studies. World J 33. Szekanecz E, Sandor Z, Antal-Szalmas P, et al. Increased production
Surg Oncol. 2019;17(1):78. of the soluble tumor-associated antigens CA19-9, CA125, and
15. Hammarstrom S. The carcinoembryonic antigen (CEA) family: struc- CA15-3 in rheumatoid arthritis: potential adhesion molecules in
tures, suggested functions and expression in normal and malignant synovial inflammation? Ann N Y Acad Sci. 2007;1108:359-371.
tissues. Semin Cancer Biol. 1999;9(2):67-81. 34. Marchesi MC, Conti MB, Pieramati C, Mangili V, Fruganti G. Assess-
16. Ledecky V, Valencakova-Agyagosova A, Lepej J, Frischova Z, ment and behavior of alphafetoprotein (AFP), antigen cancer 15/3
Hornak S, Nagy V. Determination of carcinoembryonic antigen and (CA 15/3), carcinembryonal antigen (CEA) in clinical oncology of the
cancer antigen values with the radioimmunoassay method in healthy dog: preliminary study. Vet Res Commun. 2007;31(Suppl 1):301-304.
females dogs. Vet Med. 2013;58(5):277-283. 35. Marchesi MC, Manuali E, Pacifico E, et al. Cancer antigen 15/3: pos-
17. Valencakova-Agyagosova A, Frischova Z, Sevcikova Z, et al. Deter- sible diagnostic use in veterinary clinical oncology. Preliminary study.
mination of carcinoembryonic antigen and cancer antigen (CA 15-3) Vet Res Commun. 2010;34(Suppl 1):S103-S106.
in bitches with tumours on mammary gland: preliminary report. Vet 36. Manuali E, De Giuseppe A, Feliziani F, et al. CA 15-3 cell lines and
Comp Oncol. 2014;12(3):205-214. tissue expression in canine mammary cancer and the correlation
18. Alizadeh S, Azimzadeh K. Evaluation of ultrasonography and mam- between serum levels and tumour histological grade. BMC Vet Res.
mography in diagnosis of mammary gland tumor in bitches: based on 2012;8:86.
tumor markers. Iran J Vet Surg. 2018;13(1):23-28. 37. Jain M, Ingole SD, Deshmukh RS, et al. CEA, CA 15-3, and miRNA
19. Senhorello ILS, Terra EM, Sueiro FAR, et al. Clinical value of carci- expression as potential biomarkers in canine mammary tumors.
noembryonic antigen in mammary neoplasms of bitches. Vet Comp Chromosome Res. 2021;29(2):175-188.
Oncol. 2020;18(3):315-323. 38. Terentiev AA, Moldogazieva NT. Alpha-fetoprotein: a renaissance.
20. Campos LC, Lavalle GE, Estrela-Lima A, et al. CA15.3, CEA and LDH Tumor Biol. 2013;34(4):2075-2091.
in dogs with malignant mammary tumors. J Vet Intern Med. 2012; 39. Takahashi Y, Ohta T, Mai M. Angiogenesis of AFP producing gastric
26(6):1383-1388. carcinoma: correlation with frequent liver metastasis and its inhibi-
21. Baba OK, Sood NK, Gupta K. Clinical evaluation of glycoproteins tion by anti-AFP antibody. Oncol Rep. 2004;11(4):809-813.
and inflammatory cytokines in the serum of dogs affected with 40. Liang OD, Korff T, Eckhardt J, et al. Oncodevelopmental alpha-
canine mammary cancer. Proc Natl Acad Sci India Sect B Biol Sci. fetoprotein acts as a selective proangiogenic factor on endothelial
2019;89(4):1465-1469. cell from the fetomaternal unit. J Clin Endocrinol Metab. 2004;89(3):
22. Fan Y, Ren X, Liu X, et al. Combined detection of CA15-3, CEA, and 1415-1422.
SF in serum and tissue of canine mammary gland tumor patients. Sci 41. Yamada T, Kakinoki M, Totsuka K, et al. Purification of canine alpha-
Rep. 2021;11(1):6651. fetoprotein and alpha- fetoprotein values in dogs. Vet Immunol
23. Tinder TL, Subramani DB, Basu GD, et al. MUC1 enhances tumor Immunopathol. 1995;47(1–2):25-33.
progression and contributes toward immunosuppression in a mouse 42. Kawarai S, Hashizaki K, Kitao S, et al. Establishment and characteri-
model of spontaneous pancreatic adenocarcinoma. J Immunol. 2008; zation of primary canine hepatocellular carcinoma cell lines produc-
181(5):3116-3125. ing alpha-fetoprotein. Vet Immunol Immunopathol. 2006;113(1–2):
24. Baruch A, Hartmann M, Zrihan-Licht S, et al. Preferential expression 30-36.
of novel MUC1 tumor antigen isoforms in human epithelial tumors 43. Yamada T, Fujita M, Kitao S, et al. Serum alpha-fetoprotein values in
and their tumor-potentiating function. Int J Cancer. 1997;71(5): dogs with various hepatic diseases. J Vet Med Sci. 1999;61(6):
741-749. 657-659.
25. Regimbald LH, Pilarski LM, Longenecker BM, Reddish MA, 44. Kitao S, Yamada T, Ishikawa T, et al. Alpha-fetoprotein in serum and
Zimmermann G, Hugh JC. The breast mucin MUCI as a novel adhe- tumor tissues in dogs with hepatocellular carcinoma. J Vet Diagn
sion ligand for endothelial intercellular adhesion molecule 1 in breast Invest. 2006;18(3):291-295.
cancer. Cancer Res. 1996;56(18):4244-4249. 45. Lv J, Zhou Z, Wang J, et al. Prognostic value of lactate dehydroge-
26. Fu Y, Li H. Assessing clinical significance of serum CA15-3 and carci- nase expression in different cancers: a meta-analysis. Am J Med Sci.
noembryonic antigen (CEA) levels in breast cancer patients: a meta- 2019;358(6):412-421.
analysis. Med Sci Monit. 2016;22:3154-3162. 46. Marconato L, Crispino G, Finotello R, Mazzotti S, Salerni F, Zini E.
27. Ghosh I, Bhattacharjee D, Das AK, Chakrabarti G, Dasgupta A, Serum lactate dehydrogenase activity in canine malignancies. Vet
Dey SK. Diagnostic role of tumour markers CEA, CA15-3, CA19-9 Comp Oncol. 2009;7(4):236-243.
and CA125 in lung cancer. Indian J Clin Biochem. 2013;28(1):24-29. 47. von Euler HP, Ohrvik AB, Eriksson SK. A non-radiometric method
28. Li SX, Yang YQ, Jin LJ, Cai ZG, Sun Z. Detection of survivin, carci- for measuring serum thymidine kinase activity in malignant lym-
noembryonic antigen and ErbB2 level in oral squamous cell carci- phoma in dogs. Res Vet Sci. 2006;80(1):17-24.
noma patients. Cancer Biomark. 2016;17(4):377-382. 48. Terragni R, Morselli-Labate AM, Vignoli M, Bottero E, Brunetti B,
29. Dolscheid-Pommerich RC, Manekeller S, Walgenbach-Brunagel G, Saunders JH. Is serum Total LDH evaluation able to differentiate
et al. Clinical performance of CEA, CA19-9, CA15-3, CA125 and between alimentary lymphoma and inflammatory bowel disease in a
AFP in gastrointestinal cancer using LOCI-based assays. Anticancer real world clinical setting? PLoS One. 2016;11(3):e0151641.
Res. 2017;37(1):353-359. 49. Wang B, Wang XL. Species diversity of fecal microbial flora in Canis
30. Yedema C, Massuger L, Hilgers J, et al. Pre-operative discrimination lupus familiaris infected with canine parvovirus. Vet Microbiol. 2019;
between benign and malignant ovarian tumors using a combination 237:108390.
of CA125 and CA15.3 serum assays. Int J Cancer Suppl. 1988;3: 50. Marconato L, Crispino G, Finotello R, Mazzotti S, Zini E. Clinical rele-
61-67. vance of serial determinations of lactate dehydrogenase activity
776 COLOMBE ET AL.

used to predict recurrence in dogs with lymphoma. J Am Vet Med new, fully automated non-radiometric assay. Int J Oncol. 2009;34(2):
Assoc. 2010;236(9):969-974. 505-510.
51. Walter B, Feulner H, Otzdorff C, Klein R, Reese S, Meyer- 69. Elliott JW, Cripps P, Blackwood L. Thymidine kinase assay in canine
Lindenberg A. Changes in anti-Mullerian hormone concentrations in lymphoma. Vet Comp Oncol. 2013;11(1):1-13.
bitches throughout the oestrous cycle. Theriogenology. 2019;127: 70. Kustanovich A, Schwartz R, Peretz T, Grinshpun A. Life and death of
114-119. circulating cell-free DNA. Cancer Biol Ther. 2019;20(8):1057-1067.
52. Rey R, Sabourin JC, Venara M, et al. Anti-Mullerian hormone is a 71. Bettegowda C, Sausen M, Leary RJ, et al. Detection of circulating
specific marker of sertoli- and granulosa-cell origin in gonadal tumor DNA in early- and late-stage human malignancies. Sci Transl
tumors. Hum Pathol. 2000;31(10):1202-1208. Med. 2014;6(224):224ra224.
53. Ball BA, Almeida J, Conley AJ. Determination of serum anti- 72. Thierry AR, El Messaoudi S, Gahan PB, Anker P, Stroun M. Origins,
Mullerian hormone concentrations for the diagnosis of granulosa- structures, and functions of circulating DNA in oncology. Cancer
cell tumours in mares. Equine Vet J. 2013;45(2):199-203. Metastasis Rev. 2016;35(3):347-376.
54. El-Sheikh Ali H, Kitahara G, Nibe K, et al. Plasma anti-Mullerian hor- 73. Gilson P. Enrichment and analysis of ctDNA. Recent Results Cancer
mone as a biomarker for bovine granulosa-theca cell tumors: com- Res. 2020;215:181-211.
parison with immunoreactive inhibin and ovarian steroid 74. Heidrich I, Ackar L, Mossahebi Mohammadi P, Pantel K. Liquid biop-
concentrations. Theriogenology. 2013;80(8):940-949. sies: potential and challenges. Int J Cancer. 2021;148(3):528-545.
55. Walter B, Coelfen A, Jager K, Reese S, Meyer-Lindenberg A, 75. Sun K, Jiang P, Chan KC, et al. Plasma DNA tissue mapping by
Aupperle-Lellbach H. Anti-Muellerian hormone concentration in genome-wide methylation sequencing for noninvasive prenatal, can-
bitches with histopathologically diagnosed ovarian tumours and cer, and transplantation assessments. Proc Natl Acad Sci U S A. 2015;
cysts. Reprod Domest Anim. 2018;53(3):784-792. 112(40):E5503-E5512.
56. Ano H, Hidaka Y, Katamoto H. Evaluation of anti-Mullerian hormone 76. Lehmann-Werman R, Neiman D, Zemmour H, et al. Identification of
in a dog with a Sertoli cell tumour. Vet Dermatol. 2014;25(2):142- tissue-specific cell death using methylation patterns of circulating
145. e41. DNA. Proc Natl Acad Sci U S A. 2016;113(13):E1826-E1834.
57. Holst BS, Dreimanis U. Anti-Mullerian hormone: a potentially useful 77. Hao X, Luo H, Krawczyk M, et al. DNA methylation markers for diag-
biomarker for the diagnosis of canine Sertoli cell tumours. BMC Vet nosis and prognosis of common cancers. Proc Natl Acad Sci U S A.
Res. 2015;11:166. 2017;114(28):7414-7419.
58. Bitter EE, Townsend MH, Erickson R, Allen C, O'Neill KL. Thymidine 78. Tagawa M, Shimbo G, Inokuma H, Miyahara K. Quantification of
kinase 1 through the ages: a comprehensive review. Cell Biosci. plasma cell-free DNA levels in dogs with various tumors. J Vet Diagn
2020;10(1):138. Invest. 2019;31(6):836-843.
59. Nakamura N, Momoi Y, Watari T, Yoshino T, Tsujimoto H, 79. Letendre JA, Goggs R. Measurement of plasma cell-free DNA con-
Hasegawa A. Plasma thymidine kinase activity in dogs with lym- centrations in dogs with sepsis, trauma, and neoplasia. J Vet Emerg
phoma and leukemia. J Vet Med Sci. 1997;59(10):957-960. Crit Care. 2017;27(3):307-314.
60. von Euler H, Einarsson R, Olsson U, Lagerstedt AS, Eriksson S. 80. Beffagna G, Sammarco A, Bedin C, et al. Circulating cell-free DNA in
Serum thymidine kinase activity in dogs with malignant lymphoma: a dogs with mammary tumors: short and long fragments and integrity
potent marker for prognosis and monitoring the disease. J Vet Intern index. PLoS One. 2017;12(1):e0169454.
Med. 2004;18(5):696-702. 81. Beck J, Hennecke S, Bornemann-Kolatzki K, et al. Genome aberra-
61. Sharif H, von Euler H, Westberg S, He E, Wang L, Eriksson S. A sen- tions in canine mammary carcinomas and their detection in cell-free
sitive and kinetically defined radiochemical assay for canine and plasma DNA. PLoS One. 2013;8(9):e75485.
human serum thymidine kinase 1 (TK1) to monitor canine malignant 82. Mochizuki H, Breen M. Comparative aspects of BRAF mutations in
lymphoma. Vet J. 2012;194(1):40-47. canine cancers. Vet Sci. 2015;2(3):231-245.
62. Thamm DH, Kamstock DA, Sharp CR, et al. Elevated serum thymi- 83. Tagawa M, Tambo N, Maezawa M, et al. Quantitative analysis of the
dine kinase activity in canine splenic hemangiosarcoma. Vet Comp BRAF V595E mutation in plasma cell-free DNA from dogs with
Oncol. 2012;10(4):292-302. urothelial carcinoma. PLoS One. 2020;15(4):e0232365.
63. Kiran Kumar J, Sharif H, Westberg S, von Euler H, Eriksson S. High 84. Mochizuki H, Shapiro SG, Breen M. Detection of BRAF mutation in
levels of inactive thymidine kinase 1 polypeptide detected in sera urine DNA as a molecular diagnostic for canine urothelial and pros-
from dogs with solid tumours by immunoaffinity methods: implica- tatic carcinoma. PLoS One. 2015;10(12):e0144170.
tions for in vitro diagnostics. Vet J. 2013;197(3):854-860. 85. Lavasanifar A, Sharp CN, Korte EA, Yin T, Hosseinnejad K,
64. Jagarlamudi KK, Moreau L, Westberg S, Ronnberg H, Eriksson S. A Jortani SA. Long interspersed nuclear element-1 mobilization as a
new Sandwich ELISA for quantification of thymidine kinase 1 protein target in cancer diagnostics, prognostics and therapeutics. Clin Chim
levels in sera from dogs with different malignancies can aid in dis- Acta. 2019;493:52-62.
ease management. PLoS One. 2015;10(9):e0137871. 86. Lee KH, Shin TJ, Kim WH, Cho JY. Methylation of LINE-1 in cell-free
65. Selting KA, Ringold R, Husbands B, Pithua PO. Thymidine kinase DNA serves as a liquid biopsy biomarker for human breast cancers
type 1 and C-reactive protein concentrations in dogs with spontane- and dog mammary tumors. Sci Rep. 2019;9(1):175.
ously occurring cancer. J Vet Intern Med. 2016;30(4):1159-1166. 87. Schaefer DM, Forman MA, Kisseberth WC, et al. Quantification of
66. Boye P, Floch F, Serres F, et al. Evaluation of serum thymidine kinase plasma DNA as a prognostic indicator in canine lymphoid neoplasia.
1 activity as a biomarker for treatment effectiveness and prediction Vet Comp Oncol. 2007;5(3):145-155.
of relapse in dogs with non-Hodgkin lymphoma. J Vet Intern Med. 88. Gelaleti GB, Granzotto A, Leonel C, et al. Short interspersed CAN
2019;33(4):1728-1739. SINE elements as prognostic markers in canine mammary neoplasia.
67. Saellstrom S, Sharif H, Jagarlamudi KK, Ronnberg H, Wang L, Oncol Rep. 2014;31(1):435-441.
Eriksson S. Serum TK1 protein and C-reactive protein correlate to 89. Dong H, Lei J, Ding L, Wen Y, Ju H, Zhang X. MicroRNA: function,
treatment response and predict survival in dogs with hematologic detection, and bioanalysis. Chem Rev. 2013;113(8):6207-6233.
malignancies. Res Vet Sci. 2022;145:213-221. 90. Fish EJ, Irizarry KJ, DeInnocentes P, et al. Malignant canine mam-
68. Von Euler HP, Rivera P, Aronsson AC, Bengtsson C, Hansson LO, mary epithelial cells shed exosomes containing differentially
Eriksson SK. Monitoring therapy in canine malignant lymphoma and expressed microRNA that regulate oncogenic networks. BMC Can-
leukemia with serum thymidine kinase 1 activity-evaluation of a cer. 2018;18(1):832.
COLOMBE ET AL. 777

91. Bulkowska M, Rybicka A, Senses KM, et al. MicroRNA expression pat- 107. Wright T, Brisson BA, Wood GA, et al. Flow cytometric detection of
terns in canine mammary cancer show significant differences between circulating osteosarcoma cells in dogs. Cytometry A. 2019;95(9):997-
metastatic and non-metastatic tumours. BMC Cancer. 2017;17(1):728. 1007.
92. Craig KKL, Wood GA, Keller SM, Mutsaers AJ, Wood RD. MicroRNA 108. Gabrilovich DI, Nagaraj S. Myeloid-derived suppressor cells as regu-
profiling in canine multicentric lymphoma. PLoS One. 2019;14(12): lators of the immune system. Nat Rev Immunol. 2009;9(3):162-174.
e0226357. 109. Nagaraj S, Gabrilovich DI. Myeloid-derived suppressor cells in
93. Heishima K, Ichikawa Y, Yoshida K, et al. Circulating microRNA-214 human cancer. Cancer J. 2010;16(4):348-353.
and -126 as potential biomarkers for canine neoplastic disease. Sci 110. Veglia F, Perego M, Gabrilovich D. Myeloid-derived suppressor cells
Rep. 2017;7(1):2301. coming of age. Nat Immunol. 2018;19(2):108-119.
94. Heishima K, Mori T, Ichikawa Y, et al. MicroRNA-214 and 111. Goulart MR, Pluhar GE, Ohlfest JR. Identification of myeloid derived
microRNA-126 are potential biomarkers for malignant endothelial suppressor cells in dogs with naturally occurring cancer. PLoS One.
proliferative diseases. Int J Mol Sci. 2015;16(10):25377-25391. 2012;7(3):e33274.
95. Zhang YH, Jin M, Li J, Kong X. Identifying circulating miRNA bio- 112. Sherger M, Kisseberth W, London C, Olivo-Marston S,
markers for early diagnosis and monitoring of lung cancer. Biochim Papenfuss TL. Identification of myeloid derived suppressor cells in
Biophys Acta Mol basis Dis. 2020;1866(10):165847. the peripheral blood of tumor bearing dogs. BMC Vet Res. 2012;
96. Oosthuyzen W, Ten Berg PWL, Francis B, et al. Sensitivity and spec- 8:209.
ificity of microRNA-122 for liver disease in dogs. J Vet Intern Med. 113. Mucha J, Majchrzak K, Taciak B, Hellmen E, Krol M. MDSCs mediate
2018;32(5):1637-1644. angiogenesis and predispose canine mammary tumor cells for
97. Fish EJ, Martinez-Romero EG, DeInnocentes P, et al. Circulating metastasis via IL-28/IL-28RA (IFN-lambda) signaling. PLoS One.
microRNA as biomarkers of canine mammary carcinoma in dogs. 2014;9(7):e103249.
J Vet Intern Med. 2020;34(3):1282-1290. 114. Mucha J, Rybicka A, Dolka I, et al. Immunosuppression in dogs dur-
98. Heishima K, Meuten T, Yoshida K, Mori T, Thamm DH. Prognostic ing mammary cancer development. Vet Pathol. 2016;53(6):1147-
significance of circulating microRNA-214 and -126 in dogs with 1153.
appendicular osteosarcoma receiving amputation and chemother- 115. Goulart MR, Hlavaty SI, Chang YM, et al. Phenotypic and transcrip-
apy. BMC Vet Res. 2019;15(1):39. tomic characterization of canine myeloid-derived suppressor cells.
99. Pantel K, Speicher MR. The biology of circulating tumor cells. Onco- Sci Rep. 2019;9(1):3574.
gene. 2016;35(10):1216-1224. 116. Hutchison S, Sahay B, de Mello SC, et al. Characterization of
100. Maly V, Maly O, Kolostova K, Bobek V. Circulating tumor cells in myeloid-derived suppressor cells and cytokines GM-CSF, IL-10 and
diagnosis and treatment of lung cancer. In Vivo. 2019;33(4):1027- MCP-1 in dogs with malignant melanoma receiving a GD3-based
1037. immunotherapy. Vet Immunol Immunopathol. 2019;216:109912.
101. Paoletti C, Hayes DF. Circulating tumor cells. Adv Exp Med Biol. 117. Diaz-Montero CM, Finke J, Montero AJ. Myeloid-derived suppres-
2016;882:235-258. sor cells in cancer: therapeutic, predictive, and prognostic implica-
102. Mostert B, Sleijfer S, Foekens JA, Gratama JW. Circulating tumor tions. Semin Oncol. 2014;41(2):174-184.
cells (CTCs): detection methods and their clinical relevance in breast 118. Jin G, Zhang Y, Chang X, et al. Increased percentage of mo-MDSCs
cancer. Cancer Treat Rev. 2009;35(5):463-474. in human peripheral blood may be a potential indicator in the diag-
103. da Costa A, Oliveira JT, Gartner F, Kohn B, Gruber AD, Klopfleisch R. nosis of breast cancer. Oncol Res Treat. 2017;40(10):603-608.
Potential markers for detection of circulating canine mammary tumor 119. Bergenfelz C, Roxa A, Mehmeti M, Leandersson K, Larsson AM.
cells in the peripheral blood. Vet J. 2011;190(1):165-168. Clinical relevance of systemic monocytic-MDSCs in patients with
104. da Costa A, Lenze D, Hummel M, Kohn B, Gruber AD, Klopfleisch R. metastatic breast cancer. Cancer Immunol Immunother. 2020;69(3):
Identification of six potential markers for the detection of circulating 435-448.
canine mammary tumour cells in the peripheral blood identified by
microarray analysis. J Comp Pathol. 2012;146(2–3):143-151.
105. da Costa A, Kohn B, Gruber AD, Klopfleisch R. Multiple RT-PCR
markers for the detection of circulating tumour cells of metastatic How to cite this article: Colombe P, Béguin J, Benchekroun G,
canine mammary tumours. Vet J. 2013;196(1):34-39.
Le Roux D. Blood biomarkers for canine cancer, from human
106. Marconato L, Facchinetti A, Zanardello C, et al. Detection and prog-
nostic relevance of circulating and disseminated tumour cell in dogs to veterinary oncology. Vet Comp Oncol. 2022;20(4):767‐777.
with metastatic mammary carcinoma: a pilot study. Cancer. 2019; doi:10.1111/vco.12848
11(2):163.