May 15, 2022
Mast cell tumors (MCTs) arise from malignantly transformed mast cells. In dogs, most of these tumors arise as primary tumors in the skin. They are the most common skin tumor in dogs, accounting for roughly 20% of all reported skin tumors.1 Any breed may be affected with MCTs, but certain breeds are predisposed, including golden retrievers, Labrador retrievers, Boston terriers, boxers, and pugs. Pugs are more likely to have multiple MCTs at diagnosis (56% of pugs with MCT in one study), but these tumors demonstrate more benign behavior and rarely lead to death.2 MCTs can affect dogs of any age but typically affect middle-aged to older dogs. An underlying etiology for most tumors cannot be identified. Breed predilections support some component of underlying genetic causes. Mutations in the c-kit tyrosine kinase receptor, which can lead to malignant transformation of mast cells, are found in 25%–30% of intermediate to high-grade tumors.3,4 KIT mutations will be further discussed in regards to both prognosis and treatment options for MCTs.
MCTs can be located anywhere on the body and may lie within the dermis and/or subcutis. They have a wide range of gross appearance, from raised and superficial to very deep and fixed; they may feel soft and fluctuant or firm. Most MCTs are easily diagnosed with fine needle aspiration (FNA). Infrequently, MCT granules will not stain with Diff-Quik (Jorgensen Laboratories Inc., Loveland, CO, USA) and need to be stained with a Wright’s stain. On Diff-Quik cytology, if eosinophils are seen along with large round cells that lack granules, suspicion should be raised for an MCT and the slide submitted to a clinical pathology laboratory for a non-Diff-Quik stain.5
The majority of MCTs will be cured with surgical excision.1 Prognostic factors for predicting MCTs that will exhibit a more aggressive biologic behavior – ie, tumors that will not be cured despite local excision and that will ultimately lead to the patient’s death – are varied as well as controversial. When to pursue staging tests in dogs with MCTs, which tests to perform, and treatment recommendations beyond surgery are based on the predicted biologic behavior of the tumor, with staging diagnostics and systemic therapy the recommendation for dogs with biologically aggressive MCTs.
Prognostic factors relating to history and physical examination
Some factors that can be obtained from a history and physical examination that are generally accepted to carry a more guarded prognosis in dogs with MCTs include recent, rapid tumor growth and fixed, ulcerated tumors.6–8 Although publications regarding these features are limited, one early study reported doubling of the survival percentage at 30 weeks post-MCT excision for dogs with slow-growing tumors versus (vs) those with more rapidly growing masses.8 Biologically, both the ability to grow quickly and to become fixed to deeper tissues are physical manifestations of more aggressive behavior. Tumor location on the body can also be associated with biologic behavior; this topic is more controversial and the pertinent locations and published papers are highlighted as follows.
In limited published cases, eyelid margin MCTs appeared to have relatively benign behavior and were effectively treated with local therapy, although one dog was reported to have regional lymph node (LN) metastasis.9–11 MCT of the conjunctiva may be of concern only locally, without reported metastasis in three dogs.12,13 In a paper evaluating chemotherapy for high-risk MCT patients, eleven dogs with mucous membrane MCTs (vulva, prepuce, conjunctiva, oral cavity) had significantly shorter median survival times (MST) than 50 dogs with MCTs of haired skin.14 However, a recent paper of 32 dogs with 33 conjunctival MCTs treated with surgery alone showed prolonged survival times, with only two dogs having local recurrence despite incomplete margins in 25 cases, and no dogs dying of mast cell-related disease.15
Early case reports described aggressive behavior and local metastatic disease at diagnosis in two dogs with MCT of the lip; survival times were 6 months or less.16,17 Of five dogs with MCT of the tongue, two presented with LN and/or systemic metastasis, and two of the remaining three had postoperative local recurrence leading to euthanasia.18 Larger, more recent studies confirmed that MCTs on the muzzle, perioral mucocutaneous junction, or oral mucosa have a more aggressive biologic behavior, with increased risk of locoregional LN metastasis.19–21 The rate of documented metastasis to local (mandibular) LNs was 55%–59%, compared with a <10% rate for other cutaneous sites. Despite a high rate of metastasis, the MSTs of the dogs were prolonged at 30 months, 52 months, and median not reached. Treatments varied in these cases, with many dogs receiving surgery, radiation, and chemotherapy. Dogs with LN metastasis had significantly shorter MSTs than dogs without nodal metastasis, with medians of 14 months, less than 20 months, and 9 months.19–21
Historically, the “back half” of the dog was considered to carry a worse prognosis, and although this is generally not supported, there are still concerns about MCTs located in the inguinal areas, especially involving the prepuce or scrotum. One study reported on dogs with MCT in the perineal and/or inguinal region treated intensively, most receiving trimodality therapy with surgery, radiation, and chemotherapy. The MST was 37 months, and the authors concluded that with appropriate therapy dogs with perineal or inguinal MCTs can do well.22 It is important to note that these dogs received a more aggressive treatment protocol than do the majority of dogs with MCTs at other cutaneous sites. A later study compared the outcome of dogs with inguinal or perineal MCTs with dogs with MCTs in other cutaneous locations; these dogs received a range of therapies. When the 12 dogs with preputial or scrotal MCTs were analyzed separately from other inguinal or perineal tumors, their disease-free interval was significantly shorter (4.2 months) than for the 84 dogs with tumors in noninguinal cutaneous locations (33.9 months).23 Dogs with preputial or scrotal MCTs were also significantly more likely to have received chemotherapy, thus potentially biasing the results. The authors state that the power of the study may not have been adequate to identify a difference between the locations. Scrotal and preputial MCTs may have a more malignant biological behavior; further study is needed.
Recently, several papers presented information regarding MCTs in the subcutaneous (SQ) location.24–26 As the grading system was developed only with cutaneous tumors, it cannot necessarily be applied to SQ tumors with prognostic accuracy. Survival times for dogs with the SQ tumors were found to be prolonged, with the majority of dogs in both papers (53 dogs and 306 dogs) not succumbing to mast cell disease.24,25 Out of the 306 dogs, only 8% had local recurrence, and only 4% had metastasis.25 Risk factors identified for local recurrence and metastasis included mitotic index (MI), infiltrative vs circumscribed histologic pattern, the presence of multinucleation, as well as Ki67, Ki67 + AgNOR, and KIT cellular localization pattern.25,26 As most of the dogs with SQ MCTs were cured, and MI had the strongest correlation with clinical outcomes, there seems to be little need for, or benefit in, assessment of the risk factors that require additional staining.
A histologic grading scheme (I–III) was developed for classification of MCTs affecting haired skin and is still one of the most accepted prognostic indicators of canine MCT behavior.1 However, histologic assessment is prone to operator subjectivity, giving rise to extremely variable grading results for the same tumor among different pathologists.27,28 In one study with ten pathologists evaluating the same 60 MCTs, there was agreement on grade by all pathologists for only four tumors, and six of the MCTs had all three grades assigned. Differences in the references used to grade the MCTs were the suspected reason for the disparity, with six different references being used.28 In a follow-up study involving the same tumors and pathologists, the Patnaik grading scheme29 was used by all. Although mean agreement did improve, only 16 of 60 tumors received the same grade by all, and three of 60 tumors still received all three grades.27 The subjectivity and variability between pathologists bring into question the heavy reliance placed on the grade of an MCT to predict its behavior. In addition, most MCTs are designated grade II, and most of these are cured with surgery. How to identify the subset of grade II tumors that will show that aggressive behavior is a matter of ongoing debate and study.
To address some of the grading concerns, a two-tier (high vs low grade) grading scheme was proposed in a study of 95 dogs with MCTs treated with surgical excision alone.30 High-grade tumors had one of the following criteria: MI of ≥7, three or more multinucleated cells or cells with bizarre nuclei in 10 high-power fields, or karyomegaly. Areas of highest mitoses or anisokaryosis were evaluated. MST for high grade (ten dogs) was 3.6 months vs median not reached (>2 years) for low grade (85 dogs). A retrospective study with 47 dogs attempted to validate the two-tier scheme; although some of the data reported were inconsistent (including no description of the MI of the cases), the high-grade cases did have significantly worse progression-free and overall survivals than the low-grade ones.31 Further studies validating this proposed system are warranted; currently, the author’s university pathology department provides the standard grade, an MI, and the two-tier grade for every sample. Clinically, heavy reliance is placed on the MI for behavior prediction.
Other assessments on biopsies – highlighting mitotic index
As grade is so variable and subjective, many other prognostic factors, including DNA aneuploidy, c-kit-staining pattern, presence of c-kit mutations, microvessel density, Ki67, proliferating cell nuclear antigen, and MI, have been evaluated in an attempt to better predict the behavior of canine MCTs and to pick out the “bad” grade II tumors. Two groups have evaluated the MI (total number of mitotic figures counted in ten high-power fields; fields with the highest mitoses counted) and found that it is predictive of survival time, even within the grade II tumor category. The groups did identify different values for the index, with the first paper showing an MST of >70 months for an index score ≤5 and survival <2 months for a score >5.32 The second group wrote a letter to the editor in response to the first paper and confirmed that high scores have very short survival times, with index scores broken down into three groups: MI =0, MST not reached; MI =1–7, MST =18 months; and MI >7, MST =3 months.33 The cutoff of MI ≥7 was subsequently adopted for the two-tier system.
Although MCT histologic “prognostic panels” are offered by some laboratories, no publications have shown how these panels may provide additional benefit over the MI and grade for prediction of tumor behavior.
In general, staging tests for asymptomatic dogs with cutaneous MCTs are extremely low yield. The test most often positive is regional LN aspiration. Cytologic assessment of the locoregional LN is important, even if the node is not enlarged. In a study of dogs with muzzle MCT, four of eleven LNs with metastases were of normal size.19 Another study with 55 dogs with confirmed LN metastasis and 35 dogs without metastasis showed a sensitivity of 71% and specificity of 54% for palpation as a predictor of metastasis.34 Sixteen of 35 dogs (46%) with normal size LNs on physical examination had metastasis, whereas in another study eight out of 21 (38%) normal size LNs showed metastasis.35 FNA of the LN is ideally performed prior to excision of the primary MCT, as surgical treatment of the tumor can produce confusing LN results due to local postoperative inflammation. As mast cells can be a normal feature in LNs, some MCT cases will not have a definitive answer on LN cytology. Criteria for LN involvement have been proposed and used in several subsequent reports.34–37 Prognostically, the implication of a metastatic LN is also a controversial topic. LN involvement has been associated with a worse prognosis in a number of studies.20,36–39 However, several papers also report long-term survival in dogs with LN involvement where the primary tumor and the LN are treated to achieve local control using surgery with or without radiation therapy.14,34,35,40,41 Chemotherapy was used in many of the reported cases as well, although the protocols varied, and the added benefit of chemotherapy after local control was achieved cannot be proven via the retrospective noncontrolled studies that currently exist. To summarize the information in the literature, LN involvement may carry a worse prognosis, yet dogs can still have prolonged survival with adequate treatment of the primary tumor and the metastatic node, with chemotherapy potentially having a benefit as well. Assessment of any locoregional LNs, whether normal sized or enlarged, with cytology or histopathology is critical to determine the stage of the tumor and appropriate therapy for the patient.
Other staging tests are rarely positive and may have false positive results as well. In general, staging tests other than LN aspiration are recommended in patients who have negative prognostic factors associated with their MCT. Thoracic radiographs are indicated to evaluate the sternal LN if the mass is on the ventral abdomen, or to rule out other non-MCT diseases. MCTs metastasize so rarely to the lungs that radiographs are not indicated to evaluate for pulmonary spread. Buffy coat preparation to look for circulating mast cells is a quick and easy test, but it is both insensitive and nonspecific. The bone marrow may still be infiltrated in spite of a normal buffy coat, and dogs with skin disease, parvo virus, and nonmast cell illnesses often have positive buffy coats despite not having an MCT.42–44 Bone marrow evaluation for mast cell infiltration is positive in <5% of cases, even in cases with poorly differentiated tumors, and thus is not a staging test that is recommended as standard.45
Abdominal ultrasound (US) also rarely finds evidence of MCT metastasis, although US is needed to evaluate the sublumbar LNs if the MCT is on the back half of the patient. The benefit of FNA and cytology of ultrasonographically normal-appearing liver and spleen for staging of dogs with MCTs is another topic of ongoing debate. Mast cells can be found in moderate numbers in normal canine liver and spleen, and although increased numbers were found in splenic aspirates in 51 MCT-bearing dogs with normal spleens on US compared with 32 unaffected dogs, the splenic cytology did not correlate with systemic behavior.46 In another study, 52 dogs with MCTs underwent ultrasound evaluation and FNA with cytology of their liver and spleen.47 Cytologic criteria of MCT infiltration included clustering of well-differentiated mast cells, large numbers of well-differentiated mast cells, or mast cells with atypical morphology (pleomorphic and poorly granulated). The dogs were separated into two groups: those without MCT infiltration into either organ (n=42) and those with infiltration into one or both organs (n=10); survival time between these groups was significantly different at 733 days vs 34 days, respectively (P<0.0001). This dramatic difference in survival time supports that the cytologic findings were consistent with systemic MCT. Survival time based on US appearance alone in the dogs was not significant, suggesting that cytologic evaluation of the spleen and liver may be indicated for complete staging of dogs with MCT regardless of the ultrasonographic findings.47 More recently, 19 dogs with clinically aggressive grade II or III MCTs treated with vinblastine/lomustine chemotherapy were evaluated to determine the specificity and sensitivity of US findings in determination of MCT infiltration of the liver and/or spleen, using cytologic assessment to determine infiltration.48 Seven dogs had MCT infiltration, and the sensitivity of US for detection of infiltration was 43% for the spleen and 0% for the liver. Dogs with cytologic determination of infiltration of the spleen or liver had significantly shorter survival than dogs without infiltration (100 days vs 291 days and 100 days vs 276 days, respectively, P<0.0001), this clinical survival thus corroborating the cytologic finding of systemic disease. Due to the poor sensitivity of US in detection of cytologically confirmed organ infiltration and the decreased survival associated with identified infiltration, FNA and cytology of the spleen and liver are recommended for staging of dogs with MCT thought to be at high risk for metastasis.48