2.1 Case selection

A multi-institutional, retrospective study of client-owned dogs with a diagnosis of foot pad melanoma was performed. Cases were requested using the American College of Veterinary Internal Medicine Oncology electronic mailing list. Contributors were sent abstracting forms and cases were extracted at their respective institutions. A medical records search was performed from 2002 to 2020 at 11 academic and referral institutions. Dogs were included if they had a cytologic and/or histologic diagnosis of foot pad melanoma evaluated by a board-certified clinical or anatomic pathologist. Dogs with cutaneous, oral, digital, subungual or interdigital melanomas were excluded. To be eligible for inclusion, complete medical records and follow-up information for at least 8 weeks following diagnosis were required.

2.2 Medical records review

Information obtained from the medical records included the following: age, weight, sex, breed, clinical signs at diagnosis, date of diagnosis, tumour size, presence or absence of ulceration and foot pad location. Staging diagnostics and results were recorded, when available. Stage was determined according to the WHO staging system, however complete staging was not required for study inclusion. Tumour-specific variables, when available from the histopathology reports, included mitotic count (MC), completeness of excision and histochemical and immunohistochemical staining. MC was defined as the number of mitotic figures in 10 consecutive high-powered fields. If a dog had more than one MC reported, the highest MC was recorded for use in this data set. Surgical margins were classified as complete if no tumour cells were observed at the edges of the sample and classified as incomplete if any tumour cell were observed at the edges of the sample. As pathology reports were only reviewed, techniques related to margin assessment were not routinely reported or known. Types of local and adjuvant treatments such as non-steroidal anti-inflammatories, chemotherapy, radiation therapy and ONCEPT vaccine (Merial Limited, Duluth, GA) were recorded. Outcome information, including local recurrence, distant metastasis, date of death and cause of death was obtained from the medical record or the referring veterinarian when available.

2.3 Statistical analysis

Patient outcomes were described as progression-free interval (PFI) and median survival time (MST). The PFI was defined as the time from diagnosis to the time of local recurrence or metastasis. Dogs that died and did not have local recurrence or noted metastasis at the time of death were censored from PFI. MST was defined as the time between time of diagnosis and the time of death. Melanoma-specific survival was defined as the time of diagnosis and the time of death from malignant melanoma. Kaplan–Meier analysis were used to estimate and display the distribution of PFI and ST (Prism 7, GraphPad, San Diego, California). The cause of death was determined from the medical records or after follow-up with the referring veterinarian. If cause of death was unknown, then these deaths were attributed to melanoma. Dogs were censored from outcome data if they died from unrelated causes, were alive at the time of manuscript preparation or were lost to follow-up.

2.4 Cell line validation statement

This study did not involve cell lines; therefore, cell line validation work was not carried out.

3.1 Patient characteristics

Twenty dogs met the inclusion criteria from 11 institutions (Colorado State University (n = 3), The Ohio State University (n = 3), Veterinary Emergency and Referral Center of Hawaii (n = 3), Oregon State University (n = 2), University of California Davis (n = 2), Anderson Moores Veterinary Specialists (n = 1), Blue Pearl Rockville (n = 1), Blue Pearl Specialty and Emergency Pet Hospital (n = 1), Bridge Animal Referral Center (n = 2), Central Texas Veterinary Specialty and Emergency Hospital (n = 1) and Hope Veterinary Specialists (n = 1)). The median age at diagnosis was 10.5 years (range, 5.2–15 years). The median weight was 31.7 kg (range, 7–42 kg). Weight was not available in one case. Breeds represented included Labrador retriever (n = 5), mixed breed (n = 4), miniature schnauzer (n = 3) and one each of the following, Brittany spaniel, flat-coated retriever, French bulldog, giant schnauzer, golden retriever, Norwich terrier, rottweiler and standard schnauzer. There were 13 (65%) neutered males, 6 (30%) spayed females, and 1 (5%) intact male dog. Complete patient demographics are listed in Table 1. Clinical signs were reported and available in 19/20 (95%) cases. Lameness (n = 8, 40%) and appreciation of a foot pad mass (n = 8, 40%) were the most common clinical signs. Other clinical signs noted were malodor (n = 1, 5%) and chewing of the mass (n = 1, 5%).

Concurrent morbidities included low-grade cutaneous mast cell tumours removed at the time of foot pad melanoma surgery (n = 2), renal epithelial neoplasia (n = 1), hyperadrenocorticism and prior history of a cutaneous melanoma (n = 1), splenic mantle zone lymphoma, cholangiosarcoma and trichoepithelioma (n = 1), and multicentric lymphoma (n = 2); one dog with multicentric lymphoma was enrolled into an anti-B cell clinical trial.

3.2 Tumour location and gross characteristics

Foot pad melanoma location was reported in all dogs (Table 1). A singular foot pad site was affected in 19/20 dogs (95%). One dog had two separate foot pad sites affected concurrently. The most common pad affected were the metacarpal pads,¹⁰ metatarsal pad,⁷ digital pad³ and carpal pad.¹ Tumour size was recorded in 18 of 20 dogs. Accounting for the largest tumour diameter, the median pre-surgical tumour size was 12 mm (range, 2–22 mm). Ulceration of the tumour was documented in 13 cases (65%). Four dogs had no evidence of ulceration and three had no report of ulceration status.

3.3 Initial dog staging and tumour pathologic features

Initial baseline diagnostics performed included a complete blood count in 17/20 (85%), serum biochemistry in 17/20 (85%), and urinalysis in 2/20 (10%) dogs. Thoracic imaging was performed in 18/20 dogs (90%) predominately with radiography (17/18) and with computed tomography in one case. None of the dogs with thoracic imaging had evidence of thoracic metastatic disease at diagnosis. Abdominal ultrasound was performed in seven cases (35%). One dog had a heterogenous spleen, a hypoechoic area associated with the prostate, and a right renal mass. Fine needle aspirates performed revealed lymphoid hyperplasia of the spleen and a cystic epithelial neoplasm of the kidney. The prostatic aspirates were inconclusive due to haemodilution. No other significant abnormalities were documented on abdominal ultrasound for the other six cases. Other imaging diagnostics included radiographs and focal ultrasound of the affected foot pad in one dog. On ultrasound, a focal defect of the metacarpal pad was identified. There was concurrent swelling of the pad and a focal diffuse hyperechoic area deep to the skin surface. Radiographs of the foot were not available for review.

On initial examination, regional lymph node enlargement was observed in 11/20 dogs (55%). Prominent lymph nodes affected the ipsilateral side with superficial cervical⁷ and popliteal⁴ lymphadenomegaly. The regional lymph node was evaluated in 16/20 dogs (80%) by cytology and/or histopathology. Four dogs had confirmed metastatic disease of the lymph node; two were confirmed on cytology and two were confirmed on histopathology. Of the four dogs with pathology-confirmed lymph node involvement, two dogs had an enlarged regional lymph node on examination. The regional lymph node metastatic rate for dogs was 4/20 (20%) at the time of diagnosis. In one dog, subcutaneous nodules in the left axilla were identified on physical exam on initial presentation. Histopathology of the axillary nodules revealed a moderately differentiated malignant melanoma. On review, the lesion may have represented a de novo solitary event or subcutaneous metastasis effacing a local small lymph node. However, no nodal architecture was detected.

MM diagnosis was confirmed pre-operatively by cytology in five dogs and post-operatively or via biopsy in all 20 dogs. The MC of the primary tumour was reported in 19 cases (Table 2) The median MC was 10 mitoses per 10 hpf (range, 1–48). Only one dog had the MI reported as 2 mitoses per hpf and was not included in the median. Four cases had histochemical and immunohistochemical (IHC) staining performed to confirm a melanoma diagnosis. SOX-10, S100, Melan-A, Fontana-Masson, CD18 and PLN-2 stains were utilized at the discretion of the pathologist.

3.4 Surgery and adjuvant therapy

Surgical biopsy and/or excision of the primary tumour was documented in all 20 dogs (Table 2). Complete surgical margins were reported in seven dogs, incomplete surgical margins were reported in nine dogs and margin status was not available in four dogs. Revision surgery was performed in 5/9 (55.5%) dogs with incomplete surgical margins. Surgical margin evaluation was complete in 4/5 dogs with revision surgery when reported. Adjuvant non-steroidal anti-inflammatory (NSAID) use was recorded in one dog and used in two dogs for post-operative pain relief. Records of long-term NSAID was not available for the rest of the study population.

Eleven dogs (55%) received various adjuvant therapies following surgical excision (Table 2). Chemotherapy was utilized as adjuvant therapy in 6/20 dogs (30%). Chemotherapy agents included toceranib phosphate, dasatinib, carboplatin, cyclophosphamide, gemcitabine and trametinib. Rescue chemotherapy protocols were initiated following metastatic or progressive disease in three dogs. Following incisional biopsy, one dog was treated with four doses of bleomycin and one dose of carboplatin polymer electrochemotherapy to the foot pad site as a primary local treatment. This dog also received three doses of an autologous vaccine (Torigen Pharmaceuticals, Farmington, CT) from a lymph node sample.

Seven dogs received the ONCEPT melanoma vaccine. The vaccine was administered to four dogs as the only adjuvant therapy. The median number of doses was 4 (range, 2–8). Two dogs received radiation therapy (RT). One dog was treated with RT as the sole adjuvant therapy. This dog's treatment plan included three planned fractions of 8 Gy to the metastatic lymph node, but only received one fraction due to rapid and continuous clinical decline presumably due to progressive metastatic disease. The other dog received four fractions of 8 Gy for the treatment of an incompletely excised foot pad and regional lymph node prophylactically. Out of the total 20 dogs, 9 dogs (45%) did not receive adjuvant therapy.

3.5 Post-surgical restaging and dog outcome

Fourteen of twenty (70%) dogs were restaged following surgical resection at the discretion of the attending clinician. Restaging was not routine; the earliest time of restaging was 20 days after diagnosis and varied thereafter. Diagnostics performed included regional lymph node aspirates, thoracic radiographs, abdominal ultrasound, and thoracic and abdominal CT. Ten dogs were censored from PFI analysis. Four dogs were alive at the time of analysis with no evidence of recurrence or metastasis. Three dogs died and three dogs were lost to follow-up with no evidence of recurrence or metastasis. The median follow-up time for this group was 328 days (range 155–1118 days).

Ten dogs (50%) had progressive disease including suspected or confirmed local recurrence, development of metastasis to a distant site, metastasis to a regional lymph node(s), pulmonary nodules and heart nodules. Local recurrence occurred in three dogs. At the time of surgery, margins status was incomplete, complete and not reported respectively. The most common sites of metastasis included the lungs (n = 7), regional lymph node (n = 3) and heart (n = 2). The PFI was 101 days (range, 20–960 days) (Figure 1).

Eight dogs were censored from OST analysis; five dogs were alive at the time of manuscript analysis and three dogs were lost to follow-up. The median follow-up time was 557 days (range 155–690) for this category. The median survival time (MST) for death of any cause was 240 days (range, 25–479 days) (Figure 2). MST for melanoma-specific survival was 203 days (range, 25–387 days). For dogs receiving adjuvant therapy, the MST was 159 days (range, 25–387 days).

Post-mortem evaluation was performed in three dogs. Wide-spread metastasis was present in two cases. One dog was observed to have melanoma metastasis to the inguinal and iliac lymph nodes, the left ventricle of the heart, and lungs. Progressive lymphedema of the right hindlimb and difficulty breathing was appreciated leading up to euthanasia. The other dog had metastatic lesions of the lungs and right auricle. PLN-2 IHC showed strong cytoplasmic expression within the neoplastic cells of the lungs, however no staining was present of the right auricular mass. The third dog had an enlarged left superficial cervical and left popliteal lymph node twice the size of normal as well as a 2 cm lung nodule and 3–4 cm splenic body mass. On gross examination, the cut surfaces of the lymph nodes were black, however no further histopathology was performed.

Based on low case numbers and high censorship, prognostic factors were not assessed for significance.