Incorporation of sentinel lymph node mapping in dogs with mast cell tumours: 20 consecutive procedures

Authors

  • Deanna R. Worley

    Corresponding author
    1. Department of Clinical Sciences, Flint Animal Cancer Center, Colorado State University, Fort Collins, CO, USA
    • Correspondence address:

      D. R. Worley

      Flint Animal Cancer Center

      Colorado State University

      Fort Collins, CO, USA

      e-mail: deanna.worley@colostate.edu

    Search for more papers by this author

Abstract

The study hypothesis is that incorporation of sentinel lymph node (SLN) mapping in dogs presenting for mast cell tumour (MCT) removal would impact the recommended adjuvant therapy offered. Nineteen dogs were enrolled having either spontaneously occurring or incompletely excised MCTs. Staging included regional lymph node aspiration. SLN mapping was done with regional lymphoscintigraphy combined with intra-operative lymphoscintigraphy and blue dye. Twenty MCTs in 19 dogs were excised with SLN mapping. Eight dogs had SLNs different from the closest node. Twelve dogs had metastasis in extirpated SLNs, seven occurred in MCTs with a MI ≤ 5. No correlation was noted between patient stage and the c-KIT proto-oncogene. Because of SLN staging, 8 of 19 dogs were offered additional therapy that would have otherwise been excluded. Anatomic sampling of lymph nodes in dogs with MCTs does not accurately reflect which lymph nodes are most likely to be receiving the draining tumour lymph.

Introduction

Sentinel lymph node (SLN) mapping is an accurate physiologic method for finding which lymph node or nodes receive draining tumour lymph.[1-4] A positive SLN is a lymph node having presence of metastatic cells. A negative SLN receives draining tumour lymph but does not have the presence of metastasis. The significance of a negative SLN is that the rest of the lymphatic basin, as well as the rest of the patient, is unlikely to have advanced disease for tumours known to spread via locoregional progression.[1, 3, 4] A positive SLN has paramount impact on determination of clinical stage of disease as well as prognosis for a patient. Adoption of SLN mapping has revolutionized human surgical oncology since the practice of selective lymphadenectomy was pioneered by Dr. Donald L. Morton in 1991 for melanoma.[1-5] With the observation that truncal melanomas had unpredictable lymphatic drainage patterns and typically drained to only one to two lymph nodes, Dr. Morton pioneered in cats the first SLN mapping technique using vital dyes to reliably localize the draining nodes.[2, 5] Currently, the typical method of SLN mapping used in humans combines regional lymphoscintigraphy, intra-operative lymphoscintigraphy and intra-operative vital dye injection.[6]

Many benefits have resulted for human patients from utilization of SLN mapping or selective lymphadenectomy. These include decreased morbidity, decreased postoperative pain, improved mobility, decreased occurrence of lymphoedema and decreased iatrogenic numbness.[7] With SLN mapping, fewer lymph nodes are extirpated.[6] This improves pathologic analysis as more intensive investigations can be made on fewer extirpated lymph nodes, such as one to five SLNs instead of at least 10 lymph nodes removed with complete axillary dissection using breast cancer as the example.[7, 8] Patient care has also been improved following incorporation of SLN mapping through the phenomenon of stage shift as additional therapies are offered to more patients with identified lymph node metastasis than prior to SLN mapping.[9] Inclusion of SLN mapping in surgical oncology makes use of components of tumour biology, particularly tumour-induced lymphangiogenesis, which contributes to the atypical lymphatic drainage patterns found with SLN mapping.[10] The lymphatic system also facilitates metastasis as lymphatic flow is less turbulent and the high hyaluronic content is nourishing for metastasizing cells.[10]

Mast cell tumours (MCTs) are the most common cutaneous tumours in dogs and are characterized by locoregional lymphatic metastasis.[11, 12] Presence of metastatic mast cells in draining lymph nodes impacts the prognosis and therapy offered to dogs with MCTs. The mainstay of determining the stage of an MCT in a canine patient relies heavily upon fine-needle aspiration of anatomic nearby or regional lymph nodes.[12, 13] However, not all regional lymph nodes are amenable to routine needle aspiration due to either a typically small and non-palpable size or an inaccessible anatomic location. Thus, there is considerable potential that lymph node metastasis is significantly underestimated.[14, 15] Furthermore, descriptions standardizing the surgical procedure of how best to excise the more common grade 2 MCT omit any mention of surgical lymph node staging and lymphadenectomy.[16] To date, the at-risk lymph nodes in dogs with MCTs have not been fully or accurately investigated.[17]

The purpose of this study is to describe the outcome of combined SLN mapping procedures as performed consecutively by a single surgeon in dogs presenting for surgical excision of a cutaneous or subcutaneous MCT. It was hypothesized that lymph nodes will receive draining tumour lymph in unexpected locations, different from preoperative assessment of the anatomic regional lymph nodes. It was further hypothesized that SLN mapping in these dogs will impact the recommended adjuvant therapy offered.

Materials and methods

Client-owned dogs were prospectively enrolled at the Colorado State University Flint Animal Cancer Center from 1 January 2010 to 31 March 2011 having either a cutaneous or subcutaneous MCT(s) as diagnosed via cytology or having an incompletely excised surgical scar of a grade 2 or grade 3 MCT. Dogs enrolled were presented for definitive surgical excision of their spontaneously occurring MCT through the clinical oncology service at the Flint Animal Cancer Center. Presurgical staging was completed by attending clinicians prior to enrollment and consisted of physical examination, complete blood count, serum biochemistry, fine-needle aspiration and cytology of any bulky external tumour and/or review of a histopathology report if previously excised. Fine-needle aspiration of the regional lymph node was performed at the discretion of the receiving clinician prior to enrollment and not dependent for study enrollment. Additional staging was performed at the discretion of the receiving clinician dependent upon the location of the MCT, if any comorbidities were present, or if any suspected systemic clinical signs associated with an MCT were present such as gastrointestinal upset.

Dogs were premedicated with an injectable H1 blocker prior to lymphoscintigraphy and surgical manipulation of the MCT. Regional lymphoscintigraphy was performed with the animal under the effects of preanaesthetic medication or general anaesthesia. Regional lymphoscintigraphy was accomplished with approximately 0.5 mL of 125 microcuries filtered technetium sulphur colloid that was injected in four quadrants peritumourally. Images were captured on a GE Millenium VG single photon emission computer tomography (SPECT) gamma camera using 120-s dynamic image acquisition windows serially every 5 min until the first draining lymph node(s) were visualized. Cobalt markers were used for reference and image interpretation. Orthogonal views were obtained upon clinician discretion for enhanced lymph node localization. Ion chamber measurements were obtained of each patient prior to leaving the nuclear medicine suite.

If the patient was only premedicated for regional lymphoscintigraphy, then general anaesthesia was induced and each patient was prepared for surgery. Approximately 0.4 mL of 5 mg mL−1 sterile methylene blue diluted with sterile saline was injected peritumourally in four quadrants, similar to the technetium, 5 min prior to commencing surgery to allow for peritumoural lymphatic uptake of the blue dye. A NEO3000 handheld gamma probe (Ethicon Endo-Surgery, A Johnson & Johnson Company, Cincinnati, OH, USA) with a 14-mm probe and external collimator were used to obtain background counts of the patient in a location distant to the radioactive injection site. If background counts were less than or equal to 2% of maximum value of the selected count range, a default background count of 2% was used. Definitive excision of the MCT was performed. Used surgical instruments and supplies were discarded and surgical gloves were changed. Movement of the handheld gamma probe did not exceed 1 cm s−1 when localizing any radioactive lymph nodes within the predetermined lymphatic basin. Audible noise and visible counts from the gamma probe console directed incisions over lymphatic basins. Lymphadenectomy was performed of any lymph node that was either visibly blue upon direct inspection and/or having radioactive counts greater than two times background counts. A target count was made of each lymph node ex vivo on a perpendicularly directed gamma probe. Once a SLN was excised, any additional lymph nodes in the basin having radioactive counts 10% or greater of the SLN were also extirpated, such that approximately 90% of the radioactive counts in the basin were removed. All surgical supplies and the surgical suite were assessed for any radioactive contamination. Ex vivo tissues were inked at the surgical margins (primary tumour) and placed in a buffered 10% formalin solution and held in the nuclear medicine suite to allow for radioactive decay prior to routine histopathology.

MCT tissues were processed for routine histopathology and graded according to the Patnaik system.[16, 18] A mitotic index was described for all submitted tumours, and tumours were divided into groups of mitotic indexes ≤5 or >5.[19] Extirpated lymph nodes were sectioned in half longitudinally and a single cross-sectional longitudinal slice was used for lymph node analysis. If any difficulty was encountered following haematoxylin and eosin (H&E) stains in detecting for nodal metastasis, an additional toluidine blue stain was performed on the lymph node. c-KIT immunohistochemistry staining patterns (1 through 3) were determined on all MCT tissues for which original tissue was obtained as well as any c-KIT internal tandem duplication mutations in exons 8 and 11 through the Colorado State University Clinical Immunology Laboratory of the Department of Microbiology, Immunology, and Pathology.[20, 21]

Descriptive comparisons were made of regional lymph node assessments, selective lymphadenectomy, histologic tumour grade, tumour mitotic index, presence of histopathologic lymph node metastasis, c-KIT immunohistochemistry staining patterns and presence of c-KIT mutations.

Results

Nineteen consecutive dogs with 20 spontaneously occurring cutaneous or subcutaneous MCTs were enrolled for SLN mapping following client consent. There were 11 female spayed dogs, 7 male castrated dogs and 1 male intact dog. Ages of the dogs ranged from 3.5 to 13.3 years, average age 7.8. There were five mixed breed dogs; three boxers; three Labrador retrievers; two golden retrievers; two Jack Russell terriers and one each of Scottish terrier, cocker spaniel, pug and standard poodle.

MCTs ranged in size from 7 to 30 mm in diameter in 15 of 19 dogs. Four dogs presented with surgical scars of incompletely excised MCTs which ranged from 10 to 35 mm long. MCTs were located on the hind digit, metatarsus, hock, medial crus and stifle to which the regional lymph node used for preoperative assessment was the popliteal lymph node. MCTs were also located in the caudal abdominal mammary gland, the prepuce and inguinal space to which the regional lymph node for preoperative assessment was the inguinal lymph node. Three MCTs were located in the axilla, cranial thorax and lateral thorax to which the regional lymph node was considered the axillary lymph node. Two MCTs were located on the antebrachium (distal and dorsal) to which the prescapular lymph node was considered the regional lymph node for preoperative assessment. Six MCTs were located on the head and neck to which only the mandibular lymph node was available and accessible for preoperative assessment. These included three pinnal tumours, rostral maxillary lip, the nasal planum and medial canthus. The final MCT was located in the temporal region to which the parotid lymph node was subjectively considered the regional node. It could also be considered that the three pinnal tumours were subjectively anatomically closer to the parotid lymph node. Of these 19 dogs, four dogs had incompletely excised MCT surgical scars performed at other facilities, all grade 2 (3 of 4 dogs with mitotic indices of 0, 1 and not reported per 10 high-powered fields) or grade 2/3 MCTs (1 of 4 dogs with a mitotic index of 4 per 10 high-powered fields). The original grade 2/3 MCT, also described as a high grade 2 MCT, was not re-evaluated as such information was not going to change the need for surgical re-excision. One dog was enrolled with two cutaneous MCTs, one in the caudal abdominal mammary gland and one on the distal antebrachium.

Fine-needle aspiration of regional lymph nodes was done in nine of the dogs. Regional lymph nodes that were aspirated were the popliteal, prescapular and mandibular lymph nodes. Lymph node cytology was non-diagnostic in three, suggestive for possible metastasis in one, positive for MCT metastasis in two and negative for MCT metastasis in three dogs. Fine needle aspiration was not pursued in all cases either due to an inconvenient or inaccessible location of the regional lymph node, including the axillary lymph node, or due to a non-palpable or small size of the anatomically nearest node. A specific reason was not recorded in the medical record for each unpursued regional lymph node aspiration.

Lymphoscintigraphy was accomplished with premedication in 14 dogs and with general anaesthesia in 5 dogs. Dogs having regional lymphoscintigraphy performed under general anaesthesia had MCTs located on the pinna, lip, nasal planum, medial canthus and at the metatarsus. Ion chamber measurements recorded less than 2 mRem h−1 exposure for each dog, levels below the legal limit in the state of Colorado for radioactive quarantine (Colorado Radioactive Materials License Number Colo. 002-19).[22] SLNs were identified in 18 dogs with 19 MCTs via regional lymphoscintigraphy. Eight dogs had SLNs different from the regional lymph node. One dog with a pinnal MCT did not have an identifiable SLN via regional lymphoscintigraphy. The SLN was a 6-mm diameter lateral retropharyngeal lymph node localized with intra-operative lymphoscintigraphy.

Intra-operative blue dye mapping was performed in all dogs. No dogs experienced any acute reaction following peritumoural dye injection. Extirpated SLNs were blue in 18 dogs. Intra-operative lymphoscintigraphy localized ‘hot’ SLNs in all 19 dogs (Table 1). Unexpected ‘positive’ or metastatic SLNs were found in three dogs. These included an MCT at the right axilla with histopathologic metastatic disease in both the axillary and prescapular lymph nodes, an MCT at the left pinna with histopathologic metastatic disease found within the left prescapular lymph node and an MCT at the base of the right pinna with histopathologic metastatic disease in both the right parotid and prescapular lymph nodes. Two SLNs were purposely not extirpated; these were medial iliac lymph nodes associated with tumours at the medial crus (for which an inguinal SLN was removed) and at the hind fourth digit (for which a metastatic popliteal lymph node was removed).

Table 1. Summary of MCT staging per dog grouped by the anatomically nearest lymph nodes
DogMCT locationMCT sizeAnatomic nearest LNLN FNALN FNA metastaticSLN locationSLN ‘hot’aSLN ‘blue’bSLN metastaticPrevious MCT surgerycMCT gradecMitotic indexc-KIT IHC patternc-KIT exon 8c-KIT exon 11
  1. 1 = accomplished or was blue/hot/metastatic for LN and SLN columns; 0 = not done or was not blue/hot/metastatic for LN and SLN columns; # per 10 hpf = number of mitoses per 10 high power fields; ITD, internal tandem duplication mutation.

  2. a

    Radioactive signal detected via regional scintigraphy and intra-operative lymphoscintigraphy.

  3. b

    Lymph nodes had visible blue appearance following peritumoural injection of methylene blue.

  4. c

    Grade based on Patnaik system.

  5. d

    Medial iliac lymph node purposely not extirpated, status undetermined postoperatively.

  6. e

    Only single lymph node found via regional lymphoscintigraphy, two found intra-operatively with gamma probe within lymphatic basin.

  7. f

    Intra-operative radioactive counts within SLN slightly less than 2× patient background radioactivity.

  8. g

    Lymph node cytology was negative for lymph node metastasis but suggestive.

  9. h

    Weak radioactive signal from a probable parotid lymph node too small for intra-operative localization.

  10. i

    Several non-clustering, non-invasive mast cells within lymph node, pathologists unable to rule out metastasis, reported on single node.

  11. ?Alternate possible subjective assumptions of which LN might also be considered the anatomic nearest LN for a particular MCT.

AL stifleScar 3.5 cm longL popliteal00L inguinal110Grade 2Well differentiatedNot many cells in scar
      L popliteal110      
BL hind 4th digit3 cm diameterL popliteal1Non-diagnosticL popliteal111021 per 10 hpf2No ITDNo ITD
      Medial iliac1dd      
CR caudolateral hock2 cm diameterR popliteal11R popliteal111035–6 per 10 hpf2No ITDNo ITD
DR medial crus1 cm diameterR popliteal10R inguinal 11100Low, 21 per 10 hpf2No ITDNo ITD
      R inguinal 2e10      
      R medial iliac1dd      
ER metatarsusScar 1 cm longR popliteal1Non-diagnosticR popliteal 1111Grade 22 
      R popliteal 2e11      
FL inguinal0.7 cmL inguinal00L inguinal1100Low, 21 per 10 hpf1No ITDNo ITD
GPrepuce2 cm diameterR inguinal00R inguinal111020 per 10 hpf2No ITDNo ITD
HR 4th mammary gland1.5 cm diameterR inguinal00R inguinal111022 per 10 hpf1No ITDNo ITD
IL lateral thorax2 cm diameterL axillary, L inguinal?00L axillary110020–1 per 10 hpf1No ITDNo ITD
JR cranial thorax3 cm diameterR axillary00R axillary 11110Low, 21 per 5 hpf1No ITDNo ITD
      R axillary 2101      
KR axilla2.5 cm diameterR axillary00R axillary0.5f110Low, 21 per 10 hpf2No ITDNo ITD
      R prescapular0.5f11      
LR dorsal antebrachiumScar 2.5 cm longR prescapular10gR prescapular 1111Grade 22 
      R prescapular 2110      
HR distal antebrachium3 cm diameterR prescapular11R prescapular111021 per 10 hpf1No ITDNo ITD
ML maxillary rostral lip1.5 cm diameterL mandibular1non-diagnosticL medial retropharyngeal110022 per 10 hpf3No ITDNo ITD
NL nasal planum2 cm diameterL mandibular10L mandibular 1111029 per 10 hpf2No ITDNo ITD
      L mandibular 2e11      
OL pinna1 cm diameterL mandibular, L parotid?00L prescapular111032 per 10 hpf1No ITDNo ITD
PR medial canthus1 cm diameterR mandibular10R parotid0.5h000Low, 2<1 per 10 hpf1No ITDNo ITD
QR pinnaScar 1.5 cm longR mandibular, R parotid?00R lateral retropharyngeal110Grades 2–32Rare mitoses
RR base medial pinna9 mm diameterR mandibular, R parotid?00R parotid1100High5–10 per 10 hpf2No ITDITD present
      R prescapular111      
SR temporal2 cm diameterR parotid?, R mandibular?00R parotid 1110.5i1High, 25 per 10 hpf2No ITDNo ITD
      R parotid 211i      

Twelve dogs had positive or metastatic lymph nodes identified histopathologically (Table 1). Seven dogs had lymph node metastasis associated with MCTs having a mitotic index ≤5, and five of these dogs did not have any preoperative regional lymph node cytologic assessments performed. Three metastatic lymph nodes occurred with MCTs having a mitotic index >5, and two metastatic lymph nodes were associated with scar excision for which the original tumour tissue was not available but with reported mitotic indices of ≤5. Three dogs with lymph node metastases occurred in dogs with SLNs found in unexpected locations and without regional cytologic lymph node assessments.

Of the seven dogs not having metastasis detected histologically in the SLNs, the mitotic index was 2 per 10 high-powered fields in one dog and 1 or less per 10 high-powered fields in four dogs. Two dogs not having lymph node metastases received surgical excision of an MCT scar for which the original tissue was not available but with reported mitotic indices of four and not recorded per 10 high-powered fields (Table 1).

Evaluation of the c-KIT proto-oncogene was made in 16 tumours from 15 dogs from which the original MCT tissue was available (Table 1). Of these tumours, seven had an immunohistochemical staining pattern of one, with four of these tumours associated with lymph node metastasis. Eight tumours had an immunohistochemical staining pattern of two and were associated with lymph node metastasis in all but one tumour. Only one tumour had an immunohistochemical staining pattern of three and this was not associated with any lymph node metastasis within the SLN. In all 16 same tumours internal tandem mutations within exons 8 and 11 were assessed and found in only one dog, in exon 11, and was associated with concurrent lymph node metastasis.

A SLN could not be identified intra-operatively in one dog. This occurred in a dog with a medial canthus MCT and the parotid SLN positively identified on regional lymphoscintigraphy could not be successfully identified using blue dye or intra-operative scintigraphy. In one dog, premedication with an H1 blocker was inadvertently delayed until after regional lymphoscintigraphy and this dog experienced localized MCT degranulation. Five dogs experienced reversible temporary lymphoedema following lymph node extirpation. Affected dogs had popliteal or mandibular lymph nodes extirpated. As part of the learning curve with this case series, intra-operative false positive lymphoscintigraphy was associated with regional salivary tissue in two dogs that had some affinity for radionucleotide uptake and residual “hot” or radioactive lymph fluid within a nodal basin following lymphadenectomy in another dog.

When comparing regional lymphoscintigraphy with intra-operative lymphoscintigraphy, two lymph nodes in a mandibular, inguinal, and popliteal lymphatic basin were extirpated in three dogs for which regional lymphoscintigraphy suggested a single node. Two additional dogs had two lymph nodes in an axillary and prescapular lymphatic basin which were found with regional and intra-operative scintigraphy. Intra-operative lymphoscintigraphy was particularly helpful for dogs with non-palpable and small SLNs, especially in small breed dogs such as the Jack Russell terrier having normal sized axillary lymph nodes. Intra-operative lymphoscintigraphy was also helpful for SLNs located within fatty tissue and lymph nodes that could be removed through the primary MCT incision site, such as an axillary MCT with small axillary lymph nodes for which a separate lymph node incision site would have been made too caudoproximally.

Comparisons were made to preoperative lymph node cytologic assessments (Table 1). Two dogs with lymph node aspirates negative for cytologic metastasis correlated with negative SLNs. Two dogs with cytologically positive lymph node aspirates for metastasis were found to have positive SLNs. One dog with a negative preoperative lymph node aspirate had a positive SLN. Of the four dogs with non-diagnostic regional lymph node aspirates, three dogs had SLNs positive for metastasis. Of the 10 dogs without any preoperative regional lymph node cytologic assessment, six dogs had SLNs with MCT metastasis.

Discussion

The purpose of this study was to describe the outcome of SLN mapping in a cohort of dogs having MCTs with a hypothesis that SLN mapping would impact the recommended adjuvant therapy offered. Because of SLN staging, eight of 19 dogs were offered additional adjuvant therapy that would have otherwise been excluded if regional lymph node sampling was performed alone for staging in six dogs, if a low tumour mitotic index was assessed alone in another dog with a non-diagnostic lymph node aspirate, and if a non-diagnostic lymph node aspirate was relied upon in another dog having a previously excised grade 2 MCT for which original tissue nor a mitotic index were not available. Of the remaining four metastatic lymph nodes, two dogs had preoperative cytologically positive lymph node aspirates, one lymph node aspirate was suggestive of possible metastasis for a previously excised scar, and the other had a cytologically negative lymph node aspirate but a high mitotic index; for which all four of these cases would have likely been offered adjuvant chemotherapy regardless of SLN findings.

It was also hypothesized that SLNs will be in unexpected locations differing from the regional lymph nodes. Eight SLNs were found in locations different from the regional lymph nodes. Anatomic guidelines could be made regarding where to sample which nearest lymph node for any cutaneous or subcutaneous MCT, but considering cancer biology, specifically tumour-induced lymphangiogenesis, such an anatomic guide would still miss identification of a population of at-risk lymph nodes. Tumour-induced lymphangiogenesis is a process whereby a tumour induces surrounding lymphatic vessel formation which increases and facilitates access to the lymphatic circulation by metastatic cells and can account for the unpredictable locations of affected lymph nodes.[10] The lymphatic system further aids in metastasis as the high concentration of hyaluron is nourishing to tumour cells and possesses reduced fluid sheer stress.[10, 23] In spite of usefulness of cytologic lymph node staging for metastasis diagnosis,[24] current clinical behaviour of lymph node assessment as illustrated in this small case series did not result in thorough patient staging. It may be a reasonable assumption that it is not pervasive traditional clinical practice to try to aspirate every or multiple regional lymph nodes, especially not the axillary or parotid which can be very challenging to feel without enlargement or even the medial iliac lymph nodes deep within the caudal abdomen. Perhaps more emphasis should be placed on aggressive preoperative ultrasound-guided fine needle lymph node aspiration for multiple difficult to access lymph nodes if other means of SLN mapping are not to be employed.

Incorporation of SLN mapping resulted in temporary morbidity in five of 19 dogs, specifically reversible temporary lymphoedema or stage 1 lymphoedema. Stage 1 lymphoedema is defined by complete resolution of soft pitting oedema.[25] Stages 2 and 3 are irreversible. In the five affected dogs in this case series, lymphoedema was resolved by the time of suture removal in all cases. Irreversible lymphoedema is a significant problem in cancer patients, as there is not a cure, only symptomatic support.[25] A benefit of SLN mapping in women with breast cancer is the decreased risk for postoperative lymphoedema when compared to previously performed routine completion axillary lymph node dissections.[26] It is the opinion of the author that incorporation of SLN mapping improved lymph node extirpation in this series resulting in subjectively modest incisions overlying lymph node basins and expedited intra-operative lymph node localization.

There are multiple means of performing SLN mapping in dogs having MCTs as not all veterinary facilities may be licensed or equipped for radioactive isotope use. Intra-operative blue dye can be used alone, but the sensitivity of this method in women with breast cancer was less than when combined with lymphoscintigraphy.[27, 28] There are a number of blue dyes that can be used with the more common being isosulfan blue and methylene blue. There is about a 1–3% risk of allergic reactions, including anaphylaxis, seen in people with isosulfan blue dyes[29-31] and a smaller risk for skin reactions, including dermal necrosis, in people when methylene blue is injected intradermally.[31, 32] Neither of these risks has been reported in veterinary medicine nor seen in this study. It is the personal experience of the author that either of these dyes is useful intra-operatively. A non-radioactive method for localizing a tumour-draining lymphatic basin, as an alternative to regional lymphoscintigraphy, may be the use of contrast-enhanced ultrasonography.[33-36]

The success of SLN mapping hinges on a very low false negative rate such that the minutest number of histologically positive lymph nodes is inadvertently missed with mapping. The false negative rate is determined from evaluation of extirpated SLNs to complete regional lymphadenectomy, performed sequentially in individual patients. A false negative rate was not determined in this study. An improved false negative rate resulted in multiple breast cancer clinical trials when dual modalities of SLN mapping were combined as verified from meta-analysis.[37] The false negative rate has also been used in different multi-institutional breast cancer clinical trials as a means for assessing surgical competence. A surgeon may need to accomplish as many as 20–50 SLN procedures prior to achieving an acceptable false negative rate or competency.[38-40] Competency depends upon skill of the surgeon and there is no set number mandated for breast cancer surgeons.[37] Proficiency with SLN mapping comes with practice, similar to any other skill.

Differing from typical human mammary SLN mapping practice where regional technetium lymphoscintigraphy is performed up to a day preceding surgery,[41] regional lymphoscintigraphy was performed immediately preceding definitive tumour surgery in this case series and used less technetium.[42, 43] Extrapolating from SLN mapping in breast cancer surgery, it has been calculated a non-pregnant surgeon would need to perform 500–600 intra-operative lymphoscintigraphy procedures prior to reaching 10% of their annual radiation exposure limit, a level used for guiding personnel radiation exposure monitoring.[43, 44] Thus, many human hospitals do not require surgical suite monitoring, personnel extremity dosimeters or special practices in cleaning surgical suites.[43-46] However, the concept of ‘as low as reasonably achievable’ should not be underestimated when integrating intra-operative scintigraphy, developing hospital radiation protocols and educating hospital staff involved in those procedures. Monitoring of personnel, the hospital suites and surgical supplies was implemented at this institution with coordination of campus radiation safety officers.

Radioactivity of ex vivo histopathology tissue samples following SLN mapping has also been shown to be insignificant, however, it is not uncommon for samples to be stored for additional decay until radioactivity is equivalent to background levels.[47, 48] Most samples in this study required an overnight quarantine until radioactivity was equivalent to background activity.

This is a clinically relevant patient population as assessments were made of lymph nodes for preoperative staging on routine clinical patients presenting for surgical treatment of MCTs at a single veterinary centre specializing in comprehensive cancer therapy. This study could have been designed with an inclusion criteria mandating cytologic assessment of the regional lymph node for every enrolled patient, but this would bias study results differing from the reality of typical clinical practice. The goal of this study is to describe the incorporation of SLN mapping in a cohort of dogs with MCTs. It can be safely assumed that routine and consistent regional lymph node aspirations are not performed in all canine patients presenting with MCTs for surgical removal. Furthermore, which regional lymph nodes are deemed most at-risk for locoregional spread is subjective and typically limited to easily accessible peripheral lymph nodes such as the popliteal, prescapular, and mandibular lymph nodes or palpably enlarged lymph nodes.

This population of dogs was fairly uniform as all but two dogs had grade 2 MCTs and tumour size ranged from <1 to 3 cm in diameter. A challenge with MCTs is the currently used Patnaik grading system provides limited prognostic information for intermediate or grade 2 MCTs which also accounts for more than half of all MCTs evaluated, and there is considerable variation between pathologists when grading MCTs.[18] MCT grade does not predict lymph node involvement, but the clinical stage according to the World Health Organization for dermal MCTs is correlated to prognosis. Specifically, stages 0 and 1 without local lymph node involvement have a better prognosis than higher clinical stages of disease.[11] Another challenge of MCT management is the variable and unpredictable biologic behaviour of grade 2 MCTs.[12, 49, 50] Thus, accurate lymph node staging through incorporation of intentional SLN mapping is merited.

To the author's knowledge there is no current veterinary pathology consensus statement on what constitutes a positive lymph node metastasis with mast cell neoplasia. At this institution, such a determination is guided by whether cohesive groups of cells efface the normal lymph node architecture. It is not uncommon with MCTs to have a few mast cells in the lymph node subcapsular sinus. Thus, there can be difficulty in determining if metastasis is present with increasing numbers of cells in the subcapsular sinus, especially in the absence of cellular features of malignancy. Because the goal of this project was to evaluate incorporation of SLN mapping into clinical practice, additional efforts were not made in recruiting a single pathologist to review all samples. However, future MCT SLN mapping studies should consider minimizing histopathologic variability through recruitment of a single pathologist. Future studies can also include routine multilevel step-sectioning of the SLNs to more thoroughly evaluate for metastasis as opposed to the traditional single longitudinal cross-sectional image currently used, and for determining the false negative rate of pathologists reviewing SLNs.

There may not be a direct survival advantage associated with SLN mapping because lymph node metastases do not disrupt the function of a vital organ and therefore overall survival.[51] What is not clear is if there is an incubation period where metastatic cells grow in a lymph node prior to simultaneous spread to systemic sites. If there is an incubation period of growth in a lymph node, then there may be a small window for localized disease where SLN mapping could have a direct survival advantage.[51] The indirect survival advantage of SLN mapping is clearer as more patients are offered therapies that would not have been offered following more aggressive staging.

Perhaps, with more aggressive lymph node staging, more discriminating information can be obtained regarding the vast biologic behaviour variations associated with grade 2 MCTs. It is curious when considering a previous study demonstrating the predictive value of mitiotic index (MI) for overall survival with the results of this study where lymph node metastasis was detected in seven MCTs having a MI ≤ 5.[19] The previous study found a significantly improved median survival time for cutaneous MCTs with a MI ≤ 5.[19] Yet, the presence of lymph node metastasis in clinical staging is also correlated with prognosis; whereas, traditional MCT grading does not predict lymph node metastatic potential.[11] Future prospective efforts will need to focus on longitudinal survival outcomes of dogs receiving SLN staging, especially if additional dogs have concurrent lymph node metastasis with grade 2 MCTs and a MI ≤5. It is also interesting, even though dogs in this study were not followed long term, the marked variability found in c-KIT cytoplasmic staining and lymph node metastasis and only one dog of 15 had a c-KIT mutation which was associated with lymph node metastasis. It is unknown in this patient population if there will be any correlation with response to treatment or overall survival with either c-KIT presence or mutation status as demonstrated in other studies.[20, 21] Problems still persist in how to better discriminate which MCTs may benefit from a particular therapy or will demonstrate a more aggressive biologic behaviour.

In conclusion, based on the results of this study, anatomic sampling of lymph nodes in dogs with MCTs does not accurately reflect which lymph nodes are most likely to be receiving draining tumour lymph. Eight of 19 dogs (42%) had SLNs different from the regional lymph node. Eight of 19 dogs (42%) had additional treatment recommended that would have otherwise not been offered.

Acknowledgements

This study was supported by the Colorado State University Flint Animal Cancer Center.

Ancillary