Obtaining tumor-free margins is a primary goal of surgical treatment for most neoplasms. Determining the completeness of resection is accomplished by histologic assessment of the resected tissue to determine the proximity of neoplastic cells to the surgical margins. However, routine histologic assessment only allows microscopic assessment of a small fraction of the surgical margins.1 As a result, tumors in which the narrowest histologically tumor-free margin is within a few millimeters of the surgical margin (so-called narrow margins) may be indicative of complete resection or may indicate that the tumor extends to the surgical margin in a plane of tissue adjacent to the one that was assessed. To better classify the importance of narrow margins, tumors can be defined by a histologic safety margin.1 A histologic safety margin is the minimal histologically tumor-free margin needed to decrease or eliminate the odds of local recurrence. Although the histologic safety margin has been defined for some tumors of humans,2–4 its use has not been widely evaluated in veterinary medicine.
Histologic safety margins do not directly correlate with the measured margins obtained during surgery because of shrinkage of specimens after excision and during histologic processing, which is a major disadvantage of the technique. Tissue shrinkage is a well-known phenomenon and has been reported for the skin,5–11 esophagus,12 mammary gland tissue,13 oral cavity,14 liver,15 and colon or rectum16,17 of humans. A number of studies18–25 have been conducted to evaluate skin shrinkage after excision from dogs and cats. These studies have revealed that intrinsic properties of the skin, formalin fixation, and histologic processing all contribute to changes in lateral and deep dimensions of the excised skin. In human medicine, equations have been developed to estimate appropriate preexcisional surgical margins for cutaneous melanomas on the basis of the diameter of fixed tissue specimens.9,10 However, these equations have an accuracy of only 84% to 86.5% for determining the preexcisional surgical margin to within ± 3.5 mm, which results in a ≥ 15% chance of miscalculating appropriate surgical margins. This variability may be attributable to the fact that many results were based on the assumption that an entire specimen shrinks in a uniform manner.9,10,26 However, there is evidence that neoplastic tissue shrinks less after excision and during processing than does the surrounding apparently healthy tissue. This has been determined for human basal cell carcinomas,8 breast tumors,13 and esophageal carcinomas.12 If healthy skin shrinks more than does skin containing neoplastic cells, the histologically tumor-free margin will be less than the true tumor-free margin obtained during surgery. Furthermore, if shrinkage does not occur uniformly throughout an excised specimen, the amount of shrinkage of grossly or microscopically diseased areas of the tissue will need to be factored into equations used to correlate histologically tumor-free margins with surgical margins. The first step in determining such an equation would be to determine the amount of shrinkage for the marginal skin from the initial preexcision measurements to the time of histologic assessment. That information may make it possible to derive an equation that relates surgical margins and size of postfixation specimens.
To the authors' knowledge, no studies have been conducted to determine whether cutaneous neoplasms of dogs shrink less than do grossly normal skin. The purposes of the study reported here were to assess whether cutaneous MCTs excised with curative intent had uniform shrinkage of gross tumor and grossly normal skin surrounding the tumor and to determine an equation to estimate postexcisional margins from preexcisional measurements and vice versa. We selected MCTs because they are one of the most common cutaneous tumors of dogs.27–29 The experimental hypothesis was that gross tumor would shrink less than the surrounding grossly normal skin during all stages of postexcisional processing.
Materials and Methods
Sample
Tumors of client-owned dogs with a cytologic diagnosis of ≥ 1 cutaneous MCTs that were amenable to curative-intent surgical resection, regardless of location on the body or amount of underlying fatty tissue, were included in the study. Tumors of dogs admitted for revision of incompletely resected tumors or tumors in a location where a circumferential margin of skin could not be obtained (eg, adjacent to a mucocutaneous junction) were not included in the study. Multiple MCTs on the same dog were eligible for inclusion in the study as long as they were far enough apart that each tumor could be removed with complete and separate curative-intent margins. Owners provided informed consent prior to study onset. The study was approved by the Kansas State University Institutional Animal Care and Use Committee.
Experimental procedures
Data were collected from all enrolled dogs before surgery. Data included age, sex, breed, history of prior surgery, presence of other dermal or epidermal pathological conditions, and medications administered during the 2 weeks before surgery. Location of each MCT was recorded. Tumor staging was performed on all patients, and a preanesthetic evaluation was completed at the discretion of the attending clinician. Drugs used for premedication were selected at the discretion of the attending clinician and anesthesiologist. Premedication consisted of a combination of the following agents: hydromorphone hydrochloridea (0.08 to 0.1 mg/kg, SC or IM), butorphanol tartrateb (0.2 to 0.5 mg/kg, SC or IM), acepromazine maleatec (0.05 to 0.2 mg/kg, SC or IM), midazolamd (0.2 to 0.5 mg/kg, SC or IM), and dexmedetomidinee (0.005 mg/kg, SC or IM). For dogs that required general anesthesia for tumor removal, induction was performed with propofolf (3.0 to 6.0 mg/kg, IV), and anesthesia was maintained with isoflurane in oxygen. In addition, some dogs received drugs intraoperatively, including diphenhydramine hydrochloridea (2.0 to 2.5 mg/kg, SC or IM), ephedrine sulfated (0.1 mg/kg, IV), ketamine hydrochlorideg (0.6 mg/kg/h, IV as a constant rate infusion), lidocaineh (2.4 to 3.0 mg/kg/h, IV as a constant rate infusion), or fentanyl citratea (0.005 to 0.01 mg/kg/h, IV as a constant rate infusion). Dogs for which it was anticipated that duration of surgery would be > 90 minutes received a single perioperative dose of a broad-spectrum antimicrobial (cefazolin,a 22 mg/kg, IV; or ampicillin-sulbactam,i 22 mg/kg, IV). Some dogs also received a local infusion of lidocaine (1 to 4 mg/kg) into the wound bed after the skin tumor had been completely resected.
After each dog was sedated or anesthetized, the surgical area was aseptically prepared. The grossly visible boundary of the tumor was then marked with a sterile surgical marking pen and ruler.j The attending clinician used the sterile ruler to determine proposed lateral surgical margins around the tumor, and the entire proposed surgical margin was outlined with the sterile marking pen. Proposed surgical margins were drawn at the discretion of the attending clinician and represented a circular or an elliptical incision.
Before the initial incision for tumor removal was made, the attending clinician used the sterile ruler to collect 6 measurements (time 1). Measurements included the dorsoventral (proximodistal) length of the tumor (measurement A), the craniocaudal length of the tumor (measurement B), and the distance from the visible or palpable edge of the tumor to the proposed excisional skin margins in the dorsal or proximal (measurement C), ventral or distal (measurement D), cranial (measurement E), and caudal (measurement F) directions (Figure 1). Locations where measurements were obtained were marked with a sterile marker. After those measurements were obtained, the tumor was excised en bloc with a No. 10 scalpel blade by following previously marked circles or ellipses. Electrocautery was used only after the specimen was removed. All tumors were removed with a deep margin that contained subcutaneous tissue and extended to, but did not include, muscle or muscle fascia. The defect was closed with appropriate surgical techniques. Dogs received postoperative analgesics as deemed necessary.
Schematic representation of an excised MCT and the grossly normal marginal skin and measurements obtained for the specimen. Measurements included the dorsoventral (proximodistal) length of the tumor (measurement A); the craniocaudal length of the tumor (measurement B); and the distance from the visible or palpable edge of the tumor to the proposed excisional skin margins in the dorsal or proximal (measurement C), ventral or distal (measurement D), cranial (measurement E), and caudal (measurement F) directions.
Citation: American Journal of Veterinary Research 79, 7; 10.2460/ajvr.79.7.779
Immediately after the tumor was removed, the same 6 measurements were repeated on the excised tissue by 1 of 2 investigators (DAU or EEK), who used the same surgical ruler and the same marked locations around the specimen (time 2). The investigators who performed the measurements at time 2 were not aware of the measurements obtained at time 1. The excised specimen was laid flat, and the investigator was careful to ensure the specimen was not folded or stretched while collecting the measurements. Then, a partial-thickness dermal incision was made with a No. 15 scalpel blade along the marked line that delineated the gross margin of the tumor. This technique was based on the method described in a previous study.8 The cranial, caudal, dorsal or proximal, ventral or distal, and deep margins were marked with different colors of marking dyek to facilitate identification. After the marking dye dried, specimens were placed in neutral-buffered 10% formalin (ratio, 10 parts formalin to 1 part tissue).
Each specimen was removed from the formalin within 72 hours after surgery, and 1 investigator (JNH) used the same surgical ruler and the same marked locations to repeat the same 6 measurements (time 3). The investigator was not aware of results of previous measurements obtained at times 1 and 2. The investigator was careful to again ensure that each specimen was not folded or stretched while collecting the measurements.
After measurements were obtained at time 3, the specimens were routinely processed for histologic evaluation and stained with H&E stain. Cross sections of the cranial, caudal, dorsal or proximal, and ventral or distal margins of each specimen were obtained. Tumor grade (determined by use of the Patnaik30 and Kiupel31 scales), completeness of surgical excision, presence and degree of edema (subjectively categorized as none, mild, moderate, or severe), collagenolysis, and hyalinization of the stroma were recorded.
Statistical analysis
The percentage of shrinkage was calculated between time points for each of the 6 measurements by use of the following equation: (1 – [time b/time a]) × 100, where time a is the first of the 2 time points being compared, and time b is the latter of the 2 time points being compared. Shrinkage between each set of time points was calculated for each individual measurement (measurements A through F) for each specimen. Overall shrinkage of each specimen was calculated by calculating the mean for all measurements (ie, [A + B + C + D + E + F]/6).
In addition, tumor shrinkage for each specimen was the mean for measurements A and B, and shrinkage of the grossly normal marginal skin for each specimen was the mean for measurements C, D, E, and F.
Statistical analysis was performed with a commercially available software package.l Variables assessed included age, sex, tumor location, degree of edema, deep margin, time point (1, 2, or 3), measurement (A through F), and percentage of shrinkage between times 1 and 2, 2 and 3, and 1 and 3. Correlations between variables were analyzed by means of a repeated-measures ANOVA with compound symmetry. Normality of the errors was assessed and verified by means of a normal probability plot and histogram of errors. Normality was confirmed for all data. For multiple comparisons, the type I error rate was protected by means of the Bonferroni critical P; values of P < (0.05/number of comparisons) were considered significant. There was no adjustment for multiplicity. Data were reported as mean ± SE.
Results of the present study were used to derive an equation that used the total percentage of skin shrinkage to calculate the expected margin shrinkage from the preexcisional values as follows: preexcisional margin = postformalin margin/0.244. These calculated margins were then compared with the measured margin at time 3 by use of a commercially available software program,m and goodness of fit was determined by calculating a trendline and R2 value.
Results
Tumors were obtained from 19 client-owned dogs. There were 12 spayed females and 7 males (6 castrated and 1 sexually intact). Median age of the dogs was 9 years (mean, 8.9 years; range, 4 to 14 years). Median body weight was 25.1 kg (range, 5.7 to 41.6 kg). Breeds represented included 4 Boxers, 3 Labrador Retrievers, 2 Golden Retrievers, 1 Cairn Terrier, 1 Chesapeake Bay Retriever, 1 English Bulldog, 1 Jack Russell Terrier, 1 Lhasa Apso, and 1 Pug; there also were 4 mixed-breed dogs. None of the dogs had cutaneous pathological conditions or scars from previous surgery that were in a location that interfered with the margins for the present study. Before surgery was performed, a CBC and serum biochemical analysis were performed for 10 dogs; no major abnormalities were detected for any of these 10 dogs.
All 19 tumors were cytologically diagnosed MCTs, which was confirmed during histologic examination. Seventeen MCTs were completely resected. The 2 tumors that were incompletely resected extended to 1 and 2 lateral margins, respectively. Location of the MCT was classified as abdomen (n = 9), thorax (5), hind limb (3), or forelimb (2). No tumors of the head or neck were included in the study. Percentage of skin shrinkage was not significantly correlated with location of the tumor (P = 0.251), age (P = 0.636), or sex (P = 0.697). Deep margins ranged from 0.5 to 30 mm (median, 7 mm); deep margins were not significantly (P = 0.986) correlated with the percentage of skin shrinkage. All measurements (A through F) obtained at time 1 were significantly (P < 0.001) larger than those obtained at times 2 and 3. Measurements obtained at time 2 were not significantly (P = 0.631) different from those obtained at time 3.
Histologic assessment revealed that all 19 tumors were low-grade tumors (Kiupel scale). For the Patnaik scale, 18 of 19 tumors were grade II, and the remaining tumor was grade I. No collagenolysis or hyalinization was observed in any tumor. Degree of edema was not significantly (P = 0.161) correlated with the amount of shrinkage between any pair of time points.
The mean percentage change for each of the 6 measurements between each pair of time points was determined (Table 1). All mean measurements decreased immediately after excision (time 1 to time 2). Overall shrinkage of all specimens from time 1 to time 2 was 18.06%. Mean tumor shrinkage from time 1 to time 2 was 2.29%, whereas the mean shrinkage of the marginal skin from time 1 to time 2 was 25.95%. All measurements, except for A, increased slightly during formalin fixation (time 2 to time 3), with a mean change of 4.90%. Total mean percentage change (time 1 to time 3) of all measurements was 17.70%. Mean percentage of tumor shrinkage from time 1 to time 3 was 4.45%, and mean percentage shrinkage of the marginal skin from time 1 to time 3 was 24.42%.
Mean ± SE values for percentage change of 6 measurements obtained on MCT specimens and the surrounding skin margins of dogs at various time points.
| Measurement | Shrinkage from time 1 to time 2 (%) | Shrinkage from time 2 to time 3 (%) | Total shrinkage (%) |
|---|---|---|---|
| A | 0.44 ± 7.88 | 4.50 ± 5.44 | 7.01 ± 6.88 |
| B | 4.14 ± 3.35 | −3.06 ± 4.18 | 1.89 ± 4.37 |
| C | 34.71 ± 3.01 | −13.15 ± 11.51 | 30.61 ± 3.08 |
| D | 19.18 ± 8.16 | −4.33 ± 8.27 | 18.75 ± 8.41 |
| E | 27.19 ± 6.67 | −5.12 ± 9.71 | 25.64 ± 9.05 |
| F | 22.71 ± 5.34 | −8.25 ± 11.83 | 22.67 ± 5.34 |
| Overall | 18.06 ± 6.57 | −4.90 ± 8.85 | 17.70 ± 6.81 |
Specimens were measured before the initial incision for tumor removal (time 1), after excision (time 2), and after fixation in neutral-buffered 10% formalin (time 3). Percentage of shrinkage was calculated between time points for each of the 6 measurements by use of the following equation: (1 – [time b/time a]) X 100, where time a is the first of the 2 time points being compared, and time b is the latter of the 2 time points being compared. Overall shrinkage of each specimen was calculated by calculating the mean for all measurements (ie, [A + B + C + D + E + F]/6).
See Figure 1 for remainder of key.
Mean differences in percentage change from time 1 to time 2 as well as from time 1 to time 3 between every pair of measurements were determined (Table 2). By use of a Bonferroni correction, the only significant differences were for the change from time 1 to time 2 and from time 1 to time 3 between measurements A and C (P < 0.001) and measurements B and C (P < 0.001). There was also a significant (P = 0.025) difference from time 1 to time 2 between measurements C and F. However, when measurements A and B (measurements of the gross tumor) were combined and compared with the combined measurements C through F (measurements of the marginal skin), there was a significant (P < 0.001) difference between times 1 and 2 and between times 1 and 3. No significant differences were noted between times 2 and 3 for any measurements.
Comparison between each pair of measurements for the percentage change in shrinkage of MCTs and grossly normal marginal skin of dogs at various time points.
| Difference in shrinkage from time 1 to time 2 | Difference in shrinkage from time 1 to time 3 | |||
|---|---|---|---|---|
| Measurements compared | Mean difference (%) | P value* | Mean difference (%) | P value* |
| A and B | 1.76 | 0.802 | −1.19 | 0.834 |
| A and C | 32.24 | < 0.001 | 25.63 | < 0.001 |
| A and D | 17.91 | 0.222 | 13.11 | 0.034 |
| A and E | 27.94 | 0.077 | 22.06 | 0.163 |
| A and F | 24.91 | 0.288 | 21.35 | 0.147 |
| B and C | 30.51 | < 0.001 | 26.82 | < 0.001 |
| B and D | 16.15 | 0.012 | 14.30 | 0.022 |
| B and E | 26.18 | 0.044 | 23.25 | 0.118 |
| B and F | 22.97 | 0.19 | 22.54 | 0.105 |
| C and D | −14.36 | 0.067 | −12.52 | 0.125 |
| C and E | −4.33 | 0.019 | −3.57 | 0.024 |
| C and F | −7.36 | 0.003 | −4.28 | 0.028 |
| D and E | 10.03 | 0.594 | 8.95 | 0.458 |
| D and F | 6.82 | 0.212 | 8.24 | 0.493 |
| E and F | −3.21 | 0.472 | −0.71 | 0.954 |
| A + B and C + D + E + F | 24.83 | < 0.001 | 21.81 | < 0.001 |
Values were considered significant at P < 0.003 (repeated-measures ANOVA followed by a Bonferroni correction).
See Table 1 for remainder of key.
Mean shrinkage of the marginal skin (24.42%) was used to derive an equation to determine marginal shrinkage, as follows: postformalin margin = 0.244 × preexcisional margin. This equation was used to calculate margins (measurements C through F) for all specimens; calculated values were compared with measured values, and a line of best fit was determined (Figure 2). The R2 for this correlation was 0.859. The median difference between the calculated and measured values was 9.06 (range, 0.17 to 45.12). It was found that 81.6% of calculated values differed from the measured values by greater than ± 3.5 mm, which was the margin of error cited in studies9,10 of humans; thus, accuracy of the equation to estimate postexcisional margins from preexcisional measurements was only 18.4%.
Graph of the correlation between measured shrinkage and shrinkage calculated by use of the following equation: preexcisional margin = postformalin margin/0.244. Each symbol represents the difference between the calculated and measured skin shrinkage for distance from the visible or palpable edge of the tumor to the proposed excisional skin margins in the dorsal or proximal (measurement C), ventral or distal (measurement D), cranial (measurement E), and caudal (measurement F) directions for each specimen (76 measurements total). The dotted line represents the linear regression line of the data points with an R2 value of 0.859. See Figure 1 for remainder of key.
Citation: American Journal of Veterinary Research 79, 7; 10.2460/ajvr.79.7.779
Discussion
A safety margin is the margin around a tumor required to decrease or eliminate the odds of local recurrence; it can be a surgical or histologic margin.1 Recommendations exist regarding surgical safety margins for cutaneous MCTs of dogs, with 2 cm being the most commonly accepted margin.32–34 However, attempts to determine a histologic safety margin for MCTs have provided conflicting results.26,35 A histologically tumor-free margin of 1, 2, and 5 mm has been cited as indicative of complete resection,36–39 but these margins are not correlated with local tumor recurrence rates; therefore, they do not represent true histologic safety margins. In 1 study,26 a histologic safety margin of 10 mm resulted in a risk of local recurrence or metastasis of 0%. However, investigators of another study35 found no histologic safety margin for cutaneous MCTs of dogs, with a mean histologically tumor-free margin of 4 mm for tumors that recurred and tumors that did not recur. However, for a histologic safety margin to be clinically useful, it must be directly related to surgical margins, which would require an equation that accounts for various factors, such as tumor type and degree of skin shrinkage of the tumor margins.
One purpose of the study reported here was to assess whether cutaneous MCTs excised with curative intent had uniform shrinkage between the gross tumor and grossly normal skin surrounding the tumor. Results of the present study indicated that grossly visible MCTs shrink significantly less than do the surrounding grossly normal skin when tissues are surgically removed and fixed in formalin. Therefore, the experimental hypothesis was accepted. Most of the shrinkage occurred immediately after a specimen was excised. This ostensibly is attributable to intrinsic properties of skin (eg, elasticity and mechanical creep).25 The overall amount of shrinkage of the entire specimen in the present study was 17.70%, which is within the range (13% to 42%) reported in other studies.20–24 However, tissue shrinkage was not uniform within a specimen. Mean tumor shrinkage was only 4.45%, and shrinkage of the marginal skin was 24.42%. The finding that formalin fixation caused a slight increase in measured values was slightly unexpected, but that phenomenon has been reported previously in human11 and veterinary20 medicine. The reason for this expansion is unknown, but it may have been attributable to unknown interactions between formalin and the tissue or to variability in measurements obtained at the various time points. However, the change in skin dimensions as a result of formalin fixation was not significant, which indicated a negligible effect of formalin fixation on skin shrinkage; that result is in agreement with findings of other studies.20,24,25
The second purpose of the study reported here was to determine an equation to estimate preexcisional margins on the basis of postexcisional measurements, and vice versa. Investigators attempted to determine an equation to estimate preexcisional surgical margins from the diameter of fixed tissue specimens of humans, but accuracy of the equation was only 85%.9,10 Those authors based the equation on the assumption that uniform shrinkage occurred throughout the entire specimen. However, the present study revealed that the percentage of shrinkage of the marginal skin was more than the percentage of shrinkage of the total specimen. As a result, shrinkage of marginal skin should be used when creating an equation to determine preexcisional margins on the basis of results for a formalin-fixed specimen. We attempted to create a simple equation to determine the preexcisional measurements on the basis of the mean shrinkage for specimens of the present study. There was a strong positive association (R2 = 0.859) between values calculated by use of the equation and the measured values, but use of a margin of error of ± 3.5 mm, which was reported in 2 studies9,10 of humans, resulted in only 18.4% of measurements calculated by use of the equation being within the acceptable range.
Several other variables can affect skin shrinkage and may need to be accounted for in any equation. A previous study25 by one of the authors revealed no effect of size, tension lines, or location on the percentage of skin shrinkage, but other studies20–22 have found that size and location may exert an effect, with less shrinkage for smaller specimens and specimens obtained from the head and neck. Shape of the resected tissue may also exert an influence on skin shrinkage. Investigators of a study5 of cutaneous tumors in humans found that the long axis of fusiform tissue samples tended to shrink more when the preexcision long axis measurement was larger or the preexcision short axis measurement was smaller. However, this result was not corroborated by another study6 of specimens obtained from humans.6 Because specimens in the present study ranged from circular to elliptical on the basis of surgeon preference, it is unclear whether this may have had an effect on the results. Location of tumor did not affect the percentage of skin shrinkage in the present study; however, no specimens from the head or neck were included. No effect of age or sex was found on percentage of shrinkage. The latter finding is consistent with results of another study.20 To the authors' knowledge, the effect of age on skin shrinkage has not been assessed in veterinary medicine, although there are conflicting results for studies5, 6, 9,11 of humans. Other variables, such as breed, have not been assessed and should be evaluated in future studies.
For the study reported here, the lateral margins obtained during surgery were determined by the attending surgeon. Current recommendations for lateral margins during MCT resection range from 2 to 3 cm26,33,34 to a margin equivalent to the widest margin of the tumor (proportional margins technique).36 The technique used in the present study differed among surgeons, which resulted in minimal lateral margins that ranged between 0.5 and 3 cm. The effect of differences in the size of the lateral margin on skin shrinkage is not known. Investigators of studies in veterinary22 and human5,9 medicine have correlated the size of the excised specimen to the degree of shrinkage, with larger specimens shrinking to a greater degree than smaller ones. However, investigators of 1 study25 of canine cadavers found no effect of specimen size on the degree of shrinkage. Whether size of the specimens in the present study served as a confounding variable is unknown, although there was consistently more shrinkage of the grossly normal tissue surrounding the tumor than for the gross visible tumor for each specimen, regardless of the surgical margin obtained.
For the study reported here, the deep margin included subcutaneous tissue to the level of, but not including, deep muscle fascia. This technique, although consistently used at the authors' veterinary hospital, does not meet the current dogma of obtaining at least 1 fascial plane deep to an MCT.33,34,36 Exclusion of a deep fascial plane did not affect the completeness of surgical excision in the present study because none of the specimens had a tumor that extended to the deep margin. In addition, inclusion of a deep fascial layer had no influence on the degree of skin shrinkage in excised skin specimens from dogs of another study.21 Because of the variability in the depth of excision, some specimens contained minimal amounts of subcutaneous tissue whereas others had up to 30 mm of deep adipose tissue. To the authors' knowledge, no studies have been conducted to assess the effect of the thickness of subcutaneous adipose tissue on the degree of skin shrinkage; however, no significant correlation was found between the deep margin and degree of skin shrinkage in the study reported here. Additional studies are needed to investigate this further.
An attempt was made to correlate histologic variables to the percentage of skin shrinkage. Multiple variables were assessed, including tumor grade and completeness of excision. In addition, MCTs are unique in that the cytokines released by mast cells and eosinophils within the tumor can cause substantial variability in the surrounding stroma, including various degrees of collagen breakdown (collagenolysis), intercellular edema, and protein deposition (hyalinization).40 These 3 variables were assessed. Unfortunately, the MCTs of the present study were uniform with regard to grade and had no evidence of collagenolysis or hyalinization, so the effect of these variables on shrinkage could not be assessed. Edema and completeness of excision were not significantly correlated with the amount of shrinkage.
The present study had several limitations. To prevent bias in the recording of measurements, a different observer performed measurements at each of the 3 time points, and they were not aware of previously obtained measurements. This could have introduced intraobserver variability that would lead to differences in measurements related to observer and not to true specimen shrinkage. However, this same variability occurs in practice situations because the excised sample is rarely evaluated by the same individual at the time of surgery and at the time of histologic processing. We did not account for the effect of histologic processing (eg, trimming, embedding, and sectioning) on skin shrinkage. Investigators of a previous study22 of canine cadavers found that these steps caused a formalin-fixed specimen to shrink an additional 5.1% to 19.5%. Furthermore, the delineation between tumor and surrounding skin was made on the basis of gross visual observation. Most tumors had microscopic evidence that a tumor extended beyond the grossly visible margins, although this was not specifically assessed in the present study. The effect of this microscopic disease on the amount of shrinkage within grossly normal skin margins is not known.
It is also important to mention that the histologically tumor-free margin and histologic safety margin are dependent on the amount of microscopic spread of neoplastic cells into the grossly normal skin margins. As such, correlation of a histologic safety margin with surgical margins requires at least 2 steps. First, variables affecting the amount of macroscopic shrinkage of the marginal skin must be determined, as was attempted in the study reported here. Second, the microscopic spread of neoplastic cells into marginal tissues must be determined. This latter variable was not assessed in the present study and likely will differ among tumor types. Results of the present study of MCTs should not be extrapolated to other tumor types that may have different effects on shrinkage or have a greater or lesser degree of microscopic extension into surrounding tissues.
A final limitation was the low number of specimens included in the study. This precluded the assessment of some potential variables and resulted in a higher probability of a type I or type II error.
For the study reported here, the amount of shrinkage within the grossly visible MCTs was less than that of the grossly normal skin margins. This resulted in the fact that specimens did not shrink in a uniform manner, which confounded attempts to derive an equation to calculate skin shrinkage on the basis of total specimen shrinkage. Although the percentage of marginal skin shrinkage may be more useful for creating such an equation, an accurate equation to correlate surgical margins and postexcisional specimen dimensions could not be derived because of variability in the shrinkage among specimens. Future studies should include an assessment of other factors that may contribute to skin shrinkage and incorporate these factors into a revised equation. Until a reliable equation exists to correlate the 2 margin types, veterinary surgeons and pathologists must be careful that they not extrapolate surgical margins from histologic margins or vice versa.
Acknowledgments
The authors thank Dr. Joe Hauptman for statistical assistance.
ABBREVIATIONS
| MCT | Mast cell tumor |
Footnotes
West-Ward Pharmaceutical Corp, Eatontown, NJ.
Torbutrol, Zoetis Inc, Parsippany, NJ.
Vet One, Boise, Idaho.
Akorn Pharmaceuticals, Lake Forest, Ill.
Dexdomitor, Zoetis Inc, Parsippany, NJ.
Propoflo, Zoetis Inc, Parsippany, NJ.
Ketaset, Zoetis Inc, Parsippany, NJ.
2% lidocaine hydrochloride, Hospira, Lake Forest, Ill.
Unasyn, Pfizer Inc, New York, N Y.
Cardinal convertors surgical marking pen, Cardinal Health Inc, Dublin, Ohio.
Davidson marking system, Bradley Products Inc, Bloomington, Minn.
SAS software, version 9.3, SAS Institute Inc, Cary, NC.
Excel, version 2013, Microsoft Corp, Redmond, Wash.
References
1. Kopke L, Bastos J, Filho J, et al. Safety margin—an old and relative concept. An Bras Dermatol 2005;80:279–286.
2. Breuninger H, Dietz K. Prediction of subclinical tumor infiltration in basal cell carcinoma. J Dermatol Surg Oncol 1991;17:574–578.
3. Pascal RR, Hobby LW, Lattes R, et al. Prognosis of “incompletely excised” versus “completely excised” basal cell carcinoma. Plast Reconstr Surg 1968;41:328–332.
4. Dixon AY, Lee SH, McGregor DH. Histologic features predictive of basal cell carcinoma recurrence: results of a multivariate analysis. J Cutan Pathol 1993;20:137–142.
5. Dauendorffer JN, Bastuji-Garin S, Guero S, et al. Shrinkage of skin excision specimens: formalin fixation is not the culprit. Br J Dermatol 2009;160:810–814.
6. Hudson-Peacock MJ, Matthews JN, Lawrence CM. Relation between size of skin excision, wound, and specimen. J Am Acad Dermatol 1995;32:1010–1015.
7. Gardner ES, Sumner WT, Cook JL. Predictable tissue shrinkage during frozen section histopathologic processing for Mohs micrographic surgery. Dermatol Surg 2001;27:813–818.
8. Blasdale C, Charlton FG, Weatherhead SC, et al. Effect of tissue shrinkage on histological tumour-free margin after excision of basal cell carcinoma. Br J Dermatol 2010;162:607–610.
9. Golomb FM, Doyle JP, Grin CM, et al. Determination of preexcision margins of melanomas from fixed tissue specimens. Plast Reconstr Surg 1991;88:804–809.
10. Silverman MK, Golomb F, Kopf A. Verification of a formula for determination of preexcision surgical margins from fixed tissue melanoma specimens. J Am Acad Dermatol 1992;27:214–219.
11. Kerns MJJ, Darst MA, Olsen TG, et al. Shrinkage of cutaneous specimens: formalin or other factors involved? J Cutan Pathol 2008;35:1093–1096.
12. Siu KF, Cheung HC, Wang J. Shrinkage of the esophagus after resection for carcinoma. Ann Surg 1986;203:173–176.
13. Yeap BH, Muniandy S, Lee S-K, et al. Specimen shrinkage and its influence on margin assessment in breast cancer. Asian J Surg 2007;30:183–187.
14. Mistry RC, Qureshi SS, Kumaran C. Post-resection mucosal margin shrinkage in oral cancer: quantification and significance. J Surg Oncol 2005;91:131–133.
15. Rutherford EE, Karanjia ND. The measurement of liver resection margins. HPB (Oxford) 2004;6:18–20.
16. Eid I, El-Muhtaseb MS, Mukherjee R, et al. Histological processing variability in the determination of lateral resection margins in rectal cancer. J Clin Pathol 2007;60:593–595.
17. Goldstein NS, Soman A, Sacksner J. Disparate surgical margin lengths of colorectal resection specimens between in vivo and in vitro measurements. Am J Clin Pathol 1999;111:349–351.
18. Risselada M, Mathews KG, Griffith E. Effect of feline skin specimen preparation on postexcision and postfixation tissue shrinkage. J Feline Med Surg 2016;18:970–975.
19. Jeyakumar S, Smith AN, Schleis SE, et al. Effect of histologic processing on dimensions of skin samples obtained from cat cadavers. Am J Vet Res 2015;76:939–945.
20. Miller JL, Dark MJ. Evaluation of the effect of formalin fixation on skin specimens in dogs and cats. Peer J 2014;2:e307.
21. Reimer SB, Seguin B, DeCock HE, et al. Evaluation of the effect of routine histologic processing on the size of skin samples obtained from dogs. Am J Vet Res 2005;66:500–505.
22. Reagan JK, Selmic LE, Garrett LD, et al. Evaluation of the effects of anatomic location, histologic processing, and sample size on shrinkage of skin samples obtained from canine cadavers. Am J Vet Res 2016;77:1036–1044.
23. Risselada M, Mathews KG, Griffith E. Surgically planned versus histologically measured lateral tumor margins for resection of cutaneous and subcutaneous mast cell tumors in dogs: 46 cases (2010–2013). J Am Vet Med Assoc 2015;247:184–189.
24. Risselada M, Mathews KG, Griffith E. The effect of specimen preparation on post-excision and post-fixation dimensions, translation, and distortion of canine cadaver skin-muscle-fascia specimens. Vet Surg 2016;45:563–570.
25. Upchurch DA, Malenfant RC, Wignall JR, et al. Effects of sample site and size, skin tension lines, surgeon, and formalin fixation on shrinkage of skin samples excised from canine cadavers. Am J Vet Res 2014;75:1004–1009.
26. Schultheiss PC, Gardiner DW, Rao S, et al. Association of histologic tumor characteristics and size of surgical margins with clinical outcome after surgical removal of cutaneous mast cell tumors in dogs. J Am Vet Med Assoc 2011;238:1464–1469.
27. Misdorp W. Mast cells and canine mast cell tumours. A review. Vet Q 2004;26:156–169.
28. Priester WA. Skin tumors in domestic animals. Data from 12 United States and Canadian colleges of veterinary medicine. J Natl Cancer Inst 1973;50:457–466.
29. O'Keefe DA. Canine mast cell tumors. Vet Clin North Am Small Anim Pract 1990;20:1105–1115.
30. Patnaik AK, Ehler WJ, MacEwan EG. Canine cutaneous mast cell tumor—morphologic grading and survival time in 83 dogs. Vet Pathol 1984;21:469–474.
31. Kiupel M, Webster JD, Bailey KL, et al. Proposal of a 2-tier histologic grading system for canine cutaneous mast cell tumors to more accurately predict biological behavior. Vet Pathol 2011;48:147–155.
32. Séguin B, Leibman NF, Bregazzi VS, et al. Clinical outcome of dogs with grade-II mast cell tumors treated with surgery alone: 55 cases (1996–1999). J Am Vet Med Assoc 2001;218:1120–1123.
33. Fulcher RP, Ludwig LL, Bergman PJ, et al. Evaluation of a two-centimeter lateral surgical margin for excision of grade I and grade II cutaneous mast cell tumors in dogs. J Am Vet Med Assoc 2006;228:210–215.
34. Simpson AM, Ludwig LL, Newman SJ, et al. Evaluation of surgical margins required for complete excision of cutaneous mast cell tumors in dogs. J Am Vet Med Assoc 2004;224:236–240.
35. Donnelly L, Mullin C, Balko J, et al. Evaluation of histological grade and histologically tumour-free margins as predictors of local recurrence in completely excised canine mast cell tumours. Vet Comp Oncol 2015;13:70–76.
36. Pratschke KM, Atherton MJ, Sillito JA, et al. Evaluation of a modified proportional margins approach for surgical resection of mast cell tumors in dogs: 40 cases (2008–2012). J Am Vet Med Assoc 2013;243:1436–1441.
37. Weisse C, Shofer F, Sorenmo K. Recurrence rates and sites for grade II canine cutaneous mast cell tumors following complete surgical excision. J Am Anim Hosp Assoc 2002;38:71–73.
38. Murphy SE, Sparkes AH, Smith KC, et al. Relationships between the histologic grade of cutaneous mast cell tumours in dogs, their survival and the efficacy of surgical resection. Vet Rec 2004;154:743–746.
39. Scarpa F, Sabattini S, Marconato L, et al. Use of histologic margin evaluation to predict recurrence of cutaneous malignant tumors in dogs and cats after surgical excision. J Am Vet Med Assoc 2012;240:1181–1187.
40. Gross T, Ihrle P, Walder E, et al. Mast cell tumors. In: Skin disease of the dog and cat: clinical and histopathologic diagnosis. 2nd ed. Ames, Iowa: Blackwell, 2005;853–865.
