Correlation Between Preoperative Magnetic ResonanceImaging and Surgical Margins with Modified Mohs forDermatofibrosarcoma Protuberans

CARLOS SERRA-GUILLEN, MD,� ONOFRE SANMARTIN, MD,� BEATRIZ LLOMBART, MD,�

EDUARDO NAGORE, MD,� CARLOS DELTORO, MD,y ISABEL MARTIN, MD,y

RAFAEL BORELLA-ESTRADA, MD,� CELIA REQUENA, MD,� ANTONIO MARTORELL-CALATAYUD, MD,�

JOSE CERVERA, MD,y AND CARLOS GUILLEN, MD�

BACKGROUND Dermatofibrosarcoma protuberans (DFSP) is characterized by asymmetrical and poorlydefined growth. Magnetic resonance imaging (MRI) has been proposed for the delimitation of thistumor.

OBJECTIVES To study the utility of MRI in evaluating the depth of infiltration in DFSP and to comparethe efficiency of clinical palpation with that of MRI in delimiting the invasiveness of DFSP.

METHODS Observational, prospective study of DFSP cases. The MRI scans for all cases were comparedwith the exact histological infiltration plane obtained using modified Mohs micrographic surgery (MMS).

RESULTS Forty-three DFSPs were included: 22 primary, nine recurrent, and 12 extirpated with positive mar-gins. Sensitivity for detecting deep invasion was 58% on examination using palpation and 67% using MRI.

CONCLUSIONS We present the largest series of DFSP cases studied using MRI published to date. Inprimary cases, MRI has greater sensitivity than palpation for detecting depth of infiltration (67% vs 58%).MRI seems to be useful in primary DFSP in locations other than the head, neck, and upper part of thethorax. MRI is not useful for confirming tumor persistence in extirpated DFSP with positive margins orfor studying lateral extension in primary DFSP.

The authors have indicated no significant interest with commercial supporters.

Dermatofibrosarcoma protuberans (DFSP) is an

infrequent cutaneous tumor, fibrohistiocytic

in nature, slow growing, and infiltrative.1 Clinically,

it appears as a skin-coloured plaque or tumor,

frequently located on the trunk, that may go unno-

ticed because of its indolent nature. As the tumor

grows, lateral and deep infiltration begins, and

protuberant nodules develop on the surface.2,3

Histologically, it is characterized by a uniform pro-

liferation of fusiform cells arranged in an irregular

intertwined fascicle, creating a storiform2–5 pattern.

The cellular density is much greater at the center of

the tumor than around the edges, where digitiform

projections begin to appear in the form of fibrous

tracts containing a small number of tumor cells that

can infiltrate subcutaneous cellular tissue (SCT),

muscular fascia, muscle, and even bone.2–5 These

tentacle-like tumor prolongations can extend far

from the central area; this means that DFSP has an

unpredictable subclinical extension, which in turn

means a high rate of recurrence even after broad

surgical excision.6–9 Mohs micrographic surgery

(MMS), which allows the entire margins of surgically

removed tissue to be studied histologically, is the

treatment of choice for DFSP. This surgical technique

provides much lower recurrence rates than conven-

tional surgery because it allows the morphology of

the tumor to be delimited much more precisely while

sparing much healthy peritumoral tissue.6–9

The use of imaging techniques in the study of soft

tissue tumors is a common and necessary practice in

most cases, because they provide information in

addition to that gathered in the clinical examination.

& 2011 by the American Society for Dermatologic Surgery, Inc. � Published by Wiley Periodicals, Inc. �ISSN: 1076-0512 � Dermatol Surg 2011;37:1–8 � DOI: 10.1111/j.1524-4725.2011.02077.x

1

Departments of �Dermatology and yRadiology, Instituto Valenciano de Oncologıa, Valencia, Spain

More specifically, the use of ultrasound, computed

tomography, and magnetic resonance imaging (MRI)

is standard practice for the precise diagnosis of soft

tissue tumors or for specifying their location.10

Given the characteristic clinical appearance of DFSP,

it is usually easy to diagnose using biopsy alone, and

imaging techniques are not routinely used in the

diagnostic process, but because of its infiltrative

nature, in certain cases, it may be useful to include

some sort of imaging technique to ascertain the lat-

eral extension and particularly the level of infiltra-

tion because simple clinical palpation, although

convenient, is not entirely reliable.

In 1994, Kransdorf and Meis-Kindblom11 published

the first study on the use of MRI in DFSP. In their

series of 11 people with DFSP, MRI was used in four,

in which a well-delimited, yet unspecific image of the

tumor was obtained. Various cases have subse-

quently been published in which MRI was used

preoperatively, demonstrating a certain efficacy,

particularly in cases of large DFSP tumors.12–17 The

largest series is that of Torreggiani and colleagues,17

who studied 10 cases of DFSP with MRI but were

able to distinguish only between superficial and deep

and between clearly and poorly delimited. Therefore,

there is no series of cases extensive enough to extract

valid conclusions regarding the use of MRI with

DFSP. In this sense, as Riggs and colleagues indi-

cate,12 to determine the utility of MRI in DFSP, it is

necessary to compare the MRI with the exact depth

of the tumor removed using MMS, because the de-

finitive surgical defect after MMS provides the exact

level of infiltration and an approximate silhouette of

the tumor morphology.

We present an observational and prospective study

comparing MRI and palpation of DFSP cases that

were subsequently extirpated with MMS.

Objectives

The main objective of the study was to ascertain the

reliability of MRI in cases of primary DFSP, to

evaluate its ability to specify the exact level of infil-

tration and lateral extension, and to determine

whether it provides more information than clinical

palpation of the tumor. The study also sought to

deduce the usefulness of MRI in determining the

depth of infiltration in cases of recurrent DFSP and

in establishing tumor persistence in cases of extir-

pated DFSP with affected margins.

Patients and Methods

The study included all people with DFSP seen in the

Dermatology Department at the Instituto Valenciano

de Oncologıa between July 2004 and December

2009. The study included primary tumors, recurrent

tumors, and recently extirpated tumors with positive

margins that had been referred for MMS. All cases

were diagnosed using conventional histologic anal-

ysis with hematoxylin and eosin staining, immuno-

histochemical analysis with CD 34, and positive

expression of translocation COL1A1-PDGFB on

reverse transcription and polymerase chain reaction.

Patients who had received neoadjuvant treatment

with imatinib and those who had not undergone

MMS were excluded. On the first visit, the clinical

information compiled for the study in the case of

primary DFSP included location, largest diameter,

smallest diameter, and depth of infiltration by pal-

pation. First, at least two clinicians (C.S.G., O.S.,

B.L.L.) palpated all tumors. Tumors were considered

not to be deeply infiltrated when they were mobile or

could be moved easily in a direction parallel to that

of the surface of the tumor and when, on palpation,

they seemed to be limited to the SCT. Tumors were

considered to be deeply infiltrated when they were

stationary or could not be moved easily in a direction

parallel to the surface of the tumor or when, on

palpation, they seemed to be adhered to the under-

lying muscular plane. For recurrent cases, the infor-

mation collected included the location of the tumor

and the number of previous treatments. For cases of

excised DFSP and affected margins, only clinical

data regarding location of the tumor were collected.

The MRI was done after the first visit and before

MMS using a Siemens Magnetom Symphony

D E R M AT O L O G I C S U R G E RY2

C O R R E L AT I O N B E T W E E N M R I A N D S U R G I C A L M A R G I N S I N D F S P

Maestro Class, 1.5 T with a Flex Large surface coil.

The study protocol used empty sequences followed

by injection of Prohance or Multihance paramag-

netic contrast agent in volumes that varied according

to the weight of each patient. After establishing a

localizer, sagittal T1-TSE and coronal STIR se-

quences were taken without intravenous contrast,

both with a frequency of 256, and T1-TSE and

T1-STIR axial sequences with a frequency of 516.

After injection of the intravenous contrast agent,

T1-TSE sequences were taken with axial, coronal, or

sagittal fat saturation. On the images obtained after

injection of the contrast agent, the diameters of the

lesion on the cutaneous plane were measured, and

the effect on deep structures was evaluated. Two

radiologists with expertise in soft tissue tumors

without any clinical information studied all cases

independently. The two observers jointly evaluated

uncertain cases or those with conflicting opinions.

The information gathered from the MRI in primary

DFSP cases included the largest diameter, the

smallest diameter, and the exact plane of the DFSP

infiltration, which could be SCT, muscular fascia,

muscle, periosteum, or bone. In cases of recurring

DFSP, MRI was used to determine the plane of

tumor infiltration. In patients who had recently un-

dergone surgery with affected margins, MRI was

used only to determine the persistence or absence of

the tumor.

Modified MMS18 was performed using formalin-

fixed, paraffin-embedded, horizontal sections. First,

the tumors, or the scar tissue in the case of extirpated

tumors with affected margins, were debulked, after

which the first stage of 1 cm of clinically healthy skin

was taken, reaching depths down to the totality of

the SCT. The Mohs specimen was oriented using silk

suture thread and photographed. The tissue was

divided into multiple specimens, mapped for precise

anatomic orientation, and sent to the Mohs histo-

technician for formalin fixation and paraffin em-

bedding before horizontal sections were taken. The

sections were stained with hematoxylin and eosin

and confirmed using CD34 immunostaining. Within

48 hours, the results were relayed to the dermatol-

ogist, who then marked the Mohs map. In patients

with positive margin(s), the residual tumor was

excised with additional 0.5- to 1-cm margins. The

process of excision continued until negative margins

were achieved. The skin defects were temporarily

covered with synthetic wound dressings until com-

plete excision was proven. The information gathered

from MMS included the number of stages, the largest

diameter of the final surgical defect, the smallest di-

ameter of the final surgical defect, and the micro-

scopic infiltration plane most seriously affected (SCT,

muscular fascia, muscle, periosteum, or bone). In

primary DFSP cases and recurrent cases with doubtful

interpretation of the level of microscopic infiltration,

a histologic study of all of the debulked tumor or

the entire piece of the Mohs stage was conducted to

determine the exact plane of tumor invasion. In re-

cently extirpated DFSP with affected margins, it was

also necessary to confirm whether any residual tumor

remained after the initial excision using a conven-

tional histologic study of the scar debulking.

For the MRI and the histologic infiltration, the DFSP

was considered superficial if the tumor was limited

to the SCT and deep if the tumor had reached the

muscular fascia or beyond.

The sensitivity, specificity, positive predictive value

(PPV), and negative predictive value (NPV) of clin-

ical palpation were calculated as a method for

ascertaining the depth of the primary DFSP and in

the case of MRI as a method for ascertaining the

depth of primary DFSP and the depth of recurrent

DFSP. Spearman correlation was used to determine

the correlation between clinical palpation and the

level of histologic invasion and to determine the

correlation grade between the level of depth reported

by the MRI and the level of histological invasion in

primary and recurrent DFSP.

Results

Forty-three people with DFSP were included in the

study: 22 primary cases, nine with recurrent tumors,

and 12 with recently extirpated tumors with affected

3 7 :* * : 2 0 1 1 3

S E R R A - G U I L L E N E T A L

margins. None of the 43 DFSPs in the study recurred

within a mean follow-up period of 35.8 months

(median 33 months, range 9–109 months).

Primary DFSP

Table 1 shows the clinical characteristics, the MRI

findings, and the MMS results of the 22 primary

cases (Figures 1 and 2). In 19 of the 22 primary cases

of DFSP, the size provided by the MRI image was

smaller than the clinically measured size of the

tumor. In two cases, the size of the tumor with the

clinical measurement coincided with that of the MRI

image. Only in one case did the size provided by

the MRI image exceed the clinical size. The number

of Mohs stages needed to obtain negative margins

TABLE 1. Primary Cases

Case Location

Tumor

Size,

cm

Infiltration

According

to Clinical

Palpation

MRI

Size,

cm

Infiltration

Plane in

MRI

Mohs

Stages,

n

Surgical

Defect,

cm

Histologic

Infiltration

Plane

1 Thorax 6� 3 Yes 6� 3 Fascia 2 9� 6 Muscle

2 Scapular 5� 4 No 3.8� 3 Fascia 1 6.5� 5 Fascia

3 Groin 5.5� 3 No 3.5� 3.5 SCT 1 7� 5 SCT

4 Shoulder 5� 4 No 4.2� 3.8 SCT 1 6� 5 SCT

5 Shoulder 4.5� 3 No 3.5� 2.7 SCT 1 6.5� 5 SCT

6 Scalp 5.5� 2.5 No 5� 2.8 SCT 3 12� 5.5 Periosteum

7 Shoulder 3.5� 1.5 No 2� 0.7 SCT 1 5.5� 5 Fascia

8 Lumbar 1.8� 1.5 No 1.7� 1.2 SCT 1 4� 3.5 SCT

9 Scapular 6.5� 4 No 3� 0.8 SCT 1 9� 6.5 SCT

10 Scapular 5� 2.5 No 4.6� 4.2 SCT 1 7� 4.5 SCT

11 Supraclavicular 4� 3.5 Yes 3.1� 1.2 Fascia 2 7� 4 Muscle

12 Lumbar 6.5� 4 No 5� 2.5 SCT 1 8.5� 6 SCT

13 Clavicular 2.5� 2 Yes 2.7� 1.5 Fascia 2 5.5� 4 Muscle

14 Neck 2.5� 2 Yes 1.7� 1.7 Fascia 3 7.5� 5 Muscle

15 Arm 6� 5.5 Yes 3.7� 2.7 SCT 3 8� 8 Muscle

16 Forehead 4� 3 No 3� 2 SCT 2 5.5� 5 Muscle

17 Scapular 6� 5 No 6� 3.8 Fascia 1 8� 7 Fascia

18 Thorax 8� 3.5 Yes 7.5� 3 SCT 1 11� 6 SCT

19 Scapular 2� 2 No 2� 2 SCT 1 4� 4 SCT

20 Shoulder 4.5� 4 Yes 4� 2.3 Fascia 2 7.5� 7.5 Fascia

21 Thorax 5.5� 3.5 Yes 3.9� 3.1 Fascia 2 8.5� 6 Muscle

22 Leg 3� 2 No 3.3� 3.3 SCT 1 5� 4 SCT

MRI = magnetic resonance imaging; SCT = subcutaneous cellular tissue.

Figure 1. Primary dermatofibrosarcoma protuberans (case 9). (A) Not infiltrated, 6.5-� 4-cm tumor located on the back.(B) Magnetic resonance image of a dermatofibrosarcoma protuberans limited to the subcutaneous tissue.

D E R M AT O L O G I C S U R G E RY4

C O R R E L AT I O N B E T W E E N M R I A N D S U R G I C A L M A R G I N S I N D F S P

was one in 13 cases, two in six cases, and three in

three cases. In all cases, the size of the defect created

after MMS was larger than the size of the tumor

shown by the MRI. Clinical palpation correctly de-

tected depth of infiltration in the primary DFSP cases

with a sensitivity of 58%, specificity of 90%, PPV of

87.5%, and NPV of 64.3%. The Spearman rho

correlation coefficient between clinical palpation

and histologic infiltration was 0.54 (p = .009) and

between the MRI image and the level of histologic

infiltration was 0.56 (p = .006). MRI was capable of

detecting the infiltration depth in primary DFSP

cases with a sensitivity of 67%, specificity of 100%,

PPV of 100%, and NPV of 71.4%. The results of the

clinical exploration and MRI in primary DFSP cases

are shown on Table 2.

Recurrent DFSP

Table 3 shows the location and level of infiltration

provided by the MRI and the level of histological

infiltration for the nine cases of recurrent DFSP. Of

the nine cases, three had previously undergone just

one surgical procedure, whereas in six cases, the

tumor had recurred on two or more occasions after

surgical excision (3 cases with 2 previous extirpa-

tions and 3 cases with 3 previous extirpations).

The results of the MRI as a test for detecting infil-

tration depth in recurrent DFSP were as follows:

sensitivity 60%, specificity 100%, PPV 100%, and

NPV 67% (Table 4). The Spearman rho correlation

coefficient between the infiltration depth studied

using MRI and the level of histologic infiltration was

0.51 (p = .16).

Extirpated DFSP with Positive Margins

Of the 12 recently extirpated cases of DFSP with

affected margins, the MRI reported tumor persis-

tence in five cases, tumor absence in two cases,

and uncertain interpretation in five cases. The

histologic study of these 12 cases showed tumor

persistence in 10 cases and only two cases in which

there were no tumors remain. The results of the

12 recently extirpated cases of DFSP with affected

margins and their locations are shown in Table 5.

Figure 2. Primary dermatofibrosarcoma protuberans (case 20). (A) 4.5-�4-cm infiltrated tumor according to a clinicalpalpation. (B) Magnetic resonance image of a dermatofibrosarcoma protuberans affecting the fascia.

TABLE 2. Results of Primary Cases

DiagnosticMethod

n (95% Confidence Interval)

FalseNegative

FalsePositive

TrueNegative

TruePositive Sensitivity Specificity

PositivePredictiveValue

NegativePredictiveValue

Clinical palpation 5 1 9 7 58 (30–86) 90 (71–109) 87.5 (65–110) 64.3 (39–89)Magnetic reso-

nance imaging4 0 10 8 67 (40–93) 100 (100–100) 100 (100–100) 71 (48–95)

3 7 :* * : 2 0 1 1 5

S E R R A - G U I L L E N E T A L

Discussion

Clinical Palpation

Palpation is a routine exploratory technique used in

most cutaneous tumors; it offers information on the

consistency, extension, and infiltration depth. Be-

cause of the type of infiltrative growth of DFSP, it is

particularly important to determine the value of

palpation as a method for routine examination. Ac-

cording to our study, clinical palpation has limited

value for determining infiltration depth in primary

DFSP cases. We found that palpation had a sensi-

tivity of 58% for detecting in-depth invasion; this

may be considered low, because a high percentage of

cases (42%) were deemed superficial when in fact

they were deep. Moreover, the probability that a

tumor is superficial, when using palpation and being

diagnosed as such, is 64.3% (NPV). Nevertheless,

most of the deep DFSPs were identified correctly as

deeply infiltrated by clinical examination (PPV of

87.5%). In conclusion, clinical palpation is recom-

mended as a routine and accessible exploratory

method because it provides reliable information in

the majority of cases, even though it is important

to take into account the limitations due to the in-

filtrative nature of this tumor.

Magnetic Resonance Imaging

Although MRI is the imaging test most commonly

used in soft tissue tumors,10 little is known about its

value in the study of DFSP because there have been

few studies and publications addressing this.

Delimitation of Depth: According to our results,

MRI is of greater value than clinical palpation in

detecting depth of infiltration in primary DFSP cases

(sensitivity 58% for clinical palpation vs 67% for

MRI). MRI correctly reported all cases of DFSP

limited to the SCT as superficial (100% specificity),

and all the cases reported by MRI as deeply infil-

trated were also correctly identified (8 cases, PPV

100%). We found no relationship between tumor

size and accuracy of MRI in detecting depth of in-

filtration. Although MRI lost precision in determin-

ing depth of infiltration in tumors located in the

TABLE 3. Recurrent Cases

Case Location

Infiltration Plane in Magnetic

Resonance Imaging

Histologic Infiltration

Plane

Previous

Treatments

1 Suprascapular SCT Muscle 3

2 Shoulder Muscle Muscle 1

3 Back SCT SCT 3

4 Arm SCT SCT 1

5 Buttock Lack of tumor SCT 1

6 Foot SCT Muscle 2

7 Foot SCT SCT 2

8 Scapular Fascia Fascia 3

9 Scapular Fascia Fascia 2

SCT = subcutaneous cellular tissue.

TABLE 4. Results of Recurrent Cases

Diagnostic

Method

n (95% Confidence Interval)

False

Negative

False

Positive

True

Negative

True

Positive Sensitivity Specificity

Positive

Predictive

Value

Negative

Predictive

Value

Magnetic

resonance

imaging

2 0 4 3 60 (17–103) 100 (100–100) 100 (100–100) 67 (29–104)

D E R M AT O L O G I C S U R G E RY6

C O R R E L AT I O N B E T W E E N M R I A N D S U R G I C A L M A R G I N S I N D F S P

head, neck, and upper thorax, in seven of the eight

cases in these locations (cases 1, 6, 11, 13, 14, 16, 18,

and 21), MRI reported the affected plane to be more

superficial than it was. In this regard, as various

studies have shown,19–22 DFSPs on the head and neck

are particularly susceptible to a higher rate of recur-

rence than on the rest of the body. The fact that these

locations have the anatomic peculiarity that the mus-

cular plane is located close to the cutaneous surface,

where the skin is not as thick as in other locations,

may explain the poor prognosis for DFSP on the head

and neck to a certain extent. Furthermore, DFSP

spreads microscopically through the SCT septa to the

fascia and the muscle in a less obvious way than in

other locations and therefore goes unnoticed in con-

ventional histologic studies and apparently also on

MRI. Therefore, MRI seems to be useful and reliable

in primary DFSP when the tumor is located some-

where other than the head, neck, and upper thorax.

As for the utility of MRI in cases of recurrent DFSP,

our results show that it has a value similar to that

of primary cases, although with only nine cases, no

definitive conclusions can be drawn. There appears

to be no relationship between the number of previ-

ous treatments received by each person with DFSP

and better or poorer precision in detecting infiltra-

tion by MRI. Nevertheless, in all three cases with

only one recurrence of DFSP, MRI diagnosed the

exact plane of infiltration (2 in SCT and 1 in muscle),

unlike cases with two or more recurrences, in which

MRI was accurate in four cases but diagnosed the

infiltration in a plane less than the true depth in two

cases. In any event, more cases are needed to confirm

this trend.

Detection of Tumor Persistence in DFSP Excision

with Affected Margins: In five of the 12 DFSP cases

studied, MRI provided an uncertain image and was

therefore not considered useful for scheduling a re-

exeresis of the scar tissue. In the two cases in which

MRI showed tumor absence, DFSP remains were

found in the histologic study that would certainly

have resulted in a recurrence had they not been re-

moved. Therefore, MRI is not useful for detecting

tumor persistence in cases of recent incompletely

extirpated tumors.

Lateral Delimitation: Although imaging techniques

are generally considered unnecessary for delimiting

the lateral margins of DFSP, this study shows that, in

a majority of primary cases (19/22), MRI indicated a

smaller tumor size than clinical examination deter-

mined, and a smaller defect was created after exci-

sion by MMS. Therefore, considering that the lateral

limits of DFSP can extend microscopically as far as

12 cm from the macroscopic edge of the tumor,9

MRI is not recommended for determining lateral

extension in DFSP.

The fact that this series of DFSP contains a large

number of cases in which the histologic infiltration

goes beyond the SCT (12/22 primary DFSP cases and

TABLE 5. Extirpated DFSP with Positive Margins

Case Location

Tumoral Persistence in Magnetic

Resonance Imaging

Histologic Tumor

Persistence

1 Scalp Uncertain Persistence

2 Scapular Uncertain Persistence

3 Arm Persistence Persistence

4 Thorax Uncertain Persistence

5 Scalp Persistence Persistence

6 Thorax Persistence Absence

7 Supraclavicular Absence Persistence

8 Scalp Persistence Persistence

9 Shoulder Absence Persistence

10 Infraclavicular Uncertain Persistence

11 Cheek Persistence Persistence

12 Arm Uncertain Absence

3 7 :* * : 2 0 1 1 7

S E R R A - G U I L L E N E T A L

5/9 recurrent cases) could be perceived as a non-

representative sample because the DFSP usually is

limited to the SCT in most cases.2,23 This possible

bias is probably because the cases referred to our

department were particularly complex because of the

location or size of the tumor, and therefore the

possibility of deeper infiltration was greater.

In conclusion, to the best of our knowledge, we

present the largest series to date of DFSP cases

studied using MRI. According to our results, MRI is

superior to clinical palpation in detecting depth of

infiltration of DFSP. Although the best method for

determining depth of infiltration and lateral exten-

sion of DFSP is histologic analysis using MMS, MRI

can be useful as a preoperative study in primary

cases situated in locations other than the head, neck,

and upper thorax. In DFSPs in which the tumor

has been excised with positive margins, MRI is

not useful for determining the persistence of the

tumor. MRI does not appear to be useful for

determining the lateral limits of DFSP.

References

1. LeBoit P, Burg G, Weedon D, Sarasin A. Soft tissue tumors. In:

LeBoit P, Burg G, Weedon D, et al., editors. World Health Or-

ganization Classification of Tumors. Pathology and Genetics. Skin

Tumors. Lyon: IARC; 2006. p. 229–62.

2. Gloster HM Jr. Dermatofibrosarcoma protuberans. J Am Acad

Dermatol 1996;35:355–74.

3. Sanmartin O, Llombart B, Lopez-Guerrero JA, Serra C, et al.

Dermatofibrosarcoma protuberans. Actas Dermosifiliogr

2007;98:77–87.

4. Taylor HB, Helwig EB. Dermatofibrosarcoma protuberans. A

study of 115 cases. Cancer 1962;15:717–25.

5. McPeak CJ, Cruz T, Nicastri AD. Dermatofibrosarcoma protu-

berans: an analysis of 86 casesFfive with metastasis. Ann Surg

1967;166:803–16.

6. Hobbs ER, Wheeland RG, Bailin PL, Ratz JL, et al. Treatment of

dermatofibrosarcoma protuberans with Mohs micrographic sur-

gery. Ann Surg 1988;207:102–7.

7. Gloster HM Jr., Harris KR, Roenigk RK. A comparison between

Mohs micrographic surgery and wide surgical excision for the

treatment of dermatofibrosarcoma protuberans. J Am Acad

Dermatol 1996;35:82–7.

8. Dawes KW, Hanke CW. Dermatofibrosarcoma protuberans

treated with Mohs micrographic surgery: cure rates and surgical

margins. Dermatol Surg 1996;22:530–4.

9. Ratner D, Thomas CO, Johnson TM, Sondak VK, et al. Mohs

micrographic surgery for the treatment of dermatofibrosarcoma

protuberans. Results of a multiinstitutional series with an analysis

of the extent of microscopic spread. J Am Acad Dermatol

1997;37:600–13.

10. Murphey M, Kransdorf M. Radiologic evaluation of soft tissue

tumors. In: Weiss S, Goldblum JR, editors. Enzinger and Weiss’s

Soft Tissue Tumors. 5th ed. Philadelphia: Mosby; 2008. p. 33–71.

11. Kransdorf MJ, Meis-Kindblom JM. Dermatofibrosarcoma pro-

tuberans: radiologic appearance. Am J Roentgenol 1994;

163:391–4.

12. Riggs K, McGuigan KL, Morrison WB, Samie FH, et al. Role of

magnetic resonance imaging in perioperative assessment of dermato-

fibrosarcoma protuberans. Dermatol Surg 2009;35:2036–41.

13. Chen X, Chen YH, Zhang YL, Guo YM, et al. Magnetic reso-

nance imaging and mammographic appearance of dermatofibro-

sarcoma protuberans in a male breast: a case report and literature

review. J Med Case Reports 2009;3:8246–9.

14. Laffan EE, Ngan BY, Navarro OM. Pediatric soft-tissue tumors

and pseudotumors: MRI imaging features with pathologic corre-

lation: part 2. Tumors of fibroblastic/myofibroblastic, so-called

fibrohistiocytic, muscular, lymphomatous, neurogenic, hair ma-

trix, and uncertain origin. Radiographics 2009;29:e36.

15. Bergin P, Rezaei S, Lau Q, Coucher J. Dermatofibrosarcoma

protuberans, magnetic resonance imaging and pathological cor-

relation. Australas Radiol 2007;51Spec No.:B64–6.

16. Thornton SL, Reid J, Papay FA, Vidimos AT. Childhood

dermatofibrosarcoma protuberans: role of preoperative imaging.

J Am Acad Dermatol 2005;53:76–83.

17. Torreggiani WC, Al-Ismail K, Munk PL, Nicolaou S, et al.

Dermatofibrosarcoma protuberans: MRI imaging features. Am J

Roentgenol 2002;178:989–93.

18. Hafner J, Schutz K, Morgenthaler W, Steiger E, et al. Micro-

graphic surgery (‘slow Mohs’) in cutaneous sarcomas. Dermatol-

ogy 1999;198:37–43.

19. Paradisi A, Abeni D, Rusciani A, Cigna E, et al. Dermatofibro-

sarcoma protuberans: wide local excision vs. Mohs micrographic

surgery. Cancer Treat Rev 2008;34:728–36.

20. Gayner SM, Lewis JE, McCaffrey TV. Effect of resection margins

on dermatofibrosarcoma protuberans of the head and neck. Arch

Otolaryngol Head Neck Surg 1997;123:430–3.

21. Tom WD, Hybarger CP, Rasgon BM. Dermatofibrosarcoma pro-

tuberans of the head and neck: treatment with Mohs surgery using

inverted horizontal paraffin sections. Laryngoscope

2003;113:1289–93.

22. Loss L, Zeitouni NC. Management of scalp dermatofibrosarcoma

protuberans. Dermatol Surg 2005;31:1428–33.

23. Bowne WB, Antonescu CR, Leung DH, Katz SC, et al. Dermato-

fibrosarcoma protuberans: a clinicopathologic analysis of patients

treated and followed at a single institution. Cancer 2000;88:2711–20.

Address correspondence and reprint requests to: CarlosSerra-Guillen, MD, Department of Dermatology, InstitutoValenciano de Oncologıa, C/Beltran Baguena 4, 46009Valencia, Spain, or e-mail: cserraguillen@gmail.com

D E R M AT O L O G I C S U R G E RY8

C O R R E L AT I O N B E T W E E N M R I A N D S U R G I C A L M A R G I N S I N D F S P