Dermatofibrosarcoma protuberans: Treatment

INTRODUCTION — Dermatofibrosarcoma protuberans (DFSP) is an uncommon locally aggressive cutaneous soft tissue sarcoma. Approximately 85 to 90 percent of DFSPs are low grade, while the remainder contain a high-grade sarcomatous component (which is usually a fibrosarcoma, designated DFSP-FS) and are considered to be intermediate-grade sarcomas [1]. Although they rarely metastasize (fewer than 5 percent of cases), all DFSP variants have a propensity to recur locally.

This topic review will focus on treatment of DFSP. The epidemiology, molecular pathogenesis, histology, clinical presentation, diagnosis, and staging of DFSP are discussed elsewhere, as are issues specific to DFSP arising in the head and neck region. (See "Dermatofibrosarcoma protuberans: Epidemiology, pathogenesis, clinical presentation, diagnosis, and staging" and "Head and neck sarcomas", section on 'Desmoids and dermatofibrosarcoma protuberans'.)

LOCALIZED DISEASE — The preferred initial treatment for a localized dermatofibrosarcoma protuberans (DFSP) is resection with pathologically negative margins. Tumor size and location dictate the most appropriate surgical procedure. Given that metastases to lymph nodes are extremely rare [2], there is no role for prophylactic regional node dissection. (See "Dermatofibrosarcoma protuberans: Epidemiology, pathogenesis, clinical presentation, diagnosis, and staging", section on 'Metastatic disease'.)

Importance of resection margins — One of the characteristic features of DFSP is its capacity to invade surrounding tissues in an eccentric fashion to a considerable distance from the central focus of the tumor, with tumor cells invading subcutaneous tissue in the form of irregular tentacle-like projections through the septa and fat lobules. This fact, coupled with the rarity of DFSP and diagnostic delay, often leads to inadequate initial resection. The risk of a local recurrence is as high as 50 percent with simple excision ("conservative" surgical margins), and it is even higher if the margins are positive [1,3-6]. (See "Dermatofibrosarcoma protuberans: Epidemiology, pathogenesis, clinical presentation, diagnosis, and staging", section on 'Histology'.)

The status of the surgical margins is the most important prognostic factor in patients with DFSP:

●The prognostic importance of resection margins was shown in a series of 159 patients (134 DFSP, 25 DFSP with sarcomatous transformation [DFSP-FS]) treated at Memorial Sloan-Kettering over a 48-year period [1]. The resection margins were involved in 51 patients (32 percent), close (<1 mm) in 15 (10 percent), and otherwise negative in 93 (58 percent). At a median follow-up of 57 months, there were 34 recurrences, 29 of which developed in patients with positive or close margins.

●A compilation of five literature-based series of patients with DFSP treated with undefined or conservative surgical margins (n = 116 patients) noted local recurrence rates between 26 and 60 percent, averaging 40 percent [6]. In contrast, recurrence rates in published series of patients treated with wide local excision (WLE; n = 661) ranged from 0 to 42 percent, averaging 9 percent.

Although localized recurrences can frequently be salvaged with further surgery, locally recurrent tumors have a significant propensity for deep invasion into the fascia, muscle, or bone, which increases the risk for both inadequate resection and functional, as well as cosmetic, defects. Furthermore, local recurrence also predisposes to distant metastases [1,7-9]. In the above noted series, both patients who developed fatal metastatic disease had initially incompletely resected DFSP-FS [1].

For all of these reasons, marginal resection should be avoided, and histologic assessment of the final resection margins is mandatory [8].

Role of diagnostic imaging — Given the characteristic histologic appearance of most DFSP on core needle or incisional biopsy, imaging techniques are not routinely used in the diagnostic process. (See "Dermatofibrosarcoma protuberans: Epidemiology, pathogenesis, clinical presentation, diagnosis, and staging", section on 'Histology'.)

However, because of the infiltrative nature of DFSP, simple clinical palpation is not entirely reliable for ascertaining the depth of tumor invasion. Magnetic resonance imaging (MRI) is more sensitive than palpation for assessing depth of infiltration and may be useful for certain DFSPs, particularly large or recurrent tumors [10-15]. Characteristic findings on MRI are discussed elsewhere. (See "Dermatofibrosarcoma protuberans: Epidemiology, pathogenesis, clinical presentation, diagnosis, and staging", section on 'Staging evaluation'.)

Although MRI is better than clinical palpation, the extent of local invasion may be difficult to appreciate preoperatively, even with optimal MRI evaluation, particularly in sites in the head and neck and upper thorax. The utility of MRI was shown in a study of 43 DFSPs, which included 22 primary tumors, nine recurrent tumors, and 12 tumors that were extirpated with positive margins and referred for Mohs micrographic surgery (MMS) [10]. Clinical palpation correctly predicted the depth of infiltration in the primary DFSP cases with a sensitivity of 58 percent, a specificity of 90 percent, a positive predictive value (PPV) of 88 percent, and a negative predictive value (NPV) of 64 percent. The corresponding values for MRI were 67, 100, 100, and 71 percent, respectively. MRI was of little value in determining lateral tumor extension and lost precision for determining the depth of infiltration in tumors located in the head and neck and upper thorax; in seven of eight cases, MRI reported the affected plane to be more superficial than it actually was histologically. This finding was attributed to the lower thickness of the skin at these locations and the close proximity of the muscle to the cutaneous surface, which lessened the ability of MRI to detect spread to the fascia and muscle.

For recurrent DFSP, the sensitivity, specificity, PPV and NPV of MRI were 60, 100, 100, and 67 percent, respectively. On the other hand, MRI was not accurate in cases with margin-positive excisions. Among 12 such patients, MRI suggested tumor persistence in five, tumor absence in two, and was uncertain in five. Histologic study revealed tumor persistence in 10 cases (only four of which were defined positive by MRI), and only two cases in which no tumor remained, neither of which were negative by MRI. These data suggest that MRI is beneficial for recurrent tumors but not for detecting tumor persistence in cases of incompletely extirpated tumors.

Computed tomography (CT) is only indicated in rare cases in which underlying bone involvement is suspected.

Surgical techniques — Complete surgical resection is the optimal treatment for localized DFSP. The two options are WLE and MMS.

Wide local excision — WLE is an option for surgical management and is preferred over marginal excision for both DFSP and DFSP-FS. Although local recurrence rates are lower with WLE than with conservative or marginal excision, local recurrence rates range widely with WLE (from 4 to 47 percent), but they average 6 to 8 percent [1,6,8,16-24].

Much of the variability relates to the width of the resection margin. The minimum resection margin needed to achieve local control is undefined. Mapping with MMS has revealed tentacle-like tumor projections that often extend beyond 3 cm [25]. In a series of 66 French patients treated for DFSP between 1982 and 2002, local recurrences were significantly more frequent among those with a resection margin of <3 cm, as compared with margins of 3 to 5 cm (47 versus 7 percent) [24]. Wider margins (5 cm) have been associated with even lower local recurrence rates (<5 percent) [6,9]. These data form the basis of the recommendation by some that resection margins be at least 3 cm (down to and including the fascia) and as wide as 5 cm, depending on the primary tumor size [4,8,16,19,20,22,26-29]. Consensus-based guidelines from the National Comprehensive Cancer Network (NCCN) [30] and a European interdisciplinary group [31] recommend margins of 2 to 4 cm, to the investing fascia of muscle or pericranium with clear pathologic margins, when clinically feasible. However, in practice, such margins are rarely possible without significant cosmetic and/or functional deficit, particularly in anatomically challenging areas, such as the head and neck. (See "Head and neck sarcomas", section on 'Desmoids and dermatofibrosarcoma protuberans'.)

Others suggest that very low local recurrence rates (1 percent) and a high rate of primary closure (69 percent) can be achieved with narrower margins (median 2 cm, range 0.5 to 3 cm), as long as the surgical approach is standardized and pathologic evaluation of the margins is meticulous [21,32,33]. In one multi-institutional retrospective series of 204 patients undergoing WLE for a DFSP, 81 percent achieved negative margins with only one excision, 16 percent required two surgeries, 3 percent required three, and 1 percent, four resections [32]. At a median follow-up of 64 months, only two patients (both with head and neck primaries) recurred locally (1 percent).

Frequently, reconstruction with a skin graft or flap is necessary to repair the resulting wound after resection with wide margins. However, reconstruction should be delayed until negative histologic margins (including assessment of the complete circumferential and peripheral deep margins) are verified. If positive margins are diagnosed after WLE, most guidelines (including those of the NCCN [30]) recommend reresection, followed by adjuvant radiation therapy or imatinib (see below).

Mohs micrographic surgery — An alternative to WLE, MMS is increasingly used for resection of large or recurrent DFSP and for those arising in anatomically challenging areas, such as the head or neck. In contrast to WLE, which uses representative vertical sectioning, MMS utilizes sequential horizontal sectioning during the resection with immediate microscopic examination of the resected tissue by frozen section analysis (including rapid immunostaining for CD34 [34]) until a clear margin is obtained. (See "Mohs surgery", section on 'Surgical technique' and "Dermatofibrosarcoma protuberans: Epidemiology, pathogenesis, clinical presentation, diagnosis, and staging", section on 'Immunohistochemistry (IHC)'.)

Some surgeons perform a modified "slow" MMS that uses formalin-fixed, paraffin-embedded permanent sections (with conventional immunohistochemical staining for CD34, which is more reliable than rapid CD34 immunostaining on frozen sections [6,35]) throughout the duration of the procedure [36-38]. However, in contrast to the frozen section processing time of one to two hours, paraffin sections take longer, and use of the modified MMS technique may spread the procedure over many days, thereby delaying wound reconstruction. (See "Mohs surgery", section on 'Specific indications'.)

To limit the number of stages required to clear the tumor, many Mohs surgeons initially debulk the tumor with the excision made just beyond the visible tumor margin and beneath the tumor mass; the first Mohs layer is then taken as a relatively wide (0.5 to 1.0 cm) margin of tissue excision around and under the debulked tumor [35,37,38]. The specimen is oriented using silk sutures and photographed, then the tissue is divided into multiple specimens mapped for precise anatomic orientation. Each specimen is then processed, and sections are evaluated under the microscope for evidence of tumor involvement. In patients with positive margins, the residual tumor is excised with additional 0.5 cm margins. The cyclic process of excision, mapping, and microscopic examination is repeated until tumor is no longer microscopically detected. In general, a higher number of stages are required to resect DFSP-FS than conventional DFSP due to the deeper infiltration of these lesions [39].

A high probability of cure of DFSP can be attained with MMS, as long as the final margins are negative [6,20,21,40-46]. Although randomized trials have not been conducted, the available data from nonrandomized comparative and noncomparative single-center studies suggest that local recurrence rates are lower after MMS than after WLE [6,20,21,23,27,33,36,47,48]. As an example, in systematic reviews of patients with resected DFSP, the local recurrence rate is approximately 1 to 1.5 percent for those treated with MMS and approximately 4 to 9 percent for those treated with WLE [23,49].

Furthermore, one of the main benefits of MMS is the ability to minimize surgical margins, thus preserving cosmetically and functionally vital tissue. Compared with WLE with surgical margins of ≥3 cm, MMS permits maximal conservation of normal uninvolved tissue, which results in smaller wounds needing less complex reconstruction. As noted above, while there is controversy regarding the optimal width of the surgical margins with WLE, recommended margin widths are generally 2 to 4 cm. In the systematic review described above, a median maximum lateral margin of 1 to 2 cm resulted in complete excision for most tumors treated with MMS [23].

These data have led many to consider MMS as the preferred surgical approach for DFSP because of its high cure rate and ability to preserve normal tissue [20,31,45,50]. However, some argue that MMS is histologically inadequate since even on paraffin sections, it can be difficult to differentiate minimal residual DFSP cells from scattered spindle cells in normal dermis, scar, or connective tissue [6,51]. In an attempt to minimize this problem, many Mohs surgeons send a final stage after clearing the tumor with the traditional MMS frozen sectioning for formalin-fixed, paraffin-embedded sections to confirm that the margins are in fact histologically clear.

However, there are no randomized trials that demonstrate the superiority of MMS over other surgical methods and a lack of long-term data on MMS, which is important given that the median time to recurrence is approximately three years post-surgery [23]. Furthermore, there are cost and access considerations, with the estimated cost of MMS being higher than that for WLE [52] and availability of MMS being dependent on local expertise.

Thus, although a possible benefit for MMS has been suggested in retrospective case series, the choice between MMS and WLE remains a matter of judgment based on the location and size of the tumor, perceived recurrence risk, availability, and cost.

For cosmetically sensitive areas, where achieving narrow margins is preferable, MMS represents a preferred choice for therapy, where available. In other less cosmetically sensitive areas, WLE with 2 to 4 cm margins and histologic margin confirmation is still an acceptable procedure. For DFSP-FS, we prefer WLE over MMS.

Radiation therapy — Although DFSP is a radioresponsive tumor, radiation therapy (RT) is only rarely used as a primary treatment. There are scant data pertaining to the efficacy of RT when used alone for gross disease [3,53]. In an early series, all three patients treated with RT (67 to 75 Gy) after an intralesional excision remained locally controlled at 85, 106, and 108 months posttreatment [3]. However, in a second report, a single patient who received RT alone for gross disease recurrence after surgery did not achieve disease control [53]. These data suggest that RT is not a substitute for inadequate resection.

RT following surgical resection — Despite the absence of randomized clinical trials proving benefit, adjuvant RT is commonly recommended in conjunction with wide local resection of large tumors, or when the surgical margins are close or positive and further surgery is not feasible [3,30,53-57]. Whether administered before or after surgery, RT significantly reduces the risk of a local recurrence.

There are no randomized trials addressing the benefit of RT after resection. Several retrospective case series and cohort studies have been reported, primarily of patients undergoing local excision followed by RT. A meta-analysis of twelve of these reports (totaling 167 patients) [3,21,53,55,56,58-64] came to the following conclusions [65]:

●The majority of patients (92) had positive or close resection margins, while 45 were negative, and 30 were unknown.

●Local recurrence rates after RT ranged from 0 to 25 percent (pooled estimate 11.7 percent); the rate among patients with positive or close margins was 14.2 percent, while there was no recurrence among those with negative margins.

●In the two studies that addressed this issue, the pooled estimate of local recurrence between surgery alone and surgery combined with RT (72 patients total, 45 treated with surgery alone and 27 with adjuvant RT) was potentially clinically meaningful but not statistically significant (odds ratio 0.31, 95% CI 0.09-1.08) [55,56].

These data support the view that conservative resection and postoperative RT could be considered in situations in which adequate WLE alone would result in major cosmetic or functional deficits, particularly if the expertise to perform MMS is not available.

The presence of a sarcomatous component within the tumor (ie, DFSP-FS) is not, by itself, viewed as an indication for postoperative RT. While the overall rate of distant metastases and prognosis are worse, whether tumors that contain a sarcomatous component have a worse local recurrence rate is unclear. Although some series suggest that local control rates are worse for patients with DFSP-FS compared with conventional DFSP, others do not, as long as surgical excision is adequate. (See 'Prognosis' below.)

NCCN guidelines [30] do not suggest RT for tumors with a sarcomatous component unless the tumor is large or the margins are close or positive.

The dose-fractionation schedules and treatment techniques for adjuvant RT of DFSP are similar to those used for other soft tissue sarcomas [66]. (See "Overview of multimodality treatment for primary soft tissue sarcoma of the extremities and superficial trunk".)

Prognosis — Generally, the prognosis is good. In an analysis of 7250 cases reported to the United States Surveillance, Epidemiology, and End Results program of the National Cancer Institute, the 10-year survival was 99.1 percent [67]. Predictors of higher mortality include head/neck or limb location, high mitotic index, increased cellularity, Black ethnicity, and male sex [1,37,67-69].

Although patients rarely die from DFSP, morbidity due to local recurrence remains a problem. The most important risk factor for local recurrence is the status of the surgical margins. In addition, age over 50 appears to be a risk factor for local recurrence [1] and inferior survival. Tumor size does not appear to have a significant influence on either recurrence or survival [1,6,8,20,69].

DFSP-FS is a potentially more aggressive tumor than low-grade DFSP because of larger size and frequent infiltration of the neighboring fascia and muscle [37]. However, the impact of sarcomatous progression (ie, DFSP-FS) on prognosis is debated. Most [1,7,70-75] but not all [58,68,76] series suggest a higher rate of metastases compared with low-grade DFSP. However, metastasis rates are low overall, and in some of these reports (almost all of which contained a small number of patients with the transformed DFSP), survival was not worse in patients with a sarcomatous component, as long as surgical resection margins were adequate [68,73]. Larger series with long-term follow-up have documented worse survival rates in patients with the fibrosarcomatous variant [75].

The impact of a sarcomatous component on local recurrence rates is even less clear:

●Although many early series reported higher rates of local recurrence with DFSP-FS compared with classic DFSP [7,72], the status of the surgical margins was either unreported, or a minority had received WLE. Some subsequent reports found no differences in the local recurrence rate between conventional DFSP and DFSP-FS treated with adequate surgery [22,39,58,68,70,73].

●On the other hand, others have demonstrated that having a sarcomatous variant is an adverse prognostic factor independent of surgical margin status [1,74]. The best data come from a series of 260 cases of DFSP undergoing surgical treatment in a single Chinese center over a 21-year period; there were 226 cases of classic DFSP and 34 DFSP-derived sarcomatous variants (32 fibrosarcoma and two malignant fibrous histiocytoma); 91 percent underwent WLE [74]. Compared with classic DFSP, the transformed DFSPs had a significantly higher recurrence rate after wide excision (35 versus 6 percent), a higher metastatic rate (24 versus 0.4 percent), and worse survival rates (at five years, 89.6 versus 99.6 percent; at 10 years, 66.0 versus 98.6 percent). Of the nine patients in this series who developed metastases, eight had sarcomatous transformation.

Posttreatment surveillance — The majority of local recurrences become evident within three years, but 25 to 30 percent develop after five years [1,16,24,40,77]. Thus, posttreatment follow-up must be carried out lifelong. The time to recurrence may be shorter in transformed DFSP (median 11 versus 33.5 months in one large series [75]). The disease-free interval between recurrences shortens progressively. In one report, the median time to relapse for first and second recurrences was approximately 3.5 years, but it was one year for the third relapse [56].

The optimal surveillance strategy has not been established. We and others [37,67] suggest that patients be examined every six months for the first three to five years, then annually lifelong. These guidelines are consistent with those from the NCCN [30], which recommend that the primary site be examined every 6 or 12 months and that the patient be educated as to regular self-examination.

Physical examination should focus on careful and inspection of the scar because DFSP is characterized by its capacity for local recurrence, even after WLE.

Metastases preferentially localize to the lung, but are reported at other sites including brain, bone, and heart. As only 2 to 5 percent of cases metastasize, extensive radiographic surveillance with CT, blood counts, and liver function tests are not warranted, unless there are symptoms or signs to suggest metastatic spread. Although it is difficult to determine which cases are at risk for metastasis, they generally involve recurrent lesions that have progressed for many years and when a fibrosarcomatous component is present (ie, DFSP-FS). This has led some to recommend that follow-up imaging examinations be limited to those with recurrent DFSP or a DFSP with fibrosarcomatous transformation [67].

TREATMENT OF LOCALLY ADVANCED, RECURRENT, AND METASTATIC DISEASE — As noted above, metastasis to lymph nodes is extremely rare, with only a few case reports in the literature [2]. The management strategy for patients with lymph node metastases is controversial. However, lymphadenectomy is recommended in such cases, as long-term survival is possible [78].

Distant hematogenous metastases are even rarer and are most likely in patients who have had multiple local recurrences after inadequate surgical resection. The lungs are most frequently affected, but metastases to brain, bone, and other soft tissues are reported. Resection is a reasonable approach for appropriately selected patients with a controlled primary tumor and isolated, potentially resectable sites of metastatic disease that appear resectable, especially isolated pulmonary metastatic disease. [70,79]. (See "Surgical treatment and other localized therapy for metastatic soft tissue sarcoma".)

Molecularly targeted therapy — The vast majority (possibly all) cases of dermatofibrosarcoma protuberans (DFSP) have a characteristic chromosomal translocation, t(17;22), which places platelet-derived growth factor beta (PDGFB) under the control of the collagen type I alpha 1 (COL1A1) promoter. This results in constitutive activation of the PDGF receptor beta (PDGFRB), a tyrosine kinase. (See "Dermatofibrosarcoma protuberans: Epidemiology, pathogenesis, clinical presentation, diagnosis, and staging", section on 'Molecular pathogenesis'.)

Orally active small molecule tyrosine kinase inhibitors (TKIs) have been developed that target such molecular defects. One such agent, imatinib (Gleevec), inhibits the PDGF receptor as well as other receptor tyrosine kinases such as c-kit and has revolutionized the treatment of chronic myeloid leukemia and gastrointestinal stromal tumors (GIST). (See "Overview of the treatment of chronic myeloid leukemia" and "Clinical presentation, diagnosis, and prognosis of gastrointestinal stromal tumors".)

Other agents, such as sunitinib and sorafenib, are multifunctional TKIs that inhibit the PDGF receptor, among other tyrosine kinases.

Imatinib

Advanced or metastatic disease — Accumulating reports suggest that patients with advanced DFSP may benefit from imatinib, and some have a sustained complete clinical remission [80-87]. The following represents the range of findings:

●In the largest retrospective series, 10 patients with locally advanced (n = 8) or metastatic (n = 2) DFSP received imatinib 800 mg daily [81]. All eight patients with locally advanced disease had karyotypic or fluorescence in situ hybridization (FISH) evidence of t(17;22), and all responded to therapy, two completely. The two patients with metastatic disease had more complex karyotypes. One who had the typical t(17;22) had a partial response that lasted seven months, while the second lacked the t(17;22) and did not respond.

●Two multicenter prospective phase II trials of imatinib for t(17;22)-positive locally advanced or metastatic DFSP were terminated early because of poor accrual after 2007, the year that imatinib was approved for treatment of DFSP by the US Food and Drug Administration (FDA) [84]. The North American trial (n = 8) used imatinib 400 mg daily, while the European trial (n = 16) used imatinib 400 mg twice daily. In a combined analysis of all 24 patients, there were 11 partial responses (46 percent), and nine others had stable disease. There did not seem to be an advantage for the higher imatinib dose. The median time to progression was 1.7 years.

●In published reports, the majority (12 of 17 cases) of DFSP with sarcomatous transformation (DFSP-FS) and three of three cases of DFSP with a pleomorphic sarcoma component had the fusion gene, and an excellent clinical response has been noted in such patients with imatinib [80-82,88-91], although responses may be short lived in patients with the fibrosarcomatous variant [85].

●In a systematic review of nine studies totaling 152 patients treated with imatinib either for advanced DFSP or in the primary or neoadjuvant setting for locally advanced/metastatic disease, the following were noted [87]:

•When the COL1A1-PDGFB protein translocation was reported, it was present in 91 percent of treated individuals (111 of 122).

•Complete responses were seen in 5.2 percent (8 of 152), while partial responses were seen in 55.2 percent (84 of 152); 27.6 percent had stable disease (42 of 152).

•There were no differences in response rate using a 400 or 800 mg daily dose.

•Adverse effects were present in at least 74 percent of patients, but they were serious (grade 3 or 4) in only 15 percent of cases.

It is not clear whether conventional DFSP tumors lacking t(17;22) respond as well to imatinib; some trials document a lack of response in this setting [80,81], while others have shown objective responses to imatinib in DFSPs that were negative for the t(17;22) translocation [86,92]. Although the National Comprehensive Cancer Network (NCCN) guidelines [30] suggest that molecular analysis prior to the institution of imatinib "may be useful," we would not restrict a therapeutic trial of imatinib to only those patients with a translocation-positive DFSP.

Is there a role for neoadjuvant or adjuvant imatinib? — Reports of sustained complete clinical remission in patients with metastatic DFSP have prompted interest in the use of imatinib in the neoadjuvant setting for large unresectable or recurrent tumors [81,84,86,93,94]. Early results from two prospective trials are described in detail:

●In a multicenter phase II German trial, 16 patients with locally advanced primary (79 percent) or locally recurrent (21 percent) but surgically manageable DFSP received imatinib (600 mg daily) for at least six weeks with continuation of treatment for tumor response of stable disease or better [86]. At week 12, the decision to pursue tumor surgery or continue imatinib was made at the discretion of the investigator. Treatment was stopped at any point for disease progression or intolerable side effects.

At week 12, there were seven partial responses, five with stable disease, and two with progressive disease. Of the 14 evaluable patients, definitive tumor surgery (with negative surgical margins) was carried out in 13 after a median 3.1 months of imatinib; one patient attained a clinical complete response after 16.7 months of imatinib and declined surgery. At histopathologic analysis, 5 of 11 analyzed DFSPs were COL1A1-PDGFB+, and there was no association between tumor response and the presence of the COL1A1-PDGFB fusion gene. At a median follow-up of 6.4 years, only one patient developed secondary resistance to imatinib with outgrowth of new tumor lesions after an initial partial response; this same patient developed a local recurrence after surgery with tumor-free margins and later distant metastases. No other patient with DFSP had secondary resistance, local recurrence, or metastasis.

●In a French phase II trial of neoadjuvant imatinib, 25 adults with DFSP (20 primary, five recurrence disease) received two months of imatinib (600 mg daily) prior to wide local excision [94]. The COL1A1-PDGFB fusion gene was detected in 21 of 25 patients. A clinical response was noted in only nine patients (36 percent); it is possible that a longer duration of neoadjuvant treatment might prove beneficial. Long-term results are not available from this report.

Imatinib is approved in the United States and Europe for use in adult patients with unresectable, recurrent, and/or metastatic DFSP. This broad indication covers the use of imatinib as an alternative to radiation therapy (RT) for locally advanced, unresectable disease, including cases in which wide surgical excision is not feasible or a postoperative complication or wound healing is a concern, and as an alternative to primary RT for patients with locally recurrent but potentially resectable disease for whom reresection would result in unacceptable function or cosmetic outcomes. A reduced tumor burden might permit a more conservative surgical procedure to be undertaken at a later time.

The optimal duration of neoadjuvant therapy and the appropriate imatinib dose remain unanswered questions. Several phase II studies of neoadjuvant imatinib are ongoing, and eligible patients should be encouraged to enroll [95,96].

Further studies are also needed to determine the value of imatinib in cases of positive margins after surgical excision or in patients with other risk factors for recurrence after complete resection (such as the presence of the DFSP-FS variant). However, given the lack of data regarding benefit of imatinib in any of these settings and the generally good prognosis of patients with resected DFSP, adjuvant imatinib should only be considered in the context of a clinical trial.

Other tyrosine kinase inhibitors — While it would seem likely that other TKIs that target PDGF receptor (eg, sunitinib, sorafenib, and pazopanib) would also be beneficial for DFSP, data for these agents are limited [97-99].

Conventional chemotherapy — Few data are available, but they suggest a lack of responsiveness of DFSP to the combination chemotherapy regimens that are used in other soft tissue sarcomas (eg, doxorubicin, ifosfamide) [53,91,100]. However, at least one case report suggests responsiveness of advanced DFSP to low-dose methotrexate and vinblastine, a similar regimen that is used for patients with desmoid tumors [100]. (See "Desmoid tumors: Systemic therapy", section on 'Combination chemotherapy'.)

SOCIETY GUIDELINE LINKS — Links to society and government-sponsored guidelines from selected countries and regions around the world are provided separately. (See "Society guideline links: Soft tissue sarcoma".)

SUMMARY AND RECOMMENDATIONS

●Clinical presentation – Dermatofibrosarcoma protuberans (DFSP) is an uncommon locally aggressive cutaneous soft tissue sarcoma. Although they rarely metastasize (fewer than 5 percent of cases), all DFSP variants have a propensity to recur locally. (See 'Introduction' above.)

●Principles of surgical treatment for localized disease – The optimal treatment for patients with localized DFSP is complete resection with negative margins. Every attempt should be made to achieve negative margins at the initial surgical treatment. Although many patients with locally recurrent disease can be salvaged with further surgery, the risk of functional deficits and of metastatic disease is increased in such patients. (See 'Importance of resection margins' above.)

•Preoperative diagnostic imaging – Preresection magnetic resonance imaging (MRI) is of benefit for determining the depth of invasion with large or locally recurrent tumors outside of the head and neck or upper thorax; it is of little value in detecting tumor persistence in cases of incompletely extirpated tumors. (See 'Role of diagnostic imaging' above.)

•Wide local excision – One surgical option for localized DFSP is wide local excision (WLE). The optimal width of the resection margin is controversial. Guidelines from the National Comprehensive Cancer Network (NCCN) [30] recommend margins of 2 to 4 cm, to the investing fascia of muscle or pericranium with clear pathologic margins, when clinically feasible. However, this is often compromised by the proximity of critical structures, especially in anatomically challenging areas, such as the head and neck. (See 'Wide local excision' above.)

•Mohs micrographic surgery – Mohs micrographic surgery (MMS) is an alternative surgical approach for large and/or recurrent tumors and those that arise in anatomically confined or cosmetically sensitive areas where maximal preservation of normal tissue is desirable. (See 'Mohs micrographic surgery' above.)

●Adjuvant radiation therapy – Although DFSP is a radioresponsive tumor, radiation therapy (RT) is rarely used as a primary treatment. Conservative resection with postoperative (ie, adjuvant) RT is a reasonable approach in situations where adequate WLE alone would result in major cosmetic or functional deficits, particularly if the expertise to perform MMS is not available. However, RT is not a substitute for inadequate resection. If the margins are positive, reresection is preferred, if feasible without major cosmetic or functional compromise. (See 'Radiation therapy' above.)

●Prognosis of DFSP-FS – DFSP with sarcomatous transformation (DFSP-FS) is a potentially more aggressive tumor than low-grade DFSP. Although sarcomatous transformation increases the risk of metastatic spread, whether it represents an independent negative prognostic factor for local recurrence or long-term outcomes after adequate surgery (WLE) is unclear. (See 'Prognosis' above.)

●Posttreatment surveillance – The majority of local recurrences become evident within three years, but 25 to 30 percent develop after five years. The optimal posttreatment surveillance strategy has not been established. We and others [37] suggest that patients be examined every six months for the first three years, then annually lifelong. (See 'Posttreatment surveillance' above.)

●Treatment for locally advanced and metastatic disease

•Lymph node involvement – For the rare patient with lymph node involvement, we suggest lymphadenectomy because of the possibility of long-term relapse-free survival (Grade 2C). (See 'Treatment of locally advanced, recurrent, and metastatic disease' above.)

•Surgically resectable disease – Resection is a reasonable approach for appropriately selected patients with a controlled primary tumor and isolated, potentially resectable sites of metastatic disease that appear resectable, especially isolated pulmonary metastatic disease. (See "Surgical treatment and other localized therapy for metastatic soft tissue sarcoma".)

•Recurrent, unresectable, or metastatic disease – Imatinib, an orally active small molecule tyrosine kinase inhibitor, can induce tumor regression in patients with recurrent, unresectable, and/or metastatic DFSP. Although indications for imatinib are evolving, it is approved in the United States and Europe for use in adult patients who have unresectable, recurrent, and/or metastatic DFSP. (See 'Imatinib' above.)

It is not clear whether DFSPs that lack the characteristic t(17;22) translocation respond as well to imatinib. However, objective responses have been documented in translocation-negative tumors. Although the NCCN [30] suggests that molecular testing be carried out prior to the institution of imatinib and that use of imatinib be restricted to patients who carry the characteristic t(17;22) molecular defect, we would not restrict a therapeutic trial of imatinib to only those patients whose tumors have the characteristic translocation. (See "Dermatofibrosarcoma protuberans: Epidemiology, pathogenesis, clinical presentation, diagnosis, and staging", section on 'Molecular pathogenesis'.)

●Is there a role for neoadjuvant or adjuvant imatinib? – Further studies are needed to address whether imatinib can be incorporated into surgical management of localized DFSP to permit a more conservative surgical resection, further reduce local and distant recurrence rates, and improve survival. Participation in a clinical trial is preferred. If patients are not eligible, or if a protocol is not available, it seems reasonable to try initial imatinib for patients in the following clinical situations (see 'Is there a role for neoadjuvant or adjuvant imatinib?' above):

•Inoperable presentations, including cases in which wide surgical excision is not feasible or a postoperative complication or wound healing is a concern

•Recurrent disease where unacceptable functional or cosmetic defects will result from reresection

•Extensive advanced locoregional disease, to reduce tumor burden, possibly allowing a more conservative surgical procedure

•In our view, the use of adjuvant imatinib should not be considered a standard approach and should be restricted to clinical trials until further information becomes available.