If you don't remember your password, you can reset it by entering your email address and clicking the Reset Password button. You will then receive an email that contains a secure link for resetting your password
If the address matches a valid account an email will be sent to __email__ with instructions for resetting your password
Aggressive fibromatosis, also known as desmoid-type fibromatosis (DTF) or desmoid tumor, is an uncommon locally invasive tumor. Because of its low incidence and variable behavior, DTF is often first seen by physicians who are not familiar with it, and recent advances in understanding this disease have led to changes in treatment approaches. The Wnt (β-catenin) pathway appears to play a key role in DTF pathogenesis, and recent studies of DTF biology suggest a possible model of DTF pathogenesis. Histologically, DTF shows a poorly circumscribed proliferation of myofibroblast-like cells with variable collagen deposition, similar to the proliferative phase of wound healing, and DTF has been associated with trauma and pregnancy. Desmoid-type fibromatosis may be a useful model of the tumor stroma in carcinomas as well as other fibrosing diseases such as progressive pulmonary fibrosis. The clinical course of DTF can vary greatly among patients, complicating the determination of the optimal treatment approach. Treatment options include surgery, nonsteroidal anti-inflammatory drugs with or without hormonal manipulation, chemotherapy, radiation therapy, and other forms of local therapy. Many treatments have been used, but these are not without toxicities. Because of the variable nature of the disease and the potential morbidity of treatment, some cases of DTF may do better without treatment; simple observation is often the best initial treatment. This review used a PubMed search from January 1, 1980, through October 31, 2016, using the terms fibromatosis and desmoid and discusses DTF disease characteristics, pathophysiology, and treatment options as well as examines several cases illustrating key points in the biology and treatment of this heterogeneous disease.
The clinical course of desmoid-type fibromatosis (DTF), an uncommon locally invasive tumor, can vary greatly among patients, complicating the determination of the optimal treatment approach.
The Wnt (β-catenin) pathway appears to play a key role in DTF pathogenesis.
Treatment options include surgery, nonsteroidal anti-inflammatory drugs with or without hormonal manipulation, chemotherapy, radiation therapy, and other forms of local therapy. Many treatments have been used, but these are not without toxicities.
Because of the variable nature of the disease and the potential morbidity of treatment, some cases of DTF may do better without treatment; simple observation is often the best initial treatment.
Desmoid-type fibromatosis may be a useful model of the tumor stroma in carcinomas as well as other fibrosing diseases such as progressive pulmonary fibrosis.
The term fibromatosis encompasses 2 general groups of tumors: superficial and deep fibromatoses. The superficial fibromatoses include palmar fibromatosis or Dupuytren contracture, plantar fibromatosis, and penile fibromatosis or Peyronie disease. Deep or aggressive fibromatosis, also known as desmoid-type fibromatosis (DTF) or desmoid tumor, is a clonal locally invasive tumor that does not metastasize.
This review used a PubMed search from January 1, 1980, through October 31, 2016, using the terms fibromatosis and desmoid.
Histologically, DTF shows a poorly circumscribed proliferation of myofibroblast-like cells with variable collagen deposition. These myofibroblastic cells are histologically similar to the proliferative phase of wound healing, and DTF has been associated with trauma, pregnancy, and oral contraceptive use.
and surgery may sometimes promote growth of DTF. The natural history of DTF is highly variable. This review discusses DTF disease characteristics, pathophysiology, and treatment options as well as examines several cases illustrating key points in the biology and treatment of this heterogeneous disease.
Epidemiology of DTF
Desmoid-type fibromatosis most commonly arises between the ages of 15 and 60 years, with a female predominance of 2- to 3-fold.
Familial adenomatous polyposis–associated DTF is more frequently abdominal, especially in the Gardner variant of FAP, which is characterized by intestinal polyposis, osteomas, fibromas, and epidermal inclusion (“sebaceous”) cysts.
Genetic predisposition to DTF in patients with FAP independent of germ line APC mutation has also been described, suggesting the existence of genes independent of APC that influence DTF formation in FAP.
Infantile fibromatosis (so-called diffuse or mesenchymal type of fibromatosis) is not discussed here and usually occurs before the age of 2, most commonly in the first few months of life; it may recur locally, but does not metastasize.
Histology of DTF
Histologically, DTF appears as a poorly circumscribed proliferation of myofibroblastic cells with variable collagen deposition. Typically, the margins of the tumor are difficult to assess at the time of surgery, and the final margins are often positive. Desmoid-type fibromatosis tumors are morphologically heterogeneous and may exhibit striking morphological intra- and intertumoral heterogeneity (Figure 1, A). In some areas tumors may resemble fibroblasts of inactive fibrous tissue, whereas other areas resemble the active fibroblasts of wound healing. This morphological heterogeneity covers a spectrum ranging from areas in which cells have oval nuclei containing pale-staining vesicular euchromatin and small nucleoli to areas in which cells have elongated nuclei that stain darkly with hematoxylin, reflecting heterochromatin.
Figure 1, B, shows an area that appears inactive, with sparse cells with narrow, darker-staining nuclei and few mitoses, in which in general there is more collagen deposition, imparting a more pink (collagenous) coloration to these inactive areas. Typically the areas with more “transcriptionally inactive” cells are often separated by extensive collagen.
Figure 1, C, from the same tumor shows an area that appears histologically active, characterized by cells with plump, light-staining oval nuclei, greater cell density, increased mitotic activity, and less collagen. Digital assessment of chromatin density and average nuclear size and pathological assessment of tumor activity were strongly correlated in 1 study,
with only 3 mutations reported: S45P, S45F, and T41A. S45F and T41A were the most common, with S45P seen in less than 10% of cases. Several cases of APC mutations have also been found in sporadic cases of DTF.
In a study of 17 FAP-associated DTF and 38 sporadic DTF cases using comparative genomic hybridization and multiple ligation-dependent probe amplification, a limited number of genetic changes was observed in 44% of tumors.
In addition, induction of stabilized β-catenin in a transgenic mouse model leads to hyperplastic cutaneous wounds and the development of DTF, providing further evidence that β-catenin plays a role in these fibroproliferative diseases.
Abnormal growth factor production (including transforming growth factor [TGF] and platelet-derived growth factor [PDGF]) has been associated with hereditary gingival fibromatosis and plantar fibromatosis
β-Catenin is regulated by a destruction complex (Figure 2) including APC, which has multiple β-catenin binding sites, axin, β-catenin, casein kinase 1 (CK1), glycogen synthase kinase 3β (GSK3), and protein phosphatase 2A. β-Catenin is phosphorylated in this complex by GSK3 after a “priming” phosphorylation by CK1, which leads to ubiquitination and subsequent degradation in the proteasome.
Nuclear β-catenin can act as a transcriptional activator when bound to a member of the T-cell factor/lymphocyte enhancer family, leading to the formation of nuclear β-catenin/T-cell factor/lymphocyte enhancer complexes, changing the way they bind promotor regions of DNA and altering gene transcription.
Four genes—a disintegrin and metalloproteinase gene 12 (ADAM12), fibroblast activation protein 1α (Fap-1α), Wnt 1 inducible signaling pathway protein-1 (WISP1), and SRY-box 11 (SOX11)—have been reported to be overexpressed in DTF compared with 16 nonneoplastic tissues,
). ADAM12 identifies a proinflammatory subset of PDGF receptor-α (PDGFR-α)–positive stromal cells residing in the perivascular space that can be activated by acute injury and can differentiate into myofibroblasts and act as progenitors for a large fraction of the collagen-producing cells generated in scarring; these cells are progressively eliminated during normal wound healing.
Thus, the available data suggest a possible model of DTF pathogenesis, in which an activating stimulus, such as trauma with associated inflammation and growth factor production, in the setting of deregulation of β-catenin, leads to up-regulation of β-catenin
(Figure 3, right side). Reactive oxygen species produced by neutrophils have been shown to have the potential to induce mutations in DNA. In rare cases the inciting event may stimulate a progenitor cell that does not have baseline β-catenin dysregulation (left-hand-side of the figure). β-Catenin can then translocate to the nucleus, complex to transcription factors, bind the WISP1 promotor, and increase WISP1 production. WISP1 may then bind its receptor and induce β-catenin nuclear translocation,
resulting in a prosurvival signal, and further stimulate WISP1 production, and production of ECM proteins including collagen, leading to fibrosis. WISP1 binding to the tumor cells can then further stimulate tumor growth. Myofibroblasts are functionally heterogeneous and can be generated from multiple cell types.
potentially adding nonclonal normal myofibroblasts to the tumor. These recruited cells, whether normal or part of the true clonal tumor, are Fap-1 positive and produce a number of ECM proteins, including collagen, leading to fibrosis. Although the role of SOX11 is not clear, studies have reported that SOX11 assists mesenchymal stem cell proliferation and retention of pluripotent potential.
In some cases DTF tumors, and their constituent cells, may stabilize or regress, with a decrease in expression of ADAM12, FAP-1α, WISP1, and SOX11 (Figure 4). In most cases of DTF, different areas of the tumor show either active or inactive areas,
indicating that the balance of these factors leading to progression or regression operate differently in different parts of the tumor. The mechanisms regulating these factors are unknown.
Evidence for Clinical Treatments of DTF
The natural clinical course of DTF can vary greatly among patients, complicating the determination of the optimal treatment approach. Clinical trials exhibiting the best approach in a particular patient are lacking. Treatment options include surgery, nonsteroidal anti-inflammatory drugs with or without hormonal manipulation, chemotherapy, radiation therapy, and other forms of local therapy. Many treatments have been used, but these are not without toxicities. Because of the variable course of the disease and the potential morbidity of treatment with the result that some cases of DTF may do better without treatment, Lewis et al,
However, pregnancy is also associated with changes in circulating growth factors and immune modulators, including vascular endothelial growth factor, TGF-β, and insulin-like growth factor 1; these all could also be involved.
In another study of 147 patients, 97% of whom were young women with abdominal wall DTF, 102 underwent initial observation; of these, 29 had spontaneous regression and only 16% went to surgery by 3 years.
Inflammation from other types of trauma also may augment or stimulate recurrence; however, DTF does not metastasize. Although a marginal resection is associated with a worse outcome than a complete resection, the nature of the surgical procedure is strongly influenced by tumor location and associated anatomical and functional consequences.
In another multivariate analysis of 426 cases of sporadic DTF, 87% of cases were treated surgically, and about 50% of cases recurred; only age, tumor size, and tumor site were independent prognostic factors of recurrence.
In 1 study of 6 radiation-induced sarcomas in patients with DTF whose original tumor had a mutation in CTNNB1, 3 had the same CTNNB1 mutation as the original DTF, and 3 had no CTNNB1 mutation, suggesting that some cases of DTF were not derived from the original DTF tumor clone.
Comparative evaluation of different therapies is hindered by the fact that most case series are not randomized; the variable natural history of DTF further complicates the interpretation of these studies. In cases that respond to drug, the optimal length of treatment is unknown. Treatment approaches range from holding treatment at an arbitrary time in the setting of stable disease to prolonged treatment in responders, followed by abrupt cessation of therapy or gradually weaning treatment intervals or dose.
Magnetic resonance imaging (MRI) is the best imaging technique for diagnosis and monitoring of DTF.
In some cases MRI may reveal changes associated with increased collagen deposition and decreased cellularity, such as a loss of T2 signal, suggesting either a response to treatment or a spontaneous decrease in disease activity.
In 1 study of 25 patients with DTF (8 sporadic and 17 associated with FAP), a regimen of tamoxifen (120 mg/d) and sulindac (300 mg/d) was not highly effective in preventing DTF recurrence after surgery, but was still felt to be potentially useful in other settings, in which stable disease was the most common response.
Desmoid-type fibromatosis also expresses androgen receptors; testosterone can stimulate DTF cell growth in vitro and DTF development in mouse models, suggesting androgen blockade as another potential hormonal approach.
Sarcoma Alliance for Research through Collaboration (SARC) Efficacy of imatinib in aggressive fibromatosis: results of a phase II multicenter Sarcoma Alliance for Research through Collaboration (SARC) trial.
suggesting that in some cases efficacy may be due to effects on targets other than KIT (kit proto-oncogene receptor tyrosine kinase). Trials of tyrosine kinase inhibitors in DTF are ongoing (Table). Pegylated-liposomal doxorubicin (PLD) is particularly attractive, given its efficacy and low toxicity profile and is becoming widely used.
Antibodies to WISP1 inhibit fibrosis in mouse models of bleomycin lung toxicity, suggesting this as a potential target for the future treatment of select cases of DTF as well. Similarly, the beneficial response of DTF to certain chemotherapy approaches suggests that a similar approach could be useful in severe cases of idiopathic pulmonary fibrosis. Altering Notch signaling with γ-secretase inhibition is also under study, and an adenosine monophosphate–activated protein kinase activator inhibits peritoneal fibrosis (a complication of peritoneal dialysis) in a mouse model.
Preliminary data suggest activity of a γ-secretase inhibitor in DTF. Some ongoing trials for DTF are listed in the Table.
In some cases DTF tumors, and their constituent cells, may stabilize or regress, with a decrease in expression of biochemical markers of disease activity (Figure 4). The observation that DTF tumors sometimes stabilize or regress implies that the tumor myofibroblasts retain sensitivity to a regulatory system, likely an autocrine or more likely paracrine signaling system, similar to that of wound healing. That DTF tumors may still subsequently become active again after stabilization or regression implies that a population of cells remains that retains the ability to respond to some proinflammatory and/or profibrotic stimuli, in some cases induced by trauma, associated inflammation, or other physiological conditions, such as pregnancy. The degree to which recruited normal myofibroblasts contribute to the mass of clonal myofibroblasts in an individual DTF tumor could potentially affect the tumor behavior.
Predicting DTF Behavior
Predicting which treatment is most appropriate for a particular patient, such as the observation approach, would be useful. A nomogram using tumor size, location, and patient age has been reported to be useful in predicting recurrence after surgery.
Some studies, but not all, have suggested that the location of the β-catenin mutation correlates with differences in clinical course of sporadic DTF. For example, tumors with S45F mutations in CTNNB1 may be at a higher risk of recurrence.
Correlation of CTNNB1 mutation status with progression arrest rate in RECIST progressive desmoid-type fibromatosis treated with imatinib: translational research results from a phase 2 study of the German Interdisciplinary Sarcoma Group (GISG-01).
Desmoid Working Group Management of sporadic desmoid-type fibromatosis: a European consensus approach based on patients' and professionals' expertise—a sarcoma patients EuroNet and European Organisation for Research and Treatment of Cancer/Soft Tissue and Bone Sarcoma Group initiative.
In 1 study, immunohistochemical staining for ADAM12, Fap-1α, and WISP1 correlated with nuclear chromatin density and was higher in patients with an early recurrence (<1 year after surgery compared with no recurrence at 5 years).
,p1650 Fibroblasts and myofibroblasts in neoplasms can secrete various trophic, mitogenic, and proinflammatory growth factors including hepatocyte growth factor, epidermal growth factor, TGF-β, and insulin-like growth factor 1 (reviewed in reference
), possibly influencing growth of the clonal neoplastic cells. Because of their potential contribution to tumor biology, targeting the normal stromal myofibroblasts in tumors is an interesting potential approach to cancer treatment.
As DTF closely resembles wound healing, it may be a potentially useful model to study the role of tumor stroma. Indeed, some studies suggest that gene signatures similar to that seen in DTF correlate with clinical outcome in some malignancies.
Case 1: Slow Spontaneous Regression of an Extra-Abdominal DTF in a Man
A 28-year-old man noted a small mass in his right chest near the sternum. He developed psoriatic arthritis 8 months later and began treatment with methotrexate. A year after initiating methotrexate, etanercept was added. At presentation 8 months after starting etanercept, he was not sure the mass had grown over the part year, but he now had occasional twinges of pain lasting a few seconds, from none to several times a day. Thus, the tumor progressed symptomatically while on methotrexate. Examination revealed a firm, fixed, nontender parasternal mass. Computed tomography revealed a mass involving the sternum growing through the chest wall (Figure 5, top panel) and a biopsy revealed DTF. Consultation at another institution recommended surgical removal of the chest wall mass; however, he was observed, and 3 months later his symptoms and imaging of the mass were unchanged. Ten months after the biopsy, the mass was slightly smaller and symptoms were unchanged. At 35 months after the biopsy, his symptoms had resolved and the mass was smaller (Figure 5, bottom panel). He remains symptom free 50 months after the biopsy and continues the observation. This case exhibits slow spontaneous regression of an extra-abdominal DTF in a man, not related to estrogen.
Case 2: DTF Caused by Local Trauma/Inflammation and Stable Disease After Methotrexate and Vinblastine
A 33-year-old man presented with a painful mass in the arm. He had an influenza shot 1.5 years before presentation, and shortly thereafter he became aware of a persistent discomfort in the region of the injection site that gradually progressed, and a painful mass developed. The size of the mass and degree of pain progressed markedly over the 2 months before presentation. Examination revealed a slightly tender warm 10 cm hard mass fixed to the underlying tissue in the proximal right arm. Magnetic resonance imaging revealed a 7×5×8.5 cm mass along the triceps muscle that was hyperintense on a fluid-sensitive image with some areas of heterogeneity (Supplemental Figure 1, available online at http://www.mayoclinicproceedings.org). A tru-cut biopsy revealed DTF, and an open biopsy performed to exclude a low-grade sarcoma also revealed DTF. Surgical treatment was felt to require shoulder disarticulation. As surgical treatment was felt to result in significant morbidity, chemotherapy with methotrexate and vinblastine was begun. At 6 weeks of treatment, there was no clear evidence of change in tumor size or symptoms, and he moved to a different state in which chemotherapy was continued for 2 more months without change in tumor size. The tumor was then surgically excised. He was sent for consideration of postoperative radiation therapy and thereafter lost to follow-up. This case exhibits DTF development after local trauma/inflammation, potential significant morbidity of treatment (shoulder disarticulation or more limited disfiguring surgery), and disease stabilization with methotrexate and vinblastine.
Case 3: Aggressive Multifocal DTF Controlled With PLD
A 32-year-old man with Gardner syndrome presented with painful extra-abdominal desmoid tumors as well as large intra-abdominal desmoids requiring opiates. He also had a pulmonary embolus and venous thrombosis and was taking coumadin. He had been treated a year earlier with tamoxifen for 8 weeks, but tumors grew during this period. He began treatment with PLD and had a good response. Treatment was held after 6 cycles. Desmoid-type fibromatosis progression was noted 10 months after the last chemotherapy, and he received imatinib 400 mg/d orally, but it progressed. He began treatment with PLD and noted stabilization of pain after 1 month and some tumor shrinkage at 3 months. Three months later, imaging revealed further tumor regression, and the interval between PLD treatments was increased. Subsequent imaging revealed continued gradual tumor shrinkage and then stabilization, and PLD was discontinued after 1.5 years of treatment. Imaging found stable disease at 3 years after reinitiating PLD, but 4 months later (40 months after first initiating chemotherapy) he developed increasing pain and progression of DTF on imaging, and PLD was reinitiated. One month later he developed small bowel obstruction, bacteremia, and renal failure and had a decompressive gastrostomy tube placed. His subsequent course was complicated, and he eventually entered a hospice program and died 5.3 years after initial chemotherapy. This case exhibits aggressive multifocal DTF, long-term control with PLD, and severe DTF-associated morbidity.
Controlled trials are needed to better define optimal treatment approaches. Future clinical trials must consider several aspects of DTF biology. First, because of the highly variable clinical course of DTF, patients must be carefully stratified at entry. Factors to consider include rate of tumor growth (tumor growth rate should be quantitated before treatment). Other stratification variables should include age at diagnosis, tumor location (mesenteric, abdominal wall, central extra-abdominal, and extremity), β-catenin mutation and APC mutation status, relation to pregnancy, symptoms, and tumor size. Samples should be obtained for future, more detailed genetic analysis; consideration should be given to obtaining core biopsies from different parts of the tumor, if possible, given the known intratumoral variability. Watchful waiting should be the first treatment, if possible, and when treatment is initiated, randomization to 2 treatments is needed. The “standard” treatment can be debated, but given its efficacy, tolerability, and increasing popularity, PLD would be one consideration. Finally, a decision on how long to treat a responding tumor and how to quantify tumor response must be considered. RECIST (Response Evaluation Criteria In Solid Tumors) is known to be a poor measure of response in DTF, but some measure of size (optimally careful analysis of tumor volume) and tumor “activity” (possibly determined by contrast enhancement or changes in T1/T2 signal on MRI) need to be considered. Because tumor stabilization can be a beneficial outcome, a measure of symptoms experienced by the patient should be included as well, including a quality of life assessment such as QLQ-C30 or FACTG. Perhaps better than nonlinear subjective variables such as a “pain scale” are clear measures such as the following: is a pain medication required, how much pain medication is used, does the tumor interfere with sleep (yes/no), how far can the patient walk before tumor pain limits the activity (this could be objectively determined at each clinic visit for cases with serious symptoms), and what is the range of motion of the affected body part. Only with carefully controlled trials that use careful stratification based on known variables can the best treatment approaches for DTF be determined.
Because of the heterogeneity of the biological behavior of DTF, the optimal approach to treatment is unclear. Historically, surgery was the mainstay of treatment, but recurrence after surgery is common.
In addition, trauma can stimulate DTF growth, and surgery is a form of trauma associated with inflammation and production of various growth factors important in wound healing that may also stimulate clonal DTF cells; this is especially true in the case of FAP-associated mesenteric DTF.
Although surgery remains an option for the initial treatment in which the expected morbidity is low and the chance of complete removal is high, given the usually slow progression of DTF, a regimen with low toxicity is the preferred initial approach and observation has become more standard.
Of course, the appropriate clinical decision depends on the particular case, and some require aggressive chemotherapy initially.
Desmoid-type fibromatosis is an uncommon locally invasive tumor. Because of the variable nature of the disease and the potential morbidity of treatment, some cases of DTF may do better without treatment; simple observation is often the best initial treatment.
We thank Shelly Marette, MD, for assistance in interpreting magnetic resonance images, Michael Franklin, MS, for editorial assistance, and J. Carlos Manivel, MD, for a critical review of the manuscript.
Supplemental material can be found online at: http://www.mayoclinicproceedings.org. Supplemental material attached to journal articles has not been edited, and the authors take responsibility for the accuracy of all data.