Desmoid-Type Fibromatosis—Clinical Study of an Uncommon Disease

Abstract

Desmoid-type fibromatosis is an uncommon disease which mimics sarcoma but is more locally aggressive but with less metastatic potential than sarcoma. Diagnosis is difficult and treatment protocols have changed to include more non-surgical options as compared with sarcoma. This is a retrospective study of the clinical presentation, imaging findings, treatment given, and outcomes of pathologically proven desmoid-type fibromatosis patients who presented to government medical college Kozhikode.

Introduction

Desmoid-type fibromatosis (DF) is an uncommon benign disorder due to proliferation of fibroblasts/myofibroblasts cells that can involve any anatomical location [1]. Unlike soft tissue sarcoma, it lacks metastatic potential but is notorious for local recurrence after excision causing major morbidity for the patient and is classified by the WHO as an intermediate/locally aggressive tumour [2].

Aetiology of DF remains unclear and is most often sporadic though it is associated with Familial Adenomatous Polyposis. The propensity of DF to occur following pregnancy suggests a degree of hormonal sensitivity but direct correlation with oestrogen levels has never been demonstrated [3]. The most common site of sporadic DF is the abdominal wall while the mesentery is the preferred site for FAP-associated DF.

Patients present with a painless mass in the extremity or abdominal wall. Intrabdominal desmoids may rarely present as bowel perforation or obstruction. It usually presents as a single lesion, but recurrences can be multifocal.

MRI is the imaging modality of choice for extremity DF and CT for abdominal variants. The most commonly observed MR imaging appearance of DF is a heterogeneous pattern, with signal iso- to hyperintense to skeletal muscle on T2-weighted images, and isointense to muscle on T1-weighted images [4]. The key diagnostic feature is the hypointense bands on T2W images [5, 6]. On CT scan, DF appears as a soft-tissue mass often sharply marginated in abdominal-wall tumours, or with ill-defined infiltrative margins in extra-abdominal or mesenteric tumours [7]. DF shows variable attenuation, similar to or slightly higher than that demonstrated by skeletal muscle, with hyper- and hypoattenuation, probably reflecting collagen and myxoid elements, respectively [8].

Pathology

DF is composed of fibroblasts with no malignant features embedded in a dense collagen matrix with infiltrative growth pattern. Immunostaining is positive for beta-catenin (nuclear) but is not specific for DF [9]. DF are supposed to be hormonally sensitive because young females are most affected. Oestrogen receptor positivity by immunohistochemistry has varied from 0 to 54% in various series [10–12]. However the response of these tumours to anti-oestrogens is not related to receptor status [13].

Treatment

Surgical excision with wide margins used to be the standard of care mimicking soft tissue sarcoma treatment guidelines. However achieving surgical margins are more difficult than in sarcoma because of infiltrative growth pattern. Also wider surgical margins do not correlate with lesser recurrence, and the resultant morbidity may not justify performing extensive surgery [14–17]. Observational and medical management of fibromatosis has been described by several authors with surgery reserved for those with disease progression. Medical treatments used by various investigators include anti oestrogen therapy—tamoxifen in doses from 20 mg/day [18] to 160 mg/day [19]; toremifene 200 mg/day [18], progesterone 100 mg/day, cytotoxic chemotherapy–doxorubicin with dacarbazine or doxorubicin with cyclophosphamide and vincristine or methotrexate with a vinca alkaloid [20, 21]—disease stabilization was reported in at least 50% of these patients with 20–28% of patients showing regression of their disease [22, 23].

The following are the challenges in the management of DF:

1. Establishing a diagnosis as the histologic appearance of bland fibroblasts with normal mitotic activity makes distinction from normal connective tissue difficult.

2. Given its deep seated location obtaining biopsies is tricky in case of abdominal fibromatosis

3. Lack of a characteristic immunohistochemical profile makes diagnosis difficult

4. Imaging characteristics resemble normal connective tissue or sarcoma

5. Achieving margins during surgery is difficult due to the creeping nature of infiltration

6. Monitoring patients on non-surgical treatment plan

7. Deciding between the various medical options available

Study

There were 11 cases of biopsy proven desmoid-type fibromatosis (DF) that reported to our hospital from July 2015 to July 2018. Of these, 10 were females and 1 was male. Follow-up varied from 4 to 1 year with mean follow-up of 24 months. The anatomic distribution of these lesions is given in Table 1.

Table 1.

Anatomic distribution of these lesions

Patient no	Sex	Age	Location of tumour
1	F	34	Right subscapular area
2	F	34	Right calf
3	F	40	Left cubital fossa
4	F	36	Parietal abdominal wall
5	M	14	Suprapubic area
6	F	18	Suprapubic area
7	F	20	Suprapubic area
8	F	41	Right quadratus lumborum
9	F	18	Forefoot
10	F	60	Shoulder girdle
11	F	70	Right subscapular and axillary area

Clinical Presentation

All patients with lesions of the extremities presented with swelling while pain was the presentation of those with lesions of the parietal wall of abdomen and retroperitoneum. The patient no. 3 presented with a recurrence after excision from another hospital 3 years ago and had pain, swelling, and restriction of movements. Patient 10 was managed non-operatively at a local hospital for 3 years before she presented to us. Two out of 3 lesions in the shoulder girdle had restriction of movements. Nine out of 11 patients were below 45 years of age, and only 2 were elderly. Three of them were below 20 years of age. Both patients who had lower limb disease had deformities of the involved side; 1 was affected with polio, and the other had a congenital deformity.

Imaging Diagnosis

All patients except patient no. 3 and 10 had imaging at our hospital. All extremity lesions underwent MRI while trunk and retroperitoneal lesions underwent CT scan at our centre. When hypointense bands were seen on MRI a high suspicion for desmoid tumour was suggested. Six patients underwent CT scan, and 4 patients had MRI (Table 2).

Table 2.

Radiologic diagnosis

Patient no.	Imaging done	Was imaging suggestive of DF?
1	CT	No
2	MRI	Yes
3	NA	NA
4	CT	No
5	CT	Yes
6	CT	Yes
7	CT	No
8	CT	No
9	MRI	Yes
10	MRI	Yes
11	MRI	Yes

Pathologic Diagnosis

Preoperative biopsy was attempted in all cases except patient 3 and 10. Patient 3 had undergone excision from another hospital and her slides could not be reviewed. Patient 10 had biopsy from another hospital and had received medical treatment for the same for 3 years and her slides could not be reviewed. In all except patient 8, biopsy was adequate and representative to reach a final conclusion. Except for patient 11 who underwent an open incisional biopsy for all other patients a core needle biopsy was successful. The pathologic features of DF is fibroblasts with no malignant features embedded in a dense collagen matrix with infiltrative growth pattern (Fig. 1). Immunostaining is positive for beta-catenin (nuclear) but is not specific for DF (Fig. 2). All these patients had bland fibroblasts on histologic examination. Patient 8 underwent excision, and the histology of the resected specimen was consistent with DF.

Fig. 1.

H&E stain

Fig. 2.

IHC for B-catenin

Treatment

The treatment given has been summarized in Table 3. Except for patient 1 who was our first patient, our policy has been to offer medical treatment to all patients confirmed to have DF and offer surgery only if disease progresses or becomes symptomatic.

Table 3.

Treatment outcomes

Patient no		Treatment	Whether negative margins were obtained	Recurrence if surgery was performed
1	Right subscapular area	Surgery	Yes	No
2	Right calf	Medical	NA	NA
3	left cubital fossa	Surgery	No	Yes
4	Parietal abdominal wall	Medical ➔ surgery	Yes	No
5	Suprapubic area	Medical	NA	NA
6	Suprapubic area	Medical	NA	NA
7	Suprapubic area	Medical	NA	NA
8	Right quadratus lumborum	Surgery	No	Yes
9	Forefoot	Medical	NA	NA
10	Shoulder girdle	Medical	NA	NA
11	Right subscapular and axillary area	Medical	NA	NA

Patient 1 underwent surgery at our hospital as the primary treatment as imaging revealed a well-defined swelling and excision without morbidity was possible. She is kept on follow-up only. Patient 4 was initially kept on medical management with tamoxifen but complained of increasing pain and size of the swelling. She underwent excision of the anterior abdominal wall lesion with mesh reconstruction and received postoperative radiotherapy. Surgical margins of ≥ 1 cm were obtained for both these patients. Patient 3 had undergone excision at another hospital and presented with a recurrence. Her excision was incomplete and on developing recurrence has been put on tamoxifen. Patient 8 underwent excision because biopsy was inconclusive, and imaging was suggestive of sarcoma. We could not achieve negative margins for her because of the lumbar plexus was found coursing through the lesion. She has been put on postoperative tamoxifen. The remaining 7 patients have not undergone excision and are on high dose tamoxifen of 20 mg 5 times a day with 3 monthly follow-ups. Follow-up is with clinical examination and yearly imaging with MRI. None of these patients have had progression of symptoms or size to warrant an excision.

Discussion

DF predominantly affected young females in our series (91% of affected members were female). This is consistent with other series which had 89 [15] to 68% [16] of female patients. Truncal regions were involved in 72% of our patients and extremities in 28%. Peng et al. [16] reported 68% of truncal DF. None of our patient had associated familial adenomatous polyposis. CT scan was suggestive of DF only 2 out of 6 cases. In the remaining cases, CT diagnosis was sarcoma. When MRI was performed, hypointense bands on T2 imaging were present in all the 4 cases. A larger series of 29 cases by Lee et al. had 62% of MRI having the same finding [24]. Deformity may be a risk factor for extremity DF, but the numbers are too small to draw any strong associations. Non-surgical treatment has been successful in keeping the disease at control but requires good communication between doctor and patient to convince patients who might have been advised excision from other centres. Of the 8 patients who were kept on medical treatment only 1 required surgery during follow-up (12.5%). Progression free survival of 50% at 5 years has been shown by retrospective series for asymptomatic patients managed with a frontline conservative wait and watch policy [25]. Close follow-up enables surgical intervention if disease shows progression. We have not tried second line medical management in any of our patients.

Conclusion

DF must be a differential diagnosis to be kept in mind while dealing with soft tissue swellings. Imaging suggestive of diffuse infiltration and clinical finding of restriction of movements in girdle swellings must raise a high suspicion for the same. Proper biopsy and discussion with radiologist is invaluable in reaching a diagnosis. Discussion in tumour board can give the treating doctor confidence in implementing non-surgical treatment line for this disease which will often recur with increasing morbidity if operated upon in the first instance. Surgery can be reserved for painful, progressive lesions, or in non-functional limbs.