Are You Confident of the Diagnosis?

What you should be alert for in the history

Superficial acral fibromyxoma (SAF) presents as a slow-growing solitary tumor that typically arises on the hands and feet, with a predilection for the periungual region of fingers and toes. The tumors are frequently asymptomatic, but may be uncomfortable with pressure or, on occasion, painful. The duration prior to resection has ranged from 3 months to 30 years (median approximately 3 years).

Antecedent trauma was reported in about a fifth of the originally described cases, but is not typically part of the clinical history. SAF typically affects adults, but the reported age range is 14-91 years, with a median of around 45-50 years of age.

Characteristic findings on physical examination

SAF presents as a solitary nodule, which may be exophytic, dome-shaped, polypoid or verruciform. There may rarely be ulceration. The tumors are most often located on the toes, especially the great toe, followed by the fingers, with less frequent involvement of the heel and palm. A periungual or subungual location is typical, with nailbed or near nailbed involvement reported in up to 75% of cases. This may cause nail deformity. Deformity of the underlying bone has also occurred.

Continue Reading

The size of the lesions has been reported to range from 0.5 to 5cm (median 1.5cm in one series). In two large series, the mean size was reported to be 1.75cm and 1.92cm.

Expected results of diagnostic studies

Histopathology of SAF is characterized by dermal-based, frequently exophytic, lobular non-encapsulated tumor composed of spindled and stellate cells embedded in a stroma that is variably myxoid, myxocollagenous, or collagenous (Figure 1).

Figure 1.

Scanning magnification of a dermal-based exophytic tumor with a fibromyxoid stroma, surmounted by an eroded hyperplastic epidermis (H&E)

These stromal changes vary somewhat based on site. In the original series, at least 50% of tumors with myxoid or fibromyxoid stroma arose on the hands, while all of the tumors with a predominantly collagenous matrix involved the toes. The stromal changes may represent a continuum, such that there may be myxoid and collagenous areas juxtaposed (Figure 2, Figure 3). Additionally, the stroma may be myxoid initially, and collagenous with diminished vasculature on recurrence.

Figure 2.

Areas of tumor with myxoid and collagenous stroma (10× H&E)

Figure 3.

Closer view of myxoid stroma, demonstrating blands spindled and slightly stellate cells and small capillaries in a myxoid background (H&E)

There is an accentuated microvasculature, especially if the stroma is myxoid. Vessels are thin, and minimally dilated. Occasionally, erythrocyte extravasation may be identified. Mast cell infiltration is characteristic, but otherwise, significant inflammation is not seen, unless there has been surface ulceration.

The spindled cells proliferate randomly, with areas of loosely storiform and occasionally fascicular growth. There may be intervening collagen bundles, separated by myxoid ground substance. Rare cases have been reported to have a lipomatous component, although it is possible that these intra-tumor adipocytes may represent entrapped subcutaneous tissue.

SAF is dermal-based, but subcutaneous involvement is common. The tumor may extend into the subcutis with a pushing border (Figure 4) or may have an irregular infiltrative border. Extention into fascia with erosion or focal extension into bone has rarely been described. Necrosis should be absent.

Figure 4.

low power demonstrating subcutaenous extension of SAF (2.5× H&E)

Occasional multinucleate cells may be found. Atypia and pleomorphism are usually mild, if noted at all, but occasionally moderate cytologic atypia may be encountered.

Mitotic figures are also not common (usually less than 1 per 50 high power fields). Mitotic figures that were identified in seven of 37 tumors reported in the original case series of SAF ranged from 1-7 per 50 high power fields. Only two of these tumors averaged more than 1 mitosis per 10 high power fields. Mitoses should be normal, and should not be associated with significant cytologic atypia.

Immunohistochemistry demonstrates CD34 (Figure 5, Figure 6), CD99, and EMA positivity in the lesional cells. Most of the cells stain diffusely for CD34 and CD99, but this is not uniformly observed. Cases with only focal positivity for CD34 have been reported, and occasional tumors may be negative.

Atypical multinucleate cells have been reported to be CD34 positive. EMA is typically only focally positive (Figure 7). Vimentin is usually strongly positive. The immunophenotype varies with the stromal changes. Highly collagenous tumors tend to have less EMA positivity and more prominent CD34 staining. The tumor cells also may stain positive for CD10 and nestin. Factor XIIIA stains variable fibrohistiocytic cells. Lesional cells are negative for S100, HMB45, keratin, smooth muscle actin, muscle specific actin, and desmin. Ki67 shows a proliferative fraction of 1-2%.

Figure 5.

SAF demonstrating positive staining with CD34 in most of the spindled cells (10x Immunohistochemical stain for CD34)

Figure 6.

CD34 staining of bland spindled cells (Immunohistochemical stain 20x)

Figure 7.

SAF demonstrating bland spindled and stellate tumor cells with positive staining with epithelial membrane antigen (EMA) (Immunohistochemical stain 20X)

The epidermis often shows verrucous epidermal hyperplasia, with hypergranulosis and hyperkeratosis. There may be effacement or ulceration. Overlying nailbed epithelium may show papillary hyperplasia. Ultrastructural studies demonstrate spindle cells with plump ovoid nuclei and prominent nucleoli. There is evidence of collagen production, but actin filaments, Schwannian, and neural differentiation are absent.

A single case report of a tumor described as a biphasic spindle cell tumor with prominent myxoid alternating with collagenous architecture, but that had negative staining for CD34, EMA, and CD10 was interpreted to represent SAF. Ultrastructural studies were detailed, and felt to support a purely fibroblastic cell of origin in this tumor. This immunophenotype is not typical of SAF, and since the histology is also not classic, it is unclear if it truly represents a case of SAF.

Plain radiographs demonstrate a circumscribed soft tissue tumor that may be associated with a saucer-like erosion of the bone cortex, causing a scalloped margin of the digit. A lytic lesion of the distal phalynx with sclerosis associated with a soft tissue mass has been reported. The radiologic features might suggest an epidermal inclusion cyst. A sclerotic base with buttressed margins has been described, consistent with changes of chronic pressure erosion. Buttressed margins might suggest periosteal origin, but periosteal chondromas are excluded by the absence of calcification.

Magnetic resonance imaging (MRI) may demonstrate a T2 hyperintense soft tissue tumor with changes of interosseous extension. A low T1- and T2-weighted signal intensity suggesting a fibrous nature has been described. The MRI features may suggest a differential diagnosis of giant cell tumor of tendon sheath (GCTTS), periosteal ganglion, periosteal chondroma, and glomus tumor. The T1, T2 signal intensities do not show changes of hemosiderin typical of GCTTS.

Diagnosis confirmation

Clinically, SAF may simulate acquired digital fibrokeratoma, glomus tumor, giant cell tumor of the tendon sheath, ganglion cyst, osteochondroma, ingrown nail, foreign body reaction, verruca, pyogenic granuloma, keratoacanthoma, periungual fibroma (Koenan tumor), and lipoma. Review of the histopathologic changes will differentiate SAF from most of these entities.

The histopathologic diagnosis will vary, based on the degree of myxoid and collagenous components. Evaluation of the bulk of the tumor is usually required to render a diagnosis of SAF. Superficially sampled lesions may simulate irritated verruca vulgaris or subungual keratoacanthoma if there is insufficient dermis.

Dermatofibrosarcoma protuberans (DFSP), when superficially sampled, may simulate SAF, especially if strongly CD34(+). The dendritic cells of DFSP are also CD34 positive, so deeper sampling is required to assess for the fat-trapping that is typical of DFSP and absent in SAF. DFSP is also more uniformly storiform, is negative for EMA, and does not tend to affect the digits. Documentation of the t(17;22) translocation will also confirm DFSP.

Other neoplasms that may show histologic overlap with SAF include myxoma/fibromyxoma, benign peripheral nerve sheath tumor, nodular fasciitis, myxolipoma, myxoid malignant fibrous histiocytoma, myxoid neurofibroma, acral myxoinflammatory fibroblastic sarcoma, spindle cell lipoma, superficial angiomyxoma, and liposarcoma.

Additionally, perineurioma (in particular the sclerosing variant that typically involves the digit), fibrous histiocytoma, and cellular digital fibroma may enter the differential diagnosis SAF with a predominantly collagenous component. The immunophenotype and/or acral location helps differentiate SAF from most of these spindled cell tumors.

Superficial angiomyxoma is a spindle cell proliferation in a myxoid and vascular stroma, which is occasionally positive for CD34. It is differentiated from SAF by a multilobular architecture, inflammatory cells that include neutrophils, pools of mucin with peripheral acellular cleft-like spaces, a more infiltrative growth pattern, occasionally an epithelial component, and head and neck or trunk location. Only rarely is there involvement of digits.

Spindled cell and/or fibrous tumors that typically arise on the digits may pose a challenge. Subungual perineuriomas are CD34 (+), but the existence of this has been challenged as an entity distinct from SAF. Sclerosing perineurioma is an EMA (+) tumor typically arising on the hand or digit. This can be distinguished from SAF by negative CD34 staining and an arrangement of the epithelioid and dendritic cells in an onion-bulb whorled pattern.

Acral myxoinflammatory fibroblastic tumor also arises on the digits, hands, and feet, but has a subcutaneous location. Histologically, this is distinguished from SAF by a multinodular growth pattern with hyaline areas, a prominent inflammatory component, and cells with Reed-Sternberg-like morphology. These tumors may be CD34 (+), but are EMA negative.

A CD34 collagenous spindled cell tumor with a predilection for the digits, termed cellular digital fibroma, is very similar in appearance to the collagenous presentation of SAF. Due to the clinical and histologic resemblance to SAF, cellular digital fibroma is felt by some to possibly represent a variant of SAF. SAF was not referenced in this article, so it is not clear whether the investigators had the opportunity to compare and contrast these entities.

In contrast to SAF, their cases of cellular digital fibroma had a slight female to male ratio of 4:3, with smaller lesions (less than 0.5 cm), without reported recurrences. The tumors showed a similar growth pattern as SAF, but lacked stellate cells, a vascular proliferation, and a prominately storiform pattern. In the 5 tumors evaluated, EMA was negative.

Acquired digital fibrokeratomas may show clinical and histologic overlap with SAF. Fibrokeratomas are smaller, more superficial and exophytic, with less cellular collagen that is often oriented vertical to the epidermis. A significantly myxoid component is absent. One of the originally described subtypes of acquired digital fibrokeratoma (type 2) shows similar cellularity as that seen in SAF, suggesting that it may represent an evolving lesion of SAF.

Koenan tumors (periungual/subungual fibroma) are characterized by prominent fibrous connective tissue with mild cellularity and increased vessels with variable dilation. These tumors tend to be smaller, less cellular and lack the significantly myxoid component typical of SAF, but there is enough overlap that clinical history may be required to help differentiate these entities. Periungual/subungual fibromas are frequently multiple and associated with tuberous sclerosis.

Fibrous histiocytomas lack the significant myxoid component and are more storiform, with entrapment of collagen at the interface of tumor and normal dermis. These are usually CD34 negative.

In the rare case in which significant atypia and/or mitotic activity is present, SAF will have to be differentiated from a low-grade sarcoma. In general, if atypical mitoses, increased mitotic count, and/or necrosis are identified, one should consider the tumor to represent a more aggressive neoplasm, with potential for malignant behavior.

Low-grade fibromyxoid sarcoma (also known as hyalinizing spindle cell tumor with giant rosettes) may enter in the microscopic differential, due to its bland spindle cell morphology, fibromyxoid matrix, and absent to sparse mitoses. It arises in deep soft tissue, so should not be a consideration if anatomic site is taken into account. It is also differentiated by CD34 negativity and a whorled growth pattern. If needed, one could demonstrate the balanced translocation t(7;16), which results in an FUS/CREB3L2 fusion gene, specific for low-grade fibromyxoid sarcomas.

Synovial sarcoma with monophasic spindle cell morphology with minimal atypia and hypocellular myxoid matrix may present as small tumors on the hands and feet. These show histologic overlap with SAF and are EMA positive, but can usually be differentiated by demonstrating immunoreactivity for cytokeratin.

In general, if one encounters a spindled cell tumor with a prominent myxoid matrix, or a collagneous tumor that does not easily meet the clinical-pathologic criteria of acquired digital fibrokeratoma, fibrous histiocytoma, or periungual fibroma, immunohistochemical work-up to evaluate for SAF should ensue.

If there is histologic certainty, and only confirmation is desired, positive staining with CD34 should be sufficient, presuming that the tumor lacks mitotic activity or cellular atypia of malignancy. Incorporating and demonstrating EMA and CD99 staining, especially if there is focal or negative CD34 positivity, can further support a diagnosis of SAF.

Demonstration of S100 negativity may also be desired if CD34 is initially negative. One should avoid accepting a diagnosis of myxoid tumor or fibroma ’not otherwise specified’. Confirming a diagnosis of SAF will allow appropriate management.

Who is at Risk for Developing this Disease?

There are no epidemiologic risk factors for SAF. Men are affected more than women, with a 2:1 male to female ratio. In one large case series, only men were reported in the population ages 70 and older. The male predominance is most notable with the subgroup of SAF with a predominantly collagenous matrix. Younger age groups are more likely to have SAF with a predominantly myxoid matrix (mean age 39).

What is the Cause of the Disease?

The etiology and pathogenesis of SAF have not been established. It is felt to represent a CD34+ mesenchymal neoplasm, not a reactive proliferation. A CD34 positive fibroblastic component of the periadnexal and perivascular cellular population has been proposed to be an etiologic candidate.

CD10 has been reported in the normal mesenchymal cells in the nail unit, leading some to hypothesize that the neoplastic cells may be related to the mesenchymal cells in the nail unit, known as onychoblasts.

Nestin is expressed in multipotent precursors isolated from the dermis, which can differentiate into mesodermal progeny. Nestin immunohistochemical staining has been documented in SAF, suggesting a role for these multipotent stem cells.

Trauma with heavy objects has occasionally been reported, but a strong correlation with antecedent trauma has not been established.

Systemic Implications and Complications

Two cases in the original series had more significant atypia, but with a low mitotic rate, suggesting the possibility of progression to a fibromyxoid tumor of low-grade malignant potential. Other series have confirmed the occasional occurrence of more prominent cytologic atypia in SAF. There have been no documented cases of hematologic or lymphatic spread of SAF in either the atypical or non-atypical cases of SAF. There are no known associated syndromes or medical conditions.

Local bone deformities may occur, but significant permanent disability has not been reported. There may be significant nail deformities, which may or may not reverse, depending on the proximity to the nail matrix.

Treatment Options

–Surgical excision

–Partial amputation


Optimal Therapeutic Approach for this Disease

Complete surgical excision is the recommended method of therapy. Simple excision to ensure clear margins is usually curative, but there are cases in which tumor recurred despite reportedly clear margins.

None of the series described details of surgery, nor the excision margins. Given the anatomic limitations of the digits, a wide margin often required for sarcomas does not appear to be necessary.

A few patients reported in the original series subsequently underwent amputation or partial amputation, but this form of aggressive surgery is not recommended for routine cases of SAF.

Due to the frequent subungual occurrence, nail avulsion is usually required for the surgical procedure. If the lesion is initially sampled by a partial or tangential procedure, resulting in a non-specific or non-definitive diagnosis of SAF, complete excision to allow an accurate evaluation of the tumor should be considered. If a superficial biopsy of a clinically substantial nodule or tumor yields a diagnosis of wart or other epithelial proliferation, deeper sampling or excision should be performed, since the surface of SAF may simulate a verruca.

Patient Management

Patients should be aware that SAF is considered a benign neoplasm, but that there is histologic support for possible progression to a low-grade malignancy. They should know that no cases have metastasized.

Patients should be aware of the potential for recurrence (reported to be up to 42%). Involvement of the margins does not always predict the potential for recurrence.

Recurrences tend to occur within about 1-15 months after excision, but have been reported at 38 months, 5 years, and 10 years post surgery. Thus, follow up is recommended.

As a clinician, one should avoid the temptation to accept a diagnosis of ’myxoma not otherwise specified’, for tumors arising from the digits, especially if near the nailbed, and instead, pursue a diagnosis of SAF, in order to appropriately manage these patients.

Unusual Clinical Scenarios to Consider in Patient Management

In the original description of SAF, one of the cases recurred 10 years after initial excision, and was followed for 7 years prior to re-excision, without further recurrence or evidence of widespread tumor. This indolent behavior suggests that cautious observation may be an alternative in patients who refuse to or are unable to undergo surgical excision at the time of diagnosis, presuming the diagnosis of SAF is confirmed.

What is the Evidence?

Fetsch, JF, Laskin, WB, Miettinen, M. “Superficial acral fibromyxoma: a clinicopathologic and immunohistochemical analysis of 37 cases of a distinctive soft tissue tumor with a predilection for the fingers and toes”. Hum Pathol . vol. 32. 2001. pp. 704-714. (This is the seminal article that defined superficial acral fibromyxoma. The investigators detail the clinical, pathologic and immunhistochemical features of SAF, documenting the evidence that this represents a newly described entity, and establishing its neoplastic nature.
The authors also reported comparative data for acquired digital fibrokeratomas and periungual fibromas retrieved from their consultation files. They also analyzed images from the literature from these other collagenous neoplasms.)

Prescott, RJ, Husain, EA, Abdellaoui, A, Al-Mahmoud, RMW, Khan, M, Salman, WD, Twaij, Z. “Superficial acral fibromyxoma: a clinicopathological study of new 41 cases from the UK: should myxoma (NOS) and fibroma (NOS) continue as part of 21st-century reporting?”. Br J Dermatol. vol. 159. 2008. pp. 1315-1321. (This article confirms SAF as a new entity, adding 41 additional cases to the literature. A table of the clinical features and follow-up is included. The authors report the cases to be from UK hospitals, but images from at least one of the tumors appear to represent the same tumor from the Armed Forces Institute of Pathology (AFIP; reference Al-Daraji WI & Miettinen M below), suggesting potential overlap in the case reporting.
The authors report wart-like changes in 17 of their cases. They suggest avoiding the terms myxoma ’NOS’ and fibroma ’NOS’, but instead, attempt to render a diagnosis of SAF. A differential diagnosis of SAF is discussed.)

Varikatt, W, Soper, J, Simmons, G, Dave, C, Munk, J, Bonar, F. “Superficial acral fibromyxoma: a report of two cases with radiological findings”. Skeletal Radiol. vol. 37. 2008. pp. 499-503. (This article decribes the histologic, radiologic, and MRI features of SAF. There is a review of histologic and ultrastructural features, with a discussion of the differential diagnosis of SAF.)

Tardio, JC, Butron, M, Martin-Fragueiro, LM. “Superficial acral fibromyxoma: Report of 4 cases with CD10 expression and lipomatous component, two previously under recognized features”. Am J Dermatopathol. vol. 30. 2008. pp. 431-35. (This report describes four cases of SAF, three of which were positive for CD10, and one of which had mature fat cells intermingled with the spindled cells in the tumor, a previously undescribed finding. CD10 had been reported on one case prior, and this article supports this immunophenotype as a reproducible finding in SAF.
CD10 staining is not specific for SAF, as it has been reported in other spindled cell tumors, such as atypical fibroxanthoma and spindled squamous cell carcinoma.)

Al-Daraji, WI, Miettinen, M. “Superficial acral fibromyxoma a clinicopathological analsis of 32 tumors including 4 in the heel”. J Cutan Pathol. vol. 35. 2008. pp. 1020-26. (This article adds 32 additional cases from the AFIP of SAF to the literature, further supporting SAF as a new entity. Their cases include four involving the heel, which had not previously been described. Two of their cases extended into the underlying fascia and bone. They describe verrucous and papillomatous epidermal hyperplasia with viral-type changes simulating verruca vulgaris in the epidermis of 17 cases, although the epidermis in the photograph demonstrating these features appears atrophic with hyperkeratosis that is slightly verruciform.
Similar to the original description, recurrence was noted in their population (3 of 14 with follow-up). They describe one case in which significant cytologic atypia and a single mitotic figure was seen, treated with modest excision, with no recurrences or metastases at 4 years follow-up. Immunostaining showed typical features. Additionally, they stained their tumors for factor XIII, which they reported to stain “histiocytes in varying number”, and desmin showing “delicate positivity in one tumor” but no images to detail the staining pattern were included.)

Luzar, B, Calonje, E. ” Superficial acral fibromyxoma: Clinicopathological study of 14 cases with emphasis on a cellular variant”. Histopathology. vol. 54. 2009. pp. 374-393. (In a letter to the editor, the authors report 14 cases of SAF, 10 of which were initially incompletely excised. One was re-excised with amputation of the little toe. Seven remained incompletely excised, without recurrence on follow-up (mean 18.4 months). Half of their tumors demonstrated prominent cellularity, with intersecting fascicles or bands of CD34 (+) spindled cells with areas of storiform architecture, similar to that described in cellular digital fibroma.
They describe and provide an image of a finding not previously reported of atypical multinucleated giant cells. They offer their experience in a discussion of the differential diagnosis and confirm that synovial sarcoma may arise on the hands and feet as smaller neoplasms that are difficult to distinguish from SAF.)

Oteo-Alvaro, A, Meizoso, T, Scarpellini, A, Ballestin, C, Perez-Espejo, G. “Superficial acral fibromyxoma of the toe, with erosion of the distal phalanx. A clinical report”. Arch Orthop Trauma Surg. vol. 128. 2008. pp. 128-274. (This case report published in the orthopedic literature offers images of radiologic and MRI studies from a SAF that eroded the distal phalynx. The lesion was excised and the skin healed with secondary intention.)

McNiff, JM, Subtil, A, Cowper, SE, Lazova, R, Gusac, EJ. “Cellular digital fibromas: distinctive CD34-positive lesions that may mimic dermatofibrosarcoma protuberans”. J Cut Pathol. vol. 32. 2005. pp. 413-418. (This article describes a series of 14 CD34 positive fibrous spindle cell tumors affecting the digits, which the authors termed cellular digital fibroma, distinguishing it from dermatofibrosarcoma protuberans. There is enough clinical and pathologic resemblance to SAF, that some consider this to possibly represent a variant of SAF. They do not discuss SAF in their article, but provide details of their tumor that help support cellular digital fibroma as a potentially distinct entity.)

Guitart, J, Ramirez, J, Laskin, WB. “Cellular digital fibromas: what about superficial acral firomyxoma?”. J Cutan Pathol . vol. 33. 2006. pp. 762-3. (In a letter to the editor, the authors acknowledge the value of McNiff et al.’s article (see reference above) in bringing awareness to such lesions that may simulate DFSP, causing unnecessary surgery. They report a case of a CD34 (+) and EMA (-) SAF that had features that were felt to be “essentially identical” to cellular digital fibroma. Their tumor was markedly myxoid, which differs from the collagenous stroma described in cellular digital fibroma.)

Misago, N, Ohkawa, T, Yanai, T, Narisawa, Y. “Superficial acral fibroma on the tip of the big toe: expression of CD10 and nestin”. JEADV . vol. 22. 2008. pp. 235-62. (In this letter to the editor, the authors report a case of SAF with a CD34 (+), CD99 (+), CD10 (+), nestin (30% staining+), EMA (-) immunophenotype. The authors hypothesize a pathogenesis of SAF that involves the onychoblasts and multipotent dermal stem cells, based on this immunophenotype.)