Splenic hamartoma with bizarre stromal cells: a case report and literature review

Splenic hamartomas, originally described by Rokitansky in 1861, have been documented in fewer than 200 cases to date in the literature. Splenic hamartoma is a very rare benign vascular lesion and has also been called a splenoma or spleen within a spleen. They are usually asymptomatic and discovered incidentally, particularly in the adult [1]. Splenic hamartoma consists of disorganized sinusoid-like channels such as red pulp tissue, the lining cells of which can be highlighted by CD8 immunopositivity, while no white pulp elements are observed in the lesion [2‐4].

When some large and atypical stromal cells appear in splenic harmartoma, it can be challenging to determine whether the hamartoma represents a benign or malignant lesion. Cheuk et al. first reported splenic hamartoma with bizarre stromal cells in 2005 [5]. To date, only six such cases have been described in the literature. These stromal cells with morphological diversity cause difficulties in diagnosis and raise a mimicker of malignancy. Presented here is an asymptomatic patient with a round and solid lesion in the spleen. We present our observations about morphological features and the immunohistochemical data of bizarre stromal cells in this splenic hamartoma.

Although vascular neoplasm is the most frequent primary neoplasm of the spleen, splenic hamartoma is a rare benign “tumor” (about three cases in 200,000 splenectomies) occurring in any age group, with no gender predilection and usually without symptoms [1, 6]. Splenic hamartoma is generally found incidentally during imaging tests performed for other reasons or at autopsy. Clinical manifestations associated with larger masses are more often encountered among women, suggesting a hormonal influence. In addition, symptoms are reported most frequently in pediatric patients. Although the popularity of modern radiologic imaging techniques, such as ultrasound, color Doppler ultrasound, computed tomography (CT), and MRI, make it possible for early detection of the splenic hamartoma, definitive diagnosis depends on tissue examination.

Splenic hamartoma most commonly is a solitary or multiple, round, well-circumscribed, unencapsulated, and dark red nodule compressing the adjacent normal parenchyma. The lesions vary in size ranging from a few millimeters to 20 cm maximum. Histologically, splenic hamartoma consists of unorganized sinusoid-like channels without interspersed white pulp. The lining cells of sinusoid-like channels are CD8-positive, which is a key immunohistochemical feature for splenic hamartoma [7]. The cells are also positive for CD31, factor VIII–related antigen, and vimentin, while the expression of CD34 is inconsistent in different reports [3, 5, 8‐10]. In our case, CD34 shows a focal positive for lining cells. Additionally, T lymphocytes (CD3 +), B lymphocytes (CD20 +), macrophages (CD68 +), and fibroblasts in the loose stroma can be stained by related immunohistochemical markers [2].

The pathogenesis of splenic hamartoma is controversial. Some consider hamartoma as the congenital malformation of the red pulp, excessive and disorganized growth of abnormally formed red pulp, a neoplasm, or a reactive lesion to prior trauma [4, 8, 11]. Some report that splenic hamartoma is associated with other hamartomatous lesions such as tuberous sclerosis [12‐14].

Recently, several cases of splenic hamartoma with bizarre stromal cells have been reported as a difficult-to-diagnose variant. To the best of our knowledge, seven cases, including our patient, have been documented to date (Table 2) [5, 15‐17]. The patients were 5 women and 2 men, ranging in age from 35 to 64 years (mean, 50.4 years; median, 50 years). None of the patients had evidence of recurrent disease after splenectomy. Microscopically, the large bizarre cells were distributed randomly throughout the stroma of the lesion without association with vascular lumen. These cells with morphological diversity make it possible to misdiagnose this rare benign variant as a malignancy.

For this case, the combined morphologic and immunohistochemical profile supported a diagnosis of splenic hamartoma with bizarre stromal cells that was a benign lesion. Differential diagnosis should be considered, including a group of primary or secondary lesions of the spleen presenting a pattern of spindled cells mixed with bizarre large cells and different kinds of inflammatory cells in the loose stroma. In our case, the most likely differential diagnostic considerations include IMT, follicular dendritic cell (FDC) sarcoma, angiosarcoma, and Hodgkin’s lymphoma.

Inflammatory myofibroblastic tumor is an uncommon neoplasm originally denoted as inflammatory pseudotumor [18]. IMT is a distinctive lesion composed of myofibroblastic spindle cells mixed with an inflammatory infiltration of lymphocytes, plasma cells, and eosinophils. Three basic histological patterns form in the tumor: an edematous myxoid vascular pattern resembling nodular fasciitis; a compact fascicular spindle cell pattern with variable myxoid and collagenized regions; and hypocellular plate-like collagenized pattern resembling a scar or desmoid-type fibromatosis. Immunocytochemistry shows positive staining for vimentin, SMA, desmin, ALK (50% +), and negative staining for S-100 and CD30 [19, 20]. Unlike the splenic hamartoma, myofibroblastic spindle cells of IMT are negative for CD8 and CD31.

FDC sarcoma is a neoplastic proliferation of spindled to ovoid cells dispersed within a prominent lymphoplasmacytic infiltration [21]. Typically, spindled cells show indistinct cell borders and vesicular nuclei. Some are bland-looking while others are enlarged or overtly atypical. FDC sarcoma is positive for one or more of the follicular dendritic markers, such as CD21, CD23, and CD35, and the EBER is tested by in situ hybridization [22]. However, bizarre stromal cells of splenic hamartoma are negative for all the follicular dendritic markers.

Primary splenic angiosarcoma is an extremely rare nonlymphoid malignant neoplasm that originates from the splenic sinusoidal vascular endothelium [23]. The tumor consists of irregular and anastomosing vascular channels lined by atypical endothelial cells with high nuclear grade displaying mitotic activity. Although CD8 staining of the lining cells in splenic angiosarcoma has been reported, the main endothelial cell markers, including CD31, CD34, factor VIII–related antigen, and the histiocytic marker CD68, should exhibit strong positivity [10]. Angiosarcoma has a high rate of metastasis and poor prognosis [24].

Sometimes large stromal cells with double nuclei and apparent eosinophilic nucleoli in splenic hamartoma mimic Reed-Sternberg (R-S) cells in classical Hodgkin’s lymphoma (CHL). CHL is a monoclonal lymphoid neoplasm composed of mononuclear Hodgkin cells and multinucleated R-S cells residing in an abundant admixture of infiltrative non-neoplastic inflammatory cells, histocytes, and fibroblasts [25]. R-S cells are positive for CD30 in nearly all cases and for CD15 in the majority (75-85%) of cases; they are usually negative for CD45 and CD68.

Similarly, the histogenesis of bizarre stromal cells in splenic hamartoma is still elusive, since they demonstrate no specific differentiation to epithelial, endothelial, lymphoid, histiocytic, myeloid, or melanocytic cells. For the reported six cases, the numerous markers tested are negative except a focal and faint positive for desmin in three cases, very focal and equivocal staining for SMA in one case, and a positive for keratin (CAM5.2) and CD30 in another case. According to the immunohistochemical expression, Cheuk et al. and other literature consider that these cells, which may be related to the stromal myoid cells or so-called fibroblastic reticulum cell that are usually present in the red pulp, the periarterial lymphoid sheath, and marginal zone of the spleen, represent a degenerative change accompanied by partial or complete loss of the myoid immunophenotype or undergo immunophenotypic modulation in response to physiological or pathological stimuli [5, 15]. Although bizarre stromal cells in our case failed to react with the above markers, we suspect that they are still a degenerative change. No matter what the immunohistochemical markers express, splenic hamartoma with bizarre stromal cells has benign clinical behavior because the bizarre cells do not form expansile clusters, they lack mitotic activity, and the Ki-67 index is very low.

Splenic hamartoma is a rare benign vascular proliferative lesion characterized by the CD8-positive immunophenotype of the lining endothelial cells. Bizarre stromal cells without immunohistochemical differentiation can be present in some splenic hamartomas. Awareness of this histologic variant can prevent misdiagnosis of this rare benign entity as a malignant tumor.