Metastases to soft tissue

A review of 118 cases over a 30-year period


  • Jose Antonio Plaza MD,

    1. Division of Anatomic Pathology, Department of Pathology, Ohio State University Medical Center and Arthur G. James Comprehensive Cancer Center, Columbus, Ohio
    Search for more papers by this author
  • Delia Perez-Montiel MD,

    1. Division of Anatomic Pathology, Department of Pathology, Ohio State University Medical Center and Arthur G. James Comprehensive Cancer Center, Columbus, Ohio
    2. Department of Pathology, Instituto Nacional de Cancerologia, Mexico, D.F, Mexico
    Search for more papers by this author
  • Joel Mayerson MD,

    1. Department of Orthopedic Oncology, Ohio State University, Columbus, Ohio
    Search for more papers by this author
  • Carl Morrison MD,

    1. Department of Pathology, Roswell Park Cancer Center, Buffalo, New York
    Search for more papers by this author
  • Saul Suster MD

    Corresponding author
    1. Division of Anatomic Pathology, Department of Pathology, Ohio State University Medical Center and Arthur G. James Comprehensive Cancer Center, Columbus, Ohio
    • Department of Pathology, The Medical College of Wisconsin, 9200 West Wisconsin Ave., Milwaukee, WI 53226
    Search for more papers by this author
    • Fax: (414) 805-6980.



Metastatic tumors presenting as soft tissue masses are relatively rare and can be the source of diagnostic confusion both clinically and pathologically. The authors' experience was reviewed at a large academic medical center over a 30-year period (1971–2000) with metastases to soft tissue.


The tumors in the study included mainly lesions involving skeletal muscle or skeletal muscle and subcutaneous tissue of the upper and lower limbs, trunk, shoulders, and buttocks. Direct extension from tumors originating in bone or adjacent organs, tumors involving the skin or areas known to contain abundant lymph nodes (ie, axilla, groin), and hematopoietic malignancies were excluded.


One hundred and eighteen cases were identified; 60 patients were women and 58 were men. The age range was 20 to 87 years (median of 53.5 years). The primary tumor was located in the skin (19 patients), lung (13 patients), breast (13 patients), kidney (12 patients), colon and rectum (12 patients), uterus (8 patients), ovary (5 patients), head and neck (tongue, pharynx, larynx, nasal cavity, and mandible) (5 patients), esophagus (2 patients), stomach (2 patients), cervix (2 patients), small bowel (2 patients), bone (2 patients), adrenal gland (1 patient), eye (1 patient), testis (1 patient), urinary bladder (1 patient), and salivary gland (1 patient). In 27% (32 of 118 cases) of cases, the soft tissue metastasis was the initial manifestation of the disease. In 13.5% (16 of 118 cases) of cases the primary site of origin could not be identified. The sites of metastasis included the abdominal wall (25 patients), back, including scapular region (20 patients), thigh (17 patients), chest wall (15 patients), arm (15 patients), shoulder (11 patients), buttock (5 patients), perineum (3 patients), leg (2 patients), foot (1 patient), umbilical area (1 patient), ankle (1 patient), scalp (1 patient), and elbow (1 patient). The histologic classification of the tumors included carcinoma (83 patients), malignant melanoma (20 patients), sarcoma and carcinosarcoma (9 patients), malignant mixed Mullerian tumor (2 patients), seminoma (1 patient), malignant teratoma (1 patient), malignant gastrointestinal stromal tumor (1 patient), and neuroblastoma (1 patient). Many of the tumors displayed histologic features that created difficulties for diagnosis and could be easily mistaken on routine histopathologic examination for a variety of primary soft-tissue sarcomas. Routine use of immunohistochemical stains aided in their proper recognition.


Metastases are not an infrequent finding in soft tissue and they may represent the initial manifestation of the disease. Use of a basic panel of immunohistochemical stains is recommended for defining the cell type and arriving at the correct diagnosis. Cancer 2008. © 2007 American Cancer Society.

Metastases to soft tissue are rare and can be easily misdiagnosed histologically for a primary soft-tissue sarcoma. The clinical distinction between a metastatic neoplasm to soft tissue and a primary soft-tissue sarcoma is critical because treatment and prognosis are markedly different. Metastatic neoplasms to soft tissue can also present as the initial manifestation of an occult primary malignancy, and their histologic recognition is of extreme importance. To our knowledge, only limited information is available in the literature on this phenomenon and has been mostly reported as single cases, with only a few small series available on this topic.1–9 In particular, studies utilizing modern immunohistochemical markers addressing these tumors are lacking or very sparse.

In the current study, we review our experience over a 30-year period with metastases presenting as soft tissue masses. A recommended approach using a basic panel of immunohistochemical markers for the proper identification of the tumors is presented. To the best of our knowledge, this represents the largest series of metastases to soft tissue reported in the literature.


All cases accessed under the diagnosis of metastases to soft tissue were collected from the surgical pathology files of the Ohio State University Medical Center between 1971 and 2000. Overall, 185 cases were identified from a total of 7237 soft tissue tumors accessed during the same period. Criteria for selection included location of the tumor within skeletal muscle and/or subcutaneous tissue. Metastases to lymph nodes or areas with important lymphatic drainage, such as the groin and the axilla, or tumors showing histologic evidence of lymph node remnants were excluded from the study. Direct extension from underlying or adjacent tumors, postsurgery implants, metastasis involving the epidermis, and lymphomas, leukemias, and myelomas were also excluded. After excluding cases with incomplete clinical history or inadequate pathologic material, 118 patients with metastatic tumors to soft tissue were identified and included in the study (1.6% of all soft tissue sarcomas accessed during the same period). Clinical data and follow-up information was obtained from the hospital records.

From 1 to 8 histologic sections stained with hematoxylin and eosin (H & E) were analyzed in each case. All tissues were fixed in neutral buffered formalin and embedded in paraffin for histologic processing. Immunohistochemical studies were performed in selected cases from formalin-fixed, paraffin- embedded tissue sections. Immunohistochemical staining was performed in a Dako (Carpinteria, Calif) autostainer using a standard avidin-biotin peroxidase complex technique. Heat-induced epitope retrieval was applied as pretreatment to different markers. The antibodies included: cytokeratin AE1/AE3 (1:100 dilution; Dako), carcinoembryonic antigen (CEA) (1:100 dilution; BioGenex, San Ramon, Calif), epithelial membrane antigen (EMA) (1:200 dilution; Dako), MOC31 (1:40 dilution; Dako), S-100 protein (1:4000; Dako), HMB-45, Melan-A, smooth muscle actin (SMA) (1:400; Dako), CD31, CD34, bcl-2, CD117 (C-kit), β-human chorionic gonadotropin, estrogen receptor (ER), progesterone receptor (PR), chromogranin A (1:15 dilution; Dako), and synaptophysin (1:100 dilution; Dako). The chromogen, diaminobenzidine, was utilized for antigen localization. Appropriate positive and negative controls were run concurrently for all the markers tested.


Clinical Findings

The cases were classified by anatomic site, primary origin, age, and sex (Table 1). In all, 60 patients were women and 58 were men. The age range was 20 to 87 years old (median of 53 years). The primary tumor was in the skin (19 cases), lung (13 cases), kidney (12 cases), colon and rectum (12 cases), breast (13 cases), uterus (8 cases), ovary (5 cases), head and neck (tongue, pharynx, larynx, nasal cavity, mandible) (5 cases), esophagus (2 cases), stomach (2 cases), cervix (2 cases), small bowel (2 cases), bone (2 cases), adrenal gland (1 case), eye (1 case), testis (1 case), urinary bladder (1 case), and salivary gland (1 case), and 16 cases were of unknown primary origin. The most frequent site of metastasis included the abdominal wall (25 cases), followed by the back, (including scapular region) (20 cases), thigh (17 cases), chest wall (15 cases), arm (15 cases), shoulder (11 cases), buttock (5 cases), perineum (3 cases), leg (2 cases), foot (1 case), umbilical area (1 case), ankle (1 case), scalp (1 case), and elbow (1 case). The tumors presented as the initial symptom from an occult tumor in 27.2% of cases (32 of 118 cases), as a solitary late metastasis in 70.3% (83 of 118), and as disseminated metastases in 2.5% of cases (3 of 118). In 13.5% of cases (16 of 118 cases), the primary site of origin could not be identified. The initial clinical impression was incorrect in some cases, especially in those in which the patients had no history of malignancy or had a remote history of malignancy; in such instances the tumors were clinically initially regarded as a sarcoma.

Table 1. Classification of Cases
CaseAge, ySexPrimary tumorSoft tissue site of metastasisTumor sizeMetastasis as the initial presentationTime between primary tumor diagnosis and metastasisHistologic type
  1. NA indicates not applicable; GIST, gastrointestinal stromal tumor.

175ManLungScapula2 × 2 × 1.5 cmYesAdenocarcinoma
270ManLungChest wall1.8 × 1.5 × 1.0 cmNo17 moSquamous cell carcinoma
367WomanUnknownArm15 × 8 × 4 cmYesMelanoma
476WomanColonArm3.3 × 3.0 × 2.0 cmNo4 yAdenocarcinoma
563ManKidneyPerineal2.8 × 2.5 × 2.0 cmYesNARenal cell carcinoma, clear cell
662ManLungBack3.4 × 3.0 × 1.7 cmNoNAAdenocarcinoma
755ManSkinShoulder5 × 5 × 4 cmNoNAMelanoma
847WomanSkinLeg8 × 3 × 2 cmNoNAMelanoma
982WomanSkinFoot dorsum6.5 × 6 × 4 cmNo23 moMelanoma
1051WomanOvaryPerineal4 × 4 × 3 cmNo9 moAdenocarcinoma
1142WomanSalivary glandChest wall1.8 × 1.5 × 1.0 cmNo11 moAdenocarcinoma
1254WomanSkinThigh10 × 5 × 5 cmNo2.5 yMelanoma
1351ManSkinArm12 × 10 × 7 cmNo5 yMelanoma
1463ManUnknownShoulder1.2 × 0.5 × 0.3 cmYesSquamous cell carcinoma
1553WomanUnknownUmbilical area2 × 2 × 1.5 cmYesAdenocarcinoma
1657WomanKidneyShoulder8 × 8 × 7.5 cmNo3 yRenal cell carcinoma, clear cell
1774ManColonAbdominal wall7 × 6 × 6 cmNo5 yAdenocarcinoma
1828WomanNasal cavityAbdominal wall3.5 × 3.0 × 2.7 cmNo20 moNeuroblastoma
1930WomanUnknownThigh3 × 2 × 2 cmYesSquamous cell carcinoma
2067WomanOvaryThigh6 × 6 × 5 cmNoNAMalignant teratoma
2164WomanOvaryChest wall3.5 × 3.0 × 3.0 cmNoNAAdenocarcinoma
2261ManBreastShoulder2 × 2 × 2 cmYesAdenocarcinoma
2356ManLungShoulder1.7 × 1.6 × 1.5 cmYesAdenocarcinoma
2432ManKidneyElbow2.5 × 2.0 × 2.0 cmYesRenal cell carcinoma, clear cell
2558ManUnknownLeg3 × 2 × 2 cmYesAdenocarcinoma
2673WomanBreastScapula2.1 × 2.0 × 1.0 cmNo21 moDuctal carcinoma
2747ManUnknownAnkle2.7 × 2.0 × 2.0 cmYesMelanoma
2865ManLungBack1.7 × 1.0 × 1.0 cmNo4 ySquamous cell carcinoma
2958ManKidneyThigh4 × 4 ×3.5 cmNo4 yRenal cell carcinoma, clear cell
3064ManUnknownAbdominal wall2.3 × 2.0 × 2.0 cmYesAdenocarcinoma
3171WomanSkinThigh6 × 6 × 6 cmNoNAMelanoma
3248WomanUnknownBack2 × 2 × 1.5 cmYesAdenocarcinoma
3377ManKidneyShoulder2.3 × 2.0 × 1.7 cmNANARenal cell carcinoma, clear cell
3458WomanCervixButtock3.8 × 3.5 × 3.0 cmNoNASquamous cell carcinoma
3565WomanKidneyThigh4 × 4 × 4 cmNANARenal cell carcinoma, clear cell
3658WomanBreastShoulder2 × 2 × 1.5 cmNo2 yAdenocarcinoma
3767ManSkinArm2.3 × 2.0 × 2.0 cmNANAMelanoma
3823WomanSkinScalp1.9 × 1.7 × 1.5 cmNoNAMelanoma
3967WomanBreastArm1.6 × 1.5 × 1.5 cmNANAAdenocarcinoma
4030WomanSkinChest wall0.7 × 0.7 × 0.7 cmNo2 yMelanoma
4140WomanBreastBack2 × 1.5 × 1.0 cmYesMicropapillary adenocarcinoma
4246WomanCervixThigh2.5 × 2.0 × 2.0 cmNo3.3 ySquamous cell carcinoma
4356ManLungArm1.6 × 1.5 × 1.5 cmNANASmall cell carcinoma
4437WomanUterusAbdominal wall1 to 3 cm (multiple nodules)No20 moLeiomyosarcoma
4555WomanLungChest wall2.5 × 2.5 × 2.0 cmYesAnaplastic adenocarcinoma
4654ManKidneyArm1.4 × 1.0 × 1.0 cmNANARenal cell carcinoma, clear cell
4777WomanBreastShoulder2 × 1.3 × 1.0 cmNANASmall cell carcinoma
4848WomanBreastArm1.0 × 1.0 × 1.0 cmNANADuctal carcinoma
4936ManUterusAbdominal wall1.5 × 1.0 × 1.0 cmNo2 yLeiomyosarcoma
5070ManUnknownScapula1 × 1 × 1 cmYesAdenocarcinoma
5161WomanUnknownThigh2.6 × 2.0 × 2.5 cmYesAdenocarcinoma
5348WomanBladderArm1.9 × 1.5 × 1.0 cmNANAUrothelial carcinoma
5438WomanSkinButtock4 × 3.8 × 3.5 cmNANAMelanoma
5559ManKidneyShoulder1.6 × 1.5 × 1.5 cmNANARenal cell carcinoma, clear cell
5670WomanBreastAbdominal wall2.7 × 2.5 × 2.5 cmNANADuctal carcinoma
5761WomanOvaryAbdominal wall2.5 × 2 × 2 cmNo1 yCarcinosarcoma
5857ManKidneyThigh3.5 × 3.0 × 3.0 cmNo19 moRenal cell carcinoma, clear cell
5959ManSkinBack1.7 cmNo5.2 yMelanoma
6069WomanUterusAbdominal wall4 × 3 × 3 cmNo17 moLeiomyosarcoma
6161WomanLungBack3 × 2 × 2 cmYesSmall cell carcinoma
6251ManSkinChest wallNANo4.4 yMelanoma
6370ManLungButtock5 × 4.3 × 3.0 cmYesAdenocarcinoma
6433WomanAdrenal glandBack3.5 × 2 × 2 cmNo3 yAdrenal cortical carcinoma
6553ManColonAbdominal wallNANo4 yAdenocarcinoma
6672ManColonAbdominal wall2 × 2 × 2 cmNo1 yAdenocarcinoma
6763WomanColonAbdominal wall3.2 × 3.0 × 2.5 cmNo15 moAdenocarcinoma
6866ManLungBack1.9 × 1.5 × 1.5 cmNo11 moSmall cell carcinoma
6950ManPharyngealButtock2 × 1 × 1 cmNo1 ySquamous cell carcinoma
7077WomanBreastForearm2 × 1.5 × 1.5 cmYesDuctal carcinoma
7140ManSmall bowelAbdominal wallNANo2.5 yAdenocarcinoma
7287ManSkinChest wallNANoNASquamous cell carcinoma
7368WomanLungBack2 × 1 × 1 cmYesAdenocarcinoma
7461WomanUnknownChest wall3.5 × 2.6 × 1.5 cmYesAdenocarcinoma
7551WomanColonAbdominal wall2.2 × 2.0 × 1.5 cmNo19 moAdenocarcinoma
7637WomanBreastChest wall1.8 × 1.5 × 1.0 cmNo5 yDuctal carcinoma
7755WomanStomachAbdominal wall1.6 × 1.1 × 1.0 cmNo3 yAdenocarcinoma
7862ManSkinShoulder5.0 × 2.5 × 2.0 cmNo2 yMelanoma
7953WomanColonAbdominal wall1.6 × 1.5 × 1.5 cmNo3 yAdenocarcinoma
8040ManUnknownChest wall4 × 3 × 3 cmYesSquamous cell carcinoma
8166ManUnknownThigh1.5 × 1.5 × 1.0 cmYesLarge cell carcinoma with squamous differentiation
8220ManBoneChest wall1.8 × 1.8 × 1.5 cmNo6 yOsteosarcoma
8361WomanUterusBack5 × 3 × 3 cmNo9 yLeiomyosarcoma
8456ManLungThigh4.0 × 3.5 × 3.0 cmNo7 moMucinous adenocarcinoma
8562WomanOvaryAbdominal wall2 × 2 × 2 cmNoSame timeSerous carcinoma
8655WomanBreastBack1.5 × 1.5 × 1.0 cmNo4 yDuctal carcinoma
8757ManKidneyChest wall1.0 × 1.0 × 0.8 cmNoRenal cell carcinoma, clear cell
8831WomanBreastBack2 × 2 × 1.5 cmNo3 yDuctal carcinoma
8956ManLungThigh2.2 × 2.0 × 2.0 cmNo3 yAdenocarcinoma
9057ManColonThigh1.7 × 1.5 × 1.5 cmYesAdenocarcinoma
9158WomanColonShoulder0.7 × 0.5 × 0.5 cmYesAdenocarcinoma
9233WomanSkinArm4 × 2.7 × 2.0 cmNo16 moMelanoma
9348WomanUterusAbdominal wall2.0 × 2.0 × 1.5 cmNo9 moLeiomyosarcoma
9442WomanUterusAbdominal wall1.7 × 1 × 1 cmNo4 yLeiomyosarcoma
9573WomanEsophagusAbdominal wall1.5 × 1.0 × 1.0 cmNo7 moAdenocarcinoma
9676ManTongueBack0.9 × 0.7 × 0.5 cmNo2 ySquamous cell carcinoma
9768ManEsophagusAbdominal wall2.5 × 2.0 × 1.0 cmNo16 moAdenocarcinoma, signet ring cell type
9842ManUnknownButtock0.6 × 0.5 × 0.5 cmYesAdenocarcinoma
9952ManSkinAbdominal wall3.0 × 2.5 × 2.5 cmNo1 yMelanoma
10039ManRectumPerineal3 × 3 × 2 cmNo2 yAdenocarcinoma, signet ring cell
10167ManLungScapula3.5 × 2.5 × 2.0 cmYesLarge cell carcinoma
10267ManUnknownBack2 × 2 × 1.5 cmYesAdenocarcinoma
10363ManColonChest wall3 × 3 × 2.5 cmNo4 yAdenocarcinoma
10451WomanUterusAbdominal wall (skin and soft tissue16 × 15 × 15 cmNo2 yMixed Mullerian Tumor
10566ManLarynxChest wall5 × 5 × 3 cmNo4 ySarcomatoid carcinoma
10659WomanBreastChest wall2.3 × 2.3 × 2.0 cmNo8 yDuctal carcinoma
10755WomanColonAbdominal wall7 × 5 × 5 cmNo1 yAdenocarcinoma
10819ManSkinBack1.5 × 1.5 × 1.0 cmNoNAMelanoma
10930WomanBone (femur)Thigh2.7 × 2.0 × 2.0 cmNo4 yOsteosarcoma
11052ManSmall bowelAbdominal wall4.5 × 2.8 × 2.5 cmNo3 yMalignant GIST
11168WomanUterusAbdominal wall6 × 5 × 4 cmNo5 moMixed Mullerian tumor
11266ManKidneyScapula4 × 4 × 3.5 cmYesRenal cell carcinoma, clear cell type
11346WomanMandibleArm3.0 × 2.9 × 2.5 cmNo2 ySquamous cell carcinoma
11443ManSkinThigh5 × 3 × cmYesMerkel cell carcinoma
11547ManEyeArm1.5 × 1.2 × 1.0 cmNo4 yMelanoma
11649ManSkinThighNANo1 yMelanoma, amelanotic type with rhabdoid features
11784WomanKidneyArm3 × 2.4 × 2.0 cmNo3 yRenal cell carcinoma, clear cell

Histopathologic and Immunohistochemical Findings

Metastatic carcinoma

Overall, 83 cases corresponded to metastases of carcinoma. The sites of origin are presented in Table 1. The majority of the tumors in this group originated from the lung; there were 8 adenocarcinomas (6 moderate to poorly differentiated, 1 case had mucinous features, and 1 cases had anaplastic features) (Fig. 1A). Two cases corresponded to metastases of squamous cell carcinoma (Fig. 1B), 3 corresponded to metastases of small cell carcinoma, and 1 corresponded to a metastasis of large cell carcinoma. In this group, 7 cases presented with an isolated soft tissue mass without any prior history of lung cancer. Two cases of lung metastatic carcinoma demonstrated a sarcomatoid appearance that was initially misinterpreted as a primary soft tissue sarcoma (Cases 45 and 101) (Fig. 1C and 1D). These patients were suspected of having a primary soft tissue sarcoma, both clinically and histologically. Use of a routine panel of antibodies in samples taken from the resected specimen demonstrated that the tumor cells were strongly positive for epithelial markers (cytokeratin AE1/AE3, EMA, and MOC31), prompting careful scrutiny by imaging studies that led to the discovery of the primary tumor in the lung.

Figure 1.

(A) Metastasis of anaplastic carcinoma of the lung to soft tissue showing bizarre tumor cells. (B) Metastasis of squamous cell carcinoma of bronchus to soft tissue. (C) Metastasis to soft tissue of sarcomatoid carcinoma of the lung showing an atypical spindle cell proliferation simulating fibrosarcoma. (D) Cytokeratin stain from the same case demonstrating strong cytoplasmic positivity of the spindle tumor cells.

Thirteen cases corresponded to metastases from breast cancer (Fig. 2A), and 3 presented with no prior history of breast cancer and were misinterpreted as a primary soft tissue tumor (10 cases were moderate to poorly differentiated invasive ductal carcinoma, 1 was a well-differentiated invasive ductal carcinoma, 1 was a micropapillary carcinoma, and 1 was a small cell carcinoma). Twelve cases corresponded to metastases of clear cell renal cell carcinoma from the kidney (Fig. 2B); 3 of these presented with no prior history of kidney cancer and were initially misinterpreted as a primary soft tissue tumor. Twelve cases were metastases from adenocarcinoma of the colon and rectum; 2 presented with no prior history of colorectal cancer and were initially clinically regarded as a primary soft tissue tumor (11 cases were moderate to poorly differentiated carcinoma and 1 case had signet ring cell features). Five cases corresponded to ovarian carcinoma, all of which had a prior history of ovarian tumor. Five cases corresponded to metastases from head and neck tumors; all patients had a known prior history of head and neck cancer. Two cases corresponded to esophageal carcinomas; both patients had a prior history of esophageal cancer. Two cases corresponded to carcinomas of the uterine cervix, and both patients had a prior history of cervical cancer. Two cases corresponded to small bowel tumors and both patients had prior history of carcinoma (1 case was a moderately differentiated adenocarcinoma, and the other a malignant gastrointestinal stromal tumor [GIST]). The remainder of cases included: 1 gastric carcinoma (moderately differentiated adenocarcinoma); 1 adrenocortical carcinoma (moderate to poorly differentiated); 1 bladder cancer (poorly differentiated urothelial carcinoma); 1 skin cancer (moderate to poorly differentiated squamous cell carcinoma); and 1 salivary gland carcinoma (poorly differentiated carcinoma). Case No. 21 (ovarian tumor) was clinically and histologically suspected to be a primary sarcoma given the atypia, pleomorphism, and focal spindle cell morphology (Fig. 2C); immunohistochemistry demonstrated cytokeratin positivity in the tumor cells. Case No. 96 (tongue cancer) and Case No. 97 (esophageal tumor) were suspected clinically to be a primary sarcoma and histologic confirmation prevented aggressive treatment of these lesions. Case No. 105 (larynx tumor) was suspected clinically and histologically to be a sarcoma owing to the spindle cell morphology and the increased number of mitotic figures; the clinical history was reviewed and revealed a previous diagnosis of sarcomatoid carcinoma of the larynx, prompting the pathologist to order cytokeratin stains to prove that this was a metastatic tumor from the larynx. Case No. 110 (small bowel tumor) was also confused with a sarcoma histologically, given the spindle cell morphology and the increased mitotic count; a previous clinical history of a malignant GIST led to an immunohistochemical study for C-kit (CD117), which was strongly positive in the tumor cells.

Figure 2.

(A) Metastasis of infiltrating ductal carcinoma of breast to soft tissue. (B) Metastasis of clear cell renal cell carcinoma to soft tissue. (C) Metastasis of carcinoma of the ovary with a sarcomatous spindle cell component. (D) Metastasis of small cell carcinoma to soft tissue.

Of the 16 unknown primary cases, 10 were histologically compatible with poorly differentiated adenocarcinomas, 3 cases were moderate to poorly differentiated squamous cell carcinoma, and 1 was a large cell carcinoma with squamous differentiation. Two cases corresponded to metastatic malignant melanoma. Follow-up information was not available for these 16 patients with regard to the primary source of the tumor.

Four cases corresponded to metastases of small-cell neuroendocrine carcinoma originating from the lung (3 cases) or the breast (1 case) (Fig. 2D).

Metastatic malignant melanoma

Twenty cases in our study corresponded to metastases from malignant melanoma. Histologically, 8 cases were composed of epithelioid cells; 3 were composed purely of spindle cells, 7 demonstrated a mixture of epithelioid and spindle cell components, 1 case showed rhabdoid features, and another case showed prominent balloon cell features (Fig. 3A and 3B). Eighteen patients had a known previous history of cutaneous melanoma; in 2 cases the patients had no previous history of melanoma. One patient had a history of ocular melanoma (Case 115). The amelanotic tumors with a predominantly spindle cell appearance or mixed epithelioid/spindle morphology were the most likely to be confused for a primary soft tissue sarcoma. Immunohistochemical stains in all cases showed strong positivity in the tumor cells for S-100 protein, HMB-45, and Melan-A in various proportions.

Figure 3.

(A) Metastasis to soft tissue of balloon cell melanoma simulating clear cell renal cell carcinoma. (B) Higher magnification from same case demonstrating the characteristic ‘nesting’ pattern of atypical tumor cells. (C) Metastasis of uterine leiomyosarcoma to soft tissue. (D) Metastasis of osteosarcoma to soft tissue.

Metastatic sarcomas

Eleven cases corresponded to metastases of sarcomatoid elements of primary soft-tissue sarcomas from other sites. Six cases were metastases from a uterine leiomyosarcoma. All had a prior history of uterine leiomyosarcoma. Two cases corresponded to metastases from bone osteosarcoma (Fig. 4B). Two cases corresponded to metastases of a uterine malignant mixed Mullerian tumor (MMMT); histologically, these tumors were composed of high-grade spindle cells and they all had a well-documented history of MMMT. One case corresponded to a metastasis from an ovarian carcinosarcoma; histologically the tumor was a high-grade spindle cell sarcoma similar to the primary tumor in the ovary.


Despite the finding that soft tissue comprises approximately 55% of our body mass, hematogenous metastases to these areas are rare. Direct extension of a primary tumor to soft tissue is a much more common event than distant soft tissue metastases. Several factors have been implicated in the rarity of this phenomenon such as changes in pH, accumulation of metabolites, and the local temperature at soft tissue sites.5 In addition, organs with a high predisposition for metastatic carcinomas, such as the liver or lung, are rich in capillary vessels and have a constant blood flow, whereas in soft tissues such as skeletal muscle, the blood flow is variable and is influenced by adrenergic receptors and subject to variations in tissue pressure affecting tumor implantation.5, 10 Studies have been conducted regarding whether traumatic injury to soft tissue can play a role in attracting cancer metastases.11 To the best of our knowledge, none of the cases in the current study were associated with traumatic injury at the site of the metastases. Metastases to soft tissue are usually noted in patients with advanced disease, but occasionally they can present in patients with no known primary tumor and masquerade as a soft tissue sarcoma.4 Multiple case reports and small series of patients have been reported in the literature addressing this rare phenomenon.2, 3, 8, 9, 12 Autopsy studies have also addressed the incidence of secondary malignancies involving soft tissue sites; 1 study concluded that secondary tumors were actually much more common than primary malignant tumors of soft tissue in a large autopsy series.13

We have reviewed our experience with metastases to soft tissue at a large medical center over a 30-year period (1971–2000). Rigorous criteria for the selection of patients were used to avoid cases that represented metastases to lymph nodes, particularly at sites rich in lymphatic drainage such as the groin and the axilla. The tumors in the study included mainly lesions involving skeletal muscle or skeletal muscle and subcutaneous soft tissue that could be potentially mistaken for primary soft tissue neoplasms. Overall, 118 cases were identified; the majority corresponded to metastases of carcinomas from internal organs, followed by metastases of malignant melanoma, sarcoma, and others. The most commonly affected sites included the abdominal wall, back, thigh, chest wall, and shoulder. In 27.11% of cases the metastasis to soft tissue was the initial manifestation of the disease leading to the subsequent identification of the primary tumor elsewhere. In 13.5% of cases a primary site of origin could not be identified.

The most commonly reported malignancies in the literature to result in distant soft tissue metastases are lung, kidney, and colon carcinoma.8, 14–16 Our findings paralleled those previously reported in the literature. The most common histologic diagnosis reported in the literature on distant soft tissue metastases is adenocarcinoma, followed by renal cell carcinoma of clear cell type, squamous cell carcinoma and melanoma, but many other histologic types have also been reported.14, 17, 18 The most common clinical setting in which distant soft tissue metastases have been observed is as a symptom of a previously diagnosed malignancy,2 but based on review of the literature, many metastatic tumors to soft tissue do present as occult metastases from an unrecognized primary, in particular lung.

Unlike lung cancer, renal cell carcinomas usually metastasize to soft tissue in patients with a previously known primary kidney tumor, and present clinically with a soft tissue mass from a few months up to 16 years after the initial diagnosis.2, 19 In our series, only 3 of 12 cases of metastatic renal cell carcinoma presented as an isolated soft tissue metastasis from a clinically unrecognized occult primary kidney neoplasm.

Melanoma can metastasize to any organ, and the first sign of spread is usually to the skin, subcutaneous fat, and lymph nodes followed by metastasis to lung, liver, brain, bone, and small bowel.5, 12 Melanoma specifically metastasizing to skeletal muscle is a rare event and only a few cases have been reported in the literature.5, 20 Malignant melanoma has a very unpredictable behavior and metastatic melanoma occurring without a known primary site has been very well documented. The distinction between metastatic melanoma and primary melanoma of soft parts/clear cell sarcoma may be impossible to establish on the basis of histology or special stains and requires in most instances a good clinical history documenting the prior occurrence of a primary melanoma. Two of our 20 cases of metastatic melanoma to soft tissue presented without any history or evidence of a primary melanoma elsewhere. The distinction of these cases from primary melanoma of soft parts is mostly academic, as their treatment and prognosis is likely to be the same.

The clinical features of a metastatic carcinoma to soft tissue can mimic a soft tissue sarcoma in many respects, but a painful soft tissue mass is more commonly noted in patients with soft tissue metastasis than in primary sarcomas.3 In 43 of our cases a diagnosis of primary soft tissue sarcoma was entertained clinically for patients who presented with a solitary soft tissue mass with no prior history of carcinoma. Making the distinction between a metastatic carcinoma to soft tissue and a primary sarcoma of soft tissue is of extreme importance for therapeutic stratification.

The histologic differential diagnosis of a metastatic tumor to soft tissue involves for the most part ruling out a wide variety of primary soft tissue sarcomas. If the appearance of the lesion demonstrates a poorly differentiated histology, it will most likely deserve a thorough histopathologic workup including the use of a basic panel of immunohistochemical stains to rule out a metastasis to soft tissue. The basic panel of immunohistochemical stain that we recommend in the routine workup of these tumors includes a combination of at least 2 epithelial markers (such as a broad-spectrum cytokeratin, EMA, MOC31), a neural or melanocytic marker (such as S-100 protein, HMB-45, and Melan-A), and other markers associated with various types of soft-tissue sarcomas, such as a muscle-associated antigen (SMA, desmin, calponin), vascular endothelial marker (CD31 and CD34), and bcl-2. The use of additional specific markers can be subsequently applied depending on the circumstances.

In this study the most common type of metastases to soft tissue were carcinomas (70%). The carcinoma that most often metastasized to soft tissue was lung cancer. Lung cancer was also found to account for the most common primary source of metastases to soft tissue in 1 large autopsy study.13 Histologically, metastatic lung adenocarcinomas show a well-to-moderately differentiated glandular architecture with tumor necrosis and increased mitotic activity. However, by histology alone it would be impossible to establish the exact site of origin because many carcinomas can share the same morphology. Use of a panel of markers that includes cytokeratins 7 and 20, and markers associated with either pulmonary (TTF1) or gastrointestinal differentiation (CDX2), may be of help in further defining the exact site of origin of the tumor.21, 22 Tumor cells displaying abundant clear and granular cytoplasm should be a clue for renal cell carcinoma. If the tumor has features compatible with small cell carcinoma, dot-like paranuclear staining for cytokeratin, nuclear staining for TTF-1, and cytoplasmic staining for chromogranin or synaptophysin will confirm this diagnosis.

Malignant melanoma accounted for 17% of all metastatic tumors to soft tissue in our study. These were some of the cases most likely to be confused histologically for a primary soft-tissue sarcoma due to their variegation of growth patterns, nuclear pleomorphism, and frequent spindle cell morphology. Coarse melanin pigment was seen in 7 cases. The rest of the cases were amelanotic. The amelanotic cases were the most prone to be confused for primary soft-tissue sarcomas. Immunohistochemical stains were useful for confirming the diagnosis. All of our cases were strongly positive for S-100 protein. The majority of the cases also labeled at least focally with HMB-45 and Melan-A, although 4 cases were negative for the latter 2 markers, a rare but important occurrence in metastatic melanoma.23

Metastases to distant soft tissue sites from a primary soft tissue sarcoma are very rare. In our study, they accounted for < 10% of all cases of metastases to soft tissue. This low incidence may be skewed due to the inherent difficulties in the histopathologic diagnosis of sarcoma, or due to difficulties in distinguishing clinically between metastases of sarcoma and multifocal disease. Sarcomas that metastasize to soft tissue will usually have a well-documented history of sarcoma elsewhere. Epithelioid sarcoma, a distinctive type of superficial soft tissue sarcoma of the distal extremities that may be commonly confused for a metastatic carcinoma (and which shows a high propensity to metastasize to soft tissue), was not encountered in the current study. In this study, the most common sarcoma that metastasized to soft tissue was leiomyosarcoma.

The distribution of metastases to soft tissue in our series paralleled in some cases the relative incidence of these tumors in the general population, such as lung carcinoma, which is the most common type of internal malignancy in humans and which accounted for the second-largest number of metastases to soft tissue in our study. However, in other instances there could be no correlation observed, such as with renal cell carcinoma, which was also 1 of the most commonly encountered type of tumor that metastasized to soft tissue in our study despite it being a less common type of cancer. The incidence of metastases from malignant melanoma to soft tissue, conversely, was also in proportion to the overall incidence of malignant melanoma in the general population and may simply be a reflection of the highly aggressive nature of metastatic malignant melanoma and of its propensity to spread in an unpredictable fashion to all areas of the body. In fact, malignant melanoma was the most frequent specific tumor type that metastasized to soft tissue in our study, outnumbering other site or organ-specific types of cancers such as lung, breast, and colon carcinoma. The possibility of metastatic melanoma should therefore be seriously considered by the pathologist in the differential diagnosis of a solitary soft tissue mass and should be appropriately ruled out by histopathologic examination and special stains.

In summary, we have herein reported 118 cases of metastatic tumors presenting as discrete soft tissue masses mimicking a soft tissue sarcoma. Our cases accounted for 1.6% of all soft-tissue tumors accessed at our institution over a 30-year period. Awareness of this occurrence is of extreme importance because metastatic carcinomas can often be confused clinically and histologically with primary soft tissue sarcoma, and its prognostic and therapeutic implications are quite different.