Unraveling the Mystery of a Supraclavicular Mass

July 16, 2009

In this month’s Clinical Quandaries case report, a 49-year-old man was diagnosed with an extragonadal seminoma. The diagnosis was made on a fine-needle aspiration of a 7-cm supraclavicular mass. Beta–human chorionic gonadotropin and alpha-fetoprotein levels were normal. The only cancer history for this patient was a basal cell carcinoma of the nose. The patient complaints were fatigue; sensory and motor dysfunction in the hands, legs, and abdomen; unsteady gait; and blurry vision.

In this month’s Clinical Quandaries case report, a 49-year-old man was diagnosed with an extragonadal seminoma. The diagnosis was made on a fine-needle aspiration of a 7-cm supraclavicular mass. Beta–human chorionic gonadotropin and alpha-fetoprotein levels were normal. The only cancer history for this patient was a basal cell carcinoma of the nose. The patient complaints were fatigue; sensory and motor dysfunction in the hands, legs, and abdomen; unsteady gait; and blurry vision.

The neurologic examination was extremely abnormal-the symptomatology was not consistent with an anatomic nerve pathway defect, but a predominance of sensory neuropathy was described. The working diagnosis with this physical evaluation and additional testing was the Miller-Fisher syndrome, a syndrome characterized by the triad of ophthalmoplegia, ataxia, and areflexia, but without significant motor or sensory deficit in the limbs. A complete recovery usually occurs without specific treatment.

The majority of individuals with Miller-Fisher syndrome have a unique ganglioside anti-GQ1b antibody that characterizes the disorder. Consistent with this diagnosis, a spinal tap reveals the presence of elevated protein levels. Treatments include intravenous immunoglobulin or plasmapheresis and supportive care. The ganglioside antibodies were negative, but an anti-Hu elevated titer was noted. This prompted an incisional biopsy of the supraclavicular mass.

The symptomatology thus consisted of a neoplastic mass in the supraclavicular area and a rapidly progressing neurologic syndrome related to the presence of anti-Hu antibodies.

Hu Antibodies and Cancer

Hu antibodies were discovered by an immunohistologic preparation of a Western blot analysis of proteins from the cerebrocortical neurons, serum, and cerebrospinal fluid (CSF) of a patient (Mr. Hu…, the first two letters of his last name) who was diagnosed with small-cell lung cancer (SCLC) and subacute sensory polyneuropathy.[1,2] Hu antibodies are IgGs directed against the Hu antigens, a family of proteins found in all neurons of the central and peripheral nervous system, and a few extraneuronal cells. The function of Hu antigens has not been completely elucidated, but they appear to play a role in early neuronal embryogenesis by interfering with mRNA to promote differentiation and maintenance of the neuronal phenotype. The Hu antigen is also present in most small-cell carcinomas.[3,4] For a comprehensive review of antineuronal antibody, see Senties-Madrid and Vega-Boada.[5]

The tumor most likely associated with Hu antibodies is SCLC. The specific neurologic syndrome caused by Hu antibodies is called paraneoplastic encephalomyelitis/sensory neuronopathy (PEM/SN).[6] Less than 1% of patients with SCLC develop PEM/SN. However, in a study of 196 patients with SCLC, 16% of patients produced Hu antibodies.[7] Interestingly, the anti-Hu syndrome is more common in women, despite their being less affected by SCLC.

The presence of these antibodies was associated with earlier disease stage and longer survival, as well as being a strong predictor of complete response to therapy. The presence of antibodies suggests an immune response against the antigen present in the tumor cells, thereby conferring a better prognosis. A discussion of the immunologic mechanisms involved is beyond the scope of this article, but the intensity or quality of the immune response is responsible for both the neurologic destruction and immunologic control of the tumor. In fact, most patients with Hu antibodies do not have neurologic symptoms.

The tumors associated with Hu antibodies include small-cell cancers (of the lung and other origins, 80%–90%), neuroblastoma, thymic carcinoma, synovial sarcoma, Hodgkin disease, nonseminomatous testicular germ cell tumors, adrenal tumors, lung carcinoma, prostate cancer, and one other case of chondromyxosarcoma.[5]

First Quandary: What is the specific tumor associated with the neurologic presentation of this patient?

In the reported case, the location of a supraclavicular node would have been suspicious for a metastasis of SCLC. A supraclavicular lymph node is present in 17% of patients with SCLC and is correlated with the presence of distant metastases at baseline.[8] Therefore, my working diagnosis would have been SCLC. Why was the diagnosis of extragonadal seminoma not questioned? The patient does not fit the age group, the location does not fit the tumor,[9] and the laboratory findings (completely normal tumor markers) do not fit the disease. The patient received a course of chemotherapy for seminoma, with drugs that could have been selected for the class of sarcomas including the primitive neuroectodermal tumor (see below), in which extraskeletal myxoid chondrosarcoma (EMC) probably belongs. But this treatment was not optimal.

Only after the Hu antibody titers came back positive was a new biopsy done. The diagnosis was made by a karyotype showing the classical t(9;22)(q22;q12) translocation seen in EMC.[10] This unexpected finding shows the importance of obtaining enough tissue for the diagnosis of cancer of unknown origin. Interestingly, the translocation involves the Ewing gene on chromosome band 22q12 and the CHN gene on chromosome band 9q22-31, leading to a specific fusion protein causing the development of the tumor. In the EWS-CHN fusion protein, the C-terminal RNA-binding domain of EWS is replaced by the entire CHN protein, which has a long N-terminal domain, a central DNA binding domain, and a C-terminal ligand-binding/dimerization domain.

CHN, a member of the steroid/thyroid receptor gene superfamily, acts as a transcription factor and appears to be the human homolog of the rat gene NOR1, which was recently identified as a sequence overexpressed in rat brain cells undergoing apoptosis.[10] The Ewing gene is also translocated in clear cell sarcoma, Ewing’s sarcoma/primitive neuroectodermal tumor, and desmoplastic round cell tumor, which all show involvement of chromosome 22 with formation of a hybrid gene between the Ewing’s sarcoma gene on band q12 and a transcription factor gene.

Synovial sarcoma, another sarcoma positive for Hu antibody, has recently been shown to closely resemble the molecular expression profile of malignant peripheral sheath tumors, and its line of differentiation may be determined to be primitive neural or notochordal.[11] All these sarcomas may in fact arise during neuroectodermic differentiation as illustrated by nonspecific enolase stains.

Extraskeletal Myxoid Chondrosarcoma

EMC must be differentiated from skeletal myxoid chondrosarcoma. It is a relatively rare but well recognized neoplasm, which has the pathognomonic reciprocal t(9;22), resulting in a fusion of the EWS and CHN genes. On molecular analysis, the EWS-CHN fusion RNA resulting from the t(9;22) is detected in most cases.[12] EMC was originally considered possibly a chondrosarcoma because of histologic resemblance to cartilaginous tumors and variable reactivity for S-100 protein, which is commonly expressed in cartilaginous tumors. Interestingly, a few EMCs are reactive for proteins (CD57 and GFAP or glial fibrillary acidic protein) that may also be expressed in nerve sheath tumors and synovial sarcoma.

EMCs are predominantly located in the deep soft tissues of the lower extremity (60%) and buttock (20%), with a mean tumor size around 10 to 15 cm at diagnosis, and spread distantly more often than their skeletal counterpart. Histologic grade in the EMC group correlates with survival, with 82% of high-grade tumors becoming metastatic.

Investigators looked at clinical behavior and treatment responses in a series of 87 patients.[13] The median age of the patients at the time of diagnosis was 49 years, with a male-to-female ratio of 2:1. The 5-year, 10-year, and 15-year overall survival rates were 82%, 65%, and 58%, respectively. For patients presenting without metastases, 37% developed local recurrence (median time of 3.3 years) and 26% developed distal recurrence (median time of 3.2 years). Approximately 13% of patients presented with metastases. A total of 21 patients received chemotherapy without significant radiologic or clinical responses. The median time to disease progression while receiving chemotherapy was 5.2 months, with more than 70% of patients progressing within 6 months of therapy. Aggressive control of localized disease and prevention of recurrence is the primary approach to management, because patients with systemic disease do not respond to chemotherapy.[13]

Second Quandary: What is the appropriate treatment of a paraneoplastic syndrome and its associated tumor?

Paraneoplastic syndromes often develop before the tumor becomes detectable, sometimes years before the cancer diagnosis.[14] In this case, the paraneoplastic syndrome was concomitant with the development of the supraclavicular mass. The main clinical issue is that the immunologic reaction initiated by the Hu antibody caused permanent damage to the neuronal cells.

Treatment of an autoimmune antibody-mediated syndrome consists of immunoglobulin and/or plasmapheresis followed by immunosuppressive therapy to prevent the release of additional antibodies from the B-cell clone. An attempt at immunoglobulin therapy was unsuccessful. A new protocol with tacrolimus treatment did not work either. Plasmapheresis was not tried. One issue involves the presence of the antibody in the plasma and cerebrospinal fluid, where it is secreted by B cells. While the plasma could be cleared, the CSF may not be because of the blood-brain barrier. Therefore, destruction of central neurons continues, aggravating the neurologic condition. I wonder if CSF exchange could have been performed.[15]

Treatment of the primary tumor may also prevent further stimulation of the immune reaction, but often does not help reverse the paraneoplastic syndrome. Localized EMC has a good prognosis with regional care (surgical resection followed by radiation therapy) but in general is not sensitive to chemotherapy.[12] Although mild forms of paraneoplastic neurologic syndromes may stabilize upon successful treatment of the tumor, most do not. One wonders what life this patient will have following the curative resection of his tumor. The ultimate prognosis of this patient is related to the neurologic destruction. Unfortunately, most patients affected with generalized neurologic paraneoplastic syndromes die within 1 year.[14] A gentle referral to hospice may be the best approach in this case, until we better understand how to reverse the neuronal damage.

Financial Disclosure:The author has no significant financial interest or other relationship with the manufacturers of any products or providers of any service mentioned in this article.


1. Wilkinson PC, Zeromski J: Immunofluorescent detection of antibodies against neurones in sensory carcinomatous neuropathy. Brain 88:529-583, 1965.
2. Graus F, Cordon-Cardo C, Posner JB: Neuronal antinuclear antibody in sensory neuronopathy from lung cancer. Neurology 35:538-543, 1985.
3. Dalmau J, Furneaux HM, Cordon-Cardo C, et al: The expression of the Hu (paraneoplastic encephalomyelitis/sensory neuronopathy) antigen in human normal and tumor tissues. Am J Pathol 141:881-886, 1992.
4. Dalmau J, Furneaux HM, Gralla RJ, et al: Detection of the anti-Hu antibody in the serum of patients with small cell lung cancer-a quantitative western blot analysis. Ann Neurol 27:544-552, 1990.
5. Senties-Madrid H, Vega-Boada F: Paraneoplastic syndromes associated with anti-Hu antibodies. Isr Med Assoc J 3:94-103, 2001.
6. Dalmau J, Graus F, Rosenblum MK, et al: Anti-Hu–associated paraneoplastic encephalomyelitis/sensory neuronopathy. A clinical study of 71 patients. Medicine (Baltimore) 71:59-72, 1992.
7. Graus F, Dalmau J, Reñé R, et al: Anti-Hu antibodies in patients with small-cell lung cancer: Association with complete response to therapy and improved survival. J Clin Oncol. 15:2866-2872, 1997.
8. Urban T, Chastang C, Vaylet F, et al: Prognostic significance of supraclavicular lymph nodes in small cell lung cancer. Chest 114;1538-1541, 1998.
9. Bokemeyer C, Droz JP, Horwich A, et al: Extragonadal seminoma: An international multicenter analysis of prognostic factors and long term treatment outcome. Cancer 91:1394-1401, 2001.
10. Clark J, Benjamin H, Gill S, et al: Fusion of the EWS gene to CHN, a member of the steroid/thyroid receptor gene superfamily, in a human myxoid chondrosarcoma. Oncogene 12:229-235, 1996.
11. Ishibe T, Nakayama T, Aoyama T, et al: Neuronal differentiation of synovial sarcoma and its therapeutic application. Clin Orthop Relat Res 466:2147-2155, 2008.
12. Antonescu CR, Argani P, Erlandson RA, et al: Skeletal and extraskeletal myxoid chondrosarcoma: A comparative clinicopathologic, ultrastructural, and molecular study. Cancer 83:1504-1521, 1998.
13. Drilon AD, Popat S, Bhuchar G, et al: A retrospective review from 2 referral centers emphasizing long-term outcomes with surgery and chemotherapy. Cancer 113:3364-3371, 2008.
14. Dalmau J, Rosenfeld MR: Paraneoplastic syndromes of the CNS. Lancet Neurol 7:327-340, 2008.
15. Kristof RA, Clusmann H, Koehler W, et al: Treatment of accidental high dose intraventricular mezlocillin application by cerebrospinal fluid exchange. J Neurol Neurosurg Psychiatry 64:379-381, 1998.