Poorly differentiated carcinoma of unknown primary site

J
INTRODUCTION  Cancer of unknown primary site (CUP) is a common clinical entity, accounting for 2 percent of all cancer diagnoses in the Surveillance, Epidemiology, and End Results (SEER) registries between 1973 and 1987 . Within this category, tumors from many primary sites with varying biologies are represented; this heterogeneity has made the design of therapeutic studies difficult.

Substantial improvements have been made in the management and treatment of some patients with CUP. The identification of specific subgroups of treatable patients has been made possible by the development of specialized immunohistologic techniques that can aid in tumor characterization, and by the recognition of several distinct clinical syndromes that permit prediction of chemotherapy responsiveness. (See "Overview of the classification and management of neoplasms of unknown primary site").

The typical patient with CUP presents with symptoms referable to a metastatic site. The initial work-up, including physical examination, laboratory and radiographic study, fails to identify the primary site. Light microscopic evaluation of biopsy material places the tumor into one of six histologic categories, which then guides further evaluation and treatment:

  • Adenocarcinoma  approximately 70 percent
  • Poorly differentiated carcinoma  15 to 20 percent; an additional 10 percent represent poorly differentiated adenocarcinoma
  • Squamous cell carcinoma 5 percent
  • Poorly differentiated neoplasm  less than 5 percent
  • Neuroendocrine carcinoma  less than 5 percent.

Many early empiric chemotherapy trials of CUP included patients with poorly differentiated carcinoma in addition to the more common adenocarcinomas of unknown primary origin, since these patients were assumed to have the same poor response to treatment and short survival  It is now clear that some patients with poorly differentiated carcinomas have extremely responsive neoplasms, and some are curable with combination chemotherapy.  The main goal in the work-up of these patients is to identify those who have potentially chemotherapy-responsive cancers.

PATHOLOGIC DIAGNOSIS  Examination of poorly differentiated carcinomas using light microscopy alone is inadequate. There are no light microscopic features that can reliably distinguish tumors that are chemotherapy-responsive from those that are nonresponsive  . Furthermore, even with careful retrospective review of these cases, some responsive tumors of well-defined types (eg, germ cell tumor, lymphoma) cannot be readily distinguished morphologically

Patients with the light microscopic diagnosis of poorly differentiated carcinoma of unknown primary site should undergo additional pathologic study with immunohistochemical staining similar to patients with poorly differentiated neoplasms Electron microscopy (EM) should be considered if the immunohistochemical results are inconclusive.

The frequency of identification of unsuspected tumor types (particularly lymphoma) is much lower in this group (10 to 20 percent) than in patients with poorly differentiated neoplasms (up to 66 percent) . However, unsuspected diagnoses that may provide the opportunity for reasonable therapeutic options may still be suggested by the results of immunohistochemistry and/or cytogeneteics

DIAGNOSTIC EVALUATION  A thorough history, physical examination, routine laboratory testing, and chest radiograph should be obtained in all patients. Computerized tomography (CT) of the chest and abdomen should also be performed because of the high frequency of mediastinal and retroperitoneal lymphadenopathy in these patients.

Tumor markers  Serum levels of human chorionic gonadotropin (hCG) and alpha-fetoprotein (AFP) should be obtained in all patients, since significant elevations of these tumor markers suggests the diagnosis of extragonadal germ cell tumor. This syndrome is discussed in detail elsewhere.

Other serum tumor markers, such as carcinoembryonic antigen, CA-125, CA19-9, and CA15-3, are often elevated, but are not helpful in predicting the response to therapy.

Immunohistochemistry  Immunohistochemical study is particularly valuable in narrowing the differential diagnosis. As an example, in one series of 87 patients with poorly differentiated carcinoma in whom a standard battery of immunohistochemical stains was performed, a diagnosis other than poorly differentiated carcinoma was suggested in 16 (18 percent) :

  • Melanoma  8 patients
  • Lymphoma  4 patients
  • Neuroendocrine tumor  3 patients
  • Prostate cancer  1 patient.

In a retrospective review of these patients, all four with immunohistochemical features of lymphoma (two of whom had been previously diagnosed with poorly differentiated carcinoma by EM) had clinical features compatible with lymphoma, and all were long-term survivors. Three of the eight patients in whom the diagnosis of melanoma was suggested by immunohistochemistry were long-term survivors following chemotherapy for metastatic disease.

Chromosomal studies  In addition to immunohistochemistry, some specific tumors can be defined by a characteristic chromosomal abnormality:

  • A specific chromosomal translocation (t[11;22] [q24;q12]) is found in all peripheral neuroepitheliomas and is also frequent in Ewing's sarcoma.
  • A syndrome of undifferentiated or poorly differentiated carcinoma arising in midline structures (eg, nasopharynx, mediastinum, bladder) in young patients has recently been described . These tumors are associated with a single specific chromosomal translocation t(15;19), which results in a novel fusion oncogene, BRD4-NUT. These patients have a poor prognosis despite aggressive chemoradiotherapy.
  • Many germ cell tumors have a specific 12p chromosomal abnormality (isochromosome 12p) that is diagnostic if present

TREATMENT When specialized histologic studies identify a treatable neoplasm, therapy should be administered following guidelines established for that tumor. Examples of treatable tumor types occasionally identified in these patients include neuroendocrine carcinoma, Ewing's sarcoma, and a variety of primitive sarcomas. Patients without a specific diagnosis can be divided into those with and those without characteristics of extragonadal germ cell tumors.

Patients with characteristics of extragonadal germ cell tumors  A few patients with poorly differentiated carcinoma of unknown primary site have extragonadal germ cell tumors that are unrecognizable by standard histologic criteria. These patients typically have some (but usually not all) of the following characteristics

  • Young age
  • Male gender
  • Predominant tumor location in the mediastinum or retroperitoneum
  • Marked elevation of the serum tumor markers hCG or AFP
  • Presence of 12p chromosomal gain (isochromosome 12p) on molecular genetic analysis
  • Tumor immunohistochemical staining for OCT4 (also called POU domain class 5 transcription factor 1) [5] .

Most young males with poorly differentiated carcinoma and these features of extragonadal germ cell tumor have an excellent response to chemotherapy, and some are cured when they are treated with cisplatin-based regimens. For this reason, all patients with poorly differentiated carcinoma or poorly differentiated adenocarcinoma of unknown primary site who have these characteristics should be treated according to the guidelines established for poor prognosis germ cell tumors. This syndrome is discussed in detail elsewhere.

Empiric therapy for other patients  The optimal treatment of patients with poorly differentiated carcinoma of unknown primary site who lack the characteristics of extragonadal germ cell tumor remains a subject of controversy and different regimens have been used.

  • We treated 220 patients with cisplatin-based regimens used in the treatment of advanced testicular cancer (ie, cisplatin, vinblastine, bleomycin [PVB] or cisplatin, etoposide, bleomycin [BEP), even though most of the patients in this group did not have clinical characteristics strongly suggestive of extragonadal germ cell tumor . However, many patients were young (median age 39 years), and approximately 40 percent had predominant tumor location in the mediastinum or retroperitoneum.

The overall and complete response rates were 62 and 26 percent, respectively. A minority of patients (14 percent) remained tumor-free after a minimum follow-up of eight years . In a multivariate analysis, clinical features predictive of a favorable treatment outcome included tumor location in the retroperitoneum or peripheral lymph nodes, fewer sites of metastases, younger age, and no history of cigarette smoking.

  • In another retrospective analysis of a large series of patients with poorly differentiated carcinoma, the authors could not identify a subset of patients who had long-term survival following chemotherapy. Patients with clinical features of extragonadal germ cell tumor were excluded, and only a minority received cisplatin-based regimens. Although no long-term survivors were identified, several of the same favorable prognostic factors were detected, including predominant lymphadenopathy, fewer metastatic sites, younger age, female sex, and poorly differentiated carcinoma histology (versus adenocarcinoma).

Other investigators have also documented higher response rates in patients with poorly differentiated carcinoma (versus adenocarcinoma) who were treated with cisplatin-based regimens , however, this may not translate into any survival benefit. As an example, in one randomized phase II study in which 78 patients with CUP (13 of whom had poorly differentiated tumors) were randomly assigned to cisplatin with either irinotecan or gemcitabine, objective response rates were significantly higher with the gemcitabine regimen (55 versus 38 percent), and toxicity was less pronounced. Despite this, median survivals were eight and six months, respectively.

Taxane/platinum-based chemotherapy regimens have also yielded high response rates with moderate toxicity, and are reasonable options for this group of patients . In one of these studies, paclitaxel, carboplatin, and oral etoposide was administered to 71 patients with CUP, of whom 30 had poorly differentiated carcinoma. Forty-six percent had a major objective response to treatment, and the median survival was 11 months. With 34 months of follow-up, the one, two, and three-year survival rates were 48, 20, and 14 percent, respectively.

High dose chemotherapy  The use of high dose chemotherapy with stem cell or hematopoietic growth factor support does not appear to improve the outcome of these patients . In one report, 20 patients under the age of 61 with poorly differentiated adenocarcinoma or poorly differentiated carcinoma were treated with high dose chemotherapy, hematopoietic progenitor cells and growth factor support . There was only one complete response and overall median survival was 11 months.

Recommendations  At present, a trial of combination chemotherapy should be considered for most patients with poorly differentiated carcinoma or poorly differentiated adenocarcinoma of unknown primary site. Patients with clinical features of extragonadal germ cell tumor should receive cisplatin-based chemotherapy with a regimen that is used for the treatment of germ cell tumors. Cisplatin-based regimens have also yielded relatively high response rates in the other patients in this group. However, recent experience with carboplatin-containing regimens suggests that this drug is an acceptable substitute for cisplatin . Taxane/platinum-based chemotherapy regimens are another good choice for this group of patients. By reducing the toxicity of therapy, an empiric trial of treatment becomes feasible in a larger percentage of these patients.