| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Vascular and Interventional Radiology |
1 From the Departments of Clinical Radiology (J.A.S., S.E.S.), Pathology (N.W., A.P.B.), and Gynaecology (G.L.) and the Imperial Cancer Research Fund Cancer Medicine Research Unit (T.J.P.), St James’s University Hospital, Beckett St, Leeds LS9 7TF, England. Received November 29, 2000; revision requested January 11, 2001; revision received March 12; accepted March 22. J.A.S. supported by a Pump Priming Grant from the Royal College of Radiologists. T.J.P. supported by the Imperial Cancer Research Fund. Address correspondence to J.A.S. (e-mail: wilsonspencer@compuserve.com).
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODS RESULTSDISCUSSIONREFERENCES |
|---|
MATERIALS AND METHODS: Thirty-five women (age range, 47–85 years; mean age, 69 years) prospectively identified in a gynecologic oncology center underwent 18-gauge core biopsy in omental cake (n = 25), peritoneal (n = 7), or adnexal (n = 3) sites. No complications of biopsy occurred. Standard hematoxylin-eosin analysis of the biopsy cores was supplemented by immunohistochemical markers to CA-125, carcinoembryonic antigen, cytokeratin 7, and cytokeratin 20. Diagnoses were validated with further multidisciplinary review, subsequent surgery, and response to specific chemotherapy.
RESULTS: In 27 (77%) of the 35 women, a confident primary site diagnosis was obtained with standard hematoxylin-eosin analysis of core biopsy material from the following sites: ovary (n = 22), breast (n = 2), colon (n = 2), and lymphoma (n = 1). The finding at hematoxylin-eosin analysis in another seven (20%) women was poorly differentiated adenocarcinoma with no definite primary site but with an immunohistochemical profile suggesting ovarian cancer (CA-125 positive, carcinoembryonic antigen negative, cytokeratin 7 positive, cytokeratin 20 negative). There was one false-negative biopsy result.
CONCLUSION: Image-guided peritoneal core biopsy with hematoxylin-eosin analysis supplemented with immunohistochemical analysis is a simple, safe, and accurate technique for providing site-specific diagnoses in women with undiagnosed peritoneal carcinomatosis.
Index terms: Computed tomography (CT), guidance, 791.12112, 791.12115 • Peritoneum, biopsy, 791.1261 • Peritoneum, neoplasms, 791.32, 791.33 • Ultrasound (US), guidance, 791.12985
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODS RESULTSDISCUSSIONREFERENCES |
|---|
It is now recognized that surgery, even in the most expert hands, is unlikely to achieve optimal debulking in some patients with very widespread peritoneal disease. Other patients with advanced disease are not candidates for major cytoreductive surgery. As a result, and as supported by findings of a study by the European Organisation for Research and Treatment of Cancer (2), the recent tendency has been to treat such patients with a number of cycles of primary (neoadjuvant) chemotherapy, with the aim of achieving substantial cytoreduction, and to follow this with interval debulking surgery and further cycles of chemotherapy. However, most experts regard this approach as experimental; therefore, it is the subject of further investigation in ongoing randomized clinical trials, including the Gynecologic Oncology Group Protocol 152 in the United States and the OVO6 study by the Medical Research Council in the United Kingdom. In this protocol, women may be randomly assigned to receive standard treatment with primary surgery followed by chemotherapy or with neoadjuvant chemotherapy followed by interval debulking surgery.
Prior to commencing chemotherapy, optimal management should include a firm histologic diagnosis to confirm the clinical, radiologic, and biochemical diagnoses of ovarian cancer. Metastatic breast or gastrointestinal cancer may clinically, radiologically, and even biochemically mimic that arising from ovarian cancer, and cytologic analysis of ascitic fluid rarely results in a site-specific diagnosis (3). Thus, historically, either laparoscopy or laparotomy has been required to obtain a tissue diagnosis, or a series of investigations has been performed to exclude primary tumors at other sites. The purpose of our study was to assess image-guided peritoneal core biopsy for the diagnosis of tumor type and treatment of patients with peritoneal carcinomatosis.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODS RESULTSDISCUSSIONREFERENCES |
|---|
Imaging
Core biopsy was performed after diagnostic abdominal ultrasonography (US) and computed tomography (CT), multidisciplinary case review, and aquisition of patient informed consent and local research ethics committee approval. The multidisciplinary team comprised consultants in gynecologic surgery (G.L.), radiation oncology, medical oncology (T.J.P.), palliative care, pathology (A.P.B., N.W.), and radiology (J.A.S., S.E.S.).
Diagnostic CT examinations were performed by using one of two machines (Somatom Plus S or AR Star; Siemens, Erlangen, Germany). The standard oral bowel preparation protocol included the administration of 5 mL of diatrizoate meglumine (Gastrografin; Schering, Burgess Hill, England) in 200 mL of water 6–8 hours prior to examination and 1,000 mL of 3% diatrizoate meglumine in the hour prior to examination, with the final 200 mL of this administered immediately prior to examination. A vaginal tampon was used. Contrast material was routinely administered; either 100 mL iohexol (Omnipaque 300; Nycomed, Birmingham, England) or iopromide (Ultravist 300; Schering) was intravenously administered at a rate of 3 mL/sec with a pump injector. A helical examination technique with 10-mm collimation was used in the liver and spleen to coincide with maximal portal venous enhancement; this technique was followed by examination (10-mm collimation) of the pelvis from the perineum cephalad to the upper volume. The examination was supervised by a consultant radiologist. Additional sections were obtained following a delay of 30 minutes, with further oral contrast enhancement or imaging in decubitus positions in selected cases to allow distinction of bowel masses from pelvic, peritoneal, or omental masses.
The consultant radiologist supervised trained sonographers as they performed US by using a variety of machines with curvilinear phased-array probes and variable-frequency (2.5–5-MHz) application; the patient was supine and had a filled urinary bladder.
Criteria for image-guided biopsy were the following: (a) the presence of a peritoneal mass that allowed core biopsy at diagnostic imaging; (b) no bleeding diathesis, with a platelet count of 10 x 109/L or more and an international normalized ratio of 1.4 or less; and (c) a decision, made after multidisciplinary review, that obtaining a definitive diagnosis by nonsurgical means was required to plan further treatment.
Biopsy was performed by three consultant radiologists (including J.A.S., S.E.S.) experienced in abdominal biopsy techniques. Biopsy was performed with CT control in 31 patients and with US control in four patients. Image-guided biopsy was performed as a separate procedure after multidisciplinary review of the diagnostic studies. For biopsy, CT was performed with only oral contrast material preparation, after the acquisition of a limited number of localizing sections was planned by using findings of the prior examination. An 18-gauge cutting needle with a spring-loaded device was used (Temno; Bauer Medical International, Santo Domingo, Dominican Republic). Biopsy samples were taken from the infracolic omental cake in 25 patients, other masses within the peritoneal cavity in seven patients (three right lower quadrant and four left upper quadrant), and adnexal masses in three patients. One or two cores were obtained at the discretion of the supervising radiologist. No cytologic or histopathologic review was performed at the time of biopsy.
The biopsy procedure typically lasted 15–20 minutes, was well tolerated, and resulted in no complications. No patient experienced worsening of abdominal symptoms, and no patient required blood transfusion or corrective surgery after the procedure.
Pathologic Analysis
Samples obtained at core-needle biopsy were fixed in formalin and embedded in paraffin wax; 3–4-µm-thick sections were cut. These were stained with hematoxylin-eosin. Further immunohistochemical analysis of the biopsy material was performed by using the labeled streptavidin–biotin peroxidase system with monoclonal antibodies to carcinoembryonic antigen (Dako, Carpenteria, Calif), cytokeratin 7 (Dako), cytokeratin 20 (Dako), and CA-125 (Novocastra, Newcastle, England). At the discretion of the reporting pathologist (N.W.), additional monoclonal antibodies M7047 (Dako), M3569 (Dako), and GCDFP-15 (Signet Labs, Dedham, Mass) for estrogen receptors, progesterone receptors, and breast marker BRST2, respectively, were used in selected women with previous breast cancer. Isotypes are lineage specific, and this characteristic is retained during malignant transformation and progression. Antibodies to cytokeratin 7 react with ovarian epithelia but rarely with colonic epithelia; antibodies to cytokeratin 20 react conversely (4–6). Only strong and widespread positive staining was accepted as a positive finding.
Ascites was drained in 19 patients prior to the core biopsy procedure; draining with US control was performed as a palliative procedure in 16 patients and with the biopsy procedure through the same puncture site in three. Small amounts of ascites in nine other women were not aspirated for diagnostic purposes. All cytologic analyses were performed by the same consultant cytopathologist (A.P.B.), and the core biopsy samples were all analyzed by a different consultant pathologist (N.W.). The pathologists who analyzed the specimens were blinded to the clinical and imaging findings.
Treatment
After biopsy, further multidisciplinary review was performed. If a clear histologic diagnosis was made on the basis of hematoxylin-eosin morphologic findings that were supported by results from immunohistochemical staining, it was accepted for the commencement of site-specific therapy. The findings of any subsequent surgery were correlated with those of the analysis of core biopsy samples. In women without a definitive hematoxylin-eosin morphologic diagnosis, the immunohistochemical diagnosis was further validated with the response to specific therapies or subsequent surgery.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODS RESULTSDISCUSSIONREFERENCES |
|---|
There was one false-negative biopsy result and one inconclusive core biopsy finding. In the patient with the latter, the biopsy core contained adenocarcinoma but insufficient neoplastic material to allow a confident site-specific diagnosis. This patient had convincing clinical and imaging features of ovarian cancer. Findings in a repeat biopsy sample obtained 1 week later, after further multidisciplinary review, was diagnostic of serous papillary adenocarcinoma consistent with an ovarian origin. The false-negative biopsy result occurred in a patient who had a subacute history of right iliac fossa pain with a history of endometrial cancer (mixed müllerian tumor). CT revealed infiltration of the peritoneal fat surrounding the cecum, appendix, and ileum; this finding suggested a diagnosis of recent appendicitis. She was treated conservatively, but at repeat CT, residual pericolic thickening was present, and she was referred to the multidisciplinary team. There was clinical concern about recurrent disease, but analysis of the core biopsy sample obtained in the pericolic thickening revealed granulation tissue; this biopsy was not repeated. She later underwent surgery for a distal small-bowel obstruction; during surgery, evidence of a second primary cancer, a cecal cancer (Dukes stage C) affecting the appendix and terminal ileum with local peritoneal carcinomatosis, was found.
Based hematoxylin-eosin findings, primary-site diagnoses were as follows: ovarian or primary peritoneal cancer in 22 patients, colorectal cancer in two patients (in one, recurrent), recurrent breast cancer in two patients, and lymphoma in one patient (Figs 1, 2). Thus, a specific diagnosis of the primary tumor site, with clear diagnostic features at hematoxylin-eosin analysis that were supported by findings of appropriate immunohistochemical marker expression, was obtained in these 27 patients.
|
|
|
|
Ascitic fluid was drained in 19 women. Cytologic analysis was performed by investigators who were blinded to clinical details, and histologic findings revealed diagnostic findings of adenocarcinoma of unspecified site in 11 patients; hematoxylin-eosin diagnoses were ovarian cancer (n = 8) and poorly differentiated cancer of no specific site (n = 3). Cytologic analysis revealed cancer of probable ovarian origin in three patients who had papillary clusters; hematoxylin-eosin analysis revealed were ovarian cancer (n = 2) and poorly differentiated cancer of no specific site (n = 1). Findings in the samples were nondiagnostic in five patients; diagnoses made at hematoxylin-eosin analysis were ovarian cancer (n = 3), colon caner (n = 1), and poorly differentiated cancer of no specific site (n = 1).
The results of peritoneal biopsy were used to guide management. Eight women had previous primary tumors that are known to metastasize to the peritoneum. Two patients with prior breast cancer and one with prior colon cancer had recurrent disease, which was confirmed by means of comparison with previous surgical specimens. These women were treated with appropriate chemotherapy. Four patients with prior breast cancer were found to have new ovarian cancers, and they underwent confirmatory debulking surgery followed by platinum-based chemotherapy. (Details of the patient with previous mixed müllerian endometrial cancer were described earlier in Results.)
The other 27 patients had no previous malignancy and were initially treated nonsurgically according to their biopsy results. Twenty-five patients had core biopsy findings that were consistent with ovarian cancer and received neoadjuvant platinum-based chemotherapy. In addition to the seven patients whose diagnosis was suggested on the basis of immunohistochemical findings, 18 had confident diagnoses of ovarian cancer that were based on hematoxylin-eosin findings. The histologic diagnosis was confirmed in 11 of these patients who have since undergone surgery. The single patient whose peritoneal biopsy findings revealed lymphoma received specific chemotherapy and the patient whose peritoneal biopsy findings revealed metastases from an undiagnosed colon cancer received supportive therapy.
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODS RESULTSDISCUSSIONREFERENCES |
|---|
In our study, a site-specific diagnosis was achieved in 27 (77%) of 35 women by analyzing hematoxylin-eosin–stained paraffin sections. Needle biopsy of the omentum has been previously described with both US (7) and CT (8). However, in the study of the latter (a 7-year retrospective study of patients with a variety of peritoneal diseases including metastatic adenocarcinoma), a primary site was not diagnosed in 10 (62%) of the 16 cases (8). No details of the pathologic analysis were provided. In another series of percutaneous nonvisceral abdominal biopsy (9), treatment was based on the result in 43 (83%) of 52 patients. In that study, ovarian and colonic adenocarcinoma accounted for only 11 (21%) of the 52 diagnoses, and lymphoma was the most common diagnosis. The selection process in that retrospective study was likely responsible for the findings that differed from those of our study, which had clearly defined prospective entry criteria and which was designed to address a specific and common management problem that surgical and nonsurgical gynecologic oncologists face. Nonetheless, the findings from their series and others (7–9) emphasize the ability of image-guided biopsy to aid in the diagnosis of a range of diseases that is wider than that encountered in our study.
The safety of needle biopsy is now well established (10–12). Core biopsy has been used principally in visceral and lymph node sites in the abdomen. To our knowledge, our prospective series is the largest yet in which image-guided core biopsy of omental and peritoneal masses is described. There were no complications. We obtained biopsy samples in the omental cake adherent to large and small bowel without problems. Recently, percutaneous image-guided bowel biopsy with 18-gauge core needles has been reported (13) to be a safe procedure. In our study, adnexal masses were subject to biopsy in three women with prior breast cancer: two with solid lesions and one with multicystic lesions. Although theoretical risks of cyst puncture exist, these risks probably are of limited importance in a woman with peritoneal carcinomatosis for whom chemotherapy is planned. Biopsy of adnexal masses remains controversial. A further theoretical risk is tracking of the tumor into the abdominal wall. To our knowledge, no published reports relating this to needle biopsy of ovarian cancer, and we have encountered no clinically apparent metastases in the follow-up of our 35 women. However, cutaneous metastases from ovarian cancer have been reported in laparoscopy scars (14). These may not affect overall prognosis but may produce clinically important morbidity if not surgically removed.
It is difficult to evaluate the importance of a negative image-guided biopsy. In the two patients in our series who had biopsy findings that were negative for malignancy, the decision to perform further biopsy was subject to multidisciplinary review. In one, the clinical, radiologic, and biochemical findings overwhelmingly indicated a diagnosis of disseminated ovarian cancer, and findings from repeat biopsy confirmed this. In the other, a resolving inflammatory process seemed likely, and the underlying colon cancer involving the ileocecal region and appendix was manifest only months later. It is speculated that this tumor resulted in appendicitis, because the appendix was involved at subsequent surgery.
Sampling error remains a problem for core biopsy, although it provides a greater amount of cellular material than does a cytologic preparation. There are arguments for increasing the number of cores obtained beyond the two that we routinely obtain and for sampling different sites within a mass (Fig 2). Comparison of the core biopsy findings with the histologic findings in previous surgical specimens can be helpful in patients with a cancer history. Core biopsy resulted in a firm histologic diagnosis after hematoxylin-eosin analysis in seven (88%) of the eight patients with cancer histories in our study.
Immunohistochemical techniques were valuable for the assessment of cancer in seven women whose core biopsy findings showed poorly differentiated adenocarcinoma on hematoxylin-eosin–stained sections, and they provided useful confirmatory results in other patients. All seven core biopsies had the same immunohistochemical marker profile (CA-125 positive, cytokeratin 7 positive, carcinoembryonic antigen negative, cytokeratin 20 negative), which suggested a müllerian duct origin. Clearly, in the study population, there was a high prior probability of ovarian cancer that accounts for this predominance. Further work is required to confirm that the technique can be used to reliably diagnose the whole range of tumors that can result in peritoneal carcinomatosis and to confirm that immunohistochemical techniques can be reliably used to distinguish between poorly differentiated ovarian and colon cancers. Immunohistochemical techniques have been disappointing when applied in the analysis of cytologic preparations from both effusions (3,15) and abdominal and retroperitoneal masses (16). However, in one study in which the technique was applied to cell blocks made from centrifuged fluids that were treated as histologic specimens after they were mounted on paraffin blocks, the results were more promising (17). With this approach, a confident diagnosis was made in 55% of patients with a history of prior carcinoma but only 29% with no such cancer history (17). These results are inferior to ours with core biopsy.
Image-guided peritoneal biopsy was a safe and accurate technique that provided sufficient tissue to identify the primary site and guide treatment in most women with peritoneal carcinomatosis, without recourse to laparoscopy or laparotomy. It also limited the need for investigation to exclude primary tumors at other sites. The technique was designed to provide a firm histologic diagnosis prior to neoadjuvant chemotherapy for ovarian cancer, but it has wide applications in the surgical and nonsurgical management of abdominopelvic malignancy. This approach should be used in the context of multidisciplinary management following review of the presenting clinical and biochemical features and following initial diagnostic imaging to optimize patient selection prior to the procedure. Further work is required to confirm the utility of immunohistochemical techniques in determining the primary tumor site in patients with disseminated and poorly differentiated adenocarcinoma.
|
ACKNOWLEDGMENTS |
|---|
|
FOOTNOTES |
|---|
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODS RESULTSDISCUSSIONREFERENCES |
|---|
This article has been cited by other articles:
|
|
 |
|
  J A Spencer A multidisciplinary approach to ovarian cancer at diagnosis Br. J. Radiol., October 1, 2005; 78(Special_Issue_2): S94 - S102. [Full Text] [PDF]
|
|
|
|
|
|
M Horger, M Muller-Schimpfle, I Yirkin, M Wehrmann, and C D Claussen Extensive peritoneal and omental lymphomatosis with raised CA 125 mimicking carcinomatosis: CT and intraoperative findings Br. J. Radiol., January 1, 2004; 77(913): 71 - 73. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 L. M. Ho, J. Thomas, S. A. Fine, and E. K. Paulson Usefulness of Sonographic Guidance During Percutaneous Biopsy of Mesenteric Masses Am. J. Roentgenol., June 1, 2003; 180(6): 1563 - 1566. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 G. I. Spence, J. A. Spencer, and S. E. Swift Core Biopsy in Men with Peritoneal Carcinomatosis * Drs Spencer and Swift respond: Radiology, July 1, 2002; 224(1): 294 - 294. [Full Text]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| RADIOLOGY | RADIOGRAPHICS | RSNA JOURNALS ONLINE |
