Q J Med 2000; 93: 611-615
© 2000 Association of Physicians
Bone-marrow micrometastases in patients with brain metastases from epithelial cell tumours
From the Cork Cancer Research Centre, Department of Surgery, Mercy Hospital and 1 the Departments of Medicine and Surgery, University College Cork, National University of Ireland, Cork
 |
Summary |
|---|
 Top Summary IntroductionMethodsResultsDiscussionReferences |
|---|
Carcinoma that has metastasized to the central nervous system (CNS) poses a particular clinical problem regarding confirmation of the diagnosis and subsequent management. Prior to excision, thorough evaluation for coexisting systemic disease is essential, but current imaging techniques are limited by their spatial resolution and under-stage many patients. We evaluated the potential utility of bone-marrow evaluation for micrometastatic cells in patients with CNS metastasis. Bone-marrow aspirates were examined for cytokeratin-positive cells in 12 consecutive patients who presented with symptomatic space-occupying lesions of the CNS. These patients had previously undergone surgical excision of either gastrointestinal or breast cancers. All twelve had micrometastases in their bone marrow at the time of presentation with the CNS disease and all had a fatal outcome within 13 months. In nine of the 12 patients, bone-marrow micrometastases were the only evidence for systemic spread. Three patients had elevated serum tumour markers and two of these had radiologically detectable recurrence elsewhere. Bone-marrow micrometastases indicate concurrent systemic involvement and a poor prognosis. The results suggest that bone-marrow evaluation for systemic spread is a useful diagnostic adjunct and should be performed before considering diagnostic biopsy or excision.
|
Introduction |
|---|
|
TopSummaryIntroductionMethodsResultsDiscussionReferences |
|---|
The diagnosis and management of space-occupying lesions in the central nervous system may be difficult in patients who have previous undergone surgical resection of a cancer.1 Confirmation of the metastatic nature of a solitary space-occupying lesion frequently requires stereotactic biopsy or surgical excision, with the potential for significant morbidity.2,3 Conversely, in some patients cerebral metastases are found unexpectedly where a primary tumour was unsuspected pre-operatively. These account for 15% of patients who present with cerebral metastasis.4,5 In both situations, patient prognosis is poor and the clinical challenge is to confirm the systemic/metastatic nature of the disease.1 Currently available imaging techniques have limited sensitivity and frequently under-stage tumour spread, compounding the clinical difficulty.6–8
This has prompted the assessment of more sensitive techniques to identify occult disease. Examination of haemopoietic bone marrow for micrometastatic cells provides an accurate and reproducible estimation of tumour burden in patients with normal tumour marker studies and no metastatic disease by conventional radiology.9–15 It provides a convenient window on the metastatic process, because of its mesenchymal nature, large cellular repository and rich blood supply. Epithelial metastases stain positive for cytokeratin, in contrast to bone marrow which does not constitutively express this marker.9–15 Cytokeratin-positive epithelial cells are easily detectable with a sensitivity of one epithelial cell in 1x105 marrow cells with flow cytometry.11 Bone-marrow evaluation, therefore, is a useful means of demonstrating systemic epithelial cancer which may not be detectable by conventional radiology. Similarly, bone-marrow evaluation is potentially useful in those patients with undiagnosed primary cancers. In the latter group, most cerebral metastases are adenocarcinomas4 and almost all demonstrate immunoreactivity to cytokeratin antibodies.16
We examined the potential utility of bone-marrow analysis for micrometastases in a consecutive series of patients with tumour recurrence in the central nervous system.
|
Methods |
|---|
|
TopSummaryIntroductionMethodsResultsDiscussionReferences |
|---|
Patients
In July 1993, a prospective study was initiated to assess occult bone-marrow micrometastases in patients with epithelial malignancy. All patients diagnosed with cancer had bone marrow analysed prior to and at regular intervals following surgery, and those who returned for review (who had surgery prior to 1993) were similarly analysed. Of these, a total of twelve patients who had surgery for carcinoma presented with central nervous system symptoms. Ten of these had computerized tomography of brain that confirmed space-occupying lesions (Figure 1
). The remaining two patients (11 and 12) had metastatic cells in cerebrospinal fluid, but with negative imaging. These 12 patients are the subjects of this report (Table 1).
|
|
Ten had M0 disease at the time of surgery (negative ultrasound/computerized tomography of liver, negative computerized tomography of thorax in the oesophageal group, negative laparotomy for those with GIT cancers and negative resection margins). Two further patients had liver metastases confirmed at surgery (patients 6 and 9; not detected by pre-operative imaging). Patient 6 had a solitary hepatic metastasis resected intra-operatively. Patient 2 had a single cervical lymph node containing metastatic tumour removed 48 months following initial surgery. These latter three patients were treated with 5-fluorouracil and folinic acid, and were considered clinically and radiologically to be in disease remission up to the time of presentation with their central nervous system symptoms.
Methods
At the time of presentation with central nervous system symptoms/signs, all patients were re-evaluated for disease recurrence. All had tumour markers (CEA and Ca19.9) chest radiology, ultrasound of liver, computerized tomography of abdomen and thorax (oesophageal cancers), isotope bone scan (breast cancer only) and bilateral iliac crest bone-marrow sampling.
Bone-marrow evaluation
After incising the overlying skin, to prevent contamination by skin epithelia, samples of bone marrow were taken from both iliac crests and pooled. An aliquot of marrow was fixed by addition drop-wise, with gentle shaking, into 70% ethanol, and stored at 20°C until it was used for study. A separate aliquot was used from which to separate mononuclear cells by Ficoll-Hypaque density centrifugation.
The flow cytometric technique has been previously described and validated.11,12 Bone marrow samples were washed three times in phosphate-buffered saline and stained with fluorescein isothiocyanate-conjugated monoclonal antibody to cytokeratin 18 (Sigma). The antibody was diluted 1/20 in phosphate-buffered saline and 10 µl of this stock antibody was added to 100 µl of sample. After washing in phosphate-buffered saline, propidium iodide containing ribonuclease was used to stain for deoxyribonucleic acid content. Cells were then analysed (105 events) with a cell analysis program with an Epics Coulter Elite flow cytometer. Only double-staining cells were counted as positive, thereby excluding non-specific staining and spuriously stained debris. Positive (HT29 and OC1 cells) and negative (normal bone marrow) controls were included in all analyses.11 Threshold and background fluorescence were controlled for using an irrelevant fluorescein isothiocyanate-conjugated monoclonal antibody of the same species and isotype as anti-cytokeratin-18.
The studies were confirmed using histochemistry with cytokeratin-18 staining using the alkaline phosphate anti-alkaline-phosphatase (APAP) technique.10 Approximately 0.5x106 cells were examined for each patient. With this technique, cells positive for cytokeratin stain red (Figure 1
).
|
Results |
|---|
|
TopSummaryIntroductionMethodsResultsDiscussionReferences |
|---|
At the time of presentation with central nervous system recurrence, patients were again evaluated clinically, re-staged with serum tumour markers (carcino-embryonic-antigen), liver function studies, chest radiograph, ultrasound of liver and computerized tomography of thorax and abdomen and bone marrow analysis. All twelve patients (100%) were found to have micrometastases in their bone marrow by flow cytometry and histochemistry (control data from our laboratory indicates that false-positive results occur in 2/251; 0.8%). In 9/12 patients (75%), this was the only evidence of systemic disease. Two patients (17%) had radiologically imageable disease (both had elevated serum tumour markers). One further patient had elevated tumour markers with no radiological evidence of disease. Patients 3,4 had solitary metastases, suitable for surgical excision. One refused surgery and the other succumbed prior to undergoing excision. Patients 5 and 9 were unsuitable for surgery because of their overt abdominal metastases. Patients 1, 2, 6–8 and 10 were unsuitable for excisional surgery, either because of the site of the lesion, or because of multiplicity of lesions. Patient 11 presented 6 years after mastectomy and axillary level III dissection, with severe interscapular and right upper quadrant pain. Plain X-ray of thoracic spine, isotope bone scan, myelography, CT and MR were repeatedly normal. After a delay of 1 year, bone-marrow and CSF analysis demonstrated cytokeratin-positive cells. She was treated with intrathecal methotrexate, thoracic spine irradiation and systemic chemotherapy with an excellent symptomatic response. One year later she died with gross hepatic metastatic disease.
Of the 12 patients, 10 succumbed to their disease, prior to deterioration of their central nervous system symptoms.
|
Discussion |
|---|
|
TopSummaryIntroductionMethodsResultsDiscussionReferences |
|---|
Systemic recurrence of epithelial tumours is distressing for patients, particularly if there is central nervous system involvement. The treatment of patients with central nervous system recurrence of primary epithelial cancers depends on accurate diagnosis and clinico-radiological staging.1 Histological confirmation of the diagnosis requires stereotactic or excision biopsy. Although stereotactic biopsy is safer than surgical biopsy, both procedures may cause significant morbidity and occasionally mortality.2,3 In the case of patients with overt metastatic disease, biopsy is not necessary, as prognosis for these patients is limited and they will not be offered resection.
The situation is less clear for patients with latent metastases. Several large series of patients with occult deposits of metastatic cells within bone marrow have shown that early recurrence (overt disease) occurs in those with breast, lung and gastrointestinal cancers.9–15 It appears reasonable to suggest the finding tumour cells in the marrow of patients with a suspected central nervous system metastasis provides compelling circumstantial evidence as to the nature of the central nervous system deposit. In all 12 patients described here, bone marrows were positive and a poor outcome was observed. It is unlikely that it is purely coincidental that all patients had a rapidly fatal outcome. We suggest that bone-marrow analysis should be performed in all patients and that if positive, it be considered as suggestive that the central nervous system lesion is a metastasis.
The decision to explore with a view to resection of an intracranial metastatic deposit is based on evidence for a resectable solitary lesion, a benefit of palliative excision over surgical morbidity and a potential for worthwhile symptom-free interval.1 Best palliative results are achievable where there is a protracted disease-free interval after treatment of the primary cancer and where overt systemic recurrence is excluded.1,17,18 Patient evaluation for systemic disease by clinical and conventional imaging criteria is currently sub-optimal., and for many patients excision of cerebral metastases is of little if any benefit because of a short interval to overt recurrence.1 Assessment for metastases is currently limited by spatial resolution of imaging6–8 and the restricted sensitivity and specificity of tumour markers.19–21 We have previously reported that in patients with foregut cancers, more than 30% of intraperitoneal and hepatic metastases are undetectable by conventional imaging.8 Taken together, these indicate that patients may undergo non-beneficial surgical excision in the presence of undiagnosed recurrence. This implies that for optimal management, more accurate assessment methods must be developed. Currently, the most sensitive means of examining for minimal residual disease is bone-marrow analysis. Because analysis of bone marrow for cytokeratin-positive cells has been validated as an indicator of minimal residual disease, we suggest that this should be taken into account when deciding to resect an intracranial deposit.10
Of particular interest in this report, are two patients who had spinal root symptoms due to metastatic involvement, which initially defied diagnosis by repeated clinical and radiographic evaluation. Even though clinically suspected, the first of these patients did not have the marrow evaluation until there was objective clinical evidence of root involvement and subsequent analysis of CSF confirmed the presence if intrathecal metastatic cells. In the other patient the finding of positive bone marrow prompted the investigation of the CSF. It is likely that earlier and repeated use of bone-marrow analysis in these patients would have facilitated diagnosis and institution of appropriate treatment.
While all patients with CNS metastases, in this study, had positive bone-marrow studies, false negative results do occur. These may be due to sub-optimal marrow sampling and or the presence of poorly differentiated epithelial metastases which do not express cytokeratins. Optimal studies require sampling and pooling of mesenchymal marrow (as distinct from regions populated with fat cells or venous sinuses) from multiple sites (both iliac crests), density centrifugation to concentrate cells. In essence, positive predictive value is good (false positives 0.8%) but the value of a negative result has yet to be defined.10,11
We propose that immunocytochemical examination has application in assessing patients with central nervous system metastasis. It helps confirm the diagnosis of epithelial cancer in those patients with an occult primary at the time of presentation, and improves the sensitivity of tumour burden assessment in patients with known primary disease. It also provides compelling evidence as to the nature of a suspicious lesion, thereby obviating the need for biopsy. All patients with suspected central nervous system deposits should have bone-marrow analysis performed, such that we can amass a body of data that will ultimately help in clinical decision-making.
|
Notes |
|---|
Address correspondence to Professor F. Shanahan, Department of Medicine, Cork University Hospital, Cork, Ireland. e-mail: fshanahan{at}ucc.ie
|
References |
|---|
|
TopSummaryIntroductionMethodsResultsDiscussionReferences |
|---|
1. Vecht CJ. Clinical management of brain metastasis. J Neurol1998; 245:127–31.[Web of Science][Medline]
2. Hall WA. The safety and efficacy of stereotactic biopsy for intracranial lesions. Cancer1998; 82:1749–55.[Web of Science][Medline]
3. Swain PD, Hitchon PW, Follett KA, Torner JC. Computed imaging-assisted stereotactic brain biopsy: a risk analysis of 225 consecutive cases. Surg Neurol1998; 49:640–9.[Web of Science][Medline]
4. Nguyen LN, Maor MH, Oswald MJ. Brain metastases as the only manifestation of an undetected primary tumor. Cancer1998; 83:2181–4.[Web of Science][Medline]
5. Khansur T, Routh A, Hickman B. Brain metastases from unknown primary site. J Miss State Med Assoc1997; 38:238–42.[Medline]
6.
Nelson RC, Chezmar JL, Hoel MJ, Buck DR, Sugarbaker PH. Peritoneal carcinomatosis: preoperative CT with intraperitoneal contrast material. Radiology1992; 182:133–8.
7. Soyer P, Levesque M, Elias D, Zeitoun G, Roche A. Preoperative assessment of resectability of hepatic metastases from colonic carcinoma: CT portography vs. sonography and dynamic CT. Am J Roentgenol 1992; 159:741–4.
8. O'Brien MG, Fitzgerald EF, Lee G, Crowley M, Shanahan F, O'Sullivan G. A prospective comparison of laparoscopy and imaging in the staging of esophagogastric cancer before surgery. Am J Gastroenterol1995; 90:2191–4.[Web of Science][Medline]
9. Lindemann F, Schlimok G, Dirschedel P, Witte J, Riethmuller G. Prognostic significance of micrometastatic tumour cells in bone marrow of colorectal cancer patients. Lancet1992; 340:685–9.[Web of Science][Medline]
10.
O'Sullivan G, Collins JK, Kelly J, Morgan J, Madden M, Shanahan F. Micrometastases: marker of metastatic potential or evidence of residual disease. Gut1997; 40:512–15.
11. O'Sullivan GC, Collins JK, O'Brien F, Crowley B, Murphy K, Lee G, Shanahan F. Micrometastases in bone marrow of patients undergoing curative surgery for gastrointestinal cancer. Gastroenterology1995; 109:1535–40.[Web of Science][Medline]
12. Mansi JL, Berger U, Easton D. Micrometastases in bone marrow patients with primary breast cancer: Evaluation as an early predictor of bone metastases. Br Med J1987; 295:1093–6.
13. Cote RJ, Beattie EJ, Chaiwun B, Shi SR, Harvey J, Chen SC, Sherrod AE, Groshen S, Taylor CR. Detection of occult bone marrow micrometastases in patients with operable lung carcinoma. Ann Surg1995; 222:415–23.[Web of Science][Medline]
14. Lindemann F, Schlimok G, Dirschedel P, Witte J, Riethmuller G. Prognostic significance of micrometastatic tumour cells in bone marrow of colorectal cancer patients. Lancet1992; 340:685–9.
15. Schlimok G, Funke I, Pantel K, Strobel F, Lindemann F, Witte J, Reithmuller G. Micrometastatic tumour cells in bone marrow of patients with gastric cancer: methodological aspects of detection and prognostic significance. Eur J Cancer1991; 27:1461–5.
16. Perry A, Parisi JE, Kurtin PJ. Metastatic adenocarcinoma to the brain: an immunohistochemical approach. Hum Pathol1997; 28:938–43.[Web of Science][Medline]
17. Wronski M, Arbit E, Burt M, Galicichi JH. Survival after surgical treatment of brain metastases from lung cancer: a follow-up study of 231 patients treated between 1976 and 1991. J Neurosurgery1995; 83:605–16.[Web of Science][Medline]
18. Karakousis CP. Surgical treatment of malignant melanoma. Surg Clin N Am1996; 76:1299–312.
19. Martin EW Jr, Cooperman M, Carey LC, Minton JP. Sixty second look procedures indicated by rise in serial carcinoembryonic antigen. J Surg Res1980; 28:389–94.[Web of Science][Medline]
20. Martin EW Jr, Minton JP, Carey LC. Second look surgery in the asymptomatic patient after primary resection of colorectal cancer. Ann Surg1985; 202:310–17.[Web of Science][Medline]
21. Wanebo HJ, Llaneras M, Martin T. Prospective monitoring trial for carcinoma of colon and rectum after surgical resection. Surg Gynecol Obstet1989; 169:479–85.[Web of Science][Medline]
CiteULike 
Connotea 
Del.icio.us What's this?
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||