|
Advertisement |
|
|
||
 |
|
|||||
|
|||||
|
 |
||||||
|
||||||
|
||||||
|
© 2001 American Society for Clinical Oncology
Is Immunohistochemical Analysis an Appropriate Diagnostic Technique for Bone Marrow Micrometastases?Cork University, South Infirmary Hospitals, Cork, Ireland To the Editor:Braun et al published an article entitled "Comparative Analysis of Micrometastasis to the Bone Marrow and Lymph Nodes of Node-Negative Breast Cancer Patients Receiving No Adjuvant Therapy" in the March 1, 2001, issue (J Clin Oncol 19:1468-1475, 2001).1 However, certain issues, among them the ongoing technical difficulties with immunohistochemical (ICC) detection, need to be addressed. ICC will define tissue architecture but is otherwise unreliable. It does not confer any added advantage with cytology specimens distorted during the centrifuge process. The technical difficulties encountered are many: it is time-consuming and poorly quantitative, restricts cell number analysis, encounters a high degree of observer bias, and prevents retrieval of viable tumor cells, imperative for subsequent analysis. One should interpret results from the literature with caution, as even criteria for what is considered to be positive vary between centers. For example, the status of partial or incomplete immunostaining of cells or the presence of intracytoplasmic granules is not fully defined in the methodology sections of most reputable units. Despite standardization of the technique and identification of the optimal dilution of primary and secondary antibodies, results will vary from one setting to the next. A comparative meta-analysis of results from different centers on this basis cannot be recommended. A number of issues need to be addressed in future studies. A control is imperative in all aspects of this research. With patient selection, a control subset of individuals with benign breast disease is required. The revised dual-staining technique described is decidedly different from that of single-staining protocols. Although Braun et al refer the reader to previous research performed at their center, the presumption of applying previous validation efforts using single staining as a control is worthy of critique.2 On a technical level, a control of the primary anti-human pancytokeratin antibody (which unfavorably varies between bone marrow samples [A45-B/B3] and lymph node tissue [NCL-5D3]) requires both a negative control, ie, omission of primary antibody or preferably the use of a control immunoglobulin (Ig) G of similar epitope, and a positive control, ie, a breast tumor cell line. Yet no tissue negative control has been applied that may be worthwhile. Among early-stage carcinoma patients positive for epithelial cells in bone marrow, the average tumor cell retrieval after ICC of 1 to 2 million cells of an isolated leukocyte fraction is approximately one to three cells. If such a decision is to rely on only one positive event, this emphasizes the importance of careful characterization of antibodies applied and awareness of the possibility of false-positive results. Although it is appreciated that in a previous series, only 1% of patients with benign breast disease presented with a false-positive bone marrow aspirate, such a result could lead to clinical intervention.2 Although alkaline phosphatase-anti-alkaline phosphatase is used primarily as a color substrate reaction, controversy exists about the false staining of the Ig kappa and lambda subunits of plasma cells, making interpretation difficult.3 The concept of double labeling or dual staining is appealing. Yet, anti-CD45 has been shown to identify only a minority of the A45-B/B3-AP–positive hematopoietic cells (20%).3 The diaminobenzidine reagent that relies on complete blockage of endogenous peroxidase activity creates an unacceptable level of background staining that makes the "lack" of dual staining difficult to appreciate. Also, simple morphologic features of the cell used as a form of objective malignancy grading can be difficult to visualize. Braun et al state that dual staining is performed to eliminate the possibility of registering cytokeratin-positive deposits in phagocytosing cells, a novel concept. Is this a true false-positive result, or should these cells be considered? This question addresses the significance of presence versus viability of cells. Also, where is the initial site of phagocytosis? Cells ingested at the primary tumor site may not be of relevance, but cells ingested during transit in nodal or hematogenous tissue are of indeterminate significance. They do not mention false-positive straining of the kappa units of plasma cells as a possible reason for dual staining of cells, which would be the main concern. With the variability in bone marrow aspirate volume of 3.5 to 9.5 mL, is it possible to omit the bias of technique-related dilution with peripheral-blood leukocytes? Only 2 x 106 cells per patient were analyzed despite much higher retrievals, and such numbers are small. Only 29% of bone marrow samples were positive, which is low. This may be explained by Braun et al’s sampling from only two sites at a single time point with analysis of as few as 2 million cells. They do not quantify positive cell number or tumor yield. A median of 10 lymph nodes from level I were analyzed for each patient. Sentinel lymph node sampling was not encountered. Four adjacent sections (as opposed to serial sections) were retrieved. A tumor cell within the body of the lymph node was defined as a metastasis. Although subcapsular, sinusoidal, and vascular tumor cells within the parenchyma of the lymph node were identified and are worth mentioning, results from this subcategory were not clarified at a later stage. Of the 12 distant metastases that developed, visceral sites occurred in only three patients. Multiple visceral sites with skeletal involvement occurred in nine patients. This raises the question of micro- versus macrodeposits. Do circulating tumor cells really exist, or are these subclinical, distant, end-organ deposits attached to venous sinusoids in the bone marrow circulation? It is difficult to clarify such a proposal because rib bone sections initially require fixation and decalcification before they can be embedded and cut for histologic analysis. This can significantly affect the immunostaining quality of cytokeratin-positive cells, making interpretation difficult. When processing such small numbers, can detection only be achieved at a tumor burden that will correlate with bone marrow macrometastases? It would explain the significant association between positive bone marrow status and skeletal metastases at a later date.1,4 Finally, to demote lymph node involvement as a prognostic factor, Braun et al reference several articles.5,6 Ragaz et al5 found a 40% 10-year survival rate in lymph node–positive breast cancer patients without any recurrence. Yet the notable bias of adjuvant therapy in this cohort of patients is never referred to. Thirty percent of patients with lymph node–negative disease will return with distant metastasis. Yet De Vita published this article in the New England Journal of Medicine in 1989, when the facilities for serial sectioning and immunohistochemical detection were not yet available.6 Up to 15% of patients have had their regional status upstaged with the evolution of newer technologies. Such results should be interpreted with caution. Lymph node status should not be undermined as a prognostic factor simply to validate emerging new protocols as a form of pathologic staging. The answer to micrometastasis is not simply detection, but isolation, culture, determination of genotype and phenotype, and correlation with in vitro (pharmacologic sensitivity profiling) and in vivo behavior to determine variables of predictive and prognostic significance. The "usefulness" of ICC is therefore debatable. REFERENCES
1.
Braun S, Cevatli BS, Assemi C, et al: Comparative analysis of micrometastasis to the bone marrow and lymph nodes of node-negative breast cancer patients receiving no adjuvant therapy. J Clin Oncol 19: 1468-1475, 2001
2.
Braun S, Pantel K, Muller P, et al: Cytokeratin-positive cells in the bone marrow and survival of patients with stage I, II, or III breast cancer. N Engl J Med 342: 525-533, 2000 3. Borgen E, Beiske K, Trachsel S, et al: Immunocytochemical detection of isolated epithelial cells in bone marrow: Non-specific staining and contribution by plasma cells directly reactive to alkaline phosphatase. J Pathol 185: 427-434, 1998[Medline]
4.
O’Sullivan GC, Collins JK, Kelly J, et al: Micrometastases: Marker of metastatic potential or evidence of residual disease? Gut 40: 512-515, 1997
5.
Ragaz J, Jackson SM, Le N, et al: Adjuvant radiotherapy and chemotherapy in node-positive premenopausal women with breast cancer. N Engl J Med 337: 956-962, 1997 6. De Vita VT Jr: Breast cancer therapy: Exercising all options. N Engl J Med 320: 527-529, 1989 (editorial)[Medline]
Response
Technical University, Munich, Germany In Reply:Established prognostic markers do not enable identification of patients with occult tumor cell dissemination among primary breast cancer patients who have been treated by complete tumor resection (stage R0). Even among so-called low-risk patients, such as node-negative breast cancer patients, 30% or even more will experience recurrence of incurable metastatic disease within 5 years after diagnosis. This present diagnostic dilemma motivated clinical evaluation of several approaches to identify patients with early tumor-cell dissemination, including immunocytochemical screening for occult metastatic cells in bone marrow and extended pathologic examination of lymph nodes using immunochemistry. In a passionate response to our study,1 Redmond et al raised several questions about methodologic details that could confer problems in interpreting immunocytochemical and immunohistochemical data. We agree with Redmond et al that immunocytochemistry and immunohistochemistry are time consuming and thus limited to analysis of a certain amount of cells, unless automated imaging analysis is used as it is now available.2 We further agree that any diagnostic approach that intends to draw clinical conclusions from the finding of as little as a single tumor cell against the background of 2 x 106 autochthonous bone marrow cells requires extreme precautions and care with regard to sensitivity, specificity, reliability, and validity as well as interpretation. Over the past 10 years, our careful investigations of these confounding variables showed that the antibody A45-B/B3 directed against the cytokeratin (CK) heterodimers CK8/18 and CK8/19 as well as a common CK epitope is a specific and sensitive tool to identify such micrometastatic tumor cell spread.3,4 In view of the well-known clinical evidence that approximately one third of node-negative breast cancer patients relapse at distant sites within 5 years after diagnosis, we believe that a micrometastasis prevalence of 29% in our series cannot be called too "low." The outlined immunoassay yielded a similar prevalence of tumor cells in pre- and postoperative bone marrow samples, thus indicating a high reproducibility of results.4 All studies published by our co-workers were based on data resulting from the use of an appropriate assay system, which included appropriate positive and negative control stainings, as stated explicitly in each published report so far. Redmond et al further questioned whether analysis of 2 x 106 bone marrow cells per patient as well as of few (in our series, four) adjacent lymph node sections is sufficient to eliminate concerns about a sampling error. Of course, our approach still carries the danger of a certain sampling error. Yet, there are convincing arguments in favor of the clinical relevance of our analytical approach, since the analysis of 2 x 106 bone marrow cells is obviously enough to provide clinically relevant data on prognosis of early-stage breast cancer patients.1,3 Furthermore, clinical routine diagnosis with its limited capacities also limits itself to analyzing only two histologic sections per axillary lymph node. Even though the sensitivity of reverse transcriptase polymerase chain reaction (RT-PCR) techniques for detection of single tumor cells is thought to considerably surpass that of immunocytochemistry, the RT-PCR sensitivities reported so far have been unexpectedly low due to inhomogeneity or even lack of expression of the respective tumor-associated mRNA used.5 In addition, epithelial tissue–specific mRNAs, such as CK 18, CK 19, and epithelial cell adhesion molecule, as well as tumor-associated mRNAs, such as HER2, carcinoembryonic antigen, and MUC-1, were found to be expressed in mesenchymal tissues, such as blood, bone marrow, and lymph nodes.6 Given observer expertise, which is an obvious prerequisite for every cytologic approach that is intended for clinical use, only a few staining results remain ultimately undeterminable,3,4 as is the case for every analytical approach that involves the human eye and expertise for evaluation, including Papanicolaou smears of the cervix or histopathologic grading. The robustness of the immunocytochemical approach to identifying hematogenous tumor cell spread is endorsed by the unanimous prognostic relevance shown by those recent studies that used validated immunoassays with the anti-CK antibodies. Because the overall number of patients analyzed for occult metastatic cells is still relatively small, a formal meta-analysis is not possible at present. Thus, we acknowledge the need for a multicenter prospective trial in order to obtain level I evidence for the prognostic value. (At present, there are at least three studies underway in the United States and Europe to undertake this task. Because of the limited data available so far, we added a caveat as to overinterpretation of our data in all of our reports.) Another concern raised by Redmond et al is the prognostic impact of cells that are detected but nonviable as opposed to those viable cells that are indeed capable of initiating metastases. This concern involves questions about tumor-biologic features, such as phenotype and genotype of CK-positive cells. Phenotyping and genotyping of single tumor cells at the single-cell level is not trivial from a technical point of view. So far, only limited—but nevertheless quite intriguing—data are available. CK-positive cells in bone marrow carry characteristics exclusively known for tumor cells, such as HER2 amplification and protein overexpression,7,8 as well as genomic changes that are identical to those of the respective primary tumor.9 We recently showed for the first time that HER2 expression on bone marrow micrometastases is associated with poor patient prognosis in breast cancer, independent of established prognostic factors.7 Hosch et al10 were able to demonstrate the tumorigenicity and metastatic potential of a cell line established from an immunohistochemically positive lymph node in severe combined immunodeficiency mice. In our opinion, it is not the limited information on biologic characteristics of bone marrow micrometastases that needs to be debated but the lack of specific therapeutic opportunities that exist for patients with early-stage breast cancer and micrometastatic disease. This lack of a specifically targeted therapy is true not only for the presence of bone marrow micrometastases but also for almost all other prognostic factors in primary breast cancer, except for expression of estrogen receptor and HER2 oncoprotein. In contrast to all available prognostic markers, bone marrow micrometastases can be utilized for patient monitoring during adjuvant treatment, both for conventional systemic therapies as well as for novel tumor-biologic therapies.11 Phenotyping of bone marrow micrometastases for potential therapeutic targets may even offer the unique opportunity for a more direct detection of the actual adjuvant therapy targets than analysis of primary tumor tissue. Thus, immunocytochemical analysis of bone marrow aspirates can be used to stratify patients in adjuvant therapy trials, to evaluate specifically targeted treatment strategies (eg, trastuzumab), as well as to monitor therapeutic efficacy and relieve both patients and clinicians from waiting for 5 years or more before the efficacy of adjuvant therapy can be determined. In view of the evidence discussed above, it is our conviction that immunocytochemical identification of bone marrow micrometastases offers promising opportunities for the future management of patients with breast cancer and other solid tumors. REFERENCES 1. Braun S, Cevatli BS, Assemi C, et al: Comparative analysis of micrometastasis to the bone marrow and lymph nodes of node-negative breast cancer patients receiving no adjuvant therapy. J Clin Oncol 19: 1468-1475, 2001
2.
Bauer KD, de la Torre-Bueno J, Diel IJ, et al: Reliable and sensitive analysis of occult bone marrow metastases using automated cellular imaging. Clin Cancer Res 6: 3552-3559, 2000 3. Braun S, Pantel K, Müller P, et al: Cytokeratin-positive cells in the bone marrow and survival of patients with stage I, II, or III breast cancer. N Engl J Med 342: 525-533, 2000 4. Pantel K, Schlimok G, Angstwurm M, et al: Methodological analysis of immunocytochemical screening for disseminated epithelial tumor cells in bone marrow. J Hematother 3: 165-173, 1994[Medline]
5.
Mapara MY, Körner IJ, Hildebrandt M, et al: Monitoring of tumor cell purging after highly efficient immunomagnetic selection of CD34 cells from leukapheresis products in breast cancer patients: Comparison of immunocytochemical tumor cell staining and reverse transcriptase-polymerase chain reaction. Blood 89: 337-344, 1997
6.
Zippelius A, Kufer P, Honold G, et al: Limitations of reverse transcriptase-polymerase chain reaction for detection of micrometastatic epithelial cancer cells in bone marrow. J Clin Oncol 15: 2701-2708, 1997
7.
Braun S, Jeumos I, Schlimok G, et al: ErbB2 over-expression on occult metastatic cells in bone marrow predicts poor clinical outcome of stage I-III breast cancer patients. Cancer Res 61: 1890-1895, 2001
8.
Pantel K, Schlimok G, Braun S, et al: Differential expression of proliferation-associated molecules in individual micrometastatic carcinoma cells. J Natl Cancer Inst 85: 1419-1424, 1993
9.
Klein CA, Schmidt-Kittler O, Schardt JA, et al: Comparative genomic hybridization, loss of heterozygosity, and DNA sequence analysis of single cells. Proc Natl Acad Sci U S A 96: 4494-4499, 1999
10.
Hosch S, Kraus J, Scheunemann P, et al: Malignant potential and cytogenetic characteristics of occult disseminated tumor cells in esophageal cancer. Cancer Res 60: 6836-6840, 2000 11. Braun S, Pantel K: Biological characteristics of micrometastatic cancer cells in bone marrow. Cancer Metastasis Rev 18: 75-90, 1999[Medline] This article has been cited by other articles:
|
|||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
|
|||||||||||
|
|||||||||||
|
|
Copyright © 2001 by the American Society of Clinical Oncology, Online ISSN: 1527-7755. Print ISSN: 0732-183X
|
