Obstetrics & Gynecology Track the topics, authors and articles important to you
HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
QUICK SEARCH:   [advanced]


     

Obstetrics & Gynecology 2004;103:1204-1210
© 2004 by The American College of Obstetricians and Gynecologists
This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Citing Articles
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Lentz, S. E.
Right arrow Articles by Amezcua, C. A.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Lentz, S. E.
Right arrow Articles by Amezcua, C. A.
Related Collections
Right arrow Gynecologic oncology
Right arrow Pathology

ORIGINAL RESEARCH

Identification of Micrometastases in Histologically Negative Lymph Nodes of Early-Stage Cervical Cancer Patients

Scott E. Lentz, MD*, Laila I. Muderspach, MD*, Juan C. Felix, MD{dagger}, Wei Ye, MS{ddagger}, Susan Groshen, PhD{ddagger} and Charles A. Amezcua, MD{dagger}

From the Departments of *Obstetrics and Gynecology, {dagger}Pathology, and {ddagger}Preventive Medicine, University of Southern California Keck School of Medicine, Los Angeles, California.

Address reprint requests to: Scott E. Lentz, MD, Gynecologic Oncology, Southern California Permanente Medical Group, 4900 Sunset Boulevard, Station 3C, Los Angeles, CA 90027; e-mail: Scott.E.Lentz{at}kp.org.


   ABSTRACT
TOP
ABSTRACT
MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 REFERENCES
 
OBJECTIVE: Despite histologically negative lymph nodes, approximately 15% of patients with early-stage cervical cancer will develop recurrence. Micrometastases have been shown to be important in staging and treatment of breast cancers and melanoma and have been identified by polymerase chain reaction analysis in cervical cancers. This study sought to estimate the frequency of micrometastases identified by immunohistochemistry in histologically negative lymph nodes and compare this to other known risk factors for recurrence of cervical cancer.

METHODS: Early-stage (stages IA2, IB1, and IB2) cervical cancer patients of all histologic subtypes were identified from the surgical logs of the Los Angeles County–University of Southern California Medical Center for the period 1994–2000. One hundred thirty-two patients had histologically negative lymph nodes. Immunohistochemical assay was performed on 3,106 lymph nodes by using antibodies against cytokeratins AE-1 and CAM 5.2 in combination according to standard protocols. The stained nodes were then evaluated for the presence of micrometastases and compared against the respective clinicopathologic information in each case.

RESULTS: Micrometastases were detected in 19 of 132 (15%, 95% confidence interval [CI] 9%, 22%) patients, found in 29 of the 3,106 (0.9%) lymph nodes evaluated. Vascular space invasion was seen in 50 of 132 cases (38%, 95% CI 30%, 47%) and in 8 of 19 (42%, 95% CI 21%, 66%) cases with micrometastases. Surgical margins of the resected specimen were negative in 120 of 132 cases (91%, 95% CI 84%, 95%) and in 16 of 19 (84%, 95%CI 60%, 96%) of those cases with micrometastases. Micrometastases were seen most frequently in pelvic lymph nodes (25 of 29, 86%). Patients with more than 20 lymph nodes removed were more likely to demonstrate metastasis (P < .001). There was no statistically significant association between micrometastasis and vascular space invasion or tumor volume.

CONCLUSION: Micrometastases are identifiable in histologically negative lymph nodes in 15% (95% CI 9%, 22%) of early-stage cancer patients, a frequency which approximates the recurrence rate for patients with negative nodes. In this series, patients with greater numbers of lymph nodes analyzed were more likely to have lymph node micrometastasis identified. There appears to be no relationship between tumor volume and the identification of micrometastases. Although micrometastases can be identified in histologically negative lymph nodes, their presence is not strongly associated with other known factors of cervical cancer recurrence. Further research is needed to determine whether the presence of lymph node micrometastases is associated with an unfavorable prognosis.

LEVEL OF EVIDENCE: II-3

The importance of lymph node status in cervical cancers is demonstrated in the universal requirement for complete lymphadenectomy in the surgical treatment of early-stage tumors. With the exception of stage IA1 carcinomas, most sources advocate lymphadenectomy for early-stage disease. Despite the fact that International Federation of Gynecology and Obstetrics (FIGO) staging for cervical cancers is clinical, treatment protocols all include treatment of the pelvic and/or aortic nodal regions to affect cure.1 The most significant negative prognostic factor in survival from cervix cancer is the presence of lymph node metastases.2 Patients with positive lymph nodes have decreased 3-year disease-free survival (86% versus 74%), and 5-year survival decreases dramatically as the number of positive lymph nodes increases. Patients with 5 or more positive lymph nodes have 0% 5-year survival.3 Stage IB cervical cancer patients have a 15% recurrence rate despite histologically negative nodes.3 More significantly, no patients survive their recurrence if the lymph nodes were positive at primary staging, regardless of treatment type. These same authors note that 18% of patients with negative nodes at primary treatment will survive their recurrence.4

Current practice in assessing nodal disease is performed with hematoxylin-eosin (H&E) staining. Aggressive serial sectioning of lymph nodes can increase the node positivity rate in otherwise negative-node samples.5 Conventional understanding teaches that metastasis begins with a single cell, which then expands into a visible metastasis. It is reasonable to believe that the ability to identify single-cell metastases could denote patients at risk for recurrence at their earliest state, but this is dependent on accurate and rapid evaluation of these nodes.

Breast and prostate cancers, as well as melanomas, have been studied extensively with respect to the presence of microscopically identifiable metastases, or micrometastases. These are generally regarded as single-cell metastases or clusters of cells not exceeding 2 mm.6 Occult nodal metastases in breast cancers have shown a significant association with decreased overall survival and decreased time to recurrence.7 Immunohistochemical assessment of histologically negative nodes identifies occult metastases in 10–25% of breast cancer patients, and multivariate analysis of polymerase chain reaction (PCR)–based detection of micrometastases was shown to be an independent and significant predictor of clinical outcome.8 With melanoma, studies have shown that PCR-based analysis of sentinel lymph nodes more accurately predicts clinical outcome than does conventional histopathologic examination.9 Therapeutic decisions are based on the presence or absence of micrometastases in these cancers.

Micrometastases are only now being evaluated in cervical cancers. Polymerase chain reaction–based studies have shown small-volume metastatic disease in up to 50% of otherwise negative lymph nodes. These are concisely reviewed by Van Trappen10 in 2001. To this point, few studies have evaluated the presence of micrometastases and correlated this with other known risk factors for tumor recurrence. This study evaluates the association of immunohistochemically detected micrometastases in early-stage cervical cancer patients with historically significant demographic and pathologic findings.


   MATERIALS AND METHODS
TOP
ABSTRACT
 MATERIALS AND METHODS
RESULTS
 DISCUSSION
 REFERENCES
 
Patients were identified from the surgical logs of the Division of Gynecologic Oncology at the LAC+USC Medical Center during the period 1994–2000 with local institutional review board approval. Those patients with early-stage cervical malignancy who had undergone definitive surgical treatment were eligible for inclusion. Only patients with complete pelvic/aortic lymphadenectomy as part of their treatment were considered for this study. Based on current FIGO staging, the selection of patients was confined to stages IA through IB2. Patients at stages IA through IB1 received no treatment before surgery. It is our practice to radiate (whole pelvic radiation and brachytherapy) stage IB2 patients before surgery and perform aortic lymphadenectomy at the time of adjuvant simple hysterectomy. In this small subset, only the nonirradiated aortic nodes were evaluated.

Eligible patients were required to be histologically negative with H&E staining for lymph node metastasis. The medical records of these patients were reviewed and clinical data abstracted concerning personal demographics (age, racial subgroup, and gravidity/parity), history of smoking, presence of immunosuppression, and specific procedures performed at surgery. The corresponding pathology reports were reviewed and evaluated for known pathologic risk factors. Pathology reports were evaluated for tumor type, histologic grade, size, presence of vascular space invasion, and the number of lymph nodes removed from each site. Demographic data and pathologic information were collected according to the guidelines and approval of the University of Southern California Keck School of Medicine Institutional Review Board.

Sampled lymph nodes were reexamined for each patient to confirm absence of metastasis by conventional H&E staining. Paraffin-embedded tissue blocks that contained the respective lymph node samples were cut into 5-µm sections. Immunohistochemical analysis was performed according to the protocol used by the International Breast Cancer Study Group (Cote et al7) to detect occult lymph node metastases. Sections were deparaffinized in xylene and rehydrated. The tissue sections were then soaked in 0.3% hydrogen peroxide to block endogenous peroxidase activity. Microwave antigen retrieval in pH 6.0 citrate buffer (BioGenex, San Ramon, CA) was performed for a total of 30 minutes. Specimens were then incubated overnight with a cocktail comprising murine monoclonal antibodies AE-1 (Signet Pathology Systems, Dedham, MA) and CAM 5.2 (Becton Dickinson, San Jose, CA) at 1:500 and 1:50 dilutions, respectively. A modified avidin-biotin immunoperoxidase method (Vector Laboratories, Burlingame, CA) was used for antibody detection. Sections were developed with diaminobenzidine tetrahydrochloride substrate and counterstained with hematoxylin. All slides were labeled with arbitrary designations and examined separately by each pathologist (C.A.A. and J.C.F.). Any disagreement regarding the presence of a micrometastasis was deliberated by the 2 pathologists until a final consensus was established. Lymph nodes found to have single or small groups of malignant cells (2–3 cells) showing strong dense golden-brown staining were considered to harbor an occult micrometastasis.

Fisher exact test was used to test the association between micrometastasis and other categorical clinical/tumor characteristics. The association between micrometastasis and the total number of lymph nodes resected was tested by using Cochran-Armitage trend test.19 Exact test on common odds ratio11 was performed to test the association between micrometastasis and the number of nodes resected (grouped as less than 20 and greater than or equal to 20) after adjusting for gravidity and the association between micrometastasis and gravidity after adjusting for total number of resected nodes.


   RESULTS
TOP
ABSTRACT
 MATERIALS AND METHODS
 RESULTS
DISCUSSION
 REFERENCES
 
One hundred thirty-two early-stage cervical cancer patients were identified with histologically negative lymph nodes. The patient demographics are outlined in Table 1. The median age of the patient sample was 42 years, and the vast majority were of Hispanic ethnicity (88%). Most of the patients included had squamous cell carcinomas (80%), stage IB1 was the most frequently represented stage (70%), and 46% were both FIGO stage IB1 and squamous. The median number of pregnancies was 4, and more than half of the cases (58%) had both pelvic and aortic nodes dissected. The sample of patients who had only aortic nodes dissected represents those patients with FIGO stage IB2 tumors who received preoperative radiation therapy.


View this table:
[in this window]
[in a new window]
  		Table 1. Patient Demographics (n = 132)

 

Twenty-nine micrometastases were found in a total of 3,106 (1%) nodes examined, which included 19 of 132 (15%; 95% CI 9%, 22%) patients with early-stage cervical cancer. Vascular space invasion was seen in 50 of 132 cases (38%; 95% CI 30%, 47%) and in 8 of 19 (42%; 95% CI 21%, 66%) cases with micrometastases. Surgical margins of the resected specimen were negative in 120 of 132 cases (91%; 95% CI 84%, 95%) and in 16 of 19 (84%; 95%CI 60%, 96%) of those cases with micrometastases. No single site emerged as predominant when finding micrometastases, although pelvic micrometastases were far more prevalent than aortic ones (Table 2). Micrometastases were seen most frequently in pelvic lymph nodes (25 of 29, 86%). Seventeen of the 19 patients who showed micrometastases had disease found in pelvic nodes. Three of the 19 patients showed micrometastases in the aortic nodes. Of these, 1 was radiated preoperatively and had aortic node sampling only. Of the other 2, 1 patient showed micrometastases in both pelvic and aortic nodes, whereas the final patient did show an aortic metastasis in the absence of pelvic micrometastases. Pelvic specimens typically contained more nodes than aortic specimens; therefore, micrometastases were found more frequently in the pelvic specimens. Thirteen patients demonstrated a single micrometastasis, 4 patients had 2 distinct metastases, and 2 patients showed 3 or more metastases (data not shown).


View this table:
[in this window]
[in a new window]
  		Table 2. Location of Identified Micrometastases

 

Table 3 shows the univariate analysis of the association between micrometastasis and patient/tumor characteristics. Neither age at diagnosis, clinical stage, margin status, tumor histology, the presence of vascular space invasion, preoperative radiation therapy, nor tumor volume were associated with the presence of a micrometastasis. The association between immunosuppression and micrometastasis could not be evaluated because only 3 patients had immunosuppression. The association between the presence of a micrometastasis and increasing number of lymph nodes resected was revealed by the Cochran-Armitage trend test categorizing the number of lymph nodes resected into the following groups: 1–13, 14–20, 21–30, and more than 30 (P < .001). The odds ratios of having micrometastasis are 0.05 (95% CI 0.003, –0.94), 0.12 (95% CI 0.01, 1.07), and 1.78 (95% CI 0.59, 5.34) for the 1–13, 14–20, 21–30 groups, respectively, compared with the more than 30 group. After inspection of Table 4, we found that this pattern occurred only when the patients had both pelvic and aortic dissection. It persisted when restricting the analysis to only those patients who had both pelvic and aortic nodes removed (P = .026), whereas almost no patients had micrometastasis with pelvic dissection only or aortic dissection only. There seems to have been a cut point at 20 lymph nodes resected. The odds of having micrometastasis for patients with 20 or more lymph nodes resected were 21.9 (95% CI 2.8, 169.5) times as great as that for those with less than 20 resected nodes (P < .001). Site resected and gravidity (0–3 versus 4 or more) were also significantly associated with the presence of micrometastases (P = .011 and .013, respectively). After adjusting for the number of nodes resected, the association between micrometastasis and gravidity was no longer significant (exact test of common odds ratio, P = .17, the adjusted odds ratio = 2.4, with a 95% CI 0.7, 8.2 by Mantel–Haenszel method for more than 3 gravidity versus less than or equal to 3 gravidity). The association between micrometastasis and number of nodes resected remained significant after adjusting for gravidity (exact test of common odds ratio, P < .001, the adjusted odds ratio = 15.1, with a 95% CI 2.0, 111.6 by Mantel-Haenszel method for more than 20 nodes resected versus 20 or fewer nodes resected).


View this table:
[in this window]
[in a new window]
  		Table 3. Association Between Micrometastasis and Clinicopathologic Characteristics (N = 132)

 

View this table:
[in this window]
[in a new window]
  		Table 4. Association of Micrometastasis With Site Resected and Number of Nodes Resected

 


   DISCUSSION
TOP
ABSTRACT
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
REFERENCES
 
In this study, micrometastases were found by immunohistochemical methods in 15% of early-stage cervical cancer patients who were considered node-negative by conventional histologic analysis. This number is strikingly similar to that number of patients who eventually experience recurrence after apparently negative nodes. The fact that 10–15% of patients who fail to show nodal metastases eventually experience recurrence in the lymph nodes highlights a false negative rate that is unacceptably elevated. It is tempting to speculate that these are the same patients who will eventually experience recurrence, but this study is not designed to make that assertion. However, this technique might allow us to more precisely identify those patients at risk and, thereupon, institute preventive therapy.

The paltry salvage rate of patients who have lymph node metastases at primary surgery (0%) demands accuracy and certainty about the presence (or absence) of nodal metastases. The sources of failure in conventional histology may be due to technique, interpretation, or tumor. Several investigators have tested various methods to improve detection of metastases in otherwise negative lymph nodes. These fall into 2 major categories: highly sensitive histologic methods or molecular targeting of specific features. Early work looked at L1 genes showing an increased incidence of human papillomavirus (HPV) DNA in nodes of patients with recurrence.12 Later work targeted the E6/E7 portion of HPV-16 or HPV-18 genomes with PCR-based technology. Efforts directed against DNA are limited by the presence of degraded material (tumor cell detritus). This confounder led investigators to test messenger ribonucleic acid (mRNA) as a target to focus on in viable tumor cells. Despite this, there have been no studies to date which demonstrate prognostic validity in E6/E7 mRNA, just as the DNA work failed to show prognostic utility.13,14

Current technology has focused on cytokeratin 19 mRNA. Cytokeratins offer unique targets for identifying metastases because they are specific for epithelial cells and generally not expressed in lymphoid tissue. Cytokeratin 19 is already used in breast cancers with a wide range of micrometastasis rates (9–67%).15 This disparity may be related to a limitation in quantitative measurement of the cytokeratin 19 product. This method has been evaluated in cervical cancers using a highly quantitative method that is also able to exclude potential confounding expression of cytokeratin 19 pseudogenes.10 These authors were able to identify cytokeratin 19 expression in the lymph nodes of 50% of patients in a small series. Unlike the present study, these authors did find significant associations with stage, vascular space invasion, and tumor grade. No follow-up data were offered, and the prognostic significance is uncertain.

An alternative to PCR technology is the use of immunohistochemistry to identify micrometastases. This method offers an advantage over conventional histology in that it allows for rapid assessment of nodes for distinct color differences, which localize malignant cells. The International Breast Cancer Study group recently published 2 separate reports regarding immunohistochemical sampling of negative nodes with a broad spectrum panel of anticytokeratin markers. These authors found this method an independent predictor of disease recurrence.5,16 We chose to use the same method as the International Breast Cancer Study group because of the broad specificity of the AE-1 and CAM 5.2 antibody mixture and its apparent prognostic viability.7 Antigen retrieval methods were also performed according to the methods of these authors.17

Micrometastases in malignant melanoma have been identified by using both immunohistochemistry and PCR techniques, with no clear answer regarding a preferred method. More significant is the compelling argument that pathologically staged patients show very different survival rates than those staged clinically. This led the American Joint Committee on Cancer to revise its staging system to include micrometastases in the pathologic staging of melanoma.18

That there is a direct correlation between the likelihood of finding a micrometastasis and the number of lymph nodes resected confirms the belief that "persistence pays off." Because of the significant impact of nodal status at the primary procedure, aggressive lymphadenectomy is preferred in early-stage disease rather than selective sampling.3,4 Lymph node surgery in cervical carcinoma is traditionally more meticulous than that which is practiced in ovarian or endometrial carcinomas. The therapeutic nature of lymphadenectomy in cervical cancer (as opposed to the selective resection in ovarian disease) has led to a reconsideration of the technique used in endometrial carcinoma. The increased likelihood of finding micrometastases in samples containing more than 20 nodes supports the necessity of such meticulous work. If this finding proves to be prognostically valuable, surgery that fails to remove a micrometastasis would be incomplete. The sample of patients who have only aortic specimens is somewhat small. Because aortic specimens typically yield smaller numbers of nodes than pelvic ones, the sample size may be too small to draw conclusions about the frequency of aortic node micrometastasis. Table 4 is compelling in that only 1 patient with a total node sampling of less than 20 demonstrated a micrometastasis.

Remarkably, there was no relationship between the presence of micrometastasis and other predictors of recurrence in cervical cancer. Standard measures of recurrence risk, such as vascular space invasion, tumor volume, and margin status, were not strongly associated with the presence of a micrometastasis. The absence of strong associations may be a reflection of the sample size. Small subsets may hide relationships that would come forward in a larger sample. It may be that these surrogate markers for tumor aggressiveness do not accurately measure the potential for early metastasis. However, in the absence of survival data, it is also likely that the conventional measures remain more predictive.

The present study presents a large trial of immunohistochemical analysis of otherwise negative lymph nodes in cervical cancer patients. To our knowledge, the number of patients, as well as the number of nodes sampled, exceeds other published work in this area. Our limitations include the lack of follow-up data, as well as the small numbers in various subsets. Some apparent associations fail to reach statistical significance, probably because of inadequate power. Compared with historical figures, this technique appears to accurately identify patients at risk for recurrence. The 15% incidence is more consistent with recurrence rates than the widely disparate incidences found using other techniques. This method, proven to be useful in alternate tumors, is also easily adapted to current practice. With minor alterations in usual processing, rapid and regular evaluation of micrometastases could become standard practice. Further evaluation of this patient group and a prospective trial of this technique will allow for a clear determination of the most accurate predictors of recurrence.


   Footnotes
 
Financial support provided in part by the Gynecologic Oncology Associates Charitable Foundation.

Received October 2, 2003. Received in revised form January 19, 2004. Accepted February 26, 2004.

10.1097/01.AOG.0000125869.78251.5e


   REFERENCES
TOP
ABSTRACT
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 REFERENCES
 
1. Piver MS, Marchetti DL, Patton T, Halpern J, Blumenson L, Driscoll DL. Radical hysterectomy and pelvic lymphadenectomy versus radiation therapy for small (< 3 cm) stage IB cervical carcinoma. Am J Clin Oncol 1988;11:21–4.[Medline]

2. Yuan C, Wang P, Lai C, Tsu E, Yen M, Ng M. Recurrence and survival analysis of 1,115 cervical cancer patients treated with radical hysterectomy. Gynecol Obstet Invest 1999;47:127–32.[Medline]

3. Delgado G, Bundy B, Zaino R, Sevin BU, Creasman WT, Major F. Prospective surgical-pathological study of disease-free interval in patients with stage IB squamous cell carcinoma of the cervix: a Gynecologic Oncology Group study. Gynecol Oncol 1990;38:352–7.[Medline]

4. Fuller AF Jr, Elliott N, Kosloff C, Hoskins WJ. Lewis JL Jr. Determinants of increased risk for recurrence in patients undergoing radical hysterectomy for stage IB and IIA carcinoma of the cervix. Gynecol Oncol 1989;33:34–9.[Medline]

5. Ludwig Breast Cancer Study Group. Prognostic importance of occult axillary lymph node micrometastases from breast cancers. Lancet 1990;335:1565–8.[Medline]

6. American Joint Committee on Cancer. AJCC cancer staging manual. 6th ed. New York (NY): Springer-Verlag; 2002.

7. Cote RJ, Peterson HF, Chaiwun B, Gelber RD, Goldhirsch A, Castiglione-Gertsch M, et al. Role of immunohistochemical detection of lymph-node metastases in management of breast cancer. International Breast Cancer Study Group. Lancet 1999;354:896–900.[Medline]

8. Masuda N, Tamaki Y, Sakita I, Ooka M, Ohnishi T, Kadota M, et al. Clinical significance of micrometastases in axillary lymph nodes assessed by reverse transcription-polymerase chain reaction in breast cancer patients. Clin Cancer Res 2000;6:4176–85.[Abstract/Free Full Text]

9. Blaheta HJ, Schittek B, Breuninger H, Garbe C. Detection of micrometastasis in sentinel lymph nodes of patients with primary cutaneous melanoma. Recent Results Cancer Res 2001;158:137–46.[Medline]

10. Van Trappen PO, Gyselman VG, Lowe DG, Ryan A, Oram DH, Bozse P, et al. Molecular quantification and mapping of lymph-node micrometastases in cervical cancer. Lancet 2001;357:15–20.[Medline]

11. Gart J. Point and interval estimation of the common odds ratio in the combination of 2 x 2 tables with fixed marginals. Biometrika 1970;57:471–5.[Abstract/Free Full Text]

12. Burnett AF, Barnes WA, Johnson JC, Grendys E, Willett GD, Barter JF, et al. Prognostic significance of polymerase chain reaction detected human papillomavirus of tumors and lymph nodes in surgically treated stage IB cervical cancer. Gynecol Oncol 1992;47:343–7.[Medline]

13. Baay MFD, Koudstaal J, Hollema H, Duk JM, Burger MPM, Quint WGV, et al. Detection of HPV-16 DNA by PCR in histologically cancer free lymph nodes from patients with cervical cancer. J Clin Path 1997;50:960–1.[Abstract/Free Full Text]

14. Czegledy J, Iosif C, Forslund O, Willen R, Hansson BG. Detection of human papilloma virus DNA in lymph nodes extirpated at radical surgery for cervical cancer is not predictive of recurrence. J Med Virol 1998;54:183–5.[Medline]

15. Van Trappen PO, Pepper MS. Lymph angiogenesis and lymph node microdissemination. Gynecol Oncol 2001;82:1–3.[Medline]

16. Mansi JL, Gogas H, Bliss JM, Gazet JC, Berger U, Coombes RC. Outcome of primary-breast-cancer patients with micrometastases: a long-term follow-up study. Lancet 1999;354:197–202.[Medline]

17. Shi SR, Cote RJ, Taylor CR. Antigen retrieval immunohistochemistry: past, present, and future. J Histochem Cytochem 1997;45:327–43.[Abstract/Free Full Text]

18. McNeil C. Micrometastases matter in new melanoma staging system. J Natl Cancer Inst 2000;92:1370–1.[Free Full Text]

19. Armitage P. Test for linear trend in proportions and frequencies. Biometrics 1955;11:375–86.





This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Citing Articles
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Lentz, S. E.
Right arrow Articles by Amezcua, C. A.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Lentz, S. E.
Right arrow Articles by Amezcua, C. A.
Related Collections
Right arrow Gynecologic oncology
Right arrow Pathology

HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS