Micrometastases and isolated tumor cells: relevant and robust or rubbish? (MIRROR): preliminary results of the MIRROR study from the Dutch breast cancer trialists’ group (BOOG)

Reviewer: Christine Hill, MD
The Abramson Cancer Center of the University of Pennsylvania
Last Modified: January 8, 2009

Presenter: E. Mamounas
Affiliation: NSABP Operations & Biostatistical Centers, Pittsburgh, PA


  • Decisions regarding use of chemotherapy and radiotherapy in the treatment of breast cancer patients may be partially based on axillary lymph node status.
  • In recent years, more extensive pathologic examination of lymph nodes has led to more complicated classification of lymph node status:
    • In addition to nodes being positive or negative for malignancy (disease classification pN1, pN2, or pN3 being based on number and location of positive nodes), nodes may now be classified as positive only on immunohistochemical staining (pN0(i+)), molecular study (pN0(mol+)), or with regions of microscopic metastatic disease measuring less than 2 mm (pN1mi).
  • The implications of pN0(i+), pN0(mol+), and pN1mi disease are not well understood.
  • Many early stage breast cancer patients are initially treated with lumpectomy and sentinel lymph node (SLN) biopsy. Indications for completion axillary dissection, axillary radiotherapy, and systemic treatments are based on findings from lumpectomy and SLN specimens.
  • Treatment approaches for patients with isolated tumor cells or micrometastases identified in a SLN have not been standardized.
  • This retrospective study was undertaken in order to evaluate outcomes for patients with isolated tumor cells or micrometastasis identified based on SLN biopsy, with a separate analysis performed to evaluate the impact of adjuvant systemic therapy.


  • This study was designed to examine outcomes for patients with early-stage breast cancer with favorable tumor characteristics, and to evaluate the impact of isolated tumor cells or micrometastatic lymph node disease, and treatment with or without adjuvant systemic therapy.
  • Inclusion criteria included the following:
    • Women having undergone surgical treatment for breast cancer with SLN biopsy in any Dutch hospital between 1998 and 2005, without completion axillary dissection.
    • > 35 years of age
    • Tumor size 1-3 cm, grade I-II OR tumor size < 1cm of any grade.
    • Pathologic nodal stage of pN0(i-), pN0(i+), or pNmi.
  • Patients were divided into cohorts as follows:
    • Cohort I: pN0(i-), no adjuvant systemic treatment
    • Cohort II: pN0(i+)/ pN1mi, no adjuvant systemic treatment
    • Cohort III: pN0(i+)/ pN1mi, adjuvant systemic treatment
  • All SLN specimens were centrally reviewed, and pathologic re-staging was performed according to the American Joint Committee on Cancer TNM staging system, 6th edition.


  • 3,240 women meeting study criteria were identified, and data for 1,744 were available for presentation as part of this abstract.
    • Cohort I: n = 935
      • Median age 58
      • Median tumor size 1.2 cm
    • Cohort II: n = 340
      • Median age 56 years
      • Median tumor size 1.4 cm
    • Cohort III: n = 469
      • Median age 56 years
      • Median tumor size 1.5 cm
  • Median follow-up is 5.5 years.
  • Median age and tumor size did not differ significantly between the 3 cohorts (p < 0.01 and p < 0.0001, respectively), nor did hormone receptor status.
  • Adjuvant systemic treatment in cohort III consisted of chemotherapy only in 10% of cases, hormonal therapy only in 63% of cases, and both in 27% of cases.
  • The five-year disease free survival was 84% for cohort I, 73% for cohort II, and 86% for cohort III (p = 0.003 for cohort I vs. cohort II; p < 0.0001 for cohort II vs. cohort III).
  • Results remained the same even with statistical correction for baseline characteristics.

Author's Conclusions

  • The authors conclude that patients with isolated tumor cells or micrometastatic disease in axillary lymph nodes identified as part of SLN biopsy have a significantly worse 5-year disease-free-survival than those with nodes that are pathologically negative based in immunohistochemical and molecular studies.
  • They note that disease-free survival in this population is improved with use of adjuvant systemic treatment.

Clinical/Scientific Implications

  • This study represents an important first step in improving our understanding of the potential implications of minimal disease deposits in sentinel lymph node specimens for patients with otherwise early-stage, favorable-risk breast cancer.
  • Understanding of the prognostic significance of these small areas of metastatic disease is quite limited. Most treatment paradigms for patients with early-stage breast cancer have been determined based on findings from very large, randomized trials. As pathologic evaluation of lymph node specimens has become more extensive, a cohort of patients is now identified that did not previously exist, for the most part, back when these large trials were first conducted. As these groups of patients with isolated tumor cells or microscopic tumor deposits in sentinel lymph node specimens have emerged, many questions remain regarding their optimal treatment:
    • If these patients would once have been classified as pN0, should they be treated as such?
    • Alternatively, should we treat disease that we know is present, even if minimally so, and potentially risk greater side-effects and late toxicity?
  • For many clinicians, the answers lie somewhere in the middle, but treatments for these patients are not standardized and are often debated.
  • In this trial, the authors demonstrate that disease-free survival may be negatively impacted by the presence of isolated tumor cells or microscopic nodal disease deposits. These findings represent some of the first regarding these subsets of patients, and are interesting and important. The study presented here remains limited by several factors, however:
    • Its retrospective nature is a primary limitation in and of itself. Although data from a randomized, controlled trial would be in many ways more valid and potentially practice-changing, such data will likely not be available for many years. In the meantime, the data presented here are important and hypothesis-generating.
    • In addition, the authors present data for only half of the patients identified and meeting study criteria, and they do not offer a clear explanation for this. Potentially, analysis of this relatively small fraction of eligible patients may introduce bias.
    • Patients analyzed as part of this study were treated in many different Dutch hospitals. Surgical quality and technique, particularly with regard to sentinel lymph node biopsy, would be expected to vary somewhat based on institution. The central pathologic review that was performed as part of this study serves to minimize institutional variability.
    • Finally, and perhaps most importantly, details of treatment, as well as variances compared to treatment in the United States, limit the generalizability of the data presented here.
      • First, very little information is presented regarding the specific adjuvant treatments offered. Specific information regarding chemotherapy and hormone modulating drugs used was not provided.
      • Additionally, in most United States centers, patients with pN0, pN0(i+), or pN1mi would be offered adjuvant hormonal treatments for hormone-receptor positive tumors, and many would be offered chemotherapy. The large numbers of patients not receiving these treatments in this study may represent differences in standards of care internationally.
      • Finally, the authors did not present data with regard to radiotherapy technique, or radiation utilized. One would expect that many patients analyzed in this study would have received radiotherapy, and for some, this may have been delivered with three or four radiation fields designed specifically to address the axilla. Further information regarding radiotherapy would certainly increase the applicability and completeness of the authors’ findings.
    • Despite these issues, this study is an important jumping-off point for further investigations regarding the prognostic significance of isolated tumor cells and micrometastasis in sentinel lymph node specimens. The authors’ results indicated that such pathologic findings may be associated with worse disease-free survival, and that adjuvant systemic treatment may improve upon this. Future prospective studies will hopefully help to further elucidate these implications.