Home: Educational Supplement: Appendix

Prognostic and Predictive Role of Proliferation Indices

Maria Grazia Daidone, Ph.D.

Proliferative cellular activity is one of the biological processes most thoroughly investigated in breast cancer for its association with neoplastic progression and metastatic potential. Several approaches, in addition to the mitotic activity component of all pathologic grading systems, have been used by pathologists and cell biologists to determine and quantify the whole proliferative fraction or discrete fractions of cells in specific cell cycle phases on consecutive series of clinical tumors. Such approaches are based on different rationales, including detection of proliferation-related antigens by way of the Ki-67/MIB-1 labeling index and evaluation of the S-phase cell fraction by quantifying nuclear DNA content or cells incorporating DNA precursors (labeled pyrimidine bases, such as H-thymidine, or halogenated analogs, such as bromo- or iododeoxyuridine). These approaches employ different methods of evaluation (immunocytochemistry, cytometry, or autoradiography), and each has advantages and disadvantages, including different feasibility rates. Moreover, the different measures of proliferation do not always prove to correlate with each other in terms of biological or clinical significance when comparatively analyzed on the same case series and may present slightly varying sensitivity and specificity rates.

In general, cell proliferation has proved to be associated with breast cancer prognosis, even though its prognostic power tends to decline over time, at least for the flow-cytometric S-phase cell fraction (FCM-S) (Bryant, Fisher, Gunduz, et al., 1998). In patients with node-negative breast cancer treated with local-regional therapy alone until relapse, and in the presence of traditional prognostic factors (age, tumor size, estrogen receptor [ER], and progesterone receptor [PgR]), the S-phase fraction (evaluated as H-thymidine labeling index [TLI], considered as a continuous variable and categorized by tree-structured regression analysis) can be used to identify subsets at different 8-year risk of local-regional relapse (in association with patient age) or distant metastasis (in association with tumor size and patient age). Cell proliferation is the only prognostic discriminant for intermediate-size (1-2 cm) tumors, whereas it is not predictive for contralateral cancer (Silvestrini, Daidone, Luisi, et al., 1995). In general, the prognostic information provided by FCM-S, TLI, or the Ki-67/MIB-1 index is confirmed in multivariate analyses, including DNA ploidy, p53 and bcl-2 expression, ER and PgR status, and histologic or nuclear grade (Wenger, Clark, 1998; Scholzen, Gerdes, 2000). This information helps to identify tumor phenotypes associated with a high risk of relapse (high cell proliferation, alone and in association with other unfavorable factors, such as young age, tumor size >2 cm, high pathologic grade, absence of ER or PgR, alterations in oncogenes or in tumor suppressor genes) and with low risk of relapse (low cell proliferation associated with older age, tumor size £2 cm, low pathologic grade, presence of ER or PgR, absence of genomic alterations).

All of these results, however, have been derived from investigations not specifically planned to determine the clinical utility of the biomarker and, in terms of quality of information, the outcomes of the studies can be considered to be only hypothesis-generating, with the advantage of long-term followup counterbalanced by marked heterogeneity in technical and analytical procedures. The usefulness of prognostic indicators in patient management can be tested in the context of randomized treatment protocols in which evaluation of the utility of biological information accounts for the primary or secondary objective with an improved level of evidence (LOE) of results, as in the following:

• Confirmatory studies to validate proliferative activity for identifying subsets of patients at very low risk of relapse. Evidence in favor of such a hypothesis is supported by the preliminary outcome of a large LOE I study on node-negative breast cancers (Hutchins, Green, Ravdin, et al., 1998) in which patients presenting with ER-or PgR-positive, intermediate size tumors with a low FCM-S exhibited an excellent prognosis without adjuvant treatment (5-year disease-free survival [DFS], 88 percent), similar to patients with tumors £1 cm in diameter. The result has been independently confirmed on a substantial series of node-negative tumors in a prospective investigation (Jones, et al., 1999), in which Ki-67/MIB-1 was considered in addition to FCM-S. It has also been confirmed in a study derived from cases enrolled in a large randomized clinical trial (NSABP B-14) that evaluated the effectiveness of adjuvant tamoxifen in patients with ER-positive cancers (Bryant, Fisher, Gunduz, et al., 1998). The latter study demonstrated the heterogeneous clinical outcome of patients with tumors traditionally considered at low risk and showed that FCM-S (as a continuous variable), patient age, tumor size, and PgR better differentiated the risk subsets, with the 10-year DFS rate ranging from 85 percent to 30 percent and overall survival rate from 95 percent to 40 percent. On the basis of such a clinical and pathobiological classification, it seems unlikely that the addition of adjuvant chemotherapy to tamoxifen will improve clinical outcome in women at very low risk of relapse.
  
  For high-risk (ER-negative, node-negative) tumors, however, the long-term results of a prospective randomized trial evaluating the effectiveness of adjuvant CMF confirmed the efficacy of treatment (Zambetti, Valagussa, Bonadonna, 1996). There was a benefit against both slowly and rapidly proliferating tumors that was more evident for the latter.
  
  
• Clinical utility of proliferative activity for treatment decision-making in high-risk node-negative breast cancer patients. To test the improvement in clinical outcome following adjuvant chemotherapy in high-risk cases defined on the basis of tumor proliferative activity, a prospective multicentric trial was conducted between 1989 and 1993 (Amadori, Nanni, Marangolo, et al., 2000). In that trial, patients with high-TLI tumors were randomized to receive either CMF or no further treatment following surgery ± radiotherapy. At a median followup of 80 months, relapses had occurred in 28 of the 137 patients who received CMF and in 47 of the 141 patients treated with local-regional therapy alone. A reduction in the annual relapse risk of about 40 percent with chemotherapy treatment was associated with an 11 percent absolute benefit for 5- year DFS (83 percent [95 percent CI, 77-89] for CMF-treated patients versus 72 percent [95 percent CI, 65-79] for untreated patients, p=0.028). Also shown was a reduction of both local-regional (from 6.4 percent to 2.9 percent) and distant relapses (from 21.3 percent to 12.4 percent). The benefit of CMF treatment was most evident in cases at high risk—that is, with TLI values in the second (DFS: 88 percent versus 78 percent, p=0.037) and third tertile (DFS: 78 percent versus 58 percent, p=0.024).

In summary, the results support the use of cell proliferation to select high-risk patients with node-negative tumors. The finding of a higher benefit from antimetabolite-based regimens in tumors with the highest proliferation is in keeping with the evidence from retrospective studies (to be prospectively validated) that proliferation indices may be used to help predict treatment response in adjuvant and neoadjuvant settings (Wenger, Clark, 1998).

In addition, cell proliferation can provide information regarding the efficacy of different treatment schedules. In an ancillary study analyzing 70 percent of the cases entered in a randomized treatment protocol designed to compare alternating versus sequential regimens of doxorubicin and CMF in breast cancer patients with more than three positive axillary lymph nodes, the benefit of sequential administration was mainly evident in patients with tumors with low to intermediate proliferation rates (Silvestrini, Luisi, Zambetti, et al., 2000).

Methodologically, proliferation indices in part fulfill common requirements for clinical use in terms of technical-biological effectiveness. They have been proven to describe the specific biological phenomenon, as well as to provide results that are informative and rapidly obtainable at a reasonable cost when needed for clinical decisionmaking.

However, further effort should be devoted to standardizing methodologies and interpretation criteria (mainly for FCM-S results) to improve the reliability, accuracy, and reproducibility of assay results within and among different laboratories by promoting and maintaining quality control programs (found in several countries for FCM-S and TLI), and to establishing guidelines for classifying tumors according to proliferative activity. There should also be guidelines for reporting and comparing results. All of these factors, in addition to the inherent heterogeneity of case series, could account for the variability seen in results.

In terms of clinical effectiveness, proliferation indices need to be further validated in the context of randomized trials to assess their utility to identify low-risk patients (both in the presence of traditional prognostic factors, including pathological grade, and in cases diagnosed in recent years that are possibly epidemiologically and biologically different from those diagnosed in prior decades) and to make decisions about whether to use specific adjuvant therapies.

References

Amadori D, Nanni O, Marangolo M, Pacini P, Ravaioli A, Rossi A, et al. Disease-free survival advantage of adjuvant cyclophosphamide, methotrexate, and fluorouracile in patients with node-negative rapidly proliferating breast cancer; a randomised multicenter study. J Clin Oncol 2000;18:3125-34. Abstract.

Bryant J, Fisher B, Gunduz N, Costantino JP, Emir B. S-phase fraction combined with other patient and tumor characteristics for the prognosis of node-negative, estrogen-receptor-positive breast cancer. Breast Cancer Res Treat 1998;51:239-53. Abstract.

Hutchins L, Green S, Ravdin P, et al. CMF versus CAF with and without tamoxifen in high-risk node-negative breast cancer patients and a natural history follow-up study in low-risk node-negative patients: first results of intergroup trial INT 0102. [abstract]. Proc Am Soc Clin Oncol 1998;17:1a. Jones, et al. [Abstract]. Proc Am Soc Clin Oncol 1999. Abstract.

Scholzen T, Gerdes J. The Ki-67 protein: from the known and the unknown. J Cell Physiol 2000;182:311-22. Abstract.

Silvestrini R, Daidone MG, Luisi A, Boracchi P, Mezzetti M, Di Fronzo G, et al. Biologic and clinicopathologic factors as indicators of specific relapse types in node-negative breast cancer. J Clin Oncol 1995;13:697-704. Abstract.

Silvestrini R, Luisi A, Zambetti M, Cipriani S, Valagussa P, Bonadonna G, et al. Cell proliferation and outcome following doxorubicin plus CMF regimens in node-positive breast cancer. Int J Cancer 2000;87:405-11. Abstract.

Wenger CR, Clark GM. S-phase fraction and breast cancer—a decade of experience. Breast Cancer Res Treat 1998;51:255-65. Abstract.

Zambetti M, Valagussa P, Bonadonna G. Adjuvant cyclophosphamide, methotrexate and fluorouracil in node-negative and estrogen receptor-negative breast cancer. Updated results. Ann Oncol 1996;7:481-5. Abstract.