Originally published as JCO Early Release 10.1200/JCO.2009.22.9211 on July 13 2009

This Article Full Text (PDF) Purchase Article View Shopping Cart Alert me when this article is cited Alert me if a correction is posted Services Email this article to a colleague Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Save to my personal folders Download to citation manager Rights & Permissions Google Scholar Articles by Fabian, C. J. PubMed PubMed Citation Articles by Fabian, C. J. Social Bookmarking                            What's this?

Surrogate Response Biomarkers in Prevention Research: Do They Point the Way or Lead Us Astray?

University of Kansas Medical Center, Kansas City, KS

Modulation of reversible-risk biomarkers in preliminary studies of new cancer prevention interventions is increasingly being used as a preliminary indication of efficacy before advancement to a resource-intensive cancer incidence reduction trial. However, we have little evidence from prospective studies that this approach is effective or efficient in breast cancer prevention research.

In this issue of Journal of Clinical Oncology, Decensi et al¹ correlate cancer outcomes at 5 years with modulation of two risk biomarkers—mammographic density and plasma insulin-like growth factor 1 (IGF-1)—after 2 years of treatment with low-dose tamoxifen, fenretinide, or a combination of the two. This four-arm, placebo-controlled trial in premenopausal women was designed to detect synergistic interaction of tamoxifen and fenretinide using modulation of IGF-1 and mammographic density as primary end points. Trial accrual was stopped prematurely at 235 patients because of a lack of interaction, at least for IGF-1 levels.² However, long-term follow-up of the cohort allowed investigators to address the perhaps more important question of utility of biomarker modulation in predicting cancer outcome.

Unfortunately, favorable modulation of mammographic density and IGF-1 observed with low-dose tamoxifen and a combination of tamoxifen and fenretinide did not translate into reduction in cancer events, although the hazard ratio of 0.70 for ductal carcinoma in situ (DCIS) and invasive cancer in the tamoxifen-alone arm was numerically less than that for placebo.¹ Surprisingly, fenretinide as a single agent, which was associated with minimal modulation of IGF-1 and no modulation of density when compared with placebo, was associated with significant reduction in cancer events (hazard ratio, 0.38).

How confident are we that changes (or lack thereof) in mammographic density and plasma IGF-1 are reflective of changes in breast tissue? Circulating IGF-1 is at best only modestly associated with breast cancer risk in premenopausal women.³ Reduction in plasma IGF-1 after tamoxifen may actually be reflective of estrogenic action of oral selective estrogen receptor modulators (SERMs) on liver synthesis of IGF-1, because oral estrogen also results in reduced circulating IGF-1 levels.^4,5 Because modulation of circulating IGF-1 may be related more to the pharmacodynamic properties of SERMs than to any underlying tissue change, it is a suboptimal primary surrogate response biomarker. Mammographic density, which is reflective of the amount of collagen and stroma, is a strong and well-accepted risk biomarker for premenopausal breast cancer.^6,7 Women who have density greater than 75% have a five-fold greater risk for breast cancer than women with essentially no density.⁸ It has been hypothesized that risk associated with breast density is at least partially explained by elevated systemic or tissue IGF-1 levels or by the ratio of IGF-1 to its binding proteins.^3,9,10 The increase in relative risk with 75% or higher density is similar to the five-fold increase in risk observed with histologic or cytologic evidence of atypia.^11,12 Tamoxifen consistently reduces mammographic density in premenopausal women, and both tamoxifen and fenretenide have been shown to reduce plasma IGF-1 levels.¹³ Given this seemingly strong reasoning, what explains the contradiction reported by Decensi et al¹ between biomarker results and cancer outcome?

There are multiple possibilities. First, the cohort was small, and it may have been too late in the precancer pathway for change in breast density and plasma IGF-1 to predict outcome. Second, there may have been analytic problems with IGF-1 or breast-density measurements. Third, the effects of fenretinide on progression or recurrence of in situ and invasive cancers are not mediated predominately via reduction in IGF-1 or the physiologic processes responsible for mammographic density. Fourth, tamoxifen-associated decreases in IGF-1 and mammographic density are not necessarily synonymous with a risk-reduction effect and are in part likely the results of pharmacodynamic changes common to most SERMs.¹⁴

It should be noted that this was not a typical cohort for primary prevention; 77% of the women enrolled onto this study¹ had developed a T1a invasive cancer, DCIS, or lobular carcinoma in situ within 3 years before random assignment. There is little information on whether modulation of density or plasma IGF-1 might be expected to correlate with outcome in the setting of advanced intraepithelial neoplasia and small invasive cancers. Seventy-seven percent of subsequent breast cancer events resulted from development or recurrence of DCIS or invasive cancer in the ipsilateral breast, and 98% of women who developed a cancer event had baseline in situ or invasive cancer. Consequently, this study likely examined the effect of tamoxifen and fenretinide on progression or recurrence of in situ or invasive disease. In this cohort, in which some of the premenopausal women had estrogen receptor (ER) –negative DCIS and invasive cancers, low-dose tamoxifen—which only prevents ER-positive cancer—may have been at a disadvantage compared with fenretinide, which may have activity in both ER-negative and ER-positive lesions.¹⁵

Analysis issues or technical factors may have affected the correlations of mammographic density and cancer outcome. Sixteen percent of women did not have a mammogram at baseline, and there was no attempt to control for phase of menstrual cycle in those who did. Density is known to be higher in the luteal than follicular phase of the cycle.¹⁶ In addition, there was no mandate that women had to have the same type of mammogram (eg, digital or analog) at baseline or 2 years, and according to the observations of Decensi et al,¹ digital films had an approximately 15% lower density than analog. Although an attempt was made to use a correction factor for women who had a different type of mammogram at 2 years than they did at baseline, this maneuver has not been validated. Therefore, in this study, the value of mammographic density biomarker outcome—no matter what the result—is seriously questionable, other than for hypothesis generation.

However, if we assume that mammographic density analysis issues had little bearing on outcome, we must accept the likely possibility that fenretinide does not exert its anticancer effects via molecular mechanisms associated with density modulation. There has been little investigation of the effects of fenretinide as a single agent on premenopausal breast density. In a trial in postmenopausal women receiving hormone replacement therapy, no effects of fenretinide on density were demonstrated.⁵ In preclinical models, fenretinide has reduced proliferation, inhibited angiogenesis, and potentiated apoptosis; thus it is quite conceivable that it would be active in preventing progression and recurrence of late-stage intraepithelial neoplasia.¹⁷ Fenretinide has profound effects on energy-sensing and survival (Akt) pathways.¹⁸ Fenretinide appears to improve insulin sensitivity, possibly mediated through its effects on retinol binding protein 4.^19,20 Measurements taken from women during the trial reported by Decensi et al¹ indicated that although fenretinide markedly improved insulin sensitivity in overweight women, tamoxifen reduced it.²¹ Fenretinide may also inhibit aromatase.²²

Tamoxifen treatment was associated with a reduction in cancer incidence in the IBIS-I (International Breast Cancer Intervention Study I) trial and a reduction in mammographic density in a subsequent nested study, in which baseline mammograms were estimated to contain density greater than 10%.²³ The correlation between tamoxifen and reduction in density was strongest in women younger than age 45 years, with little reduction in density for women older than age 55 years. Because tamoxifen is an effective drug for reduction of breast cancer risk in older women, even tamoxifen must be acting through additional molecular mechanisms other than those responsible for breast density. Reduction in density may be part of the pharmacodynamic profile of tamoxifen, and it is premature to base prevention treatment decisions on reduction in density after tamoxifen initiation, although density continues to be a good surrogate for testing SERMs in premenopausal women.^14,24

It is becoming clear that a single risk biomarker will not be an appropriate response indicator for all cohorts, stages of preinvasive neoplastic development, or interventions. Interestingly, the aromatase inhibitor letrozole does not seem to modulate IGF-1 or mammographic density, yet it is more effective than tamoxifen in both the metastatic and adjuvant settings.^25–27 One must be particularly cautious about the choice of biomarker when two agents with dissimilar mechanisms of action are being compared or combined. Molecular biomarkers obtained from breast tissue that are reflective of proliferation and altered gene expression, important in the promotion and progression of neoplasia, and consistent with the mechanism of action of the intervention are more likely than current imaging or blood markers to be informative in early-phase prevention trials.^28,29 The success of this approach has been demonstrated in neoadjuvant treatment trials.³⁰ The conundrum, of course, is that tissue sampling is invasive and expensive and may be difficult to apply in large prevention trials. This obstacle might be overcome with the development of inexpensive functional imaging reflective of underlying physiologic processes. In addition, validation of tissue or imaging biomarkers as response end points in prospective breast cancer prevention trials such as the one reported by Decensi et al¹ has yet to be accomplished, and a considerable expenditure of resources will be required to do so. Despite the expense, modulation of these response biomarkers, once validated, could efficiently point the way to effective prevention interventions.

AUTHOR'S DISCLOSURES OF POTENTIAL CONFLICTS OF INTEREST

The author(s) indicated no potential conflicts of interest.

REFERENCES

1. Decensi A, Robertson C, Guerrieri-Gonzaga A, et al: Randomized double-blind 2 x 2 trial of low-dose tamoxifen and fenretinide for breast cancer prevention in high-risk premenopausal women. J Clin Oncol 27:3749–3756, 2009.[Abstract/Free Full Text]

2. Guerrieri-Gonzaga A, Robertson C, Bonanni B, et al: Preliminary results on safety and activity of a randomized, double-blind, 2 x 2 trial of low-dose tamoxifen and fenretinide for breast cancer prevention in premenopausal women. J Clin Oncol 24:129–135, 2006.[Abstract/Free Full Text]

3. Schernhammer ES, Holly JM, Pollak MN, et al: Circulating levels of insulin-like growth factors, their binding proteins, and breast cancer risk. Cancer Epidemiol Biomarkers Prev 14:699–704, 2005.[Abstract/Free Full Text]

4. Pollak MN, Huynh HT, Lefebvre SP: Tamoxifen reduces serum insulin-like growth factor I (IGF-I). Breast Cancer Res Treat 22:91–100, 1992.[CrossRef][Medline]

5. Decensi A, Bonanni B, Baglietto L, et al: A two-by-two factorial trial comparing oral with transdermal estrogen therapy and fenretinide with placebo on breast cancer biomarkers. Clin Cancer Res 10:4389–4397, 2004.[Abstract/Free Full Text]

6. Santen RJ, Boyd NF, Chlebowski RT, et al: Critical assessment of new risk factors for breast cancer: Considerations for development of an improved risk prediction model. Endocr Relat Cancer 14:169–187, 2007.[Abstract/Free Full Text]

7. Martin LJ, Boyd NF: Mammographic density: Potential mechanisms of breast cancer risk associated with mammographic density—Hypotheses based on epidemiological evidence. Breast Cancer Res 10:201; 2008.[CrossRef][Medline]

8. Boyd NF, Guo H, Martin LJ, et al: Mammographic density and the risk and detection of breast cancer. N Engl J Med 356:227–236, 2007.[Abstract/Free Full Text]

9. Fehringer G, Ozcelik H, Knight JA, et al: Association between IGF1 CA microsatellites and mammographic density, anthropometric measures, and circulating IGF-I levels in premenopausal Caucasian women. Breast Cancer Res Treat epub ahead of print on September 11, 2008.

10. Tamimi RM, Cox DG, Kraft P, et al: Common genetic variation in IGF1, IGFBP-1, and IGFBP-3 in relation to mammographic density: A cross-sectional study. Breast Cancer Res 9:R18; 2007.[CrossRef][Medline]

11. Hartmann LC, Sellers TA, Frost MH, et al: Benign breast disease and the risk of breast cancer. N Engl J Med 353:229–237, 2005.[Abstract/Free Full Text]

12. Fabian CJ, Kimler BF, Zalles CM, et al: Short-term breast cancer prediction by random periareolar fine-needle aspiration cytology and the Gail risk model. J Natl Cancer Inst 92:1217–1227, 2000.[Abstract/Free Full Text]

13. Decensi A, Veronesi U, Miceli R, et al: Relationships between plasma insulin-like growth factor-I and insulin-like growth factor binding protein-3 and second breast cancer risk in a prevention trial of fenretinide. Clin Cancer Res 9:4722–4729, 2003.[Abstract/Free Full Text]

14. Fabian CJ, Kimler BF: Selective estrogen receptor modulators for primary prevention of breast cancer. J Clin Oncol 23:1644–1655, 2005.[Free Full Text]

15. Wang TT, Phang JM: Effect of N-(4-hydroxyphenyl)retinamide on apoptosis in human breast cancer cells. Cancer Lett 107:65–71, 1996.[CrossRef][Medline]

16. Ursin G, Astrahan MA, Salane M, et al: The detection of changes in mammographic densities. Cancer Epidemiol Biomarkers Prev 7:43–47, 1998.[Abstract]

17. Patel JB, Mehta J, Belosay A, et al: Novel retinoic acid metabolism blocking agents have potent inhibitory activities on human breast cancer cells and tumour growth. Br J Cancer 96:1204–1215, 2007.[CrossRef][Medline]

18. Simeone AM, Li YJ, Broemeling LD, et al: Cyclooxygenase-2 is essential for HER2/neu to suppress N- (4-hydroxyphenyl)retinamide apoptotic effects in breast cancer cells. Cancer Res 64:1224–1228, 2004.[Abstract/Free Full Text]

19. Yang Q, Graham TE, Mody N, et al: Serum retinol binding protein 4 contributes to insulin resistance in obesity and type 2 diabetes. Nature 436:356–362, 2005.[CrossRef][Medline]

20. Coward P, Conn M, Tang J, et al: Application of an allosteric model to describe the interactions among retinol binding protein 4, transthyretin, and small molecule retinol binding protein 4 ligands. Anal Biochem 384:312–320, 2009.[CrossRef][Medline]

21. Johansson H, Gandini S, Guerrieri-Gonzaga A, et al: Effect of fenretinide and low-dose tamoxifen on insulin sensitivity in premenopausal women at high risk for breast cancer. Cancer Res 68:9512–9518, 2008.[Abstract/Free Full Text]

22. Andrews WJ, Winnett G, Rehman F, et al: Aromatase inhibition by 15-deoxy-prostaglandin J(2) (15-dPGJ(2)) and N-(4-hydroxyphenyl)-retinamide (4HPR) is associated with enhanced ceramide production. J Steroid Biochem Mol Biol 94:159–165, 2005.[CrossRef][Medline]

23. Cuzick J, Warwick J, Pinney E, et al: Tamoxifen and breast density in women at increased risk of breast cancer. J Natl Cancer Inst 96:621–628, 2004.[Abstract/Free Full Text]

24. Fabian CJ, Kimler BF: Mammographic density: Use in risk assessment and as a biomarker in prevention trials. J Nutr 136:2705S–2708S, 2006 (suppl.[Free Full Text]

25. Harper-Wynne C, Ross G, Sacks N, et al: Effects of the aromatase inhibitor letrozole on normal breast epithelial cell proliferation and metabolic indices in postmenopausal women: A pilot study for breast cancer prevention. Cancer Epidemiol Biomarkers Prev 11:614–621, 2002.[Abstract/Free Full Text]

26. Fabian CJ, Kimler BF, Zalles CM, et al: Reduction in proliferation with six months of letrozole in women on hormone replacement therapy. Breast Cancer Res Treat 106:75–84, 2007.[CrossRef][Medline]

27. Cigler T, Yaffe MJ, Johnston D, et al: A placebo-controlled trial examining the effects of letrozole on mammographic breast density and bone and lipid metabolism. Breast Cancer Res Treat 106:S112; 2007 (suppl 1) abstract 2082.

28. Fabian CJ, Kimler BF, Mayo MS, et al: Breast-tissue sampling for risk assessment and prevention. Endocr Relat Cancer 12:185–213, 2005.[Abstract/Free Full Text]

29. Fabian CJ, Kimler BF: Use of biomarkers for breast cancer risk assessment and prevention. J Steroid Biochem Mol Biol 106:31–39, 2007.[CrossRef][Medline]

30. Dowsett M, Ebbs SR, Dixon JM, et al: Biomarker changes during neoadjuvant anastrozole, tamoxifen, or the combination: Influence of hormonal status and HER-2 in breast cancer—A study from the IMPACT trialists. J Clin Oncol 23:2477–2492, 2005.[Abstract/Free Full Text]

CiteULike    Complore    Connotea    Del.icio.us    Digg    Facebook    Reddit    Technorati    Twitter    What's this?

This Article Full Text (PDF) Purchase Article View Shopping Cart Alert me when this article is cited Alert me if a correction is posted Services Email this article to a colleague Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Save to my personal folders Download to citation manager Rights & Permissions Google Scholar Articles by Fabian, C. J. PubMed PubMed Citation Articles by Fabian, C. J. Social Bookmarking                            What's this?