Are We Ready to Use Surrogate End Points and Surrogate Tissues to Evaluate Response to Chemopreventive and Therapeutic Intervention?

2126 Vol. 6, 2126–2128, June 2000 Clinical Cancer Research

Editorial Are We Ready to Use Surrogate End Points and Surrogate Tissues to Evaluate Response to Chemopreventive and Therapeutic Intervention?

Reuben Lotan1 intermediate biomarkers in the oral mucosa of the liarozole- Department of Thoracic/Head and Neck Medical Oncology, The treated patients is that liarozole failed to increase endogenous University of Texas M. D. Anderson Cancer Center, Houston, Texas ATRA levels to those required to affect the expression of the 77030 biomarkers. Previous studies have shown that liarozole failed to inhibit the proliferation of breast and prostate carcinoma cell In this issue of Clinical Cancer Research, Zeng et al. (1) lines or to enhance the differentiation of mouse embryonal describe the analysis of putative intermediate biomarker gene carcinoma cells in vitro (2, 3). In patients with NSCLC, liaro- expression in oral cavity mucosa biopsies from 17 patients with zole (300 mg twice daily p.o.) caused side effects usually seen advanced malignancies who were enrolled in a Phase I clinical with ATRA but was well tolerated. However, no objective trial of liarozole. This agent was administered p.o. at three doses tumor response was seen in any of the 14 patients with stage (75, 150, and 300 mg twice daily), and biopsies were collected IIIB and IV NSCLC (4). In vitro studies have shown that in from the oral cavity mucosa at baseline and after 4 weeks of breast and embryonal carcinoma cells, liarozole enhanced the treatment. Total RNA was analyzed for the expression of the effects of exogenous ATRA on growth inhibition and differen- intermediate biomarkers EGF-R,2 TGF-␣, nuclear RAR-␤, and tiation (2, 3). Likewise, a combination of ATRA and liarozole cellular RA-binding protein II using a quantitative RT-PCR. The showed a 10ϫ greater induction of the RAR-␤2-lacZ reporter results show no statistically significant effect of liarozole on the (5). In human epidermis, topical liarozole enhanced responses to expression of any of the biomarkers in the oral mucosa biopsies. low levels of retinol or ATRA by inhibiting 4-hydroxylase (6). However, a trend toward an increase in RAR-␤ expression was Similarly, oral administration of liarozole at 300 mg 1 h before observed in the liarozole-treated group. Although this study an oral dose of ATRA (45 mg/m2) partially reversed the decline failed to detect modulation of the putative intermediate biomar- in ATRA level (due to oxidation) observed in patients treated kers, it has demonstrated that quantitative PCR could be a useful with ATRA as a single agent (7). Thus, an improved strategy for method for analysis of biomarkers in small biopsies, and it raises using liarozole in clinical trials may be to combine it with and addresses several issues that merit further discussion. Four ATRA. of these issues, namely, the use of liarozole, the use of surrogate Biomarkers related to cancer can be defined as phenotypic biomarkers, the use of surrogate tissues, and the use of quanti- or genotypic characteristics that are altered during carcinogen- tative RT-PCR, merit further consideration in the context of esis and/or during response to chemopreventive or therapeutic cancer chemoprevention and therapy. intervention. Such biomarkers can be used for detection, differ- Liarozole, a novel benzimidazole derivative, is a potent ential diagnosis, assessment of disease, and assessment of effi- inhibitor of cytochrome P450-dependent 4-hydroxylation of en- cacy and toxicity of preventive or therapeutic agents. Because dogenous ATRA, thereby increasing ATRA levels in both intervention trials that rely on cancer development or recurrence plasma and certain tissues (2, 3). Because prolonged use of RAs or on patient survival as end points are very prolonged and may leads to side effects, Zeng et al. (1) have chosen to use liarozole require large populations, biomarkers are needed to serve as to determine whether it can elevate endogenous ATRA levels surrogate or intermediate end points. Such biomarkers should with less severe side effects than seen with retinoids. The substitute for clinical end points such as decreased cancer inci- possible reasons why no effect was observed on the levels of the dence, delay of recurrence, or increased survival. Ideally, the surrogate end point is on the pathway of cancer development and is modulated by intervention. Biomarkers that change during the carcinogenesis process can indicate how far a lesion has advanced. Other biomarkers can indicate whether a chemopre- Received 2/23/00; accepted 2/24/00. The costs of publication of this article were defrayed in part by the ventive or therapeutic agent has reached the target tissue and payment of page charges. This article must therefore be hereby marked induced one or more cellular or molecular effects predicted on advertisement in accordance with 18 U.S.C. Section 1734 solely to the basis of the agent’s mechanism of action. Biomarkers that indicate this fact. are restored to normal expression level by a chemopreventive or 1 To whom requests for reprints should be addressed, at Department of therapeutic agent can indicate potential response to the inter- Thoracic/Head and Neck Medical Oncology, The University of Texas M. D. Anderson Cancer Center, 1515 Holcombe Boulevard, Houston, vention. To establish that a biomarker is a valid surrogate end TX 77030. Phone: (713) 792-7480; Fax: (713) 794-0209; E-mail: point requires large clinical trials using cancer development as [email protected]. an end point in chemoprevention trials and recurrence or me- 2 The abbreviations used are: EGF-R, epidermal growth factor receptor; tastasis as an end point in therapeutic adjuvant trials. However, TGF, transforming growth factor; RAR, retinoic acid receptor; RT-PCR, reverse transcription-PCR; ATRA, all-trans-retinoic acid; NSCLC, non- at present, no biomarker has been fully validated. Therefore, in small cell lung cancer; HNSCC, head and neck squamous cell carci- the meantime, we have to use unvalidated biomarkers as surro- noma; RA, retinoic acid. gates for clinical efficacy. Understanding the biology of the Downloaded from clincancerres.aacrjournals.org on October 1, 2021. © 2000 American Association for Cancer Research. Clinical Cancer Research 2127

cancer may help in the identification of such markers. Indeed, organs such as the breast or prostate may have limitations Zeng et al. (1) have based their choice of biomarkers on their because the mechanisms of carcinogenesis in these organs are previous demonstration that both EGF-R and its ligand TGF-␣ distinct from the oral mucosa in terms of the etiology and the are elevated in HNSCCs and in histologically normal mucosal differentiation and proliferation signaling pathways that are samples from patients with HNSCC compared with control abrogated during carcinogenesis. normal mucosa from patients without cancer. Furthermore, Zeng et al. (1) have demonstrated the analysis of RNA Grandis et al. (8–10) have shown that both EGF-R and TGF-␣ isolated from whole biopsies by quantitative RT-PCR assay, could be down-regulated by ATRA in HNSCC cell lines. The using a primer dripping method to measure the expression level intermediate biomarker RAR-␤ was chosen because its mRNA of biomarkers relative to a control housekeeping gene. They levels have been reported to decrease during oral carcinogenesis recommend that methods suitable for small quantities of RNA even in premalignant lesions and to be up-regulated by 13- should be considered and devised for use in the analysis of small cis-RA in vivo (11). Likewise, cellular RA-binding protein II biopsies. Castillo et al. (18) have previously described a fast and has been implicated in ATRA metabolism and response to sensitive method based on combining RT-PCR with a colori- ATRA (12, 13). metric ELISA. They have demonstrated the use of this method The choice of these biomarkers appears to be plausible for the analysis of RAR-␤ mRNA using total RNA extracted because they have been modulated by ATRA in vitro and in from small laryngeal biopsies and related it quantitatively to a vivo. They were not increased by liarozole in oral mucosa ␤2 microglobulin internal control (18). In their method, PCR biopsies, possibly because the ATRA concentration required to products are labeled with digoxigenin during amplification, and modulate these biomarkers is higher than that achieved by the amplified DNA is hybridized with a biotinylated probe. The liarozole treatment. Other studies have shown that the levels of hybrid is then captured on a streptavidin-coated microtiter plate, one of the biomarkers used here, RAR-␤, could be increased and detection is accomplished using antidigoxigenin antibodies in vivo in patients treated with pharmacological doses of 13- conjugated with peroxidase and a peroxidase substrate (18). The cis-RA (11, 14). methods described by both Zeng et al. (1) and Castillo et al. (18) Analysis of expression of intermediate biomarkers is best can be improved by a combination of laser capture microdis- accomplished with tissues that are accessible directly for biopsy section under direct microscopic visualization (19) and real-time or brushing (e.g., premalignant and malignant tissues in skin, PCR (20). By extracting total intact RNA from the whole biopsy oral cavity, cervix, bladder, and colon) or with secretions (sali- without first separating the minor subpopulation of epithelial va, sputum, bronchial lavage, and bladder washing) or body cells from the stromal cells, Zeng et al. (1) may have missed fluids (blood and urine). However, many cancers arise in tissues changes in biomarker(s) that are expressed differentially at that are inaccessible. It has been proposed that some accessible baseline or after liarozole treatment in epithelial and mesenchy- tissues may serve as surrogates for inaccessible tissues. Re- mal cells. For example, ATRA has been reported to up-regulate cently, Kopelovich et al. (15) have suggested that cancer risk in EGF-R in fibroblasts (21) but to decrease EGF-R in epithelial inaccessible target organs can be assessed by analysis of acces- cells (10). Laser capture microdissection under direct micro- sible surrogate tissues. They proposed that a requirement from a scopic visualization permits rapid one-step isolation of selected surrogate tissue is that it be in either anatomical continuum or cell populations from a section of heterogeneous tissue. This functional (mechanistic) continuum (if noncontiguous) with the technique has been used successfully in conjunction with PCR inaccessible organ. Their assumption is that genetic changes that amplification of RNA from dissected cells (19). The real-time can be detected in the surrogate site may predict risk in the quantitative PCR method measures PCR product accumulation inaccessible site. It is not clear whether the concept can also be by following a dual-labeled fluorogenic probe (i.e., TaqMan applied to modulation of biomarkers by chemopreventive or Probe) progressively during the exponential phase of the reac- therapeutic agents because those responses depend on uptake of tion so that quantitative PCR is accomplished. This method the agents, metabolism of the agent, and regulation of expres- provides very accurate and reproducible quantitation of tran- sion of different genes that may be distinct in different tissues. scripts (mRNA), even if the starting amount of RNA is in the For example, Ayoub et al. (14) compared RAR-␤ mRNA ex- nanogram range, as in the present study. Real-time quantitative pression in brushings from the oral cavity (palate) and bronchi PCR is extremely accurate and is less labor intensive than from 40 smokers and found complete agreement in only 68% of current quantitative PCR methods because it can be automated the subjects. Thus, even in an epithelium that is continuous and and used for high throughput analysis of multiple samples exposed to the same carcinogenic insult, there is substantial (20, 22). difference in the expression of a specific biomarker. Sun et al. In conclusion, the development of quantitative methods to (16) have shown that NSCLC cell lines are generally resistant to improve the use of intermediate end point biomarkers to predict the growth-inhibitory effects of ATRA, whereas many HNSCC response to therapy is clearly an important endeavor. Zeng et al. cell lines are susceptible (17). Thus, response to ATRA in the (1) have described one method to analyze material extracted oral cavity may not predict response in the lung. Zeng et al. (1) from small biopsies for expression of intermediate end point used oral cavity mucosa biopsies as surrogate tissue to assess biomarkers, and other improvements are suggested above. response to liarozole among patients with advanced malignan- These methods can be applied for the very important assessment cies at sites that were not specified in the report. Therefore, it is of response to chemopreventive and therapeutic agents in the not possible to comment on the suitability of the oral biopsy to short term. However, long-term studies with clinical end points serve as a surrogate for those types of malignancy. The use of are needed to validate specific biomarkers and the use of acces- oral mucosa biopsies as a surrogate for tumors that develop in sible surrogate tissues to avoid equivocal results.

References 12. Delva, L., Bastie, J. N., Rochette-Egly, C., Kraiba, R., Balitrand, N., 1. Zeng, Q., Smith, D. C., Suscovich, T. J., Gooding, W. E., Trump, Despouy, G., Chambon, P., and Chomienne, C. Physical and functional D. L., and Rubin-Grandis, J. Determination of intermediate biomarker interactions between cellular retinoic acid binding protein II and the expression levels by quantitative RT-PCR in oral mucosa of cancer retinoic acid-dependent nuclear complex. Mol. Cell. Biol., 19: 7158– patients treated with liarozole. Clin. Cancer Res., 6: 2245-2251, 2000. 7167, 1999. 2. De Coster, R., Wouters, W., Van Ginckel, R., End, D., Krekels, M., 13. Jing, Y., Waxman, S., and Mira-y-Lopez, R. The cellular retinoic Coene, M. C., and Bowden, C. Experimental studies with liarozole (R acid binding protein II is a positive regulator of retinoic acid signaling 75,251): an antitumoral agent which inhibits retinoic acid breakdown. J. in breast cancer cells. Cancer Res., 57: 1668–1672, 1997. Steroid Biochem. Mol. Biol., 43: 197–201, 1992. 14. Ayoub, J., Jean-Francois, R., Cormier, Y., Meyer, D., Ying, Y., 3. Van Heusden, J., Wouters, W., Ramaekers, F. C., Krekels, M. D., Major, P., Desjardins, C., and Bradley, W. E. C. Placebo-controlled trial Dillen, L., Borgers, M., and Smets, G. The antiproliferative activity of of 13-cis-retinoic acid activity on retinoic acid receptor-␤ expression in all-trans-retinoic acid catabolites and isomers is differentially modu- a population at high risk: implications for chemoprevention of lung lated by liarozole-fumarate in MCF-7 human breast cancer cells. Br. J. cancer. J. Clin. Oncol., 17: 3546–3552, 1999. Cancer, 77: 1229–1235, 1998. 15. Kopelovich, L., Henson, D. E., Gazdar, A. F., Dunn, B., Srivastava, 4. O’Byrne, K. J., Han, C., Mitchell, K., Lane, D., Carmichael, J., S., Kelloff, G. J., and Greenwald, P. Surrogate anatomic/functional sites Harris, A. L., and Talbot, D. C. Phase II study of liarozole in advanced for evaluating cancer risk: an extension of the field effect. Clin. Cancer non-small cell lung cancer. Eur. J. Cancer, 34: 1463–1466, 1998. Res., 5: 3899–3905, 1999. 5. Sonneveld, E., van den Brink, C. E., van der Leede, B. J., Maden, M., 16. Sun, S-Y., Yue, P., Shroot, B., Dawson, M. I., Lamph, W. W., and van der Saag, P. T. Embryonal carcinoma cell lines stably trans- Chandraratna, R., Shudo, K., Hong, W. K., and Lotan, R. Differential fected with mRAR␤2-lacZ: sensitive system for measuring levels of effects of synthetic nuclear retinoid receptor-selective retinoids on the active retinoids. Exp. Cell Res., 250: 284–297, 1999. growth of human non-small cell lung carcinoma cells. Cancer Res., 57: 6. Kang, S., Duell, E. A., Kim, K. J., and Voorhees, J. J. Liarozole 4931–4939, 1997. inhibits human epidermal retinoic acid 4-hydroxylase activity and dif- 17. Sun, S-Y., Yue, P., Mao, L., Dawson, M. I., Shroot, B., Lamph, ferentially augments human skin responses to retinoic acid and retinol in W. W., Heyman, R. A., Chandraratna, R. A. S., Shudo, K., Hong, W. K., vivo. J. Invest. Dermatol., 107: 183–187, 1996. and Lotan, R. Identification of receptor-selective retinoids that are 7. Miller, V. A., Rigas, J. R., Muindi, J. R., Tong, W. P., Venkatraman, potent inhibitors of the growth of human head and neck squamous cell E., Kris, M. G., and Warrell, R. P., Jr. Modulation of all-trans retinoic carcinoma cells. Clin. Cancer Res., 6: 1563–1573, 2000. acid pharmacokinetics by liarozole. Cancer Chemother. Pharmacol., 34: 18. Castillo, L., Milano, G., Santini, J., Demard, F., and Pierfite, V. 522–526, 1994. Analysis of retinoic acid receptor ␤ expression in normal and malignant 8. Grandis, J. R., and Tweardy, D. J. Elevated levels of transforming laryngeal mucosa by a sensitive and routine applicable reverse transcrip- growth factor ␣ and epidermal growth factor receptor messenger RNA tion-polymerase chain reaction enzyme-linked immunosorbent assay are early markers of carcinogenesis in head and neck cancer. Cancer method. Clin. Cancer Res., 3: 2137–2142, 1997. Res., 53: 3579–3584, 1993. 19. Emmert-Buck, M. R., Bonner, R. F., Smith, P. D., Chuaqui, R. F., 9. Grandis, J. R., Melhem, M. F., Gooding, W. E., Day, R., Holst, V. A., Zhuang, Z., Goldstein, S. R., Weiss, R. A., and Liotta, L. A. Laser Wagener, M. M., Drenning, S. D., and Tweardy, D. J. Levels of TGF-␣ capture microdissection. Science (Washington DC), 274: 998–1001, and EGFR protein in head and neck squamous cell carcinoma and 1996. patient survival. J. Natl. Cancer Inst., 90: 824–832, 1998. 20. Heid, C. A., Stevens, J., Livak, K. J., and Williams, P. M. Real time 10. Grandis, J. R., Zeng, Q., and Tweardy, D. J. Retinoic acid normal- quantitative PCR. Genome Res., 6: 986–994, 1996. izes the increased gene transcription rate of TGF-␣ and EGFR in head 21. Jetten, A. M. Retinoids specifically enhance the number of epider- and neck cancer cell lines. Nat. Med., 2: 237–240, 1996. mal growth factor receptors. Nature (Lond.), 284: 626–629, 1980. 11. Lotan, R., Xu, X. C., Lippman, S. M., Ro, J. Y., Lee, J. S., Lee, J. J., 22. Bieche, I., Onody, P., Laurendeau, I., Olivi, M., Vidaud, D., Lide- and Hong, W. K. Suppression of retinoic acid receptor-␤ in oral pre- reau, R., and Vidaud, M. Real-time reverse transcription-PCR assay for malignant lesions and its upregulation by isotretinoin. N. Engl. J. Med., future management of ERBB2-based clinical applications. Clin. Chem., 332: 1405–1410, 1995. 45: 1148–1156, 1999.

Downloaded from clincancerres.aacrjournals.org on October 1, 2021. © 2000 American Association for Cancer Research. Are We Ready to Use Surrogate End Points and Surrogate Tissues to Evaluate Response to Chemopreventive and Therapeutic Intervention?

Reuben Lotan

Clin Cancer Res 2000;6:2126-2128.

Updated version Access the most recent version of this article at: http://clincancerres.aacrjournals.org/content/6/6/2126

Cited articles This article cites 21 articles, 12 of which you can access for free at: http://clincancerres.aacrjournals.org/content/6/6/2126.full#ref-list-1

E-mail alerts Sign up to receive free email-alerts related to this article or journal.

Reprints and To order reprints of this article or to subscribe to the journal, contact the AACR Publications Subscriptions Department at [email protected].

Permissions To request permission to re-use all or part of this article, use this link http://clincancerres.aacrjournals.org/content/6/6/2126. Click on "Request Permissions" which will take you to the Copyright Clearance Center's (CCC) Rightslink site.