Journal of Nuclear Medicine, published on June 12, 2009 as doi:10.2967/jnumed.108.060772 INVITED PERSPECTIVE Aiming for a Direct Hit: Combining Molecular Imaging with Targeted Cancer Therapy

nohistochemical staining or probing for tary and synergistic; in fact, one clinical Oncogene addiction is an intrigu- HER2 gene amplification by fluores- trial has examined this combination for ing new term that has emerged in the cence in situ hybridization. There is treatment of HER2-positive breast can- lexicon of cancer biology (1). This generally good concordance in HER2 cer, with promising results (13). term implies that tumor cells may have positivity between primary and meta- Fundamental to the success of any an extraordinary, almost complete de- static lesions, but some discrepancies therapeutic strategy aimed at disrupt- pendence on the function of one or have been noted (8). Patients with ing the influence of the HER2 onco- more genes (oncogenes) for their HER2-amplified tumors (immunohis- gene are an accurate assessment of the survival, whereas these same genes tochemical staining 31 or HER2–to– level of HER2 density in tumors and expressed in normal cells are not chromosome-17 control gene sequence the ability to sensitively measure a treat- absolutely essential. The human epi- ratios . 2.2) are expected to benefit ment effect on HER2 expression. In this dermal growth factor receptor type 2 most from trastuzumab (9). issue of The Journal of Nuclear Medi- (HER2) is one such oncogene that A novel alternative strategy to in- cine, Kramer-Marek et al. report that encodes a transmembrane receptor terfere with the function of the HER2 small-animal PET with 18F-labeled tyrosine kinase overexpressed in about oncogene is the use of heat-shock ZHER2:342 Affibody (Affibody AB) spe- protein (Hsp) inhibitors, in particular cific for HER2 can sensitively provide those aimed at Hsp90: 17(allylamino)- information on these 2 phenomena (14). See Page 1131 17-demethoxygeldanamycin (17-AAG; Affibody is a new class of targeting tanespimycin) and its orally bioavailable vehicle generated from a phage display 20% of breast cancers, primarily be- analog, 17(dimethlyaminoethylamino)- library of Staphylococcus protein A cause of gene amplification (2). HER2 17-demethoxygeldanamycin (17-DMAG). variants in which substitutions were overexpression confers an aggressive Hsp90 is a molecular chaperone that made to a 13-amino-acid sequencewithin tumor phenotype that correlates with complexes with Hsp70 and Hsp40, as the Z-domain (15). Initial panning of the hormone insensitivity (3), chemother- well as with cochaperone proteins (p23 library for binding to the HER2 extracel- Cdc37 apy resistance (4), and poor long-term and p50 ), to mediate the correct lular domain yielded ZHER2:4 Affibody survival (5). Trastuzumab (Herceptin; folding of numerous proteins in the with good specificity but relatively low Roche Pharmaceuticals) is a humanized cell, thereby protecting them from affinity (dissociation constant, 50 nM) IgG1 monoclonal antibody approved routing to the proteasome for degra- (16). Subsequent modifications and re- for the treatment of HER2-amplified dation (10). 17-AAG and 17-DMAG screening (affinity maturation) identified metastatic breast cancer and as an ad- block the function of Hsp90 by pre- the extremely high affinity ZHER2:324 juvant treatment for early-stage disease venting its conversion to the active Affibody (dissociation constant, 22 pM) (6). The mechanisms through which adenosine triphosphate–bound form used in this study (17). Successful im- trastuzumab exerts its antitumor effects (11). Hsp90 inhibitor treatment of aging of HER2-overexpressing human are not completely understood, but an breast cancer is based on the premise tumor xenografts in mice has been important predictor of response is tu- that the cell-surface expression of achieved with Affibody labeled with mor HER2 overexpression (7). HER2 HER2 can be diminished by rerouting 111In, 99mTc, or 18F(18), and these novel positivity is assessed in a biopsy sam- newly synthesized receptors to the pro- radiopharmaceuticals were recently stud- ple of primary breast cancer by immu- teasome for proteolytic destruction. Re- ied for the first time in humans (19). liance on Hsp90 may in itself represent Kramer-Marek et al. showed that the 18 a form of oncogene addiction (10). exquisite affinity of the F-ZHER2:324 Moreover, there is evidence that Hsp90 Affibody for HER2 provided excellent Received Jan. 3, 2009; revision accepted permits recycling of internalized HER2 tumor uptake (#21% injected dose Jan. 7, 2009. For correspondence or reprints contact: Raymond back to the cell surface, thus preserving [%ID]/g) whereas the low molecular M. Reilly, University of Toronto, 144 College St., overexpression (12). Because trastuzu- weight (6 kDa) caused rapid elimination Toronto, Ontario, Canada. E-mail: [email protected] mab has promoted HER2 internalization from the blood. The resulting tumor-to- COPYRIGHT ª 2009 by the Society of Nuclear in some studies (7), its combined use with blood ratios exceeded 20:1 at only 1–2 h Medicine, Inc. DOI: 10.2967/jnumed.108.060772 Hsp90 inhibitors would be complemen- after injection, thus making it feasible

IMAGING AND TARGETED CANCER THERAPY • Reilly 1017

jnm060772-sn n 6/10/09 to use the 110-min half-life positron- clone18 and BT-474 xenografts pro- directly (26). These biomarkers, al- emitter, 18F, as a radiolabel. vided an opportunity to interrogate the though convenient, have yielded in- There was a direct relationship be- effects of 17-DMAG Hsp90 inhibitor consistent results that do not appear 18 tween uptake of F-ZHER2:324 Affi- therapy on tumor HER2 expression. to correlate with tumor response body and tumor HER2 density in Tumor-bearing mice received a base- (13,26,27). Molecular imaging using 18 a panel of 5 subcutaneous human line PET scan and were then treated F-ZHER2:324 Affibody may provide breast cancer xenografts in athymic intravenously with 4 doses of 40 mg/ a more sensitive, direct, and highly mice, except for MCF-7/clone18 tu- kg of 17-DMAG over a 60-h period, feasible means of evaluating tumor mors. Uptake was lower for MCF-7/ followed by repeated imaging 12 h response in situ to these new therapies; clone18 tumors than for BT-474 xeno- after the final dose. Region-of-interest it could also be invaluable in optimiz- grafts (12 %ID/g vs. 20 %ID/g, re- analysis revealed a 71% and 33% ing the dose. Phase I clinical trials of spectively), despite 1.5-fold higher significantly lower uptake of 18F- 17-AAG have used conventional dose- HER2 expression measured ex vivo ZHER2:324 Affibody in the BT-474 and escalation protocols to reach the max- by enzyme-linked immunosorbent as- MCF-7/clone18 tumors, respectively, imally tolerated dose for subsequent say. MCF-7/clone18 is a HER2 gene– after 17-DMAG treatment. There was phase II studies, yet this may not be transfected variant of MCF-7 cells, virtually complete agreement between the lowest or optimal dose required to which naturally have low HER2 den- decreased tumor uptake of 18F- block Hsp90. These drugs are associ- sity. This discrepancy was attributed to ZHER2:324 Affibody and ex vivo anal- ated with significant toxicities, includ- possible differences in the microenvi- ysis of tumor HER2 expression by ing hepatotoxicity (26,27), and thus, ronment of MCF-7/clone18 xenografts enzyme-linked immunosorbent assay identifying the minimum dose that is that may have affected the delivery of and Western blot. Unexpectedly, im- effective in blocking Hsp90 would be 18 18 F-ZHER2:324 Affibody. Interestingly, munohistochemical staining for HER2 desirable. PET with F-ZHER2:324 we recently reported that strong asso- (HercepTest score; Dako) in the ex- Affibody could probe the downstream ciations between tumor HER2 density planted tumors did not reveal decreases effect of increasing doses of 17-AAG and uptake of 111In-labeled trastuzu- in HER2 positivity after 17-DMAG or 17-DMAG in patients enrolled in mab in a panel of 5 HER2-positive treatment, indicating that this tech- trials of Hsp90 inhibitors for the breast cancer xenografts in athymic nique does not appear to have the treatment of HER2-positive tumors. mice were obtained only if tumor up- sensitivity required to detect these Evaluation of the kinetics of HER2 take was corrected for nonspecific IgG changes. We have also recently found downregulation by Hsp90 inhibitors localization or circulating blood-pool that immunohistochemical staining through repeated imaging either pre- radioactivity, factors that presumably does not have sufficient sensitivity to de- clinically or clinically would be reflect differences in the tumor micro- tect HER2 downregulation in athymic facilitated by the rapid biologic elim- 18 environment (20). Reasonably good mice bearing subcutaneous human ination of F-ZHER2:324 Affibody and correlations between epidermal growth breast cancer xenografts treated with the short physical half-life of the factor receptor density and the accumu- trastuzumab, whereas micro-SPECT radionuclide. The results of such lation of 64Cu-labeled anti–epidermal with 111In-labeled pertuzumab (a HER2 molecular imaging of HER2 expres- growth factor receptor cetuximab antibody that recognizes an epitope sion may be particularly profound for antibodies in epidermal growth factor different from trastuzumab) was capable combinations of trastuzumab and 17- receptor–positive xenografts in mice of monitoring such changes (23). Imag- AAG (13), in which there may be have similarly been described (21). ing of decreased tumor HER2 expres- dramatic downregulation due to the However, Aerts et al. (22) have reported sion caused by Hsp90 inhibition was potential synergy between these 2 disparities in uptake of 89Zr-labeled first reported by Smith-Jones et al. using drugs as discussed earlier. Molecular 68 cetuximab by tumors and epidermal Ga-labeled trastuzumab F(ab9)2 frag- imaging is a sensitive tool that clearly growth factor receptor density, again ments (24); this approach proved to be has an important future role to play in possibly due to differences in tumor an earlier predictor of tumor response predicting and monitoring response to physiology that control radiophar- to 17-AAG therapy than was 18F-FDG new targeted cancer therapies that aim maceutical delivery. Kramer-Marek PET (25). to exploit oncogene addiction (28). et al. (14) visualized all breast cancer It is important to appreciate that the The study by Kramer-Marek et al. (14) xenografts by small-animal PET us- effectiveness of 17-AAG in down- stimulates our imagination about what 18 ing F-ZHER2:324 Affibody, but the regulating tumor HER2 expression in could be achieved by combining these tumor signal ranged from low for clinical trials is most commonly mea- 2 exciting and continually evolving MCF-7 xenografts to intermediate for sured by Western blot analysis of fields. MCF-7/clone18 tumors and strong readily accessible peripheral blood for BT-474 tumors. lymphocytes for Hsp90, Hsp70, or Raymond M. Reilly The avid accumulation of 18F- client proteins (e.g., RAF-1, cdk4, or University of Toronto ZHER2:324 Affibody in the MCF-7/ LCK) rather than by sampling tumors Toronto, Ontario, Canada

1018 THE JOURNAL OF NUCLEAR MEDICINE • Vol. 50 • No. 7 • July 2009

jnm060772-sn n 6/10/09 REFERENCES 10. Workman P, de Billy E. Putting the heat on cancer. labeled Affibody molecule [abstract]. Eur J Nucl Nat Med. 2007;13:1415–1417. Med Mol Imaging. 2006;33(suppl 2):S91. 1. Weinstein IB, Joe A. Oncogene addiction. Cancer 11. Isaacs JS, Xu W, Neckers L. Heat shock protein 90 20. McLarty K, Cornelissen B, Scollard DA, Done SJ, Res. 2008;68:3077–3080. as a molecular target for cancer therapeutics. Chun K, Reilly RM. Associations between the 2. Owens MA, Horten BC, Da Silva MM. HER2 111 Cancer Cell. 2003;3:213–217. uptake of In-DTPA-trastuzumab, HER2 density amplification ratios by fluorescence in situ hybrid- 12. Austin CD, De Mazie`re AM, Pisacane PI, et al. and response to trastuzumab (Herceptin) in athymic ization and correlation with immunohistochemis- Endocytosis and sorting of ErbB2 and the site of mice bearing subcutaneous human tumour xenografts. try in a cohort of 6556 breast cancer tissues. Clin action of cancer therapeutics trastuzumab and Eur J Nucl Med Mol Imaging. 2009;36:81–93. Breast Cancer. 2004;5:63–69. geldanamycin. Mol Biol Cell. 2004;15:5268–5282. 21. Cai W, Chen K, He L, Cao Q, Koong A, Chen X. 3. Konecny G, Pauletti G, Pegram M, et al. 13. Modi S, Stopeck AT, Gordon MS, et al. Combi- Quantitative PET of EGFR expression in xeno- Quantitative association between HER2/neu and 64 nation of trastuzumab and tanespimycin (17-AAG, graft-bearing mice using Cu-labeled cetuximab, steroid hormone receptors in hormone receptor- KOS-953) is safe and active in trastuzumab- a chimeric anti-EGFR monoclonal antibody. Eur J positive primary breast cancer. J Natl Cancer Inst. refractory HER-2-overexpressing breast cancer: Nucl Med Mol Imaging. 2007;34:850–858. 2003;95:142–153. a phase I dose-escalation study. J Clin Oncol. 22. Aerts HJWL, Dubois L, Perk L, et al. Disparity 4. Colomer R, Montero S, Lluch A, et al. Circulating 89 2007;25:5410–5417. between in vivo EGFR expression and Zr-labeled HER2 extracellular domain and resistance to J Nucl Med. 14. Kramer-Marek G, Kiesewetter DO, Capala J. cetuximab uptake assessed with PET. chemotherapy in advanced breast cancer. Clin 2009;50:123–131. Changes in HER2 expression in breast cancer Cancer Res. 2000;6:2356–2362. 23. McLarty K, Cornelissen B, Scollard D, Reilly RM. xenografts after therapeutic intervention can be 5. Slamon DJ, Clark GM, Wong SG, Levin WJ, 111In-pertuzumab is a sensitive probe of trastuzumab- quantified using PET and 18F-labeled Affibody Ullrich A, McGuire WL. Human breast cancer: mediated HER2 downregulation in vitro and in vivo molecules. J Nucl Med. 2009;50:1131–1139. correlation of relapse and survival with amplifica- [abstract]. J Nucl Med. 2008;49(suppl):105P. 15. Nord K, Gunneriusson E, Ringdahl J, Sta˚hl S, tion of the HER-2/neu oncogene. Science. 1987; 24. Smith-Jones PM, Solit DB, Akhurst T, Afroze F, Uhle´n M, Nygren P-A˚ . Binding proteins selected 235:177–182. Rosen N, Larson SM. Imaging the pharmaco- from combinatorial libraries of an a-helical bacterial 6. Smith I, Procter M, Gelber RD, et al. 2-year dynamics of HER-2 degradation in response to Hsp90 follow-up of trastuzumab after adjuvant chemo- receptor domain. Nat Biotechnol. 1997;15:772–777. inhibitors. Nat Biotechnol. 2004;22:701–706. therapy in HER2-positive breast cancer: a random- 16. Wikman M, Steffen A-C, Gunneriusson E, et al. 25. Smith-Jones PM, Solit D, Afroze F, Rosen N, ized controlled trial. Lancet. 2007;369:29–36. Selection and characterization of HER2/neu-binding Larson SM. Early tumor response to Hsp90 ther- 7. Baselga J, Albanell J. Mechanism of action of anti- Affibody ligands. Protein Eng Des Sel. 2004;17: apy using HER2 PET: comparison with 18F-FDG HER2 monoclonal antibodies. Ann Oncol. 2001; 455–462. PET. J Nucl Med. 2006;47:793–796. 12(suppl 1):S35–S41. 17. Orlova A, Magnusson M, Eriksson T, et al. Effect 26. Banerji U, O’Donnell A, Scurr M, et al. Phase I 8. Zidan J, Dashkovsky I, Stayerman C, Basher W, of affinity maturation on biodistribution of radio- pharmacokinetic and pharmacodynamic study of Cozacov C, Hadary A. Comparison of HER-2 iodinated anti-HER2 Affibody molecules [abstract]. 17-allylamino, 17-demethoxygeldanamycin in pa- overexpression in primary breast cancer and meta- Eur J Nucl Med Mol Imaging. 2005;32(suppl 1):S78. tients with advanced malignancies. J Clin Oncol. static sites and its effect on biological targeting of 18. Tolmachev V. Imaging of HER-2 overexpression in 2005;18:4152–4161. metastatic disease. Br J Cancer. 2005;93:552–556. tumors for guiding therapy. Curr Pharm Des. 27. Goetz MP, Toft D, Reid J, et al. Phase I trial of 17- 9. Wolff AC, Hammond EH, Schwartz JN, et al. 2008;14:2999–3019. allylamino-17-demethoxygeldanamycin in patients American Society of Clinical Oncology/College of 19. Baum RP, Orlova A, Tolmachev V, Feldwisch J. A with advanced cancer. J Clin Oncol. 2005;23:1078–1087. American Pathologists guideline recommendations novel molecular imaging agent for diagnosis of 28. McLarty K, Reilly RM. Molecular imaging as for human epidermal growth factor receptor 2 testing recurrent HER2 positive breast cancer: first time in a tool for personalized and targeted anticancer in breast cancer. J Clin Oncol. 2007;25:118–145. human study using an indium-111- or gallium-68 therapy. Clin Pharmacol Ther. 2007;81:420–424.

IMAGING AND TARGETED CANCER THERAPY • Reilly 1019

jnm060772-sn n 6/10/09