Prospective study of the impact of the Prosigna assay on adjuvant clinical decision-making in unselected patients with estrogen receptor-positive, HER2-negative, node-negative early-stage breast cancer

Depósito de investigación de la Universidad de Sevilla https://idus.us.es/ Esta es la versión aceptada del artículo publicado en: This is a accepted manuscript of a paper published in: Current Medical Research and Opinion (2015) DOI: 10.1185/03007995.2015.1037730 Copyright: © 2015 Informa UK Ltd www.cmrojournal.com El acceso a la versión publicada del artículo puede requerir la suscripción de la revista. Access to the published version may require subscription. “This is an Accepted Manuscript of an article published by Taylor & Francis in Current Medical Research and Opinion on 23 Apr 2015, available at: https://doi.org/10.1185/03007995.2015.1037730” 1 Prospective study of the impact of the Prosigna assay on adjuvant clinical decisionmaking in unselected patients with estrogen receptor-positive, HER2-negative, node-negative early-stage breast cancer. Authors: Miguel Martín1,2; Milagros González-Rivera2; Serafín Morales3; Juan de la Haba4; Lucía González-Cortijo5; Luis Manso6; Joan Albanell7; Antonio González- Martín8; Sónia González9; Angels Arcusa10; Luis de la Cruz-Merino11; Federico Rojo12; María Vidal13; Patricia Galván14; Elena Aguirre3; Cristina Morales4; Sean Ferree15; Kristen Pompilio15; Maribel Casas16; Rosalía Caballero16; Uxue Goicoechea16; Eva Carrasco16; Steven Michalopoulos17; John Hornberger17,18; and Aleix Prat13,14,19 Affiliations: 1. Department of Medical Oncology, H. Gregorio Marañón, Madrid, Spain; 2. Laboratory of Translational Oncology, H. Gregorio Marañón, Madrid, Spain; 3. H. Arnau de Vilanova, Lleida, Spain; 4. H. Reina Sofía, Córdoba, Spain; 5. H. Quirón, Madrid, Spain; 6. H. Doce Octubre, Madrid, Spain; 7. H. del Mar, Barcelona, Spain; 8. MD Anderson Cancer Center, Madrid, Spain; 9. H. Mutua de Terrassa, Barcelona, Spain; 10. Consorci Sanitari de Terrassa, Barcelona, Spain; 11. H. Vírgen de la Macarena, Sevilla, Spain; 12. Department of Pathology, Fundación Jiménez Díaz, Madrid, Spain; 13. H. Vall d'Hebrón, Barcelona, Spain; 14. Translational Genomics Group, Vall d´Hebron Institute of Oncology (VHIO); 15. NanoString Technologies, Inc., Seattle, WA; 2 16. Spanish Breast Cancer Group (GEICAM), San Sebastian de los Reyes, Madrid, Spain; 17. Cedar Associates, Menlo Park, CA, USA; 18. Stanford University School of Medicine, Palo Alto, CA, USA; 19. Hospital Clínic i Provincial, Barcelona, Spain. Corresponding Author: Miguel Martin Professor of Medicine Head, Medical Oncology Service Hospital General Universitario Gregorio Marañón Dr. Esquerdo 46 28007 Madrid (Spain) Phone (and fax number): +34 91 586 81 15 3 ABSTRACT Purpose: Improved understanding of risk of recurrence (ROR) is needed to reduce cases of recurrence and more effectively treat breast cancer patients. The purpose of this study was to examine how a gene-expression profile (GEP), identified by Prosigna, influences physician adjuvant treatment selection for early breast cancer (EBC) and the effects of this influence on optimizing adjuvant treatment recommendations in clinical practice. Methods: A prospective, observational, multicenter study was carried out in 15 hospitals across Spain. Participating medical oncologists completed pre-, post-, and follow-up questionnaires recording their treatment recommendations and confidence in these recommendations, before and after knowing the patient’s ROR. Patients completed questionnaires on decision-making, anxiety, and health status. Results: Between June 2013 and January 2014, 217 patients enrolled and a final 200 were included in the study. Patients were postmenopausal, estrogen receptor-positive, human epidermal growth hormone factor negative, and node-negative with either stage 1 or stage 2 tumors. After receiving the GEP results, treatment recommendations were changed for 40 patients (20%). The confidence of medical oncologists in their treatment recommendations increased in 41.6% and decreased in 6.5% of total cases. Patients reported lower anxiety after physicians made treatment recommendations based on the GEP results (p < 0.05). Conclusions: GEP results influenced the treatment decisions of medical oncologists and their confidence in adjuvant therapy selection. Patients’ anxiety about the selected adjuvant therapy decreased with use of the GEP. Keywords: Prosigna, PAM50, intrinsic subtypes, genomic, adjuvant, breast cancer. 4 INTRODUCTION Breast cancer occurs in 1.7 million women worldwide per year.[1] Approximately 60 percent of patients are diagnosed with early stage disease.[2] Clinicopathological evaluation is recommended to facilitate treatment decisions, such as whether to undergo adjuvant hormonal therapy alone or in combination with adjuvant chemotherapy. Gene-expression profiling (GEP) is recommended in clinical guidelines to assess the risk of distant recurrence and response to chemotherapy.[3] The Prosigna Breast Cancer Assay (NanoString Technologies® Inc., Seattle, WA), which uses the PAM50 gene signature, is a newer standardized test that measures the expression levels of 50 classifier genes from formalin-fixed paraffin-embedded (FFPE) breast tumor samples. It provides intrinsic subtype classification based on the fundamental biology of an individual patient’s tumor. Prosigna’s clinically validated prognostic risk of recurrence (ROR) score categorizes patients as low, intermediate or high risk and predicts the probability of distant recurrence over 10 years.[4, 5] In the TransATAC study, a trial comparing anastrozole to tamoxifen as adjuvant therapy for early stage breast cancer, Prosigna provided more prognostic information than the Oncotype DX multi-gene- expression assay and categorized fewer patients as intermediate risk.[4] The difference in prognostic performance was particularly pronounced in the clinically relevant subset of patients with estrogen receptor (ER)-positive, human epidermal receptor negative (HER2-) disease. The Prosigna ROR score has also been shown to provide clinically meaningful prognostic information about the risk of late distant recurrence in patients with ER-positive tumors after 5 years of endocrine therapy.[6, 7] The objective of this study was to examine whether Prosigna influences adjuvant treatment selection, beyond standard immunohistochemistry (IHC) testing and standard 5 clinicopathological variables, and whether it may be used by oncologists to optimize adjuvant treatment recommendations in clinical practice. The study also sought to assess whether Prosigna could be performed reliably in hospital laboratories and to compare measures of reliability with those obtained using IHC. PATIENTS AND METHODS Study design This was a prospective, observational, multicenter study carried out in 15 hospitals across Spain affiliated with the Spanish Breast Cancer Group (GEICAM). The primary objective of the study was to characterize the impact of Prosigna on adjuvant therapy decision recommendations made by medical oncologists. Secondary study objectives were to elicit information on (1) the confidence of medical oncologists in their treatment recommendation before and after Prosigna, according to cancer recurrence risk, (2) the rate of chemotherapy-related adverse events, and (3) patients’ decisional conflict status, anxiety levels, and functional status before and after test results. Additionally, the analytical performance of the assay, when deployed as a decentralized test, was analyzed. The study was approved by all institutions’ ethical review boards. Patient population A group of postmenopausal women with histologically proven early stage breast cancer with T1 or T2 tumors (< 5 cm), negative lymph nodes (N0) and absence of metastasis were eligible for this study. Eligibility criteria also included: Eastern Cooperative Oncology Group score of 0 or 1, ER-positive, and HER2- by IHC and in-situ hybridization (ISH), and no contraindications for receiving adjuvant chemotherapy. 6 Participants provided written consent and were able to complete patient reported outcome surveys without assistance. To minimize selection bias, all consecutively seen postmenopausal women meeting the inclusion criteria were screened and asked to take part in the study by the participating medical oncologists. Tumor sample assessments FFPE surgical specimens were reviewed and prepared in tumor sections for Prosigna analysis in a GEICAM central pathology laboratory (Fundación Jiménez Díaz, Madrid, Spain) according to the manufacturer specifications.[8] RNA was extracted and tested with Prosigna on the NanoString nCounter® Analysis System in a central laboratory (IiSGM Translational Oncology laboratory, Madrid, Spain) following the manufacturer guidelines.[8] Patients were classified according to the intrinsic subtype (Luminal A, Luminal B, HER2-enriched, basal-like) and ROR risk group (low risk, 0-40; intermediate risk, 41-60; and high risk, 61-100).[4, 9, 10] A second set of FFPE tumor sections was subsequently analyzed with Prosigna in an independent replication laboratory (VHIO Translational Genomics Group, Barcelona, Spain) to assess concordance with the central laboratory. For IHC-subtyping of ER, progesterone receptor (PR), HER2 and antigen Ki67 (Ki67) samples, retrospective assessment was done in a third set of FFPE tumor sections in the central pathology laboratory, following the American Society of Clinical Oncology and the College of American Pathologists guidelines.[11-13] ER and PR were analyzed by IHC using anti-ERα specificity protein 1 clone antibody and anti-PR (1E2) clone antibody (Ventana Medical System-Roche, Tucson, AZ, USA), respectively; HER2 status was determined by IHC using Herceptest (Dako, Glostrup, Denmark) and 7 confirmed by fluorescence ISH (FISH) using Pathvysion (Abbott Molecular, IL, USA) when indicated; and Ki67 was assessed by IHC using anti-Ki-67 MIB1 clone antibody (Dako, Glostrup, Denmark). Prosigna subtypes were compared to IHC intrinsic subtypes following the St. Gallen 2013 criteria: Luminal A (ER+, HER2-, PR > 20%, Ki67 < 20%); Luminal B (Luminal B1: ER+, HER2-, Ki67 ≥ 20% and/or PR ≤ 20%; Luminal B2: ER+, HER2+, any PR, any Ki67); HER2 positive (ER-, PR-, HER2+); and Triple Negative (ER-, PR-, HER2-)[9, 14]. Physician questionnaires Participating medical oncologists completed a pre-test questionnaire recording their initial treatment recommendation, based upon standard clinical and pathological treatment variables as well as the IHC results, and their confidence in this recommendation before knowing the patient’s ROR and intrinsic tumor subtype. After receiving the test results and at a 6-month follow up, oncologists stated their final treatment recommendation and their confidence in their decision. Patient questionnaires Participating patients completed three standardized questionnaires at pre- and postassessment of Prosigna: the Decisional Conflict Scale (DCS)[15] to assess perceived level of decisional conflict; the State-Trait Anxiety Inventory (STAI)[16] to differentiate between temporary “state anxiety”, long-lasting “trait anxiety,” and depression; and the Functional Assessment of Cancer Therapy-General, version 4 (FACT-G v.4)[17] to assess patients’ quality-of-life and health status over time. 8 Statistical analyses All variables and summarized distributions were plotted with means or proportions, standard deviations, and 95% confidence intervals. The primary endpoint of the study was the proportion of adjuvant treatment recommendations that changed [e.g., from hormonal therapy only (HT) to chemotherapy plus hormonal therapy (CHT)] after physicians received Prosigna reports; the test statistic was the Fisher Exact test. A sample size of 200 patients was calculated in order to have a one-sided 95% lowerlimit confidence interval, with a 0.050 distance from the sample proportion to the lower limit when the sample proportion was 0.250 (Clopper-Pearson). The concordance of Prosigna subtypes and ROR scores between central and replication laboratories was determined, and score differences greater than 6.75 were considered to be clinically meaningful.[5] Prosigna subtypes assessed by central laboratory were compared to IHC intrinsic subtypes using the St. Gallen 2013 criteria.[9] Responses to DCS, STAI, and FACT-G questionnaires administered before and after Prosigna were compared using paired Student's t-tests. Changes in decision conflict, anxiety levels and functional status responses were examined through analysis of variance (ANOVA). 15 Patient-reported outcomes Patients reported lower state anxiety after physicians’ treatment recommendations based on Prosigna results (p < 0.01). In a subgroup analysis, only low risk patients decreased state anxiety (p < 0.01). No significant changes in trait anxiety, decisional conflict or functional status were observed (Supplementary Table 1). However, there was a statistically significant association between changes in state and changes in trait anxiety for patients in the ROR groups (p < 0.01) (Supplementary Table 2). Prosigna tumor subtyping and ROR concordance between central and replication laboratories Subtype results obtained by the central and replication laboratories were concordant in 190 samples (95%) (Kappa = 0.89) (Supplementary Table 3). Prosigna subtype was discrepant in 10 patients (3 luminal A to B, 6 luminal B to A and 1 basal-like to HER2- enriched). Nine of them were classified as intermediate risk and 1 as low risk, according to central laboratory results. Only 4 risk group discrepancies resulted from tests carried out in the replication laboratory (intermediate to low risk) (Supplementary Table 4). Risk groups were concordant in 89% of samples between central and replication laboratories (Kappa=0.81) (Supplementary Table 4). Risk categorization by the central and replication laboratories was discrepant in 22 patients (4 low to intermediate risk, 8 intermediate to low risk, 1 intermediate to high risk, and 8 high to intermediate risk). Fifty percent of the patients with discrepant risk group results demonstrated an ROR difference greater than 6.75. No samples were discordant from low-to-high risk. 16 Concordance between centrally determined IHC and Prosigna intrinsic subtypes Intrinsic subtypes defined by IHC, according to St. Gallen 2013 criteria, in a central laboratory agreed with Prosigna in 51.6% of luminal A patients and 75.4% of luminal B patients (Table 4). The concordance of luminal A or B (HER2-enriched included) compared to IHC-luminal A or B between the central IHC subtypes and Prosigna intrinsic subtypes was 60% (Kappa=0.2365, only fair agreement). Table 4 Prosigna subtypes across IHC subtypes (St. Gallen 2013 guidelines) IHC Subtype St. Gallen 2013 Prosigna Subtype Luminal A n (%) Luminal B n (%) HER2- Enriched n (%) Basal n (%) Total IHC- Luminal A 64 (51.6) 15 (24.6) 0 (0.0) 0 (0.0) 79 IHC- Luminal B 60 (48.4) 46 (75.4) 2 (66.7) 2 (100.0) 110 IHC-Triple negative 0 (0.0) 0 (0.0) 1 (33.3) 0 (0.0) 1 Total 124 61 3 2 190 Abbreviations: IHC, immunohistochemistry testing, HER2-, human epidermal receptor negative. 17 DISCUSSION Over the past decade, several multi-gene expression assays—including Oncotype Dx[18], Mammaprint[19], EndoPredict[20, 21], and Prosigna[4]—have been progressively incorporated into the clinical decision-making process. This is the first study to test Prosigna’s influence on treatment recommendations for adjuvant therapy in postmenopausal patients with HER2-, node-negative breast cancer in a clinical setting. Among postmenopausal breast cancer patients with ER-positive tumors, the addition of adjuvant chemotherapy to adjuvant endocrine therapy alone has less than a 5% increase on the otherwise predicted 15-year survival period.[23] This marginal survival benefit means that 20 patients will need to be treated with chemotherapy in order to prevent a single death. The absolute gain in survival from the addition of adjuvant chemotherapy depends on the absolute risk of recurrence with endocrine therapy alone; therefore, accurate tools for establishing patient prognosis on endocrine therapy alone are crucial. Conventional clinicopathological parameters used in the assessment of ER-positive postmenopausal patients are suboptimal predictors of both relapse risk and chemotherapeutic benefit. In particular, immunohistochemical tests are characterized by several sources of variability (e.g., different time of fixation and type of fixatives, different antibodies, variation in techniques of deparaffinization, antigen retrieval, and antibody staining, etc.) due to the non-automatized nature of the techniques. In addition, they offer only semi-quantitative results, establish somewhat artificial cut-off points of positivity, and depend largely on the skills of pathologists and technicians.[24] 18 Several studies have shown that the modern multi-gene-expression assays provide prognostic information beyond that of classical clinicopathological parameters and can influence the decision-making process for adjuvant therapy (Mammaprint[19, 25, 26], Oncotype Dx[18, 27, 28], Prosigna[4, 20, 29-31], EndoPredict[21, 32, 33]). In our study, the investigators changed adjuvant treatment recommendations in 20% of patients after knowing the results of Prosigna. This study was designed with consecutive enrollment of patients according to protocol, not taking into consideration the investigators’ perception of risk of relapse or confidence regarding therapeutic decisions made prior to obtaining Prosigna results. As a result, approximately half of the patients enrolled in the study were in the low risk group, according to both the Prosigna ROR score and the investigator´s perception before Prosigna; only 33 patients (16.5%) included in the study were in the high-risk group. Prosigna also influenced the confidence of medical oncologists in their treatment recommendation. After receipt of the Prosigna results, physician confidence increased in 41.6% of cases and decreased in 6.5%. Overall, considering the high proportion of patients with clear clinicopathological low-risk included in the study, the Prosigna results reinforced the confidence of clinicians on the accuracy of adjuvant therapy selected for the patients. In addition, Prosigna had a significant influence on patient's anxiety about the recommended adjuvant therapy. After receiving the Prosigna results, patients reported significantly lower levels of anxiety regarding their treatment selection (p < 0.05). 19 In our study, the Prosigna results showed 95% concordance between the central and replication laboratories, confirming its analytical validity. We found that the intrinsic subtype classification based on IHC or ISH, as suggested by the St. Gallen Consensus Conference [9], is not an adequate proxy for the real genomic subtypes as determined by Prosigna and previously shown by other studies.[12, 34] Twenty-five percent of patients classified as luminal A by IHC or ISH were accurately re-identified as luminal B by Prosigna and should be considered for adjuvant chemotherapy. In addition, 48% of the patients with tumors classified as luminal B by IHC/ISH were re-identified as luminal A by Prosigna and could be spared chemotherapy. Our study supports the clinical validity and clinical utility of Prosigna in real-world settings. Physicians’ knowledge of Prosigna results increased their confidence in prognosis and influenced adjuvant therapy recommendations as a result. Acknowledgement This work was supported by Nanostring Technologies, Inc. 20 REFERENCES 1. IARC Global cancer burden rises to 14.1 million new cases in 2012: Marked increase in breast cancers must be addressed. 2014. 2. Society, A.C. Cancer Facts & Figures 2014. 2014; Available from: http://www.cancer.org/acs/groups/content/@research/documents/webcontent/acspc- 042151.pdf. 3. NCCN. NCCN Guidelines Version 3.2014 Breast Cancer. 2014; Available from: http://www.nccn.org/professionals/physician_gls/f_guidelines.asp. 4. Dowsett, M., et al., Comparison of PAM50 risk of recurrence score with oncotype DX and IHC4 for predicting risk of distant recurrence after endocrine therapy. J Clin Oncol, 2013. 31(22): p. 2783-90. 5. Nielsen, T., et al., Analytical validation of the PAM50-based Prosigna Breast Cancer Prognostic Gene Signature Assay and nCounter Analysis System using formalin-fixed paraffin-embedded breast tumor specimens. BMC Cancer, 2014. 14: p. 177. 6. Sestak, I., et al., Factors predicting late recurrence for estrogen receptorpositive breast cancer. J Natl Cancer Inst, 2013. 105(19): p. 1504-11. 7. Sestak, I., et al., Prediction of Late Distant Recurrence After 5 Years of Endocrine Treatment: A Combined Analysis of Patients From the Austrian Breast and Colorectal Cancer Study Group 8 and Arimidex, Tamoxifen Alone or in Combination Randomized Trials Using the PAM50 Risk of Recurrence Score. J Clin Oncol, 2014. 8. Prosigna [Package Insert], 2013, NanoString Technologies, Inc: Seattle, WA. 9. Perou, C.M., et al., Molecular portraits of human breast tumours. Nature, 2000. 406(6797): p. 747-52. 10. Goldhirsch, A., et al., Personalizing the treatment of women with early breast cancer: highlights of the St Gallen International Expert Consensus on the Primary Therapy of Early Breast Cancer 2013. Ann Oncol, 2013. 24(9): p. 2206-23. 11. Wolff, A.C., et al., Recommendations for human epidermal growth factor receptor 2 testing in breast cancer: American Society of Clinical Oncology/College of American Pathologists clinical practice guideline update. J Clin Oncol, 2013. 31(31): p. 3997-4013. 12. Hammond, M.E., et al., American Society of Clinical Oncology/College Of American Pathologists guideline recommendations for immunohistochemical testing of estrogen and progesterone receptors in breast cancer. J Clin Oncol, 2010. 28(16): p. 2784-95. 21 13. Harris, L., et al., American Society of Clinical Oncology 2007 update of recommendations for the use of tumor markers in breast cancer. J Clin Oncol, 2007. 25(33): p. 5287-312. 14. Prat, A., et al., Prognostic significance of progesterone receptor-positive tumor cells within immunohistochemically defined luminal A breast cancer. J Clin Oncol, 2013. 31(2): p. 203-9. 15. O'Connor, A.M., Validation of a decisional conflict scale. Med Decis Making, 1995. 15(1): p. 25-30. 16. Spielberger, C.D., Gorsuch, R. L., Lushene, R., Vagg, P. R., & Jacobs, G. A, Manual for the State-Trait Anxiety Inventory. , in Consulting Psychologists Press.1983: Palo Alto, CA. 17. Montazeri, A., Health-related quality of life in breast cancer patients: a bibliographic review of the literature from 1974 to 2007. J Exp Clin Cancer Res, 2008. 27: p. 32. 18. Paik, S., et al., A multigene assay to predict recurrence of tamoxifen-treated, node-negative breast cancer. N Engl J Med, 2004. 351(27): p. 2817-26. 19. van 't Veer, L.J., et al., Gene expression profiling predicts clinical outcome of breast cancer. Nature, 2002. 415(6871): p. 530-6. 20. Filipits, M., et al., A new molecular predictor of distant recurrence in ER- positive, HER2-negative breast cancer adds independent information to conventional clinical risk factors. Clin Cancer Res, 2011. 17(18): p. 6012-20. 21. Martin, M., et al., Clinical validation of the EndoPredict test in node-positive, chemotherapy-treated ER+/HER2- breast cancer patients: results from the GEICAM 9906 trial. Breast Cancer Res, 2014. 16(2): p. R38. 22. Adaniel, C., et al., Genome-based risk prediction for early stage breast cancer. Oncologist, 2014. 19(10): p. 1019-27. 23. Effects of chemotherapy and hormonal therapy for early breast cancer on recurrence and 15-year survival: an overview of the randomised trials. Lancet, 2005. 365(9472): p. 1687-717. 24. Allred, D.C., Issues and updates: evaluating estrogen receptor-alpha, progesterone receptor, and HER2 in breast cancer. Mod Pathol, 2010. 23 Suppl 2: p. S52-9. 25. van de Vijver, M.J., et al., A gene-expression signature as a predictor of survival in breast cancer. N Engl J Med, 2002. 347(25): p. 1999-2009. 26. Buyse, M., et al., Validation and clinical utility of a 70-gene prognostic signature for women with node-negative breast cancer. J Natl Cancer Inst, 2006. 98(17): p. 1183-92. 22 27. Albain, K.S., et al., Prognostic and predictive value of the 21-gene recurrence score assay in postmenopausal women with node-positive, oestrogen-receptor-positive breast cancer on chemotherapy: a retrospective analysis of a randomised trial. Lancet Oncol, 2010. 11(1): p. 55-65. 28. Dowsett, M., et al., Prediction of risk of distant recurrence using the 21-gene recurrence score in node-negative and node-positive postmenopausal patients with breast cancer treated with anastrozole or tamoxifen: a TransATAC study. J Clin Oncol, 2010. 28(11): p. 1829-34. 29. Nielsen, T.O., et al., A comparison of PAM50 intrinsic subtyping with immunohistochemistry and clinical prognostic factors in tamoxifen-treated estrogen receptor-positive breast cancer. Clin Cancer Res, 2010. 16(21): p. 5222-32. 30. Martin, M., et al., PAM50 proliferation score as a predictor of weekly paclitaxel benefit in breast cancer. Breast Cancer Res Treat, 2013. 138(2): p. 457-66. 31. Gnant, M., et al., Predicting distant recurrence in receptor-positive breast cancer patients with limited clinicopathological risk: using the PAM50 Risk of Recurrence score in 1478 postmenopausal patients of the ABCSG-8 trial treated with adjuvant endocrine therapy alone. Ann Oncol, 2014. 25(2): p. 339-45. 32. Dubsky, P., et al., The EndoPredict score provides prognostic information on late distant metastases in ER+/HER2- breast cancer patients. Br J Cancer, 2013. 109(12): p. 2959-64. 33. Kronenwett, R., et al., Decentral gene expression analysis: analytical validation of the Endopredict genomic multianalyte breast cancer prognosis test. BMC Cancer, 2012. 12: p. 456. 34. Prat, A. and C.M. Perou, Deconstructing the molecular portraits of breast cancer. Mol Oncol, 2011. 5(1): p. 5-23.