Altria Group, Inc. · NYSE:MO
Altria — CORESTA Smoke-Techno Conference 2019
26 slides from Altria's conference presentation, October 2019
Slide 1: HPHC Market Map Study for US Machine-Made Cigars – Part 2 Predictive Modeling
HPHC Market Map Study for US Machine-Made Cigars – Part 2 Predictive Modeling Michael J. Morton, Karl A. Wagner, Raquel M. Olegario, Lara L. Baker, Jennifer H. Smith CORESTA Smoke-Techno Conference 6-10 October 2019 Hamburg, Germany Michael Morton | Regulatory Sciences | Altria Client Services | COR
Slide 2: Study Overview
Study Overview ▪ Part 1: market map overview1 • Discuss the inherent variability of cigars • Describe the products • Present physical properties and abbreviated HPHCs for filler and smoke under CORESTA, ISO and Intense smoking regimes • Compare the physical properties and HPHC variability of cigars
Slide 3: Market mapping
Market mapping ▪ Sampled 24 U.S. machine-made cigars • Diameter: 7.8 mm – 16 mm • Length: 95 mm – 158 mm • Product weight: 1.1 g – 8.0 g • 13 Untipped, 8 plastic / wood tipped, 3 filter tip • ~38 % market share for cigarillos • ~11 % market share for filter tips ▪ Products tested for smoke and fille
Slide 4: Objectives
Objectives ▪ Compare three different smoking regimes: CORESTA, ISO, Intense ▪ Examine correlations of TPM, Tar, and CO to smoke constituent yields and effect on market mapping prediction intervals ▪ Examine cigar filler manufacturing variability Michael Morton | Regulatory Sciences | Altria Client S
Slide 5: Comparison of Smoking Regimes
Comparison of Smoking Regimes ▪ Short-term relative standard deviations (rep-to-rep variation) ▪ Overall yield comparisons ▪ Consistency of constituent smoke yield orderings • Are the relative rankings close to the same for the different smoking regimes? Michael Morton | Regulatory Sciences | Altria
Slide 6: Smoke Chemistry Variability Comparison
Smoke Chemistry Variability Comparison ▪ Smoking regimes: CORESTA, ISO, Intense ▪ 18 constituents on FDA abbreviated constituent list plus tar and TPM ▪ 7 replicates for each product for each constituent for each regime ▪ Relative standard deviations were averaged across all 24 products for each con
Slide 7: Relative variation (%) averaged across products, analytes
Relative variation (%) averaged across 24 products, 20 analytes Cigars Analyte CORESTA ISO Intense 1-Aminonaphthalene 11.8 14.1 11.8 2-Aminonaphthalene 11.2 13.6 11.9 4-Aminobiphenyl 10.5 13.1 11.2 The short-term relative standard Acetaldehyde 10.9 9.3 8.4 deviations are comparable among the three A
Slide 8: Measured diameter and associated CORESTA Puff Volume
Measured diameter and associated CORESTA Puff Volume CORESTA CORESTA Diameter Puff Diameter Puff Product Product mm Volume mm Volume mL mL 1 10.7 20 13 10.7 20 2 12.7 22 14 15.7 34 3 10.7 20 15 9.4 20 4 9.6 20 16 10.4 20 5 9.6 20 17 10.4 20 6 9.6 20 18 10.6 20 7 9.5 20 19 10.3 20 8 7.9 20 20 7.9 20
Slide 9: Overall tar yield comparison
Overall tar yield comparison Generally Intense > ISO > CORESTA Michael Morton | Regulatory Sciences | Altria Client Services | CORESTA SSPT 6-10 October 2019 | ST37 9
Slide 10: Yield ratios can be different for analytes
Yield ratios can be different for analytes Ratio of Intense/CORESTA Ratio of ISO/CORESTA Similar to findings in cigarettes by Counts et al. Reg Tox Pharm 41 (2005) pp 185-227 Michael Morton | Regulatory Sciences | Altria Client Services | CORESTA SSPT 6-10 October 2019 | ST37 10
Slide 11: Consistency of yield rankings
Consistency of yield rankings Generally the rankings are consistent between the methods. Note most of these cigars had the same puff volume under CORESTA and ISO correlation CORESTA. ■ CORESTA and Intense correlation ▲ ISO and Intense correlation Michael Morton | Regulatory Sciences | Altria Clien
Slide 12: Objectives
Objectives ▪ Compare three different smoking regimes: CORESTA, ISO, Intense ISO 20778 (Intense) ▪ Examine correlations of TPM, Tar, and CO to smoke constituent yields and effect on market mapping prediction intervals ▪ Examine cigar filler manufacturing variability Michael Morton | Regulatory Scienc
Slide 13: Correlations between analyte and TPM, Tar, CO
Correlations between analyte and TPM, Tar, CO Gas phase compounds often better correlated to CO. Particulate phase typically better correlated to tar or TPM. TPM ■ Tar ▲ CO Michael Morton | Regulatory Sciences | Altria Client Services | CORESTA SSPT 6-10 October 2019 | ST37 13
Slide 14: Prediction Intervals
Prediction Intervals ▪ Common outputs from market maps or benchmarking studies are prediction intervals giving ranges within which future test results are expected to lie. • Commonly indexed by Tar, TPM, or CO • May incorporate filler analyte concentration Michael Morton | Regulatory Sciences | Altr
Slide 15: Prediction Intervals Benzene in Smoke
Prediction Intervals Benzene in Smoke Prediction interval Prediction interval Prediction intervals ignoring CO assuming CO=100 mg/cigar are narrower with knowledge of CO. The better the correlation of CO to the constituent yield, the more the narrowing. R2 = 0.84 Because of the variability of cigars
Slide 16: Prediction Intervals NNN in Smoke
Prediction Intervals NNN in Smoke R2 = 0.29 Michael Morton | Regulatory Sciences | Altria Client Services | CORESTA SSPT 6-10 October 2019 | ST37 16
Slide 17: Tobacco NNN varies from product to product
Tobacco NNN varies from product to product Michael Morton | Regulatory Sciences | Altria Client Services | CORESTA SSPT 6-10 October 2019 | ST37 17
Slide 18: Alternative Model
Alternative Model ▪ Intuitively, because of the large product-to-product NNN differences, one would expect the tobacco NNN to affect NNN smoke yields in addition to TPM ▪ Simple Model Incorporating TPM and Filler Analyte Concentration: 𝑁𝑁𝑁 = 𝑎 + 𝑏 𝑇𝑜𝑏𝑁𝑁𝑁 ∗ 𝑇𝑃𝑀 , where TobNNN is the…
Slide 19: Prediction Intervals NNN in Smoke
Prediction Intervals NNN in Smoke R2 = 0.84 Michael Morton | Regulatory Sciences | Altria Client Services | CORESTA SSPT 6-10 October 2019 | ST37 19
Slide 20: Prediction Intervals
Prediction Intervals ▪ The better the correlation, the narrower are the prediction intervals. ▪ The prediction intervals incorporating tobacco NNN concentration are roughly half the width of the intervals using TPM alone ▪ Because of the very high variation in cigar smoke yields, the prediction inte
Slide 21: R2 Values Incorporating Tobacco Characteristics
R2 Values Incorporating Tobacco Characteristics Tar or TPM Tar or TPM Smoke analyte Regime only and Tobacco The model for formaldehyde is NNN CORESTA 0.464 0.843 y=a + b1*Tar + b2*(TobAmm*Tar) (with TPM) ISO 0.293 0.842 Intense 0.511 0.871 (TPM not captured with carbonyls) NNK CORESTA 0.402 0.634 (w
Slide 22: Objectives
Objectives ▪ Compare three different smoking regimes: CORESTA, ISO, Intense ISO 20778 (Intense) ▪ Examine correlations of TPM, Tar, and CO to smoke constituent yields and effect on market mapping prediction intervals ▪ Examine cigar filler manufacturing variability Michael Morton | Regulatory Scienc
Slide 23: Filler Temporal Variation
Filler Temporal Variation ▪ Most products in this market map were tested twice for filler analytes • Some products from the same production batch and some from two different production batches Michael Morton | Regulatory Sciences | Altria Client Services | CORESTA SSPT 6-10 October 2019 | ST37 23
Slide 24: Percentage Difference in Two NNN Filler Results
Percentage Difference in Two NNN Filler Results This shows moderate differences when the same batch is tested at different times and often much larger differences when testing different batches Not available There was not enough of products 4, 5, and 6 remaining for filler re-testing Michael Morton
Slide 25: Summary and Conclusions
Summary and Conclusions ▪ Cigars are a more diverse product category than cigarettes ▪ All three smoking regimes gave similar levels of variability and generally ranked the smoke yields comparably ▪ Smoke yields can be benchmarked with prediction intervals using TPM, Tar, or CO yield, and, for some
Slide 26: Questions?
Thank You Questions? For copies of this presentation visit the Altria’s Science Website at www.altria.com/alcs-science Altria has a new Altria Science Twitter account Follow us @AltriaScience Michael Morton | Regulatory Sciences | Altria Client Services | CORESTA SSPT 6-10 October 2019 | ST37 26