e-cigarettes use: appraising relative risk and harm … › 2015 › 03 › uk...the analytical...

E-Cigarettes use:

Appraising relative risk and harm reversal

Prof. Riccardo Polosa Institute of Internal Medicine

University of Catania ITALY

All Party Parliamentary Group on E-‐cigare5es | 18 March 2015 | House of Commons, London

Representative PTR-MS mass spectra of VOCs released in a single exhaled breath

O’Connell G. et al SRNT-USA 26/02/2015

SRNT-USA 26 February 2015

Indoor air quality and exhaled breath composition after use of nicotine delivery products

Visit our Scientific Research website: www.imperialtobaccoscience.com

1. Introduction An electronic cigarette (e-cigarette) air quality study was conducted by a leading independent UK accredited laboratory with recognised expertise in air quality measurements and analyses for Imperial Tobacco to assess the concentration of nicotine, propylene glycol and glycerol (the main components of e-cigarette liquid) in the ambient air before, during and after use of the Puritane™ 16 mg disposable e-cigarette (manufacturer Fontem Ventures B.V.) in an office environment. A schematic representation of the office layout, the two independent sampling locations and the positions of the e-cigarette users and non-users is shown in Figure 1. To investigate potential changes in indoor air quality, the ambient air was analysed before, during and after a 165 min vaping session. Sampling times are shown in Figure 2. The average puff rate over the three e-cigarette users during the 165 min vaping session was 3.2 puffs per minute. This level of product use may have been influenced by the no-vaping restriction during the first hour. Given the puffing frequency and 0.8 air changes per hour air exchange rate, it is likely that findings in this study may be an overestimate. Table 1 summarises the results for airborne concentrations of nicotine, propylene glycol and glycerol before, during and after the vaping session. As would be anticipated, the concentration of propylene glycol in the indoor ambient air, the major constituent of the e-liquid, was higher during the vaping session relative to the background and no vaping control period but remained within the workplace exposure limit (WEL) set for this chemical. Following cessation of vaping, there was a substantial decrease in the concentration of propylene glycol in the indoor ambient air. By contrast, there was no measurable increase in the airborne concentration of nicotine during use of the e-cigarette in the office space (limit of detection [LOD] for nicotine, 7 μg/m3). Due to the LOD for glycerol (150 to 350 μg/m3), glycerol was not detected in any of the samples taken, with the results being < 250 μg/m3 for the vaping samples.

2. Air quality testing in an office before, during and after use of an electronic cigarette Table 1 Analysis of nicotine, propylene glycol and glycerol in indoor ambient air before, during and after a vaping session (average from the two sampling locations)

- [1] WHO Conference of the Parties to the WHO Framework Convention on Tobacco Control. FTCT/COP/6.10. Sixth session. Provisional agenda item 4.4.2. apps.who.int/gb/gctc/PDF/cop6/FCTC_COP6_10-en.pdf [2] UK Health and Safety Executive. EH40/2005 Workplace exposure limits. www.hse.gov.uk/pubns/books/eh40.htm [3] Colard, S. et al. (2015) Electronic Cigarettes and Indoor Air Quality: A Simple Approach to Modelling Potential Bystander Exposures to Nicotine. Int. J. Environ. Res. Public Health. 12, 282-299 [4] O’Connell, G. et al. (2015) Real-time analysis of exhaled breath following the use of a range of nicotine delivery products by PTR-MS: proof of concept study. Technical Report. Access at www.imperialtobaccoscience.com [5] McNeill, A. et al. (2014) A critique of a WHO-commissioned report and associated article on electronic cigarettes. Addiction. DOI:10.111/add.12730

References

Electronic cigarettes (e-cigarettes) and heated tobacco (Heat-not-Burn) products are gaining acceptance with consumers as alternatives to traditional tobacco products. Consequently, there is a growing interest from regulators and public health organisations on whether the aerosol exhaled from such products has implications for the quality of air breathed by bystanders. There is currently an absence of robust scientific evidence on the potential impact of exhaled aerosol on indoor air quality in everyday environments, like homes and offices. Nonetheless, there are calls, including by some by government bodies, to prohibit the use of e-cigarettes in workplaces and enclosed public spaces [1]. In the first part of our work we aimed to perform an assessment of indoor air quality by analysing the airborne concentrations of nicotine, propylene glycol and glycerol (the major components of e-cigarette liquids) before, during and after use of e-cigarettes in ‘real-life’ conditions. As there are no general indoor air quality guidelines or standards for nicotine, propylene glycol or glycerol, a comparison of the findings to UK workplace exposure limits (WELs) is made to provide an indication of potential bystander air quality [2].

As the quality of indoor air is influenced by the chemical composition of exhaled breath, in the second part of our work we aimed to determine whether Proton Transfer Reaction-Mass Spectrometry (PTR-MS) may be a suitable technique for the real-time analysis of chemicals released in exhaled breath following use of a range of nicotine delivery products. Please refer to our second SRNT-USA 2015 poster presented today for more information from our PTR-MS pilot studies [session 2; poster #54].

Figure 1 The layout of the office, the sampling locations and the positions of the e-cigarette users and non-users during the meeting.

Figure 2 Timeline showing when participants entered and exited the office, when e-cigarette use was and was not permitted and the sampling times.

3. Analysis of VOCs released in exhaled breath following use of nicotine delivery products 4. Conclusions & future work During the use of the Puritane™ 16 mg disposable e-cigarette in the small office space indoor air quality study, the concentration of propylene glycol measured in the office air, and therefore breathed by bystanders, was significantly lower than the UK WEL. Exposure of bystanders to indoor ambient air following exhalation of this chemical at the levels seen in this study within the e-cigarette aerosol would not be anticipated to cause health problems, a conclusion in agreement with [5]. There was no measureable increase in the concentration of nicotine in the indoor ambient air during vaping. To explore this finding further, we aim to determine (i) the quantity of nicotine retained by the e-cigarette user (i.e. the fraction not exhaled into the ambient air); and (ii) whether any potential nicotine in the exhaled aerosol is deposited to various surfaces. As may be expected from the tobacco basis of conventional cigarettes and heated tobacco (Heat-not-Burn), many more chemical components are detected in exhaled breath compared to simple electronic vapour products. Of note, substantially more nicotine is present in the exhaled breath following use of the tobacco based products. Due to the wide range of chemical species detected in the exhaled breath following use of the heated tobacco product, it is likely use of such products could impact indoor air quality in a similar way that has been reported for conventional cigarettes. As such, this is an important area for additional research. The indoor air quality experimental design and methodology used in our work may be employed to evaluate the indoor ambient air quality assessment of other chemicals or particulates. Moreover, our proof-of-concept PTR-MS work showed the potential of this technology to be used as a technique to monitor the emissions from a range of nicotine delivery products and quantify released VOCs in real-time under a range of conditions and determine the impact on indoor air quality.

Grant O’Connell1, Kerstin Burseg2, Kostiantyn Breiev3, John D. Pritchard1, Stefan Biel2, Xavier Cahours4 and Stéphane Colard1,4

1 Imperial Tobacco Limited, Winterstoke Road, Bristol, BS3 2LL, UK 2 Reemtsma Cigarettenfabriken GmbH, Imperial Tobacco Group, Albert Einstein Ring 7, 22761, Hamburg, Germany

3 IONICON Analytik GmbH, Eduard-Bodem-Gasses 3, 6020 Innsbruck, Austria

4 SEITA, Imperial Tobacco Group, 48 rue Danton, 45404 Fleury-les-Aubrais, France

Ion

yiel

d (a

rbitr

ary

units

)

Mass (m/z)

10-3

10-2

10-1

100 101

a) Conventional cigarette

20 40 60 80 100 120 140 160 180

Nicotine

20 40 60 180

b) Heated tobacco (Heat-not-Burn)

10-4

10-3

10-2

10-1

100

80 100 120 140 160

Nicotine

c) E-cigarette

20 40 60 80 100 120 140 160 180

10-4

10-3

10-2

10-1

100

Nicotine

Mass (m/z)

10-3

10-2

10-1

100

101

20 40 60 80 100 120 140 160 180

d) Nicotine inhalator

Nicotine

Figure 3 Representative PTR-MS mass spectra of VOCs released in a exhaled breath following use of (a) a conventional cigarette (0.6 mg nicotine [ISO smoking regime]), (b) heated tobacco device (Heat-not-Burn; iQOS with regular heatsticks) (c) electronic cigarette (20 mg/mL nicotine Puritane rechargeable e-cigarette device) and (d) 15 mg nicotine inhalator (Nicorette® Inhalator). Black peaks, VOCs released in normal exhaled breath (background control); red peaks, VOCs released in exhaled breath following product use. Results shown here are the output from a single exhalation event. Specific compound (ion trace) at m/z 163 is nicotine and is labelled with arrowhead. PTR-MS identification of nicotine at m/z 163 is shown elsewhere [4]. Three volunteers participated in this study and each volunteer used each of the four products described above. For each of the products tested: five blank breath measurements were taken directly before product use (background control) and following this the volunteer was given the product to use and become familiar with. Following this, the volunteer used the product ad libitum five times and exhaled into the PTR-MS each time allowing analysis on a per puff basis.

Chemical

Background

(before participants enter room)

[µg/m3]

Room occupied

(NO VAPING)

[µg/m3]

Room occupied

(VAPING PERMITTED)

[µg/m3]

Room unoccupied

(after participants leave room)

[µg/m3]

Workplace exposure

limit (WEL)

(8 h mean)

[µg/m3]

Comments

Measurement 1

Measurement 2

Measurement 3

Measurement 4

Nicotine < LOD < LOD < LOD < LOD 500

No measurable increase during vaping relative to background and no vaping control;

below the WEL

Propylene glycol

< LOD < LOD 204 10.2

474000

(total vapour and particulates)

Increase during vaping relative to background and no vaping control

period; substantial decrease with cessation

of vaping; below the WEL

Glycerol < LOD < LOD < LOD < LOD 10000

Glycerol not detected in any sample; due to

large limit of detection, a more sensitive

analytical method is required

Note: LOD, limit of detection

The analytical technique PTR-MS (Proton Transfer Reaction-Mass Spectrometry) is a sensitive tool for the simultaneous real-time monitoring of volatile organic compounds (VOCs) with high sensitivity. PTR-MS is a tool that does not require sample preparation and so can be used for rapid determination of exhaled breath profiles e.g. in medical diagnostics. We recently published an indoor air quality mathematical model to predict potential bystander exposures to exhaled e-cigarette aerosol constituents [3]. Here we identified ‘quantity of chemical constituent exhaled’ as the most important factor influencing indoor air quality and bystander exposure. Therefore, it is essential that precise measurements are made regarding the quantity of compounds exhaled by the e-cigarette user (e.g. nicotine) when determining potential bystander exposure. As the composition of the exhaled breath will influence the quality of indoor ambient air, PTR-MS may be used as part of an assessment scheme for indoor air quality. In this proof-of-concept study we aimed to identify and determine the breath concentrations of nicotine following use of a range of nicotine delivery products. Representative data presented in Figure 3 shows mass spectrometric profiles of exhaled breath following a single exhalation event after product use (red) and comparison with blank control breath (black). The peaks on mass 19 and 37 m/z (and their isotopes) represent the reagent ions (H3O+) and their clusters. The PTR-MS has been calibrated for nicotine (m/z 163; see arrowheads) [4]; all other red peaks correspond to compounds released following use of the specific nicotine delivery product; their identities remain to be determined in future work. Following use of a conventional cigarette and heated tobacco product, a large number of different chemicals are released in the exhaled breath, as shown by the red spectra across a range of masses. With regards to exhaled nicotine, 1150 ppb (parts per billion) nicotine were detected in the exhaled breath following use of the conventional cigarette (a) and 1840 ppb nicotine following use of the heated tobacco device (b). In contrast, with the non-tobacco products, nicotine was detected in the exhaled breath at 7 ppb following use of the e-cigarette (c) and 1 ppb nicotine following use of the nicotine inhalator (d).

Declaration This project was supported by Imperial Tobacco Group. The e-cigarette used in this study was manufactured by Fontem Ventures, a fully owned subsidiary of Imperial Tobacco Group.

Black peaks, VOCs released in exhaled breath (background control) Red peaks, VOCs released in exhaled breath following product use. Specific compound (ion trace) at m/z 163 is nicotine and is labelled with arrowhead.

ECs have a more favorable toxicity profile than tobacco cigare5es

Hect SS, et al. Nico.ne Tob Res 2015

ND

pmol/m

L

pmol/m

L

Carcinogen metabolites levels in the urine of EC users and cigarette smokers

(adjusted for age and sex)

DJ Nutt, LD Phillips, D Balfour, HV Curran, M Dockrell, J Foulds, K Fagerstrom, K Letlape, A Milton, R Polosa, J Ramsey, D Sweanor. Estimating the harms of nicotine-containing products using the MCDA approach. Eur J Addiction 2014

NicoJne containing products – risk esJmates 100

4 E-cig

6,5 3,7 6,1 1 9,5

119,2

11,3

334,6

0

50

100

150

200

250

300

350

400

Formaldehyde

ND

µg/10 puff

s

6.5 wa5s 7.5 wa5s 9 wa5s 10 wa5s

Atomizer/double wicks

Atomizer/single wick

Important factors when interpreting EC data

Findings with the product under invesFgaFon cannot be extended to other models

6,5 3,7 6,1 1 9,5

119,2

11,3

334,6

0

50

100

150

200

250

300

350

400

6,5 3,7 6,1 1 9,5

119,2

11,3

334,6

0

50

100

150

200

250

300

350

400

Formaldehyde

ND

µg/10 puff

s

6.5 wa5s 7.5 wa5s 9 wa5s 10 wa5s



74 µg/10 puffs

Threshold value for tobacco cigarettes

6,5 3,7 6,1 1 9,5

119,2

11,3

334,6

0

50

100

150

200

250

300

350

400

Formaldehyde

ND

µg/10 puff

s

6.5 wa5s 7.5 wa5s 9 wa5s 10 wa5s



Dry Puff Detec.on

Dry Puff Detec.on

74 µg/10 puffs


0,2 0,8 0,2 0,8 3,5

58,9

4,5

206,03

0

50

100

150

200

250

ND ND

Acetaldehyde µg/10 puff

s



6.5 wa5s 7.5 wa5s 9 wa5s 10 wa5s

0,2 0,8 0,2 0,8 3,5

58,9

4,5

206,03

0

50

100

150

200

250

ND ND

Acetaldehyde µg/10 puff

s



6.5 wa5s 7.5 wa5s 9 wa5s 10 wa5s

1240 µg/10 puffs


Dry Puff Detec.on

Dry Puff Detec.on

E-Cig: a product in (fast) evolution

ECLAT was planned in the 2009 when E-‐cig technology was in its infancy

The future A.D. 2009

Future studies will have to focus on improved technology and

overall user saJsfacJon

A.D. 201X

Improved product reliability Better taste More nicotine delivery

Randomised Controlled Trials

•  ‘Categoria’ 24mg nicotine EC vs. 18mg nicotine EC vs. no nicotine EC

•  300 smokers (unwilling to quit) •  1 year abstinence rates: 13%, 9% and 4% •  good tolerability (Caponnetto et al. Plos One 2013)

•  ‘Elusion’ 16mg nicotine EC vs. nicotine patch vs. no nicotine EC

•  657 smokers (motivated to quit) •  6 month abstinence rates: 7.3%, 5.8% and 4.1% •  good tolerability (Bullen et al. Lancet 2013)

STUDY ASSESSMENTS Procedure BL

Visit Wk2 Wk4 Wk6 Wk8 Wk10 Wk12 Wk24 Wk52

Visit 1 Visit 2 Visit 3 Visit 4 Visit 5 Visit 6 Visit 7 Visit 8 Visit 9 Informed consent

X

Sociodemografic factors X Medical history X Drug history X Physical examination X X X X Vital signs – HR & BP X X X X X X X X X Weight - Kg X X X X Smoking Hx X BDI and BAI X FTND X

S M O K E

C H A R T

eCO X X X X X X X X X GN-SBQ X NO and spirometry X X X X X X Saliva collection for cotinine X X Give Study Diary X X X X X X Collect Study Diary X X X X X X Craving/VAS X X X X X X X X X MNWS (past 2 weeks) X X X X X X X X X MNWS (past 24 hrs) X X X X X X X X X Adverse events X X X X X X X X X E-cig training and dispense of E-cig kit

X

Dispense study cartridges X X X X X X Cartridges use record X X X X X X Smokers’ preference X X X

Exploring harm reducFon/reversal potenFal (e.g. reducFon in biomarkers used as proxy for risk predicFon in CVD)

Long effect of smoking absJnence/reducJon on BP and HR in smokers switching to ECs

SBP changes at Week 52 from baseline

Long effect of smoking absJnence/reducJon on BP and HR in smokers switching to ECs

SBP changes at Week 52 from baseline

Harm Reversal!

BLOOD PRESSURE CONTROL IN E-‐CIG USERS K. Farsalinos et al. Int. J. Environ. Res. Public Health 2014

(N = 2162)

Dual users Single users




X


S M O K E

C H A R T


X


Exploring harm reducFon/reversal potenFal (e.g. reducFon in biomarkers used as proxy for risk predicFon in CVD and metabolic diseases)

Post CessaFon Weight Gain

Long effect of smoking absJnence/reducJon on weight gain in smokers switching to ECs

2,9

4,2

4,7

2,4

2,9

2,5

Week-‐12 Week-‐24 Week-‐52

Weight in kg

QuiGers’ weight

62 previous studies Our study

Post CessaFon Weight Gain in QuiYers: Cochrane vs ECLAT

Learning Points

•  Smoking absFnence by using ECs may lower elevated systolic BP

•  Decreases were reported also in reducers •  ECs may be a helpful alternaFve to cigareYes in smokers with elevated BP

Burj al Arab Hotel, Dubai

The World’s Highest Tennis Court On top of Burj al Arab Hotel, Dubai

70 m2




X


S M O K E

C H A R T


X


Exploring harm reducFon/reversal potenFal (e.g. reducFon in biomarkers used as proxy for risk predicFon in COAD)

80 85 90 95 100 105 110 115 120

FEV 1 (%

of p

redicted

)

BL W12 W24

2_QuiYer 1_Reducer 0_Failure

W52 Test mulJvariaJa EffeYo Valore F Ipotesi df Gradi di libertà dell'errore Sig. Eta quadrato parziale FEV1 Traccia di Pillai ,282 50,308b 1,000 128,000 ,000 ,282

Lambda di Wilks ,718 50,308b 1,000 128,000 ,000 ,282 Traccia di Hotelling ,393 50,308b 1,000 128,000 ,000 ,282 Radice di Roy ,393 50,308b 1,000 128,000 ,000 ,282

FEV1 * ClassifatW52 Traccia di Pillai ,045 2,987b 2,000 128,000 ,054 ,045 Lambda di Wilks ,955 2,987b 2,000 128,000 ,054 ,045 Traccia di Hotelling ,047 2,987b 2,000 128,000 ,054 ,045 Radice di Roy ,047 2,987b 2,000 128,000 ,054 ,045

a. Disegno: InterceYa + ClassifatW52 Disegno entro soggek: FEV1 b. StaFsFca esaYa

Chronic effect of absJnence/reducJon on spirometry in smokers switching to ECs

FEV1 data

60

70

80

90

100

110

120

FEF 2

5-‐75% (%

predicted

)

BL W12

2_QuiYer 1_Reducer 0_Failure

W24 W52

Test mulJvariaJa EffeYo Valore F Ipotesi df Gradi di libertà dell'errore Sig. Eta quadrato parziale FEF2575 Traccia di Pillai ,601 192,975b 1,000 128,000 ,000 ,601

Lambda di Wilks ,399 192,975b 1,000 128,000 ,000 ,601 Traccia di Hotelling 1,508 192,975b 1,000 128,000 ,000 ,601 Radice di Roy 1,508 192,975b 1,000 128,000 ,000 ,601

FEF2575 * ClassifatW52 Traccia di Pillai ,362 36,320b 2,000 128,000 ,000 ,362 Lambda di Wilks ,638 36,320b 2,000 128,000 ,000 ,362 Traccia di Hotelling ,567 36,320b 2,000 128,000 ,000 ,362 Radice di Roy ,567 36,320b 2,000 128,000 ,000 ,362

a. Disegno: InterceYa + ClassifatW52 Disegno entro soggek: FEF2575 b. StaFsFca esaYa

Chronic effect of absJnence/reducJon on spirometry in smokers switching to ECs

FEF25-75 data

q  Increased prevalence of asthma q  More incident asthma q  Increased asthma morbidity and mortality

q  Greater asthma severity q  More uncontrolled asthma q  Accelerated decline in lung funcJon q  Persistent airway obstrucJon q  CorJcosteroid insensiJvity

Effect of smoking cessaJon on asthma outcomes

-‐ Improvements in QoL score -‐ ReducJons in use of rescue b2-‐agonists, doses of ICS, and dayJme asthma symptoms -‐ Improvement in AHR

-‐ Improvement in AHR to direct and indirect bronchoprovocaJon tests

-‐ Improvement in lung funcJon -‐ ReducJon in sputum neutrophil count

Tønnesen et al. Nico.ne Tob Res 2005

Piccillo G, et al.Respir Med 2008

Chaudhuri R, et al.AJRCCM 2006

Quitting

HOW?

EC use and smoking cessaJon: a meta-‐analysis

0.20 overall pooled Effect Size

Rahman MA et al. Plos One 2015 (in press).

•  RCTs in healthy smokers have shown that ECs are effecFve and safe

•  No data about EC use among vulnerable populaFons, including people with asthma

•  We invesFgated changes in subjecFve and objecFve asthma outcomes as well as safety in smoking asthmaFcs who switched to EC.

Smoking habit and asthma exacerbations

Parameter Baseline 1st follow-up visit (6 months ± 1)

2nd follow-up visit (12 months ± 2)

p value to Baseline

p value to Baseline

Cigarettes/day 21.9 (±4.5) 5.0 (±2.6) <0.001 3.9 (±1.0) <0.001

Exacerbations 1.17 (±0.9) 0.87 (±0.7) 0.296 0.78 (±0.7) 0.153

Frequent exacerbators (≥ 2 exacerbaFons; n=6) halved their exacerbaFons at both follow-‐up visits

1st F/up Visit

Assessment Timepoints

Baseline Pre-Baseline

For

ced

Exp

irato

ry V

olum

e in

1 s

econ

d (L

)

3.0

3.2

3.4

3.6

3.8

2nd F/up Visit

**

FEV1 Improvement from baseline at 12 months

p=0.005 mean increase of 100mls

Harm Reversal!

Regular EC use

FE

F25

-75

(L/s

ec)

2.4

2.6

2.8

3.0

3.2

3.4

1st F/up Visit



2nd F/up Visit

**

***

FEF25-‐75 Improvement from baseline at 6 and 12 months

p=0.006 mean increase of 250mls/sec

p=0.001 mean increase of 360mls/sec

Harm Reversal!

Regular EC use

Pre-Baseline

Met

hach

olin

e P

C20

(mg/

mL)

1.0

1.5

2.0

2.5

3.0

3.5

Baseline 2nd F/up Visit

1st F/up Visit


**

Methacholine PC20 Improvement from baseline at 12 months

p=0.003 mean increase of 1.2 DD

Harm Reversal!

Regular EC use

AC

Q s

core

s

1.2

1.4

1.6

1.8

2.0

2.2

2.4

1st F/up Visit



2nd F/up Visit

***

***

Juniper’s ACQ Improvement from baseline at 6, 12 months

p=0.001 mean decrease of 0.43

p=0.001 mean decrease of 0.56

Harm Reversal!

Regular EC use

Safety and Tolerability

•  No severe adverse reactions or acute exacerbations of asthma requiring hospitalisation/ITU admissions.

•  ECs were well tolerated with dry mouth and throat irritation occasionally reported.

(N = 1173)

(N = 1062)

Dual users Single users

RESPIRATORY SYMPTOMS IN E-‐CIG USERS K. Farsalinos et al. Int. J. Environ. Res. Public Health 2014

Learning Points

•  EC use improves lung funcFon, respiratory symptoms, subjecFve asthma outcomes

•  Improvements were reported also in dual users •  Exposure to e-‐vapour in this vulnerable populaFon did not trigger acute symtoms

•  ECs are a safe alternaFve to cigareYes in smokers with chronic airways disease

o  Avoid precauFonary principle o  Think of unintended consequences o  Beware of science overload o  Focus on quality and safety product standards o  Careful monitoring of smoking prevalence

o  Cost-‐effecFveness impact on NHS

FDA and EU TPD Approach to E-‐Cig RegulaFon:

DaunFng?

“E-‐cigs greatest health advance since vaccinaJons” Prof. David NuG BBC Radio 5 live’s Shelagh Fogarty 4 February 2014

Professor David Nu5, Former government's chief drug adviser

e-cigarettes use: appraising relative risk and harm … › 2015 › 03 › uk...the analytical...

Documents