Design of the study
We performed a randomised, two-period cross-over, pilot study. Subjects were randomised into smoking two cigarettes or no smoking. Subjects refrained from smoking during nine days before each study period, verified by exhaled carbon monoxide (CO < 6 ppm) and urinary cotinine (< 25 ng/ml). The time interval between the two study periods varied between 9 to 20 days. Measurements of exhaled CO, exhaled Nitric Oxide (NO), blood sampling and Forced Expiratory Volume in 1 second (FEV1) were performed immediately before (baseline) and 1, 3, 6, 12, 24, 48, 96 and 192 hours after smoking and at the same time points in the no smoking period. Sputum was induced at 3, 6, 12, 24, 48, 96, 192 hours after smoking and no smoking. All subjects smoked two cigarettes from the same brand within 30 minutes and were encouraged to inhale deeply (Caballero unfiltered cigarettes, tar 12 mg, nicotine 1.0 mg, commercially obtained, no gifts). Adequacy of smoke inhalation was verified by the investigator. The working groups sputum induction from the ERS stated recently that sputum inductions should not be repeated within 48 hours to avoid carry over effects . Taking this into account, we used a cross-over design (including no smoking) in this study to correct for this carry over effect. We have analysed the results of the control arm as a separate study in order to investigate the induction and resolution of the inflammatory response generated by repeated sputum inductions .
Sixteen healthy intermittent smokers were recruited by advertisements in the local newspaper. Intermittent smoking was defined as smoking more than one cigarette a month, but not daily, during the last 6 months. We chose to investigate intermittent smokers because they are able to refrain from smoking for a certain time period (in contrast to most current smokers) and they are used to inhale smoke (in contrast to non-smokers). Included were subjects older than 40 years, with normal lung function (prebronchodilator FEV1/IVC [Inspiratory Vital Capacity] > 89% of predicted for women and > 88% of predicted for men  and a prebronchodilator FEV1 > 1 litre). Excluded were subjects with: 1) a history of asthma, allergic rhinitis, or allergic eczema; 2) atopy, confirmed by a positive skin prick test; 3) any current respiratory disease, symptoms of cough or sputum production; 4) a respiratory tract infection within the preceding 8 weeks or a nasal infection within the preceding 4 weeks; 5) treatment with glucocorticosteroids within the preceding 8 weeks; 6) use of aspirin, NSAIDs, paracetamol or antihistamines within the preceding 4 weeks. Subjects were asked to avoid places with high environmental tobacco smoke exposure during the study periods. The study was approved by the medical ethics committee of the University Medical Center Groningen, the Netherlands. Written informed consent was obtained from all subjects.
Pulmonary function, exhaled NO and CO
FEV1 and IVC were measured according to the guidelines of the European Respiratory Society , using a pneumotachograph (Jaeger, Wurzberg, Germany). Exhaled NO levels were determined according to the guidelines of the American Thoracic Society , exhaling with a flow of 100 ml/sec against a resistance between 5 and 20 cm H2O, using a chemiluminescence analyser (Ecophysics CLD 700 AL). Exhaled CO levels were measured using an infrared CO analyser (UNOR 6N, Maihak AG, Hamburg, Germany) .
Blood differential cell counts were analysed automatically with a haematology flow cytometer (Coulter-STKS, Beckman Coulter, Miami, USA). Flow cytometric analysis was performed on blood cells using peridinin chlorophyll protein (PerCP) labelled anti-human leukocyte antigen (HLA)-DR, phycoerythrin (PE) labelled anti-CD11b, allophycocyanin (APC) labelled CD14 and fluorescein-isothiocyanate (FITC) labelled CD63 monoclonal antibodies (Becton Dickinson, Franklin Lakes, NJ USA). HLA-DR, CD63 and CD11b are activation markers for respectively monocytes and granulocytes. CD14 is used to discern between monocytes and granulocytes. Functional assays were performed on unstimulated and lipopolysacharide (LPS, 1 ng/ml, BioWhittaker, Walkerville, USA) stimulated blood cells, measuring tumor necrosis factor (TNF)-α, interleukin (IL)-1β, IL-8 and IL-10 by ELISA (Sanquin, Amsterdam, the Netherlands).
Sputum induction and processing
Sputum was induced according to a modified standard technique , using 4.5% hypertonic saline. Whole sputum was processed within 120 minutes according to the modified method of Rutgers and colleagues . The cell-free supernatant was collected and stored in aliquots at -80°C pending analysis of soluble mediators.
Flow cytometric analysis was performed on sputum cells using PerCP labelled anti-HLA-Dr, PE labelled anti-CD11b, APC labelled CD14 and FITC labelled CD63 monoclonal antibodies (Becton Dickinson, Franklin Lakes, NJ USA). Immunocytology was performed to quantify the percentage of inducible NO synthase (iNOS) positive macrophages. Cytospins were double stained with a monoclonal antibody against CD68 (IgG1 isotype, Dako, Glostrup, Danmark) as a marker for macrophages and rabbit polyclonal antiserum against iNOS (Transduction Laboratories, Lexington, KY, USA).
The following soluble mediators were measured in sputum supernatant. NO2
- was measured using the Griess reaction, eosinophilic cationic protein (ECP) using the fluorenzyme immunoassay UniCAP ECP (Pharmacia, Uppsala, Sweden). IL-8 and Leukotriene B4 (LTB4) were measured by a commercial ELISA (IL-8: Sanquin, Amsterdam, the Netherlands, LTB4: Amersham Biosciences, UK). Matrix metalloproteinase-9 (MMP-9) was measured by gelatine zymography , and tissue inhibitor of metalloproteinase-1 (TIMP-1) by ELISA (R&D, Abingdon, UK). Neutrophil elastase (NE) activity was measured by chromogenic substrate assay (N-methoxysuccinyl-ala-ala-pro-val-p-nitoanilide, Sigma, UK)] .
Before inhalation of smoke (or control), a urine portion was collected to measure urine cotinine. Cotinine was measured by gaschromatography-mass-spectrometry (Pharmacy Department, Groningen, the Netherlands). Furthermore, urine was collected over 24 hours in five consecutive fractions: 0–1 hour, 1–3 hours, 3–6 hours, 6–12 hours and 12–24 hours from all subjects to assess leukotriene E4 levels (ELISA, Amersham Biosciences, UK).
Since the start and duration of the acute effects of smoking of two cigarettes on our parameters were unknown, time series of all variables were plotted. Based on visual inspection of these plots the time intervals to be analysed were selected. The slopes of parameters were estimated using linear mixed effect models  by including the variables time (hours), smoking (yes or no) and their interaction. For the sputum parameters no baseline values were present, therefore time point 192 hours was used as baseline value. After log-transformation of all blood and sputum variables, the residuals of the models were normally distributed. All analyses were performed in S-plus 2000 (Insightful Corporation, Seattle, WA, USA). A p value <0.05 was considered statistically significant.