Exhaled breath contains hundreds of volatile organic compounds (VOCs). Several independent researchers point out that the breath of lung cancer patients shows a characteristic VOC-profile which can be considered as lung cancer signature and, thus, used for diagnosis. In this regard, the analysis of exhaled breath with gas sensor arrays is a potential non-invasive, relatively low-cost and easy technique for the early detection of lung cancer. This clinical study evaluated the gas sensor array response for the identification of the exhaled breath of lung cancer patients. This study involved 146 individuals: 70 with lung cancer confirmed by computerized tomography (CT) or positron emission tomography—(PET) imaging techniques and histology (biopsy) or with clinical suspect of lung cancer and 76 healthy controls. Their exhaled breath was measured with a gas sensor array composed of a matrix of eight quartz microbalances (QMBs), each functionalized with a different metalloporphyrin. The instrument produces, for each analyzed sample, a vector of signals encoding the breath (breathprint). Breathprints were analyzed with multivariate analysis in order to correlate the sensor signals to the disease. Breathprints of the lung cancer patients were differentiated from those of the healthy controls with a sensitivity of 81% and specificity of 91%. Similar values were obtained in patients with and without metabolic comorbidities, such as diabetes, obesity and dyslipidemia (sensitivity 85%, specificity 88% and sensitivity 76%, specificity 94%, respectively). The device showed a large sensitivity to lung cancer at stage I with respect to stage II/III/IV (92% and 58% respectively). The sensitivity for stage I did not change for patients with or without metabolic comorbidities (90%, 94%, respectively). Results show that this electronic nose can discriminate the exhaled breath of the lung cancer patients from those of the healthy controls. Moreover, the largest sensitivity is observed for the subgroup of patients with a lung cancer at stage I.

Objective: In this study, we analyze our experience so far with robotic pulmonary lobectomy, compare it with published data, and suggest a learning curve for the operation. Methods: Ninety-one patients with suspected or proven clinical stage I–III lung cancer underwent robotic lobectomy. Selection criteria included lesion ≤5 cm and normal respiratory function. One surgeon performed the operations using the da Vinci system with three ports and a 3-cm utility thoracotomy. Results: Median duration of operation was 239 (range 85–411) minutes, 260 minutes in the first 18 patients and 221 minutes in the remaining 73 cases (P = 0.01). Median hospitalization declined from 6 days in the first 18 cases to 5 days in the remaining cases (P = 0.002). Conversion rate and number of complications reduced nonsignificantly from the initial to later series. Major complications occurred in 11% of the first 18 cases and 4% of the later cases. The number of lymph nodes removed did not change over the two series. There was no 30-day postoperative mortality. After a median follow-up of 24 months, 80 of 91 patients were alive with no sign of disease. Conclusions: Our data suggest that about 20 operations are required to achieve surgical competence. Robotic lobectomy appears safe, oncologically radical, and associated with shorter postoperative hospitalization than open surgery.