In these case report, the results of late dead space fraction (fDlate), end-tidal alveolar dead space fraction (AVDSf), arterial-alveolar gradient CO2 [P(a-et)CO2], and slope phase 3 of spirogram of two patients who underwent thromboendarterectomy for pulmonary embolism (PE) are shown. PE was diagnosed by pulmonary scintigraphy, helical tomography, and pulmonary angiography. The calculation of fDlate, AVDSf and P(a-et)CO2 was based on volumetric capnography associated with arterial blood gas analysis. The pre-operative fDlate of the first patient was 0.16 (cutoff 0.12) and AVDSf was 0.30 (cutoff 0.15). However, the fDlate of the second patient was false-negative (0.01) but, the AVDSf was positive (0.28). Postoperative fDlate of the first patient was -0.04 and AVDSf was 0.16; for the second patient, the values were 0.07 and 0.28, respectively. The association of these capnographic variables with image exams reinforces the importance of this noninvasive diagnosis method.
Este relato de dois casos com os resultados da fração tardia de espaço morto (fDlate), fração do espaço morto alveolar end-tidal (AVDSf), gradiente artério-alveolar de CO2 [P(a-et)CO2] e slope da fase 3 do espirograma, submetidos à tromboendarterectomia pulmonar por tromboembolismo pulmonar (TEP). O TEP foi diagnosticado pela cintilografia pulmonar, tomografia helicoidal computadorizada e por arteriografia pulmonar. O cálculo da fDlate, AVDSf e P(a-et)CO2 baseou-se na capnografia volumétrica associada à gasometria arterial. A fDlate pré-operatória do primeiro paciente foi de 0,16 (cutoff de 0,12) e a AVDSf = 0,30 (cutoff de 0,15). Já a fDlate do segundo paciente resultou falso-negativa (0,01), embora a AVDSf resultasse positiva (0,28). A fDlate pós-operatória do primeiro paciente foi de -0,04 e a AVDSf de 0,16; a fDlate do segundo paciente foi de 0,07 e a AVDSf = 0,28. A associação destas variáveis com os exames por imagem reforça a importância deste método como ferramenta diagnóstica não-invasiva no diagnóstico de TEP.
It is well-known that unexpected death can occur in consequence of pulmonary thromboembolism (PTE) and anticoagulation is often effective in reducing the possibility of a new embolic event and death. For this reason, in patients in whom PTE is suspected, noninvasive methods would be preferred and available to be incorporated as part of bedside evaluation. Bedside techniques to evaluate patients with PTE are based upon a few respiratory parameters derived from the alveolar dead space. However, these variables have some limitations due to the difficulty to distinguish patients with PTE from those with other chronic obstructive pulmonary diseases (COPD). In order to overcome this difficulty, Eriksson et al.  described a diagrammatical method to extrapolate arterial-alveolar gradient [P(a- ET)CO2] to a late virtual expiration. This variable is called late dead space fraction (fDlate). The authors have performed a study enrolling 38 patients with suspected PTE and observed that fDalte was above than 0.12 in normal individuals, and in patients with COPD fDlate was below than 0.12. Another capnographic variable, the end-tidal alveolar dead space fraction (AVDSf)  calculate from the equation PaCO2 - PetCO2 / PaCO2, "corrected" fDalte falsenegative result. In the present two-case report, it was possible to correlate the result of pulmonary artery pre and post-thromboendarterectomy scan with fDlate, AVDSf, and the CO2 arterial-alveolar gradient [P(a-ET)CO2].
A male patient, aged 69-year-old, was admitted to general ward with a history of dyspnea, palpitation, and dry cough, and later referred to the ICU; he denied pneumopathy and smoking. Physical examination revealed arrhythmic pulse and subtle hepatomegaly. Arterial blood gases revealed hypoxemia (PaO2 = 48.8 mmHg) and hypocapnia (PaCO2 = 31mmHg). Echocardiogram showed mild increase of the right ventricle, pulmonary artery hypertension with an estimated systolic pressure of 76 mmHg and thrombus size of 30 x 20 mm in the right atrium. With PTE as a diagnostic hypothesis, a complete anticoagulation with unfractioned heparin was started. The patient was maintained on oxygen therapy through an air-entrainment (Venturi) mask. Ventilation/perfusion pulmonary scintigraphy showed multiple hypoperfusion areas compatible with difuse pulmonary embolization. Helical computed tomography (HCT) showed the presence of thrombi in the right and left pulmonary arteries spanning to their posterior segmental arteries. Taking into consideration the patient's history of over a two-week evolution and the maintenance of hemodynamics and clinical stability, extensive chronic embolism was suspected and chemical thrombolysis was contraindicated. After hemodynamic evaluation, a pulmonary thromboendarterectomy was performed .
A 42-year-old male heavy smoker patient (30 cigarettes a day over 15 years) presented on physical examination antiphospholipid antibody syndrome dyspnea to minor exertions and normothermia. Arterial blood gases revealed hypoxemia with PaO2 = 64.9 mmHg, PaCO2 = 31.4 mmHg, and sO2 = 92.9%. Echocardiogram showed pulmonary artery hypertension with an estimate systolic pressure of 75 mmHg and a large thrombus (17 x 67 mm) adhered to right ventricle (RV) anterior wall occluding its outflow tract. Pulmonary scintigraphy showed perfusion failure on right and left lungs basal segments. HCT showed right ventricular filling pressure failure compatible with left descending interlobar artery thromboembolism with areas of pulmonary infarction in lower lobe of left lung. This patient was considered for surgical treatment because of right ventricle occlusion, besides presenting right pulmonary artery thromboembolism, which was highlighted by arteriography. Left pulmonary artery was embolus-free.
Volumetric capnography was recorded using a respiratory profile monitor CO2MO Plus 8100® (Dixtal/ Novametrix). Capnographic data were recorded for a 3 to 5- minute time at room temperature. Arterial blood gases were obtained afterwards . fDlate was calculated after determining late PetCo2, i.e., extrapolated to 15% of Total Lung Capacity (TLCp) according to Eriksson et al.  using the following equation:
fDlate = PaCO2 - Pet(15% TLCp)CO2/PaCO2
where PaCO2 is the partial pressure of carbon dioxide tension of arterial blood; Pet (15% TLCp) CO2 is the partial pressure of CO2 in the air expired extrapolated to 15% TLCp. TLCp (Total Lung Capacity) is obtained through the previous published Table and it is based on patient's age, weight, and height .
AVDSf  was calculated using the following equation:
AVDSf = PaCO2 - PetCO2/PaCO2
On patient 1, pulmonary arteriotomy with thromboendarterectomy of its right and left branches was performed using cardiopulmonary bypass support and deep hypothermic circulatory arrest. After arteriorrhaphy and warming of the patient, the cardiopulmonary bypass was interrupted. The patient was discharged from hospital on postoperative day 6. On patient 2, a bulky thrombus with a diameter size of nearly 70 mm, which was occluding the right ventricle outflow tract, was removed and right pulmonary artery thromboendarterectomy using cardiopulmonary bypass support without deep hypothermia. Patient was discharged from hospital on postoperative day 13. Before hospital discharge both patients underwent a new pulmonary perfusion scintigram, which highlighted a significant improvement on patient 1. On patient 2, it was observed a subtle significant improvement in middle and superior thirds of both lungs; worsening in inferior lobe segments, and improvement of dyspnea on minor exertion. fDlate, AVDSf, P(a-et)Co2, and slope 3 pre and postoperative values can be observed on Table 1.
Verschuren et al.  observed that after chemical thrombolysis to PTE, Slope 3 value increased, what reflected perfusion improvement and consequent reduction of both dead space fraction and P(a-et)CO2. In another study, Thys et al.  observed AVDSf and P(a-et)CO2 improvement after the same treatment. With surgical treatment, we observed the same as for patient 1 with confirmation by control scintigram as for patient 2. Patient 2 presented capnographic consistent variables when compared to control scintrigram, even though clinical improvement was referred.
In summary, patient 1 presented a PTE diagnosis by fDlate e AVDSf values, which were confirmed by preoperative images and improvement of postoperative images as well as the capnographic variables. Patient 2 presented a PTE diagnosis by both two capnographic variables studied and preoperative image methods; however, on postoperative period, fDlate presented improvement and AVDSf did not, which was confirmed by image methods. Consequently, there was a false fDlate improvement. Thus, the two capnographic variables should be used asd a whole to achieve a specificity and sensitivity method improvement.
Paschoal et al. , comparing capnographic data of a sample comprising 108 patients with suspected PTE and 114 healthy volunteers, identified a volunteer Slope 3 value of 7.80±2.36. The first patient presented a mean Slope 3 value (Table 1) slightly higher than the control group, what can explain the fact of AVDSf variable turns to be higher when compared to fDlate. The second patient presented a double Slope 3 value, what explains the fact of a negative fDlate. fDlate is direct related to Slope 3, what does not happen to AVDSf and P(a-et)CO2 variables. Such a situation suggests that when a patient presenting Slope 3 higher than the control group value, AVDSf e a P(a-et)CO2 must be considered, in order to avoid a likely fDlate false-negative outcome. In any case, by applying the two variables, a 100% PTE sensitivity was obtained.
In a recent study using clinical data D-dimer and AVDSf, Rodger et al.  stated that by applying two variables of this protocol in patients presenting to the Emergency Room with PTE suspected, one can eliminate the need for diagnostic imaging by 36%.
On the other hand, in developing countries, the diagnostic imaging test is not always available. For this reason, noninvasive methods to rule out the possibility of PTE would considerably reduce the number of patients unnecessarily undergoing diagnostic imaging test even in hospitals where these tests are available. Noninvasive tests could be used also in small hospitals where these diagnostic imaging tests are not available aiming at to select the patients to be transferred to a referral facility to perform the diagnostic imaging test.
In the present case reports, it was presented two patients with PTE diagnosis confirmed by diagnostic imaging tests and volumetric capnography whenthe later was judiciously used taking into consideration the information known up to the present time.
No doubts, the small number of patients (n=2) is a limitation, but what is being discussed in the present study is a diagnosis proposal that need to be better study in the future.