There are three main conclusions that follow from the present results. Firstly, the sPESI is more useful for predicting the long term adverse outcomes. Secondly, a combination model of hsTnT levels ≥ 0.014 ng/mL with a sPESI ≥ 1 may be used for the short term mortality risk. Thirdly, sPESI ≥ 1 showed the high negative predictive value (100%) for 30-day mortality but PE patients had various clinical presentations leading to different prognostic and therapeutic approaches. Accurate risk stratification with precise diagnostic tools is of crucial importance.
Right ventricular dysfunction on echocardiography/computed tomography as well as cardiac markers, including troponins and clinical scores, are important tools for accurate risk stratification. There is only one study comparing the sPESI with any troponins(including hsTnT) levels, but this study population was composed of patients with massive PE. Another study only compared the prognostic role of the hsTnT assay and of the sPESI. Thus, comparisons of 2 cardiac troponins (hsTnT and cTnT) combined with the sPESI or alone are limited in the literature.
In our study three patients in the group with negative cTnT levels died in the early period compared to only one patient who died among the groups with negative hsTnT levels. The negative predictive value for the sPESI was 100% for a 30-day mortality. We agree with Lankeit et al. who found a 100% NPV of PE for the 30-day mortality. In addition, no patient in the group with a combination of negative hsTnT levels and sPESI < 1 died. As a result it was found that, although the negative hsTnT level had a similar performance as the low sPESI for predicting adverse outcomes, it was obviously superior to the cTnT assay.
Mortality of PE has been defined to be <1% in patients without RVD on echocardiography or computerized tomography, and also with no elevations in biomarkers. However, advanced age and associated comorbidities may increase the mortality for PE. Therefore, the addition of the sPESI to the biomarkers and to RVD may provide more accurate risk stratification of PE. Any patient with PE and one of the following factors (age over 80, presence of cardiopulmonary diseases and presence of cancer) has been classified as high risk patient, thus causing an overestimation of the risk stratification for these PE patients. Therefore, specificity of the high risk sPESI is lower than that of cardiac biomarkers and of right ventricular dysfunction on echocardiography for PE mortality[17, 19]. The specificity of echocardiography decreases from 78% to 17–22% in PE patients with accompanying comorbidities. However, a favourable prognosis of PE can be easily estimated when the sPESI is low. According to the sPESI model the 30-day overall mortality in the low risk group has been reported to be between 0% and 2%[16, 17]. Fourteen patients with a positive hsTnT assay had a low sPESI in our study. However, none of these PE patients with positive ahsTnT level had an adverse event. Relatively to the good prognosis for PE patients with a low sPESI, the hsTnT assay may be an unnecessary marker, especially in those patients who are in the low sPESI group. Interestingly, in the present study the 30-day mortality was found to be 0.8 for patients that only had an hsTnT ≥ 0.014 ng/mL or a sPESI ≥ 1. Lankeit et al. found a 3.6% adverse outcome in the same patients population. However, one third of their study population had renal failure which could be associated with positive troponin values in that study. Therefore combination models may provide much more information about the prognosis of PE patients and prevent unnecessary interventions in these groups of patients. Outpatient management of low risk patients with PE may improve the quality of life, and provide reductions of cost.
Combinations of prognostic tools including multiple biomarker assay, biomarker plus right ventricular dysfunction (both on echocardiography/tomography) and PESI (not simplified) plus shock index were found to be more predictive than biomarkers and PESI only for the risk stratification of PE patients[21–24]. In patients with hsTnT level ≥ 0.014 ng/mL plus a sPESI ≥ 1 there was a 94% risk of 30 day all-mortality and a mortality rate 12.4%. Lankeit et al. reported the 30 day adverse outcome as 10.4% in the same group. In SWIVTER study it was found that patients with a high sPESI plus any positive troponin test (conventional troponin T or I, highly sensitive troponin T) had a mortality rate of 10.3%. Moreover, the mortality rate was found to be 15.4% in a group where the PESI (not simplified) was combined with troponin-I in another study. It remains unclear whether thrombolysis may improve the early and long-term clinical outcomes of selected normotensive patients with a high risk score and/or with biomarker positivity. The PEITHO trial, a prospective, multicenter, international, randomized, double-blind study is currently comparing thrombolysis with tenecteplasevs placebo in a normotensive patient group with confirmed PE (NCT00639743).
Of course there are some limitations in the present study. Firstly, our study population is relatively small, and secondly, we did not study hsTnT testing at the 3rd hour of admission which has 100% negative predictive value for the exclusion of myocardial infarction. In addition, when interpreting our results it should also be considered that no autopsy was performed. The third concern may be related to the recurrent PE which was diagnosed by the presence of a new intraluminal filling defect or an extension of a previous filling defect on computed tomography pulmonary angiography. Because the whole diagnosis of PE patients was initially confirmed by contrast-enhanced computerized tomographic pulmonary angiography in the present study, we did not use perfusion lung scan which could be feasible for the recurrent PE and for the follow up of those PE patients.