The Age of Non-Invasive Measurements: Echocardiographic Equations Methods toDetermine Variables in the Pulmonary Vascular Reactivity Test

Introduction and objectives: Right Heart Catheterization (RHC) is a gold method to perform the Pulmonary Vascular Reactivity Test (PVRT), however echocardiography has proven effective in estimating different hemodynamic variables. This study to demonstrate that with the Doppler echocardiographic index: Tricuspid Regurgitation Maximum Velocity (TRV)/Time-Velocity Integral of the Right Ventricle Outflow Tract (TVIrvot) we will obtain a Pulmonary Vascular Resistance Value (PVR) comparable to the invasive method. We evaluate its applicability in the PVRT. Methods: Prospective, double-blind, observational study in 30 patients, performed in 2 stages: A) We obtained simultaneous measurements of PVR by invasive method and echocardiography. TRV/TVIrovt index measurements were correlated with invasive PVR using the analysis of linear regression. An equation was modeled to calculate PVR in Wood Units (WU) by echocardiography; the results were compared with invasive RVP measurements using the Bland-Altman analysis. B) TRVP was performed by simultaneous measurement of PVR and Pulmonary Arterial Mean Pressure (MPAP) with the echocardiographic and invasive method. We use the equation formulated by Abbas (RVPa) and another modified one proposed in this study (RVPt). Results: In the first stage, the TRV/TVIrotv ratio correlated well with invasive PVR measurements (R 2=0.92). The Bland Altman analysis using RVP=10 × VRT/TVIrvot+0.36 showed satisfactory agreement limits (mean ± 0.36, L: 0.12-0.61). In the TRVP there was a high correlation of PVR between the two methods using the two equations. We also found high correlation with MPAP. Conclusions: The Doppler echocardiogram offers a reliable and non-invasive method to measure the PVR and MPAP in the PVRT.


Introduction
Pulmonary vascular resistance is a hemodynamic variable very useful for the management of patients with advanced pathologies cardiovascular and pulmonary, to evaluate the therapeutic response in patients with congestive heart failure [1,2]. Evaluation of patients candidates for cardiac and liver transplant [3][4][5][6][7], is a prognostic parameter for patients with congenital diseases [8], is considered a critical parameter for evaluating PH independent of etiology [9]. The PVR is calculated invasively by the relation of the pressure gradient and the flow transpulmonar [10], Abbas and collaborators [11,12] proposed to estimate the measurement echocardiographic of the PVR by the relation TRV/TVI rvot representing the method most used currently, among others proposed for its calculation [13][14][15][16][17][18]. On the other hand, the PVRT has very precise indications according to the European Society Cardiology guidelines [9] for the diagnosis, treatment and prognosis of the PH. Right catheterization is considered as the gold standard for the definitive diagnosis of PAH and the method of choice for PVRT [19]. There aren't reports of studies comparing in the PVRT simultaneous measurements of PVR and MPAP with the invasive method and the Echocardiography Doppler. In this investigation we raised 3 objectives: 1) Revalidate Abbas's equation in a group of patients [20] with indication of Right Heart Catheterization (RHC) diagnosis. 2) Compare the invasive and non-invasive measurements of MPAP and PVR in a group of patients who underwent the PVRT and verify the effectiveness of the echocardiographic measurements. 3) Demonstrate in this group of patients the utility and efficiency of a novel Echocardiographic equation to calculate the PVR when compared with the Abbas's equation [10] and the RHC.

Echocardiography
The Doppler and two-dimensional measurements were performed with the Philips Sonos 7500 S3 and Philips IE33 SE1 equipment According to American Society Cardiology guidelines [20]. The TRV/TVI rvot index was determined by Doppler echocardiography using the view of the short parasternal axis at the level of large vessels. The TVI rvot (cm) was obtained with pulsed Doppler by

Methods Population and sample
It's a prospective, double-blind, observational study that was divided into two stages, the first part was selected 20 patients. 10 patients with admission to Intensive Care Unit (ICU) and placement of Swan Ganz catheter for presenting medical pathologies that merited measurement of hemodynamic variables; 10 patients in the Center's cardiology department consultation or hospitalization, with indication of right catheterization to confirm diagnosis clinical and echocardiographic of PH. No distinction was made for age group, sex, comorbidity, or medication was omitted ( Table 1). All patients entered our center in the period from February to August 2008.

Exclusion criterion
Regurgitation tricuspid >2+. The clinical, hemodynamic and demographic characteristics are shown in table 1.
In the second part of the study were included 11 patients referred to the medical consultation, of pneumonology, pediatric cardiology and adult cardiology of the same center with presumptive diagnosis of pH, of different etiology, degree of severity and echocardiographic criteria for Perform the pulmonary vascular reactivity test. These patients were studied from July to December 2009 (Table 2).
Exclusion criterion: Patients with hemodynamic instability, chronic or acute hypoxemia, coagulation disorders.
One patient was removed from the study because of the inability to perform measurements with the catheter and in 2 patients only one measurement was made. All patients in the Department of Cardiology (both stages) signed informed consent. Table 3 shows the values obtained in the 10 patients of the PVRT of the TRV/TVI rvot index, the first Abbas equation, applying the Abbas algorithm and RHC.    In both parts of the work, measurements were performed 3 times and averaged. The measurement of the Ejection Fraction (EF) was determined by the method of Simpson [20]. Measurement of the Right Ventricle Diameter (RVD) was done in the long parasternal axis by evaluating the proximal or subvalvular RVOT measured from the angle formed by the junction of the aortic ring with the basal septum and the free wall of the RV at the end of the diastole [20].

Right catheterization
In the intensive care unit, patients in the first group [10] were placed Swan Ganz catheters arrow, 6 or 7 French (F) to obtain pulmonary pressure and flow measurements. The Seldinger technique was used to approach the Internal Jugular Vein (IJV) or Subclavia [24]. Cardiac Output (CO) was determined by the technique of thermodilution and PVR with the equation: PVR: MPAP-PCP/CO [24][25][26][27]. In the Hemodynamics laboratory, Judkins catheter 6 F was used in all patients, with femoral venous approach, guided by fluoroscopy. The PVRT was performed in the hemodynamics laboratory with central venous approach and placing, Swan Ganz catheter. 20 micrograms of inhaled illoprost were administered with simultaneous measurements of the variables of TRV, TVI rvot , MPAP. CO and Pulmonary Capillary Pressure (PCP) were obtained by thermodilution and wedging respectively. Measurements were performed at 0, 5, 10, 15, 30 minutes and recovery [28,29].
We take as a criterion of positivity of the PVRT a decrease >20% of the MPAP and the PVR without diminishing the CO [30].

Statistical analysis
We used software MedCalc statitic 2019 version 18.11.3/14.0 SPSS and Real statistics Excel. Performed linear regression analysis between the invasive method PVR measured in Woods Units (WU), and noninvasive TRV/TVIrvot in both parts of the study. This analysis was also used between the values of PVR obtained by echocardiography using the algorithm of Abbas AE, et al. [12] and the equation proposed in this work. The Pearson correlation coefficient was determined in all cases, and a regression equation was derived. The calculated values were then studied using the Bland-Altman analysis. Using the ROC curve, a cutting value was obtained for TRV/TVI rvot and PVR (catheterization) to predict high values of PVR>4.5 WU with balanced values of sensitivity and specificity [31].
The images were reassessed to quantify the reliability of the intraobserver and interobserver. The Intraclass Correlation Coefficient (ICC) was determined for variability inter and intraobservers; the relationship between the force of agreement and ICC were analyzed. The Kappa coefficient was also estimated to assess the degree of interobserver concordance.

Results
The first group found analyzed a PCP greater than 12 mmHg was found in 7 patients. 10 patients presented left heart disease of different etiology ( Table 1). The linear correlation analysis between PVR cath and TRV/TVI rvot , showed high correlation (R 2 =0.95.95% CI). The equation derived from linear regression was PVR echo =10 × TRV/TVI rvot +0.36 ( Figure 1).
Applying the analysis of Bland-Altman, the values of PVR of this equation showed satisfactory limits in accordance with the PVR cath , with average value 0.36 (SD: 0.13-0.6) ( Figure 2).
In the second group analyzed ( In the second stage of the study, the analysis of linear correlation between PVR cath and TRV/TVI rvot basal showed good correlation (R 2 =0.70, 95% CI) that was exceeded by excluding the highest values of PVR cath (R 2 =0.92, 95% CI) (Figures 3 and 4).       There was also a report of high correlation between the two methods by excluding the 3 highest values at 30 minutes of the test and in the recovery stage (R 2 =0.89 in both cases) (Figures 5 and 6). Prior to exclusion, the correlation coefficient was R 2 =0.56 with the first equation of Abbas and R 2 =0.64 when applying its algorithm (Figure 7). The Youden's index was estimated obtaining a cut value in the ratio TRV/TVI rvot of 0.24 provides 95% sensitivity and 100% of specificity to determine PVR cath of 4.67 WU in the basal stage of the PVRT ( Figure  8). Receiver Operating Characteristic (ROC) curve. In addition to the above, it shows AUC of 1.95% confidence interval between 0.692-1 and significance nivel p<0.0001 (Figure 9).

Sci Forschen
We found very high correlation (R 2 =0.97) when comparing values obtained from PVR by the Abbas algorithm (if TRV/TVI rvot <0.275 index, PVR echo =TRV/TVI rvot × 10+0.16. If TRV/TVI rvot ≥ 0.275 index, PVRABBAS=TRV 2 /TVI rvoy × 5) and the equation proposed in this work: PVRt=(TRV × RVD/TVI rvot ) × Cardiac output correction factor (cfCO TVI rvot ) ( Figure 10). The correlation obtained when both equations were compared with the values of RHC were high (R 2 =0.78 and R 2 =0.84, respectively) (Figures 11 and 12) Bland Altman's analysis showed satisfactory agreement limits between the methods with average differences similar between the methods (Figures 11 and 12).      cases, with average differences between the RHC of: 2, -0.75 and -1 ( Figure 14). The TRVP reported negative in all patients. The CCI was 0.78 between observers and 0.90 observers with excellent agreement force. The Kappa coefficient inter observer was 0.83 with an almost perfect concordance force.

Discussion
15 years have passed since Abbas AE, et al. [11] published a simple Doppler echocardiographic equation: PVR=0.16+TRV/TVI rvot × 10. As we know, the PVR is directly related to the Pressure Gradient (∆P) and inversely with cardiac output (CO), according to this author; its equivalents echocardiographic would be represented by the TRV as a measure of pressure gradient and TVI as a Cardiac flow measurement. Unfortunately the work presented had a great limitation because it excluded all patients with moderate and severe tricuspid regurgitation representing the population that could be most favored. In successive years, emerged publications aimed at giving validity to this equation [32][33][34][35][36], discuss its content 16-18 or establishing different approaches and hemodynamic variables for the calculation of PVR [13][14][15][16]37,38].
One of the works proposed by another variable echocardiographic to determine the PVR is that presented by Gurudevan SV, et al. [14], they concluded that there is an inverse correlation between systolic velocity of Tricuspid Annulus (tSm) and PVR, determining that a velocity <10 cm/s equals a PVR cath >12.5 UW. Haddad F, et al. [13] proposed an index to obtain PVR echo =SPAP/(HR × TVI rvot ) where a cut value of 0.076 provided 86% sensitivity and 82% specificity to determine IPVR>15 WU/m 2 . He also mentioned that patients with elevated values of PVR didn't correlated well using the equation proposed by Abbas (R 2 =0.46).but with the lowest values yes.
An index proposed by Scapellato F, et al. [15], requires the measurement of the Pulmonary Preeyective Period (PPEP), Pulmonary Acceleration Time (PAT) and Total Systolic Time (TST) (PVR=PPEP/PAT/TST; reported high correlation with PVR obtained by catheterization Cardiac(R 2 =0.96). An index>2.6 predicts a PVR>2.5 WU when the resistances are between 0-8 WU. This equation didn't apply in the work of Haddad F, et al. [13] by low correlation (R 2 =0.30).
Vlahos AP, et al. [32] conducted a prospective study in 12 patient's candidates for liver transplantation and obtaining PVR by right catheterization. They analyzed the index TRVTVI rvot and TRV/TVI rvot corrected by the diameter of the RVOT, finding that both correlated well with PVR cath (R 2 =0711 and R 2 =0731, respectively), even in patients with elevated values of PVR. They concluded in their study that the relationship TRVTVI rotv with a value of 0.38 can provide a specificity of 100% for PVR of 8 WU.
Opotoswky AR, et al. [17], in 2013 published a paper where they derive 2 equations to estimate the PVR from the relationship between SPAP and TVI rvot , validated these equations and compared with the equation of Abbas AE, et al. [11] (model 1). The derived models were: PVR=1.2 × SPAP/TVI rvot (Model 2) and PVR=(SPAP/TVI rvot )+3, if the systolic notch is present (Model 3). They found that model 1 consistently underestimated the PVR estimated by catheterization, especially for those with high PVR. The derivative models showed no bias, model 3 was best correlated with PVR cath (R 2 =0.80 vs R 2 =0.73 and R 2 =0.77 for Models 1 and 2, respectively). This approach caused discomfort in Abbas AE, et al. [11], because precisely in that year, they published an article 12 proposing a new equation: PVR echo =TRV 2 /TVI rvot × 5-0.4   Two relevant findings were evidenced with the calculations: 1. The measurements of MPAP obtained by this formula were sustained similar to those reported by the catheterization in all phases of measurement of the test (t0, T30 and TR) with high linear correlation (t0 R 2 =0.87, 0.99 and 0.98, respectively). 2. The correlation coefficient obtained when comparing the PVR echo and PVR cath values ascended from R 2 =0.70 to R 2 =0.78 when we applied the Abbas algorithm and in the cases with VRT/TVI rvoy index ≥ 0.275 it increased even more when omitting the highest values , R 2 =0.92 (t0), R 2 =0.89 (t30 and tR). The cut-off point (TRV/TVI rvot ≥ 0.275), specifies when in the linear regression analysis the correlation ceases to be linear to define the relationship between variables by an exponential growth curve, expressing the basis of the Abbas algorithm in its second work.
It's evident analyzes these finding that the TVI rvot variable of the TRV/TVI rvot index, is responsible for opening the widest gap between invasive and non-invasive methods. Very possibly by not taking into account the dilation of the right ventricle present in all cases of patients with elevation PVR [40,41]. It's thus that this work proposes an equation with reasoning more pathophysiologic than statistical, incorporating in its calculation. We also include a correction factor of the CO applicable in all patients according to the value of TVI rvot that we call cfCOTVI rvot suggested by the repeated observation of its decrease proportional to the degree of dilatation and systolic dysfunction of the RV. This corresponds approximately to the near value of the CO if we divide the TVI between 2 when its value is below 10 cm, above this value, we will find normal values of CO (4-6 Lt/min) allowing to standardize it to 5. At the other end, TVI rvot values below 6 would correspond to a very low CO estimate with probable severe right heart failure or shock (heart rate (CI) <2.2 lt/min/mt 2 ).
The proposed equation is as follows: The linear regression analysis between the values obtained from PVR with this equation and that of Abbas with his algorithm reported a high correlation, R 2 =0.97, but when comparing the algorithms of Abbas, the PVRt with the RHC, we obtained a greater correlation with our proposal (R 2 =0.84).
Limitations: low number of patients to perform the TRVP, however it is important to note that a total of 50 measurements were made. In all cases where it is not possible to measure the TRV or TVI rvot , it will not be possible to perform the PVR calculations, not representing the case of this investigation.

Conclusions
In this work we conclude: 1. The algorithm of equations proposed by Abbas to determine PVR according to the relationship TRV/TVI rvot is ≥ 0275 is useful and easy to estimate [2]. The equation proposed in this work: PVR echo =(TRV × RVD/TVI rvot ) × cfCOTVI rvot , is practical and applicable in all cases of pulmonary hypertension, representing an alternative to that offered by Abbas AE, et al. [13]. It's possible to perform the PVRT, TRVP through of Doppler echocardiography based on the high correlation found in this work between the values of MPAP and PVR when compared with the invasive method and echocardiographic.

Recommendations
Further studies of TRVP are recommended by comparing both methods (echocardiography, RHC) to give more statistical weight and validity to echocardiographic methods and therefore, the applicability in the test.

Conflicts of Interest
The authors of the present work do not declare conflicts of interest.