Original Article

Aortic Valve Annular SizingCLINICAL PERSPECTIVE

Intraoperative Assessment Versus Preoperative Multidetector Computed Tomography

Isaac George, Laura C. Guglielmetti, Nicolas Bettinger, Andrew Moss, Catherine Wang, Nathan Kheysin, Rebecca Hahn, Susheel Kodali, Martin Leon, Vinayak Bapat, Michael A. Borger, Mathew Williams, Craig Smith, Omar K. Khalique
https://doi.org/10.1161/CIRCIMAGING.116.005968
Circulation: Cardiovascular Imaging. 2017;10:e005968
Originally published May 9, 2017

Jump to

Abstract

Background—Appropriate valve sizing is critical in aortic valve replacement. We hypothesized that direct intraoperative valve sizing results in smaller aortic annular diameters compared with sizing based on systolic-phase multidetector computerized tomographic (MDCT) imaging.

Methods and Results—We retrospectively analyzed 78 patients undergoing surgical aortic valve replacement for severe aortic stenosis between 2012 and 2014 at our institution. Preoperative MDCT measurements of the aortic annulus served as basis for assignment to a theoretical surgical valve size, which was then (1) compared to the implanted valve size and (2) to a theoretical transcatheter aortic valve replacement valve size. To quantify the resulting differences, geometric orifice areas (GOA) were calculated. MDCT-based sizing produced the same valve size for n=34 patients (group CT-same), a larger valve with a 25% increased GOA in n=32 patients (group CT-Lg) and a smaller GOA by 22% in n=12 patients (group CT-Sm). On the basis of MDCT measurements, 41% of valves implanted were undersized. The comparison of intraoperative implanted to a theoretical transcatheter aortic valve replacement valve size resulted in GOAs 25% larger for patients in group CT-same, 40.6% larger in group CT-Lg and 14.6% larger in group CT-Sm.

Conclusions—Preoperative MDCT measurements differ substantially from direct intraoperative assessment of the aortic annulus. Implanted surgical aortic valve replacement valves were smaller relative to MDCT-based sizing in 41% of patients, and the potential GOA was between 25% and 40.6% larger if patients had undergone transcatheter aortic valve replacement.

Over the past 40 years, surgical aortic valve replacement (SAVR) has become one of the most commonly performed cardiac procedures worldwide. In the last decade, transcatheter aortic valve replacement (TAVR) has emerged as an alternative treatment for patients with intermediate and high perioperative risk. For any approach, appropriate valve sizing is paramount to avoid paravalvular leakage, annular rupture, or other complications.1,2 Placing the largest valve that can be safely implanted is advocated to optimize left ventricular mass regression and derive maximal clinical benefit.3,4

See Editorial by Benton and Litwin

See Clinical Perspective

Sizing strategies differ fundamentally between SAVR and TAVR: in SAVR, the surgeon chooses the subjectively best fitting valve size using manufacturer-specific annular valve sizers, whereas in TAVR sizing relies entirely on cardiac imaging. Early TAVR-sizing experience was based on transesophageal echocardiography but ECG-gated multidetector computerized tomographic (MDCT) imaging has gained popularity and has been proven to be the superior, noninvasive option to assess aortic annular size, coronary height, and root geometry.1,5

We hypothesized that direct, intraoperative valve sizing for SAVR results in smaller aortic annular diameter measurements when compared with systolic-phase MDCT sizing, resulting in smaller postoperative geometric aortic valve areas than by preoperative MDCT sizing or TAVR. In this study, we compare direct surgical measurements to hypothetical sizes established through preoperative MDCT scans for patients with symptomatic aortic stenosis undergoing SAVR and quantify the differential area. Then, we calculate the valve areas after implantation of a balloon-expandable valve (TAVR) and compare these valve areas. These data may yield important insights into the real and perceived hemodynamic differences between SAVR and TAVR, yet also highlight how surgical decision making may be improved by advanced preoperative imaging.

Methods

All patients who underwent isolated SAVR with a tissue valve and received preoperative MDCT imaging between June 2012 and December 2014 at our institution were analyzed in this study (n=101). In brief, all patients had symptomatic severe aortic stenosis, had an at least intermediate surgical risk treated as part of the PARTNER II or SURTAVI trial (STS score 3% to 8%),6,7 or treated commercially with SAVR, and the native aortic annulus size was ≥18 mm or ≤27 mm as measured by echocardiography. Important exclusion criteria were patients with bicuspid aortic valve anatomy, end-stage renal disease, cardiogenic shock, active infection, and recent stroke. Additional exclusion criteria for this analysis were aortic root enlargement (n=2) and missing (n=5) or inadequate MDCT images (n=8) and valve-in-valve procedures (n=8). A total of n=78 patients were included in this analysis. Clinical information was collected in patient charts and surgical records; CT images were analyzed for all patients. Postoperative transthoracic echocardiograms were performed before discharge. The study was approved by the Columbia University Institutional Review Board and need for informed consent was waived.

Aortic Annulus Measurements and Calculations

To detect the most precise measurement, end-systolic and end-diastolic MDCT scans were analyzed by a cardiologist experienced in CT imaging and blinded to clinical data. The aortic annulus was defined as the plane of the virtual circumferential ring containing the basal attachment points of the 3 aortic valve leaflets. The MDCT Acquisition Protocol is depicted in the Data Supplement. Ellipticity was assessed using the manually assessed minimal and maximal diameter using the following formula8: 1−(Dmin/Dmax). See also Figure I in the Data Supplement for the full nomenclature and Table IA and 1B in the Data Supplement for all measurements.

Surgical Aortic Valve Replacement

All operations were performed using a standard, median sternotomy approach, using cardiopulmonary bypass with mild systemic hypothermia (30–34°C). Myocardial protection was achieved with del Nido solution (650–1000 mL antegrade single dose) cardioplegia. Intraoperative assessment of the aortic annulus size was conducted after resection of the aortic valve cusps and complete debridement of the calcifications. The annular end of the sizing obturator was inserted into the annulus and the complete fit of the largest possible sizer was defined as optimal valve choice. The replica end of the sizer (where applicable) was used to verify the correct valve size. No patient underwent an aortic root enlargement procedure, as previously noted. The annular end of the valve sizer was measured for each valve (see Table IIa in the Data Supplement); for both Mosaic and Magna-Perimount valve sizers, the annular end of the sizer corresponds to the outer diameter of the stent frame of the surgical valve.

Valve Sizing Scale for MDCT Measurements

Each valve sizer for each manufacturer and valve size were individually measured, and it was found that the outer dimension of the annular portion of the sizer corresponds to the outer stent diameter of a valve and also corresponded to the labeled valve size. Therefore, if an annulus allowed a 21 sizer to be placed, a #21 valve of that manufacturer would be chosen, which represents the outer stent frame diameter in millimeters; the outer stent diameter for both Edwards and Medtronic valves and the outer dimension of their respective valve sizers are equivalent in size. The MDCT measurements served as basis for assignment to a theoretical surgical valve size according to the following scale based on the physical sizer measurements:

MDCT diameter 19.0 to 20.9 mm: valve size 19

MDCT diameter 21.0 to 22.9 mm: valve size 21

MDCT diameter 23.0 to 24.9 mm: valve size 23

MDCT diameter 25.0 to 26.9 mm: valve size 25

MDCT diameter 27.0 to 28.9 mm: valve size 27

The sizing chart above assumes that the annular end of the sizer (the sizer end that fits entirely through the annulus) cannot be forced into an annulus smaller than the sizer. For example, a 21.5-mm annulus would be sized to a size 21 valve rather than a size 23 valve.

After assessment of the strongest correlation of MDCT measurement–derived diameters at the annulus with the implanted valve size and in concordance with recent literature,1,9,10 SAP (Systolic, at Annulus, Perimeter derived diameter) was used for further analysis. See Table IIIA and IIIB in the Data Supplement for results of the correlation and Figure II in the Data Supplement displaying the diameter predicted by MDCT-SAP measurements by implanted valve size.

Patients were then stratified according to suggested valve size by SAP in relation to actual valve size: group CT-same, MDCT valve size, and implanted valve size in accordance, group CT-Lg) larger valve size by preoperative MDCT measurement than actual implanted valve, and group CT-Sm) smaller valve size by MDCT than implanted valve.

Comparison of Valve Sizes: Implanted (GOASAVR) Versus MDCT (GOASAVR-CT) Versus TAVR (GOATAVR)

Although outer stent diameters may be equivalent, valve sizes from different manufacturers have unequal inner stent diameters; therefore, we calculated the geometric orifice area (GOA) and indexed GOA (GOA/body surface area) to quantify the difference between the choice of valve size for intraoperative versus MDCT-based sizing. The GOAs are calculated from the true inner diameter11 of the valves, which is smaller than the reported inner stent diameter because this value does not include the space occupied by the valve leaflets if these are attached inside the valve (see Table IIB in the Data Supplement and compare to internal stent diameters in Table IIC in the Data Supplement). GOA was calculated as follows: (true internal diameter/2)2×π. The GOA is, therefore, specific for each valve brand and size.

In a second step, we determined the SAPIEN 3 (Edwards Lifesciences, Irvine, CA) valve size for each patient based on SAP-derived measurements and compared the theoretical difference with corresponding values of the implanted surgical valve. See Table IIE in the Data Supplement displaying the mean nominal GOA for each TAVR valve size.

The SAPIEN 3 is a balloon-expandable stented valve used for TAVR. We assumed complete expansion of the chosen valve through balloon dilation.12 As opposed to SAVR, the SAPIEN 3 valve will theoretically expand to the size of the native aortic annulus dimensions (as assessed by MDCT). We calculated the achieved area by TAVR accordingly (GOA derived from SAP diameter, GOATAVR) and compared these values with the GOAs from the implanted valve (GOASAVR) and the theoretical SAVR valve after MDCT sizing (GOASAVR-CT). To account for true inner stent diameter of the SAPIEN 3 valves, we assumed a total leaflet and frame thickness of 2 mm (reducing SAP d for each SAPIEN 3 valve size by 2 mm), and GOATAVR was calculated as follows: ((SAP d−2 mm)/2)2×π. See Table IIE in the Data Supplement.

Statistical Analysis

Descriptive statistics were used to summarize patients’ characteristics. Continuous variables were reported as mean and SD or median and range or interquantile range and compared between the 2 groups using 2-sample independent t tests or Mann–Whitney U test (non-normal data). Categorical variables were summarized as frequencies (%) and compared using Pearson χ2 test or Fisher exact test where applicable. For pairwise comparisons, a paired t test or a Wilcoxon test was used. The Pearson correlation of all diameters acquired by MDCT imaging with the intraoperatively chosen valve size was assessed. SPSS version 22 (IBM Corp, Armonk, NY) was used for data analysis. P values ≤0.05 were considered statistically significant; group CT-Lg and CT-Sm were compared with group CT-same as reference, all P values are labeled as such (CT-same versus CT-Lg or CT-same versus CT-Sm).

Results

Patient Demographics and Baseline Characteristics

As displayed in Table 1, baseline demographics and comorbidities were comparable between the groups. No patients underwent previous cardiac surgery. The Mosaic valve (Medtronic, Minneapolis, MN) was predominant in group CT-Sm, whereas the Carpentier Edwards valves were the most commonly implanted valves in groups CT-same and CT-Lg (Figure 1).

View this table:
Table 1.

Baseline Demographics

Figure 1.
Figure 1.

Valve type according to groups. In groups CT-same and CT-Lg Carpentier Edward valves (Perimount, Magna, and Magna Ease) were predominantly used (Table 1), whereas group CT-same consists of patients who received mostly Medtronic valves (Mosaic and Mosaic Ultra). CT indicates computed tomography.

Intraoperative Surgical Versus MDCT-Based Valve Sizing

The GOAs based on implanted surgical valve size (GOASAVR) and indexed GOASAVR were comparable across all groups (Table 2). The mean diameter (SAP) for group CT-same, CT-Lg, and CT-Sm in mm was 23.85±2.04, 25.42±2.26, and 22.16±0.99. The GOA for a given valve sized by MDCT (GOASAVR-CT) for group CT-same, -Lg, and –Sm, respectively, was as follows: 3.06±0.69, 3.90±0.83, 2.21±0.30, and indexed GOASAVR-CT was 1.68±0.34, 2.09±0.35, and 1.18±0.25. The implantation of the same valve type was assumed for GOASAVR-CT, and only the sizes were changed according to MDCT measurements. As shown by the ratio GOASAVR-CT/GOASAVR, MDCT-based valve sizing produced the same-sized valve in 34 patients (group CT-same, GOASAVR-CT/GOASAVR=1.0), a larger sized valve with a 25% increased GOA in 32 patients (group CT-Lg, GOASAVR-CT/GOASAVR=1.25), and a smaller GOA by 22% in 12 patients (Group CT-Sm, GOASAVR-CT/GOASAVR=0.78, all P<0.001). In conclusion, >41% of valves implanted using intraoperative sizing techniques were undersized, based on MDCT measurements.

View this table:
Table 2.

Valve Sizing (SAP): Implanted, MDCT-Based SAVR, and MDCT-Based TAVR Sizing

Comparison of Valve Sizes: Implanted Versus MDCT Versus TAVR

Next, we sought to quantify the calculated potential difference in GOA when comparing direct surgical intraoperative measurements of valve size, to a theoretical SAPIEN 3 TAVR valve, assuming that the SAPIEN 3 can expand to the full patient native annular size (as measured by MDCT: SAP). The hypothetical GOA after implantation of a transcatheter SAPIEN 3 valve (GOATAVR) compared with the GOASAVR is 25.3% larger for group CT-same, 40.6% larger for group CT-Lg, and 14.6% larger for group CT-Sm (GOATAVR/GOASAVR; Table 2). Notably, GOATAVR was 4.96% to 9.06% smaller than GOASAVR for n=3 patients in Group CT-Sm. See Figure 2A through 2C for graphical comparison of GOA SAVR, GOA SAVR-CT, and GOA for each group.

Figure 2.
Figure 2.

A–C, Box-and-whiskers plot displaying GOASAVR, GOASAVR-CT, and GOATAVR for group CT-same, -Lg and -Sm, respectively. In (A), as previously defined for group CT-same, GOASAVR equals GOASAVR-CT. (*P<0.05 vs GOA SAVR, #P<0.05 vs GOA SAVR-CT). CT indicates computed tomography; GOA, geometric orifice area; SAVR, surgical aortic valve replacement; and TAVR, transcatheter aortic valve replacement.

The mean difference between GOASAVR and GOATAVR is 0.871. Considering that the resulting GOA depends not only from the chosen valve size but also from the implanted valve type (Medtronic versus Carpentier Edwards), the increase of one valve size leads to a GOA that is 0.25 to 0.82 cm2 larger depending on valve size and type.

The mean difference of GOASAVR to GOATAVR equals a valve between 1.1 (0.871/0.82) and 3.5 (0.871/0.25) valve sizes larger than the implanted valve after intraoperative sizing (Table IIE in the Data Supplement for mean GOATAVR per SAPIEN 3 valve size).

Echocardiographic Findings

The preoperative and postoperative echocardiographic (Echo) data for all patients are shown in Table 3. Preoperative left ventricular ejection fraction was significantly higher in group CT-Sm compared with group CT-same (P<0.001) but was equivalent at the predischarge Echo (group CT-same versus CT-Sm, P=0.109). All other Echo data were comparable between the 3 groups, in particular preoperative and postoperative valve gradients (postoperative mean gradient [mm Hg], mean±SD: group CT-same 13.5±5.4 versus group CT-Lg 14.1±5.8 versus group CT-Sm 14.3±4.5; P value CT-same versus CT-Lg=0.661 and CT-same versus CT-Sm=0.660). Postoperative peak and mean gradients were similar between Medtronic and Edwards Lifesciences’ valves (peak 28.6±2.2 versus 25.7±1.7 mm Hg; P=0.27, mean 14.6±1.0 versus 13.5±0.9; P=0.43, respectively; data not displayed in tables). For comparison, the mean gradient from the 1-year PARTNER 3 High Risk Registry for the SAPIEN 3 valve was 12.9 mmHg for a size 23 valve, 10.6 mmHg for a size 26 valve, and 9.1 mmHg for a size 29 valve at 30 days13; the mean gradient from the 3-year CoreValve Pivotal High Risk Trial similarly showed a mean composite gradient of all valve sizes of 7.62±3.57 mm Hg.14

View this table:
Table 3.

Echocardiographic Data

There was no significant difference in ellipticity between the groups.

Postoperative Outcomes

There was only one case of mild paravalvular leak and one patient who died within 30 days after the procedure, both in group CT-Sm. All postoperative variables, including postoperative cerebrovascular accident and postoperative acute kidney injury, were comparable between the 3 groups. Postoperative outcomes are displayed in Table IV in the Data Supplement.

Discussion

Optimal valve sizing is paramount in aortic valve replacement—be it SAVR or TAVR—for procedural success and optimal hemodynamics. Patients undergoing both conventional surgical and transcatheter valve replacement may suffer clinical consequences of annular undersizing, specifically paravalvular leak, leaflet dysfunction, worsened hemodynamics, or possible patient-prosthesis mismatch. Conversely, significant annular oversizing may result in inability to seat the valve or coronary obstruction in SAVR, or central regurgitation or annular rupture in TAVR.15 The importance of annular sizing in TAVR to achieve optimal outcomes has been widely established in the literature and in clinical practice,16 and MDCT has currently become the standard of care for optimal annular sizing. Systolic-phase MDCT can be used to determine the optimal deployment area preprocedure; however, the same sizing imaging algorithm and rigor has not been applied to patients undergoing SAVR. As a result, optimal sizing in SAVR has relied on intraoperative measurements after calcium debridement, which may or may not correlate with MDCT-based sizing. The ability to optimally size surgical prostheses has taken on much greater importance in recent years after recognition that hemodynamics after SAVR may be inferior to TAVR, mainly because of geometric constraints of the prostheses themselves. The primary objective of the current study was to provide a detailed anatomic, comparative analysis of annular sizing of patients undergoing SAVR using both preoperative MDCT and intraoperative annular sizing, with the hypothesis that intraoperative sizing may yield smaller annular sizes compared with MDCT-based sizing.

In the current study of a cohort of SAVR patients with preoperative MDCT imaging, we found that (1) 43.6% of patients received valves of the same size as predicted by MDCT (group CT-same), 41% of patients received a valve smaller than that predicted by MDCT (group CT-Lg), and 15.4% of patients received a valve larger than that predicted by MDCT; (2) systolic perimeter derived D at the nadir of the annulus (SAP) provided the closest estimation of valve prosthesis chosen in both TAVR and SAVR; (3) no short-term clinical or echocardiographic consequence was apparent from undersizing surgical valves; and (4) theoretical placement of a balloon-expandable valve expanded to native patient annular area was associated with significantly larger GOA than either intraoperative valve sizing or MDCT-based SAVR sizing in all patients.

For quantitative comparisons of the resulting differences in size among different valve brands and between SAVR and TAVR valves, we calculated the respective GOAs from true inner diameters.11 The differences in GOAs if TAVR valves were implanted are significant between all groups, and the mean achievable GOAs are significantly larger compared with the GOAs by SAVR. Interestingly, GOATAVR was slightly smaller than GOASAVR-CT for 3 patients in group CT-Lg and between 4.96% and 9.06% smaller than GOASAVR for another 3 patients in group CT-Sm. GOATAVR in group CT-same was 25.3% larger, and 40.6% larger in group CT-Lg compared with GOASAVR. Remarkably, even in group CT-Sm where MDCT suggests a smaller valve size, the mean GOATAVR was still 14.6% larger than GOASAVR. The differences in GOATAVR to GOASAVR correspond to 1.1 to 3.5 SAVR valve sizes. Group CT-Sm consists of the patients who would have received a smaller valve by MDCT sizing (GOASAVR-CT <GOASAVR), and most patients in this group (n=10, 83.3%) had Medtronic valves implanted (Table 1). These have a smaller GOA per valve size compared with the Carpentier Edwards valves used in our study population. Therefore, the resulting GOATAVR/GOASAVR-CT ratio is largest for group CT-Sm (Table 2; Table IIA through IIE). Although the immediate postoperative echocardiographic parameters were comparable between the groups with the implanted valve sizes, the clear differences in GOATAVR suggest significantly better hemodynamic results, if these patients had undergone TAVR and were sized by MDCT.

In our findings, MDCT measurements of the aortic annulus differ substantially from intraoperative direct measurements, and intraoperative evaluation of the aortic root dimensions resulted in smaller dimensions in most cases. This can be explained by many factors. First, the complex shape of the aortic annulus is not assessed in its physiological state intraoperatively, because the left ventricle is arrested and decompressed. Precise intraoperative sizing can depend on numerous structural conditions including angle of the aorta, adequacy of exposure, compliance of the annulus, left ventricular outflow tract calcium, and degree and quality of calcium debridement. A frequent limitation in intraoperative valve sizing is the presence of a calcified sinotubular junction or aortic sinus wall, which prevents the rigid valve sizer from entering the sinus without risk of aortic injury; instead, the surgeon may choose a smaller valve to minimize surgical risk. Second, there are key differences in the exact location at which intraoperative annular sizing occurs. The aortic annulus complex is a 3-dimensional crown-shaped fibrous structure defined by the outline of each leaflet cusp (Figure III in the Data Supplement). Intraoperatively, the annulus is identified as a ridge of fibrous tissue onto which sutures are placed. The annular portion of the sizer is grossly inserted into the annulus such that the sizer just fits beyond this plane bounded by both the commissures and nadir of the ridges. In contrast, a basal ring is defined for MDCT TAVR sizing as a virtual ring located in a single plane bounded by the nadir of each cusp. This plane can be located multiple millimeters below the surgical fibrous annulus described above. It is likely that the surgical annulus measured by the annular sizer and the MDCT-defined basal ring do not correspond in many patients, such as those with severe calcifications of the left ventricular outflow tract, eccentric calcium, or a severely elliptical annular shape. Moreover, the surgical sizer likely underestimates the true annular perimeter by virtue of its inability to accurately account for coverage of perimeter that exists at the top of each commissure, above the basal ring (Figure IV in the Data Supplement). Third, the discrete, ordinal nature of SAVR valve sizing further exacerbates annular undersizing: valve sizes are typically provided in sizes 19 to 29 mm with valves available every 2 mm. For native annular sizes in between a particular valve size, the default practice for many cardiac surgeons is to choose the smaller size rather than attempt to force a larger valve. Both balloon-expandable and self-expanding TAVR delivery systems, however, allow expansion of the valve frame to at least native annular area/perimeter in most cases (dependent on calcium burden) and is not bound by manufacturing constraints. This difference, as well as the absence of a thick sewing ring that occupies vital space in the inner stent frame, accounts for a large part of the discrepancy in GOA between SAVR and TAVR. Finally, the phenomenon of stretching an aortic annulus is not possible in SAVR with a rigid frame, other than to perform an aortic root enlargement procedure, whereas postdilatation of TAVR routinely is performed to oversize a valve. This is particularly relevant in the SAPIEN 3 balloon-expandable valve, which allows for overexpansion via foreshortening of the cobalt chromium frame. ECG-gated MDCT is comparable to a live imaging of the aortic annulus, whereas general anesthesia, calcified sinotubular junction, and the limitations mentioned above of a rigid sizer do not allow such precise measurements in a physiological state of the heart. To account for the fact that the sewing ring of the surgical valve can rest higher in the aortic sinus and thus allow a larger valve, the aortic annulus at 2 mm above the anatomic annulus was also measured and included in the correlation between the method of annular sizing (area, perimeter, and manually assessed) and surgical valve size to see what method showed strongest correlation with intraoperative sizing (see Data Supplement).

In our cohort, 15.4% of patients received a smaller valve when compared with MDCT measurements (group CT-Sm), whereas 41% received a larger valve size (group CT-Lg). Review of the literature reveals few studies on this subject that are limited by small case numbers, lack of advanced imaging, and reports before the advent of TAVR. Dashkevich et al17 found no systematic difference between intraoperative diameters assessed with a Hegar dilator and MDCT measurements of the aortic annulus by calculating the diameter from minimal and maximal measured diameters. However, pre- and intraoperative transesophageal echocardiography in a small sample group guided this study. Wang et al8 compared aortic annulus diameters measured by intraoperative transesophageal echocardiography versus preoperative CT images with direct intraoperative sizing of the aortic annulus diameter with sizers in 227 consecutive patients undergoing proximal aortic surgery. Annular sizes by CT sizing were larger in 72.2% of cases. In 46.3% of these cases, the difference was more than one TAVR valve size. Sizing was based on the effective diameter calculated from the basal ring area, which corresponds to SAA or DAA in our nomenclature (see Figures I and IV in the Data Supplement). Because only a small portion of CT scans was ECG-gated, these results are not directly comparable to our numbers and correlation coefficients.

Although intraoperative measurements are usually considered the gold standard for comparison with CT measurements, we suggest that CT leads to more precise measurements of aortic annulus dimensions and should be considered the reference standard instead. MDCT measurements lead to 25% larger GOAs in 41% of the patients (n=32) and 22% smaller GOAs for 15.4% of the study population (n=12). Preoperative MDCT assessment may not be necessary or cost effective for all SAVR patients but may be particularly helpful in certain subgroups of patients at risk for patient-prosthesis mismatch including those with small body size, predilection for extreme calcium (ie, end-stage renal disease patients), or obese patients. It is our belief that preoperative MDCT sizing may be a useful decision making tool for annular root enlargement to optimize hemodynamics in SAVR, particularly in patients at risk for patient-prosthesis mismatch.

Because future perspectives with MDCT sizing becoming increasingly relevant, the possibility of 3-dimensional printing to aid intraoperative planning may become more widespread. As already used by some specialties (eg, orthopedic surgery, congenital cardiac surgery),18,19 3-dimension–printed models of the aortic root may aid in assessment of valve anatomy, sizing, orientation, and other difficulties associated with SAVR. The resolution of current printers has improved and is actually higher than MDCT resolution, is cheap relative to the intervention, and can be performed rapidly (ie, within 1 hour). Future advances in 3-dimensional printing may include integration of calcification by imaging, variation of biomaterials in the model, and prototyping for practice procedures.

Limitations

The incidence of patient-prosthesis mismatch could not be assessed in our cohort because the effective orifice area was not calculated; in clinical practice, effective orifice area represents a smaller value than GOA despite the best predictive estimates. In our study and in theory, the GOA represents a simple measurement to compare different valve sizes of different valve types. We used it as basis to quantify the hypothetical hemodynamic benefits of larger valves. The clinical impact of these theoretical results remains unknown, but they underline the potential to optimize hemodynamics by TAVR in patients eligible for both procedures.

Conclusions

Preoperative MDCT measurements differ significantly from intraoperative direct measurements of the aortic annulus, and these differences can be quantified by comparison of GOAs. Implanted SAVR valves based on intraoperative sizing are smaller relative to MDCT-based sizing in 41% of patients, and the potential GOA was between 25.3% and 40.6% larger if patients had undergone TAVR. The clinical impact of these theoretical results requires further studies with systematic clinical and echocardiographic follow-up.

Disclosures

Dr Hahn is a consultant with St. Jude Medical and a speaker for GE Medical, Abbott Vascular, and Boston Scientific. Dr Borger receives speakers’ honoraria (<$10,000) from Edwards Lifesciences, St. Jude, Medtronic, and CryoLife. Dr George: Consultant for Medtronic and Edwards Lifesciences. Dr Kodali is a consultant for Medtronic. Drs Leon and Smith are national co-primary investigators of the PARTNER (Placement of Aortic Transcatheter Valves) and PARTNER 2 trials.

  • Received November 23, 2016.
  • Accepted March 9, 2017.

References

  1. 1.↵
    1. Hamdan A,
    2. Guetta V,
    3. Konen E,
    4. Goitein O,
    5. Segev A,
    6. Raanani E,
    7. Spiegelstein D,
    8. Hay I,
    9. Di Segni E,
    10. Eldar M,
    11. Schwammenthal E
    . Deformation dynamics and mechanical properties of the aortic annulus by 4-dimensional computed tomography: insights into the functional anatomy of the aortic valve complex and implications for transcatheter aortic valve therapy. J Am Coll Cardiol. 2012;59:119–127. doi: 10.1016/j.jacc.2011.09.045.
    OpenUrlFREE Full Text
  2. 2.↵
    1. Tamborini G,
    2. Fusini L,
    3. Muratori M,
    4. Cefalù C,
    5. Gripari P,
    6. Ali SG,
    7. Pontone G,
    8. Andreini D,
    9. Bartorelli AL,
    10. Alamanni F,
    11. Fiorentini C,
    12. Pepi M
    . Feasibility and accuracy of three-dimensional transthoracic echocardiography vs. multidetector computed tomography in the evaluation of aortic valve annulus in patient candidates to transcatheter aortic valve implantation. Eur Heart J Cardiovasc Imaging. 2014;15:1316–1323. doi: 10.1093/ehjci/jeu157.
    OpenUrlAbstract/FREE Full Text
  3. 3.↵
    1. Willson AB,
    2. Webb JG,
    3. Labounty TM,
    4. Achenbach S,
    5. Moss R,
    6. Wheeler M,
    7. Thompson C,
    8. Min JK,
    9. Gurvitch R,
    10. Norgaard BL,
    11. Hague CJ,
    12. Toggweiler S,
    13. Binder R,
    14. Freeman M,
    15. Poulter R,
    16. Poulsen S,
    17. Wood DA,
    18. Leipsic J
    . 3-dimensional aortic annular assessment by multidetector computed tomography predicts moderate or severe paravalvular regurgitation after transcatheter aortic valve replacement: a multicenter retrospective analysis. J Am Coll Cardiol. 2012;59:1287–1294. doi: 10.1016/j.jacc.2011.12.015.
    OpenUrlFREE Full Text
  4. 4.↵
    1. Pibarot P,
    2. Weissman NJ,
    3. Stewart WJ,
    4. Hahn RT,
    5. Lindman BR,
    6. McAndrew T,
    7. Kodali SK,
    8. Mack MJ,
    9. Thourani VH,
    10. Miller DC,
    11. Svensson LG,
    12. Herrmann HC,
    13. Smith CR,
    14. Rodés-Cabau J,
    15. Webb J,
    16. Lim S,
    17. Xu K,
    18. Hueter I,
    19. Douglas PS,
    20. Leon MB
    . Incidence and sequelae of prosthesis-patient mismatch in transcatheter versus surgical valve replacement in high-risk patients with severe aortic stenosis: a PARTNER trial cohort–a analysis. J Am Coll Cardiol. 2014;64:1323–1334. doi: 10.1016/j.jacc.2014.06.1195.
    OpenUrlFREE Full Text
  5. 5.↵
    1. Jilaihawi H,
    2. Kashif M,
    3. Fontana G,
    4. Furugen A,
    5. Shiota T,
    6. Friede G,
    7. Makhija R,
    8. Doctor N,
    9. Leon MB,
    10. Makkar RR
    . Cross-sectional computed tomographic assessment improves accuracy of aortic annular sizing for transcatheter aortic valve replacement and reduces the incidence of paravalvular aortic regurgitation. J Am Coll Cardiol. 2012;59:1275–1286. doi: 10.1016/j.jacc.2011.11.045.
    OpenUrlFREE Full Text
  6. 6.↵
    1. Leon MB,
    2. Smith CR,
    3. Mack MJ,
    4. Makkar RR,
    5. Svensson LG,
    6. Kodali SK,
    7. Thourani VH,
    8. Tuzcu EM,
    9. Miller DC,
    10. Herrmann HC,
    11. Doshi D,
    12. Cohen DJ,
    13. Pichard AD,
    14. Kapadia S,
    15. Dewey T,
    16. Babaliaros V,
    17. Szeto WY,
    18. Williams MR,
    19. Kereiakes D,
    20. Zajarias A,
    21. Greason KL,
    22. Whisenant BK,
    23. Hodson RW,
    24. Moses JW,
    25. Trento A,
    26. Brown DL,
    27. Fearon WF,
    28. Pibarot P,
    29. Hahn RT,
    30. Jaber WA,
    31. Anderson WN,
    32. Alu MC,
    33. Webb JG
    . Transcatheter or surgical aortic-valve replacement in intermediate-risk patients. N Engl J Med. 2016;374:1609–1620.
    OpenUrlCrossRefPubMed
  7. 7.↵
    1. van Mieghem NM,
    2. Head SJ,
    3. van der Boon RM,
    4. Piazza N,
    5. de Jaegere PP,
    6. Carrel T,
    7. Kappetein AP,
    8. Lange R,
    9. Walther T,
    10. Windecker S,
    11. van Es GA,
    12. Serruys PW
    . The SURTAVI model: proposal for a pragmatic risk stratification for patients with severe aortic stenosis. EuroIntervention. 2012;8:258–266.
    OpenUrlCrossRefPubMed
  8. 8.↵
    1. Wang H,
    2. Hanna JM,
    3. Ganapathi A,
    4. Keenan JE,
    5. Hurwitz LM,
    6. Vavalle JP,
    7. Kiefer TL,
    8. Wang A,
    9. Harrison JK,
    10. Hughes GC
    . Comparison of aortic annulus size by transesophageal echocardiography and computed tomography angiography with direct surgical measurement. Am J Cardiol. 2015;115:1568–1573. doi: 10.1016/j.amjcard.2015.02.060.
    OpenUrl
  9. 9.↵
    1. Blanke P,
    2. Russe M,
    3. Leipsic J,
    4. Reinöhl J,
    5. Ebersberger U,
    6. Suranyi P,
    7. Siepe M,
    8. Pache G,
    9. Langer M,
    10. Schoepf UJ
    . Conformational pulsatile changes of the aortic annulus: impact on prosthesis sizing by computed tomography for transcatheter aortic valve replacement. JACC Cardiovasc Interv. 2012;5:984–994. doi: 10.1016/j.jcin.2012.05.014.
    OpenUrlAbstract/FREE Full Text
  10. 10.↵
    1. von Aspern K,
    2. Foldyna B,
    3. Etz CD,
    4. Hoyer A,
    5. Girrbach F,
    6. Holzhey D,
    7. Lucke C,
    8. Grothoff M,
    9. Linke A,
    10. Mohr FW,
    11. Gutberlet M,
    12. Lehmkuhl L
    Effective diameter of the aortic annulus prior to transcatheter aortic valve implantation: influence of area-based versus perimeter-based calculation. Int J Cardiovasc Imaging. 2015;31:163–169.
    OpenUrlCrossRefPubMed
  11. 11.↵
    1. Bapat V.
    Valve-in-valve apps: why and how they were developed and how to use them. EuroIntervention. 2014;10(suppl U):44–51.
    OpenUrl
  12. 12.↵
    1. Willson AB,
    2. Webb JG,
    3. Gurvitch R,
    4. Wood DA,
    5. Toggweiler S,
    6. Binder R,
    7. Freeman M,
    8. Madden M,
    9. Hague C,
    10. Leipsic J
    . Structural integrity of balloon-expandable stents after transcatheter aortic valve replacement: assessment by multidetector computed tomography. JACC Cardiovasc Interv. 2012;5:525–532. doi: 10.1016/j.jcin.2012.03.007.
    OpenUrlAbstract/FREE Full Text
  13. 13.↵
    1. Herrmann HC,
    2. Thourani VH,
    3. Kodali SK,
    4. Makkar RR,
    5. Szeto WY,
    6. Anwaruddin S,
    7. Desai N,
    8. Lim S,
    9. Malaisrie SC,
    10. Kereiakes DJ,
    11. Ramee S,
    12. Greason KL,
    13. Kapadia S,
    14. Babaliaros V,
    15. Hahn RT,
    16. Pibarot P,
    17. Weissman NJ,
    18. Leipsic J,
    19. Whisenant BK,
    20. Webb JG,
    21. Mack MJ,
    22. Leon MB
    ; PARTNER Investigators. One-year clinical outcomes with SAPIEN 3 transcatheter aortic valve replacement in high- risk and inoperable patients with severe aortic stenosis. Circulation. 2016;134:130–140. doi: 10.1161/CIRCULATIONAHA.116.022797.
    OpenUrlAbstract/FREE Full Text
  14. 14.↵
    1. Deeb GM,
    2. Reardon MJ,
    3. Chetcuti S,
    4. Patel HJ,
    5. Grossman PM,
    6. Yakubov SJ,
    7. Kleiman NS,
    8. Coselli JS,
    9. Gleason TG,
    10. Lee JS,
    11. Hermiller JB Jr.,
    12. Heiser J,
    13. Merhi W,
    14. Zorn GL 3rd.,
    15. Tadros P,
    16. Robinson N,
    17. Petrossian G,
    18. Hughes GC,
    19. Harrison JK,
    20. Maini B,
    21. Mumtaz M,
    22. Conte J,
    23. Resar J,
    24. Aharonian V,
    25. Pfeffer T,
    26. Oh JK,
    27. Qiao H,
    28. Adams DH,
    29. Popma JJ
    ; CoreValve US Clinical Investigators. 3-Year outcomes in high-risk patients who underwent surgical or transcatheter aortic valve replacement. J Am Coll Cardiol. 2016;67:2565–2574. doi: 10.1016/j.jacc.2016.03.506.
    OpenUrlFREE Full Text
  15. 15.↵
    1. Barbanti M,
    2. Yang TH,
    3. Rodès Cabau J,
    4. Tamburino C,
    5. Wood DA,
    6. Jilaihawi H,
    7. Blanke P,
    8. Makkar RR,
    9. Latib A,
    10. Colombo A,
    11. Tarantini G,
    12. Raju R,
    13. Binder RK,
    14. Nguyen G,
    15. Freeman M,
    16. Ribeiro HB,
    17. Kapadia S,
    18. Min J,
    19. Feuchtner G,
    20. Gurtvich R,
    21. Alqoofi F,
    22. Pelletier M,
    23. Ussia GP,
    24. Napodano M,
    25. de Brito FS Jr.,
    26. Kodali S,
    27. Norgaard BL,
    28. Hansson NC,
    29. Pache G,
    30. Canovas SJ,
    31. Zhang H,
    32. Leon MB,
    33. Webb JG,
    34. Leipsic J
    . Anatomical and procedural features associated with aortic root rupture during balloon-expandable transcatheter aortic valve replacement. Circulation. 2013;128:244–253. doi: 10.1161/CIRCULATIONAHA.113.002947.
    OpenUrlAbstract/FREE Full Text
  16. 16.↵
    1. Schmidkonz C,
    2. Marwan M,
    3. Klinghammer L,
    4. Mitschke M,
    5. Schuhbaeck A,
    6. Arnold M,
    7. Lell M,
    8. Achenbach S,
    9. Pflederer T
    . Interobserver variability of CT angiography for evaluation of aortic annulus dimensions prior to transcatheter aortic valve implantation (TAVI). Eur J Radiol. 2014;83:1672–1678. doi: 10.1016/j.ejrad.2014.06.001.
    OpenUrlCrossRefPubMed
  17. 17.↵
    1. Dashkevich A,
    2. Blanke P,
    3. Siepe M,
    4. Pache G,
    5. Langer M,
    6. Schlensak C,
    7. Beyersdorf F
    . Preoperative assessment of aortic annulus dimensions: comparison of noninvasive and intraoperative measurement. Ann Thorac Surg. 2011;91:709–714. doi: 10.1016/j.athoracsur.2010.09.038.
    OpenUrlCrossRefPubMed
  18. 18.↵
    1. Sodian R,
    2. Schmauss D,
    3. Markert M,
    4. Weber S,
    5. Nikolaou K,
    6. Haeberle S,
    7. Vogt F,
    8. Vicol C,
    9. Lueth T,
    10. Reichart B,
    11. Schmitz C
    . Three-dimensional printing creates models for surgical planning of aortic valve replacement after previous coronary bypass grafting. Ann Thorac Surg. 2008;85:2105–2108. doi: 10.1016/j.athoracsur.2007.12.033.
    OpenUrlCrossRefPubMed
  19. 19.↵
    1. Schmauss D,
    2. Schmitz C,
    3. Bigdeli AK,
    4. Weber S,
    5. Gerber N,
    6. Beiras-Fernandez A,
    7. Schwarz F,
    8. Becker C,
    9. Kupatt C,
    10. Sodian R
    . Three-dimensional printing of models for preoperative planning and simulation of transcatheter valve replacement. Ann Thorac Surg. 2012;93:31–33.
    OpenUrl

CLINICAL PERSPECTIVE

Appropriate valve sizing is critical in aortic valve replacement. Patients undergoing aortic valve replacement may experience clinical consequences with annular undersizing, specifically paravalvular leak, leaflet dysfunction, worsened hemodynamics, or possible patient-prosthesis mismatch. Multidetector computed tomography (MDCT) has currently become the standard of care for optimal annular sizing in transcatheter aortic valve replacement. For patients undergoing surgical aortic valve replacement (SAVR), sizing has relied on intraoperative measurements after calcium debridement, which may or may not correlate with MDCT-based sizing. The ability to optimally size surgical prostheses has taken on much greater importance in recent years after recognition that hemodynamics after SAVR may be inferior to transcatheter aortic valve replacement, mainly because of geometric constraints of the prostheses themselves. Preoperative MDCT offers the potential to improve sizing during SAVR by understanding annular and valvular geometry before surgery. In the current study of a cohort of SAVR patients with preoperative MDCT imaging, we found that 41% of patients received a valve smaller than predicted by MDCT, and the potential geometric orifice areas was between 25.3% and 40.6% larger if patients had undergone transcatheter aortic valve replacement. Preoperative MDCT assessment may not be necessary or cost effective for all SAVR patients and may be a useful decision-making tool for annular root enlargement to optimize hemodynamics in SAVR, particularly in patients at risk for patient-prosthesis mismatch.

View Abstract

This Issue

Jump to

Article Tools

Share this Article

  • Email
  • Aortic Valve Annular SizingCLINICAL PERSPECTIVE
    Isaac George, Laura C. Guglielmetti, Nicolas Bettinger, Andrew Moss, Catherine Wang, Nathan Kheysin, Rebecca Hahn, Susheel Kodali, Martin Leon, Vinayak Bapat, Michael A. Borger, Mathew Williams, Craig Smith and Omar K. Khalique
    Circulation: Cardiovascular Imaging. 2017;10:e005968, originally published May 9, 2017
    https://doi.org/10.1161/CIRCIMAGING.116.005968
    del.icio.us logo Digg logo Reddit logo Twitter logo CiteULike logo Facebook logo Google logo Mendeley logo

Related Articles

Cited By...

Subjects