<abstract xmlns="http://www.w3.org/1999/xhtml">

<p>We set out to assess whether the preoperative prediction of patient-prosthesis mismatch by an algorithm popularised by Pibarot et al. is accurate compared to PPMM defined as AVA &lt;0.85 cm<sup>2</sup>/m<sup>2</sup> (‘any PPMM’) or as AVA &lt;0.65cm<sup>2</sup>/m<sup>2</sup> (‘severe’ PPMM).</p>
<p>We studied 37 patients (16 female), with mean age 72.3 +/− 8.4 years, who had unplanned SAVR during 2019. By echo, 17 (19.7%) patients had severe PPMM (mean AVA (SD) = 0.55 (0.04) cm<sup>2</sup>). For the 0.85cm<sup>2</sup>/m<sup>2</sup> cut-off, 14/37 patients (37%) had PPMM (mean AVA (SD) 0.70 (0.15) cm<sup>2</sup>. By the Pibarot algorithm, 21 (56.7%) patients were predicted to have any PPMM and 10 (27%) severe PPMM.</p>
<p>The preoperative algorithm for the avoidance of patient prosthetic mismatch does not accurately identify patients who have postoperative PPMM as defined by echocardiography. Further refinements of the algorithm seem necessary before it can be used widely in clinical practice.</p>
</abstract>

We set out to assess whether the preoperative prediction of patient-prosthesis mismatch by an algorithm popularised by Pibarot et al. is accurate compared to PPMM defined as AVA <0.85 cm2/m2 (‘any PPMM’) or as AVA <0.65cm2/m2 (‘severe’ PPMM).
We studied 37 patients (16 female), with mean age 72.3 +/− 8.4 years, who had unplanned SAVR during 2019. By echo, 17 (19.7%) patients had severe PPMM (mean AVA (SD) = 0.55 (0.04) cm2). For the 0.85cm2/m2 cut-off, 14/37 patients (37%) had PPMM (mean AVA (SD) 0.70 (0.15) cm2. By the Pibarot algorithm, 21 (56.7%) patients were predicted to have any PPMM and 10 (27%) severe PPMM.
The preoperative algorithm for the avoidance of patient prosthetic mismatch does not accurately identify patients who have postoperative PPMM as defined by echocardiography. Further refinements of the algorithm seem necessary before it can be used widely in clinical practice.

<div xmlns="http://www.w3.org/1999/xhtml">
<sec id="j_rjc-2022-0033_s_001"><div>Introduction</div>
<p>Although initially controversial [<a ref-type="bibr" href="#j_rjc-2022-0033_ref_001">1</a>], the notion that an aortic replacement valve that is too small for the size of the patient (i.e. patient-prosthesis mismatch or PPMM) [<a ref-type="bibr" href="#j_rjc-2022-0033_ref_002">2</a>] is associated with adverse outcomes after surgical aortic valve replacement (SAVR) is now generally accepted and backed up by data [<a ref-type="bibr" href="#j_rjc-2022-0033_ref_003">3</a>].</p>
<p>An algorithm for prevention of PPMM has been proposed [<a ref-type="bibr" href="#j_rjc-2022-0033_ref_004">4</a>]. This involves multiplying the body surface area of each patient by 0.85, which gives the minimal effective valve area that would not produce PPMM. The surgeon is informed of this measure and told to try to avoid using a valve smaller than this cutoff [<a ref-type="bibr" href="#j_rjc-2022-0033_ref_005">5</a>]. The performance, applicability and clinical impact of the proposed anti-PPMM algorithm ‘in the real world’ are not clear.</p>
<p>Patients undergoing unplanned, urgent or emergency SAVR are a group at high risk for adverse outcomes: preoperative planning and investigations are often curtailed by the need to resolve clinical instability rapidly by offering SAVR. Moreover, the surgeons would want to ‘keep it short’ and thus, for instance, avoid time-consuming LIMA harvesting as opposed to saphenous vein graft in patients undergoing CABG [<a ref-type="bibr" href="#j_rjc-2022-0033_ref_006">6</a>].</p>
<p>In patients undergoing unplanned SAVR we set out to investigate:
<list list-type="custom">
<list-item><label>–</label><p>The proportion of patients that receive artificial valves which are mismatched based on the AVA reported in the literature [<a ref-type="bibr" href="#j_rjc-2022-0033_ref_007">7</a>] for the specific model used in each patient.</p></list-item>
<list-item><label>–</label><p>The proportion of valves that are mismatched based on the effective AVA calculated by the continuity equation.</p></list-item></list></p>
</sec>
<sec id="j_rjc-2022-0033_s_002"><div>Setting</div>
<p>Morriston Cardiac Centre is a regional tertiary centre in West Wales, UK, serving a population of about 1 million. There are 5 cardiac surgeons operating in 2 cardiac theatres, performing approximately 150 surgical aortic valve replacement (SAVR) operations/year. The echo department performs approximately 11,000 transthoracic echocardiograms using the minimum dataset as recommended by the BSE [<a ref-type="bibr" href="#j_rjc-2022-0033_ref_008">8</a>].</p>
</sec>
<sec id="j_rjc-2022-0033_s_003"><div>Methods</div>
<p>Using our surgical database, we identified patients who had unscheduled, isolated SAVR for AS during the last year before the pandemic (2019). These patients also had documented procedural and echocardiographic parameters from their last preoperative, and first postoperative scans available (<a ref-type="table" href="#j_rjc-2022-0033_tab_001">Table 1</a>). We focused on unplanned operations because such patients often do not benefit from the same amount of careful preoperative planning as those having elective SAVR, and – because of the high risk associated with clinical instability – procedures such as aortic annular enlargement are unlikely to be performed. These features render them particularly susceptible to the deleterious effects of PPMM.</p>
<table-wrap id="j_rjc-2022-0033_tab_001" position="float"><label>Table 1</label><caption><p>Clinical and echocardiographic characteristics. NYHA – New York Heart Association; CCS – Canadian cardiac Society class; CVRFs – cardiovascular risk factors (smoking, diabetes, systemic arterial hypertension, hypercholesterolaemia, premature family history of coronary artery disease)</p></caption>
<table frame="hsides" rules="groups">
<tbody>
<tr>
<td align="center" valign="middle">Angina</td>
<td align="center" valign="middle">Absent</td>
<td align="center" valign="middle">28</td>
</tr>
<tr>
<td align="center" valign="middle"></td>
<td align="center" valign="middle">Present</td>
<td align="center" valign="middle">9</td>
</tr>
<tr>
<td align="center" valign="middle">Heart failure</td>
<td align="center" valign="middle">Absent</td>
<td align="center" valign="middle">14</td>
</tr>
<tr>
<td align="center" valign="middle"></td>
<td align="center" valign="middle">Present</td>
<td align="center" valign="middle">13</td>
</tr>
<tr>
<td align="center" valign="middle">Syncope</td>
<td align="center" valign="middle">Absent</td>
<td align="center" valign="middle">34</td>
</tr>
<tr>
<td align="center" valign="middle"></td>
<td align="center" valign="middle">Present</td>
<td align="center" valign="middle">3</td>
</tr>
<tr>
<td align="center" valign="middle">Number of symptoms</td>
<td align="center" valign="middle">0</td>
<td align="center" valign="middle">1</td>
</tr>
<tr>
<td align="center" valign="middle"></td>
<td align="center" valign="middle">1</td>
<td align="center" valign="middle">32</td>
</tr>
<tr>
<td align="center" valign="middle"></td>
<td align="center" valign="middle">2</td>
<td align="center" valign="middle">4</td>
</tr>
<tr>
<td align="center" valign="middle">NYHA Class</td>
<td align="center" valign="middle">1</td>
<td align="center" valign="middle">2</td>
</tr>
<tr>
<td align="center" valign="middle"></td>
<td align="center" valign="middle">2</td>
<td align="center" valign="middle">12</td>
</tr>
<tr>
<td align="center" valign="middle"></td>
<td align="center" valign="middle">3</td>
<td align="center" valign="middle">16</td>
</tr>
<tr>
<td align="center" valign="middle"></td>
<td align="center" valign="middle">4</td>
<td align="center" valign="middle">7</td>
</tr>
<tr>
<td align="center" valign="middle">CCS Class</td>
<td align="center" valign="middle">1</td>
<td align="center" valign="middle">23</td>
</tr>
<tr>
<td align="center" valign="middle"></td>
<td align="center" valign="middle">2</td>
<td align="center" valign="middle">6</td>
</tr>
<tr>
<td align="center" valign="middle"></td>
<td align="center" valign="middle">3</td>
<td align="center" valign="middle">7</td>
</tr>
<tr>
<td align="center" valign="middle"></td>
<td align="center" valign="middle">4</td>
<td align="center" valign="middle">1</td>
</tr>
<tr>
<td align="center" valign="middle">Number of CVRFs</td>
<td align="center" valign="middle">0</td>
<td align="center" valign="middle">4</td>
</tr>
<tr>
<td align="center" valign="middle"></td>
<td align="center" valign="middle">1</td>
<td align="center" valign="middle">9</td>
</tr>
<tr>
<td align="center" valign="middle"></td>
<td align="center" valign="middle">2</td>
<td align="center" valign="middle">16</td>
</tr>
<tr>
<td align="center" valign="middle"></td>
<td align="center" valign="middle">3</td>
<td align="center" valign="middle">6</td>
</tr>
<tr>
<td align="center" valign="middle"></td>
<td align="center" valign="middle">4</td>
<td align="center" valign="middle">2</td>
</tr>
<tr>
<td align="center" valign="middle">BSA (m<sup>2</sup>)</td>
<td align="center" valign="middle">mean</td>
<td align="center" valign="middle">1.93</td>
</tr>
<tr>
<td align="center" valign="middle"></td>
<td align="center" valign="middle">SD</td>
<td align="center" valign="middle">0.22</td>
</tr>
<tr>
<td align="center" valign="middle">AVA (cm<sup>2</sup>)</td>
<td align="center" valign="middle">mean</td>
<td align="center" valign="middle">1.63</td>
</tr>
<tr>
<td align="center" valign="middle"></td>
<td align="center" valign="middle">SD</td>
<td align="center" valign="middle">0.41</td>
</tr>
<tr>
<td align="center" valign="middle">Indexed AVA (cm<sup>2</sup>/m<sup>2</sup>)</td>
<td align="center" valign="middle">mean</td>
<td align="center" valign="middle">0.85</td>
</tr>
<tr>
<td align="center" valign="middle"></td>
<td align="center" valign="middle">SD</td>
<td align="center" valign="middle">0.22</td>
</tr>
<tr>
<td align="center" valign="middle">Prosthesis size (mm)</td>
<td align="center" valign="middle">mean</td>
<td align="center" valign="middle">23.88</td>
</tr>
<tr>
<td align="center" valign="middle"></td>
<td align="center" valign="middle">SD</td>
<td align="center" valign="middle">2.40</td>
</tr>
<tr>
<td align="center" valign="middle">Stroke volume (ml)</td>
<td align="center" valign="middle">mean</td>
<td align="center" valign="middle">33.80</td>
</tr>
<tr>
<td align="center" valign="middle"></td>
<td align="center" valign="middle">SD</td>
<td align="center" valign="middle">7.88</td>
</tr>
<tr>
<td align="center" valign="middle">Stroke volume index</td>
<td align="center" valign="middle">Mean</td>
<td align="center" valign="middle">52.20</td>
</tr>
<tr>
<td align="center" valign="middle">(ml/m<sup>2</sup>)</td>
<td align="center" valign="middle">SD</td>
<td align="center" valign="middle">11.89</td>
</tr>
</tbody>
</table>
</table-wrap>
<p>We defined patient-prosthesis mismatch (PPMM) as an aortic valve area (AVA) &lt;/= 0.85cm<sup>2</sup>/m<sup>2</sup> of body surface area (BSA), and severe PPMM as AVA &lt;0.65cm<sup>2</sup>/m<sup>2</sup>, in accordance with current guidelines [<a ref-type="bibr" href="#j_rjc-2022-0033_ref_005">5</a>].</p>
<p>We then calculated the ‘ideal’ minimal area of an implanted valve that would avoid the occurrence of any (AVAi>0.85cm<sup>2</sup>/m<sup>2</sup>) or of severe (AVAi>0.65cm<sup>2</sup>/m<sup>2</sup>) patient-prosthesis mismatch as follows: we calculated the estimated body surface area (BSA) with the Mosteller formula for each patient and multiplied it by 0.85 (for the limit of any mismatch) or by 0.65 (for a severe mismatch). For example, assuming a BSA of 2m<sup>2</sup>, any valve with a continuity area &lt; 2 × 0.85 = 1.6cm<sup>2</sup> would have a mismatch, while if the area was &lt; 2 × 0.65 = 1.3cm<sup>2</sup>, then a severe mismatch would be present.</p>
<p>We compared the effective AVA calculated by the continuity equation with the literature-indicated minimal AVA for each size and model of valve used in this patient population, in order to establish whether mismatch was accurately predicted by the areas quoted in the literature. For instance, if the normal range for valve areas in a size 23 bioprosthetic AVR was 2.3 +/− 0.5cm<sup>2</sup>, we calculated the minimal normal area as 2.3 – 0.5 = 1.8cm<sup>2</sup>, and compared it to the AVA measured by continuity equation on the postoperative echocardiogram. We used an online calculator [<a ref-type="bibr" href="#j_rjc-2022-0033_ref_009">9</a>] for the chi-square statistic, with a significance level set at p&lt;0.05.</p>
</sec>
<sec id="j_rjc-2022-0033_s_004"><div>Results</div>
<p>37 patients (16 female), mean age 72.3 +/− 8.4 years, had unplanned SAVR during 2019. Selected clinical and echo pre-operative features are listed in <a ref-type="table" href="#j_rjc-2022-0033_tab_001">Table 1</a>. The mean BSA (SD) was 1.93 (0.23) m<sup>2</sup>.</p>
<p>The following bioprosthetic valves (BPVs) were used: Carpentier Edwards – 17; Perimount – 6; Intuity and Inspiris – 4 of each; Trifecta – 3; Regent – 2; Perceval −1. The average BPV size (SD) was 23.8 (2.4) mm.</p>
<sec id="j_rjc-2022-0033_s_004_s_001"><label>1</label><div>Patient-prosthesis mismatch</div>
<p>We found that 7/37 (19.7%) patients had severe PPMM (AVA&lt;0.65cm<sup>2</sup>/m<sup>2</sup> of BSA), with a mean AVA (SD) of 0.55 (0.04) cm<sup>2</sup> and range 0.48 – 0.63 cm2. In the remaining 30 patients without severe PPMM, AVA (SD) was 0.91 (0.19) cm<sup>2</sup> with range 0.66 – 1.35cm<sup>2</sup>. Applying the 0.85cm<sup>2</sup>/m<sup>2</sup> cut-off, we found that 14/37 patients (37%) had PPMM (AVA (SD) 0.70 (0.15) cm<sup>2</sup>, range 0.48 – 0.85 cm<sup>2</sup>). Meanwhile, in those without PPMM by the 0.65 cm<sup>2</sup>/m<sup>2</sup> criterion, the mean AVA (SD) was 1.09 (0.15) cm<sup>2</sup>, with a range of 0.88–1.55 cm<sup>2</sup>.</p>
<p>By using the lower limit of the normal AVA provided by the literature for each valve type and size used, we identified that 21/37 (56.7%) of patients would have had any PPMM (AVA&lt;0.85cm<sup>2</sup>/m<sup>2</sup>), and that 18/37 (51%) would have had severe PPMM (AVA&lt;0.65cm<sup>2</sup>/m<sup>2</sup>). The numbers illustrating the discrepancy we found between PPMM as predicted by the minimal literature-provided AVA and the AVA calculated by the continuity equation is illustrated in <a ref-type="table" href="#j_rjc-2022-0033_tab_002">Table 2</a>.</p>
<table-wrap id="j_rjc-2022-0033_tab_002" position="float"><label>Table 2</label><caption><p>Actual and algorithm-predicted cases where any patient-prosthesis mismatch was present</p></caption>
<table frame="hsides" rules="groups">
<thead>
<tr>
<th align="center" valign="middle"><bold>Actual AVA (by continuity)</bold></th>
<th align="center" valign="middle" colspan="2"><bold>AVA predicted by the algorithm</bold></th>
<th align="center" valign="middle"><bold>Total</bold></th>
</tr>
</thead>
<tbody>
<tr>
<td align="center" valign="middle"></td>
<td align="center" valign="middle">AVA&lt;0.85×BSA</td>
<td align="center" valign="middle">AVA>0.85×BSA</td>
<td align="center" valign="middle"></td>
</tr>
<tr>
<td align="center" valign="middle">AVA&lt;0.85×BSA</td>
<td align="center" valign="middle">14</td>
<td align="center" valign="middle">9</td>
<td align="center" valign="middle"><bold>23</bold></td>
</tr>
<tr>
<td align="center" valign="middle">AVA>0.85×BSA</td>
<td align="center" valign="middle">8</td>
<td align="center" valign="middle">6</td>
<td align="center" valign="middle"><bold>14</bold></td>
</tr>
<tr>
<td align="center" valign="middle"><bold>Total</bold></td>
<td align="center" valign="middle"><bold>22</bold></td>
<td align="center" valign="middle"><bold>15</bold></td>
<td align="center" valign="middle"><bold>37</bold></td>
</tr>
</tbody>
</table>
</table-wrap>
<table-wrap id="j_rjc-2022-0033_tab_003" position="float"><label>Table 3</label><caption><p>Actual and algorithm-predicted cases where severe patient prosthesis mismatch was present. p>0.05 for all cells</p></caption>
<table frame="hsides" rules="groups">
<thead>
<tr>
<th align="center" valign="middle"><bold>Actual AVA (by continuity)</bold></th>
<th align="center" valign="middle" colspan="2"><bold>AVA predicted by the algorithm</bold></th>
<th align="center" valign="middle"><bold>Total</bold></th>
</tr>
</thead>
<tbody>
<tr>
<td align="center" valign="middle"></td>
<td align="center" valign="middle">AVA&lt;0.65×BSA</td>
<td align="center" valign="middle">AVA>0.65×BSA</td>
<td align="center" valign="middle"></td>
</tr>
<tr>
<td align="center" valign="middle">AVA&lt;0.65×BSA</td>
<td align="center" valign="middle">2</td>
<td align="center" valign="middle">5</td>
<td align="center" valign="middle"><bold>7</bold></td>
</tr>
<tr>
<td align="center" valign="middle">AVA>0.65×BSA</td>
<td align="center" valign="middle">7</td>
<td align="center" valign="middle">23</td>
<td align="center" valign="middle"><bold>30</bold></td>
</tr>
<tr>
<td align="center" valign="middle"><bold>Total</bold></td>
<td align="center" valign="middle"><bold>9</bold></td>
<td align="center" valign="middle"><bold>28</bold></td>
<td align="center" valign="middle"><bold>37</bold></td>
</tr>
</tbody>
</table>
</table-wrap>
</sec>
</sec>
<sec id="j_rjc-2022-0033_s_005"><div>Discussion</div>
<p>It is generally accepted today that PPMM should be avoided if possible, as it is associated with worse outcomes [<a ref-type="bibr" href="#j_rjc-2022-0033_ref_010">10</a>]. This statement is extensively proven for AVR, but there is data suggesting a negative impact on outcomes for MVR as well [<a ref-type="bibr" href="#j_rjc-2022-0033_ref_011">11</a>].</p>
<p>The ideal strategy would be to identify not only patients at risk of PPMM (e.g. those with large BMI or small LVOT) but also the ‘ideal’ valve size above which PPMM would not occur. The group from Montreal has proposed an algorithm, which recommends using the EROA from the literature to decide the cut-off valve size for specific BSAs in patients.</p>
<p>There is little data available on how this algorithm fares ‘in the real world,’ but in one of the very few published series [<a ref-type="bibr" href="#j_rjc-2022-0033_ref_012">12</a>], the best predictive accuracy for PPMM was not by using the AVA reported in the literature, but AVA's normal ranges defined in the echo laboratory where the study was conducted.</p>
<p>We found that PPMM diagnosed on the pre-discharge echocardiogram was present in patients for whom the algorithm did not predict it, and, conversely, that patients in whom the algorithm predicted PPMM did not display it. The reasons for this discrepancy are only speculated, but are probably mostly related to the relatively arbitrary limits of normality identified for prosthetic valve effective orifice area, as well as the intrinsic variability of echo measurements [<a ref-type="bibr" href="#j_rjc-2022-0033_ref_013">13</a>].</p>
</sec>
<sec id="j_rjc-2022-0033_s_006"><div>Limitations</div>
<p>The main limitation is the very small number of patients we studied. This was a pilot study and it demonstrated that PPMM is common and not easily avoidable. We are planning further large-scale studies in order to delineate the most effective strategy for PPMM avoidance. Another limitation relates to the timing of the echo study soon after surgery. Ideally, the baseline postoperative study should be obtained at the 1<sup>st</sup> clinic visit. However, in our case, some such studies were inaccessible as they were performed in other hospitals, so we decided to use the pre-discharge echocardiogram.</p>
</sec>
<sec id="j_rjc-2022-0033_s_007"><div>Conclusions</div>
<p>Patient-prosthetic mismatch is common after isolated surgical aortic valve replacement. Current algorithms based on the reference ranges for the effective valve area published in cardiology literature do not have good accuracy for predicting PPMM.</p>
</sec>
</div>

Patient Prosthesis Mismatch in SAVR: How Avoidable is It in the ‘Real World’?

Romanian Journal of Cardiology

This work is licensed under the Creative Commons Attribution 4.0 International License.

We set out to assess whether the preoperative prediction of patient-prosthesis mismatch by an algorithm popularised by Pibarot et al. is accurate compared to...

Patient Prosthesis Mismatch in SAVR: How Avoidable is It in the ‘Real World’?

Article Category: Original Article

Data publikacji: 31 gru 2022

Zakres stron: 189 - 191

DOI: https://doi.org/10.2478/rjc-2022-0033

Słowa kluczowe
aortic replacement valve, mismatch, echocardiography

© 2022 Amir Mushtaq et al., published by Sciendo

This work is licensed under the Creative Commons Attribution 4.0 International License.

Patient Prosthesis Mismatch in SAVR: How Avoidable is It in the ‘Real World’?

Article Category: Original Article

Data publikacji: 31 gru 2022

Zakres stron: 189 - 191

DOI: https://doi.org/10.2478/rjc-2022-0033

Słowa kluczoweaortic replacement valve, mismatch, echocardiography

© 2022 Amir Mushtaq et al., published by Sciendo

This work is licensed under the Creative Commons Attribution 4.0 International License.

Słowa kluczowe
aortic replacement valve, mismatch, echocardiography