Aortic patient-prosthesis mismatch - does it matter? A review for cardiologists and cardiac surgeons

PPM is an early example of what has now become a very fashionable notion—personalized medicine¹⁴. The original definition¹ does not state this explicitly, but subsequent clinical applications of PPM make it clear that the artificial valve is considered mismatched when it is too small for the individual patient into which it has been implanted.

How do we define ‘too small’ for a given individual? The function of a heart valve is to convey the cardiac output unidirectionally, from the heart to the rest of the body, with minimal energy ‘loss’, i.e., with the smallest possible pressure drop and least turbulence associated with the flow of blood through the valve. ‘Too small for the patient’ thus translates as ‘too small for the effective and efficient delivery of a patient’s cardiac output to the body’. How do we decide what is the normal cardiac output of an individual patient? In the echo lab and by the bedside we need an easily accessible surrogate marker and although the supportive evidence is of poor quality, BSA is universally used, based on the observation that BSA correlates (albeit poorly) with cardiac output¹⁵.

As a result, all the accepted definitions of PPM use indexation of the valve area (AVA_i) to BSA, with the thresholds mentioned earlier (moderate – EOA_i <0.85-0.65 cm²/m², and severe – EOA_i <0.65cm²/m²). For obese patients (BMI≥30) the cutoff values are 0.70 cm²/m²and 0.55cm²/m², respectively (Table 2).

However, there are multiple problems with using BSA as the denominator for the indexation of an artificial valve’s EOA_i, as summarized recently in a paper addressing native aortic stenosis (AS) but making arguments which are also relevant to PPM¹⁶: -

The relationship between cardiac output (CO) and BSA is not linear in obese patients;

Severe PPM was associated with reduced regression of LV hypertrophy (LVH), more symptomatic heart failure, and more cardiac events after SAVR⁴, while Ruel et al.¹⁸ reported that, in 1563 patients undergoing SAVR, PPM (defined as EOA <0.80 cm²/m²) was an independent predictor of postoperative congestive heart failure, and larger prosthetic valves had a beneficial effect on symptoms, but PPM had no effect on overall survival. Conversely, in 2005, Koch et al., reporting on data collected in 1998 from 1014 patients undergoing SAVR and who completed a postoperative functional status questionnaire, found that ‘neither indexed orifice area, nor standardized orifice size was associated with functional recovery’, and concluded that ‘factors other than prosthesis–patient size influence functional quality of life early after SAVR’¹⁹.

He at al²⁰ reported that in 447 patients receiving SAVR size 21 or less, PPM was associated with increased mortality only in patients who also received coronary artery bypass grafting (CABG) at the time of SAVR.

Blais et al⁵ found a prevalence of moderate or severe PPM of 38% among 1,266 consecutive patients who had undergone SAVR. Mortality at one month postoperative was 4.6% and the relative risk of death increased by 11-fold in those with severe PPM, modulated by the left ventricular ejection fraction (LVEF), with the highest mortality in patients with severe PPM and LVEF <40% (Figure 2). Similarly, a report from the STS found significantly higher operative mortality in patients receiving small artificial valves²¹.

Figure 2

Conversely, Medalion et al.²² followed up 892 adults receiving a mechanical (n = 346), pericardial (n = 463), or allograft (n = 83) valve for aortic stenosis for up to 20 years (mean, 5.0 ± 3.9 years) after primary isolated aortic valve replacement. PPM was diagnosed if the indexed internal orifice area was <1.5cm²/m². After adjusting for multiple variables, indexed AVA had no impact on mortality at any time point after SAVR. It is notable how different the threshold for PPM is in this paper compared to the definitions accepted currently.

Blackstone et al.²³ studied 13,258 patients who had SAVR with various types of prosthetic valves and, after adjustment for multiple risk factors, found that valve size had no effect on survival up to 15 years.

Hanayama et al.²⁴ studied prospectively 1,129 consecutive patients undergoing aortic valve replacement between 1990 and 2000. Patient-prosthesis mismatch was defined as indexed EOA below the 10^th percentile (< 0.60 cm²/m²). On multivariate analysis, there was no significant difference in survival between normal and abnormal gradient groups and patients with and without patient–prosthesis mismatch were similar with respect to postoperative left ventricular mass index, 7-year survival (95.1% ± 1.3% versus 94.7% ± 3.0%; p = 0.54), and 7-year freedom from severe heart failure (NYHA >3) (79.3% ± 6.6% versus 74.5% ± 2.5%; p = 0.40).

In an editorial²⁵, Tirone David reviewed the salient papers on the clinical impact of PPM and stated that ‘although the majority of these studies used multivariate analysis to determine the incremental risk of small prosthetic valves on operative mortality, the reality is that patients with small aortic annulus are usually older, have more severe left ventricular hypertrophy with impaired diastolic function, and frequently have coronary artery disease²⁵—in other words, the small aortic root that leads to the occurrence of PPM is a surrogate marker for patient characteristics associated with adverse outcomes, rather than PPM being a causal factor for such outcomes. Nevertheless, the editorial recommends trying to avoid PPM in young, active individuals.

It is obvious that the literature on PPM is problematic: definitions are inconsistent, the quality of the data is questionable, and methodologies are often suboptimal (such as using the ‘labeled size’ of the valve as equivalent to the effective valve area). In cardiology, this is the moment when meta-analyses are usually called to the rescue.

Head et al.²⁶ identified 34 studies comprising 27,186 patients and 133,141 patient-years of follow-up. Using the 0.85cm²/m² cutoff, 44% had PPM; 10% of the total number had severe PPM (AVA_i<0.65cm²/m²). The presence of any degree of PPM increased all-cause mortality (HR 1.34 CI: 1.18 – 1.51) but not cardiac mortality, while severe PPM also markedly increased cardiac mortality (and HR = 6.46, 95% CI: 2.79–14.97). This may be interpreted as evidence that PPM is a marker for patient features associated with increased mortality, rather than a causal factor for adverse events (if PPM increased mortality through mechanisms similar to those by which AS increases mortality, then we would expect to see cardiac, rather than just general mortality going up, in the presence of PPM). Perhaps only severe PPM is relevant from the perspective of cardiac mortality. Nevertheless, the authors concluded that ‘current efforts to prevent PPM should receive more emphasis and a widespread acceptance to improve long-term survival after AVR’.

Dayan et al.²⁷ extracted data from 58 articles including 40,381 patients (of whom 813 were TAVR). Severe PPM increased both perioperative and overall mortality, whereas moderate PPM was associated with increased risk of perioperative mortality but not of overall mortality.

The authors concluded that PPM increases perioperative and overall mortality proportionally to its severity, more so below the age of 70, and in patients receiving concomitant CABG, and suggest that predictors identified in their meta-analysis (older age, female sex, hypertension, diabetes, renal failure, larger BSA, larger body mass index, and the utilization of a bioprosthesis) can ‘flag’ patients at risk of PPM.

In 2019, De Oliviera Sa at al.²⁸ produced what is probably the definitive meta-analysis on the topic of PPM after SAVR, including 70 articles with over 100,000 patients included, with an incidence of PPM after SAVR of 53.7%. Perioperative, 1, 5, and 10-year mortality were increased in patients with both moderate and severe PPM, to a similar extent at each time-point, but more so in severe PPM (Figure 3). The authors conclude that surgical strategies should be implemented, to prevent PPM and to decrease mortality after SAVR. Similar conclusions are offered in a recent patient-level meta-analysis of outcomes after SAVR from the same group²⁹.

Figure 3

Of course, analysis of a dataset of >100,000 patients sounds impressive and conclusions derived will be regarded as authoritative. However, regardless of how large the number of patients included was, the validity of the conclusions is dependent more on the quality of the data than on their quantity. These meta-analyses have important weaknesses: -

The quality of many of the studies included in the meta-analyses is poor. A large proportion used methods for estimating the AVA which are wholly discredited, such as using the labeled valve size, or the reported in-vitro geometric area, rather than the echo-calculated AVA in patients; moreover, in all the meta-analyses the authors note the inhomogeneity of the studies included.

Flow has become central in the assessment of native AS since Hachicha et al.’s seminal paper on paradoxical low-flow, low-gradient AS³⁵. Abbas et al³⁶ reported over 1,600 patients after TAVR or SAVR, from the PARTNER trials. They demonstrated clinical utility of TAVR when compared to SAVR, focusing on PPM in conjunction with low flow (stroke volume index <35ml/m²) across the artificial valve, and its impact on outcomes after TAVR and SAVR (Table 3).

There were markedly fewer cases of severe PPM after TAVR than after SAVR in all patients (9% versus 28%, P<0.0001), and PPM was associated with a high risk for readmission to hospital after both TAVR and SAVR, while the combination of severe PPM with low stroke volume index (a marker of low flow) after TAVR was associated with mortality (OR: 1.85 (CI 1.06 – 3.23)). For SAVR, the combination of severe PPM, low flow and low LVEF was associated with both mortality and readmission to a hospital. Somewhat surprisingly, severe PPM with normal flow was not associated with clinical outcomes despite higher gradients and smaller valves compared with severe PPM in low flow. The authors concluded that severe PPM occurs more often after SAVR than after TAVR and increases mortality when associated with low flow after TAVR, and with low flow and low LVEF after SAVR. This is intriguing (why would LVEF have an impact on mortality after SAVR, but not after TAVR?) but may represent the effect of incomplete data (LVEF was reported only in 64% of the SAVR group with severe PPM and low flow, compared to 84% in the TAVR group).

Ternacle et al.³⁷ also reported on PPM in the PARTNER trials. In a cohort of cclude. 1,000 patients, they demonstrated that both moderate and severe PPM, in both SAVR and TAVR, were significantly less common if, instead of using the continuity equation-measured AVA in each individual patient, the predicted AVA (provided by the manufacturer) was used. According to the predicted EOAi, 18% had moderate PPM and 2% had severe PPM (overall 20%). After adjustment for obesity, severe PPM was observed in only 1% of patients (P < 0.001 vs. unadjusted cutoff values of EOAi). The incidence of PPM was thus markedly lower when using predicted EOAi vs. measured EOAi. Eighty-three percent (n = 401/482) of patients with any degree of PPM and 76% (n = 140/185) of patients with severe PPM based on measured EOAi were reclassified to no PPM with the use of predicted EOAi. In the TAVR sub-cohort, the incidence of severe PPM by predicted area adjusted for obesity was exceptionally low: 0.1%, vs. 15% for PPM defined by measured AVA. Severe PPM by predicted area was strongly and independently associated with adverse outcomes in SAVR, after adjustment for sex and STS score (HR: 3.18; 95% CI: 1.69-5.96; P < 0.001), but there was no association between PPM (regardless of the method used to identify it) and outcomes after TAVR (Table 4).

Sa et al²⁹ reported a patient-level meta-analysis from 23 studies of TAVR, which included 81,969 patients; 19,612 of these patients had PPM. Moderate or severe PPM increased the risk of death compared to no PPM (HR: 1.09 [95% CI: 1.04-1.14]; P < 0.001). Beyond 30 months there was a reversal of the risk of death (0.83 [95% CI: 0.68-1.01]; P = 0.064), but without statistical significance. There was no difference for mortality between moderate PPM and no PPM.

LVOT diameter is the parameter most susceptible to an operator error. The technique for measuring LVOT diameters by echo has been described in detail³⁸ and should be adhered to meticulously. For TAVR there are a few technical modifications of the standard method, also described in detail³². In cases where the LVOT diameter cannot be measured accurately on echo, a CT-measured diameter or the internal dimension of the valve (as described by the manufacturer) should be used³⁹. A reality check is essential—LVOT diameters <20mm are rare, and efforts should be made to ensure the cross-section where the diameter is imaged genuinely aligns with the maximal dimension of the LVOT.

Inaccurate placement of the sample volume (SV) in the LVOT. The pulse wave Doppler trace of LVOT flow is also susceptible to measurement error and operator variation. This error could be augmented by an inaccurate measurement of the LVOT diameter, due to the squaring of this parameter in the continuity equation. The placement of the SV should be achieved under guidance with color flow mapping, avoiding areas of flow turbulence and aliasing. A trace with minimal spectral width is preferable to one with a wide distribution of the velocities (indicating turbulence). Note that in the presence of systolic anterior motion (SAM) of the MV, or of other causes of flow acceleration in the subaortic region, the continuity equation becomes inaccurate.

There has been renewed interest in this topic due to its significance in the TAVR era⁴⁰, although its effect of potentially severe (up to 80%) overestimation of pressure gradients across artificial valves by echo has been known for decades⁴¹. AVA will be measured as spuriously reduced in patients with significant pressure recovery, and this effect can be mitigated by using the energy loss index (ELI) in patients with a sino-tubular junction diameter <30mm⁴². If needed, pressure gradients can be safely elucidated invasively in mechanical prosthetic valves by crossing the valve with a coronary pressure wire⁴³ while bioprostheses can be crossed safely with standard J-wires and pigtail catheters.

This is an elementary error, and although in hypertrophic obstructive cardiomyopathy (HOCM) the distinction can be difficult, with artificial valves, and in the absence of MV SAM, the distinction should be straightforward. The AV continuous wave (CW) Doppler signal is triangular, dagger-shaped, and bracketed by a short, dense opening and closing linear artefacts (in the mechanical heart valves) as well as sounds harsh on the audio channel; the MV signal is rounded, usually less dense, with a ‘puffing’ quality of the audio signal and overlaps the opening and closure artefacts.

Any condition that increases the LV stroke volume will increase the gradient across the AVR, and these should be eliminated before incriminating the valve. Anaemia, volume overload, and hyperthyroidism are relatively frequent causes.

This is a highly significant diagnosis, because it may be due to a life-threatening condition, and it is often treatable, usually with a significant degree of urgency. Causes of valve obstruction include those that interfere with the motion of the occluders/leaflets (such as thrombus, vegetations, or structural valve deterioration in bioprostheses) and those where cclude/leaflet motion is maintained (pannus). The distinction is not absolute, because obstructive thrombus and subvalvular pannus often coexist⁴⁴. Hypoattenuated leaflet thickening (HALT) and hypoattenuation affecting motion (HAM) are new entities (due to lamellar thrombus on bioprostheses) that have been identified due to intensive CT surveillance of TAVR valves, as part of research studies⁴⁵ and subsequently of SAVRs⁴⁶. Through a judicious combination of imaging tests—including TTE, fluoroscopy for assessing cclude motion in mechanical valves, TOE and CT (gold standard for HALT, HAM, and for pannus)—the correct diagnosis can usually be made. A simple and pragmatic algorithm for the echocardiographic diagnosis of valve obstruction and its differentiation from patient-prosthesis mismatch is presented in (Figure 4).

Figure 4

Preoperative prosthetic valve sizing has been advocated by Pibarot et al.³ and involves preoperative measurements of the patient’s BSA and the LVOT dimensions using echocardiography. Using the patient’s BSA, the surgeon should firstly calculate the minimal valve EOA needed to ensure an iEOA >0.85 cm²/m². Using widely available charts from prosthetic valve manufacturers and the literature^47,48, the surgeon should then identify the minimum size of the given prosthetic valve intended to be used that would avoid PPM and can be implanted in the measured LVOT. This should be regarded as the first step in the preoperative planning and should help the surgeon identify patients at risk of moderate or severe PPM, as well as guide the surgeon further in their decision-making.

This strategy has not, to our knowledge, ever been tested in the ‘real world’. Its implementation is likely to meet with some difficulties because surgeons often have their own idiosyncratic preferences for a specific valve type, or model, or manufacturer, and may not be as flexible as needed to conform to the choice mandated by the algorithm. Moreover, it is possible that even with the best intentions, a valve size that would avoid PPM may just not exist, for instance in a large patient who happens to have a small LVOT (Figure 5).

Figure 5

All stented prosthetic valves incorporate a sewing cuff into their design which will render the true EOA smaller than the absolute size of the sewing ring. Yet, the newer, modern bioprostheses have been designed with lower profiles for complete supra-annular placement, allowing for larger sized valves with lower pressure gradients and improved haemodynamics, thus reducing the incidence of PPM^49,50. But despite the improved haemodynamics across these modern bioprosthetic valves, the main concern with these valves remains the early structural valve deterioration, often leading to the need for further intervention. However, the trends in using bioprostheses in younger populations have seen an increase in recent years, with recent reports showing a 28% increase in bioprosthesis use in 2018 in the United Kingdom in patients under the age of 50 and further after^51,52.

By contrast, stentless and sutureless bioprosthetic valves lack the rigid sewing ring and allow a higher EOA for a similar size when compared to the stented biopresthetic valves. Multiple studies have shown lower rates of PPM when these types of valves were used^53–55. Although the use of sutureless valves is advocated by some authors to reduce the incidence of PPM in patients with a small aortic root, these valves have been associated with an increased risk of conduction issues and paravalvular regurgitation⁵⁵. Similarly, when the aortic valve is replaced through a transcatheter approach, the rates of PPM are lower, owing to the profile of the percutaneous-deployable valves, and the smaller footprint required, allowing for larger valve areas, especially when dealing with small aortic roots⁵⁶. In a large study from the STS/ACC TVT Registry, looking at almost 64,000 patients receiving de novo TAVR or valve-in-valve TAVR, the incidence of severe PPM in the de novo TAVR group was only 5.3%⁵⁷.

Despite the lower incidence of PPM after TAVR or rapiddeployment valves, the use of any bioprosthetic valves in young and middle-aged adults should be discouraged due to the inevitable risk of re-operation and poorer mid and long-term outcomes, particularly in the context of a small aortic annulus^58,59. Not only does a re-operation pose technical challenges, but the indication for re-intervention only appears when the patient develops severe aortic stenosis again, thus translating into further myocardial strain, LV dysfunction, and heart failure.

Lastly, the most important prosthetic option to be considered in patients at risk of moderate or severe PPM is the mechanical prosthesis. These have shown to have better haemodynamics with higher EOA than the equivalent in size bioprostheses. As such, this type of valve prostheses has been associated with reduced rates of PPM, even in patients with small aortic annuli^60,61. Moreover, some of the newer generation bileaflet mechanical valves are associated with very low incidence of thrombotic events and have been approved in use with lower INR targets^62–64. The Sorin Bicarbon Overline (currently manufactured by Corcym, Corcym UK Limited, London, United Kingdom) is one of them and offers even better EOA in a supra-annular position⁶⁵. In spite of lower thrombogenicity, better haemodynamics, and EOA compared to bioprosthetic valves and more importantly, lifelong durability, mechanical valves have seen a fall in their in recent years, which could probably be related to the need for life-long anticoagulation with warfarin⁵².

A few examples of the newer generation prosthetic valves are shown in (Figure 6).

Figure 6

Different tools (Cardiovalve, Digimednet; Valve PPM, KRUTSCH or PPM calculator, Medtronic) are now available online and on smartphones and will allow the surgeon to predict the new iEOA following implantation of a prosthetic valve, and even more, some of these applications can, in fact, recommend different types, manufacturers, and minimum sizes to prevent moderate or severe PPM based on databases of referenced estimated EOA from various manufacturers. If patients are identified to be at risk of PPM irrespective of valve manufacturers and models, particularly with echocardiographic data available preoperatively, these patients should be counseled and offered an alternative approach in order to prevent PPM.

Different algorithms and flowcharts have previously been proposed to aid the surgeon in preoperative decision making ^3,66. We also present a flow-chart on the approach of the patient requiring aortic valve replacement (Figure 7).

Figure 7

Although there is no consensus to date regarding a cutoff value for the definition of a small aortic annulus, mainly due to the inconsistency of definitions and prostheses sizes by the manufacturers, in multiple series^66–68, the aortic annulus was considered ‘small’ when this could not accommodate a minimum size 21 valve. When this is the case, in particular in patients at risk of severe PPM, the surgeon should utilize another strategy to minimize the risk by employing different techniques of aortic root enlargement or aortic root replacement.

Regardless of the ARE technique involved, there is some anxiety among surgeons due to controversy over the possible negative impact of these techniques on outcomes⁶⁹. As expected, adding a technique for annular enlargement to the SAVR will increase the cardiopulmonary bypass and cross-clamp times. In a large recent meta-analysis⁷⁰, SAVR with ARE was associated with increased operative mortality, but this was likely attributed to the increased risk from combined procedures like MV surgery or CABG, rather than in isolated SAVR with ARE. However, despite some of the reported risk for peri-operative adverse outcomes, most of the studies have shown that the long-term survival of SAVR with ARE remains similar to the one after SAVR without ARE^69–71, but minimizes the risk of PPM and provides improved overall haemodynamics with the implantation of a larger prosthetic valve. In one of the largest studies to date⁷², which included over 7,000 patients, Ouzounian and her team showed that ARE was not associated with an increased risk of mortality or adverse perioperative events and is in fact a safe adjunct to conventional SAVR. Their study emphasized the safety and the benefits of ARE and these attainable results should be the focus of all surgeons.

Several techniques for ARE can be used, but they could be grossly classified into anterior and posterior techniques. Most commonly performed techniques are, however, the posterior ones.

The main anterior annular enlargement technique is the Konno-Rastan procedure, also known as aortoventriculoplasty. This was first described separately by Konno in 1975⁷³ and Rastan in 1976⁷⁴, and was first designed to relieve subvalvar, valvar, and supravalvar stenosis in the paediatric population with congenital aortic stenosis. A longitudinal aortotomy is performed which is continued caudally into the interventricular septum (Konno incision), passing 4-5mm left to the origin of the right coronary artery. A further curved incision is then carried from the level of the aortic annulus into the right ventricular outflow tract, thus exposing both the left ventricular outflow tract and the aortic root. A separate incision is then required to open the LVOT in the basal septum. Once the larger sized prosthesis is sutured to the intact part of the aortic annulus, the incisions are closed using Dacron patches. The most common technique for closing the incisions is the use of two separate triangular-shaped Dacron patches. The first patch is sutured onto the interventricular septum and to the LVOT, with the widest part of the patch now covering and sutured to the aortic prosthesis. The second patch is used to close the right ventricular outflow tract. Out of the ARE techniques, the anterior approach is considered to bring the highest degree of annular enlargement. However, this comes at a cost of increased morbidity. Higher incidence of intraventricular conduction problems, injury to septal perforators or intracameral fistulae have been reported⁷⁵. This is mainly due to the Konno incision within the inter-ventricular septum. Furthermore, the complexity of this procedure makes its use very limited and only by experienced surgeons, mainly in a small, tubular LVOT.

On the other hand, posterior annular enlargement techniques can be more easily and safely performed, without increasing morbidity and as described by some, mortality^71,72,76. In the order of the degree of enlargement offered, these techniques are the Nicks, Nunez, and Manouguian/Rittenhouse (Figure 8). These techniques can enlarge the aortic annulus by one or two sizes of the prosthetic valve.

Figure 8

Nicks technique⁷⁷ uses an oblique aortotomy that continues inferiorly towards the nadir of the non-coronary sinus. The incision is continued across the aortic annulus up to the origin of the mitral valve. A patch is sutured to the fibrous origin of the mitral valve using two mattress sutures which are continued to either side of the incision. The larger prosthesis is then implanted with part of the sewing ring being sutured to the patch from the outside. The upper part of the patch is used within the aortotomy closure thus allowing supra-annular enlargement of the aortic root.

Manouguian’s technique⁷⁸ also uses an oblique aortotomy that continues into the commissure between the left and the non-coronary cusps and extends inferiorly past the aortic annulus and the fibrous origin of the mitral valve and into the anterior leaflet of the mitral valve up to a few millimetres from the hinge point, while posteriorly this extends into the roof of the left atrium. A tear-drop shaped patch is then used to augment the incision within the anterior leaflet of the mitral valve and similarly to the Nicks technique, each running suture is continued superiorly into the aortic annulus and finally into the aortotomy closure. The roof of the left atrium is then separately sutured to the patch. Rittenhouse⁷⁹ and his colleagues described a similar technique whereas the roof of the left atrium is closed using a separate triangular-shaped patch. Although this allows for the highest degree of annular augmentation based on the length of the incision, this technique is associated with increased risk of post-operative mitral prolapse or aorto-atrial fistulae⁸⁰.

The Nunez technique⁸¹ uses the same oblique aortotomy as the Manouguian-Rittenhouse technique but stops before the level of the fibrous origin of the mitral valve. Although limited to the level of the aortic annulus, this technique can upsize the prosthetic valve by up to two or three sizes. Furthermore, it offers the advantage of having strong anchoring structures for the closure patch and avoids the opening of the left atrial roof.

The need for larger prostheses to avoid PPM and the controversy around the risk-benefits of standard ARE techniques have led surgeons to seek further technical methods of ARE. One of these novel techniques was recently described by Yang and his colleagues⁸², using a Y incision and a rectangular-shaped patch rather than an oval shape to enlarge the aortic annulus by up to two to four sizes. An oblique aortotomy is performed similarly to the Nunez technique, and this is continued into the aorto-mitral curtain and close to the roof of the left atrium. From this point, the incision is extended into a “Y” fashion across the level of the aortic annulus and towards nadirs of both the left coronary sinus and the non-coronary sinus. A rectangular-shaped patch is then sutured to the fibrous mitral annulus and continued on each side to the undermined aortic annulus, then sutured along the longitudinal length of the patch to the point of the incised commissure. Once the bigger prosthesis is implanted and secured in place, the aortotomy is closed with the tapered end of the patch. Although there is limited data with this novel technique, the simplicity and safety of it compared to the Manouguian or Konno techniques along with the better efficiency in upsizing compared to the Nicks technique, makes this procedure a promising option for PPM prevention in the small aortic root.

In search of the ideal substitute for the diseased aortic valve, Ozaki and his team published a new technique aimed at reconstructing the aortic valve leaflets using glutaraldehyde-treated autologous pericardium⁸³. In a recent meta-analysis, Mylonas et al.⁸⁴ commented on the safety of the Ozaki procedure in the patient with a small aortic root citing outstanding results and very low risk for PPM with this procedure. But it would be safe to say that this procedure is still a novelty, with only a limited number of cases being performed globally. As such, more data and follow-up are required before strong conclusions can be drawn.

Another option in selected cases, although inevitably increasing the complexity of SAVR, is the full replacement of the aortic root using a valved conduit, also called the Bentall procedure⁸⁵, allowing the insertion of a true supra-annular aortic valve, integrated into the Teflon conduit. The procedure was first described by Bentall and De Bono in 1968⁸⁵, and has seen different modifications throughout the years^86,87. In patients with a small aortic annulus, the composite graft can be carefully oversized thus allowing for better haemodynamics and therefore significantly reducing the risk of PPM⁸⁸.

In rare patients who require combined aortic and MVR, an elective commando procedure can be performed. This fibrous skeleton reconstruction technique is well-described in patients with infective endocarditis. It has recently been reported as an elective approach for double valve replacement allowing for patch enlargement and subsequently larger sized valves⁸⁹.

Finally, particularly in younger adults, a pulmonary autograft or the Ross procedure should be considered in patients at risk of PPM. This was first described back in 1967 by Donald Ross⁹⁰, and has since been used as a first option in the paediatric population.

More recently, this procedure has gained popularity as a natural complement in aortic valve repair programmes, when the aortic valve cannot be repaired. Studies from the largest registry to date, the German-Dutch Ross Registry have shown a mortality rate similar to the one of the general population’s, however, the survival advantage disappears around the age of 60, making this procedure suitable for young and selected middle aged adults^91,92.

The pulmonary valve has been shown to have ideal haemodynamics when transposed into the aortic position, with little or no structural valve deterioration owing to the ‘living nature’ of the native valve. There is increasing emerging evidence that supports the stance that the Ross procedure should be considered as the first option in the young adult who presents with aortic valve disease and when the valve is not amenable to repair^93–95.

On top of the favorable haemodynamic performance and durability offered by the autograft, translated into low transvalvular gradients and excellent left ventricular mass regression, the fact that the living valve ‘grows with the patient’ also significantly reduces the risk of PPM.

The benefits of the Ross procedure also include improved quality of life, avoidance of lifelong anticoagulation and the complications associated with it, particularly in women of childbearing age where anticoagulation with vitamin K antagonists is contraindicated. Furthermore, in young and middle-aged adults, bioprosthetic valve have a prohibitive risk of structural deterioration leading to re-operation, while mechanical prostheses are noisy, with an impact on quality of life in some patients.

The main limitation of the Ross procedure comes however from the fact it replaces a single-valve pathology with a double-valve pathology, introducing the added potential risk of needing further interventions on the homograft used in the pulmonary position. Moreover, the pulmonary autograft in the aortic position is at risk of dilatation over time⁹², adding a degree of angst among surgeons, on top of requiring a high level of expertise due to the complexity of this procedure. As such, the Ross operation is absolutely contraindicated in patients with diagnosed connective tissue disorders or familial aortopathies due to a high risk of autograft dilatation and failure⁹⁶. Furthermore, the Ross procedure is also contraindicated in patients with a life expectancy of less than 15 years or patients with certain autoimmune conditions (e.g., rheumathoid arthritis or lupus erythematousus).

With its increasing popularity, increasing numbers of surgical groups are reporting different variations of the original procedure and methods to reduce the risk of complications following the Ross procedure⁹⁶. Most of these modern revisions focus on stabilizing the new pulmonary autograft and preventing future dilatation of the neoroot (Figure 9). One of the most common modifications is the semiinclusion technique, as it was described by an Australian group led by Skillington⁹⁷, where the pulmonary autograft is included within the native aortic root. Using this technique, the rate of dilatation of the pulmonary autograft at 15 years dropped to only 5%.

Figure 9

The landing zone of the pulmonary autograft does, however, need size-matching to the pulmonary autograft, meaning the native aortic annulus needs often to be enlarged, or rarely, plicated. The Ross procedure can be easily combined with the Kono enlargement technique⁹⁸, and depending on the extension of the enlarging incision, this is classed as mini-Konno (only the fibrous annulus is divided) or a true Ross-Konno (the annular enlargement incision extends into the body of the interventricular septum).

Newer technologies have permitted surgeons to advance the Ross procedure even further, with bespoke personalized external aortic root support⁹⁹. Although this is still a new addition to the surgical armamentarium with little supporting data, the Ross-PEARS procedure promises to counteract the issues of autograft dilatation by including it into a bespoke custom-made mesh, tailored-fitted through magnetic resonance or tomographic 3D reconstruction.