Influence of Prosthesis-Patient Mismatch on Survival with Aortic Valve Replacement

Prosthesis-patient mismatch (PPM) was first described over 30 years ago (Rahimtoola, 1978) for aortic valve replacement: when the in vivo effective orifice area (EOA) of the prosthetic valve is less than that of the native, non-diseased, human valve. Extensive documentation on the role of PPM after aortic valve replacement (AVR) particularly addresses left ventricular mass regression and patient survival. Controversy continues about the influence of PPM on patient survival, both early and late mortality. Many studies (Pibarot and Dumesnil, 2000; Muneretto et al., 2004; Mohty et al., 2006; Tasca et al., 2006; Moon et al., 2006; Florath et al., 2008; Mohty et al., 2009; Blais et al., 2003) report PPM to be an independent predictor of mortality while others (Jamieson et al., 2010; Kato et al., 2007; Vicchio et al., 2008; Mascherbauer et al., 2008; Monin et al., 2007) showed no significant effect of PPM on patient outcome. There is also debate about whether the control of PPM reduces congestive heart failure and regression of the left ventricular mass, thereby contributing to improved survival. Several Canadian centers have been actively involved in this area of research, namely the Laval University group led by P. Pibarot, J.G. Dumesnil and D. Mohty, the UBC group led by W.R.E. Jamieson, and the University of Ottawa group led by M. Ruel and A. Kulik. PPM is categorized by Pibarot and Dumesnil (2000), Mohty et al. (2009), and Jamieson et al. (2010) as normal (EOA index (EOAI) of > 0.85 cm2 / m2), mild-to-moderate (> 0.65 cm2 / m2 to ≤ 0.85 cm2 / m2), and severe (≤ 0.65 cm2 / m2). Tasca et al. (2006) defined PPM as an EOAI of ≤ 0.80 cm2 / m2, Moon et al. (2006) as an EOAI of < 0.75 cm2 / m2, while Ruel et al. (2004), Kulik et al. (2006), Kato et al. (2007), and Monin et al. (2007) as EOAI of ≤ 0.85 cm2 / m2; Florath et al. (2008) and Vicchio et al. (2008) chose 0.60 cm2 / m2 as the cutoff between moderate and severe PPM. As can be seen, there is no clear consensus on the exact definition of PPM; this lack of consensus may contribute at least in part to the observed discrepancies in the conclusions of the studies. The studies also differ in the length of their patient followup. Jamieson et al. (2010) report survival to 15 years, Moon et al. (2006) and Mohty et al. (2009) to 12 years, and the majority of the other publications on the topic of PPM report survival from 4 to 8 years (Mothy et al. 2006; Tasca et al., 2006; Florath et al., 2008; Kato et al., 2007; Mascherbauer et al., 2008; Monin et al., 2007). These differences may also contribute to the different conclusions reached.


Introduction
Prosthesis-patient mismatch (PPM) was first described over 30 years ago (Rahimtoola, 1978) for aortic valve replacement: when the in vivo effective orifice area (EOA) of the prosthetic valve is less than that of the native, non-diseased, human valve. Extensive documentation on the role of PPM after aortic valve replacement (AVR) particularly addresses left ventricular mass regression and patient survival. Controversy continues about the influence of PPM on patient survival, both early and late mortality. Many studies (Pibarot and Dumesnil, 2000 I t s h o u l d b e n o t e d t h a t t h e i n d i c a t i o n f o r s u r g i c a l m a n a g e m e n t o f a o r t i c s t e n o s i s i s symptomatic severe aortic stenosis (< 1.0 cm 2 valve area). In the majority of patients, this is equivalent to an EOAI at or below the level of severe mismatch by our definition.

The influence of PPM on postoperative patient outcomes
The objective of our study (Jamieson et Figure 1). There is a misconception with the Carpentier-Edwards supra-annular aortic valve for the early version (prior to 1985) of the mitral valve failed because of stent-post dehiscence due to excessive trimming of the aortic wall; however, this failure mode was identified in only one aortic prosthesis before the manufacturing trimming was changed (Jamieson et al., 2005;Jamieson et al, 2009). The level of PPM was classified for each patient based on reference EOAs and size for each prosthesis in the published literature. The patients considered for the study had their first aortic valve replacement. Patients who had a subsequent valvular replacement were censored alive on the date of the reoperative procedures. This concept was to avoid a hemodynamically different prosthesis at the time of reoperative explantation.

Carpentier-Edwards
Carpentier Overweight or mild-to-moderately obese patients had a lower risk of early mortality, while underweight and severely obese patients had a higher risk of late mortality. When patients were analyzed as normal/underweight or overweight/obese, those with a normal EOAI had better 15-year survival than those with severe PPM. After adjusting for EOAI, age > 60 years and EF ≤ 50% indicated a higher risk of overall (early + late) mortality within BMI categories. These results suggest that BMI is associated with survival after AVR and that PPM may modify the effect. EOAI was also evaluated as a continuous variable (along with other variables except EF), as well as a categorical variable, which revealed that EOAI was not an independent risk fac tor f or la te (> 30 day s) or ov era ll mo rta li ty. Th e p red i ct ors , o the rwi se , wer e n ot different from the categorical modeling except for the elimination of valve size and the addition of BMI for early mortality. Valve type was eliminated for late mortality and overall mortality. The survival curves in Jamieson et al. show that severe PPM (EOAI of ≤ 0.65 cm 2 /m 2 ) reduces survival for patients > 60 years old but not for patients ≤ 60 years old, that severe PPM reduces survival for patients with a BMI ≥ 25kg/m 2 but not for those with a BMI < 25 kg/m 2 , and that severe PPM reduces survival for patients with an ejection fraction > 50% but not for those with an EF ≤ 50% ( Figure 5). In comparison, Mohty et al. found that severe PPM was associated with increased mortality in patients < 70 years old but not in older patients, and that it significantly affected survival in patients with a BMI < 30kg/m 2 but not in those with a BMI ≥ 30kg/m 2 ( Figure 6A, 6B, 6C, 6D). They also found moderate-to-severe PPM to be an independent predictor of late mortality in patients with a pre-operative LVEF < 50% but not in those with preserved LV systolic function ( Figure 6E, 6F). With regard to these discrepancies, it is worth noting that there were only 21 patients in the Jamieson et al. BMI < 25 kg/m 2 severe PPM group and 39 patients in the LVEF ≤ 50% severe PPM group, while for the severe PPM subset of the Mohty cohort, there were fewer than 20 patients in each of the < 70 years old, ≥ 70 years old, BMI < 30 kg/m 2 , and BMI ≥ 30kg/m 2 subgroups. We therefore believe that the discrepancies in the above results may be purely due to random variations in the small data sets, and that if given an adequate number of cases in each of the categories, there may be no differences in the results between the Jamieson et al. and the Mohty et al. groups. Ruel et al. (2006) found that PPM primarily affected patients with impaired left ventricular function at the time of AVR, and patients in whom PPM was associated with decreased overall long-term survival, lower freedom from heart failure, and diminished left ventricular mass regression. Also, an EOAI ≤ 0.85 cm 2 / m 2 did not have a significantly detrimental effect in patients with normal preoperative left ventricular function. However, the authors pointed out that PPM might have been found to have a significant effect in the normal LV function cohort had they evaluated cases with severe mismatch (≤ 0.65 cm 2 / m 2 ). An earlier study by Ruel et al. (2004) had shown that although PPM had significant effects on cardiac end points (occurrence of congestive heart failure, etc), it had no effect on overall survival after AVR. Kulik et al. (2006) found that patients with low-gradient aortic stenosis (LGAS, defined as an aortic valve area of < 1.2cm 2 , a mean transvalvular pressure gradient of < 40 mmHg, and a LVEF of < 50%) have worse long-term outcomes after AVR, and that PPM further adversely affects the long-term outcomes of LGAS patients and should therefore be avoided in this population.    (Higgins et al., 2011) that evaluated the influence of gender on early, late, and overall survival reported that the predictors of mortality after AVR for aortic stenosis differed between male and female patients. Female gender was a predictor of early mortality while male gender was a predictor of late (but not early or overall) mortality. Male gender increased the risk of late mortality, and a valve size ≤ 21 mm increased the risk of early and overall mortality among male patients only. These differences need to be taken into consideration preoperatively and require consideration during operative management. The Jamieson et al. analysis indicated that severe PPM identified with an EOAI < 0.65 cm 2 /m 2 is not an independent predictor of early mortality, late mortality, or overall mortality after AVR. These findings have been discussed in perspective with other studies that have and have not provided evidence of PPM as an independent predictor of survival. The independent influence of bioprostheses as a risk factor of late and overall mortality also needs extensive evaluation because currently bioprostheses are recommended for patients ≥ 60 years old to minimize serious valve-related morbidity and provide a relatively acceptable degree of valve-related reoperation for structural valve deterioration. Valve-related mortality is not influenced by valve type (bioprosthesis or mechanical prosthesis). The documented finding that AVR does not provide the same age/gender matched survival as in the general population allows this lower age threshold for bioprostheses in AVR (van Geldrop et al., 2009). This earlier failure threshold may be related to residual systolic dysfunction and more likely related to diastolic dysfunction concomitant with PPM (Nozohoor et al., 2008).

A suggested approach to PPM
Because the negative impact of severe PPM on postoperative survival, it is crucial to avoid leaving patients with severe PPM after valvular surgery. Pibarot and Dumesnil (2000) presented a 3-step approach for preventing PPM: [1] calculate the patient's body surface area from weight and height; [2] using a BSA versus EOAI table, find the minimal valve EOA (in cm 2 ) that will allow a given patient to have proper (ideally > 0.85 cm 2 / m 2 ) EOAI after surgery; and [3] select the type and size of prosthesis that has EOA reference values equal to or greater than the minimal valve EOA value obtained in step 2. The occurrence and severity of postoperative PPM can also be predicted before the operation from the patient's BSA and the reference EOA value of the selected prosthesis (Pibarot et al., 2001;Urso et al., 2010;Dumesnil and Pibarot, 2010).
In agreement with the above, despite failing to find severe PPM (< 0.65 cm 2 / m 2 ) as an independent predictor of early, late, or overall mortality after AVR, we recommend that surgeons do not leave patients with a severe mismatch (especially for bioprostheses, which may develop degenerative changes over time that would further reduce the EOAI). Surgeons should maintain a prospective strategy of implanting an adequately sized aortic prosthesis that will preclude patients from being in the category of severe mismatch (near equivalent to indications for intervention in severe aortic stenosis). However, a significant portion of patients undergoing AVR will have some level of mild-to-moderate PPM owing to the intrinsic obstructive nature of most prostheses, and Jamieson et al. (2010) should provide some confidence to surgeons and cardiologists that mild-to-moderate PPM is unlikely to be detrimental to survival. Other than selecting a prosthesis with sufficient EOA, as described above, there are several more intraoperative options available to surgeons to prevent the occurrence of severe PPM. Aortic root enlargement may be considered in patients with an elevated risk of developing moderate-to-severe PPM at time of valvular replacement surgery (Mohty et al., 2006). Kulik et al. (2008) were able to insert larger prosthetic valves and achieve lower PPM by doing aortic root enlargement (ARE) at the time of AVR. They reported that the addition of an ARE to AVR increased the aortic cross-clamp time by 9.9 minutes, on average, and that there was no significant increase in perioperative morbidity or mortality associated with the added ARE. However, the lower incidence of PPM did not significantly affect long-term outcomes in their AVR + ARE cohort, once again coming back to the question of whether PPM significantly affects survival. The third option is a total aortic root replacement. Compared with a traditional stented bioprosthesis, total root replacement allows for optimal hemodynamics with no significant aortic regurgitation, improved regression of the LV mass, and less PPM in the small aortic root ( However, 5 of the 9 patients who underwent concomitant major cardiac procedures at the time of valvular replacement died in-hospital, so there is a risk to reoperation. The benefit of relief from PPM must be weighed carefully against the risks of reoperation, and must be assessed on a patient-by-patient basis. When evaluating patients with mild-to-moderate PPM for the possibility of reoperation, we suggest that surgeons take into account the Jamieson et al. (2010) finding of the unlikelihood of mild-to-moderate PPM contributing to worse survival. From the accumulated data from published literature, it is easy to see that the topic of prosthesis-patient mismatch remains controversial. The issue is further complicated by the fact that there are several levels of PPM (nonsignificant, mild, moderate, or severe), with different studies showing different outcomes for each level of PPM. There is also currently no clear consensus on the exact definitions of PPM and its categories. A sensible approach to the issue of PPM is that we should avoid generalizations for any given level of PPM except for severe PPM, for which the data in the existing literature is more consistent; therefore, proactive measures should be taken to prevent its occurrence. For other levels of PPM, it is reasonable to evaluate each patient on an individual basis (i.e., moderate PPM being more acceptable for a sedentary elderly patient, but less so for someone who is younger and more active), and for surgical and postoperative management options to be dependent on the individualized assessment.