Quality of life of elderly patients following valve surgery forchronic organic mitral regurgitation§

Francesco Maisano *, Giorgio Vigano, Chiara Calabrese, Maurizio Taramasso,Paolo Denti, Andrea Blasio, Andrea Guidotti, Ottavio Alfieri

Cardiac Surgery, San Raffaele University Hospital, Via Olgettina 60, 20132 Milano, Italy

Received 29 August 2008; received in revised form 9 February 2009; accepted 18 February 2009; Available online 15 April 2009

Abstract

Objective: Mitral valve surgery for organic mitral regurgitation (MR) in the elderly has been debated. In the elderly, quality of life is a betterindicator of surgical success than survival. We assessed quality of life of elderly patients submitted to surgery for MR using the Minnesota Livingwith Heart Failure (MLHF) questionnaire. Methods: Between August 2003 and August 2006, 225 consecutive patients >70 years old underwentsurgical treatment of organic MR. Mean age was 77 � 3.3 years (range 71—87 years). Mean EF was 50 � 11%. Degenerative disease was the mostprevalent (77%) etiology. CABG was associated in 25% of patients. Mean Charlson score was 4.3 � 1.5 and 101 patients (45%) were NYHA class IIIand IV. Hospital survivors were followed up and quality of life by MLHF score was assessed. Results: Mitral repair and replacement were equallydistributed in this population. Hospital mortality was 2.7%. Late survival was 91 � 1.9% at 3 years. MLHF was obtained from 204 patients at mean2 � 1 years of follow-up. MLHF score was 38 � 18; there were 135 (66%) patients with MLHF >30. MLHF tended to increase with age at follow-up(p = 0.007). Multivariable predictors of MLHF were preoperative atrial fibrillation ( p = 0.019), diabetes ( p = 0.03), higher creatinine level(p = 0.0009), higher EuroSCORE ( p = 0.02), residual mitral regurgitation grade at follow-up echocardiography (p < 0.0001) and systolicpulmonary artery pressure at follow-up (p = 0.04). Type of surgical treatment (repair vs replacement and choice of prosthesis) did not predictMLHF at follow-up, although those who had recurrent MR after repair had the highest scores compared to patients who had repair and durableresult and those treated by replacement (MLHF was 51 � 21, vs 34 � 16, vs 39 � 18, respectively, p = 0.0013). Conclusions: Quality of lifefollowing mitral valve surgery is suboptimal in more than half of elderly patients. MLHF score at follow-up is mostly related to preoperativeconditions. Type of surgery does not influence MLHF score, however, quality of life is worse in patients with recurrent/residual MR followingrepair.# 2009 European Association for Cardio-Thoracic Surgery. Published by Elsevier B.V. All rights reserved.

Keywords: Mitral regurgitation; Mitral valve surgery; Mitral valve replacement; Mitral valve repair; Surgery in the elderly; Quality of life

www.elsevier.com/locate/ejctsEuropean Journal of Cardio-thoracic Surgery 36 (2009) 261—266

1. Introduction

Mitral valve surgery in the elderly is a debated issue.Although an increasing number of elderly patients arereferred for mitral valve surgery, due to the higher risk ofmorbidity and mortality following cardiac surgery, manyelderly patients are refused or not even referred forsurgery [1]. Since survival benefit is difficult to demon-strate, relief of symptoms and improvement of quality oflife are the main objectives of surgery in this subgroup ofpatients. Quality of life after mitral valve surgery in theelderly has not been sufficiently investigated yet. Weretrospectively reviewed our experience on mitral surgeryfor organic mitral valve regurgitation in patients older

§ Presented at the 22nd Annual Meeting of the European Association forCardio-thoracic Surgery, Lisbon, Portugal, September 14—17, 2008.* Corresponding author. Tel.: +39 02 26437111; fax: +39 02 26437125.E-mail address: francesco.maisano@hsr.it (F. Maisano).

1010-7940/$ — see front matter # 2009 European Association for Cardio-Thoracicdoi:10.1016/j.ejcts.2009.02.055

than 70 years of age to evaluate clinical outcomes aswell as to prospectively determine late quality of life usingthe Minnesota Living with Heart Failure (MLHF) question-naire [2].

2. Methods

The study included 225 elderly patients (age olderthan 70 years) who underwent mitral valve surgeryfor severe organic mitral regurgitation (MR) betweenAugust 2003 and August 2006, either as an isolatedprocedure or in combination with aortic valve replacementto treat concomitant aortic regurgitation. Patients withpreoperative aortic stenosis and/or functional mitralregurgitation have been excluded from the analysis. Themean age was 77 � 3.2 years. There were 179 patients withan age between 71 and 79, and 46 patients aged 80 years ormore. Clinical characteristics of the overall population are

Surgery. Published by Elsevier B.V. All rights reserved.

F. Maisano et al. / European Journal of Cardio-thoracic Surgery 36 (2009) 261—266262

Table 1Clinical characteristics of the overall population and of the two age subgroups: septuagenarians (71—79 years old) and octogenarians (>80 years old).

Overall 71—79 >80 p

N 225 179 46Age 77 � 3 75.7 � 2.1 81.9 � 1.9 <0.0001Gender M/F 110/115 88/91 21/25 0.64BMI 24 � 3.3 24.3 � 3.4 23.5 � 3.1 0.18Smoke history 79 (35%) 66 (36%) 13 (28%) 0.29Current smoker 7 (3%) 7 (4%) 0 (0%) 0.07CVA history 12 (5%) 8 (4%) 4 (9%) 0.28AFib 58 (25%) 43(24%) 15 (33%) 0.24AMI history 29 (13%) 25 (14%) 4 (9%) 0.33Peripheral vascular disease 47 (21%) 38 (21%) 9 (20%) 0.8Acute endocarditis 4 (1%) 4 (2%) 0 0.17Diabetes 30 (13%) 27 (15%) 3 (6%) 0.10Serum creatinine 1.06 � 0.39 1.07 � 0.41 1.04 � 0.26 0.6Hypertension 163 (72%) 127 (71%) 36 (78%) 0.31Hyperlipidemia 97 (43%) 80 (44%) 17 (37%) 0.34

COPD 0.02No 179 (80%) 146 (82%) 33 (72%)Mild 16 (7%) 14 (8%) 2 (4%)Moderate 26 (11%) 15 (8%) 11 (24%)Severe 4 (2%) 4 (2%) 0

NYHA class 0.8NYHA I 19 (8%) 15 (8%) 4 (9%)NYHA II 105 (47%) 86 (48%) 19 (41%)NYHA III 90 (40%) 69 (39%) 21 (46%)NYHA IV 11 (5%) 9 (5%) 2 (4%)

EuroSCORE 7.7 � 2.6 7.4 � 2.5 8.7 � 2.5 0.002Charlson score 4.3 � 1.5 4.3 � 1.5 4.2 � 1.2 0.8EF (%) 50 � 11.3 50.6 � 11.1 50.1 � 11.8 0.80PAPs 46.8 � 17.2 46 � 15.0 50.0 � 23.9 0.16MR grade moderately severe 15 (7%) 10 (6%) 5 (11%)MR grade severe 210 (93%) 169(94%) 41 (89%) 0.2Moderately severe or severe AR 58 (25%) 46 (26%) 12(26%) 0.95

CAD 0.16No CAD 160 (71%) 128 (71%) 33 (72%)1 VD 25 (11%) 17 (9%) 9 (20%)2 VD 29 (13%) 26 (14%) 3 (6%)3 VD 11 (5%) 10(5%) 1 (2%)

Tricuspid regurgitation 0.45No 44 (19%) 32 (18%) 12 (26%)Mild 67 (30%) 59 (33%) 8 (17%)Moderate 51 (23%) 38 (21%) 13 (28%)Severe 63 (28%) 50 (28%) 13 (28%)

Table 2Procedural characteristics of the overall population and of the two agesubgroups: septuagenarians (71—79 years old) and octogenarians (>80 yearsold).

Overall 71—79 >80 p

MV repair 119 (53%) 97 (52%) 22 (48%) 0.4

MVR 106 (47%) 82 (46%) 24 (54%)Tissue 95 72 23 0.3Mechanic 11 10 1

Associated AVR 48 (21%) 36 (20%) 12 (26%) 0.38Associated TV surgery 94 (41%) 74 (41%) 20 (43%) 0.8Associated CABG 57 (25%) 47 (26%) 10 (22%) 0.5Urgent surgery 9 (4%) 8 (4%) 1 (2%) 0.4

shown in Table 1. There were no significant differencesamong the younger and older subpopulation, with theexception of a higher incidence of respiratory failure in theoctogenarians. Surgery was performed under electiveconditions in 96% of the cases. Mitral reconstructive surgerywas performed in about one half of the population(Table 2). The patients who were treated by valvereplacement received a tissue valve in the great majorityof cases. Patients undergoing valve replacement had thefollowing statistically significant correlations: female( p > 0.0001), preoperative AFib (0.001), higher pulmonarysystolic pressure ( p = 0.006), higher EuroSCORE( p < 0.0001), rheumatic etiology ( p < 0.0001), associatedaortic valve replacement ( p = 0.03), associated tricuspidrepair ( p = 0.01). Combined tricuspid surgery was per-formed in 94 patients (41%), while aortic valve replacementand CABG were performed in 21% and 25% of cases,respectively.

2.1. Data management and statistical analysis

The study has been approved by the local ethicalcommittee. Preoperative characteristics, operative details

F. Maisano et al. / European Journal of Cardio-thoracic Surgery 36 (2009) 261—266 263

Fig. 1. Mid-term survival of patients according to age at operation: nostatistical difference was found in the 5-year survival between septuagenar-ians and octogenarians ( p = 0.16).

Fig. 2. Relation between MLHF score and NYHA class at follow-up: the dotsrepresent the individual MLHF scores according to the NYHA class group. Thediamonds show the mean and the standard deviations.

and in-hospital events data have been entered prospectivelyin an institutional database. These data were harvestedretrospectively at the time of follow-up, on August 2008.Follow-up has been performed via a telephone interview andby the request of recent (within 3 months) echocardiographicdata. Quality of life has been quantified by the MinnesotaLiving with Heart Failure.

Statistical analysis has been conducted using the JMP5.1.1 software (SAS institute Inc, Cary, NC, USA). Continuousvariables are presented as mean � SD or as median whenappropriate, and categorical variables are expressed asproportions. Univariable comparisons have been performedusing either the unpaired t-test and ANOVA for parametric orthe Wilcoxon test for nonparametric continuous data andlikelihood ratio for categorical data. Survival has beenreported using Kaplan—Meier method, using the log-rankmethod for comparisons. The level of significance has beenset at p = 0.05. Multivariable analysis has been conductedusing the stepwise regression method. A p < 0.25 level toenter and a p > 0.1 to leave setting was adopted in a mixedforward backward direction. To obtain odds-ratios (OR) for anevent, the nominal logistic regression analysis has beenselected and OR was reported with the confidence intervals(CI). In multivariable analysis, the NYHA class at follow-upwas not added to the model since it is very stronglyassociated to MLHF scores and it would dampen down thevalue of other factors.

3. Results

Follow-up was 98% completed with 5 patients lost atfollow-up. Mean follow-up time was 2.8 � 1.2 years (up to 5years), for a total of 682 patient years. Hospital (30-day)mortality in the overall cohort was 2.7%: 5 deaths were due tolow cardiac output syndrome, in two cases associated tosepsis and in one case associated to stroke; one additionaldeath was due to hemorrhagic shock. Hospital mortality wasnot predicted by preoperative age ( p = 0.7), gender( p = 0.07), body mass index ( p = 0.6), smoking ( p = 0.5),AFib ( p = 0.6), serum creatinine level ( p = 0.9), hypertension( p = 0.5), dyslipidemia ( p = 0.2), respiratory disease( p = 0.4), NYHA class ( p = 0.5), EuroSCORE ( p = 0.2), Charl-son score ( p = 0.4), Lee score (0.8), pulmonary hypertension( p = 0.1), degree of mitral regurgitation ( p = 0.2), coronarydisease ( p = 0.6), tricuspid regurgitation ( p = 0.2), urgentoperation ( p = 0.4). Hospital mortality was predicted bylower preoperative EF ( p = 0.02, r2 = 0.1). There was nostatistical difference in-hospital mortality between patientsundergoing repair or replacement ( p = 0.5). Time in the ICUwas longer in patients older than 80 ( p = 0.02), mean ICU LOSin patients younger than 80 was 3.5 � 5.5 days, versus6.1 � 11.2 days in patients older than 80 years. There was nodifference in length of stay in the ICU between patientstreated by repair or by replacement ( p = 0.7). Median staywas 2 days for patients younger than 80 and 3 days forpatients older than 80 ( p = 0.07). In-hospital complicationsincluded 22 cases (10%) of acute renal failure, with 4 patientsrequiring temporary ultrafiltration, 18 cases (7%) of high doseinotropic support, 17 cases (7%) of sepsis, 15 cases (6%) ofrespiratory failure, with two patients requiring tracheost-

omy, 10 cases (4%) of chest re-entry for excessive bleeding, 7(3%) perioperative acute myocardial infarctions, 5 cases (2%)of minor (type II) neurologic events. Total postoperativelength of stay was 6.2 � 6.0 days (range 0—42 days), with amedian of 5 days. Postoperative length of stay wassignificantly higher in patients who had the valve replacedcompared to the repaired patients (mean length of stay was7.5 � 7.2 days vs 4.9 � 3.3 days, p = 0.0007).

Late survival in the overall population was 92% � 1.7% and91% � 1.9% at 1 and 3 years, respectively. Long-term survivalwas similar in patients younger or older than 80 ( p = 0.16,Fig. 1); there was one reintervention for mitral prosthesisacute endocarditis, while no reinterventions were requiredfor failed repairs. There were 8 cerebrovascular events, 5occurred in patients after valve repair and 2 in patients withtissue valve replacement. Late functional class assessmentrevealed that 41, 130, 32 and 1 patient(s) were in NYHA classI, II, III and IV, respectively. At latest follow-up, overall MLHFscore was 38.2 � 18.0, ranging from 4 to 124. MLHF score washighly related to NYHA at follow-up (Fig. 2), as expected [3].

F. Maisano et al. / European Journal of Cardio-thoracic Surgery 36 (2009) 261—266264

Fig. 3. Relation between MLHF score and cardiac rhythm at follow-up. Thedots represent the individual MLHF scores according to the rhythm group. Thediamonds show the mean and the standard deviations. AFib: atrial fibrillation;PM: pace maker dependent rhythm; Sinus: normal sinus rhythm.

Fig. 4. MLHF at follow-up in patients with failed mitral valve repair (MVrepwith MR), compared to patients with stable hemodynamic results (MVrep noMR) and patients with mitral valve replacement (MVR). The dots represent theindividual MLHF scores according to the type of surgery and the presence ofresidual mitral regurgitation. The diamonds show the mean and the standarddeviations. Differences among the groups are presented on top. On the right isdisplayed the comparison circles plot, a visual representation of group meancomparisons, demonstrating minimal overlapping between the groups, as forstatistically significant differences.

There were 135 patients (66%) who had a score higher than30. All the factors influencing MLHF scores are reported inTable 3. Rhythm at follow-up was also associated to MLHFscore ( p = 0.02), with patients in sinus rhythm having thelowest average score (35 � 16), followed by the patients inAFib (39 � 18) and those with a pacemaker (43 � 19), asshown in Fig. 3. The risk of having a pacemaker inducedrhythm at follow-up was higher in patients who had AFib atadmission ( p = 0.02), or those who received prostheticvalve replacement ( p = 0.02). We did not find a statisticallysignificant difference ( p = 0.36) in average MLHF score inpatients with a repair (37 � 18, CI: 22—40) versus patientswith a replacement (39 � 18, CI: 35—43). However, patientswho had recurrent MR after repair had the highest scorescompared to patients who had repair and durable result andthose treated by replacement (MLHF was 51 � 21, vs34 � 16, vs 39 � 18, respectively, p = 0.0013), as shownin Fig. 4. At multivariable analysis (excluding NYHA atfollow up), MLHF score was influenced by: preoperativeatrial fibrillation, diabetes, higher creatinine level, higherEuroSCORE, mitral regurgitation grade at follow-up echo-cardiography and systolic pulmonary artery pressure atfollow-up. A MLHF score greater than 30 was predicted at

Table 3Univariable and multivariable analysis of preoperative and postoperativefactors associated with MLHF score.

r2 p uni p multi

Age at follow-up 0.03 0.007Preoperative AFib 0.07 <0.0001 0.019Diabetes 0.03 0.009 0.03Preoperative serum creatinine level 0.1 <0.0001 0.0009Preoperative NYHA class 0.05 0.01EuroSCORE 0.07 <0.0001 0.02Charlson score 0.06 <0.0001Preoperative PAPs 0.02 0.09EF at follow-up 0.11 <0.0001Sinus rhythm vs AFib or PM at follow-up 0.02 0.01LVEDD at follow-up 0.07 0.0001LVESD 0.08 0.0001PAPs at follow-up 0.13 <0.0001 0.04MR grade at follow-up 0.07 0.0001 0.002

multivariable nominal logistic fit analysis by age at follow-up (OR: 0.11; CI: 0.02—0.75; p = 0.02), preoperativediabetes (OR: 3.9; CI: 1.14—17.16; p = 0.04), preoperativeNYHA class (OR: 5.9; CI: 1.4—29.7; p = 0.02), EF at follow-up(OR: 18.7; CI: 0.91—405.0; p = 0.05).

4. Discussion

A significant number of elderly patients show high MLHFscores late after mitral valve surgery for mitral regurgitation.Scores are higher in those with older age, but a number ofother clinical and hemodynamic factors influence quality oflife.

Mitral valve surgery in the elderly is often denied for theperceived higher risk and lower benefit expectation fromcorrection of MR. Mirabel et al. from the Euro Heart Survey onvalve disease, reported that in a group of 396 patients withsevere symptomatic MR, surgery was denied in 49%: age andcomorbidities played a major role in the decision not tooperate [4]. At multivariable analysis, older age [OR = 1.40per 10-year increase, 95% CI (1.15—1.72), p = 0.001],increased Charlson comorbidity index [OR = 1.38 per 1 pointincrease, 95% CI (1.12—1.72), p = 0.004]. As a result, decisionto operate was undertaken in 42% of patients with an agebetween 70 and 79 years, and only in 15% of patients olderthan 80 years. However, octogenarians accounted for alimited (10%) proportion of the overall population in thissurvey.

The main reason for denying mitral valve surgery in theelderly is the higher risk of the operation. Operativemortality in patients older than 75 years ranges between 5and 35% [5—7]. Recent data suggest that operative risk hasreduced in the last decade. One component of the decreaseof mortality is the growing percentage of patients beingsubmitted to valve repair as opposed to replacement. In

F. Maisano et al. / European Journal of Cardio-thoracic Surgery 36 (2009) 261—266 265

patients aged 75 years or older, mitral repair is associatedwith a lower risk of mortality, postoperative stroke, andprolonged intensive care unit and hospital stay comparedwith mitral replacement [8]. Detaint et al. [7] reported thatoperative mortality in the elderly has declined considerablyfrom 27% to 5% from the mid 1990s to recent era. The authorsalso stated that repair feasibility in the elderly is similar totheir younger counterparts reporting a 93% chance of repairin patients with degenerative etiology. In the present study,the repair was far lower, about 50%, possibly due to a higherprevalence of non-degenerative MR etiology and of patientsrequiring aortic valve replacement for aortic regurgitation,but also to surgeons or patients preference. Although therewere no differences in-hospital mortality or length of stay inthe ICU, total length of stay was significantly longer aftervalve replacement. However, mean EuroSCORE was higher inthe patients who underwent replacement ( p < 0.0001):there was one point difference, possibly related to the higherprevalence of female gender in the replacement group. Thisdata could be influenced by a selection bias, which was notrecognized at retrospective analysis. There were manypatients with comorbidities in our cohort, which could haveinfluenced surgical decision. Mean comorbidity Charlsonscore [9] was 4.3 � 1.5, with no differences in the repair andreplacement group. In the Euro Heart survey report, whenthe Charlson index was higher than 3, chance of beingsubmitted to surgery was less than 35%. In our population,only 25% of the patients had a Charlson index of 3 or less.However, with a careful and systematic approach to riskmanagement, hospital mortality and morbidity have beenlow and below the predicted mortality.

Although increasing life span is not the main objective insurgery for the elderly, our data confirm that survival aftersurgery is excellent both in the septuagenarians and in thepatients older than 80, with no statistical difference in thetwo groups. It is difficult to demonstrate however, if surgeryprovides a survival benefit compared to medical treatment.Detaint et al. [7] have been able to demonstrate that aftersurgery, life expectancy is restored in elderly patients to asimilar degree of the younger population. Quality of lifeimprovement and or preservation is the most appropriateobjective of cardiac surgery in the elderly. Data have beenaccumulated which supports the role of surgery in elderlypatients who can undergo cardiac surgical procedures at areasonable risk. Elderly patients can benefit from improvedfunctional status and quality of life after cardiac proce-dures [10]. Goldsmith et al. studied the hospital outcomesand quality of life following mitral valve surgery in a seriesof 41 elderly patients (median age 77 years), using medicaloutcomes trust short form 36-item health survey (SF-36).They reported a high hospital mortality both after repair(23%) and replacement (38%), but the hospital survivorsshowed a marked improvement in energy and rolelimitation due to physical and mental health at follow-up, compared to baseline [11]. The role of valve repairversus replacement in the long-term quality of life has beenaddressed by few studies, with contradictory results.Sedrakyan et al. [12] prospectively studied 25 patientswith mitral valve replacement (MVR) and 45 patients withvalve repairs measuring quality of life by SF-36 question-naire at baseline and after 18 months. The study was

underpowered to identify differences between the groups.However, quality of life scores improved after surgery inmost domains, and were comparable to age-adjusted U.S.norms in the valve repair group. In the multivariableanalysis, mitral valve repair recipients had higher socialfunctioning compared with patients who received valvereplacement ( p = 0.04). The authors concluded that valverepair is associated with better quality of life. Morerecently, Jokinen et al. [13] compared the quality of life of184 patients undergoing either mitral repair or replacementfor MR, with or without concomitant coronary artery bypassgrafting, in a retrospective propensity matched analysis.The authors were unable to demonstrate difference inquality of life in the two groups. All patients had lowerenergy and mobility scores than an age- and sex-matchedreference population, using the Nottingham health profilequality-of-life analysis. Age, NYHA class at follow-up,coronary artery disease, other severe disease were allincremental risk factors for suboptimal quality of life aftermitral valve surgery. In this study, the mean age of thepatients was 62 years, and 75% of patients were in sinusrhythm at time of operation. On the other hand,concomitant coronary artery bypass grafting was done inabout 60% of the patients. Goldsmith et al. [14] reported aprospective study on 61 consecutive patients undergoingeither valve repair (40 patients) or replacement (21patients). Quality of life was assessed by SF-36 at baselineand at 3 months after surgery. Mean scores for all patientsfollowing mitral valve surgery showed clinical and statis-tically significant improvement in most parameters.Following repair there was significant improvement inseven of eight quality of life parameters and following MVRthere was significant improvement in three of eight qualityof life parameters. In addition, patients with impaired LVfunction and those with MR with end-systolic dimensions�45 mm did not show improvement in quality of lifecompared to baseline. Immer et al., on the contrary, did notfind any difference between QoL after mitral valvereplacement or repair in a series of 115 patients at 3 yearsafter surgery. However, they reported a higher in-hospitalmortality in replacement than in repair patients (6.4% vs0%; p < 0.01) [15].

In our study, the role of surgical technique did not influencequality of life at follow-up; however, patients with failedmitral valve repair (classified as having evidence of recurrentMR grade 2/4 or more) had the highest scores, compared toeither patients submitted to mitral replacement or patientswith mitral repair and stable hemodynamic result. Othercardiovascular factors involved in the MLHF score include theatrial fibrillation at baseline or rhythm other than sinus atfollow-up (either chronic atrial fibrillation or pace-makerdependent rhythm), baseline and follow-up pulmonary arterypressure, NYHA class, and ejection fraction and left ventri-cular dimensions at follow-up. The highest predictor orsuboptimal quality of life at follow up (MLHF score >30)was EF at follow-up (OR 18.7, p = 0.05). NYHA class at follow-up was directly related to MLHF scores (r2 = 0.4, p < 0.0001).This finding is in line with previous evidence and it isintrinsically related to the scoring model [3]. For this reason,the variable was not added to the multivariable models toenable other variables to show their role.

F. Maisano et al. / European Journal of Cardio-thoracic Surgery 36 (2009) 261—266266

Our findings suggest a role of extra-cardiac factors in thedeterminationof theMLHF scoreat follow-up.BothEuroSCOREand Charlson comorbidity index were significantly associatedtoMLHF at follow-up. Serumcreatinine levelwas the strongestpredictor of MLHF score in the multivariable analysis anddiabetes gave the highest odd ratio (OR 3.9, p = 0.04) fordetecting a MLHF score greater than 30 at follow-up. Inaddition, age at follow-up was a good predictor (r2 = 0.03,p = 0.007) of quality of life. These findings reinforce theconcept that quality of life in the elderly is the result ofmultiple factors, cardiac and non-cardiac, and that attemptsat identifying any difference in quality of life outcomesrelative to surgical techniques are challenging. On the otherhand, earlier interventions [16], as well asminimization of theinvasiveness may play a role in the improvement of long-termresults of mitral valve surgery for organic mitral regurgitationin the elderly. Late quality of life is strongly related toventricular function and to the hemodynamic state, and anearlier intervention, if operative risk is reasonable, mayenable better preservation of left ventricular function, andrestore active life expectancy, also in the elderly population.

4.1. Limitations of the study

Being retrospective, some assumptions of this work maybe affected by a selection bias and by incomplete datareporting and retrieval: as an example, we did not evaluatethe influence of surgical ablation on the outcome of patientswith preoperative AFib because of unavailability of such datain the dedicated database. Additionally, it was impossible toidentify the reasons for the decision between MVR and repairin the individual patients. Quality of life has been assessedsolely by the MLHF score at follow-up. With the study beingretrospective in nature, it was not feasible to retrieve thepreoperative quality of life score. Additionally, MLHF scorewas determined by telephone interview rather than during anoutpatient visit, although this has been utilized in the pastwith reliable results [17].

References

[1] Mirabel M, Iung B, Baron G, Messika-Zeitoun D, Detaint D, VanoverscheldeJ, Butchart EG, Vahanian A. What are the characteristics of patients withsevere, symptomatic, mitral regurgitation who are denied surgery? EurHeart J 2007;28:1358—65.

[2] Rector TS. A conceptual model of quality of life in relation to heartfailure. J Card Fail 2005;11:173—6.

[3] Kubo SH, Schulman S, Starling RC, Jessup M, Wentworth D, Burkhoff D.Development and validation of a patient questionnaire to determine NewYork Heart Association classification. J Card Fail 2004;10:228—35.

[4] Mirabel M, Iung B, Baron G, Messika-Zeitoun D, Detaint D, VanoverscheldeJL, Butchart EG, Ravaud P, Vahanian A. What are the characteristics ofpatients with severe, symptomatic, mitral regurgitation who are deniedsurgery? Eur Heart J 2007;28:1358—65.

[5] Mehta RH, Eagle KA, Coombs LP, Peterson ED, Edwards FH, Pagani FD,Deeb GM, Bolling SF, Pragel RL. Influence of age on outcomes in patientsundergoing mitral valve replacement. Ann Thorac Surg 2002;74:1459—67.

[6] Craver JM, Puskas JD, Weintraub WW, Shen Y, Guyton RA, Gott JP, JonesEL. 601 octogenarians undergoing cardiac surgery: outcome and compar-ison with younger age groups. Ann Thorac Surg 1999;67:1104—10.

[7] Detaint D, Sundt TM, Nkomo VT, Scott CG, Tajik AJ, Schaff HV, Enriquez-Sarano M. Surgical correction of mitral regurgitation in the elderly:outcomes and recent improvements. Circulation 2006;114:265—72.

[8] Ailawadi G, Swenson BR, Girotti ME, Gazoni LM, Peeler BB, Kern JA,Fedoruk LM, Kron IL. Is mitral valve repair superior to replacement inelderly patients? Ann Thorac Surg 2008;86:77—85.

[9] Charlson ME, Pompei P, Ales KL, MacKenzie CR. A new method of classify-ing prognostic comorbidity in longitudinal studies: development andvalidation. J Chronic Dis 1987;40:373—83.

[10] Fruitman DS, MacDougall CE, Ross DB. Cardiac surgery in octogenarians:can elderly patients benefit? Quality of life after cardiac surgery. AnnThorac Surg 1999;68:2129—35.

[11] Goldsmith I, Lip GY, Kaukuntla H, Patel RL. Hospital morbidity andmortality and changes in quality of life following mitral valve surgeryin the elderly. J Heart Valve Dis 1999;8:702—7.

[12] Sedrakyan A, Vaccarino V, Elefteriades JA, Mattera JA, Lin Z, RoumanisSA, Krumholz HM. Health related quality of life after mitral valve repairsand replacements. Qual Life Res 2006;15:1153—60.

[13] Jokinen JJ, Hippelainen MJ, Pitkanen OA, Hartikainen JE. Mitral valvereplacement versus repair: propensity-adjusted survival and quality-of-life analysis. Ann Thorac Surg 2007;84(August (2)):451—8.

[14] Goldsmith RA, Lip GYH, Patel RL. A prospective study of changes in thequality of life of patients following mitral valve repair and replacement.Eur J Cardiothorac Surg 2001;20:949—55.

[15] Immer FF, Donati O, Wyss T, Immer-Bansi AS, Schmidli J, Berdat PA, CarrelTP. Quality of life after mitral valve surgery: differences between recon-struction and replacement. J Heart Valve Dis 2003;12:162—8.

[16] St. John Sutton MG, Gorman RC. Surgery for asymptomatic severe mitralregurgitation in the elderly. Early surgery or wait and watch? Circulation2006;114:258—60.

[17] Bennet SJ, Oldridge NB, Eckert GJ, Embree JL, Browning S, Hou N, Chui M,Deer M, Murray MD. Comparison of quality of life measures in heartfailure. Nurs Res 2003;52:207—16.

Appendix A. Conference discussion

Dr P. Perier (Bad Neustadt/Saale, Germany): You state in your conclusionssomething which is somewhat disappointing, which is that in half of thepatients the quality of life after surgery is suboptimal. Nevertheless, thesurvival is not affected. It is disappointing because those people most ofthe time are requiring a lot of care, attention, concentration, dedication,and it may also comfort the strategy of our colleagues, cardiologists, whodeny surgery so often in patients with severe mitral valve regurgitation. Myquestion is wouldn’t it be more interesting to compare the quality of lifebefore surgery and after surgery to really grasp and understand the impact ofsurgery?

Second, if we want to improve things and if we try to think strategically, wemight imagine that the quality of life after surgery is depending, first, as youhave said, on the clinical status of the patient preoperatively, and factors likeneurocognition, pulmonary artery pressure, atrial fibrillation are important,but it may also be dependent on the trauma of surgery, trauma which veryoften we underestimate. So my second question would be, since timing ofoperation is so important for the resultant quality of life, what should be thecriteria which would help to time surgery not too early but also not too late tohave themost optimal result? And second, if we want to reduce the trauma, wehave one means, which is minimally invasive surgery. What do you think shouldbe the place of minimally invasive surgery for this category of patient?

Dr Vigano: Certainly one limit of our study was that it was carried out as aretrospective study and we couldn’t assess preoperatively patients’ quality oflife with the Minnesota score. So in the future it would be advisable to conducta prospective study in which it will be possible to compare patients’ quality oflife before and after surgery.

Moreover, timing of surgery in this subgroup of patients, seems to be veryimportant especially for the severity of inherent comorbidities. Consideringthe results that we observed, that almost half of the patients have asuboptimal quality of life after mitral valve surgery, perhaps we can say thatsurgery was performed too late and these patients could profit by earliersurgery in terms of less in-hospital complications and late results.

Actually minimally invasive surgery is performed in the young population,but perhaps in the future it would have a role even in the elderly, because itcould be able to reduce surgical trauma, hospital length of stay and in-hospitalcomplications, like respiratory failure, improving the clinical outcome. So it’sreasonable that minimally invasive surgery could have a relevance in thiscategory of patients.