Eur J Cardiothorac Surg 2007;32:596-603. doi:10.1016/j.ejcts.2007.06.044
Copyright © 2007, European Association for Cardio-Thoracic Surgery. Published by Elsevier B.V. All rights reserved
Mitral annulus calcification: determinants of repair feasibility, early and late surgical outcome
Cosimo d’Alessandro,
Nicola Vistarini,
Stéphane Aubert,
Frédérique Jault,
Christophe Acar*,
Alain Pavie,
Iradj Gandjbakhch
Département de Chirurgie Cardiovasculaire, Institut de Cardiologie, Hôpital Pitié Salpétrière, 50-52 Bd Vincent Auriol, 75013 Paris, France
Received 1 March 2007;
received in revised form 13 June 2007;
accepted 15 June 2007.
* Corresponding author. Tel.: +33 142 16 56 85; fax: +33 142 16 56 78. (Email: c.acar{at}psl.aphp.fr).
 |
Abstract
|
|---|
Objective: The aim of this study was to determine the factors influencing the feasibility of valve repair and the surgical outcome in patients with mitral annulus calcification. Methods: In 124 patients with mitral annulus calcification undergoing surgery, two entities were distinguished: Barlow disease (myxomatous leaflets, n
= 60) and fibroelastic deficiency (FED) (normal leaflets, n
= 64). The calcification score was lower (1.9 vs 2.8); the annulus was more dilated (ring 35 vs 32 mm) and ruptured chordae were more frequent (77% vs 37%) in Barlow than in FED (p
< 0.001). The clinical profile was different: age (60 ± 14 vs 73 ± 8 years, p
< 0.001), systemic hypertension (22% vs 70%, p
< 0.001), chronic renal insufficiency (5% vs 22%, p
< 0.01), cancer (7% vs 25%, p
< 0.01). Multifocal atherosclerosis was less frequent in Barlow than in FED: carotid disease (17% vs 54%, p
< 0.001), aortic atheroma (21% vs 51%, p
< 0.001) and coronary disease (22% vs 56%, p
< 0.01). Echocardiography showed two different patterns in Barlow and FED: aortic valve stenosis (1.7% vs 31%), left atrial diameter (54 vs 49 mm), left ventricular end-diastolic diameter (62 vs 54 mm), interventricular septal thickness (11 vs 13 mm), and systolic pulmonary pressure (40 vs 56 mmHg), respectively (p
< 0.001). Bacterial endocarditis was observed in 24 cases (19%). Results: The surgical technique was a valve repair in 68% and a replacement in 32%. The repair rate depended upon the extent of annulus calcifications (p
< 0.001) and the type of degenerative disease (95% vs 44% in Barlow and FED p
< 0.001). In-hospital mortality was 14% (Barlow: 5% vs FED: 23%, p
< 0.01). The mean follow-up was 50 ± 41 months. Overall 5-year year survival was 76% (Barlow: 90% vs FED: 64%, p
< 0.001) and survival free from cardiac event was 69% at 5 years (Barlow: 87% vs FED: 52%, p
< 0.001). Five-year survival was higher following repair than replacement (84% vs 64% p
< 0.001). Chronic renal insufficiency and bacterial endocarditis were two predictors of early and late death (p
< 0.01). Conclusions: The aetiopathogeny of the degenerative mitral disease responsible for annulus calcifications corresponded to distinct anatomical, clinical and echographic patterns. It was a main determinant of repair feasibility, early and late surgical outcome.
Key Words: Mitral annulus calcification • Mitral valve repair • Mitral valve replacement
|
1. Introduction
|
|---|
Mitral valve repair has emerged as the ideal treatment for degenerative mitral insufficiency and the possibilities of valve repair have reached 95% for this aetiology in experienced teams. Undoubtedly, calcifications of the mitral annulus represent the main anatomical obstacle to valve repair in this indication. The decalcification with sliding plasty technique described by Carpentier et al. [1] has considerably increased the possibilities of surgical reconstruction, but valve replacement still remains necessary in some cases. The aim of this study was to identify the anatomical, clinical and echographic factors influencing the feasibility of the repair and the determinants of the early and late surgical outcome in case of mitral annulus calcification.
|
2. Methods
|
|---|
2.1 Patient population
One hundred twenty-four patients with mitral annulus calcification underwent surgery between 1995 and 2005. Patients’ age was 66 ± 13 years (23–88 years). Fifty-five patients (44%) were in functional class NYHA I or II and 69 patients (56%) were in functional class NYHA III or IV. The operation was performed in emergency in 11 cases (9%). Five patients had undergone a previous valve surgery (aortic valve replacement n
= 4, failed attempt at mitral valve repair n
= 1). ECG showed a sinus rhythm (n
= 67, 54%), sinus rhythm with paroxysmal atrial fibrillation (n
= 32, 26%) or permanent atrial fibrillation (n
= 25, 20%). Eight patients (6%) were paced with a permanent stimulator. Marfan syndrome was present in four cases. Chronic renal insufficiency was noted in 17 cases (14%): creatinine level >200 µg/l (n
= 9) or haemodialysis (n
= 8). Systemic hypertension was observed in 58 cases (49%). Twenty patients (19%) had been treated for a cancer: digestive (n
= 6), gynaecologic (n
= 4), urologic (n
= 4), pulmonary (n
= 2) carcinoma or malignant hemopathy (n
= 4). Other comorbidities included: cerebral stroke (n
= 14, 11%), chronic obstructive pulmonary disease or severe asthma (n
= 19, 15%), diabetes mellitus (n
= 16, 13%), arteritis (n
= 6), liver cirrhosis (n
= 5), autoimmune systemic disease (n
= 3), meningioma (n
= 2). Bacterial endocarditis complicated mitral annulus calcification in 24 cases (19%).
Preoperative vascular echo Doppler study of the supraaortic vessels performed in 93 cases (75%) showed a carotid disease in 33 cases (35%): calcified plaque with stenosis <30% (n
= 21), stenosis 30–50% (n
= 10) and stenosis >50% (n
= 2). Echographic examination of the thoracoabdominal aorta in 79 cases (64%) showed an aortic wall atheroma in 28 cases (35%): protruding calcified plaque (n
= 26) or aneurysm (n
= 2). A preoperative angiogram in 108 cases (87%) showed a coronary artery disease in 43 cases (40%): previous stenting (n
= 5), <50% stenosis (n
= 18) or >50% stenosis (n
= 20).
Echocardiography showed a mitral regurgitation grade 1/2 in 12 patients (10%) and 3/4 in 112 patients (90%). In 11 cases (9%) an associated mitral stenosis was also noted (mean transvalvular gradient: 14 ± 6 mmHg, mean surface valve area: 0.91 + 0.26 cm2). Mitral annulus calcification was detected on the preoperative echography in 85 cases (69%). It was overlooked and revealed during surgery in 39 cases (31%). Echography showed an aortic valve disease in 43 cases (35%): isolated cusp calcifications (n
= 22, 18%), or with stenosis (n
= 21, 17%) (mean transvalvular gradient: 35 ± 13 mmHg and surface valve area: 0.82 ± 0.26 cm2) including four previously inserted prosthetic valve where three were obstructive. Two patients had a hypertrophic subaortic stenosis. An aortic valve insufficiency was observed in 47 cases (38%): grade 1 (n
= 27), grade 2 (n
= 19) and grade 3 (n
= 1). The mean ascending aorta diameter was 33 ± 5 mm. The left ventricle dimensions were: end-diastolic diameter (58 ± 8 mm), end-systolic diameter (35 ± 8 mm), ejection fraction (63 ± 9%), and diastolic septal thickness (12 ± 3 mm). The left atrial diameter in four-chamber view was 51 ± 8 mm. A grade 2 or more tricuspid insufficiency was noted in six cases. The systolic pulmonary pressure using Doppler was evaluated in 108 patients (87%) and its average value was 48 ± 16 mmHg.
According to Carpentier, an aetiopathogenic classification of the type of degenerative mitral disease was established based on the anatomy of the valve and two groups were distinguished [1–3]. The Barlow disease was characterised by the presence of a severe myxomatous degeneration of the leaflets and chordae with excess tissue. Conversely, the diagnosis of fibroelastic deficiency was carried when the leaflets appeared normal or thinner [1,2]. In the rare intermediary forms in which the sole prolapsed area presented some degree of myxomatous involvement, the size of the mitral annulus was also taken into consideration [3]. The localisation as well as the extent of the annulus calcification graded from 1 (limited to one segment of the mitral annulus) to 5 (involving the three posterior segments, at least 1 commissural in addition to the anterior part) was noted. Ruptured chordae were observed in 70 cases (56%) involving in five cases two separate parts of the valve: posterior leaflet (n
= 65), anterior leaflet (n
= 6), commissure (n
= 4). A large cavity containing caseous necrosis was found in three cases.
2.2 Surgical technique
Mitral valve repair was performed in 85 cases (68%). In case of a calcification localised to one segment, it was either removed together with the prolapsed area or left in place when it involved a commissure (Fig. 1
). When the calcium bar was extended to two segments or more of the mitral annulus, the Carpentier decalcification technique [1] was used (n
= 34). It comprised a disinsertion of the posterior leaflet with en bloc removal of the calcium bar and posterior leaflet sliding plasty (Fig. 2
). The other techniques of repair were: posterior leaflet quadrangular resection (n
= 78), annulus plication (n
= 54), chordae transposition on the anterior leaflet (n
= 13), commissural closure (n
= 9), pericardial patch (n
= 3) or direct (n
= 1) closure of a perforation. All caseous or abscessed collections were evacuated and the cavity was deterred and closed directly except in one case in which a pericardial patch was used. Prosthetic ring annuloplasty was accomplished in all cases with either a Carpentier physio (n
= 24) or a Duran ring (n
= 61). The mean intercommissural distance of the ring was 34.2 ± 2.6 mm. Intraoperative echocardiography control of the repair was routinely performed.
View larger version (34K):
[in this window]
[in a new window]
|
Fig. 1. (A) Localised calcification of a commissural area away from the site of prolapse (P2). (B) The calcification process involves the underlying chordae and the papillary muscle. (C) The repair is performed leaving alone the calcific nodule.
|
|
View larger version (37K):
[in this window]
[in a new window]
|
Fig. 2. Carpentier repair technique of annulus decalcification. Part 1 (sketches): (A) Extensive calcification of the annulus. (B) Disinsertion of the posterior leaflet, note that one commissure involved by the calcifying process as well as the clefts separating P1 from P2 and P2 from P3 have been closed. Removal en bloc of the calcium bar. (C) The posterior annuloplasty sutures are placed and the posterior leaflet is reattached. (D) The sliding plasty is completed. (E) Prosthetic ring annuloplasty. Part 2 (operative views): (A) Barlow disease with extensive annulus calcification (arrow). (B) Quadrangular resection of P2 together with the calcium bar. (C) Leaflet reconstruction using the sliding plasty. (D) Prosthetic ring annuloplasty.
|
|
Mitral valve replacement was achieved in 39 cases (32%) with a bioprosthesis (n
= 34) or a mechanical valve (n
= 5). The technical details are described on Fig. 3
. The mean diameter of the prostheses was 28.9 ± 2.4 mm. The associated procedures included: aortic valve replacement in 23 cases (18%) (Bioprosthesis n
= 20, mechanical n
= 3), septal myectomy (n
= 2), tricuspid valve annuloplasty (n
= 3), pulmonary veins isolation (n
= 2) and coronary artery grafting in 18 cases (14%). The average bypass time was 83 ± 29 min and the average ischemic time was 73 ± 25 min.
View larger version (27K):
[in this window]
[in a new window]
|
Fig. 3. Technique of valve replacement in mitral annulus calcification. (A) The fixation sutures are passed around the annulus calcification around the whole posterior leaflet kept intact with its chordae through the adjacent left atrial wall. (B) The left atrial wall has been plicated in between the calcification and the prosthesis-sawing ring. The calcified aortic valve is then replaced. (C) In exuberant calcification, the sawing ring is extended using a circumferential Teflon felt patch.
|
|
2.3 Follow-up
All patients underwent an echocardiographic control at the end of the hospitalisation stay. In addition, information concerning the clinical status was obtained by phone directly from the patient or from his general practitioner or cardiologist at a follow-up of 50 ± 41 months (from 3 to 125 months).
2.4 Statistical analysis
Descriptive data for continuous variables were presented as means with standard deviation. Patient groups were compared by chi-square test for categorical values and Student's t-test for continuous variables and for comparison of groups including small samples. In-hospital morbidity included all complications occurring during the hospitalisation stay, beyond that time only cardiac related events were recorded. Survival rates were estimated according to the Kaplan–Meier method and were compared using the log-rank test.
|
3. Results
|
|---|
3.1 Aetiopathogenic classification
Mitral annulus calcifications were found to be associated in 60 (48%) patients with Barlow disease and in 64 (52%) others with a fibroelastic deficiency. Preoperative patients’ characteristics of the two groups are shown in Table 1
. In fibroelastic deficiency, patients were older and had more severe heart failure symptoms, the calcifications were more extended, dilatation of the annulus was less pronounced as shown by the smaller rings used for repair and ruptured chordae were less frequent than in Barlow disease. As shown in Fig. 4
describing the localisation of the calcifications, the P2 and P3 as well as the anterior annular segments were involved more frequently in fibroelastic deficiency than in Barlow disease. The clinical profile (Table 1) was more severe of fibroelastic deficiency in which systemic hypertension, chronic renal insufficiency, cancer or haemopathy and diabetes mellitus were more frequent than in Barlow disease. In addition, as shown by the vascular echo Doppler and angiographic studies (Table 1), multifocal atheroma was markedly more frequent in fibroelastic deficiency than in Barlow disease, with a higher incidence of carotid artery disease, aortic wall atheroma and coronary artery disease. Similarly, two different patterns of echocardiographic data were observed according to the type of degenerative disease (Table 1, Fig. 5
). In fibroelastic deficiency, the amount of mitral regurgitation was less, the frequency of an associated degenerative mitral stenosis was higher, the ability of echography to detect annulus calcification was better and the incidence of aortic valve calcifications without or with stenosis was higher than in Barlow disease. In fibroelastic deficiency, the diameter of the ascending aorta was smaller, the left atrium was less dilated as well as the left ventricle, the interventricular septum was thicker, the ejection fraction was slightly lower although within the normal range and the systolic pulmonary pressure was more markedly elevated than in Barlow disease (Table 1, Fig. 5).
View this table:
[in this window]
[in a new window]
|
Table 1 Preoperative characteristics according to the type of degenerative mitral valve disease responsible for annulus calcifications
|
|
View larger version (14K):
[in this window]
[in a new window]
|
Fig. 4. Localisation of the annulus calcifications in Barlow versus fibroelastic deficiency (FED). C: Commissural annulus segment, P1, P2 and P3: annulus segment at the base of the corresponding scallop of the posterior leaflet, A: anterior annular segment.
|
|
View larger version (27K):
[in this window]
[in a new window]
|
Fig. 5. Typical patterns of the two types of degenerative valve disease associated with mitral annulus calcifications: Barlow (left) and fibroelastic deficiency (FED) (right).
|
|
3.2 Repair feasibility
The feasibility rate of valve repair correlated with the type of degenerative disease, higher in Barlow than in fibroelastic deficiency (n
= 57 (95%) vs n
= 28 (44%), respectively) (<0.001). The extent of the annular calcium process also influenced the repair feasibility and the calcification severity score was lower in the repair than in the replacement group (1.8 ± 1.0 vs 3.6 ± 1.2, respectively) (p
< 0.001). Aortic valve replacement was performed more often in fibroelastic deficiency (n
= 20, 31%) than in Barlow disease (n
= 3, 5%) (p
< 0.001) the same was true for coronary bypass (n
= 13, 20% vs n
= 4, 7%, respectively (p
< 0.05). As a result, bypass time was longer in fibroelastic deficiency than in Barlow disease (88 ± 33 vs 77 ± 23 min, respectively) (p
< 0.05). Overall, eight patients (6%) required a second pump run. Of six patients presenting with a residual mitral dysfunction, the valve was preserved in four cases: a smaller prosthetic ring was inserted in case of a central lack of coaptation (n
= 3) and a larger ring was placed in case of a SAM (n
= 1). Mitral valve replacement was required in two cases in which the mechanism of the residual insufficiency remained unclear. Two other patients in whom an extensive decalcification has been performed presented with ECG signs of myocardial ischemia and a hypokinetic lateral wall, a circumflex artery plication was suspected and a coronary bypass was constructed whose postoperative course was simple. There was no haemorrhagic complication due to annulus injury following either repair or replacement in this series.
3.3 In-hospital mortality and morbidity
Overall, in-hospital mortality included 18 cases (14%) whose causes are listed in Table 2
. The type of mitral degenerative disease was a predictor of in-hospital mortality (5% vs 23% for Barlow and fibroelastic deficiency respectively, p
< 0.01). Increased in-hospital mortality was also observed in chronic renal insufficiency (35%) (p
< 0.01) and bacterial endocarditis (29%) (p
< 0.05). The repair group had a lower in-hospital mortality than the replacement group (8.2% vs 28%, respectively) (p
< 0.01). Perioperative complications occurring in 21 patients (17%) are listed in Table 2 (some patients presented more than one complication), haemorrhagic complication (n
= 3) included: mediastinal bleeding, rectus major haematoma and gastric ulcer haemorrhage. Echocardiography at the end of the hospital stay showed in mitral valve repair: no or grade 1 residual insufficiency (n
= 71) or grade 2 insufficiency (n
= 7). No periprosthetic leak was noted following valve replacement. The mean transvalvular gradient was lower after repair (3.6 ± 1.1 mmHg) than after replacement (5.1 ± 1.9 mmHg) (p
< 0.001).
3.4 Follow-up
Overall predicted survival at 5 years was 76% and was higher in Barlow (90%) than in fibroelastic deficiency (63%) (p
< 0.001) (Fig. 6A). Five-year survival was also higher for valve repair as compared to replacement (84% vs 64%, respectively, p
< 0.001) (Fig. 6B). A lower survival rate at 50 months was observed in patients with chronic renal insufficiency/dialysis (29%, p
< 0.01) and those with bacterial endocarditis (46%, p
< 0.05).
View larger version (21K):
[in this window]
[in a new window]
|
Fig. 6. Survival in mitral annulus calcification: for each category the upper curve shows overall survival and the lower curve shows survival free from cardiac event. (A) Fibroelastic deficiency (FED) versus Barlow disease. (B) Mitral valve replacement (MVR) versus repair.
|
|
Of the 106 survivors, 52 patients (49%) were under chronic anticoagulation therapy and 45 (42%) were receiving low dose aspirin. Throughout the follow-up period haemorrhagic complication was noted in five cases (digestive n
= 2, urologic n
= 2, retroperitoneal haematoma n
= 1); if one includes the postoperative bleedings and the early onset of tamponade, the total haemorrhagic complication rate concerned 11 patients (9%) where 2 were fatal. A cerebral stroke occurred in three cases. There were four reoperations due to: early repair failure from 2 to 11 months postoperatively (n
= 3) or infective endocarditis on a prosthesis with obstructive vegetations without dehiscence at 22 months (n
= 1). One other patient suffered recurrence of bacterial endocarditis on a valve repair at 19 months treated medically. Survival free from cardiac event was 69% at 5 years and was higher in Barlow than in fibroelastic deficiency (87% vs 52%, respectively, p
< 0.001) (Fig. 6A). Similarly, survival free from major cardiac event at 5 years was higher after repair than after replacement (75% vs 54% respectively, p
< 0.001) (Fig. 6B).
|
4. Discussion
|
|---|
4.1 Anatomical findings
Two types of degenerative diseases responsible for annulus calcification were distinguished according to Carpentier based on the intraoperative macroscopic analysis of the mitral valve [1,2]. The Barlow disease was characterised by the presence of a myxomatous involvement of the leaflets and chordae with excess tissue. Conversely, the diagnosis of fibroelastic deficiency was carried when the leaflets had a normal structure or appeared thinner (Fig. 5). For the unusual intermediary cases in which the myxomatous degeneration was strictly localised to the prolapsed area, where the remnant mitral valve appeared normal, it has been demonstrated that only quantitative histology was able to distinguish the two nosological entities [3]. Such an analysis was not possible in this study. According to the recommendations of this author for litigious cases, the classification took into consideration the dimensions of the mitral annulus (size 34 and above: Barlow disease, size 32 and below: fibroelastic deficiency) [3]. The anatomical lesions observed in patients with Marfan syndrome were typical of a remarkably severe Barlow disease. In contrast, the leaflets were usually free from any myxomatous lesion in patients with chronic renal insufficiency and these patients were most frequently categorised as fibroelastic deficient.
Chordal ruptures were observed in a vast majority of Barlow disease cases involving preferentially the P2 portion of the posterior leaflet, in the remaining patients the regurgitation was usually linked to a bileaflet prolapse due to chordal elongation with severe annulus dilatation. Chordal rupture was observed only in one third of the fibroelastic deficiency cases. The calcium bar was commonly located posteriorly and invaded a commissural area in 40% of the cases. It was more extended in fibroelastic deficiency in which the P2 and P3 annular segments were more frequently involved than in Barlow disease (Table 1, Fig. 4). The anterior annular part was usually spared in Barlow disease, whereas it was occasionally calcified particularly in the trigone areas in fibroelastic deficiency (Fig. 4).
4.2 Clinical and echographic aspects
Although Barlow disease can occur in young adults and even in the childhood, the annulus calcification is a slowly evolving process and the average age of Barlow cases in this series was 60. By contrast, fibroelastic deficiency preferentially develops in the elderly and patients’ age in this group was 73. Not surprisingly, comorbidities such as cancer or diabetes mellitus were more frequent in this group. An accelerated calcification of the mitral annulus in adults in their third or fourth decade was observed in patients with a Marfan syndrome and in those under chronic dialysis.
Nowadays many patients with severe mitral insufficiency are operated upon while still asymptomatic based on echocardiographic criteria [4]. Thus, the Barlow disease group in this series included a substantial number of asymptomatic patients (Table 1). However in the elderly, signs of poor clinical tolerance are required before referring the patient to surgery and many patients suffering fibroelastic deficiency were in NYHA class III or IV. Overall, preoperative permanent or paroxystic atrial fibrillation was observed in almost half of the cases. The ability of preoperative echocardiography to detect annulus calcification depended upon the extent of the calcium bar; echo diagnosis was particularly efficient in fibroelastic deficiency (80%). When compared with the operative findings, its sensitivity was only 57% in Barlow disease probably because a localised calcified nodule was hardly more distinguishable at the base of a thickened myxomatous leaflet.
The two anatomical entities of mitral degenerative disease appeared to be associated with two distinct echographic patterns whose characteristics are summarised in Table 1 and Fig. 5. In case of Barlow disease, the myocardial changes were solely due to the chronically regurgitant mitral valve which was invariably severe (grade 3/4). The left atrial cavity was markedly enlarged allowing an easy access to the mitral orifice. The left ventricle whose thickness was normal was almost invariably dilated. Thanks to these compensation mechanisms, the pulmonary pressures were only mildly elevated at the time of surgery. Even in the absence of Marfan syndrome, a dystrophy of the aortic root was occasionally observed as suggested by the slight increase in aortic diameter whereas the aortic valve was rarely calcified.
The echographic presentation of fibroelastic deficiency was strikingly different because it very frequently occurred in the setting of a hypertensive cardiomyopathy (Fig. 5). The amount of regurgitation was frequently less but a genuine degenerative stenosis was occasionally observed. The left atrium had a normal or slightly increased size sometimes making the surgical exposure difficult. The left ventricular wall was hypertrophied whereas its cavity was only mildly dilated. The aortic valve was calcified in half of the cases (Mönckeberg disease) and required an aortic valve replacement in one third of the cases. Mitral annulus calcification being a well-known marker of atheroma, [5,6] not surprisingly the elderly population with fibroelastic deficiency had frequently multifocal disease with a carotid disease or an aortic wall atheroma. Similarly, an associated ischaemic heart disease was noted in over half of the cases (Table 1). Due to the lack of significant cardiac chamber dilatation, the pulmonary pressures were notably increased resulting in a disabling dyspnoea as mentioned above. In between these two typical patterns, intermediary cases were observed: Barlow disease with systemic hypertension in an elderly patient or early phase of a fibroelastic deficiency with a calcific nodule at the base of a single myxomatous leaflet segment. Whatever the type of degenerative disease, the ejection fraction of the left ventricle most frequently remained within the normal range although slightly lower in the fibroelastic deficiency.
4.3 Techniques of mitral repair
The possibilities of repair were evaluated directly during the procedure based on two criteria: (1) the extent of the annular calcification (2) the amount of leaflet tissue available for reconstruction. Hence, a mitral valve repair was achieved in the vast majority of patients with a Barlow disease (95%) even in the case of extensive annular calcifications. In fibroelastic deficiency, the lack of excess tissue limited the possibilities of mitral reconstruction and valve repair was restricted to the cases with more localised calcifications (44%).
As opposed to previous publications focused exclusively on the surgical challenge raised by extensive annular calcifications [7–9], this series also deals with the less invasive forms. When localised, the calcification frequently involved the annulus at the insertion of the prolapsed area. The repair comprised, in addition to the resection of the myxomatous posterior leaflet segment, a removal of the calcific nodule so as to allow an annulus plication with leaflet suture or a reconstruction with a sliding plasty. Nevertheless, in fibroelastic deficiency, when the calcified nodule was located away from the site of prolapse, close to a commissure and invaded the underlying chordae or even the papillary muscle, it was left in place (Fig. 1). Indeed, the cure of such a lesion would have required in addition to the prolapsed area, a resection of the commissure [1] whose reconstruction could have been problematic in the absence of excess tissue.
Whenever the calcifications were extended to two segments or more of the mitral annulus, the repair technique included a disinsertion of the posterior leaflet where the most myxomatous portion was excised followed by an en bloc resection of the calcium bar according to Carpentier [1] (Fig. 2). In case of a large leaflet removal, a plication of the P2 portion of the annulus was achieved. The posterior sutures used for ring annuloplasty were placed at that stage and maintained under traction so as to produce some degree of annular contraction in order to facilitate the reapproximation of the leaflet remnants (Fig. 2). Then the posterior leaflet was reattached using large bites on the atrial wall in such way that the prosthetic ring, once in place, would cover the whole suture line. In our experience, it matters to preserve at least a 1 cm height of tissue along the three posterior leaflet segments so as to provide a large coaptation area. Due to the infiltration by the annular calcifications, the amount of healthy tissue is often insufficient at the level of the clefts separating P1 from P2 and P2 from P3 (Fig. 2). Thus, the clefts were systematically closed before disinserting the posterior leaflet. For the same reason, whenever the calcification was extended to a commissural area, the corresponding commissure was closed by means of a few separated sutures before removing the calcium bar. In our experience, these technical details seemed to significantly decrease the risk of residual insufficiency.
Although the atrioventricular continuity was completely disconnected, no serious intraoperative haemorrhage occurred in this series. The proximity of the circumflex artery, particularly in its proximal segment corresponding to the anterolateral commissure and to the P1 segment of the mitral annulus represents another potential source of complication. In two cases with a left dominant circumflex artery, the detection of ECG changes in the lateral wall and a segmental dysfunction on the intraoperative echography led to the construction of coronary bypasses where postoperative courses were uneventful.
In case of an extensive annulus involvement, some authors have advocated the creation of a double mitral orifice leaving the calcifications in place [10]. Our limited experience with this technique resulted in residual insufficiency and a complete decalcification with sliding plasty or a valve replacement appeared to us a more reliable option.
A large collection of caseous necrosis developed in contact with the calcification or was observed separately in a few cases in this series [11]. These formations mimicking pus were constantly sterile and their surgical treatment was easy because a fibrotic sheath that could be sutured directly once the collection had been evacuated surrounded them.
The systematical use of intraoperative transoesophageal has avoided the need for an early operation in six cases in this series. A second pump run was instituted during which the mitral valve was most frequently preserved. The prosthetic ring was replaced by a larger ring in case of SAM or by a smaller ring in case of central lack of coaptation. In two cases in which the mechanism of repair failure remained unclear, mitral valve replacement was performed. Unfortunately, intraoperative echography has not completely abolished the onset of early repair failure (three cases in these series), probably because it had not always been realised under physiological haemodynamic conditions [12]. Filling with a large volume and if necessary intravenous vasoconstrictors should be used in order to maintain a normal afterload (systolic systemic pressure >120 mmHg) so as to reveal any residual insufficiency [12].
4.4 Valve replacement on a calcified mitral annulus
In fibroelastic deficiency with circumferential calcifications, a genuine degenerative mitral stenosis was occasionally observed. The combination of protrusive posterior calcifications, a calcific invasion of the commissures and of the base of the anterior leaflet interfering with the valve opening, together with a narrow mitral orifice, was responsible for a high diastolic gradient. The associated mitral insufficiency was usually mild to moderate. As a rule, a valve replacement was then performed (Fig. 3).
When valve replacement was considered, a complete excision of the anterior leaflet and of the commissures was realised whereas the whole posterior leaflet with its chordae and the adjacent annular calcifications were preserved. However a partial decalcification was accomplished in some rare cases: marked protrusion of the calcifications into the ventricular cavity or failed attempt at valve repair. Considering the age and the comorbidities in patients requiring valve replacement, a limited life expectancy could be expected and the usual choice as a valve substitute was a bioprosthesis.
In this series the total haemorrhagic complication rate was 9%. Although an anticoagulation therapy was necessary due to atrial fibrillation in half of the cases, the level of anticoagulation required (INR: 2) was lower than for a mechanical mitral prosthesis (INR: 3–4). The treatment could even be discontinued in case of necessity (extracardiac surgery), which in addition, facilitated the management of potential bleeding.
Surgery carries a high risk in patients requiring chronic haemodialysis [13]. Mitral annulus calcification is a sign of diffuse nephrocalcinosis that is associated with a poor prognosis [14]. According to a recent series, the lesser longevity of bioprostheses in these patients has been disproved [15]. Pericardial effusion is a common complication in chronic renal insufficiency and can be further aggravated by the anticoagulation required during the dialysis periods leading to tamponade. The adjustment of the anticoagulation level was facilitated by the choice of a biological valve substitute that allowed to lower or to discontinue the anticoagulation therapy in between the dialyses. The early and late mortality of patients with severe renal insufficiency (creatinine level >200 µM/l or chronic dialysis) was high in our series, mostly from extracardiac causes and only 29% of these patients were still alive at 5 years.
Concerning the surgical technique, the size of the prosthesis was carefully chosen so as to fit the mitral orifice exactly. In our experience, the insertion of an undersized prosthesis on a non-pliable annulus has been a frequent source of periprosthetic leak. Some authors have proposed to reconstruct the annulus with pericardium once decalcified and to insert the prosthesis on the pericardial patch [8]. Others have chosen to leave the calcium bar and use an intra-atrial implantation of the prosthesis extended with a piece of Dacron [16].
The fixation of a valve in the anatomical position firmly anchored to the calcifications left intact seemed to offer a more reliable solution. No periprosthetic leak was detected during the early and late follow up in this series. This option did not increase the risk of annular rupture and there has not been any intraoperative bleeding due to annulus injury.
In order to improve the compliance of the insertion site, the left atrial tissue adjacent to the mitral annulus with the posterior leaflet was applied together with the annular calcifications on the sawing ring of the valve (Fig. 3). Simple, interrupted and numerous braided 2/0 sutures with large uppercut needles were used for fixation. Occasionally the sawing ring was extended by means of a circumferential patch of Teflon felt so as to fit exactly the shape of the calcifications (Fig. 3).
In conclusion, the aetiopathogeny of the mitral disease with annulus calcifications corresponded to distinct anatomical, clinical and echographic patterns. It was a main determinant of the repair feasibility as well as the surgical outcome. A repair was performed in almost all cases of Barlow disease patients and the long-term results have remained excellent (Fig. 6A). In fibroelastic deficiency, a repair has been possible in less than half of the cases. Due to the patients’ older age and to the frequent extracardiac associated diseases, in-hospital mortality was higher than in the Barlow group. Similarly overall survival and cardiac event free survival at 5 years was lower (Fig. 6A). Chronic renal insufficiency and bacterial endocarditis were two strong predictors of early and late death when associated with mitral annulus calcifications requiring surgery. The surgical technique itself seemed to influence the prognosis and early or late survivals were higher following mitral valve repair than following replacement (Fig. 6B).
|
References
|
|---|
- Carpentier AF, Pellerin M, Fuzellier JF, Relland JY. Extensive calcification of the mitral valve annulus: pathology and surgical management. J Thorac Cardiovasc Surg 1996;111:718-730.[Abstract/Free Full Text]
- Acar J, Caramanian M, Perrault M, Luxereau P, Arnaud JC. Les insuffisances mitrales par rupture de cordages d’origine dégénérative. Arch Mal Cœur 1968;62:1724.
- Fornes P, Heudes D, Fuzellier JF, Tixier D, Bruneval P, Carpentier A. Correlation between clinical and histologic patterns of degenerative mitral valve insufficiency: a histomorphometric study of 130 excised segments. Cardiovasc Pathol 1999;8:81-92.[CrossRef][Medline]
- Enriquez-Sarano M, Tajik AJ, Schaff HV, Orszulak TA, Bailey KR, Frye RL. Echocardiographic prediction of survival after surgical correction of organic mitral regurgitation. Circulation 1994;90(2):830-837.[Abstract/Free Full Text]
- Allison MA, Cheung P, Criqui MH, Langer RD, Wright CM. Mitral and aortic annular calcification is highly associated with systemic calcified atherosclerosis. Circulation 2006;113:861-866.[Abstract/Free Full Text]
- Sgorbini L, Scuteri A, Leggio M, Leggio F. Association of mitral annulus calcification, aortic valve calcification with carotid intima media thickness. Cardiovasc Ultrasound 2004;2:19.[CrossRef][Medline]
- Bichell DP, Adams DH, Aranki SF, Rizzo RJ, Cohn LH. Repair of mitral regurgitation from myxomatous degeneration in the patient with a severely calcified posterior annulus. J Card Surg 1995;10(4 Pt 1):281-284.[CrossRef][Medline]
- Feindel CM, Tufail Z, David TE, Ivanov J, Armstrong S. Mitral valve surgery in patients with extensive calcification of the mitral annulus. J Thorac Cardiovasc Surg 2003;126:777-782.[Abstract/Free Full Text]
- Ng CK, Punzengruber C, Pachinger O, Nesser J, Auer H, Franke H, Hartl P. Valve repair in mitral regurgitation complicated by severe annulus calcification. Ann Thorac Surg 2000;70:53-58.[Abstract/Free Full Text]
- Maisano F, Caldarola A, Blasio A, De Bonis M, La Canna G, Alfieri O. Midterm results of edge-to-edge mitral valve repair without annuloplasty. J Thorac Cardiovasc Surg 2003;126:1987-1997.[Abstract/Free Full Text]
- Harpaz D, Auerbach I, Vered Z, Motro M, Tobar A, Rosenblatt S. Caseous calcification of the mitral annulus: a neglected, unrecognized diagnosis. J Am Soc Echocardiogr 2001;14:825-831.[CrossRef][Medline]
- Maurer G, Czer LS, Chaux A, Bolger AF, DeRoberts M, Resser K, Kass RM, Lee ME, Matloff JM. Intraoperative Doppler colour flow mapping for assessment of valve repair for mitral regurgitation. Am J Cardiol Aug 1987;60(4):333-337.[CrossRef][Medline]
- Jault F, Rama A, Bonnet N, Reagan M, Nectoux M, Petitclerc T, Pavie A. Gandjbakhch Cardiac surgery in patients receiving long-term hemodialysis. Short and long-term results. J Cardiovasc Surg (Torino) 2003;44:725-730.[Medline]
- Bittrick J, D’Cruz IA, Wall BM, Mansour N, Mangold T. Differences and similarities between patients with and without end-stage renal disease, with regard to location of intracardiac calcification. Echocardiography 2002;19:1-6.[CrossRef][Medline]
- Chan V, Jamieson WR, Fleisher AG, Denmark D, Chan F, Germann E. Valve replacement surgery in end-stage renal failure: mechanical prostheses versus bioprostheses. Ann Thorac Surg 2006;81:857-862.[Abstract/Free Full Text]
- Nataf P, Pavie A, Jault F, Bors V, Cabrol C, Gandjbakhch I. Intra-atrial insertion of a mitral prosthesis in a destroyed or calcified mitral annulus. Ann Thorac Surg 1994;58:163-167.[Abstract]
Top
Abstract
1. Introduction
