Get documents about "An evAluAtion of mitrAl vAlve procedures using the europeAn system
Document Sample
Scandinavian Journal of Surgery 97: 254–258, 2008
An evAluAtion of mitrAl vAlve procedures using
the europeAn system for cArdiAc operAtive
risk evAluAtion
t. kaartama, l. heikkinen, A. vento
Department of Cardiothoracic Surgery, Helsinki University Central Hospital, Helsinki, Finland
ABstrAct
Background and Aims: this study was undertaken in order to evaluate the usefulness of
the euroscore in the choice and outcome of mitral valve procedures undertaken at the
helsinki university central hospital.
Material and Methods: data from 378 patients was collected. predicted mortalities were
calculated for all patients using the european system for cardiac operative risk evalu-
ation and different mitral valve procedures were compared with 30-day mortality, length
of hospital care and rate of post-operative complications.
Results: the mortality rate in the mitral valve repair (mvp) group decreased gradu-
ally from 5.9% (in 1999) to 2.2% (2003). the variation of annual mortality was higher in
the mitral valve replacement (mvr) group. the predicted mortality given by euroscore
increased over the years in both groups. the mortality in the mvr group was nearly four
times higher than in the mvp group.
the length of both intensive and overall hospital stay decreased in patients with mvp
procedures. post-operative survival was 89% in the mvp patients and 74% in mvr pa-
tients after three years.
Discussion: the results of mitral valve operations have improved. this is observed as
decreased mortality rates and lengths of hospital care in the mvp group, although the
predicted mortality rate was increased.
Key words: Mitral valve repair; Euroscore; risk evaluation; mitral valve reconstruction; annuloplasty; valve
surgery
INTRODUCTION light the liability of the procedure in various situa-
tions (1, 2, 3). The main purpose of this study was to
There has been a pivotal change in mitral valve oper- evaluate the type and results of mitral valve proce-
ations, with a move from mitral valve replacement dures undertaken using the European System for
(MVR) to mitral valve repair (MVP) over the last few Cardiac Operative Risk, and to contribute to the dis-
decades. Several studies have already reported con- cussion on how this scoring system can effect the
vincing long-term results with MVP’s which high- choice of procedure – i.e. whether to repair or re-
place.
Correspondence:
Antti Vento, M.D.
Helsinki University Central Hospital MATERIAL AND METHODS
Department of Cardiothoracic Surgery
PL 340 The data of 378 patients operated on between January 1999
FIN - 00029 HUS, Helsinki and December 2003 at Helsinki University Central Hospital
Finland was collected retrospectively, 227 patients had undergone
E-mail: antti.vento@hus.fi MVP and 151 MVR. Carpentier techniques were used in
Risk evaluation in mitral valve surgery 255
mitral valve repair (4) – including a quadrangular resection used to draw up survival curves. Statistical analyses were
and sliding-technique in posterior leaflet operations, a tri- carried out using the SPSS 12.0.1 for Windows program
angular resection in anterior leaflet, ring annuloplasty, package. Student’s t-test was used for comparative statis-
chordal-shortening, artificial chords and chordal-replace- tics of normally distributed data, and the Mann-Whitney
ments. Additional methods used were closure of clefts, a U-test for non-normally distributed data. Fischer’s exact
widening of the anterior leaflet with a pericardium-patch test was applied in analysing the incidence of mortality and
and support of the annulus with a posterior Gore-Tex-band. complications. .
Both biological and mechanical prostheses were used to
replace the valve. In addition, surgery for atrial fibrillation
was performed on 26 patients (5). Simultaneous coronary RESULTS
artery bypass grafting (CABG), aortic valve replacement
(AVR) and/or tricuspid valve operations were performed
as needed. Vancomysin 1g and kefuroxime 1.5g were given Preoperative data has been collected into Table 1.
at the beginning of anaesthesia and 8 hours later as an an- Three hundred and seventy-eight (239 male and 139
tibiotic prophylaxis. Low molecular weight heparin was female) patients were operated during this time pe-
given postoperatively until INR values were at the treat- riod. Although left ventricle ejection fraction (EF) did
ment level after the beginning of warfarin treatment. Pa- not differ between the two groups, patients in the
tients in sinus rhythm who had MVP or MVR with bio- MVR group had significantly higher mean systolic
logical valve were treated with warfarin for three months pulmonary artery pressures (syst. pPA) .
postoperatively, while those with atrial fibrillation and/or Of the mitral valve plasties (MVP) performed, the
who had MVR with mechanical prostheses take warfarin posterior leaflet alone was repaired in 56.3%, and the
for life. Complications included infections, valve-related anterior leaflet alone in 15.3%. The sliding-technique
problems (valve regurgitations or paravalvular leakage),
haemorrhage, transient neurological disorders (TIA, de- was used in 20.5% of the posterior leaflet procedures.
layed wakening, convulsions, deterioration, confusion, An anterior leaflet was repaired with simultaneous
agitation) and permanent neurological disorders (stroke, aortic root reconstruction by homograft in one pa-
hypoxic ischemic encephalopathy, paraplegia). Diagnoses tient, and pericardium patches were used in 13 cases
were made by a neurologist and was completed by CT-scan of anterior leaflet repair. Both leaflets were repaired
or other neurological interventions according to a neurolo- in 11.3% of the cases. In 17.1% of the patients ring
gist. The length of ICU and hospital stay was recorded. annuloplasty only was performed. Of the valve re-
Transthoracic or transesophageal echo was used to assess placement operations (MVR) performed, biological
heart function and postoperative result, cardiac rhythm prostheses were used in 44.7% of the cases. Simulta-
was recorded and the MB-fraction of creatine kinase (s-CK-
MB) was measured postoperatively. neous CABG was performed in 31.3% (71) of the MVP
group and 34.4% (52) of the MVR group. A concomi-
tant aortic valve operation was performed in 10.6%
STATISTICAL ANALySIS: (40) of the operations, and a tricuspid valve proce-
We calculated risk factors for all patients using the Euros-
dure in 5.3%(22), respectively. Surgery for atrial fibril-
core and compared them to actual 30-day mortalities (6). To lation was performed simultaneously with mitral
define the possible change in the choice of mitral valve valve operations in 7.5% (17) of the MVP group and
procedures during the followed period, we made a year- in 6.0% (9) of the MVR group in order to restore si-
by-year study of mortality rates, risk and lengths of post- nus-rhythm. MVP was performed on 10 (59%) and
operative hospital stay. The Kaplan-Meier method was MVR on 7 (41%) patients with endocarditis.
TABLE 1
Preoperative data of the patients
MVP MVR P
Gender 170(m)/57(f) 69(m)/82(f) P = .000
(female = 25.11%) (female = 54.30%)
Age 60.78y (SD = 12.46) 64.87y (SD = 11.02) P = .001
Logistic EuroSCORE 5.63% (SD = 7.78) 11.66% (SD = 11.90) P = .000
BMI 25.26 (SD = 3.93) 25.37 (SD = 4.13) P = .776
BSA 187.88 (SD = 20.88) 179.64 (SD = 22.28) P = .000
EF 58.68% (SD = 13.03) 58.74% (SD = 12.41) P = .971
Pre-operative endocarditis 4.50% 4.67% P = .942
Calcified annulus 5.86% 21.48% P = .000
Systolic pPA 44.39 mmHg (SD = 18.45) 52.63 mmHg (SD = 17.70) P = .000
MVP = 176
MVR=109
pPCW (mean) 18.53 mmHg (SD = 7.76) 22.02 mmHg (SD = 7.75) P = .000
MVP = 159
MVR = 104
BMI = Body Mass Index, BSA = Body Surface Area, pPA = pulmonary artery pressure, pPCW = Pulmonary Capillary Wedge Pressure
256 T. Kaartama, L. Heikkinen, A. Vento
TABLE 2
Annual change of predicted and actual mortality, ICU and hospital stay of the MVP group:
MVP:
year Euro-SCORE(%) 30-day mortality(%) Difference ICU-stay (d) Hospital stay (d) n
1999 3.61 5.88 –2.27 4.69 11.25 17
2000 6.24 5.88 +0.36 5.57 11.31 51
2001 6.98 3.51 +3.47 5.96 12.16 57
2002 5.66 1.75 +3.91 4.50 10.82 57
2003 4.00 2.22 +1.78 4.04 9.38 45
P = .276 P = .761 P = .335 P = .162
TABLE 3
Annual change of predicted and actual mortality, ICU and hospital stay of the MVR group.
MVR:
year Euro-SCORE(%) 30-day mortality(%) Difference ICU-stay (d) Hospital stay (d) n
1999 10.06 12.50 –02.44 5.95 11.55 40
2000 09.80 12.90 –03.10 8.13 14.53 31
2001 18.13 35.29 –17.16 7.76 14.71 17
2002 11.89 18.18 –06.29 8.30 14.97 33
2003 11.74 00.00 +11.74 6.27 12.80 30
P = .164 P = .018 P =. 463 P = .446
TABLE 4
Predicted and actual mortality of mitral valve procedures with and without CABG
MVP: MVR:
Euroscore 30-day difference Euroscore 30-day Difference n
mortality mortality
MV(–) 4.77% 1.28% +2.32 10.13% 8.08% +1.93 MVP (– CABG) = 156
CABG MVP + CABG = 71
MV(+) 7.54% 8.45% –0.68 14.64% 25.00% –10.36 MVR (– CABG) = 99
CABG MVR + CABG = 52
The presence of a calcified annulus was a major was within its expected range. In contrast, the mortal-
determinant for choosing the procedure. More ity in MVR operations was much higher then ex-
than 24.4% of the patients in the MVR group had pected.
annulus calcification, but only 5.9% in the MVP The mean length of ICU-stay for MVP patients in-
group. Table 1. creased initially during the period from 1999 to 2001,
Overall mortality was 3.5% for MVP operations but then decreased to below its starting level by 2003.
and 13.9% for MVR operations. A gradual decrease The average hospital stay had also decreased by the
was observed in the mortality rate of MVP group end of 2003. The average lengths of ICU and hospital
from 1999 to 2003 (Table 2). Simultaneously, the mean stay for MVR patients has varied a good deal, and –
logistic EuroSCORE at first increased up to 2001 and although they have decreased overall – they were
then decreased in 2003. The difference between pre- slightly higher at the end of the follow-up period
dicted and actual mortality became more favourable than at the beginning.
for MVP patients, while the mortality in the MVR Patients in the MVR group had less transient neu-
group displayed a much wider variation (Table 3). rological disorders and haemorrhages than those in
Nevertheless, apart from in 2001, there has been a the MVP group (Table 5). The amount of permanent
notable rise in the mean logistic EuroSCORE among neurological disorders and valve related problems
patients chosen for MVR between 1999 and 2003 by was the same in both groups. Furthermore, patients
almost two percentage units. in the MVR group suffered more infections and re-
The mortality in MVP and MVR procedures with- operations. Although, statistically could only be
out CABG was lower than predicted (Table 4). The found within infections and mortality of patients
mortality in MVP operations combined with CABG with predicted preoperative mortality rates of 5–10%.
Risk evaluation in mitral valve surgery 257
TABLE 5 Eight (47.1%) of the 17 patients in the MVP group,
Postoperative results for mitral valve procedures. who had had a simultaneous MAZE-procedure had
sinus-rhythm when leaving hospital. In the MVR
MVP MVR P group, two (22.2%) of the nine patients had sinus-
n = 227 n = 151 rhythm at the end of the hospital care, six had atrial
30-day mortality 3.52% 13.91% P = .000
fibrillation and one died during surgery.
In the MVP group, eight patients (3.5%) operated
Infections 3.10% 5.30% P = .284 on between 1999 and 2003 required a subsequent re-
Haemorrhage 5.78% 4.00% P = .442 operation of the valve. Two of them had re-MVP and
Transient neurological 11.01% 6.62% P = .149 six underwent MVR. In addition four other patients
disorders operated on before 1999 were included in the valve
Permanent neurological 2.20% 1.99% P = .886 repair reoperation group. In the MVR group, seven
disorders (4.6%) patients needed a re-operation. Overall, 26
Valve related problems 2.20% 1.99% P = .886
(17.2%) of the MVR procedures were re-operations
due to previous MV plasty (12) or MV replacement
Reoperations 3.52% 5.30% P = .401 (14).
The survival curve displayed a higher mortality
rate in the MVR group than in the MVP group during
the first postoperative year, 80% versus 92%. At three
years the survival rates were 74% and 89%, respec-
The differences were in favour of the patients in the tively. Five-year survival was 86.6% in the MVP
MVP group. All patients with endocarditis who were group and 69.8% in the MVR group (Fig. 1).
operated on survived.
A total of 12.3% of the patients in the MVP group
and 16.6% of the patients in the MVR group had post- DISCUSSION
operative CK-MB levels over 100 µ/ml. The mean EF
decreased peri-operatively in 7.1% (MVP) and 5.9 There has been a change in the procedure of choice
(MVR), respectively. In the MVP group, 71.7% of the during the study period, over which time our teams
patients had sinus-rhythm pre-operatively. Of these, have moved towards repairing the mitral valve with-
57.0% had atrial fibrillation at discharge, while in the out prosthesis. Before the year 2000, MVR was the
MVR group 64.5% of the patients who had preopera- first choice for treating patients with mitral valve fail-
tively been in sinus-rhythm had atrial fibrillation as ures, but after this there have been more repair op-
a post-operative complication erations performed annually. The MVP/MVR ratio
Fig. 1. Five-year survival of
mitral valve repair and re-
placement generated with
Kaplan-Meier method
258 T. Kaartama, L. Heikkinen, A. Vento
has now settled at a rate of 9:1 (MVP/MVR). The during the study period and continue to improve.
good results obtained have encouraged us to operate The decreased mortality rates of the MVR patients
ever more complex cases of mitral failures. together with their increased Euroscore points dem-
The actual mortality in both the MVP and MVR onstrated how the technique of mitral valve replace-
groups was close to the predicted mortality risk for ment procedures has improved. Mitral valve surgery
the procedure. In single MVP and MVR procedures should be concentrated amongst as small a number
the risk calculator overestimates the logistic risk for of surgeons as possible in order to maintain quality.
30-day mortality. Other than isolated coronary sur- The low mortality rate, low incidence of complica-
gery, each surgical procedure tends to increase the tions, low frequency of re-operations, the short peri-
predicted risk by 0.63 percentage units. The mortality ods of hospital stay and the good middle-term sur-
for the MVP + CABG group was in the expected vival confirm that MVP is a safe method for treating
range. The combined surgery with prosthetic valve patients who have mitral valve diseases with less se-
increased the actual mortality much higher than ex- vere symptoms, as estimated by EuroSCORE. The
pected. The numbers of cases were few and the pa- treatment of these patients with MVP allows us to use
tients were in worse condition than MVP patients the MVR procedures only in patients in more severe
preoperatively. ICU stay was somewhat prolonged in preoperative condition.
our hospital compared to data from other studies (7,
8, 9). The reason for this being that there is no step-
down unit in our hospital. If patients require rhythm REFERENCES
observation or observation for another reason, they
therefore needed to be on the ICU. 01. Braunberger E, Deloche A, Berrebi A, Abdallah F, Celestin JA,
Meimoun P, Chatellier G, Chauvaud S, Fabiani JN, Carpentier
The exceptionally high (35 %) mortality in the MVR A: Very long-term results (more than 20 years) of valve repair
group in 2001 can be partially explained by the fol- with carpentier’s techniques in nonrheumatic mitral valve in-
lowing factors: First, we had only 17 patients under- sufficiency. Circulation 2001;104(12 Suppl 1):8–11
going MVR that year (compared to over 30 patients 02. David TE, Omran A, Armstrong S, Sun Z, Ivanov J: Long-term
for every other year). Second, these patients were at results of mitral valve repair for myxomatous disease with and
without chordal replacement with expanded polytetrafluoro-
a very high risk, with a mean logistic EuroSCORE of ethylene sutures. J Thorac Cardiovasc Surg 1998;115(6):1279–
18.1%. Third, although the length of the hospital stay 1285
in the MVR group rose from that in 2000 to 2001, the 03. Zegdi R, Debieche M, Latremouille C, Lebied D, Chardigny C,
length of ICU stay fell. This suggests that the patients Grinda JM, Chauvaud S, Deloche A, Carpentier A, Fabiani JN:
Long-term results of mitral valve repair in active endocarditis.
who underwent MVR in 2001 were in poor pre- Circulation 2005;111(19):2532–2536
operative condition and they either died during or 04. Carpentier A: Cardiac valve surgery––the “French correction”.
almost instantly after surgery, or required long re- J Thorac Cardiovasc Surg 1983;86(3):323–337
covery periods in the hospital. 05. Gillinov AM, Cosgrove DM 3rd: Current status of mitral valve
repair. Am Heart Hosp J 2003;1(1):47–54
When the two different operation methods were 06. Roques F, Michel P, Goldstone AR, Nashef SA: The logistic
compared, we found that the mortality for the MVR EuroSCORE. Eur Heart J 2003;24(9):882–823.
group was nearly four times higher than for the MVP 07. Cosgrove DM 3rd, Sabik JF, Navia JL: Minimally invasive valve
group. However, the patients in the MVR group had operations. Ann Thorac Surg 1998;65(6):1535–8; discussion
1538–1539
almost double the higher preoperative risk points. 08. Cohn LH, Adams DH, Couper GS, Bichell DP, Rosborough
Although the patients in the MVR group had a calci- DM, Sears SP, Aranki SF: Minimally invasive cardiac valve
fied annulus nearly five times more often then pa- surgery improves patient satisfaction while reducing costs of
tients in the MVP group, there was no difference in cardiac valve replacement and repair. Ann Surg 1997;226(4):421–
the incidence of permanent neurological disorders. 6; discussion 427–428
09. Goldsmith I, Lip Gy, Kaukuntla H, Patel RL: Hospital morbid-
Indeed, the patients in the MVR group had signifi- ity and mortality and changes in quality of life following mi-
cantly less transient neurological disorders. The high tral valve surgery in the elderly. J Heart Valve Dis 1999;8(6):702–
mortality rate in the MVR group may also conceal a 707
higher incidence of neurological complications. The 10. Gillinov AM, Faber C, Houghtaling PL, Blackstone EH, Lam
BK, Diaz R, Lytle BW, Sabik JF 3rd, Cosgrove DM 3rd: Repair
five-year survival was higher in MVP group, as has versus replacement for degenerative mitral valve disease with
previously been shown in other studies (10, 11). coexisting ischemic heart disease. J Thorac Cardiovasc Surg
The benefits of MVP and biological valve prosthe- 2003;125(6):1350–1362
ses compared to mechanical prostheses are the lack 11. Thourani VH, Weintraub WS, Guyton RA, Jones EL, Williams
WH, Elkabbani S, Craver JM: Outcomes and long-term sur-
of lifelong warfarin treatment. As biological valve vival for patients undergoing mitral valve repair versus re-
prostheses tend to degenerate, however, these pa- placement: effect of age and concomitant coronary artery by-
tients may need to have a re-operation within a mean pass grafting. Circulation 2003;108(3):298–304
of 15 years (12). Therefore, mechanical prostheses are 12. Hammermeister K, Sethi GK, Henderson WG, Grover FL,
Oprian C, Rahimtoola SH: Outcomes 15 years after valve re-
favoured in patients under 65 years of age who need placement with a mechanical versus a bioprosthetic valve:
to undergo MVR. Final report of the Veterans Affairs randomized trial. J Am Coll
In conclusion, the decreased mortality rate and the Cardiol 2000;36(4):1152–1158
decreased lengths of both ICU and hospital stay of
the patients in the MVP group showed that the re- Received: October 24, 2007
sults of mitral valve repair procedures have improved Accepted: July 9, 2008
