Key words
LGE MRI - ischemic mitral regurgitation - myocardial infarction - papillary muscle
Introduction
Myocardial infarction (MI) is frequently associated with ischemic mitral regurgitation
[1]. Acute mitral regurgitation necessitating immediate surgery due to complete or sequential
papillary muscle (PM) rupture is a rare major adverse cardiac event [2]
[3]. In the majority of patients, ischemic mitral regurgitation develops gradually due
to remodeling of the ischemic myocardium and altered hemodynamic conditions in the
left ventricle [4]. Additionally, PM infarction may contribute to the development of ischemic mitral
regurgitation by tethering of the chordae and subsequent dysfunction of the PM-chorda-mitral
valve complex [5]. Ischemic mitral regurgitation is associated with cardiac morbidity and mortality
and is therefore regarded as an unfavorable prognostic factor [6].
Although PM viability is regarded as a prognostic factor when considering mitral valve
replacement or repair, its impact on the development of ischemic mitral regurgitation
is uncertain [3]
[7]
[8]
[9]
[10]. In comparison to intraoperative findings, the extent of PM damage is frequently
underestimated by echocardiography [11]. Thus, patients with ischemic mitral regurgitation may benefit from magnetic resonance
imaging (MRI) prior to mitral valve repair or replacement as it provides reliable
information on left ventricular function, myocardial viability and PM morphology [8]
[12]
[13]
[14].
The aim of the presented study was to correlate the frequency of PM involvement depicted
by LGE MRI in patients with chronic MI with the presence and extent of ischemic mitral
regurgitation assessed by echocardiography.
Method and Materials
Patient group
280 patients with MI depicted by LGE-MRI undergoing cardiac MRI for clinical reasons
were retrospectively screened. Ultimately 48 patients met the inclusion criteria.
Patient characteristics and medication are tabulated in [Table 1].
Table 1
Risk factors for myocardial infarction and cardiac-related medication in the study
group.
Tab. 1 Risikofaktoren für einen Myokardinfarkt und medikamentöse Therapie der Studienpopulation.
|
29 patients with mitral regurgitation
|
19 patients without mitral regurgitation
|
p-value
|
|
age (years)
|
59.9 ± 11.8
|
63 ± 9.2
|
0.36
|
|
male sex
|
24
|
17
|
0.55
|
|
female sex
|
5
|
2
|
0.55
|
|
arterial hypertension
|
20
|
15
|
0.46
|
|
diabetes mellitus
|
5
|
4
|
0.75
|
|
nicotine abuse
|
16
|
9
|
0.61
|
|
familial disposition
|
13
|
8
|
0.86
|
|
hyperlipoproteinemia
|
19
|
11
|
0.61
|
|
BMI (kg/m2)
|
25.6 ± 2.6
|
25 ± 3
|
0.77
|
|
CKmax (U/l)
|
1835.8 ± 1912.5
|
1347 ± 1622.7
|
0.37
|
|
thrombocyte aggregation inhibitor
|
27
|
18
|
0.84
|
|
vitamin K antagonist
|
1
|
4
|
0.07
|
|
beta blocker
|
27
|
17
|
0.69
|
|
Ca antagonist
|
1
|
0
|
0.56
|
|
ACE inhibitor
|
21
|
15
|
0.62
|
|
angiotensin antag.
|
4
|
3
|
0.86
|
|
antianginosum
|
5
|
4
|
0.75
|
|
diuretic
|
17
|
10
|
0.69
|
|
CSE inhibitor
|
27
|
19
|
0.34
|
|
other
|
21
|
10
|
0.18
|
Most patients were treated with a combination of antihypertensive and anticoagulant
drugs. Non-cardiac-related medication indicated as other includes medication for diabetes,
thyroid dysfunction, reflux esophagitis, and gout. BMI – body mass index, CK – creatine
kinase, MI – myocardial infarction, Ca – calcium, ACE – angiotensin converting enzyme,
CSE – cholesterol synthesis enzyme, angiotensin antag – angiotensin antagonists. Antianginosum
refers to nitrate.
Die meisten Patienten wurden mit einer Kombination aus blutdrucksenkenden und gerinnungshemmenden
Medikamenten behandelt. Die nicht Herz bezogene Medikation wird untere sonstige angegeben
und beinhaltet Medikamente zur Behandlung von Diabetes, Schilddrüsenfunktionsstörungen,
Refluxösophagitis, Gicht. BMI – Body Mass Index, CK – Creatinkinase, MI – Myokardinfarkt,
Ca – Kalzium, ACE – Angiotensin Converting Enzyme, CSE – Cholesterol Synthesis Enzyme,
angiotensin antag – Angiotensin Antagonisten. Antianginosum bezieht sich auf Nitrate.
Inclusion criteria
The inclusion criteria were: proven single event of MI clearly related to one cardiac
region and more than three months prior to cardiac MRI, current echocardiography and
patient age between 18 and 80 years.
Exclusion criteria
The exclusion criteria were: aortic valve insufficiency grade 2 or higher, MI involving
more than one cardiac region, known mitral valve leaflet pathology, ventricular or
atrial septum defect, and general contraindications for MRI.
Ultimately 48 patients meeting the inclusion criteria could be included retrospectively.
The reason for the exclusion of 232 patients was incompleteness of essential data:
unknown date of MI, no CK values of acute MI on record, no contemporary echocardiography,
no assessment of mitral valve function on record.
Patient characteristics
Clinical data including arterial hypertension, hyperlipoproteinemia, diabetes mellitus,
nicotine consumption and body mass index (BMI) were recorded according to the patient
file. The maximum serum creatinine kinase (CKmax) recorded during acute MI was correlated with the extent of chronic MI depicted by
LGE MRI. CKmax was defined as the highest creatine kinase value on record measured in blood samples
taken from the patient during hospitalization for acute myocardial infarction. The
CKmax value was considered representative for the size of the MI.
MR imaging
All MR examinations were performed on a 1.5 T imager (Magnetom Sonata / Avanto Siemens
Healthineers, Erlangen, Germany), using a dedicated multichannel receiver coil. The
MR imaging protocol included cine steady-state free-precession (SSFP) sequences in
a 4-chamber view, 2-chamber view as well as in short-axis views from mitral valve
to apex. The sequence parameters were as follows: TR 39.9 ms, TE 1.12 ms, slice thickness
6 mm, field of view 300 – 360 mm, flip angle 70°, matrix 192*100 and a T1w inversion
recovery (IR) GRE sequence in corresponding slice location and orientation to the
cine SSFP sequences (TR 11 ms, TE 44 ms, slice thickness 6 mm, flip angle 25°, baseline
matrix 265) 15 minutes after intravenous contrast medium (0.15 mmol Gadobutrol/kg
bodyweight, Bayer Vital, Leverkusen, Germany) injection. By using an inversion time
localizer sequence, the individual inversion time (TI) was determined to optimally
suppress the signal from the normal myocardium. The mean interval between MI and MR
examination was 1051 ± 1728 days. The mean interval between MRI and echocardiography
was 81 ± 71 days.
Image interpretation
MR images were assessed by two independent experienced readers for the presence, extent
and location of PM infarction (CB: eight years of cardiac MRI experience, BK: thirteen
years of cardiac MRI experience). Quantitative assessment of left ventricular function
and size of the MI was performed by two readers in consensus (CB, HH).
Left ventricular function
Dedicated software (ARGUS®, Siemens Healthineers, Erlangen, Germany) was used for post-processing and determination
of cardiac functional parameters (ejection fraction (EF), end-diastolic volume (EDV),
end-systolic volume (ESV), stroke volume (SV)). For the assessment of functional parameters,
the modified Simpson’s rule was used, based on the manually segmented left ventricular
contours in end-diastolic and end-systolic short-axis images [15]. All parameters are given as normalized values (normalized on patient’s body surface
area (BSA)).
Myocardial viability
Image evaluation included detection of PM involvement and transmural extent of MI.
The myocardium was defined as infarcted if regional late contrast enhancement was
observed 15 minutes after the injection of contrast medium. PM involvement was assessed
regarding the extent and location. PM infarction was differentiated as anterior and
posterior papillary muscle infarction and complete or partial infarction ([Fig. 1]). Hyperenhancement on LGE images of the whole cross-sectional area of the involved
PM was classified as complete papillary muscle infarction since viability of the PM-chorda
unit was not preserved in its continuity ([Fig. 1b]). Hyperenhancement on LGE images of a part of the cross-sectional area of the involved
PM was classified as partial papillary muscle infarction with preserved continuity
of the viability of the PM-chorda unit ([Fig. 1a]). Hyperenhancement of the PM surface only was not regarded as PM infarction as this
could arise from enhancement of the normal endocardium which is also present in patients
without MI. For quantitative assessment of MI, manual planimetry of hyperenhancement
representing MI on LGE images was performed on all slices on which MI was identified
and multiplied by the distance between the slice center positions to calculate volumes
[15].
Fig. 1 a The schematic drawing depicts a partial infarction of the anterior PM (black) with
preserved continuity of viable tissue on the base of a subendocardial MI of the anterior
wall. b The schematic drawing depicts complete infarction of the posterior PM on the base
of a transmural MI of the posterior wall. The continuity of viable tissue in the PM-chorda-mitral
valve complex is not preserved.
Abb. 1 a Die Schemazeichnung zeigt einen teilweisen Infarkt des vorderen Papillarmuskels (schwarz)
mit erhaltener Kontinuität vitalen Gewebes auf dem Boden eines subendokardialen Vorderwandinfarkts.
b Die Schemazeichnung zeigt einen vollständigen Infarkt des hinteren Papillarmuskels
auf dem Boden eines transmuralen Hinterwandinfarkts. Die Kontinuität vitalen Gewebes
der Papillarmuskel-Chorda-Mitralklappen Einheit ist nicht erhalten.
Statistical analysis
Data were statistically evaluated using dedicated software (JMP, SAS Institute Inc.,
North Carolina, USA). If not otherwise indicated, results are given as mean ± standard
deviation. To test for a correlation of PM infarction and MI size as well as CKmax, both parameters were compared in patients with and without PM involvement. As data
did not show normal distribution, logarithmic values were used. To compare patients
with and without PM involvement, Student`s t-test was applied. The correlation index
for CKmax values and MI size depicted by LGE was calculated. Left ventricular functional parameters
including EF, EDV, ESV, and SV were compared to test for differences between patients
with and without PM involvement. Functional parameters were normalized to body surface
to reduce a bias arising from patient size. To test for differences in left ventricular
function and patient age, Student`s t-test was applied. These functional parameters
were also compared for patients with and without mitral regurgitation to analyze the
impact of ischemic mitral regurgitation on left ventricular function assessed by MRI.
To test for differences in the frequency and degree of ischemic mitral regurgitation
and transmural extent of MI between patients with and without PM infarction, the Mann-Whitney
U-test was applied. P-values < 0.05 were considered significant.
Results
The mean age was 59.4 ± 13.2 years for patients with PM infarction and 62 ± 10.1 years
for patients without PM infarction, p = 0.55, which is not a significant difference.
Left ventricular function
Results of the functional analysis are tabulated in [Table 2], [3]. There are no significant differences for patients with and without PM infarction
regarding left ventricular function. The mean left ventricular ESV is moderately increased
in patients with PM infarction ([Table 2]). In contrast, the comparison of functional parameters between patients with and
without mitral regurgitation reveals a significantly increased ESV in patients with
mitral regurgitation whereas the other parameters do not differ significantly ([Table 3]). The degree of ischemic mitral regurgitation did not differ significantly between
patients with and without PM involvement ([Table 2]). However, regarding only patients with PM infarction, ischemic mitral regurgitation
is present in four of four patients (100 %) with complete PM infarction, whereas ischemic
mitral regurgitation in patients with partial PM infarction is present in four of
seven patients (57 %). Ischemic mitral regurgitation was found in 22 of 37 patients
without PM infarction (59 %).
Table 2
There is no significant difference regarding frequency and degree of ischemic mitral
regurgitation as well as size and transmurality of myocardial infarction and CKmax values between patients with (PM+) and without (PM–) papillary muscle involvement.
The comparison of the left ventricular function parameters indicates no significant
difference. Only the mean left ventricular endsystolic volume is moderately increased
in patients with papillary muscle involvement.
Tab. 2 Es besteht kein signifikanter Unterschied bezüglich Häufigkeit und Grad der Mitralinsuffizienz
sowie der Infarktgröße und Transmuralität sowie der CKmax Werte zwischen Patienten mit (PM+) und ohne (PM–) Papillarmuskel Beteiligung. Der
Vergleich der linksventrikulären Funktionsparameter zeigt keinen signifikanten Unterschied,
lediglich das linksventrikukäre endsystolische Volumen ist bei Patienten mit PM Infarkt
leicht erhöht.
|
PM infarction
|
no PM infarction
|
p-value
|
|
mitral regurgitation
|
8 (73 %)
|
22 (59 %)
|
0.44
|
|
grade 1
|
5
|
15
|
0.690
|
|
grade 1 – 2
|
3
|
4
|
0.205
|
|
grade 2
|
0
|
3
|
0.103
|
|
EDV (ml)/ BSA
|
79 ± 31
|
66 ± 20
|
0.205
|
|
ESV (ml)/ BSA
|
50 ± 25
|
38 ± 18
|
0.069
|
|
EF (%)
|
50 ± 11
|
52 ± 10
|
0.424
|
|
SV (ml)/ BSA
|
40 ± 9
|
36 ± 9
|
0.194
|
|
MI size (ml)
|
14.7 ± 12.8
|
12.1 ± 8.1
|
|
|
logarithmic MI size
|
2.5 ± 0.7
|
2.2 ± 0.3
|
0.351
|
|
transmural MI
|
8
|
22
|
0.440
|
|
CKmax (U/l)
|
1924 ± 2344
|
1551 ± 1628
|
0.835
|
Table 3
The mean left ventricular ESV is significantly increased in patients with ischemic
mitral regurgitation compared to patients without mitral regurgitation. There is no
significant difference regarding the other functional parameters.
Tab. 3 Das mittlere linksventrikuläre ESV ist bei Patienten mit ischämischer Mitralinsuffizienz
signifikant erhöht gegenüber Patienten ohne Mitralinsuffizienz. Die übrigen Funktionsparameter
weisen keine statistisch signifikanten Unterschiede auf.
|
mitral regurgitation
|
no mitral regurgitation
|
p-value
|
|
EDV (ml)/ BSA
|
69 ± 28
|
69 ± 13
|
0.94
|
|
ESV (ml)/ BSA
|
47 ± 23
|
32 ± 10
|
0.007
|
|
EF (%)
|
50 ± 9
|
54 ± 11
|
0.120
|
|
SV (ml)/ BSA
|
37 ± 9
|
37 ± 8
|
0.94
|
Myocardial viability
Results of the myocardial viability analysis are tabulated in [Table 2]. PM infarction depicted by LGE was found in 11 of 48 patients ([Fig. 2], [3], [4]). Neither the size of the MI depicted by LGE MRI nor transmural extent nor CKmax values differed significantly between patients with and without PM involvement ([Table 2]). A moderate correlation of r = 0.407 (p = 0.0045) was found between mean CKmax and mean infarction size. The distribution and extent of papillary muscle infarction
are tabulated in [Table 4].
Fig. 2 a Frequency of ischemic mitral regurgitation correlated with the number of patients
with complete, partial and no PM infarction indicated as number of patients. b Proportion of patients with ischemic mitral regurgitation indicated as percentage
of 100 % for each group (complete, partial and no PM infarction).
Abb. 2 a Häufigkeit der ischämischen Mitralinsuffizienz bei Patienten mit vollständigem, teilweisen
und keinem Papillarmuskelinfarkt angegeben als Patientenzahl. b Anteil der Patienten mit ischämischer Mitralinsuffizienz angegeben als Prozentsatz
von 100 % für jede Patientengruppe (vollständiger, teilweiser, kein Papillarmuskelinfarkt).
Fig. 3 LGE short-axis views a–c and long-axis view d acquired in a 44-year-old male patient with transmural myocardial infarction of the
apex (d, asterisk). The posterior papillary muscle exhibits superficial contrast enhancement
(a–d, arrows) with a viable core (a–c, dashed arrows) representing partial ischemic necrosis.
Abb. 3 Die LGE Kurzachsenschnitte a–c und Langachsenschnitt d eines 44-jährigen Patienten zeigen einen transmuralen Myokardinfarkt der Herzspitze
(d, Stern). Der hintere Papillarmuskel weist eine oberflächlich vermehrte Kontrastmittelaufahme
(a–c, Pfeile) bedingt durch ischämische Nekrose mit einem intakten Kern (a–c, gestrichelte Pfeile) auf.
Fig. 4 LGE short-axis views a–c and cine SSFP long-axis view d acquired in a 74-year-old male patient with transmural myocardial infarction of the
inferior and inferolateral wall. The posterior papillary muscle exhibits contrast
enhancement in all parts (a–c, arrows) representing complete ischemic necrosis. A first grade mitral regurgitation
is displayed by a central systolic regurgitation jet on the cine SSFP image (d, dashed arrow).
Abb. 4 Die LGE Kurzachsenschnitte a–c und der cine SSFP Langachsenschnitt d eines 74-jährigen Patienten mit transmuralem Myokardinfarkt der inferioren und inferolateralen
Wand zeigen eine vollständige ischämische Nekrose des hinteren Papillarmuskels (a–c, Pfeile). Der systolische Regurgitationsjet in der cine SSFP Bildgebung (d, gestrichelter Pfeil) zeigt eine erstgradige Mitralinsuffizienz.
Table 4
The distribution analysis of PM infarction indicates a preference for the posterior
PM which is more frequently affected. In two patients both the anterior and posterior
PM were affected thus resulting in four patients with anterior PM and nine patients
with posterior PM involvement and in summary eleven patients with PM infarction.
Tab. 4 Die Verteilungsanalyse zeigt eine Bevorzugung des hinteren PM der häufiger betroffen
ist. Bei zwei Patienten sind sowohl der hintere wie auch der vordere PM betroffen
wodurch sich eine Gesamtzahl von elf Patienten mit PM Infarkt bei vier Patienten mit
vorderer und neun Patienten mit hinterer PM Beteiligung ergibt.
|
anterior PM
|
posterior PM
|
|
patients
|
4 (36 %)
|
9 (82 %)
|
|
complete MI
|
0
|
4 (44 %)
|
|
partial MI
|
4 (100 %)
|
5 (56 %)
|
|
mitral regurgitation
|
2 (50 %)
|
5 (56 %)
|
Discussion
Ischemic mitral regurgitation is associated with increased morbidity and mortality
and is consequently regarded as an unfavorable prognostic factor [6]. Whether ischemic mitral regurgitation arises from dysfunction of the mitral valve
complex due to PM infarction or is an effect of left ventricular remodeling is uncertain
[5]. In the present study no difference between patients with and without PM infarction
could be found regarding frequency and degree of ischemic mitral regurgitation as
well as left ventricular function and myocardial viability. Patients with MI of the
complete cross-sectional area of PM were more frequently affected by ischemic mitral
regurgitation than patients with only partial PM infarction although there was no
difference in its severity. Patients with partial PM infarction did not differ from
patients with MI other than a lack of PM involvement regarding ischemic mitral regurgitation.
Compared to results published by Tanimoto et al. [17] reporting a prevalence of 40 %, PM infarction was observed less frequently in the
present study with a prevalence of 23 %. These results are more in line with a frequency
of 19 – 32 % as published in autopsy studies [18]. As PM involvement correlates with the size of the MI, the explanation for the higher
prevalence of PM infarction in the Tanimoto study is probably the larger size of MI
reflected by markedly higher CKmax values. Tanimoto et al. conclude that ischemic mitral regurgitation is brought about
by left ventricular remodeling, and there is no evidence of a contribution of PM infarction.
However, they assessed only the presence of PM infarction and the size of the MI,
whereas the present study analyzed the extent of PM infarction in order to assess
its impact on PM and mitral valve function. In all four patients with complete PM
infarction ([Fig. 4]), mitral regurgitation was present, compared to only four of seven patients with
partial PM infarction ([Fig. 3]). The difference between patients with MI of the complete cross-sectional area and
patients with only partial PM infarction suggests a correlation with the development
of ischemic mitral regurgitation. As there is no difference in the frequency of ischemic
mitral regurgitation between patients with partial PM infraction (57 %) and patients
without PM involvement (59 %), the presence of partial PM infarction but preserved
viable mitral-chorda-papillary muscle continuity seems to have no adverse effect on
mitral valve function. These results suggest a functional impairment of the mitral-chorda-papillary
muscle complex if no viable continuity is preserved due to complete PM infarction
although the necrotic PM is still in place. As the degree of ischemic mitral regurgitation
did not differ between patients with complete, partial or no PM infarction, complete
PM infarction seems to be associated with an increased prevalence of mitral regurgitation
but does not result in high-grade mitral regurgitation. A potential reason is that
complete PM infarction impairs the function of the PM-chorda-mitral valve complex
due to discontinuity of viable tissue but the fact that the necrotic tissue is still
in place provides residual function, thus preventing high-grade mitral regurgitation
which in contrast occurs in the event of PM rupture.
These findings are supported by the fact that preservation of the sub-valvular apparatus
and papillary muscle continuity contributes to the improvement of the prognosis in
patients with ischemic mitral regurgitation [19]. Consequently, LGE-MRI could be a suitable approach for identifying patients at
risk for the development of ischemic mitral valve regurgitation, based on the extent
of PM involvement [11].
In contrast to the current literature, the size of the MI as well as CKmax did not differ significantly between patients with and without PM involvement although
both were slightly higher in patients with PM involvement [20]. This might be a result of the number of patients with PM involvement being too
small to reach significance.
Ischemic mitral regurgitation was found in 73 % of patients with PM infarction and
in 59 % of patients with MI but no sign of PM involvement. Ischemic mitral regurgitation
after MI is frequently observed and is usually mild [8]
[21]
[22]. In our study group, second-grade mitral regurgitation was found in 3 of 29 patients
with ischemic mitral regurgitation compared to 10 % in the published data [8]. Yet, in patients with ischemic mitral regurgitation, even a mild degree of mitral
regurgitation is strongly associated with an unfavorable prognosis. On the other hand,
mild mitral regurgitation is frequently observed in about 45 % of people older than
50 years [23]. Consequently, difficulties in differentiating between ischemic mitral regurgitation
after MI and preexisting mitral regurgitation may arise as the mean age of patients
was 61 years and mitral regurgitation could therefore be expected in a significant
percentage.
Ischemic mitral regurgitation is believed to initiate myocardial remodeling associated
with left ventricular dilatation due to increased diastolic wall stress and decreased
contractility with a consecutively increased end-systolic volume [24]
[25]. Our results are in concordance with these reports, as the end-systolic volume was
significantly increased in patients with ischemic mitral regurgitation.
Published data indicate that PM infarction is usually the result of ischemia of the
inferior and inferolateral left ventricular myocardium [17]
[26]. Due to the vascular anatomy of the PM, the anterior PM is much less frequently
affected [11]
[27]. The posterior PM was also more frequently affected, with nine patients compared
to four with anterior PM involvement. Yet, the difference was lower compared to the
published data, which referred however to a rupture of the PM, reporting a three-fold
higher frequency.
The limitations of our study are the small number of patients identified to have PM
infarction, thus restricting the potential to derive statistically valid conclusions
regarding the contribution of PM infarction to the development of ischemic mitral
regurgitation and to differentiate between ischemic mitral regurgitation due to left
ventricular remodeling after MI. Secondly, there is no surgical or histopathological
confirmation for the presence of PM infarction depicted by LGE MRI. However, as the
published data give evidence of an excellent correlation of LGE imaging of myocardial
viability for the presence and extent of MI with histopathological results, this constitutes
a minor limitation and justifies the assumption that LGE of PM represents ischemic
necrosis due to MI [12]
[28]
[29]. Moreover, no patients with high-grade mitral regurgitation were included, thus
representing a statistical bias. High-grade mitral regurgitation in PM infarction
usually results from acute PM rupture and is a major adverse cardiac event [30]. As it is a rare event necessitating immediate mitral valve repair or replacement,
these patients are not referred to MRI. In the case of limited PM infarction, especially
if restricted to the superficial part, difficulties may arise with respect to the
differentiation of this from normal enhancement of the endocardium on LGE images.
In conclusion, there is no difference in the overall frequency of ischemic mitral
regurgitation between patients with and without PM infarction. Compared to patients
with partial or no PM infarction, the frequency of mitral regurgitation is increased
in patients with complete PM infarction, but the severity does not differ. A potential
reason is the impaired PM function due to the discontinuity of viable tissue in spite
of preserved continuity of the PM-chorda-mitral valve complex resulting in increased
frequency of light ischemic mitral regurgitation.
