Scand J T h o r Cardiovasc Surg 10: 53-62, 1976
PULMONARY VEIN FLOW PATTERN IN MITRAL STENOSIS BEFORE AND AFTER COMMISSUROTOMY Erling Skagseth From Surgical Depariment A , Rikshospitalet (University Hospital), Oslo, Norway
Scand Cardiovasc J Downloaded from informahealthcare.com by Osaka University on 02/12/15 For personal use only.
(Submitted for publication October 28, 1974)
Abstract. Pulmonary vein flow (PVF) pattern was studied with an electromagnetic flowmeter in 15 patients with mitral stenosis. ECG, left atrial and left ventricular pressure were recorded simultaneously. Commissurotomy was performed in 8 patients and prosthetic valve replacement in the others. Measurements were performed before and after the intracardiac procedure. The predominating forward flow occurred in ventricular diastole, in contrast to ventricular systole in normal hearts. A small retrograde PVF was observed in 11/15 patients in early ventricular systole, coinciding with the c-wave in left atrial pressure (LAP). The reversed flow did not exceed 13% and showed no correlation to a slight regurgitation demonstrated at left ventricular angiography. The maximum delay in forward systolic PVF after onset of ventricular systole was 0.10 sec. The PVF pattern remained unchanged after commissurotomy in the majority of patients, even after reduction of the mitral valve gradient to less than 5 mmHg. However, holosystolic reversed PVF occurred in one patient, indicating a severe mitral regurgitation. The method of PVF recording is simple and proved useful for haemodynamic analysis and as a control after mitral valve surgery for evaluating the degree of mitral regurgitation.
The normal pulmonary vein flow (PVF) pattern in man during thoracotomy was described in previous study (Skagseth, 1975). A constant a n d inverse relationship between the profile of left atrial pressure (LAP) and PVF pattern w a s found (Fig. l), making a biphasic pattern with the predominating forward flow occurring in ventricular systole (S-wave). Morkin, Collins, Goldman & Fishman (1965), Morgan, Abel, Mullins & Guntheroth (1966) and Dixon, Stanton & Morrow (1971) found similar PVF characteristics in animal experiments, which closely resembled the flow patterns obtained from superior vena c a v a in man during thoracotomp (Fmysaker,
1972). T h e present study concerns the PVF pattern before and after commissurotomy in patients with fiiitral stenosis with o r without a n insignificant r'e-
gurgitation. T h e author gave a preliminary report o n the P V F pattern before and after mitral valve surgery in 1973. Book (1975) described the flow pattern in a single patient with pure mitral stenosis before and after prosthetic valve replacement. A further a i m of t h e study w a s to develop a more reliable intra-operative method for quantitation of mitral regurgitation after comrnisurotomy than the surgeon's finger.
M A T E R I A L AND M E T H O D S This study was performed on 15 patients with rheumatic mitral stenosis. Left ventricular angiography revealed no regurgitation of contrast medium to the left atrium in 7 patients, while in the others a small leakage was observed without opacification of the atrium. This was the smallest degree of regurgitation according to the grading described by Eie, Semb & Efskind (1972). The age of the patients, diagnosis, and operative procedures are listed in Table I. Patients 1 1 and 12 had an additional lesion of the aortic valves, necessitating prosthetic valve replacement in No. 12. Three patients (Nos. 3, 4 and 8) had mitral restenosis. Valve replacement was performed in cases 4 and 8, whereas another commissurotomy was performed in case No. 3 . The pre-operative evaluation included right and left heart catheterization, and the data are presented in Table 11. Seven patients had sinus rhythm, while the others had atrial fibrillation. Mean pulmonary capillary wedge pressure (PCW) was only I I mmHg in patient No. 13, but increased significantly after exercise, as in the other patients. Mean PCW at rest exceeded 20 mmHg in 9 of the patients. Only 2 patients (Nos. 11 and 13) had a v-wave of less than 20 mmHg. The mean diastolic mitral valve gradient was less than 10 mmHg in4 patients, 10-14 mmHg in 2, and exceeded 15 mmHg in 9 patients. Mean pulmonary arterial pressure was more than 50 mmHg in patients Nos. 2, 6 and 7, in whom pulmonary arterial resistance exceeded 500 d y n . s e ~ - ~Cardiac . index H B S below 2.5 l/min/m2in 7 patients (mean 2.7 l/min/m2). The 8 patients undergoing mitral commissurotomy were operated upon through a left thoracotomy, uhile the S c a d J Thor Cardiovusc Surg 10
54
E . Skagseth
ECG
PAF Scand Cardiovasc J Downloaded from informahealthcare.com by Osaka University on 02/12/15 For personal use only.
S
PVF
I
'
I
I
' A S ' VS
LAP Fig. 1 . Diagram of normal pulmonary vein flow (PVF) correlated to ECG, pulmonary arterial flow (PAF) and left atrial pressure (LAP). Inverse relation between PVF and LAP, and predominating S-wave are noted. AS=atrial systole. VS/VD=ventricular systole/diastole.
Fig. 2. Case 4. Heart rate ( H R ) 90lrnin. Atrial fibrillation (AF). Biphasic pattern recorded from superior pulmonary vein ( S P V ) is noted. Major forward flow occurs in VD. Early systolic reversed PVF of 1 1 % coincides with c-wave in LAP. Forward systolic PVF is delayed 0.10 sec.
Table I. Clinical data o n diagnosis and operative procedure MS/MRS=mitral stenosislrestenosis. MI=mitral incompetence. AS/AI=aortic stenosis/incompetence. AVR/MVR= aortic/mitral valve replacement. MC=mitral commissurotomy. , ++=slight, moderate stenosis/incompetence (see text).
+
Case No.
Age
Diagnosis
Operation
S.O. s. L. E. T. O.V. T.G. A.K. B.H. G. N . M . R . A . E. K. T. T. 11. H . A . J . 12. B. L . A.
37 51 56 44 55 46 52 53 53 49 56 41
MS MS MRS MRS MS MS MS MRS MS MS MS MS
13. G . B . 14. E . H . L . 15. 0. M . H .
37 47 54
MS MS MS
MC MC MC MVR Bjork-Shiley No. 29 MVR Bjork-Shiley No. 31 MVR Bjork-Shiley No. 29 MVR Lillehei-Kaster No. 22 MVR Bjork-Shiley No. 29 MC MC MC MVR Lillehei-Kaster No. 22 AVR Lillehei-Kaster No. 18 MVR Lillehei-Kaster No. 22 MC MC
1. 2. 3. 4. 5. 6. 7. 8. 9. 10.
Scand J Thor Cardiovasc Surg 10
MI+ MI+ MI+ MI+ MI+ MI+ MI+ AS+ AI++
MI+
Scand Cardiovasc J Downloaded from informahealthcare.com by Osaka University on 02/12/15 For personal use only.
56
E. Skugseth
Table 111. Intru-operutive pressures before mitrul w l v e surgery correluted to PVF patterri, reversed PVF ( R F ) expressed us N percentage of ,forward ,flow, and deluy o f fiwward systolic PVF ufter onset o f ventriculur systole +=present and forward flow. ++=predominating forward flow wave. -=reversed flow. Other symbols as in Fig. 1 and Table 11. LAP (mmHg)
Scand Cardiovasc J Downloaded from informahealthcare.com by Osaka University on 02/12/15 For personal use only.
Case No. 1 2 3 4 5 6 7 8 9 10 It 12 13 14 15
v-wave
M
33 42 40 45
28 35 33 32 13 30 50 26 30 27 29 21 30 34 27
16
38 60 34 43 37 40 26 37 41
38
PVF pattern AMV (mmHg) 10 25 25 20 8 18
30 18 15 18 14 12 18 20 20
S
-+
-+ -+ -+ ++ + -+ + -+ -+ -+ -+ + -+ -+
+ + + + + + + + + + + + (monphasic) + +
mitral valve replacements were performed through a median sternotomy using cardiopulmonary bypass. In the prosthetic valve group, the right superior pulmonary vein was dissected free for application of the flowprobe, while the flow pattern was recorded from the left superior vein in the commissurotomy group. The pulmonary veins were enlarged and distended without visible pulsatory changes of the vein diameter. Nycotron cuff type flow probes, 10-20 mm ID, were used. A close vessel-to-probe contact was easily obtained in all the patients. Because of the shortness of the accessible part of the pulmonary veins, mechanical zero flow was obtained only in 2 patients without impairing the veinlprobe contact. The mechanical zero flow signal corresponded well with the baseline obtained with the probe submerged in saline. Provided that a close-fitting probe was used, a good agreement between in vivo and in vitro baselines has previously been found (Fraysaker, 1972; Skagseth, 1975). The flow measurements were performed with a Nycotron flowmeter model 372 with simultaneous recordings of ECG, LAP and left ventricular pressure (LVP), using the same technique as previously described (Skagseth, 1975). Pulmonary arterial pressure (PAP) and pulmonary arterial flow (PAF) were also recorded in a few patients. On completion of the intracardiac procedure and achievement of a stable haemodynamic state with approximately the same central venous pressure as at the first recordings, the measurements were repeated in all the patients. In case No. I , however, technical failure deprived us of complete measurements. No complications occurred or could be referred to the flowlpressure recordings. Reversed PVF (RF) was calculated by planimetry of flow patterns from 5 consecutive head heats and is preScand J Thor Cardiovasc Surg I0
D
V
+ ++ ++ ++ + ++ ++ ++ ++ ++ ++ ++ + ++ ++
RF 7%
7 6 4 11 0 0
7 0 13 13 4 3 0 4 6
Forward systolic PVF delay (sec)
0.04 0.04 0.06 0.10 0 0 0.10 0 0.08 0.08 0.06 0.08 0.07 0.10 0.10
sented as a percentage of forward flow. The accuracy of RF is highly dependent on the baseline, as a very small deviation from zero will give a relatively large error in the estimated RF. In the search for a more suitable parameter, delay in forward systolic PVF, defined as the time in sec from commencement of mechanichal ventricular systole to onset of forward systolic PVF, was calculated. The PVF patterns recorded after mitral valve replacements will be discussed in a later paper.
RESULTS The intra-operative recordings of flow, pressures and calculated data are presented in Table 111. Because of the unstable situation during the operation with positive pressure ventilation, there are some differences in the recorded pre-operative and intraoperative data. However, both Tables I1 and 111 show that patients Nos. 1 and 5 had moderate mitral stenosis with diastolic gradients of less than 10 mmHg. The maximum difference in gradient was 7 mmHg (patient No. 13). Except for patient No. 3, who had nodal rhythm during the recordings, all the patients retained the same heart rhythm as preoperatively (Table 11). The recordings of PVF pattern in mitral stenosis reveal some main characteristics (Figs. 2-9). The predominating forward PVF occurs in ventricular diastole in 13 out of 15 patients (Table III), in con-
57
Scand Cardiovasc J Downloaded from informahealthcare.com by Osaka University on 02/12/15 For personal use only.
Pulmonary vein.ftow in mitrul stenosis
Fig. 5. Case 9. HR 70/min. Sinus rhythm. Slightly increased regurgitation was palpated after limited splitting of posterior commissure, which reduced gradient from 15
to 12 mmHg. Reversed PVF during atrial systole vanes, but there is slightly increased reversal in VS. PVF delay is increased from 0.08 to 0.10 sec.
trast to the predominating S-wave in normal hearts. This seems to be independent of heart rhythm, as the D-wave was predominating in both sinus and nodal rhythm and atrial fibrillation. In patient No. 1, who had a mitral valve gradient of 10 mmHg, the D-wave was less predominating than in the other
patients with more severe stenosis. The PVF pattern in patient No. 5 (Fig. 3) had normal characteristics with a predominating S-wave. The normal flow pattern was not unexpected in this patient because of the slightly elevated LAP and a gradient of only 8 mmHg. Scand J Thor Cardiova.sc SurE 10
E . Skagseth
Scand Cardiovasc J Downloaded from informahealthcare.com by Osaka University on 02/12/15 For personal use only.
58
Fig. 6 . Case 14. HR 140/min before and 130/min after commissurotomy. MV gradient reduced from 20 t o 12 mmHg. RF increased from 4 to l o %, but PVF delay was unchanged 0.10 sec. PVF pattern is unchanged and LAP remains elevated.
The PVF pattern was biphasic in all the patients except No. 13. The V-wave always coincided with the v in LAP. The monophasic pattern in patient No. 13 (Fig. 4) may be explained by the constant c-v level in LAP associated with tachycardia. In early ventricular systole, PVF was reversed in 11 patients and a minimum forward flow was reached in the other 4. Maximum RF was 13%. Four of the 8 patients with a slight regurgitation at angiography had no retrograde PVF. In sinus rhythm, atrial systole may produce reversed PVF as seen in Fig. 5. Planimetry of the flow pattern in sinus rhythm may therefore give a higher RF than if the patient had atrial fibrillation. The delay in forward systolic PVF ranged from 0 to 0.10 sec (mean 0.06 sec). There was no increased PVF delay in those patients with a slight angiographic mitral regurgitation. RF and PVF delay corresponded to some extent (Table 111). PVF pattern after conzmissurotomy
Fig. 7. Case 3 . HR 118/min. Nodal rhythm. Before commissurotomy. Predominating forward flow occurs in VD. Small reversed flow in early systole coincides with c-wave. PAP is probably damped. LPA=left pulmonary artery. Scand J fhor Cardiovasc Surg 10
Technically successful recordings were obtained in 7 patients. The PVF patterns before and after commissurotomy are diagrammatically displayed in Table IV. In spite of splitting of the commissures, no decrease in diastolic mitral valve gradient was measured in patients Nos. 2 and 11. In patients Nos. 9, 14 and 15, a moderate reduction of the gradient was obtained, but the PVF patterns re-
Scand Cardiovasc J Downloaded from informahealthcare.com by Osaka University on 02/12/15 For personal use only.
Pulmonary w i n flow in mitral stenosis
59
Fig. 8. Case 3 . HR 114/min. Nodal rhythm. Splitting of anterior com-
missure caused rupture of a chorda tendineae, and large regurgitation was palpated. This is reflected in PVF which is reversed during entire VS. R F increased from 4 to 15%, and PVF dealy from 0.06 to 0.29 sec. PAP increased considerably, while LAP probably is damped after commisurotomy.
mained almost unchanged. The reduction of the gradient from 18 to less than 5 mmHg in patient 10 did not normalize the pattern as the D-wave remained predominating (Fig. 9). Left ventricular diastolic pressure increased considerably after commissurotomy in this patient and LAP remained at a high level, probably because of some degree of overtransfusion and left ventricular failure. In patient 3 (Figs. 7 and S), who had nodal rhythm during the recordings, rupture of a corda tendinae occurred during commissurotomy , giving the surgeon an impression of a large regurgitation. Left ventricular pressure recording was omitted. After commissurotomy , a holosystolic reversed flow was observed in PVF pattern, and RF increased from 4 to 15 %. Prosthetic valve replacement of the mitral valves was desirable, but the heart-lung machine was not available. The surgeon also felt increased regurgitation after commissurotomy in patients Nos. 9 and 14, which corresponded well to the increased RF and delay of forward PVF, while these parameters were unchanged or decreased in patients Nos. 2 , 10, 11, 15, in whom the surgeon estimated that the slight regurgitation before commissurotomy was unchanged afterwards. The influence of positive pressure ventilation on
the mean flow and flow pattern could be noted in only a few patients, and was considerably smaller than found in normal hearts (Abel & Waldhausen, 1969; Skagseth, 1975).
DISCUSSION The pulmonary vein flow pattern in mitral stenosis is studied to a limited extent. Efskind & Cappelen (1965) found that the functional condition of mitral valves was reflected in PVF pattern, and that flowmetry might be used in studies of this function. Dixon et al., who experimentally produced mitral regurgitation in dogs, demonstrated reversed PVF in early ventricular systole. Results from the present series correspond well to our preliminary results from 7 flow recordings in patients with mitral stenosis. The forward PVF was found biphasic in 14 of 15 patients with a small S-wave, and with a predominating V-wave in the 13 patients 1, in whom the mean diastolic gradient exceeded 10 mmHg. Book demonstrated similar PVF characteristics in a patient who had severe mitral stenosis. He also found early systolic reversed flow coinciding with “the first peak of the v-wave”, which indicates Scand J Thor C a r d i o v a ~ cS i i r , v 10
60
E. SkagJeth
Scand Cardiovasc J Downloaded from informahealthcare.com by Osaka University on 02/12/15 For personal use only.
I
Fig. 9. Case 10. HR 120/min before and 108/min after
commisurotomy. AF. In spite of reduction of gradient from 18 to less than 5 mmHg, D-wave is still predominating. R F decreased from 13 to 2 %, while PVF delay was unchanged 0.08 sec. LAP remained elevated.
closure of the mitral valves and by us named the c-wave. In less severe mitral stenosis, a relatively larger part of the forward flow occrirred during ventricular systole. In patient No. 5, who had a mitral gradient of only 8 mmHg, the S-wave was found predominating as in normal hearts. The V-wave present in the biphasic patterns coincided with v in LAP. The downward deflection of V was usually more marked in heart rates below 100/min than at higher rates. a s seen in Figs. 4 and 5. The prominence of v, calculated as the difference between the v and mean LAP, ranged between 3 and 13 mmHg. We found no correlation between this difference and degree of downward deflection of the V-wave. However. a prominent c , which more often was connected with slower heart rates, was usually foundincases with amarked V-wave (Fig. 5 ) . Neither the reversed PVF in early systole nor the delay in forward systolic P V F showed anp correlaS a n d J Thor Cardiovusc Surg 10
tion with a slight mitral regurgitation demonstrated at angiography, and the prominence of the v-wave in L A P was not increased in these patients. Therefore it seems likely that the small regurgitation, which causes no opacification of the left atrium, has an insignificant haemodynamic effect. Minimum forward o r maximum reversed PVF coincided with the peak of c, and the degree of R F seemed to have an inverse relation with the prominence of c . The impaired emptying of the left atrium in mitral stenosis causes a n elevated L A P in late ventricular diastole, depending of the severity of the stenosis, cardiac output and heart rate. Closure of the mitral valve in ventricular systole causes an additional increase in LAP, depending on the stiffness and ballooning of the mitral leaflets, producing a c-wave as high as the v in 5/15 patients. The maximum reversed PVF caused by the prominent c-wave was 13 % in our recordings, while Book recorded 19% R F in his patient with pure mitral stenosis. An accurate baseline is a prerequisite for calculating R F by planimetry if a significant error is to b e avoided. In spite of a good in v i t w and in vitrn zero flow correlation, a small baseline deviation may occur, leading to erroneous calculation of RF. The reversed flow, which is sometimes caused by atrial systole in sinus and nodal rhythm, will confluence with the early systolic reversal, and comparison of R F in those rhythms with R F in atrial fibrillation may be misleading. Therefore we have found it advantageous to calculate the delay in systolic forward PVF, which is a more convenient and simple parameter for clinical use, and eliminates the problems associated with different heart rhythms. As the length of ventricular systole varies only slighly with different heart rates, we have not found it opportune to present the delay in percentage of the systolic period. The delay did not exceed 0.10 sec in pure mitral stenosis, in contrast to the increased delay in severe mitral regurgitation (Fig. 8), which will be discussed in a later paper. The small number of recordings after commissurotomy do not permit definite conclusions to be drawn, particularly because the mitral valve gradient was significantly reduced only in one patient. The PVF pattern remained pathological in all patients, even after a successfiil commissurotomy as in patient No. 10. Normalization of LAP, however, did not occur in this patient. In normal hearts, w e found the S-wave predominating, irrespective of L A P up to a mean pressure of 16 nimHg. In canine
Pulmonary vein flow in mitral stenosis
61
Table IV. Diagram of PVF pattern before (unbroken line) and ajter mitral valve commissurotomy (dotted line) correlated to left atrial pressure and mean diastolic niitral valve grctdient Symbols as in Fig. 1 and Table I1 Before
PVF pattern VS VD
LAP (mmHg)
Scand Cardiovasc J Downloaded from informahealthcare.com by Osaka University on 02/12/15 For personal use only.
Case No.
v-wave
M
AMV (mmHg)
9
43
30
15
10
31
21
11
40
14 15
After LAP (mmHg)
M
AMV (mmHg)
38
26
12
18
28
21
29
14
30
23
14
41
34
20
36
30
12
38
27
20
58
30
15
-
AS
v-wave
bA
experiments, Dixon et al. showed that increasing mean L A P above 10 mmHg induced a relatively larger P V F in ventricular diastole, but the S-wave was still predominating and the flow was forward throughout the heart cycle. A further increased L A P in mitral disease will usually be associated with left ventricular failure with increased diastolic filling pressure, causing a resistance to the emptying of the left atrium and leaving a n elevated presystolic LAP. Even a small residual mitral valve gradient tends to accentuate this elevation, and closure of the mitral valves causes an additional increase (Fig. 9). The gradient between the pulmonary capillary bed and the left atrium is therefore reduced with a small S-wave as a consequence. The pulmonary veins serve not only as a conduit, but also as a reservoir for blood, and may be considered an extension of the left atrium. The reservoir function seems relatively larger in the congested lungs in mitral valve stenosis, and the compliance of the veins undoubtedly plays a part in modulating the flow pattern. Peroperatively it was observed that the left atrium was markedly enlarged with a thick unyielding wall with small contractions, even in
PULMONARY VEIN FLOW PATTERN IN MITRAL STENOSIS BEFORE AND AFTER COMMISSUROTOMY Erling Skagseth From Surgical Depariment A , Rikshospitalet (University Hospital), Oslo, Norway
Scand Cardiovasc J Downloaded from informahealthcare.com by Osaka University on 02/12/15 For personal use only.
(Submitted for publication October 28, 1974)
Abstract. Pulmonary vein flow (PVF) pattern was studied with an electromagnetic flowmeter in 15 patients with mitral stenosis. ECG, left atrial and left ventricular pressure were recorded simultaneously. Commissurotomy was performed in 8 patients and prosthetic valve replacement in the others. Measurements were performed before and after the intracardiac procedure. The predominating forward flow occurred in ventricular diastole, in contrast to ventricular systole in normal hearts. A small retrograde PVF was observed in 11/15 patients in early ventricular systole, coinciding with the c-wave in left atrial pressure (LAP). The reversed flow did not exceed 13% and showed no correlation to a slight regurgitation demonstrated at left ventricular angiography. The maximum delay in forward systolic PVF after onset of ventricular systole was 0.10 sec. The PVF pattern remained unchanged after commissurotomy in the majority of patients, even after reduction of the mitral valve gradient to less than 5 mmHg. However, holosystolic reversed PVF occurred in one patient, indicating a severe mitral regurgitation. The method of PVF recording is simple and proved useful for haemodynamic analysis and as a control after mitral valve surgery for evaluating the degree of mitral regurgitation.
The normal pulmonary vein flow (PVF) pattern in man during thoracotomy was described in previous study (Skagseth, 1975). A constant a n d inverse relationship between the profile of left atrial pressure (LAP) and PVF pattern w a s found (Fig. l), making a biphasic pattern with the predominating forward flow occurring in ventricular systole (S-wave). Morkin, Collins, Goldman & Fishman (1965), Morgan, Abel, Mullins & Guntheroth (1966) and Dixon, Stanton & Morrow (1971) found similar PVF characteristics in animal experiments, which closely resembled the flow patterns obtained from superior vena c a v a in man during thoracotomp (Fmysaker,
1972). T h e present study concerns the PVF pattern before and after commissurotomy in patients with fiiitral stenosis with o r without a n insignificant r'e-
gurgitation. T h e author gave a preliminary report o n the P V F pattern before and after mitral valve surgery in 1973. Book (1975) described the flow pattern in a single patient with pure mitral stenosis before and after prosthetic valve replacement. A further a i m of t h e study w a s to develop a more reliable intra-operative method for quantitation of mitral regurgitation after comrnisurotomy than the surgeon's finger.
M A T E R I A L AND M E T H O D S This study was performed on 15 patients with rheumatic mitral stenosis. Left ventricular angiography revealed no regurgitation of contrast medium to the left atrium in 7 patients, while in the others a small leakage was observed without opacification of the atrium. This was the smallest degree of regurgitation according to the grading described by Eie, Semb & Efskind (1972). The age of the patients, diagnosis, and operative procedures are listed in Table I. Patients 1 1 and 12 had an additional lesion of the aortic valves, necessitating prosthetic valve replacement in No. 12. Three patients (Nos. 3, 4 and 8) had mitral restenosis. Valve replacement was performed in cases 4 and 8, whereas another commissurotomy was performed in case No. 3 . The pre-operative evaluation included right and left heart catheterization, and the data are presented in Table 11. Seven patients had sinus rhythm, while the others had atrial fibrillation. Mean pulmonary capillary wedge pressure (PCW) was only I I mmHg in patient No. 13, but increased significantly after exercise, as in the other patients. Mean PCW at rest exceeded 20 mmHg in 9 of the patients. Only 2 patients (Nos. 11 and 13) had a v-wave of less than 20 mmHg. The mean diastolic mitral valve gradient was less than 10 mmHg in4 patients, 10-14 mmHg in 2, and exceeded 15 mmHg in 9 patients. Mean pulmonary arterial pressure was more than 50 mmHg in patients Nos. 2, 6 and 7, in whom pulmonary arterial resistance exceeded 500 d y n . s e ~ - ~Cardiac . index H B S below 2.5 l/min/m2in 7 patients (mean 2.7 l/min/m2). The 8 patients undergoing mitral commissurotomy were operated upon through a left thoracotomy, uhile the S c a d J Thor Cardiovusc Surg 10
54
E . Skagseth
ECG
PAF Scand Cardiovasc J Downloaded from informahealthcare.com by Osaka University on 02/12/15 For personal use only.
S
PVF
I
'
I
I
' A S ' VS
LAP Fig. 1 . Diagram of normal pulmonary vein flow (PVF) correlated to ECG, pulmonary arterial flow (PAF) and left atrial pressure (LAP). Inverse relation between PVF and LAP, and predominating S-wave are noted. AS=atrial systole. VS/VD=ventricular systole/diastole.
Fig. 2. Case 4. Heart rate ( H R ) 90lrnin. Atrial fibrillation (AF). Biphasic pattern recorded from superior pulmonary vein ( S P V ) is noted. Major forward flow occurs in VD. Early systolic reversed PVF of 1 1 % coincides with c-wave in LAP. Forward systolic PVF is delayed 0.10 sec.
Table I. Clinical data o n diagnosis and operative procedure MS/MRS=mitral stenosislrestenosis. MI=mitral incompetence. AS/AI=aortic stenosis/incompetence. AVR/MVR= aortic/mitral valve replacement. MC=mitral commissurotomy. , ++=slight, moderate stenosis/incompetence (see text).
+
Case No.
Age
Diagnosis
Operation
S.O. s. L. E. T. O.V. T.G. A.K. B.H. G. N . M . R . A . E. K. T. T. 11. H . A . J . 12. B. L . A.
37 51 56 44 55 46 52 53 53 49 56 41
MS MS MRS MRS MS MS MS MRS MS MS MS MS
13. G . B . 14. E . H . L . 15. 0. M . H .
37 47 54
MS MS MS
MC MC MC MVR Bjork-Shiley No. 29 MVR Bjork-Shiley No. 31 MVR Bjork-Shiley No. 29 MVR Lillehei-Kaster No. 22 MVR Bjork-Shiley No. 29 MC MC MC MVR Lillehei-Kaster No. 22 AVR Lillehei-Kaster No. 18 MVR Lillehei-Kaster No. 22 MC MC
1. 2. 3. 4. 5. 6. 7. 8. 9. 10.
Scand J Thor Cardiovasc Surg 10
MI+ MI+ MI+ MI+ MI+ MI+ MI+ AS+ AI++
MI+
Scand Cardiovasc J Downloaded from informahealthcare.com by Osaka University on 02/12/15 For personal use only.
56
E. Skugseth
Table 111. Intru-operutive pressures before mitrul w l v e surgery correluted to PVF patterri, reversed PVF ( R F ) expressed us N percentage of ,forward ,flow, and deluy o f fiwward systolic PVF ufter onset o f ventriculur systole +=present and forward flow. ++=predominating forward flow wave. -=reversed flow. Other symbols as in Fig. 1 and Table 11. LAP (mmHg)
Scand Cardiovasc J Downloaded from informahealthcare.com by Osaka University on 02/12/15 For personal use only.
Case No. 1 2 3 4 5 6 7 8 9 10 It 12 13 14 15
v-wave
M
33 42 40 45
28 35 33 32 13 30 50 26 30 27 29 21 30 34 27
16
38 60 34 43 37 40 26 37 41
38
PVF pattern AMV (mmHg) 10 25 25 20 8 18
30 18 15 18 14 12 18 20 20
S
-+
-+ -+ -+ ++ + -+ + -+ -+ -+ -+ + -+ -+
+ + + + + + + + + + + + (monphasic) + +
mitral valve replacements were performed through a median sternotomy using cardiopulmonary bypass. In the prosthetic valve group, the right superior pulmonary vein was dissected free for application of the flowprobe, while the flow pattern was recorded from the left superior vein in the commissurotomy group. The pulmonary veins were enlarged and distended without visible pulsatory changes of the vein diameter. Nycotron cuff type flow probes, 10-20 mm ID, were used. A close vessel-to-probe contact was easily obtained in all the patients. Because of the shortness of the accessible part of the pulmonary veins, mechanical zero flow was obtained only in 2 patients without impairing the veinlprobe contact. The mechanical zero flow signal corresponded well with the baseline obtained with the probe submerged in saline. Provided that a close-fitting probe was used, a good agreement between in vivo and in vitro baselines has previously been found (Fraysaker, 1972; Skagseth, 1975). The flow measurements were performed with a Nycotron flowmeter model 372 with simultaneous recordings of ECG, LAP and left ventricular pressure (LVP), using the same technique as previously described (Skagseth, 1975). Pulmonary arterial pressure (PAP) and pulmonary arterial flow (PAF) were also recorded in a few patients. On completion of the intracardiac procedure and achievement of a stable haemodynamic state with approximately the same central venous pressure as at the first recordings, the measurements were repeated in all the patients. In case No. I , however, technical failure deprived us of complete measurements. No complications occurred or could be referred to the flowlpressure recordings. Reversed PVF (RF) was calculated by planimetry of flow patterns from 5 consecutive head heats and is preScand J Thor Cardiovasc Surg I0
D
V
+ ++ ++ ++ + ++ ++ ++ ++ ++ ++ ++ + ++ ++
RF 7%
7 6 4 11 0 0
7 0 13 13 4 3 0 4 6
Forward systolic PVF delay (sec)
0.04 0.04 0.06 0.10 0 0 0.10 0 0.08 0.08 0.06 0.08 0.07 0.10 0.10
sented as a percentage of forward flow. The accuracy of RF is highly dependent on the baseline, as a very small deviation from zero will give a relatively large error in the estimated RF. In the search for a more suitable parameter, delay in forward systolic PVF, defined as the time in sec from commencement of mechanichal ventricular systole to onset of forward systolic PVF, was calculated. The PVF patterns recorded after mitral valve replacements will be discussed in a later paper.
RESULTS The intra-operative recordings of flow, pressures and calculated data are presented in Table 111. Because of the unstable situation during the operation with positive pressure ventilation, there are some differences in the recorded pre-operative and intraoperative data. However, both Tables I1 and 111 show that patients Nos. 1 and 5 had moderate mitral stenosis with diastolic gradients of less than 10 mmHg. The maximum difference in gradient was 7 mmHg (patient No. 13). Except for patient No. 3, who had nodal rhythm during the recordings, all the patients retained the same heart rhythm as preoperatively (Table 11). The recordings of PVF pattern in mitral stenosis reveal some main characteristics (Figs. 2-9). The predominating forward PVF occurs in ventricular diastole in 13 out of 15 patients (Table III), in con-
57
Scand Cardiovasc J Downloaded from informahealthcare.com by Osaka University on 02/12/15 For personal use only.
Pulmonary vein.ftow in mitrul stenosis
Fig. 5. Case 9. HR 70/min. Sinus rhythm. Slightly increased regurgitation was palpated after limited splitting of posterior commissure, which reduced gradient from 15
to 12 mmHg. Reversed PVF during atrial systole vanes, but there is slightly increased reversal in VS. PVF delay is increased from 0.08 to 0.10 sec.
trast to the predominating S-wave in normal hearts. This seems to be independent of heart rhythm, as the D-wave was predominating in both sinus and nodal rhythm and atrial fibrillation. In patient No. 1, who had a mitral valve gradient of 10 mmHg, the D-wave was less predominating than in the other
patients with more severe stenosis. The PVF pattern in patient No. 5 (Fig. 3) had normal characteristics with a predominating S-wave. The normal flow pattern was not unexpected in this patient because of the slightly elevated LAP and a gradient of only 8 mmHg. Scand J Thor Cardiova.sc SurE 10
E . Skagseth
Scand Cardiovasc J Downloaded from informahealthcare.com by Osaka University on 02/12/15 For personal use only.
58
Fig. 6 . Case 14. HR 140/min before and 130/min after commissurotomy. MV gradient reduced from 20 t o 12 mmHg. RF increased from 4 to l o %, but PVF delay was unchanged 0.10 sec. PVF pattern is unchanged and LAP remains elevated.
The PVF pattern was biphasic in all the patients except No. 13. The V-wave always coincided with the v in LAP. The monophasic pattern in patient No. 13 (Fig. 4) may be explained by the constant c-v level in LAP associated with tachycardia. In early ventricular systole, PVF was reversed in 11 patients and a minimum forward flow was reached in the other 4. Maximum RF was 13%. Four of the 8 patients with a slight regurgitation at angiography had no retrograde PVF. In sinus rhythm, atrial systole may produce reversed PVF as seen in Fig. 5. Planimetry of the flow pattern in sinus rhythm may therefore give a higher RF than if the patient had atrial fibrillation. The delay in forward systolic PVF ranged from 0 to 0.10 sec (mean 0.06 sec). There was no increased PVF delay in those patients with a slight angiographic mitral regurgitation. RF and PVF delay corresponded to some extent (Table 111). PVF pattern after conzmissurotomy
Fig. 7. Case 3 . HR 118/min. Nodal rhythm. Before commissurotomy. Predominating forward flow occurs in VD. Small reversed flow in early systole coincides with c-wave. PAP is probably damped. LPA=left pulmonary artery. Scand J fhor Cardiovasc Surg 10
Technically successful recordings were obtained in 7 patients. The PVF patterns before and after commissurotomy are diagrammatically displayed in Table IV. In spite of splitting of the commissures, no decrease in diastolic mitral valve gradient was measured in patients Nos. 2 and 11. In patients Nos. 9, 14 and 15, a moderate reduction of the gradient was obtained, but the PVF patterns re-
Scand Cardiovasc J Downloaded from informahealthcare.com by Osaka University on 02/12/15 For personal use only.
Pulmonary w i n flow in mitral stenosis
59
Fig. 8. Case 3 . HR 114/min. Nodal rhythm. Splitting of anterior com-
missure caused rupture of a chorda tendineae, and large regurgitation was palpated. This is reflected in PVF which is reversed during entire VS. R F increased from 4 to 15%, and PVF dealy from 0.06 to 0.29 sec. PAP increased considerably, while LAP probably is damped after commisurotomy.
mained almost unchanged. The reduction of the gradient from 18 to less than 5 mmHg in patient 10 did not normalize the pattern as the D-wave remained predominating (Fig. 9). Left ventricular diastolic pressure increased considerably after commissurotomy in this patient and LAP remained at a high level, probably because of some degree of overtransfusion and left ventricular failure. In patient 3 (Figs. 7 and S), who had nodal rhythm during the recordings, rupture of a corda tendinae occurred during commissurotomy , giving the surgeon an impression of a large regurgitation. Left ventricular pressure recording was omitted. After commissurotomy , a holosystolic reversed flow was observed in PVF pattern, and RF increased from 4 to 15 %. Prosthetic valve replacement of the mitral valves was desirable, but the heart-lung machine was not available. The surgeon also felt increased regurgitation after commissurotomy in patients Nos. 9 and 14, which corresponded well to the increased RF and delay of forward PVF, while these parameters were unchanged or decreased in patients Nos. 2 , 10, 11, 15, in whom the surgeon estimated that the slight regurgitation before commissurotomy was unchanged afterwards. The influence of positive pressure ventilation on
the mean flow and flow pattern could be noted in only a few patients, and was considerably smaller than found in normal hearts (Abel & Waldhausen, 1969; Skagseth, 1975).
DISCUSSION The pulmonary vein flow pattern in mitral stenosis is studied to a limited extent. Efskind & Cappelen (1965) found that the functional condition of mitral valves was reflected in PVF pattern, and that flowmetry might be used in studies of this function. Dixon et al., who experimentally produced mitral regurgitation in dogs, demonstrated reversed PVF in early ventricular systole. Results from the present series correspond well to our preliminary results from 7 flow recordings in patients with mitral stenosis. The forward PVF was found biphasic in 14 of 15 patients with a small S-wave, and with a predominating V-wave in the 13 patients 1, in whom the mean diastolic gradient exceeded 10 mmHg. Book demonstrated similar PVF characteristics in a patient who had severe mitral stenosis. He also found early systolic reversed flow coinciding with “the first peak of the v-wave”, which indicates Scand J Thor C a r d i o v a ~ cS i i r , v 10
60
E. SkagJeth
Scand Cardiovasc J Downloaded from informahealthcare.com by Osaka University on 02/12/15 For personal use only.
I
Fig. 9. Case 10. HR 120/min before and 108/min after
commisurotomy. AF. In spite of reduction of gradient from 18 to less than 5 mmHg, D-wave is still predominating. R F decreased from 13 to 2 %, while PVF delay was unchanged 0.08 sec. LAP remained elevated.
closure of the mitral valves and by us named the c-wave. In less severe mitral stenosis, a relatively larger part of the forward flow occrirred during ventricular systole. In patient No. 5, who had a mitral gradient of only 8 mmHg, the S-wave was found predominating as in normal hearts. The V-wave present in the biphasic patterns coincided with v in LAP. The downward deflection of V was usually more marked in heart rates below 100/min than at higher rates. a s seen in Figs. 4 and 5. The prominence of v, calculated as the difference between the v and mean LAP, ranged between 3 and 13 mmHg. We found no correlation between this difference and degree of downward deflection of the V-wave. However. a prominent c , which more often was connected with slower heart rates, was usually foundincases with amarked V-wave (Fig. 5 ) . Neither the reversed PVF in early systole nor the delay in forward systolic P V F showed anp correlaS a n d J Thor Cardiovusc Surg 10
tion with a slight mitral regurgitation demonstrated at angiography, and the prominence of the v-wave in L A P was not increased in these patients. Therefore it seems likely that the small regurgitation, which causes no opacification of the left atrium, has an insignificant haemodynamic effect. Minimum forward o r maximum reversed PVF coincided with the peak of c, and the degree of R F seemed to have an inverse relation with the prominence of c . The impaired emptying of the left atrium in mitral stenosis causes a n elevated L A P in late ventricular diastole, depending of the severity of the stenosis, cardiac output and heart rate. Closure of the mitral valve in ventricular systole causes an additional increase in LAP, depending on the stiffness and ballooning of the mitral leaflets, producing a c-wave as high as the v in 5/15 patients. The maximum reversed PVF caused by the prominent c-wave was 13 % in our recordings, while Book recorded 19% R F in his patient with pure mitral stenosis. An accurate baseline is a prerequisite for calculating R F by planimetry if a significant error is to b e avoided. In spite of a good in v i t w and in vitrn zero flow correlation, a small baseline deviation may occur, leading to erroneous calculation of RF. The reversed flow, which is sometimes caused by atrial systole in sinus and nodal rhythm, will confluence with the early systolic reversal, and comparison of R F in those rhythms with R F in atrial fibrillation may be misleading. Therefore we have found it advantageous to calculate the delay in systolic forward PVF, which is a more convenient and simple parameter for clinical use, and eliminates the problems associated with different heart rhythms. As the length of ventricular systole varies only slighly with different heart rates, we have not found it opportune to present the delay in percentage of the systolic period. The delay did not exceed 0.10 sec in pure mitral stenosis, in contrast to the increased delay in severe mitral regurgitation (Fig. 8), which will be discussed in a later paper. The small number of recordings after commissurotomy do not permit definite conclusions to be drawn, particularly because the mitral valve gradient was significantly reduced only in one patient. The PVF pattern remained pathological in all patients, even after a successfiil commissurotomy as in patient No. 10. Normalization of LAP, however, did not occur in this patient. In normal hearts, w e found the S-wave predominating, irrespective of L A P up to a mean pressure of 16 nimHg. In canine
Pulmonary vein flow in mitral stenosis
61
Table IV. Diagram of PVF pattern before (unbroken line) and ajter mitral valve commissurotomy (dotted line) correlated to left atrial pressure and mean diastolic niitral valve grctdient Symbols as in Fig. 1 and Table I1 Before
PVF pattern VS VD
LAP (mmHg)
Scand Cardiovasc J Downloaded from informahealthcare.com by Osaka University on 02/12/15 For personal use only.
Case No.
v-wave
M
AMV (mmHg)
9
43
30
15
10
31
21
11
40
14 15
After LAP (mmHg)
M
AMV (mmHg)
38
26
12
18
28
21
29
14
30
23
14
41
34
20
36
30
12
38
27
20
58
30
15
-
AS
v-wave
bA
experiments, Dixon et al. showed that increasing mean L A P above 10 mmHg induced a relatively larger P V F in ventricular diastole, but the S-wave was still predominating and the flow was forward throughout the heart cycle. A further increased L A P in mitral disease will usually be associated with left ventricular failure with increased diastolic filling pressure, causing a resistance to the emptying of the left atrium and leaving a n elevated presystolic LAP. Even a small residual mitral valve gradient tends to accentuate this elevation, and closure of the mitral valves causes an additional increase (Fig. 9). The gradient between the pulmonary capillary bed and the left atrium is therefore reduced with a small S-wave as a consequence. The pulmonary veins serve not only as a conduit, but also as a reservoir for blood, and may be considered an extension of the left atrium. The reservoir function seems relatively larger in the congested lungs in mitral valve stenosis, and the compliance of the veins undoubtedly plays a part in modulating the flow pattern. Peroperatively it was observed that the left atrium was markedly enlarged with a thick unyielding wall with small contractions, even in
