Nine patients were included in the study (67% females) with an age of 33.1±13.9 years (range 15 – 59). All patients were assessed, cannulated, and retrieved on ECMO from other hospitals in Sweden and northern Europe by our mobile ECMO service. The median mechanical ventilation time at the start of ECMO was 2 days (range 1 – 14 days). Mean VIS was 43±37. Additional data at admission is available in Table 1. Seven patients were initiated on venoarterial (VA) ECMO, one on veno-venoarterial (VVA), and one was commenced on venovenous (VV) ECMO (Table 2). From here on VVA mode will be included in the VA group.
Table 2
Modes during extracorporeal support
|
No.
|
T0
|
T25
|
T50
|
T75
|
T100
|
Outcome
|
1
|
V-VA
|
V-VA
|
V-VA
|
V-VA
|
VV
|
Survived
|
2
|
VA
|
VA
|
VA
|
VA
|
VA
|
†; L-Tx, cerebral bleeding one week after transplant
|
3
|
VA
|
VV
|
VV
|
VV
|
VV
|
L-Tx (and subsequent renal Tx)
|
4
|
VA
|
VA
|
VA
|
VA
|
VA
|
†; L-Tx, died from intrathoracic bleeding 1st postop
|
5
|
VA
|
VA
|
VA
|
VA
|
VA
|
L-Tx
|
6
|
VV
|
VV
|
VV
|
VA
|
VA
|
†;
|
7
|
VA
|
VV
|
VV
|
VV
|
VV
|
Survived
|
8
|
VA
|
VA
|
VA
|
VA
|
VA
|
†; ECMO withdrawn due to futility
|
9
|
VA
|
VA
|
VA
|
VA
|
VV
|
Survived
|
Abbreviations: VV, venovenous; V-VA, veno-venoarterial; VA, venoarterial; †, deceased; L-Tx, lung transplanted from extracorporeal membrane oxygenation
Table 3
Comparison between survivors and deceased prolonged ECMO at T0, at T25 and at T100
|
|
Alive
|
Deceased
|
p-value
|
Male sex, n(%)
|
2(40)
|
1(25)
|
1.0
|
Age, years
|
31±18
|
35±9
|
0.71
|
T25
|
79±25
|
53±16
|
0.11
|
T100
|
72±18
|
51±17
|
0.11
|
Percent weight loss since Day 1 (T0), T25
|
8.7±16
|
13±14
|
0.80
|
T100
|
16±14
|
15±12
|
0.89
|
Body mass index, kg/m2, T0
|
28±6
|
24±3
|
0.26
|
Days from intubation to ECMO
|
1(1.0-12)
|
5(3.2-14)
|
0.73
|
Days at 25% of ECMO treatment
|
19(12-38)
|
17(9-58)
|
0.83
|
Days to 25% echocardiography
|
19(14-36)
|
17(10-56)
|
0.83
|
Days of ECMO
|
75(43-154)
|
65(34-232)
|
0.80
|
Days to last echocardiography
|
54(36-104)
|
50(30-219)
|
0.57
|
Time% to last echocardiography of total time on ECMO
|
71(69-81)
|
88(77-97)
|
0.10
|
Mean arterial blood pressure, mmHg
|
65±9
|
74±11
|
0.24
|
FiO2
|
1.0(1.0-1.0
|
1.0(1.0-1.0)
|
1.0
|
P/F ratio, mmHg
|
51±17
|
44±12
|
0.50
|
SAPS 3 score
|
56±9
|
82±12
|
0.006
|
RESP score (VV ECMO)
|
5±1.4 (n=2)
|
-3 (n=1)
|
n/a
|
SAVE score (VA ECMO)
|
-3.7±2.9 (n=3)
|
-8.3±6.5 (n=3)
|
0.32
|
VA ECMO, n(%), T0
|
5(100)
|
3(75)
|
0.44
|
T25
|
3(60)
|
3(75)
|
1.0
|
T100
|
1(20)
|
4(100)
|
0.048
|
Number of conversions
|
3
|
1
|
0.52
|
p-Lactate, mmol/L, T0
|
4.0(2.4-6.4)
|
2.4(1.0-11)
|
0.85
|
T25
|
1.9(1.0-2.6)
|
1.0(0.7-2.0)
|
0.32
|
T100
|
0.9(0.8-1.0)
|
0.95(0.4-2.4)
|
0.51
|
SOFA score, T0
|
11.6±0.9
|
14.3±4.1
|
0.20
|
T25
|
13.2±3.0
|
12.5±3.7
|
0.76
|
T100
|
10.0±3.8
|
14±2.6
|
0.12
|
VIS, T0
|
31±21
|
57±45
|
0.30
|
T25
|
14±13
|
7.2±6.6
|
0.37
|
T100
|
2.8±4.1
|
7.5±5.7
|
0.19
|
Hemoglobin, g/L, T0
|
126±11
|
123±19
|
0.72
|
T25
|
116±4.4
|
114±8.9
|
0.78
|
T100
|
101±12
|
109±9.1
|
0.29
|
White blood cell count, 10^9/L, T0
|
8.0(1.3-11.6)
|
13.3(4.2-14.2)
|
0.35
|
T25
|
21.8(16.8-36.9)
|
8.4(7.2-16.6)
|
0.06
|
T100
|
16.7(19.6-22.8)
|
9.6(6.2-15.4)
|
0.15
|
p-C-reactive protein, T0
|
202±155
|
263±52
|
0.48
|
T25
|
174±114
|
148±58
|
0.69
|
T100
|
90±65
|
103±23
|
0.70
|
p-Procalcitonine, T0
|
78(1.1-306)
|
9.6(1.1-1583)
|
0.43
|
T25
|
2.0(1.2-6.2)
|
1.2(0.43-5.9)
|
0.68
|
T100
|
0.51(0.43-0.78)
|
0.70(0.34-1.0)
|
0.65
|
p-Bilirubin, T0
|
21(17-68)
|
23(16-28)
|
0.37
|
T25
|
50(25-80)
|
24(16-37)
|
0.20
|
T100
|
30(12-55)
|
37(33-69)
|
0.41
|
p-Creatinine, T0
|
149±46
|
159±109
|
0.85
|
T25
|
108±57
|
84±17
|
0.44
|
T100
|
63±29
|
94±45
|
0.25
|
Ongoing CRRT n, (%), T0
|
0, (0)
|
0, (0)
|
1.0
|
T25
|
4, (80)
|
4, (100)
|
1.0
|
T100
|
2, (40)
|
4, (100)
|
0.17
|
paO2, kPa, T0
|
6.2(5.2-9.8)
|
4.8(4.3-6.2)
|
0.15
|
T25
|
5.4(4.8-7.0)
|
5.0(4.5-5.9)
|
0.43
|
T100
|
5.6(5.3-6.4)
|
5.8(4.8-6.7)
|
0.97
|
paCO2, kPa, T0
|
6.3±1.6
|
10.8±4.4
|
0.07
|
T25
|
5.4±0.4
|
5.2±0.2
|
0.40
|
T100
|
5.9±0.5
|
5.6±0.5
|
0.40
|
SpreO2, %, T0
|
74(63-82)
|
68(49-80)
|
0.48
|
T25
|
65(61-86)
|
69(65-77)
|
0.86
|
T100
|
73(71-83)
|
70(70-81)
|
0.60
|
Abbreviations: SOFA, Sequential Organ Failure Assessment score; SAVE, Survival after Veno-Arterial ECMO score; RESP, Respiratory Extracorporeal Membrane Oxygenation Survival Prediction score; SAPS3, Simplified Acute Physiology Score 3; ALT, alanine aminotransferase; AST, aspartate aminotransferase; FiO2, fraction of inspired oxygen; P/F, ratio of arterial oxygen partial pressure to fractional inspired oxygen; CRRT, continuous renal replacement therapy; paO2, arterial oxygen partial pressure; paCO2, arterial carbon dioxide partial pressure; SpreO2, pre oxygenator saturation.
Basic data of survivors and non-survivors
There were no differences between survivors and non-survivors concerning age, days from intubation to ECMO (1(1.0-12) vs. 5(3.2-14) days, p=0.73), length of ECMO treatment (75(43-154) vs. 65(34-232) days, p=0.80), (Table 3). However, at admission there was a difference in SAPS 3 score between survivors (56±9) and non-survivors (82±12, p=0.006). The P/F ratio, SOFA and SAVE scores, plasma lactate, number of vasoactive agents, and VIS did not differ between groups. On group level there were no differences regarding biochemistry or blood gas parameters, nor were there any difference in these parameters between the groups over the course of treatment, (Table 3).
Over the course of treatment there was a decrease of VA ECMO support from 89% at T0 to 55% at T100. At this final point, 80% of the survivors were supported with VV ECMO whilst all non-survivors were on VA ECMO (p=0.048) (Table 2). There was no significant change over time in ECMO blood flow, nor was there any significant difference in ECMO flow between survivors and non-survivors (Table 4). The trajectory showed that the patients in both groups were more awake at later time points. The survivors experienced spontaneous breathing more frequently than the non-survivors. At the onset of ECMO treatment (T0), the tidal volumes differed between survivors (344±214 mL) and non-survivors (68±64 mL, p=0.043). Furthermore, there was no progression or change in tidal volumes within the respective groups. Hence, among survivors, tidal volumes were preserved at higher volumes over the whole course of ECMO treatment, in contrast to the deceased group (Table 4). Comparing progress of disease, there were improvements concerning infectious, renal, and circulatory status. However, the groups did not differ on any single parameter at any of the analyzed timepoints (T0, T25 and T100) (Table 3).
Computed tomography
Regarding respiratory progress using contrast enhanced CT scans, echocardiography, tidal volume and level of extracorporeal support, there were discrepancies between the clinical picture where tidal volumes tended to increase but this could not be independently detected by CT exams (Table 3). Concerning viable lung tissue, there was no difference between survivors (75±43%) and non-survivors (70±42%, p=0.91) at initiation of ECMO nor at T100 (72±20 vs. 53±32%, respectively, (p=0.34)).
Table 4
Trajectories of physiology during extracorporeal support
|
Item
|
Group
|
T0
|
T25
|
T50
|
T75
|
T100
|
p-value
|
Comments
|
Venoarterial support
|
Pooled, N=9
|
89
|
67
|
67
|
88
|
44
|
0.59
|
|
(%)
|
survivors, n=5
|
100
|
60
|
60
|
60
|
20
|
|
|
|
deceased, n=4
|
25
|
25
|
25
|
0
|
0
|
|
|
|
|
|
|
|
|
|
|
|
Richmond Agitation-
|
Pooled
|
-3(-3.5;-3)
|
-2(-2.6;-0.5)
|
-1(-3;0)
|
0(-1.5;0)
|
0(0;0)
|
0.008
|
T0 vs. T100, p=0.0036
|
Sedation Scale (RASS)
|
survivors
|
-3(-3;-3)
|
-2(-3;-2)
|
-3(-3;0)
|
0(-2;0)
|
0(0;0)
|
0.055
|
T0 vs. T100, p=0.012
|
|
deceased
|
-4(-4;-3.8)
|
-1.5(-2.5;-0.5)
|
-0.5(-1.2;-0.25)
|
-0.5(-1.5;0)
|
-0.5(-1.8;0)
|
|
|
|
p-value S. vs. D.
|
0.12
|
0.57
|
0.22
|
1.0
|
0.18
|
|
|
|
|
|
|
|
|
|
|
|
Spontaneous breathing
|
Pooled
|
33
|
67
|
56
|
44
|
67
|
0.55
|
|
(%)
|
survivors
|
40
|
80
|
80
|
80
|
100
|
|
|
|
deceased
|
25
|
50
|
25
|
0
|
25
|
|
|
|
p-value S. vs. D.
|
1.0
|
0.52
|
0.21
|
0.048
|
0.048
|
|
|
|
|
|
|
|
|
|
|
|
Tidal volume (mL)
|
Pooled
|
221±214
|
152±150
|
146±132
|
179±163
|
260±221
|
0.63
|
No difference in tidal
|
|
survivors
|
344±214
|
220±170
|
212±142
|
300±105
|
440±65
|
0.13
|
volume between T0 and
|
|
deceased
|
68±64
|
68±61
|
63±56
|
28±34
|
35±57
|
|
T100 in any group
|
|
p-value S. vs. D.
|
0.043
|
0.14
|
0.09
|
0.0017
|
<0.0001
|
|
|
|
|
|
|
|
|
|
|
|
Viable lung
|
Pooled
|
73±37
|
56±29
|
58±21
|
57±21
|
65±25
|
0.83
|
No difference in tidal
|
parenchyma (%)
|
survivors
|
75±43#
|
66±28
|
59±18
|
59±16
|
72±20
|
|
volume between T0 and
|
|
deceased
|
70±42#
|
37±21*
|
57±29*
|
53±32*
|
53±32*
|
|
T100 in any group; *) missing
|
|
p-value S. vs. D.
|
0.91
|
0.15
|
0.89
|
0.74
|
0.34
|
|
data n=1; #) missing data n=2
|
|
|
|
|
|
|
|
|
|
Extracorporeal blood
|
Pooled
|
4.2±0.8
|
4.0±1.0
|
4.2±0.8
|
3.9±0.8
|
3.7±0.8
|
0.66
|
No difference in tidal
|
flow (L/min)
|
survivors
|
3.9±1.0
|
4.1±1.4
|
4.4±1.0
|
3.8±1.0
|
3.8±0.8
|
|
volume between T0 and
|
|
deceased
|
4.4±0.6
|
4.0±0.1
|
4.0±0.5
|
4.0±0.4
|
3.8±0.8
|
|
T100 in any group
|
|
p-value S. vs. D.
|
0.42
|
0.86
|
0.56
|
0.67
|
0.94
|
|
|
|
|
|
|
|
|
|
|
|
Extracorporeal blood
|
Pooled
|
136±21
|
121±27
|
124±23
|
114±24
|
107±21
|
0.13
|
T0 vs. T100, p=0.011
|
flow of calculated
|
survivors
|
133±28
|
117±37
|
120±29
|
102±24
|
101±17
|
0.38
|
T0 vs. T100, p=0.07
|
cardiac output (%)
|
deceased
|
140±10
|
126±6
|
128±15
|
128±17
|
119±16
|
|
T0 vs. T100, p=0.08
|
|
p-value S. vs. D.
|
0.66
|
0.64
|
0.61
|
0.10
|
0.15
|
|
|
|
|
|
|
|
|
|
|
|
Right ventricular
|
Pooled
|
61±28
|
68±32
|
64±22
|
63±14
|
49±23
|
0.55
|
No difference in tidal
|
systolic pressure
|
survivors
|
62±38 *
|
70±45
|
63±30
|
56±16
|
40±26
|
|
volume between T0 and
|
(mmHg)
|
deceased
|
61±10 *
|
65±7.1
|
66±12
|
70±4.7
|
60±16
|
|
T100 in any group
|
|
p-value S. vs. D.
|
0.96
|
0.82
|
0.85
|
0.14
|
0.23
|
|
*) missing data n=1
|
|
|
|
|
|
|
|
|
|
Systolic pulmonary
|
Pooled
|
3.7±1.6
|
5.3±3.0
|
5.6±4.2
|
5.2±3.2
|
4.8±4.8
|
0.85
|
T0 vs. T100, p=0.57
|
blood flow
|
survivors
|
3.8±2.1
|
7.0±2.6
|
7.2±4.2
|
7.1±2.1
|
7.9±4.3
|
|
T0 vs. T100, p=0.12
|
(L/min)
|
deceased
|
3.5±1.0
|
3.0±1.6
|
3.4±3.4
|
2.9±2.7
|
1.0±1.1
|
|
T0 vs. T100, p=0.029
|
|
p-value S. vs. D.
|
0.78
|
0.03
|
0.19
|
0.03
|
0.02
|
|
|
|
|
|
|
|
|
|
|
|
Pulmonary vascular
|
Pooled
|
18.6±12
|
19.4±20.2
|
20.1±16.2
|
13.7±9.9
|
11.7±13.2
|
0.75
|
No difference in tidal
|
resistance
|
survivors
|
18.8±16.3*
|
10.0±4.8
|
12.5±12.7
|
8.2±2.2
|
6.4±3.1
|
|
volume between T0 and
|
(mmHg*min/mL)
|
deceased
|
18.4±5.9*
|
31.1±27.0
|
30.9±16.5
|
42.1±39.7
|
64.0±42.7
|
|
T100 in any group
|
|
p-value S. vs. D.
|
0.97
|
0.12
|
0.10
|
0.09
|
0.02
|
|
*) missing data n=1
|
Echocardiography
Regarding echocardiographic findings there was no significant difference in PBF at T0 between survivors (3.8±2.1 L/min) and non-survivors (3.5±1.0 L/min, p=0.78). Survivors had no significant change over time concerning PBF (3.8±2.1 at T0 compared to 7.9±4.3 L/min at T100, p=0.12). However, among the non-survivors there was a significant deterioration in PBF from 3.5±1.0 to 1.0±1.1 L/min, p=0.029. In detail, a difference between the groups was indicated already at T25 (7.0±2.6 vs. 3.0±1.6 L/min, p=0.03). Right ventricular pressure did not differ at any time point between survivors and non-survivors (Table 4). However, PVR tended to diverge incrementally between the groups from T50 and on (T50, 12.5±12.7 vs. 30.9±16.5 mmHg min/L (p=0.10); T75, 8.2±2.2 vs. 42.1±39.7 mmHg min/L (p=0.09); and T100, 6.4±3.1 vs. 64.0±42.7 mmHg min/L, (p=0.02), respectively).
Of the nine included patients five survived (56%) to hospital discharge and were still alive at 5-year follow up. Four patients were subjected to lung transplantation, two among the survivors and two of the non-survivors. The survivors from lung transplantation presented themselves before transplantation with maintained tidal volumes and PBF in contrast to the transplanted who did not survive. The non-survivors succumbed from uncontrollable intrathoracic bleedings within days after lung transplantation. When examining the perioperative reports, these two patients exhibited a generalized tendency of bleeding when suturing the new organ to the recipient.