Left ventricular overloading is the leading mechanism in extubation failure of patients at high-risk of weaning-induced pulmonary edema.

Dear Editor,

We read with interest the correspondence by Sanfilippo et al. [1] who suggest performing further analysis for our recently reported study [2]. As per their thoughtful suggestion, we first redefined the two study groups as “weaning success” (n = 43) when the patient could be extubated without reinstitution of ventilator support within 48 h after extubation, and as “weaning failure” (n = 16) whenever one the following occurred: (i) failed spontaneous breathing trial (SBT) (n = 12); (ii) reintubation and/or resumption of ventilator support within 48 h after extubation (n = 3); or (iii) death in the 48 h after extubation (n = 1) [3]. When compared to patients who succeeded weaning, patients who failed had greater tachycardia and exhibited a higher E wave maximal velocity (103 cm/s [84-140] vs. 81 cm/s [61-89]: p = 0.002), an increased E/A ratio (1.4 [0.9–2.3] vs. 0.9 [0.7–1.1]: p = 0.002), a shorter E wave deceleration time (110 ms [97-158] vs. 173 ms [130-212]: p = 0.014) and a higher tricuspid regurgitation peak velocity (3.37 m/s [2.70–3.87] vs. 2.83 m/s [2.57–3]: p = 0.034), while baseline left ventricular (LV) ejection fraction and LV outflow tract velocity–time integral were lower (29% [18-36] vs. 35% [26-40]: p = 0.056; and 14.5 [8.5–19.1] vs. 16.7 [13.5–21.1]: p = 0.034, respectively). In contrast, no significant difference in the E′ wave maximal velocity and E/E′ ratio was noted (Table 1). During SBT, variations of Doppler parameters were more marked in patients who failed than in those who succeeded weaning: higher increase of E wave maximal velocity (122 cm/s [92-160] vs. 93 cm/s [73-105]: p = 0.002) and of E/A ratio (1.7 [1–3.5] vs. 0.9 [0.7–1.3]: p = 0.001), and greater shortening of E wave deceleration time (88 ms [74-141] vs. 147 ms [105-174]: p = 0.024). Again, no significant difference was observed for the E′ wave maximal velocity and E/E′ ratio (Table 1). These data confirm those reported initially when dichotomizing patients according to passed or failed SBT [2].

Table 1

Hemodynamic parameters and echocardiographic findings obtained immediately before and during the first spontaneous breathing trial (SBT) according to weaning success or failure

	Weaning success (n = 43)	Weaning failure (n = 16)	P value
Before SBT
Hemodynamic parameters
HR, bpm	90 (79–102)	114 (100–123)	0.001
sBP, mmHg	142 (130–151)	134 (130–151)	0.500
dBP, mmHg	78 (71–86)	81 (71–88)	0.511
mBP, mmHg	97 (87–107)	98 (85–109)	0.993
Echocardiography findings
E, cm/s	81 (61–89)	103 (84–140)	0.002
A, cm/s	86 (70–106)	63 (44–125)	0.290
E/A	0.9 (0.7–1.1)	1.4 (0.9–2.3)	0.002
DTE, ms	173 (130–212)	110 (97–158)	0.014
E′ lateral, cm/s	9.6 (7.4–12.2)	11.1 (8.2–15.4)	0.230
E/E′	7.9 (5.6–11)	9.2 (6.6–12.5)	0.170
MR/LA area	0.2 (0.17–0.25)	0.15 (0.13–0.2)	0.170
LVEF, %	35 (26–40)	29 (18–36)	0.056
LVOT VTI, cm	16.7 (13.5–21.1)	14.5 (8.5–19.1)	0.069
TR peak velocity, m/s	2.83 (2.57–3)	3.37 (2.7–3.87)	0.034
Left ventricular diastolic dysfunction*
Grade 1	25 (58%)	3 (19%)	0.008
Grade 2	16 (37%)	9 (56%)
Grade 3	2 (5%)	4 (25%)
During SBT
Hemodynamic parameters
HR, bpm	96 (84–110)	121 (109–132)	0.002
sBP, mmHg	147 (126–156)	143 (120–166)	0.946
dBP, mmHg	78 (70–92)	89 (70–98)	0.314
mBP, mmHg	103 (90–110)	112 (89–116)	0.177
Echocardiography findings
E, cm/s	93 (73–105)	122 (92–160)	0.002
A, cm/s	86 (75–112)	64 (33–117)	0.127
E/A	0.9 (0.7–1.3)	1.7 (1–3.5)	0.001
DTE, ms	147 (105–174)	88 (74–141)	0.024
E′ lateral, cm/s	9.9 (8–11.3)	12.5 (8–16.1)	0.079
E/E′	9.5 (7.2–12.7)	9.6 (7.8–13.5)	0.500
MR/LA area	0.21 (0.16–0.23)	0.26 (0.14–0.39)	0.330
LVEF, %	33 (25–43)	27 (15–37)	0.087
LVOT VTI, cm	17.8 (14.4–20.8)	12.9 (8.2–18.9)	0.012
TR peak velocity, m/s	3 (2.75–3.3)	3.44 (3.1–4.16)	0.047
Left ventricular diastolic dysfunction*
Grade 1	19 (44%)	3 (19%)	0.007
Grade 2	22 (51%)	6 (38%)
Grade 3	2 (5%)	7 (43%)

HR heart rate, sBP systolic blood pressure, dBP diastolic blood pressure, mBP mean blood pressure, DTE deceleration time of mitral E wave, MR mitral regurgitation, LA left atrium, LVEF left ventricular ejection fraction, LVOT left ventricular outflow tract, VTI velocity–time integral, TR tricuspid regurgitation

*According to [4]

Second, we investigated if the grade of LV diastolic dysfunction was associated with weaning failure [4]. Not surprisingly, the proportion of higher grades of LV diastolic dysfunction was significantly higher in patients who failed weaning, both at baseline and during SBT (Table 1). At baseline, grade 1 was more frequent in the “weaning success” group, whereas grades 2 and 3 were more prevalent in the “weaning failure” group (25/43 [58%] vs. 3/16 [19%] and 13/16 [81%] vs. 18/43 [42%], respectively: p = 0.007). This highlights the facilitating role of underlying diastolic dysfunction in precipitating LV overload at the origin of weaning-induced pulmonary edema [5].

Third, we performed a subgroup analysis in patients with chronic obstructive pulmonary disease (n = 15). In this subset of patients, those who failed weaning (n = 4) tended to have a higher E wave maximal velocity (119 cm/s [58-144] vs. 87 cm/s [61-96]: p = 0.177) and exhibited increased E/A ratio (1.2 [0.9–3.5] vs. 0.9 [0.7–1]: p = 0.04) at baseline.

Overall, these additional analyses confirm the leading role of LV overload in the weaning failure of patients at high risk of developing a weaning-induced pulmonary edema. Whether a therapeutic strategy based on such hemodynamic assessment could facilitate weaning in this targeted population remains to be determined.