Cardiogenic shock is a deadly disorder and temporary mechanical circulatory support devices (tMCS) have been introduced to improve survival. However, criteria and strategies for timely placement and escalation are lacking. A major limitation of previous research is the focus on a single intervention or device; in clinical practice, these patients are transitioned through multiple configurations of tMCS. Accordingly, we have applied the generalized framework of a multistate model to enhance our understanding of the disease process. A multistate survival analysis was conducted on the Altshock-2 registry to describe the proportions, over time, of patients alive without tMCS (starting state), and being treated with tMCS until heart replacement therapy (disease transitions) or, ultimately, death (absorbing state). Among 544 included patients 199 (36.6%) did not receive any tMCS, 252 (46.3%) received a single device (81% intra-aortic balloon pump [IABP]), and 93 (17.1%) underwent escalation. Considering only patients with IABP (281 patients), 77 (27.4%) underwent escalation whereas 204 received IABP as the only tMCS, with in-hospital death of 59.1% and 40.5%, respectively (p <0.01). IABP escalation was associated with increased mortality compared with IABP alone only in the acute coronary syndrome (ACS) group (RR 1.74, 95% CI 1.29 to 2.35), whereas no difference was observed in the heart failure (HF) group (RR 0.66, 95% CI 0.28 to 1.58), also after adjustment for SCAI stage at 24 hours from admission. In conclusion, the need to escalate from IABP to other tMCS devices is associated with increased mortality in the setting of ACS-CS. Timely implantation of the right tMCS is a key factor in CS treatment.
Cardiogenic shock (CS) is a complex condition with multiple competing events and absorbing states that make it challenging to analyze a single status or device without considering the entire disease process. ,, Each patient has a unique trajectory that is dictated by multiple factors (i.e., etiology) and events (i.e., complications). Accordingly, the temporary mechanical circulatory support (tMCS) platform can change over time and nomenclature related to escalation and bridging has been implemented in clinical practice to capture the process of care in the treatment of CS and the ultimate outcome. Escalation, defined as the process of implanting a tMCS device able to provide greater systemic flow and/or biventricular support, appears a key issue. Nevertheless, criteria and strategies for timely escalation are lacking in clinical practice and the outcomes of escalation strategies are not systematically reported. The patient’s trajectory has not been the focus of previous research, even though it is crucial to improving short-term survival.
The aim of this study was to assess the patients’ trajectory and related outcome of patients prospectively enrolled in the Altshock-2 Italian registry, looking at escalation using a multistate model that goes beyond a single medical treatment or device, to enhance our understanding of the disease process. , In this perspective, we decided to select the intra-aortic balloon pump (IABP) as first-line treatment because, even if no trial have demonstrated increased effectiveness of IABP in both acute coronary syndrome (ACS) and heart failure (HF) related CS, IABP is widely used and we question whether its use might delay other therapies and ultimately drive mortality.
Materials and Methods
Data source
The Altshock-2 registry is a multicenter prospective registry (ClinicalTrials.gov Identifier: NCT04295252) that is part of the Italian Altshock-2 program. This study was approved by the local ethics committee of Milano Area 3 of the ASST Grande Ospedale Metropolitano Niguarda (Piazza Ospedale Maggiore 3, 20162 Milano) and conducted in accordance with ethical principles based on the Helsinki Declaration. The Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) guidelines were followed for reporting the findings.
Study population
CS patients enrolled between March 2020 and December 2023 were included. CS was defined as: (1) systolic blood pressure (SBP) <90 mmHg or mean arterial pressure (MAP) <60 mmHg, after an appropriate fluid challenge if there is no sign of overt fluid overload, OR need of vasoactive agents to maintain SBP >90 mmHg or MAP >60 mmHg, OR need of MCS; (2) at least one of the following criteria/signs of overt hypoperfusion: mixed venous oxygen saturation <60%; arterial lactates >2 mmol/L; oliguria <0.5 ml/Kg/h for at least 6 hours. The only exclusion criteria was to be younger than 18 years old upon admission.
Statistical analysis
Continuous variables were reported as median and interquartile range (IQR) and compared using Mann-Whitney U test. Categorical variables were reported as absolute numbers and percentages and compared using the chi-square test.
Multistate model
Time-to-event analysis in CS patients has been traditionally focused solely on the association of one single MCS (IABP, Impella, extracorporeal membrane oxygenation [ECMO]) with the primary outcome (death) while ignoring the influence of time to tMCS and possible transitions between different tMCS before an event of interest (e.g., heart transplant [HT] or durable left ventricular assist device [LVAD] implantation) or death. A multistate model is a generalized framework that describes longitudinal changes in the condition of patients during the course of a complex disorder by estimating the probability in time of visiting multiple states and quantifying the rates of possible transitions between them, accounting for a more granular description and deeper and more comprehensive evaluation of the disease phenomenon.
Multistate models are applied to enhance understanding of the disease process by gathering as much information as possible on the observed patient journeys. The multistate model we considered for Altshock-2 patients is depicted in the Figure 1 . A transient state means a patient can transition out of that state (e.g., from one MCS to another), whereas an absorbing state means a transition out of that state is impossible (e.g., death).
Cardiogenic shock: disease trajectories throughout a multistate approach. In the center is a space diagram representing Altshock-2 patients. Each box is a state in which a patient can exist at a point in time, and each arrow is a possible transition, that is, the direction of moving from one state to another.
All patients were entered at time 0 (which represents the time of admission to the Cardiac Intensive Care Unit [CICU]) as “Alive, no MCS”. The transition to a tMCS was coded accordingly (e.g., when a patient is treated with IABP, the patient moved to the state “Alive, after IABP” at that time) as it was the eventual transition to further tMCS. Patients may also have moved from one MCS state to another if they were treated with multiple MCS. Finally, the patient could move to HT or left ventricular assist device (LVAD) (HRT) and, ultimately, to death. Transition intensities were estimated using the Aalen-Nelson estimator; state probabilities (i.e., the proportion of patients expected in each state at each time) were estimated using the Aalen-Johansen estimator and depicted using stacked plots ( Fig. 2–4 ) The state probabilities at 1 year are reported in Supplementary Tables 1–3 . Multistate analysis was performed in the overall cohort and in subgroups according to CS etiology and SCAI stage at 24 hours after admission. The multistate analysis was performed using R software (version 4.3.2) with mstate package.
State probabilities over time in the overall sample ( N = 544). ECMO = extracorporeal membrane oxygenation; HRT = heart replacement therapy; IABP = intra-aortic balloon pump; MCS = mechanical circulatory support.
State probabilities over time in patients with HF-CS ( N = 157). MCS = mechanical circulatory support; IABP = intra-aortic balloon pump; ECMO = extracorporeal membrane oxygenation; HRT = heart replacement therapy.
State probabilities over time in HF-CS patients in 24-h SCAI stage C ( N = 80). MCS = Mechanical circulatory support; IABP = intra-aortic balloon pump; ECMO = extracorporeal membrane oxygenation; HRT = heart replacement therapy.
Results
Trajectories in the multistate framework
A total of 725 patients have been included in the Altshock-2 registry up until December 2023. Among them, 544 patients for whom admission data, event data, data on escalation, and follow-up data at 1-year follow-up were available, were included in the multistate framework.
Strategies of escalation, use of heart replacement therapies, in-hospital death and the time in days from the first device to escalation are reported in Supplementary Table 4 .
We classified patients into three groups based on the sequence of MCS received: “Escalation” (if a patient received Impella after IABP, or ECMO after IABP/Impella), “Single tMCS” (if a patient received only one MCS), and “No MCS” if the patient was never treated with MCS ( Supplementary Table 1 ). A total of 199 patients (36.6%) did not receive any MCS, of whom 97 died in hospital (48.7%). A total of 252 patients (46.3%) received a single device (204 [81%] IABP, 23 [9.1%] Impella, and 25 [9.9%] V-A ECMO), and significant differences in mortality was observed among the three device types (40.5%, 26.1%, and 60%, respectively; p = 0.048). Ninety-three patients (17.1%) underwent escalation, and in-hospital mortality among these patients was 59.1%.
IABP escalation
Considering the high prevalence of IABP use (281 patients [79%]), we focused on this group. Among the patients receiving IABP, 77 (27.4%) underwent escalation. Table 1 describes the baseline characteristics of the group treated with only IABP versus the IABP + escalation group. Escalation was more common in patients with acute coronary syndrome (ACS), in younger patients, and in patients with greater hemodynamic severity or worsening hemodynamic severity (increased lactate levels, organ damage, mostly in SCAI C patients). Overall, in-hospital mortality was 40.5% in the IABP only group versus 59.1% in the IABP + escalation group (p <0.01).
Table 1
Characteristics of patients treated with IABP by MCS sequence (IABP then escalation vs IABP only)
| Variables |
Overall
N = 281 |
IABP then escalation
N = 77 |
IABP only
N = 204 |
p | |
|---|---|---|---|---|---|
| Clinical characteristics | |||||
| Etiology, no. (%) | ACS-CS | 174 (62%) | 50 (65%) | 124 (61%) | 0.008 |
| HF-CS | 67 (24%) | 13 (17%) | 54 (26%) | ||
| Other | 40 (14%) | 14 (18%) | 26 (13%) | ||
| Female sex, no. (%) | 60 (21%) | 15 (19%) | 45 (22%) | 0.744 | |
| Age (IQR), yr | 64 (55–72) | 60 (52–66) | 66 (57–74) | <0.001 | |
| BMI [IQR), kg/m 2 | 25.9 (23.5–28.1) | 26.1 (24.2–28.5) | 25.69 (23.4–27.9) | 0.249 | |
| Smoke, no. (%) | 84 (30%) | 20 (26%) | 64 (32%) | 0.47 | |
| Hypertension, no. (%) | 150 (53%) | 39 (51%) | 111 (54%) | 0.667 | |
| Diabetes, no. (%) | 94 (33%) | 17 (22%) | 77 (38%) | 0.019 | |
| Dyslipidemia, no. (%) | 120 (43%) | 26 (34%) | 94 (46%) | 0.079 | |
| Stroke/TIA, no. (%) | 12 (4.3%) | 2 (2.6%) | 10 (4.9%) | 0.597 | |
| Peripheral artery disease, no. (%) | 49 (17%) | 10 (13%) | 39 (19%) | 0.302 | |
| Asthma, no. (%) | 28 (10%) | 6 (7.8%) | 22 (11%) | 0.592 | |
| Atrial fibrillation, no. (%) | 60 (21%) | 12 (16%) | 48 (24%) | 0.192 | |
| Chronic kidney disease, no. (%) | 40 (14%) | 8 (10%) | 32 (16%) | 0.339 | |
| Anemia, no. (%) | 29 (10%) | 3 (3.9%) | 26 (13%) | 0.051 | |
| Hepatopathy, no. (%) | 12 (4.3%) | 4 (5.2%) | 8 (3.9%) | 0.889 | |
| Thyroid disorders, no. (%) | 32 (11%) | 9 (12%) | 23 (11%) | 1 | |
| Prior PCI, no. (%) | 70 (25%) | 19 (25%) | 51 (25%) | 1 | |
| Prior CABG, no. (%) | 24 (8.6%) | 4 (5.3%) | 20 (9.9%) | 0.323 | |
| Prior cardiomyopathy, no. (%) | 87 (31%) | 24 (31%) | 63 (31%) | 1 | |
| Hemometabolic characteristics | |||||
| Lactate T0 (IQR), mmol/L | 2.60 [1.60–5.27) | 3.50 (2.08–7.12) | 2.30 (1.50–4.38) | 0.001 | |
| Lactate 24 h (IQR), mmol/L | 1.60 (1.12–2.40) | 2.10 (1.30–3.32) | 1.50 (1.10–2.10) | 0.002 | |
| AST T0 (IQR), U/I | 99 (33–352) | 207 (64–736) | 68.50 (28–274) | <0.001 | |
| AST 24 h (IQR), U/I | 139 (40–415) | 271.5 (57–723) | 102 (37–330) | 0.008 | |
| AST last (IQR), U/I | 36 (19–57) | 45 (25–82) | 32 (19–50) | 0.08 | |
| ALT T0 (IQR), U/I | 59 (27–162) | 123.5 (42–292) | 47.50 (23–125) | 0.001 | |
| ALT 24h (IQR), U/I | 69 (31–204) | 134 (44–242) | 59 (30–155) | 0.015 | |
| ALT last (IQR), U/I | 36 (21–77) | 56 (18–90) | 34 (22–64) | 0.237 | |
| Creatinine T0 (IQR), mg/dL | 1.27 (0.99–1.72) | 1.30 (1.06–1.69) | 1.25 (0.97–1.74) | 0.432 | |
| Creatinine 24 h (IQR), mg/dL | 1.38 (1.02–2.10) | 1.46 (1.07–2.10) | 1.35 (1.00–2.08) | 0.539 | |
| Creatinine last (IQR), mg/dL | 1.10 (0.84–1.58) | 1.27 (0.90–1.67) | 1.08 (0.80–1.56) | 0.322 | |
| eGFR T0 (IQR), mL/min/1.73 m 2 | 61.0 (43.0–81.0) | 65.5 (50.2–81.2) | 58.0 (39.0–80.0) | 0.123 | |
| eGFR 24 h (IQR), mL/min/1.73 m 2 | 55.0 (35.7–81.0) | 63.0 (41.0–85.0) | 53.0 (35.0–80.0) | 0.09 | |
| eGFR last (IQR), mL/min/1.73 m 2 | 69.0 (46.0–99.0) | 70.0 (47.7–102.0) | 68.0 (46.0–99.0) | 0.58 | |
| SAP T0 (IQR), mm Hg | 93 (80–110) | 90 (80–101) | 95 (82–115) | 0.027 | |
| SAP 24 h (IQR), mm Hg | 105 (92–115) | 100 (85–115) | 105 (95–115) | 0.065 | |
| SAP last (IQR), mm Hg | 105 (95–117) | 100 (85–112) | 105 (96–118) | 0.045 | |
| RAP T0 (IQR), mm Hg | 10 (7–14) | 10 (8–13) | 10 (7–14) | 0.831 | |
| RAP 24 h (IQR), mm Hg | 9 (6–12) | 10 (7–12) | 8 (5–11) | 0.055 | |
| Echocardiographic characteristics and score | |||||
| LVEF T0 (IQR), (%) | 25 (18–30) | 16 (10–25) | 25 (20–30) | <0.001 | |
| LVEF 24 h (IQR), (%) | 25 (20–30) | 17 (11–25) | 25 (20–35) | <0.001 | |
| TAPSE 0 h (IQR), mm | 16 (13–18) | 15 (10–18) | 16 (14–19) | 0.038 | |
| TAPSE 24 h (IQR), mm | 17 (14–18) | 15 (12–17) | 17 (15–18) | 0.071 | |
| sPAP T0 (IQR), mm Hg | 40 (34–50) | 41 (35–50) | 40 (33–50) | 0.709 | |
| sPAP 24 h (IQR), mm Hg | 36 (30–50) | 32 (30–42) | 39 (31–50) | 0.264 | |
| SOFA score T0 (IQR) | 7 (4–9) | 8 (6–9) | 6 (4–9) | <0.001 | |
| SOFA score 24 h (IQR) | 7 (5–9) | 8 (6–10) | 6 (4–8) | <0.001 | |
| SOFA score last (IQR) | 3 (1–5) | 4 (2–7) | 2 (1–4) | 0.01 | |
| SAPS score T0 (IQR) | 43.0 (34.0–56.0) | 47.5 (39.0–57.0) | 41.0 (31.0–54.0) | 0.028 | |
| SAPS score 24 h (IQR) | 38.0 (29.0–48.0) | 42.0 (36.0–50.0) | 36.0 (28.0–46.0) | 0.005 | |
| SAPS score last (IQR) | 27.0 (18.0–34.0) | 28.0 (21.0–39.0) | 27.0 (18.0–34.0) | 0.259 | |
| SCAI T0, no. (%) | A | 6 (2.3%) | 1 (1.3%) | 5 (2.6%) | <0.001 |
| B | 22 (8.3%) | 3 (4%) | 19 (10%) | ||
| C | 142 (53%) | 20 (27%) | 122 (64%) | ||
| D | 67 (25%) | 27 (36%) | 40 (21%) | ||
| E | 29 (11%) | 24 (32%) | 5 (2.6%) | ||
| SCAI 24h, no. (%) | A | 13 (5.2%) | 1 (1.4%) | 12 (6.6%) | <0.001 |
| B | 30 (12%) | 3 (4.2%) | 27 (15%) | ||
| C | 145 (57%) | 28 (39%) | 117 (65%) | ||
| D | 46 (18%) | 24 (34%) | 22 (12%) | ||
| E | 18 (7.1%) | 15 (21.1%) | 3 (1.7%) | ||
| Inotropic score (IQR) | 15 (6–31) | 22 (10–45) | 14 (6–25) | 0.001 | |
| Respiratory support | |||||
| Noninvasive ventilation, no. (%) | 39 (14%) | 1 (1.3%) | 38 (19%) | <0.001 | |
| Mechanical ventilation, no. (%) | 169 (60%) | 72 (93%) | 97 (47%) | <0.001 | |
| MCS sequence | |||||
| IABP only, no. (%) | 204 (73%) | 0 (0.0) | 204 (100%) | ||
| IABP → ECMO, no. (%) | 48 (17%) | 48 (62%) | 0 (0.0) | ||
| IABP → ECMO → ECPella, no. (%) | 12 (4.3%) | 12 (16%) | 0 (0.0) | ||
| IABP → Impella, no. (%) | 12 (4.3%) | 12 (16%) | 0 (0.0) | ||
| IABP → Impella → ECMO, no. (%) | 5 (1.8%) | 5 (6.5%) | 0 (0.0) | ||
| Time to escalation after IABP, no. (%) | |||||
| Same day as IABP | 62 (80%) | ||||
| 1 day after IABP | 3 (3.9%) | ||||
| 2–7 days after IABP | 5 (6.5%) | ||||
| >7 days after IABP | 7 (9.1%) | ||||
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
