Abstract
Objectives:
The objective of this study was to describe practice patterns of anesthetic management during pericardial window creation.
Design:
Retrospective observational cohort study.
Setting:
Single tertiary cancer center.
Participants:
A total of 150 patients treated for cancer between 2011 and 2015 were included in the study.
Measurements and Main Results:
The primary objective was to evaluate anesthetic management in pericardial window creation. Secondary outcomes were 30-day mortality and overall survival after pericardial window creation. Thirty-day mortality was 19.3%, and median survival was 5.84 months. Higher American Society of Anesthesiologists (ASA) physical status of patients was associated with preinduction arterial line placement (51% ASA 3 vs. 79% ASA 4; p=0.002) and use of etomidate for anesthetic induction (34% ASA 3 vs. 60% ASA 4; p=0.003). However, there was no association between anesthetic management and presence of tamponade in these patients. Cardiac aspirate volume (per 10 mL: OR, 1.02 [95% CI, 1.0–1.04]; p=0.026) and intraoperative dysrhythmia (atrial fibrillation: OR, 6.76 [95% CI, 1.2–37.49]; p=0.029; sinus tachycardia: OR, 4.59 [95% CI, 1.25–16.90]; p=0.022) were independently associated with increased 30-day mortality. High initial heart rate (per 10 bpm: HR, 1.18 [95% CI, 1.05–1.33]; p=0.005) in the operating room and intraoperative sinus tachycardia (HR, 1.86 [95% CI, 1.15–3.03]; p=0.012) were independently associated with worse overall survival.
Conclusions:
Risk of death after pericardial window creation remains high in patients with cancer. Variations in anesthetic management did not affect survival in oncologic patients with pericardial effusions.
Keywords: Cancer, pericardial effusion, subxiphoid pericardial window creation, anesthetic management, survival
Introduction
Malignancy is a common cause of pericardial effusion, which is often a sign of advanced disease and poor prognosis.1 In patients with cancer, pericardial effusion may develop through several different mechanisms: (1) direct extension or metastatic spread of the underlying malignancy, (2) complication of systemic treatment (such as immunotherapy or chemotherapy), and (3) opportunistic infection in the setting of an immunocompromised state after antineoplastic therapy.1, 2
Previous studies have investigated survival and rates of effusion recurrence after different surgical interventions for pericardial effusion in patients with cancer,3–13 but no study to date has examined the effect of intraoperative anesthetic management on risk of death in these patients. The aim of this study was to evaluate anesthetic management during subxiphoid pericardial window creation, as well as 30-day mortality and overall survival, in patients with cancer presenting with pericardial effusion.
Methods
We retrospectively reviewed the records of patients with cancer who underwent surgical intervention for pericardial effusion in a tertiary cancer center between January 2011 and December 2015. Patient information was maintained in an institutional database, with codified identifiers to protect confidentiality. A waiver of informed consent was approved by the hospital institutional review board due to the retrospective nature of the study.
Patient Population
All patients with malignancy requiring surgical drainage for pericardial effusion in our institution during the study period were included. We excluded patients with incomplete intraoperative anesthetic records as well as patients who had previously undergone pericardial drainage in the operating room.
Intraoperative Management
Standard anesthetic management at our institution consisted of standard ASA monitoring once the patient entered the operating room. Vital signs were automatically recorded into the electronic anesthetic record. The patient was prepped and draped while awake, with at least 2 peripheral IVs placed. An indwelling radial arterial catheter was placed before induction of anesthesia or soon thereafter in a majority of cases. General anesthesia was induced, with the surgeon in the room, with induction and paralytic agent use determined by the anesthesia attending. Vasopressors were used as necessary to maintain hemodynamic stability. A midline subxiphoid incision was made, with elevation of the xyphoid process, until the underlying pericardium was identified. The pericardium was opened sharply, allowing release of pericardial fluid. A pericardial window was created by removing at least a 2- to 3-cm portion of pericardium, and fluid was sent off for cytologic examination. A #24 Blake chest drain was brought out through the skin, placed in the posterior pericardium, and connected to bulb suction. The incision was then closed in the standard fashion.
Data Collection
Data from the electronic health record of the institution was extracted from electronic medical records retrospectively for the following preoperative factors: sex, age, body mass index, primary malignancy site and stage. The presence of effusions as well as diagnosis of tamponade were based preoperative echocardiogram and cardiology attending note prior to surgery.
Intraoperative data, such as timing of arterial line insertion, use of vasopressors and inotropes, selection of induction and paralytic agents, surgery time, hemodynamic variables, and dysrhythmias, were collected from the anesthetic record. For the purpose of this analysis, we defined survival as both short-term, 30-day mortality, and overall survival. Thirty-day mortality was defined as any death occurring within 30 days from the date of surgery. Overall survival was measured from the date of surgery to the date of death and was censored at the time of the last follow-up.
Statistical Analysis
Descriptive statistics, such as median and interquartile range (IQR), were used for continuous variables, and groups were compared using the Wilcoxon rank sum test. Count and percentage were presented for categorical variables, and groups were compared using Fisher’s exact test. A logistic regression model was used for 30-day mortality. Cox proportional hazards models were used for overall survival. Backward selection was used for multivariable analysis, and a factor was entered if it had p<0.1 in the univariable analysis. SAS version 9.4 (SAS institute, Cary, NC) was used for all analyses. All tests were 2-sided, and p<0.05 was considered to indicate statistical significance.
Results
Preoperative Characteristics
In total, 150 patients were included in the final analysis treated for pericardial effusion with the creation of a pericardial window. Of these 150 patients, 102 (68%) were women, and 48 (32%) were men. The median age was 59 years. Of the patients with metastatic cancer (83%), the most common primary malignancies were lung (37%) and breast (24%) cancer (Table 1). The median age of patients varied by site of the primary malignancy: 62 years for patients with lung cancer, 54 years for patients with breast cancer, 46 years for patients with hematologic cancer, and 63 years for other sites of cancer. Preoperative transthoracic echocardiogram showed that 55% had evidence of tamponade. The majority of patients (65%) were assigned American Society of Anesthesiologists (ASA) physical status of 4, and the other 35% were assigned ASA 3. Baseline patient characteristics are shown in Table 1.
Table 1.
Patient Characteristics (N=150)
| Characteristic | No. (%) or Median (IQR) |
|---|---|
|
| |
| Sex | |
| Female | 102 (68) |
| Male | 48 (32) |
| Age, years | 59 (48–67) |
| BMI | 24 (21–27) |
| ASA physical status | |
| ASA 3 | 53 (35) |
| ASA 4 | 97 (65) |
| Metastatic | |
| No | 26 (17) |
| Yes | 124 (83) |
| Primary cancer | |
| Breast | 36 (24) |
| GI/GU | 21 (14) |
| Hematologic | 14 (9) |
| Lung | 56 (37) |
| Other | 23 (15) |
| Tamponade (N=143)a | |
| No | 65 (45) |
| Yes | 78 (55) |
| Time from cancer diagnosis to surgery, months | 27 (10–72) |
ASA, American Society of Anesthesiologists; BMI, body mass index; GI/GU, gastrointestinal/genitourinary; IQR, interquartile range.
Seven patients did not have preoperative echocardiogram.
Intraoperative Findings
Median time from diagnosis of cancer to surgery for pericardial effusion was 27 months. Median duration of surgery was 44 min, and median duration of anesthesia was 116 min. Twenty-six cases (17%) took place between the hours of 8 pm and 7 am. There were no intraoperative deaths.
Median initial heart rate was 103 bpm. Forty-two cases (28%) had initial heart rate higher than 110 bpm. Forty-nine cases (33%) had hemodynamic instability with SBP <80 mmHg for more than 5 min intraoperatively or required repeated boluses of vasopressors. Forty-five cases (30%) had end-tidal CO2 level <30 mmHg for >15 min. Dysrhythmias were common (65%, n=98), with sinus tachycardia (heart rate >100 bpm) recorded most often (55%). Median estimated volume of pericardial fluid evacuated was 500 mL. Malignant cells were found in pericardial fluid in 73 cases (49%) (Table 2).
Table 2.
Univariable and Multivariable Logistic Regression for 30-Day Mortality
| Variable | Deceased (N=29), No. (%) or Median | Alive (N=121), No. (%) or Median (IQR) | OR (95% CI) | P* |
|---|---|---|---|---|
|
| ||||
| Univariable | ||||
| Sex | ||||
| Female | 15 (51.7) | 87 (71.9) | REF | |
| Male | 14 (48.3) | 34 (28.1) | 2.39 (1.04–5.47) | 0.040* |
| Age (for every additional year) | 60 (50–69) | 58 (48–65) | 1.00 (0.98–1.03) | 0.877 |
| BMI | 23 (21–27) | 25 (21–27) | 0.97 (0.89–1.05) | 0.433 |
| Time from diagnosis to surgery | 15.7 (9.8–27.4) | 31.8 (12.2–73.7) | 0.99 (0.98–1.00) | 0.127 |
| Preoperative tamponade on echocardiogram | 18 (62.1) | 60 (52.6) | 1.47 (0.64–3.40) | 0.364 |
| Metastatic cancer | ||||
| No | 6 (20.7) | 20 (16.5) | REF | |
| Yes | 23 (79.3) | 101 (83.5) | 0.76 (0.27–2.10) | 0.596 |
| Cancer type | ||||
| Lung cancer | 12 (41.4) | 44 (36.4) | REF | |
| Hematologic cancer (vs lung cancer) | 2 (6.9) | 12 (9.9) | 0.61 (0.12–3.11) | 0.553 |
| Breast cancer (vs lung cancer) | 4 (13.8) | 32 (26.4) | 0.46 (0.14–1.55) | 0.210 |
| GI/GU cancer (vs lung cancer) | 8 (27.6) | 13 (10.7) | 2.26 (0.76–6.70) | 0.143 |
| Other cancer (vs lung cancer) | 3 (10.3) | 20 (16.5) | 0.55 (0.14–2.17) | 0.393 |
| ASA Physical Status | ||||
| ASA 3 | 4 (13.8) | 48 (39.7) | REF | |
| ASA 4 | 25 (86.2) | 72 (59.5) | 4.25 (1.39–12.99) | 0.011* |
| Malignant cardiac aspirate | 18 (66.7) | 55 (50) | 2 (0.83–4.84) | 0.124 |
| Initial heart rate, per 10 bpm | 11.4 (9.4–12.1) | 10.2 (8.8–11.6) | 1.15 (0.90–1.46) | 0.253 |
| Starting mean blood pressure (mmHg) | 88 (79–97) | 93 (84–103) | 0.99 (0.96–1.01) | 0.384 |
| Anesthesia duration (for every additional 5 h) | 23.2 (20.0–27.0) | 23.0 (19.8–27.8) | 1.00 (0.98–1.01) | 0.550 |
| Surgery duration (min) | 41 (36–52) | 45 (37–59) | 0.98 (0.96–1.00) | 0.103 |
| Overnight surgery | 5 (17.2) | 21 (17.4) | 0.99 (0.34–2.90) | >0.95 |
| Estimated blood loss (mL) | 10 (0–100) | 20 (5–50) | 1.00 (1.00–1.00) | 0.410 |
| Intravenous fluid (mL) | 775 (450–1225) | 800 (600–1000) | 1.00 (1.00–1.00) | 0.910 |
| Minimum systolic blood pressure (mmHg) | 106 (95–117) | 99 (79–118) | 1.00 (0.99–1.02) | 0.771 |
| Minimum mean blood pressure (mmHg) | 79 (73–89) | 76 (59–88) | 1.01 (0.99–1.04) | 0.315 |
| Minimum diastolic blood pressure (mmHg) | 69 (57–75) | 63.5 (51–74) | 1.02 (0.99–1.05) | 0.199 |
| End-tidal CO2 <30 mmHg | 12 (41.4) | 33 (27.3) | 1.88 (0.81–4.36) | 0.140 |
| Inotrope infusion | 5 (17.2) | 22 (18.2) | 0.94 (0.32–2.73) | 0.906 |
| Phenylephrine administration | 10 (34.5) | 48 (39.7) | 0.80 (0.34–1.87) | 0.607 |
| Ephedrine administration | 6 (20.7) | 22 (18.2) | 1.17 (0.43–3.22) | 0.756 |
| Preinduction arterial line (vs post-induction) | 24 (82.8) | 81 (67.5) | 2.31 (0.82–6.52) | 0.113 |
| Dysrhythmia | ||||
| Atrial fibrillation | 5 (17.2) | 8 (6.6) | 7.50 (1.65–34.05) | 0.009* |
| PACs and PVCs (n=3) | 1 (3.4) | 2 (1.7) | 6.00 (0.44–81.45) | 0.178 |
| Sinus tachycardia | 19 (65.5) | 63 (52.1) | 3.62 (1.16–11.34) | 0.027* |
| Induction agent | ||||
| Etomidate | 19 (65.5) | 57 (47.5) | REF | |
| Propofol (vs etomidate) | 5 (17.2) | 32 (26.7) | 0.47 (0.16–1.37) | 0.168 |
| Ketamine (vs etomidate) | 3 (10.3) | 3 (2.5) | 3.00 (0.56–16.14) | 0.201 |
| Combo1 (vs etomidate) | 2 (6.9) | 27 (22.5) | 0.22 (0.05–1.02) | 0.054 |
| Paralytic agent | ||||
| Succinylcholine only | 12 (44.4) | 40 (36) | REF | |
| Vecuronium (vs succinylcholine) | 4 (14.8) | 20 (18) | 0.67 (0.19–2.33) | 0.526 |
| Rocuronium (vs succinylcholine) | 7 (25.9) | 32 (28.8) | 0.73 (0.26–2.07) | 0.552 |
| Succinylcholine and vecuronium | 3 (11.1) | 17 (15.3) | 0.59 (0.15–2.35) | 0.453 |
| SBP <80 mmHg for >5 min | 2 (6.9) | 4 (3.3) | 1.09 (0.93–1.28) | 0.275 |
| Multivariable | ||||
| Dysrhythmia | ||||
| Atrial fibrillation | 5 (17.2) | 8 (6.6) | 6.76 (1.22–37.49) | 0.029* |
| PACs and PVCs (n=3) | 1 (3.4) | 2 (1.7) | 7.69 (0.25–233.62) | 0.241 |
| Sinus tachycardia | 19 (65.5) | 63 (52.1) | 4.59 (1.25–16.90) | 0.022* |
| Cardiac aspirate (mL) | 600 (450–800) | 500 (400–600) | 1.02 (1.00–1.04) | 0.026* |
ASA, American Society of Anesthesiologists; BMI, body mass index; CI, confidence interval; IQR, interquartile range; OR, odds ratio; PAC, premature atrial contraction; PVC, premature ventricular contraction.
Significant P value.
Anesthetic Management
Etomidate was used as an induction agent in 76 cases (51%), with a median dose of 20 mg, followed by propofol in 37 cases (25%), with a median dose of 95 mg. A combination of induction agents was used in 29 cases (19%), and 6 cases (4%) used only ketamine for induction. Only ten cases (7%) maintained spontaneous ventilation during induction.
Initial paralysis (n=138) was accomplished with succinylcholine in 72 cases (52%), 20 of which (14%) received rocuronium or vecuronium for subsequent reparalysis. Thirty-nine cases (28%) received rocuronium as the sole paralytic agent, and 24 cases (17%) received vecuronium as the sole paralytic agent. Only 3 patients received cisatracurium, due to renal failure or insufficiency.
One hundred four cases (69%) were performed with a preinduction arterial line, 28 cases (19%) were performed with a postinduction arterial line, and 18 cases (12%) were performed without an arterial line. Surgical incisions were made within 5 min after anesthesia induction in 71 cases (47%) (Table S1).
Only 2 patients required intraoperative epinephrine administration, 58 patients (39%) required phenylephrine, with a median dose of 120 mcg, and 28 patients (19%) required ephedrine, with a median dose of 10 mg. The median volume of crystalloid administered intraoperatively was 800 mL. Fentanyl was administered to most patients (n=148 [99%]), with a median dose of 200 mcg, compared with hydromorphone (n=44 patients [29%]), with a median dose of 1 mg (Table S1).
More patients with ASA 4 status received etomidate for induction, compared with patients with ASA 3 status (60% vs 34%; p=0.003). A higher number of patients with ASA 4 status received preinduction arterial line, compared with patients with ASA 3 status (79% vs 51%; p=0.002) (Table 3). No statistically significant difference in anesthetic management was seen between patients with preoperative echographic evidence of tamponade and those without tamponade (Table 3) or with high baseline heart rates (Table S2).
Table 3.
Anesthetic Management by ASA Status and Presence of Tamponade
| Management | ASA 3 (N=53) | ASA 4 (N=97) | P | Tamponade (N=78) | No Tamponade (N=65) | P |
|---|---|---|---|---|---|---|
|
| ||||||
| Induction agent | 0.003 | 0.114 | ||||
| Etomidate | 18 (34) | 58 (60) | 47 (60) | 27 (42) | ||
| Propofol | 21 (40) | 16 (17) | 13 (17) | 19 (30) | ||
| Ketamine | 1 (2) | 5 (5) | 4 (5) | 2 (3) | ||
| Combinationa | 12 (23) | 17 (18) | 14 (18) | 15 (23) | ||
| None | 1 (1) | 0 (0) | 0 (0) | 1 (2) | ||
| Paralytic agent(s) | 0.081 | 0.912 | ||||
| Succinylcholine | 13 (27) | 39 (44) | 30 (41) | 21 (36) | ||
| Vecuronium | 14 (29) | 10 (11) | 10 (14) | 11 (19) | ||
| Rocuronium | 14 (29) | 25 (28) | 20 (27) | 17 (29) | ||
| Succinylcholine and vecuronium or succinylcholine and rocuronium | 7 (14) | 13 (15) | 12 (16) | 8 (14) | ||
| Cisatracurium | 1 (2) | 2 (2) | 1 (1) | 1 (2) | ||
| Arterial line | 0.002 | 0.235 | ||||
| Preinduction arterial line | 27 (51) | 77 (79) | 59 (76) | 45 (69) | ||
| Postinduction arterial line | 16 (30) | 12 (12) | 14 (18) | 10 (15) | ||
| No arterial line | 10 (19) | 8 (8) | 5 (6) | 10 (15) | ||
| Phenylephrine required | 21 (40) | 37 (38) | 0.863 | 33 (42) | 24 (37) | 0.607 |
| Median dose, mcg | 120 (80–160) | 120 (80–320) | 0.571 | 80 (80–300) | 120 (80–240) | 0.799 |
| Ephedrine required | 9 (17) | 19 (20) | 0.827 | 13 (17) | 14 (22) | 0.523 |
| Median dose, mg | 15 (10–15) | 10 (10–25) | >0.95 | 15 (10–25) | 10 (5–15) | 0.202 |
Data are no. (%) or median (interquartile range). Comparisons between groups used Fishers exact test for categorical variables and Wilcoxon rank sum for continuous.
Survival Outcomes
Median survival was 5.84 months. Twenty patients died before discharge from the hospital, with in-hospital mortality of 13.3% (n=20), 30-day mortality of 19.3% (n=29), and 90-day mortality of 36.7% (n=55). No statistically significant difference in 30-day or overall survival was observed by site of primary cancer. Evidence of cardiac tamponade on preoperative echocardiography also was not statistically significantly associated with 30-day, or overall survival. Gastrointestinal and genitourinary cancer were associated with the shortest median survival in our study (2.24 months [95% CI, 0.49–6.71]), followed by other cancer sites (3.78 months [95% CI, 1.38–5.66]), lung cancer (5.69 months [95% CI, 2.86–8.39]), hematologic cancer (8.22 months [95% CI, 1.88–28.68]), and breast cancer (13.19 months [95% CI, 7.57–18.03]) (Table S3 and Figure 1).
Figure 1.
Overall survival by site of the primary malignancy. GI/GU, gastrointestinal/genitourinary.
Thirty-day mortality was higher for male sex (48% vs 28%) (OR, 2.39 [95% CI, 1.04–5.47]; p=0.040), ASA 4 (86% vs 60%) (OR, 4.25 [95% CI, 1.39–12.99]; p=0.011), higher cardiac aspirate volume (median, 600 vs. 500) (OR per 10-mL increase, 1.02 [95% CI, 1.00–1.04]; p=0.026), atrial fibrillation (17% vs 7%) (OR, 7.50 [95% CI, 1.65–34.05]; p=0.009), and sinus tachycardia (66% vs 52%) (OR, 3.62 [95% CI, 1.16–11.34]; p=0.027). Age, tamponade, time from diagnosis to surgery, metastatic disease, initial heart rate, hemodynamic instability, end tidal CO2 <30 mmHg for 15 min, and presence of a preinduction arterial line were not associated with 30-day mortality. However, in the final multivariable analysis for 30-day mortality, only cardiac aspirate per 10-mL increase (OR, 1.02 [95% CI, 1.00–1.04]; p=0.026), atrial fibrillation (OR, 6.76 [95% CI, 1.22–37.49]; p=0.029), and sinus tachycardia (OR, 4.59 [95% CI, 1.25–16.90]; p=0.022) were independently associated with 30-day mortality (Table 2).
On univariable analysis for overall survival, metastatic disease (HR, 1.77 [95% CI, 1.07–2.93]; p=0.026), presence of malignant cells in pericardial effusion (HR, 1.50 [95% CI, 1.03–2.17]; p=0.032), higher initial heart rate (per 10 bpm: HR, 1.19 [95% CI, 1.08–1.31]; p<0.001), and intraoperative sinus tachycardia (HR, 1.76 [95% CI, 1.19–2.59]; p=0.004) were associated with shorter overall survival (Table 4), and overall survival was longer when two induction agents were used in combination (HR, 0.61 [95% CI, 0.38–0.98]; p=0.039) (propofol with etomidate, ketamine with propofol, or ketamine with etomidate) or when propofol was used alone (HR, 0.62 [95% CI, 0.40–0.96]; p=0.031) than when etomidate was used alone. On multivariable analysis (Table 4), higher initial heart rate (HR, 1.18 [95% CI, 1.05–1.33]; p=0.005) and intraoperative sinus tachycardia (HR, 1.86 [95% CI, 1.15–3.03]; p=0.012) were associated with shorter overall survival. Each increase in initial heart rate of 10 bpm was associated with an 18% increase in risk of death.
Table 4.
Univariable and Multivariable Cox Model for Overall Survival
| Variable | HR (95% CI) | P |
|---|---|---|
|
| ||
| Univariable | ||
| Preoperative tamponade on echocardiogram | 1.41 (0.98–2.01) | 0.064 |
| Male (vs female) | 1.28 (0.88–1.85) | 0.194 |
| Age (for every additional year) | 1.00 (0.99–1.01) | 0.737 |
| BMI | 0.97 (0.94–1.01) | 0.100 |
| Metastatic cancer (vs nonmetastatic) | 1.77 (1.07–2.93) | 0.026 |
| Hematologic cancer (vs lung cancer) | 0.74 (0.38–1.42) | 0.357 |
| Breast cancer (vs lung cancer) | 0.70 (0.44–1.11) | 0.133 |
| GI/GU cancer (vs lung cancer) | 1.25 (0.73–2.15) | 0.414 |
| Other cancer (vs lung cancer) | 1.07 (0.63–1.80) | 0.803 |
| Malignant cardiac aspirate | 1.50 (1.03–2.17) | 0.032 |
| Initial heart rate, per 10 bpm | 1.19 (1.08–1.31) | <0.001a |
| Starting mean blood pressure | 1.00 (0.98–1.01) | 0.375 |
| Anesthesia duration (for every additional 5 h) | 1.01 (0.98–1.03) | 0.570 |
| Surgery duration | 1.00 (0.99–1.00) | 0.152 |
| Overnight surgery | 1.43 (0.92–2.22) | 0.112 |
| Estimated blood loss | 1.00 (1.00–1.00) | 0.890 |
| Intravenous fluid | 1.00 (1.00–1.00) | >0.95 |
| Minimum systolic blood pressure | 0.99 (0.99–1.00) | 0.101 |
| Minimum mean blood pressure | 1.00 (0.99–1.01) | 0.445 |
| End-tidal CO2 <30 mmHg | 1.25 (0.85–1.84) | 0.248 |
| Inotrope infusion | 1.22 (0.78–1.90) | 0.387 |
| Preinduction arterial line (vs post-induction) | 1.23 (0.83–1.82) | 0.301 |
| Dysrhythmia | ||
| Atrial fibrillation | 1.71 (0.88–3.34) | 0.115 |
| PACs and PVCs (N=3) | 2.17 (0.67–7.04) | 0.196 |
| Sinus tachycardia | 1.76 (1.19–2.59) | 0.004a |
| Induction agent (vs etomidate) | ||
| Propofol | 0.62 (0.40–0.96) | 0.031 |
| Ketamine | 1.50 (0.65–3.48) | 0.343 |
| Combo1 | 0.61 (0.38–0.98) | 0.039 |
| Paralytic agent (vs succinylcholine) | ||
| Cisatracurium | 1.96 (0.60–6.37) | 0.262 |
| Rocuronium | 0.80 (0.51–1.25) | 0.331 |
| Vecuronium | 0.74 (0.43–1.25) | 0.254 |
| Succinylcholine and vecuronium or succinylcholine and rocuronium | 0.70 (0.39–1.25) | 0.222 |
| SBP <80 mmHg for >5 min | 1.07 (0.98–1.16) | 0.127 |
| Multivariable | ||
| Dysrhythmia | ||
| Reference (NSR) | 1 | — |
| Atrial fibrillation | 0.96 (0.42–2.22) | 0.924 |
| PACs and PVCs (n=3) | 50.74 (5.75–448.0) | <0.001 |
| Sinus tachycardia | 1.86 (1.15–3.03) | 0.012 |
| Higher initial heart rate, per 10 bpm | 1.18 (1.05–1.33) | 0.005 |
BMI, body mass index; CI, confidence interval; GI/GU, gastrointestinal/genitourinary; HR, hazard ratio; NSR, normal sinus rhythm; PACs, premature atrial contractions; PVCs, premature ventricular contractions; SBP, systolic blood pressure.
Associations remain statistically significant after final multivariable analysis (Table 6).
Discussion
We examined several perioperative markers of patient acuity, such as presence of tamponade, ASA status, and initial tachycardia that may influence an anesthesiologist’s perioperative management of pericardial effusions in patients with cancer. ASA status was statistically significantly associated with choice of induction agent and use of preinduction arterial line, whereas other patient acuity factors, such as the presence of tamponade and high initial heart rate, were not associated with anesthetic choices. Although all patients who require pericardial window should be classified as ASA 4 status, more than a third of such patients in our cohort were assigned ASA 3 status. Other important aspects of anesthetic management, such as choice of paralytic agent, dosage of phenylephrine and ephedrine required, and volume of intravenous fluid, were not found to be statistically significantly different between ASA groups, groups with or without tamponade, or groups with different starting heart rates.
We also examined the influence of patient and operative factors on survival outcomes. On multivariable analyses, high initial heart rate and intraoperative sinus tachycardia at the time of presentation to the operating room were independent predictors of worse overall survival. Higher resting heart rate has been shown to be an independent predictor of death in patients with pancreatic, colorectal, and non-small cell lung cancer.14 We observed that every 10-bpm increase in initial heart rate was associated with an 18% increase in risk of death. The higher initial heart rate may be due to a greater degree of external compression of cardiac tamponade at presentation and may represent more-severe systemic oncologic disease or an overall poorer compensatory mechanism. However, preoperative tamponade on echocardiogram alone was not independently associated with overall survival. We also found no statistically significant correlation between hypotension (systolic blood pressure <80 mmHg) and overall survival, which differs from the published literature.15, 16
On univariable analysis, use of propofol or a combination of two induction agents (propofol with etomidate, ketamine with propofol, or ketamine with etomidate) was associated with longer survival, compared with use of etomidate alone. Although etomidate has a favorable hemodynamic profile, a single dose has been associated with adrenal insufficiency, increased hospital/ICU days, and mortality.17, 18 However, this association was not statistically significant on multivariable analysis. Despite the large variation in anesthetic management and the high incidence of tamponade in our cohort, there were no intraoperative deaths.
Despite the availability of treatment options, long-term prognosis remains poor for patients with malignant pericardial effusion. In the literature, median survival after intervention for pericardial effusion ranges from 6 to 40 weeks for patients with cancer.4, 6, 8–12 Sanchez-Enrique et al. reported that, at 10 years postintervention, 89% of patients in the neoplastic group had died, compared with 48% in the nonneoplastic group.5 In the present study, median survival was 23.4 weeks (5.84 months). Previous studies have shown that the presence of malignant cells in pericardial effusion4, 6, 11, 13 and the site of primary cancer6, 7 are the factors most commonly associated with an increased risk of death after an intervention for malignant pericardial effusion. In our study, neither the presence of malignant cells in pericardial effusion nor the site of primary cancer was found to be associated with 30-day mortality. This is consistent with the findings of Jeon et al.,11 who found that overall survival for patients with cancer with pericardial effusion requiring intervention was not associated with the site of primary malignancy.
Unsurprisingly, metastatic disease was associated with shorter survival on univariable analysis. Studies have found that pericardial metastasis is associated with worse survival in patients with pericardial effusion, but no association has been found between presence of distant metastasis and survival in these patients.11, 12 Age was not associated with prognosis in our study or in the study from Jeon et al.11 In contrast, Celik et al. found an association between increasing age and shorter survival.12 We showed that high initial heart rate and intraoperative sinus tachycardia were associated with worse survival after pericardial window creation. Similarly, Anker and colleagues14 showed that increased resting heart rate in patients with oncologic tumors, such as colorectal and pancreatic tumors, compared with healthy patients, was a predictor of poor survival. Elevated resting heart rate was also associated with a higher rate of recurrence of adenoma in colorectal cancer survivors.19
The current study has several limitations. First, this is a retrospective observational study and is therefore prone to misclassification or recall bias, although all data were collected from a prospectively maintained electronic health record. Second, the present study also did not collect information on less-well-studied variables affecting survival, such as presence of a mass on computed tomographic scan3 or postoperative paradoxical hemodynamic instability after decompression.6 Finally, with the small number of deaths at 30 days (n=29), our final multivariable analysis of 30-day mortality may be underpowered. Although difficult to obtain, larger sample sizes who have undergone this uncommon surgical procedure would be needed. The major contribution of this study is that it not only contains one of the largest cohorts of oncologic patients with pericardial effusion but also examines perioperative anesthetic management, which has not previously been studied.
Conclusions
Pericardial effusion in patients with cancer remains a sign of poor prognosis associated with high mortality. Anesthetic perioperative management was affected by ASA status but ultimately did not influence the survival of oncologic patients with pericardial effusions.
Supplementary Material
Acknowledgments:
David B. Sewell, of the Department of Surgery, Memorial Sloan Kettering Cancer Center, provided editorial assistance.
Funding:
This work was supported, in part, by the National Institutes of Health/National Cancer Institute (Cancer Center Support Grant P30CA008748). The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Institutes of Health. The funder played no role in any aspect of the study.
Footnotes
Declarations of Interest: Dr. Afonso is a consultant with Pacira and Merck.
References
- 1.Burazor I, Imazio M, Markel G, et al. Malignant pericardial effusion. Cardiology 2013;124:224–232. [DOI] [PubMed] [Google Scholar]
- 2.Maisch B, Ristic A, Pankuweit S. Evaluation and management of pericardial effusion in patients with neoplastic disease. Prog Cardiovasc Dis 2010;53:157–163. [DOI] [PubMed] [Google Scholar]
- 3.Mirhosseini SM, Fakhri M, Mozaffary A, et al. Risk factors affecting the survival rate in patients with symptomatic pericardial effusion undergoing surgical intervention. Interact Cardiovasc Thorac Surg 2013;16:495–500. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Gornik HL, Gerhard-Herman M, Beckman JA. Abnormal cytology predicts poor prognosis in cancer patients with pericardial effusion. J Clin Oncol 2005;23:5211–5216. [DOI] [PubMed] [Google Scholar]
- 5.Sanchez-Enrique C, Nunez-Gil IJ, Viana-Tejedor A, et al. Cause and long-term outcome of cardiac tamponade. Am J Cardiol 2016;117:664–669. [DOI] [PubMed] [Google Scholar]
- 6.Wagner PL, McAleer E, Stillwell E, et al. Pericardial effusions in the cancer population: prognostic factors after pericardial window and the impact of paradoxical hemodynamic instability. J Thorac Cardiovasc Surg 2011;141:34–38. [DOI] [PubMed] [Google Scholar]
- 7.Takayama T, Okura Y, Okada Y, et al. Characteristics of neoplastic cardiac tamponade and prognosis after pericardiocentesis: a single-center study of 113 consecutive cancer patients. Int J Clin Oncol 2015;20:872–877. [DOI] [PubMed] [Google Scholar]
- 8.Mizukami Y, Ueda N, Adachi H, et al. Long-term outcomes after video-assisted thoracoscopic pericardiectomy for pericardial effusion. Ann Thorac Cardiovasc Surg 2017;23:304–308. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Apodaca-Cruz A, Villarreal-Garza C, Torres-Avila B, et al. Effectiveness and prognosis of initial pericardiocentesis in the primary management of malignant pericardial effusion. Interact Cardiovasc Thorac Surg 2010;11:154–161. [DOI] [PubMed] [Google Scholar]
- 10.Kazantzis T, Bibas BJ, Dela-Vega AJ, et al. Predictors of hospital discharge in cancer patients with pericardial effusion undergoing surgical pericardial drainage. J Surg Oncol 2019;119:143–147. [DOI] [PubMed] [Google Scholar]
- 11.Jeon HW, Cho DG, Park JK, et al. Prognostic factors affecting survival of patients with cancer-related pericardial effusion managed by surgery. World J Surg Oncol 2014;12:249. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Celik S, Celik M, Aydemir B, et al. Surgical properties and survival of a pericardial window via left minithoracotomy for benign and malignant pericardial tamponade in cancer patients. World J Surg Oncol 2012;10:123. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.He B, Yang Z, Zhao P, et al. Cytopathologic analysis of pericardial effusions in 116 cases: Implications for poor prognosis in lung cancer patients with positive interpretations. Diagn Cytopathol 2017;45:287–293. [DOI] [PubMed] [Google Scholar]
- 14.Anker MS, Ebner N, Hildebrandt B, et al. Resting heart rate is an independent predictor of death in patients with colorectal, pancreatic, and non-small cell lung cancer: results of a prospective cardiovascular long-term study. Eur J Heart Fail 2016;18:1524–1534. [DOI] [PubMed] [Google Scholar]
- 15.Mascha EJ, Yang D, Weiss S, et al. Intraoperative mean arterial pressure variability and 30-day mortality in patients having noncardiac surgery. Anesthesiology 2015;123:79–91. [DOI] [PubMed] [Google Scholar]
- 16.Monk TG, Bronsert MR, Henderson WG, et al. Association between intraoperative hypotension and hypertension and 30-day postoperative mortality in noncardiac surgery. Anesthesiology 2015;123:307–319. [DOI] [PubMed] [Google Scholar]
- 17.Chan CM, Mitchell AL, Shorr AF. Etomidate is associated with mortality and adrenal insufficiency in sepsis: a meta-analysis. Crit Care Med 2012;40:2945–2953. [DOI] [PubMed] [Google Scholar]
- 18.Hildreth AN, Mejia VA, Maxwell RA, et al. Adrenal suppression following a single dose of etomidate for rapid sequence induction: a prospective randomized study. J Trauma 2008;65:573–579. [DOI] [PubMed] [Google Scholar]
- 19.Park J, Kim JH, Park Y, et al. Resting heart rate is an independent predictor of advanced colorectal adenoma recurrence. PLoS One 2018;13:e0193753. [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.