Risk factors for the development of postoperative atrial fibrillation after esophagectomy for esophageal cancer

Abstract

Postoperative atrial fibrillation (POAF) is one of the most complications after esophagectomy. Thus, our study was conducted to explore risk factors that are associated with POAF following esophagectomy. In the current study, we retrospectively evaluated 511 patients with esophageal cancer who underwent esophagectomy at our center between May 2018 and December 2020. The incidence of POAF and clinical variables were collected and analyzed. POAF occurred in 62 (12.13%) of 511 patients. Increasing age (P = .004) and lower preoperative albumin level (P = .028) was significantly associated with POAF. The length of stay was significantly increased in patients with POAF when compared to patients without POAF (P = .003). Multivariate analysis suggested that age (odds ratio [OR] = 1.049, 95% confidence interval [CI] = 1.008–1.093, P = .020), hypertension (OR = 2.207, 95% CI = 1.221–3.987, P = .009), respiratory complications (OR = 2.015, 95% CI = 1.130–3.591, P = .018) and Ivor Lewis approach (OR = 3.001, 95% CI = 1.032–8.723, P = .044) were independent risk factors for POAF following esophagectomy. Increasing age, preoperative hypertension, respiratory complications and Ivor Lewis approach are independent risk factors for POAF after esophagectomy. POAF is associated with prolonged length of stay. This study suggests that older patients, patients with hypertension or patients underwent Ivor Lewis approach should be monitored more closely during the postoperative period.

1. Introduction

Esophageal cancer is the 7th cause of global cancer incidence and 6th in mortality worldwide,^[1] with esophageal squamous cell carcinoma being the most common histological type in Asian countries and adenocarcinoma in western countries.^[2] The main curative treatment for patients with esophageal cancer is esophagectomy.^[3] Despite advances in surgical procedures and perioperative managements, esophagectomy remains correlated with a high incidence of postoperative complications.^[4,5] Complications after esophagectomy include pneumonia, anastomotic leakage, recurrent laryngeal nerve injury, and atrial fibrillation.

A number of studies have reported the incidence of postoperative atrial fibrillation (POAF) ranging from 9.2% to 45.5%.^[6–8] It was reported that POAF potentially increase the risk of thromboembolic events and mortality.^[9] Recent studies have also showed that POAF might be related to other postoperative complications such as pneumonia and anastomotic leakage through decreased organ blood flow.^[10] Therefore, understanding the risk factors of POAF is of clinical importance. Several studies have investigated the risk factors of POAF following esophagectomy.^[6,7,11,12] However, the results were inconsistent because of the different size of study population and analysis methods. A meta-analysis was conducted to investigate the risk factors and subsequent events associated with the onset of POAF following esophagectomy.^[13] The authors found that coronary artery disease and hypertension were associated with POAF. However, these findings should be interpreted with caution because of the high risk of bias and few randomized trials in this study. Here, our study was conducted to explore risk factors that are associated with POAF after esophagectomy.

2. Materials and Methods

2.1. Patient selection

Patients who underwent potentially radical esophagectomy for esophageal cancer at Department of thoracic surgery, Tongji Hospital, Tongji Medical college, Huazhong University of Science and Technology between May 2018 and December 2020 were included. Patients who receive neoadjuvant therapy, R2 resection, with history of atrial fibrillation before the operation or insufficient clinical data were excluded. Eventually, 511 patients were enrolled in this study. Clinical data were retrospectively collected from electronic medical records and reviewed by 2 doctors. The tumor stage was staged according to the 8th edition of the TNM classification of the International Union Against Cancer.

Several Inflammation Indexes and Nutrition Indexes, such as neutrophil-to-lymphocyte ratio (NLR), platelet-to-lymphocyte ratio (PLR), systemic immune-inflammation index (SII), prognostic nutritional index (PNI), and geriatric nutritional risk index (GNRI) were calculated by the relevant clinical variables. The NLR was calculated by diving absolute neutrophil count to absolute lymphocyte count. The PLR was calculated by diving platelet count to absolute lymphocyte count. The SII was defined as follows: platelet count × neutrophil count/lymphocyte count.^[14] The PNI was calculated as 10 × serum albumin level (g/dL) + 0.005 × total lymphocyte count (cells/μL).^[15] The GNRI was calculated using the following equation: 14.89 × serum albumin level (g/dL) + 41.7 × (present body weight/ideal body weight). The ideal body weight was calculated using the following Lorenz equation: height – 100 - [(height – 150)/4] for men, and height – 100 - [(height – 150)/2.5] for women. The units were expressed as kilograms for weight and centimeters for height.^[16]

This retrospective study was approved by the ethics committee of Tongji Hospital, Tongji Medical college, Huazhong University of Science and Technology (TJ-IRB20221308), and was conducted in accordance with the principles of the Declaration of Helsinki. Written informed consent was waived because of the retrospective nature of this study.

2.2. Surgical procedure

Briefly, in the Mckeown approach (cervical, right thoracic and abdominal 3-incision surgery), anastomosis was performed on the left side of the neck. In the Ivor Lewis approach (abdominal and right thoracic 2-incision surgery), anastomosis was performed in the right upper mediastinum. In the Sweet approach (left postero-lateral or left thoraco-abdominal incision), anastomosis was performed in the left thoracic cavity. Minimally invasive esophagectomy was defined as performance of the thoracic and abdominal parts of the surgery by endoscopy.^[17] Any part of surgical procedure without endoscopy was defined as open esophagectomy.

2.3. Definition of complications

New-onset AF was diagnosed as an irregular rhythm without discernible P waves and documented by ECG monitor device. All patients had telemetry monitoring for a minimum of 3 days postoperatively. Treatment for AF included amiodarone, digoxin, or a combination. Anastomotic leakage was defined based on clinical suspicion and confirmation by upper gastroenterography, endoscopy, drainage of ingested materials or saliva into the chest tube or in the cervical wound. respiratory complications were defined as the development of infiltrates on chest radiography with signs of increased inflammation and administration of antibiotics for pneumonia, pleural effusion that needed drainage, pleural empyema and prolonged ventilation for > 5 days or reintubation. Recurrent laryngeal nerve injury was defined as any dysmotility in the vocal cords by fiberoptic bronchoscopy.

2.4. Statistical analysis

The continuous variables are expressed as means ± standard deviation or median. The categorical variables are expressed as frequencies and proportions (count and percent). The patients were divided into 2 groups: the POAF group and no POAF group. Pearson Chi-square test was used to compare the categorical variables and Student t test or Mann–Whitney U test was used to compare the continuous variables between the 2 groups. Multivariate analysis was used to determine risk factors for POAF. All data were analyzed with SPSS version 25.0 software (IBM Corp., Armonk), and P value < .05 was considered significant.

3. Results

The demographics and clinical characteristics of the 511 patients in our study are shown in Table 1. POAF occurred in 62 (12.13%) patients. The median age of the patients was 62.00 years (35–82), with a male preponderance (n = 422, 82.58%). Most patients (n = 443, 86.69%) had esophageal squamous cell carcinoma (n = 443, 86.69%). The majority of patients had a middle third tumor location (n = 331, 64.77%), stage II and III (n = 374, 73.19%). The length of stay was significantly increased in patients with POAF when compared to patients without POAF (P = .003). Increasing age (P = .004) and lower preoperative albumin level (P = .028) was significantly associated with POAF, while histology (P = .408) and TNM stage (P = .669) was not.

Table 1.

Patient characteristics.

	No POAF (N = 449)	POAF (N = 62)	P value
Age (yr)	61.33 ± 7.31	64.21 ± 7.07	.004
Sex			.175
Male	367 (81.74%)	55 (88.71%)
Female	82 (18.26%)	7 (11.29%)
Tumor location			.997
Upper third	42 (9.35%)	6 (9.68%)
Middle third	291 (64.81%)	40 (64.52%)
Lower third	116 (25.84%)	16 (25.81%)
Histology			.408
Squamous cell carcinoma	390 (86.86%)	53 (85.48%)
Adenocarcinoma	50 (11.14%)	9 (14.52%)
Other	9 (2.00%)	0 (0.00%)
TNM stage			.669
I	83 (18.49%)	8 (12.90%)
II	151 (33.63%)	23 (37.10%)
III	176 (39.20%)	24 (38.71%)
IV	39 (8.69%)	7 (11.29%)
Length of stay	29 (12–125)	34.50 (15–108)	.003

POAF = postoperative atrial fibrillation.

Patient comorbidities are displayed in Table 2. Most patients had a smoking history (n = 330, 64.58%) or an alcohol drinking history (n = 268, 52.45%). Smoking history (P = .785), alcohol drinking history (P = .889) and family history of cancer (P = .265) were not associated with onset of POAF. Hypertension was associated with onset of POAF (P = .003), while Diabetes mellitus (P = .366), CAD (P = .377) was not associated with onset of POAF.

Table 2.

Patient comorbidities.

	No POAF (N = 449)	POAF (N = 62)	P value
Hypertension	91 (20.27%)	23 (37.10%)	.003
Diabetes mellitus	23 (5.12%)	5 (8.06%)	.366
Coronary artery disease	12 (2.67%)	0 (0.00%)	.377
COPD	6 (1.34%)	0 (0.00%)	1.000
Others	33 (7.35%)	6 (9.68%)	.518
Smoker	289 (64.37%)	41 (66.13%)	.785
Alcohol drinker	236 (52.56%)	32 (51.61%)	.889
Family history of cancer	84 (18.71%)	8 (12.90%)	.265

COPD = Chronic obstructive pulmonary disease, POAF = postoperative atrial fibrillation.

Patient surgery data are shown in Table 3. Mckeown esophagectomy (n = 322, 63.01%) was the most common approach performed in the cohort, followed by Sweet (n = 168, 32.88%) and Ivor Lewis (n = 21, 4.11%). Among all the patients, open esophagectomy was performed in 295 patients (57.73%) and minimally invasive esophagectomy (MIE) was performed in 216 patients (42.27%). Surgical type was associated with onset of POAF (P = .034). Further analysis suggested that Ivor Lewis approach was associated with onset of POAF when compared with Mckeown approach (P = .010), but not when compared with Sweet approach (P = .074).

Table 3.

Patient surgery data.

	No POAF (N = 449)	POAF (N = 62)	P value
Surgical Approach			.249
OE	255 (56.79%)	40 (64.52%)
MIE	194 (43.21%)	22 (35.48%)
Surgical type			.034
Mckeown	289 (64.37%)	33 (53.23%)
Ivor Lewis	15 (3.34%)	6 (9.68%)
Sweet	145 (32.29%)	23 (37.10%)
Operation time (min)	345 (120–660)	335 (150–750)	.339

MIE = minimally invasive esophagectomy, OE = Open esophagectomy, POAF = postoperative atrial fibrillation.

Postoperative complications are shown in Table 4. The most common postoperative morbidities for patients undergoing esophagectomy were respiratory complications (n = 137, 26.81%), recurrent laryngeal nerve injury (n = 86, 16.83%), atrial fibrillation (n = 62, 12.13%), and anastomotic leakage (n = 54, 10.57%). Respiratory complications were associated with onset of POAF (P = .004), while anastomotic leakage (P = .129) and recurrent laryngeal nerve injury (P = .875) were not.

Table 4.

Postoperative complications.

	No POAF (N = 449)	POAF (N = 62)	P value
Respiratory complications	111 (24.72%)	26 (41.94%)	.004
Anastomotic leakage	44 (9.80%)	10 (16.13%)	.129
Recurrent laryngeal nerve injury	76 (16.93%)	10 (16.13%)	.875

POAF = postoperative atrial fibrillation.

Inflammation Indexes and Nutrition Indexes are displayed in Table 5. We found that the level of albumin was associated with onset of POAF (P = .028), so we investigate whether NLR, PLR, SII, PNI, and GNRI were associated with onset of POAF. The results of Mann–Whitney U test suggested that the inflammation indexes such as NLR, PLR, and SII were not associated with onset of POAF. Regarding nutrition indexes, although the median of PNI and GNRI was lower in the patients with POAF compared with the patients without POAF, the differences were not statistically significant.

Table 5.

Inflammation Indexes and Nutrition Indexes.

	No POAF (N = 449)	POAF (N = 62)	P value
Neutrophil count	3.32 (1.04–15.60)	3.58 (1.07–10.98)	.243
Lymphocyte count	1.54 (0.49–3.50)	1.54 (0.44–3.85)	.851
Platelet count	209 (26–540)	211.5 (25–341)	.941
Albumin	41.4 (28.3–50.9)	40.15 (31.2–48.9)	.028
NLR	2.15 (0.59–24.76)	2.25 (0.72–24.95)	.274
PLR	139.13 (33.33–419.23)	136.68 (15.82–297.73)	.769
SII	440.94 (83.67–5200.00)	467.18 (81.32–3269.05)	.740
PNI	49.30 (32.70–64.75)	48.23 (35.50–63.45)	.078
GNRI	105.13 (78.75–126.43)	104.98 (78.75–126.43)	.263

GNRI = geriatric nutritional risk index, NLR = neutrophil-to-lymphocyte ratio, PLR = platelet-to-lymphocyte ratio, PNI = prognostic nutritional index, POAF = postoperative atrial fibrillation, SII = systemic immune-inflammation index.

Multivariate analysis was used to determine risk factors for POAF (Table 6). Age [odds ratio (OR) = 1.049, 95% confidence interval (CI) = 1.008–1.093, P = .020), hypertension (OR = 2.207, 95% CI = 1.221–3.987, P = .009), respiratory complications (OR = 2.015, 95% CI = 1.130–3.591, P = .018), Ivor Lewis approach (OR = 3.001, 95% CI = 1.032–8.723, P = .044) were independent risk factors for POAF following esophagectomy, while the level of albumin was not (OR = 0.890, 95% CI = 0.773–1.024, P = .103).

Table 6.

Multivariate analysis of risk factors for the onset of POAF.

	Odds ratio	95% CI	P value
Age	1.049	1.008–1.093	.020
Hypertension	2.207	1.221–3.987	.009
Respiratory complications	2.015	1.130–3.591	.018
Ivor Lewis approach	3.001	1.032–8.723	.044

CI = confidence interval, POAF = postoperative atrial fibrillation.

4. Discussion

The incidence of POAF following esophagectomy varies greatly, ranging from 9.2% to 45.5%.^[6–8] Several risk factors for POAF have been investigated, but the results were inconsistent. Our study showed that approximately 1st to 8th (12.13%) of patients in this cohort developed POAF following esophagectomy. The length of stay was significantly increased in patients with POAF when compared to patients without POAF. Multivariate analysis suggested that age, hypertension, respiratory complications and Ivor Lewis approach are independent risk factors for POAF after esophagectomy.

The present study was the first to reveal that Ivor Lewis approach is significantly associated with onset of POAF in esophageal cancer. multivariate analysis suggested that Ivor Lewis approach is an independent risk factor for POAF following esophagectomy. Previous study has shown an association between transthoracic approach and POAF after esophagectomy.^[18] An inflammatory response after surgical trauma near the sympathovagal nerve fibers supplying the heart may induce the onset of POAF after esophagectomy. In the Ivor Lewis approach (abdominal and right thoracic 2-incision surgery), anastomosis was performed in the right upper mediastinum. Inflammation, sympathetic activation and cardiac ischemia are considered to combine to trigger atrial fibrillation.^[19] We speculate that the surgical procedure produces an excessive inflammatory response in the right upper mediastinum, which is nearer the atrial comparing with the Mckeown and Sweet approach, and induce the onset of POAF.

In the current study, minimally invasive esophagectomy was not associated with the onset of POAF compared with open esophagectomy. This result is consistent with the past reports.^[18] Day et al^[20] also investigated the difference incidence of POAF between the 2 groups. They did not find an increased rate of POAF in the minimally invasive esophagectomy group compared with prior reported rates in open esophagectomy populations. On the contrary, Chen et al^[11] reported that MIE was negatively associated with the onset of POAF in a dataset of 575 patients. Different size of study population and analysis methods may be the potential reasons for these inconsistent findings.

Several reports demonstrated that age was a risk factor for the onset of POAF. In our study increasing age was significantly associated with POAF, and multivariate analysis suggested that age was one of independent risk factors for POAF following esophagectomy. This result is consistent with the previous reports. Day et al^[20] investigated 121 patients underwent MIE and found that patients with POAF were significantly older than those without POAF. Mc Cormack et al^[21] also suggested that a significant association was confirmed for age.

We also found that hypertension was an independent risk factor for the onset of POAF after esophagectomy. This result is consistent with the previous reports. Schizas et al^[13] investigated the postoperative complications associated with the onset of POAF after esophagectomy in 53 studies including 9087 patients and found that hypertension was associated with POAF. Recent study also suggested that POAF was significantly associated with preoperative hypertension.^[22] Hypertension enhanced the renin angiotensin aldosterone system and the ensuing cardiac remodeling, which may influence the development of atrial fibrillation postoperatively.^[23] Interestingly, Pu et al^[24] found that an elevated preoperative BNP level was a significant risk factor for POAF.

In the current study, the postoperative respiratory complications were associated with the onset of POAF. This finding is consistent with previous studies. Seesing et al suggested that POAF was associated with pulmonary complications and anastomotic leak.^[10] Schizas et al^[13] found that patients with POAF had a significantly higher risk of postoperative complications than those without POAF, including anastomotic leak and pneumonia. However, we did not find the association between anastomotic leak and POAF following esophagectomy in our study. Additional large population studies should be considered to determine the association between POAF and postoperative complications, especially respiratory complications and anastomotic leak.

Our results suggested that the level of albumin was associated with onset of POAF. However, Multivariate analysis suggested that the level of albumin was not an independent risk factor for POAF. Albumin is an inflammation and nutrition indicator. Zhao et al^[25] reported that low albumin level was independently associated with atrial fibrillation in their cohort with 2000 patients. In our study, we explored whether inflammation and nutrition indexes were associated with onset of POAF. The results suggested that the inflammation indexes such as NLR, PLR and SII were not associated with onset of POAF. Regarding Nutrition indexes, although the median of PNI and GNRI was lower in the patients with POAF compared with the patients without POAF, the differences were not statistically significant. These results suggested that preoperative nutrition may contribute to the onset of POAF in the esophageal cancer patients as a cofactor. Further studies are warranted to identify the role of nutrition indexes in the onset of POAF after esophagectomy in large cohorts.

There are several limitations in the current study. The retrospective nature of data limits the generalizability of the results and conclusions. Although patient characteristics in our study are with those from previously reports, the study sample is confined to a single center. Additionally, subtle factors that may have influenced the onset of POAF such as hypoxia, hypoglycemia were not available in the current database to be analyzed. In order to ascertain the risk factors associated with the onset of POAF, a prospective multicenter study could be designed.

5. Conclusion

Increasing age, preoperative hypertension, respiratory complications and Ivor Lewis approach are independent risk factors for POAF after esophagectomy. POAF is associated with postoperative respiratory complications and prolonged length of stay. This study suggests that older patients, patients with hypertension or patients underwent Ivor Lewis approach should be monitored more closely during the postoperative period.

Acknowledgments

We thank PhD. Xinquan Zhou for her advice in statistical analysis.

Author contributions

Conceptualization: Feng He, Fan Li, Yu Deng.

Data curation: Xi Wu, Ziheng Yang, Dehao Tu, Fan Li.

Formal analysis: Ziheng Yang, Dehao Tu.

Methodology: Dehao Tu.

Validation: Xi Wu, Fan Li.

Writing – original draft: Feng He, Yu Deng.

Writing – review & editing: Feng He, Yu Deng.