Patient characteristics
A total of 157 patients with atrial fibrillation after lung cancer surgery were enrolled in this study, including 86 males (54.8%) and 71 females (45.2%), aged 48-88 (67.31±7.13) years. The average follow-up time of all patients was 11.14±7.70 months, among which 23 patients had complex adverse prognosis events (14.6%). Baseline characteristics are summarized in Table 1. Postoperative atrial fibrillation lasted (1.84, 12.66) hours, with a median of 7.14 hours. Postoperative hospital stay was 5.70±2.58 days. Thoracoscopic surgery was performed in 154 cases (98.1%), and thoracotomy in 3 cases (1.9%). 119 patients (75.8%) underwent surgical resections of one site, and 38 (24.2%) underwent surgical resections of two or more sites. Wedge resection was performed in 26 cases (16.6%), segmental resection in 11 cases (7.0%), lobectomy in 94 cases (59.9%), and combined lobectomy in 26 cases (16.6%). The operation lasted (1.92±1.09) hours, and the closed thoracic drainage tube was retained for 4.58±2.48 days. 138 cases (87.9%) were adenocarcinoma, 17 cases (10.8%) were squamous cell carcinoma, 1 case (0.6%) was small cell carcinoma, and 1 case (0.6%) was large cell carcinoma. 150 patients (95.5%) underwent lung cancer surgery for the first time, and 7 patients (4.5%) underwent lung cancer surgery for the second time. 11 patients (7.0%) were in stage 0, 133 patients (84.7%) in stage I, 7 patients (4.5%) in stage II, 5 patients (3.2%) in stage III, and 1 patient (0.6%) in stage IV.
Table 1 Patients characteristics
Variables
|
N/Median±Deviation
|
%/IQR
|
Female
|
71
|
45.2
|
Age (years)
|
67.31±7.13
|
/
|
Duration of atrial fibrillation (minutes)
|
7.14
|
1.84-12.66
|
Mode of operation (Thoracoscopic surgery)
|
154
|
98.1
|
Surgical site (one)
|
119
|
75.8%
|
Surgical site (left)
|
64
|
40.8
|
Scope of operation
|
|
|
wedge resection
|
26
|
16.6
|
segmental resection
|
11
|
7.0
|
lobectomy
|
94
|
59.9
|
combined lobectomy
|
26
|
16.6
|
Lymph node dissection or sampling (yes)
|
139
|
88.5
|
BMI (kg/m2)
|
24.96±3.50
|
/
|
Postoperative pathological diagnosis
|
|
|
adenocarcinoma
|
138
|
87.9
|
squamous cell carcinoma
|
17
|
10.8
|
small cell carcinoma
|
1
|
0.6
|
large cell carcinoma
|
1
|
0.6
|
The number of lung cancer surgeries (one)
|
150
|
95.5
|
Anticoagulation (yes)
|
137
|
87.3
|
T stage
|
|
|
This
|
11
|
7.0
|
T1a
|
30
|
19.1
|
T1b
|
67
|
42.7
|
T1c
|
33
|
21.0
|
T2a
|
9
|
5.7
|
T2b
|
2
|
1.3
|
T3
|
3
|
1.9
|
T4
|
2
|
1.3
|
N stage
|
|
|
N0
|
150
|
95.5
|
N1
|
4
|
2.5
|
N2
|
2
|
1.3
|
N3
|
1
|
0.6
|
M stage
|
|
|
M0
|
156
|
99.4
|
M1
|
1
|
0.6
|
Neoplasm staging
|
|
|
0 stage
|
11
|
7.0
|
Ⅰ stage
|
133
|
84.7
|
Ⅱ stage
|
7
|
4.5
|
Ⅲ stage
|
5
|
3.2
|
Ⅳ stage
|
1
|
0.6
|
Postoperative hospitalization days
|
5.7±2.57
|
/
|
CHA2DS2-VASC score
|
2.29±2.85
|
/
|
INR
|
0.95±0.10
|
/
|
CEA
|
4.06
|
1.49-3.46
|
HASBLED score
|
1.38±1.00
|
/
|
Number of lymph node dissection groups
|
4.62±2.81
|
/
|
The operation time
|
1.92±1.09
|
/
|
Preoperative lung function
|
2.26±0.59
|
/
|
Preoperative actual/prediction of lung function
|
88.70±18.59
|
/
|
FEV1/FVC
|
81.76±10.49
|
/
|
Preoperative ejection fraction
|
58.21±1.46
|
/
|
E/e '
|
7.51±2.42
|
/
|
Indwelling time of thoracic closed drainage tube
|
4.58±2.48
|
/
|
Maximum tumor diameter (cm)
|
1.98±1.24
|
/
|
PNI on the last day before the surgery
|
50.36±5.20
|
/
|
NLR on the last day before the surgery
|
2.22
|
1.27-2.54
|
PLR on the last day before the surgery
|
132.73±55.06
|
/
|
PNI on the day of the surgery
|
42.08±4.26
|
/
|
NLR on the day of the surgery
|
11.98
|
6.44-13.11
|
PLR on the day of the surgery
|
216.61±96.94
|
/
|
Abbreviations: IQR, interquartile range, BMI, Body Mass Index, CHA2DS2-VASC, age in years, sex, history of congestive heart failure, history of hypertension, history of stroke/transient ischemic symptoms/thromboembolic events, history of vascular disease, history of diabetes mellitus, INR, International Standard Ratio, CEA, Carcinoembryonic antigen, HASBLED, High blood pressure, kidney and liver dysfunction, stroke, bleeding, INR, age, drug or alcohol use, FEV1/FVC, Ratio of forced expiratory pressure to forced vital capacity in one second, E/e’, early transmitral velocity/tissue Dopplermitral annular early diastolic velocity, PNI, Prognostic Nutritional Index, NLR, neutrophil to lymphocyte ratio, PLR, platelet to lymphocyte ratio.
Determination of PNI, NLR, PLR high-low group boundary value
The specific grouping criteria for high and low groups of PNI, NLR, and PLR
The ROC curve analysis results are shown in Figure 1. It can be seen from the figure that the ROC curve area of the PNI on the last day before the surgery and the day of the surgery is less than 0.5, so the patients are grouped according to the median. The ROC curve area of the NLR on the last day before the surgery and the day of the surgery, and the ROC curve area of the PLR on the last day before the surgery and the day of the surgery > 0.5, so we grouped them according to the best cutoff value. Table 2 shows the specific grouping criteria for each group.
Table 2 AUC and grouping values of PNI, NLR, and PLR
Variables
|
AUC
|
P
|
95% CI
|
Optimal cut-off value/median
|
PNI on the last day before the surgery
|
0.421
|
0.225
|
0.289-0.552
|
50.322)
|
NLR on the last day before the surgery
|
0.648
|
0.023
|
0.528-0.768
|
2.261)
|
PLR on the last day before the surgery
|
0.601
|
0.121
|
0.463-0.739
|
173.551)
|
PNI on the day of the surgery
|
0.415
|
0.193
|
0.287-0.543
|
41.702)
|
NLR on the day of the surgery
|
0.539
|
0.555
|
0.418-0.659
|
8.731)
|
PLR on the day of the surgery
|
0.632
|
0.043
|
0.515-0.750
|
195.651)
|
Note: 1) indicate it was grouped according to optimal cut-off value=sensitivity+specificity-1, 2) indicate it was grouped according to the median.
Abbreviations: PNI, Prognostic Nutritional Index, NLR, neutrophil to lymphocyte ratio, PLR, platelet to lymphocyte ratio, AUC, area under ROC curve.
K-M survival curve of patients with atrial fibrillation complicated with adverse prognostic events after lung cancer surgery in different PNI, NLR, and PLR groups
Prognostic events in patients with atrial fibrillation after lung cancer surgery Kaplan Meier survival analysis and log-rank test were used to evaluate the relationship between PNI, NLR, PLR, and prognosis of patients. The results showed that patients with lung cancer who had atrial fibrillation in the high NLR group on the last day before the surgery (P=0.007), the high PLR group on the last day before the surgery (P=0.001), and the high PLR group on the day of the surgery (P=0.022) were more likely to have complex adverse prognostic events. The high PNI group on the last day before the surgery (P=0.150), the high PNI group on the day of the surgery (P=0.076), and the high NLR group on the day of the surgery (P=0.077) were not related to the occurrence of complex adverse prognostic events in patients with lung cancer complicated with atrial fibrillation after surgery, as shown in Figure 2.
Analysis of influencing factors of complex adverse prognostic events in patients with atrial fibrillation after lung cancer surgery
COX univariate regression analysis found that gender, age, thyroid history, history of malignant tumor, history of arrhythmia, operation mode, actual/predicted preoperative pulmonary function, FEV1/FVC, postoperative pathological diagnosis, several lung cancer operations, T stage, the high PNI group on the last day before the surgery (PNI>50.32), the high NLR group on the last day before the surgery (NLR>2.26), the high PLR group on the last day before the surgery (PLR>173.55), the high PNI group on the day of the surgery (PNI>41.70), the high PLR group on the day of the surgery (NLR>8.73), the high PLR group on the day of the surgery (PLR>195.65), the NLR on the last day before the surgery, and the PLR on the last day before the surgery are risk factors of the occurrence of complex adverse prognostic events in patients with atrial fibrillation after lung cancer surgery (P<0.2), as shown in Table 3.
Table 3 COX univariate analysis of complex adverse prognostic events in patients with atrial fibrillation after lung cancer surgery
Variables
|
Univariate analysis
|
HR
|
95% CI
|
P values
|
Gender (male vs. female)
|
1.818
|
(0.767,4.307)
|
0.174*
|
Age (per year)
|
1.061
|
(0.997,1.129)
|
0.060*
|
Duration of atrial fibrillation (per hour)
|
1.008
|
(0.993,1.023)
|
0.311
|
Postoperative hospitalization days (per day)
|
1.018
|
(0.871,1.189)
|
0.824
|
Thyroid History (yes vs. no)
|
2.537
|
(0.751,8.564)
|
0.134*
|
History of Malignant tumor (yes vs. no)
|
2.979
|
(1.009,8.792)
|
0.048**
|
Smoking history (yes vs. no)
|
1.605
|
(0.694,3.712)
|
0.269
|
Alcohol (yes vs. no)
|
0.603
|
(0.215,2.443)
|
0.603
|
History of respiratory diseases (yes vs. no)
|
0.047
|
(0.000,1969.763)
|
0.574
|
History of Heart Disease (yes vs. no)
|
1.623
|
(0.639,4.125)
|
0.309
|
History of arrhythmia (yes vs. no)
|
2.664
|
(0.784,9.054)
|
0.117*
|
History of hypertension (yes vs. no)
|
0.785
|
(0.332,1.852)
|
0.580
|
History of diabetes (yes vs. no)
|
1.628
|
(0.641,4.134)
|
0.305
|
CHA2DS2-VASC score (per unit)
|
1.036
|
(0.896,1.197)
|
0.636
|
INR (per unit)
|
5.454
|
(0.340,87.613)
|
0.231
|
CEA (per ng/ml)
|
0.997
|
(0.951,1.044)
|
0.895
|
HASBLED score (per unit)
|
1.089
|
(0.729,1.626)
|
0.677
|
Family history of cancer (yes vs. no)
|
1.579
|
(0.369,6.764)
|
0.538
|
History of surgery (yes vs. no)
|
0.646
|
(0.280,1.494)
|
0.307
|
Mode of operation (thoracotomy vs. thoracoscopic surgery)
|
4.483
|
(1038,19.353)
|
0.044**
|
Surgical site (more than one vs. one)
|
0.692
|
(0.256,1.870)
|
0.468
|
Surgical site
|
|
|
|
left
|
—
|
—
|
1.000
|
right vs. left
|
1.007
|
(0.441,2.302)
|
0.986
|
left+right vs. left
|
0.000
|
—
|
0.986
|
Scope of operation
|
|
|
|
wedge resection
|
—
|
—
|
0.764
|
segmental resection vs. wedge resection
|
—
|
—
|
0.979
|
lobectomy vs. wedge resection
|
0.637
|
(0.229,1.770)
|
0.387
|
combined lobectomy vs. wedge resection
|
0.481
|
(0.115,2.021)
|
0.318
|
Lymph node dissection or sampling (yes vs. no)
|
0.852
|
(0.252,2.876)
|
0.796
|
Number of lymph node dissection groups (per unit)
|
1.007
|
(0.863,1.175)
|
0.929
|
The operation time (per hour)
|
1.069
|
(0.710,1.610)
|
0.750
|
Height (per m)
|
3.295
|
(0.015,718.241)
|
0.664
|
Weight (per kg)
|
0.992
|
(0.957,1.028)
|
0.653
|
BMI ((per kg/m2)
|
0.952
|
(0.839,1.081)
|
0.449
|
Preoperative lung function (per unit)
|
0.722
|
(0.345,1.510)
|
0.387
|
Preoperative actual/prediction of lung function (per unit)
|
0.982
|
(0.960,1.004)
|
0.109*
|
FEV1/FVC (per unit)
|
0.968
|
(0.931,1.006)
|
0.096*
|
Preoperative ejection fraction (per unit)
|
0.890
|
(0.713,1.112)
|
0.305
|
E/e’ (per unit)
|
0.905
|
(0.742,1.104)
|
0.325
|
Postoperative pathological diagnosis
|
|
|
|
adenocarcinoma
|
—
|
—
|
0.473
|
squamous cell carcinoma vs. adenocarcinoma
|
2.229
|
(0.827,6.008)
|
0.113*
|
small cell carcinoma vs. adenocarcinoma
|
—
|
—
|
0.989
|
large cell carcinoma vs. adenocarcinoma
|
—
|
—
|
0.991
|
Indwelling time of thoracic closed drainage tube (per unit)
|
1.049
|
(0.903,1.219)
|
0.531
|
The number of lung cancer surgeries (not the first time vs. The first time)
|
3.509
|
(0.810,15.199)
|
0.093*
|
Maximum tumor diameter (per unit)
|
1.024
|
(0.757,1.386)
|
0.878
|
Anticoagulation (yes vs. no)
|
2.254
|
(0.303,16.759)
|
0.427
|
T stage
|
1.209
|
(0.919,1.591)
|
0.174*
|
N stage
|
1.152
|
(0.417,3.180)
|
0.784
|
M stage
|
0.049
|
(0.000,35595165795.515570)
|
0.829
|
Neoplasm stage
|
1.250
|
(0.620,2.521)
|
0.533
|
the high PNI group on the last day before the surgery (>50.32 vs.≤50.32)
|
0.537
|
(0.227,1.268)
|
0.156*
|
the high NLR group on the last day before the surgery (>2.26 vs.≤2.26)
|
2.963
|
(1.299,6.761)
|
0.010**
|
the high PLR group on the last day before the surgery (>173.55 vs.≤173.55)
|
3.821
|
(1.571,9.297)
|
0.003**
|
the high PNI group on the day of the surgery (>41.70 vs. ≤41.70)
|
0.456
|
(0.187,1.109)
|
0.083*
|
the high NLR group on the day of the surgery (>8.73 vs. ≤8.73)
|
2.384
|
(0.883,6.436)
|
0.087*
|
the high PLR group on the day of the surgery (>195.65 vs. ≤195.65)
|
2.834
|
(1.117,7.193)
|
0.028**
|
Note: Asterisks (*) indicate P<0.2;Asterisks (**) indicate P<0.05
Abbreviations: CHA2DS2-VASC, age in years, sex, history of congestive heart failure, history of hypertension, history of stroke/transient ischemic symptoms/thromboembolic events, history of vascular disease, history of diabetes mellitus, INR, International Standard Ratio, CEA, Carcinoembryonic antigen, HASBLED, High blood pressure, kidney, and liver dysfunction, stroke, bleeding, INR, age, drug or alcohol use, BMI, Body Mass Index, FEV1/FVC, Ratio of forced expiratory pressure to forced vital capacity in one second, E/e’, early transmitral velocity/tissue Dopplermitral annular early diastolic velocity, PNI, Prognostic Nutritional Index, NLR, neutrophil to lymphocyte ratio, PLR, platelet to lymphocyte ratio.
To avoid the absence of important influencing factors, the influencing factors of univariate analysis P<0.2 are included in the COX regression model for forward LR analysis. COX multivariate regression analysis found that when PNI, NLR, and PLR were analyzed by categorical variables, the history of malignant tumor and the high PLR group on the last day before the surgery (PLR>173.55) were independent risk factors for the occurrence of complex adverse prognostic events in patients with atrial fibrillation after lung cancer surgery, as shown in Table 4.
Table 4 COX multivariate analysis of complex adverse prognostic events in patients with atrial fibrillation after lung cancer surgery
Risk factors
|
B
|
SE
|
Wald
|
P
|
HR
|
95% CI
|
History of malignant tumor (yes vs. no)
|
1.266
|
0.563
|
5.066
|
0.024
|
3.548
|
1.178-10.690
|
the high PLR group on the last day before the surgery (>173.55 vs. ≤173.55)
|
1.441
|
0.460
|
9.808
|
0.002
|
4.225
|
1.715-10.410
|