Therapeutic evaluation of postoperative chemotherapy and chemoradiotherapy on cardiopulmonary complications in patients with esophageal squamous cell carcinoma

Background : Postoperative chemotherapy (pCT), together with chemoradiotherapy (pCRT), is well used for the treatment of squamous cell carcinoma in the esophagus. However, these adjuvant therapies may cause serious adverse effects resulting in higher postoperative morbidity. Methods : In this study, we aim to assess and compare the postoperative cardiopulmonary complications as a result of pCT and pCRT in patients with esophageal cancer. PCT ( n = 182) or pCRT ( n = 153) was conducted to treat patients with esophageal cancer. Results : A significantly higher percentage of pneumonia was diagnosed in patients with the pCRT treat-ment compared with pCT group ( P < 0.01). In comparison to the pCT treated group, pCRT treatment patients showed significantly decreased early diastolic filling velocity ( P < 0.01) and higher plasma level of N-terminal pro - brain natriuretic peptide (NT - proBNP) ( P < 0.01). Moreover, pCRT treatment was associated with higher pneumonia risk and increased NT - proBNP level significantly by a multivariate analysis. Conclusion : pCRT treatment increased the percentage of cardiopulmonary complications in comparison with pCT treatment.

sophageal cancer causes the highest mortality rate in the whole world (1,2). Surgical resection, together with postoperative chemotherapy (pCT) or radiotherapy (pCRT), is the present treatment strategy for locoregional esophageal squamous cell carcinoma (ESCC) (3). The 5-year overall survival (OS) percentage of patients after surgery is around 20%-50% (4,5). Due to developments in anesthesia, imaging, and surgical procedures recently, the surgical resection rate has improved, and the morbidity and mortality of surgery have decreased significantly. However, as the OS of resectable ESCC patients has not been markedly increased, more than 50% of patients eventually decease from metastasis or recurrence of tumor (6,7). Regarding the merits of postoperative therapy, controversial results have been reported in limited studies for patients with esophagectomy (8)(9)(10). Importantly, adenocarcinoma prevails in developed countries; in contrast, ESCC is one of the main histological types of esophageal carcinoma. Thus treatment protocols including guidelines from the National Comprehensive Cancer Network may not be suitable for ESCC cases. Hence, it is very important to investigate the effects of present postoperative adjuvant therapies on patients carrying resectable ESCC.
Regarding long-term survival, oncological values of radiotherapy have been compromised by a significant higher mortality rate caused by severe adverse events as well as postoperative complications within the chemoradiotherapy (CRT) arm in the first year after randomization. Additionally, postoperative complications caused by chemoradiotherapy were more serious (11). These results suggest that the application of radiotherapy seems to be affecting the heart and lungs adversely and significantly, may be because of the radiation. Pneumonia is known to 2 be a severe medical complication as well as a severe surgical complication after esophagectomy (12). Pneumonia impacts OS negatively after esophagectomy (13). In addition, it has been demonstrated recently that the diastolic and systolic functions of the left ventricle decreased acutely following radiotherapy; however, the results are not consistent (14). Further, N-terminal pro-B-type natriuretic peptide (NT-proBNP), which is widely used as a parameter for heart failure diagnosis and prognosis, has been found to be unregulated following radiotherapy (15,16). The occurrences of cardiac complications following esophagectomy are reported to be within 15%-30% for either chemo-or chemoradiotherapy (17). However, how pCT and pCRT affect the function of the lungs and heart of patients with esophageal cancer after esophagectomy has not been fully studied.

Patient characteristics
A total of 335 patients were enrolled in the study during 2018 to 2021. Patients diagnosed with histologically proven and resectable esophageal carcinoma who were treated with pCT or pCRT were included. However, patients with preexisting diseases in the heart or lungs or prior cardiopulmonary radiotherapy treatment are excluded. Based on adjuvant treatments, 182 and 153 patients were included in the pCT and pCRT treatment groups, respectively. The following data were collected from patients in each group: therapeutic information, tumor characteristics, demographic characteristics, comorbidity, adjuvant treatment, aroused complications, and survival data. This study was conducted under the guidelines in the approved protocol by the Ethics Committee of University de Brasilia. Written information and informed consent were provided to all patients.

Surgery
Esophagectomy was conducted in all patients. Every patient with a cardiopulmonary history was evaluated by cardiologists or pulmonologists for perioperative therapeutic care. Two proficient surgeons performed transthoracic esophagectomy including two-field lymphadenectomy in all patients. A left thoracolaparotomy was applied to reach tumors around the gastroesophageal junction, while a right-sided procedure was utilized to reach more esophageal tumors in the cranial location.

Adjuvant treatment
Chemotherapy was applied for three cycles in 21 days. Cisplatin with a dose of 100 mg/m 2 was taken by patients on the first day and 5-fluorouracil with a dose of 750 mg/ m 2 per 24 h was taken on day 1-5 post esophagectomy.
Cisplatin was replaced by carboplatin or oxaliplatin to avoid renal dysfunction or hearing impairment. A reduction in dose was allowed if side effects occur. For chemoradiotherapy treatment, a total dose of 40 Gy radiotherapy was concomitantly administered in the second and third cycles with 2 Gy per day and 5 fractions/week. A 3D treatment planning system based on computer tomography (EclipseTM, Varian Medical systems, Palo Alto) was used for radiotherapy planning. A multiple-field technique was used to administrate the treatment with optimized direction of beam entry and weights so as to achieve an even dose to the planned target volume as well as minimize the dose in unspecific risking organs (18).

Outcomes
Complications in hospital admission post esophagectomy were sorted into cardiac, pulmonary, or other complications. The criteria described previously were used to score complications, except for that pneumonia was supplemented by antibiotic use with clinical indications (19). The score of the American Society of Anesthesiologists (ASA) was applied to classify the comorbidity, with ASA 1 for very good condition and ASA 5 for a moribund patient (20). All patients were subjected to a cardiac examination. The parameters of echocardiography were selected according to the guidelines of the European Association of Echocardiography. More than one loops with three heartbeats in a role were documented online as a view and the best heart beat cycle was selected for examination post processing. Simpson biplane was used to calculate ejection fraction (EF). Pulsed wave Doppler was used to measure mitral inflow. Pulsed Doppler was used to measure E-velocity, which is the peak velocity during the early rapid filling phase through ventricle relaxation as well as A-velocity, which is the peak velocity in the late filling phase causing by atrial contraction across the mitral valve through diastole. For the assay of NT-proBNP levels in the serum, venous blood was drawn and collected in EDTA tubes. Then a NT-proBNP ELISA Kit (Biocompare, CA) with a detection limit of 0.31 ng/ mL was used for this assay according to the manufacturer's protocol.

Statistical analysis
Data were shown as mean ± standard deviation and analyzed with SPSS software. Categorical variables were demonstrated as frequency with percentage analyzed by Fisher's exact test or chi-squared test where appropriate. Student's t-test was used for comparisons between the two groups. Multivariate analysis was performed to determine the effect of postoperative complications by pCRT.

Patient characteristics
In total, 335 patients (n = 182 and 153 for pCT and pCRT groups, respectively) were included in this study. Table 1 shows the clinicopathological parameters of all patients. No significant differences in disease-specific and demographic characteristics were observed between the two groups. Patients matching in ischemic heart diseases, medical history, gender, age, body mass index (BMI), tumor type and localization, clinical T stages and ASA classification were considered for comparisons. Table 2 shows that 30-day mortality did not have a significant difference between pCRT (5.0%) and pCT (4.9%) groups. Additionally, there was no significant difference in general morbidity rates between pCRT (76.5%) and pCT (73.1%) groups.

Impact of pCRT and pCT on pulmonary complications
Within both groups, cases of pulmonary complications occurred frequently. The most commonly observed complication was pneumonia, which occurred more frequently in the pCRT group (56.2%) compared to the pCT group (25.8%) (P < 0.01). As shown in Table 2, there were no significant differences in other kinds of pulmonary complications, such as acute respiratory distress syndrome (ARDS), atelectasis, pleural effusion, pulmonary embolism, respiratory failure, and reintubation.

Impact of pCRT and pCT on cardiac complications
As shown in Table 2, no significant differences in EF were observed with postoperative treatment in both groups. In the pCRT group, E velocity decreased significantly (P < 0.01) and A velocity was unaltered in comparison to the pCT group. Additionally, in the pCRT group, E/A ratio was reduced significantly (P < 0.01) due to a significant reduction in E velocity. Moreover, a significant increase of NT-proBNP was observed after adjuvant treatment in the pCRT group than in the pCT group (P < 0.01). Thirty-eight arrhythmia cases occurred in 24.8% patients in the pCRT group versus 41 cases in 22.5% patients in the pCT group with no significant differences. There were no statistically significant differences in other complications, including systemic inflammatory response syndrome (SIRS), anastomotic leakage, sepsis, chylothorax, renal failure, wound infections, and liver failure.

Multivariate analysis
The multivariate analysis was performed including pneumonia, pleural effusion, NT-proBNP level, and

4
arrhythmia so as to determine the effects of postoperative complication occurrences by pCRT. Table 3 shows that pCRT treatment was significantly and positively associated with pneumonia development (OR 3.74, CI 1.36-6.85, P < 0.01) as well as NT-proBNP plasma level (OR 2.65, CI 1.12-5.37, P < 0.01). No significant associations were observed between pCRT and pleural effusion or arrhythmia.

Discussion
After oesophagectomy, 30%-50% of patients developed postoperative complications (21). Although surgical procedures as well as perioperative management have made some improvements, esophagectomy has been reported as associated with increased morbidity and high mortality rate (22). A likely negative prognostic effect on postoperative complications in the respiratory system was suggested by Kinugasa et al. (23). Other reports have indicated that there are negative impacts on postoperative complications with long-term effects (24). In addition, assessment of cardiac characteristics with adjuvant chemoradiotherapy during the esophageal cancer treatment is important and necessary because they could potentially direct anesthetic management implications. Our study demonstrates a bigger postoperative pneumonia incidence with the treatment of pCRT in comparison with pCT. Our results also indicated that pCRT applied for esophagus cancer treatment causes an acute heart function impairment, but no similar deficiency was seen by pCT treatment. So far, the function of adjuvant CRT in the treatment of esophageal cancer has been investigated and debated for decades. Part of the prolonged discussion was focused on the generation of adverse cardiopulmonary events and thus a higher postoperative morbidity risk and mortality rate (25), inflammation, and anastomotic leakage also occurred in addition to cardiopulmonary complications (26). Several studies have reported the pathophysiological association between the dose of radiation and lung

5
functions or pulmonary complication risk. (27,28). Wang et al. (28) reported that only the lung volume spared from ≥5 Gy doses was an independent dosimetric parameter for pneumonia risk or occurrence of ARDS. It is suggested that lower radiotherapy doses within multimodal treatments lead to irradiated lung volume minimization and might diminish the pulmonary complication risk. The dose of radiation on the heart in adjacency with esophageal cancer causes substantial effects in spite of the comparatively low total dose of radiation. Radiation on the pericardium upregulates the pericardial effusion risk, seemingly dose dependently (29). Lately, Bosch et al. observed a cardiopulmonary complication induction in adjuvant pCRT treated patients (17). Consistent with prior studies, in the current study, we observed significantly more pneumonia cases occurred in the pCRT group in comparison to the pCT group. Moreover, pCRT was found associated with a higher pneumonia risk significantly. Hence, it is important to decrease the radiation amount on cardiopulmonary organs to avoid compromising the favorable effects of radiotherapy. Advanced radiation tools utilizing intensity-modulated radiotherapy improve and are promising, with further research warranted. Chemoradiotherapy has been reported to decrease the velocities of blood flow across the mitral valve through the fast and passive filling period of the left ventricle (E), with an unaffected blood flow through atrial contraction (A) and as a result a reduced E/A (30). These results suggest a diastolic function deficiency resulting from weakened left ventricle relaxation. Hatakenaka et al. also observed a deficiency of relaxation of the left ventricle following radiotherapy (30). Increased NT-proBNP level was seen with adjuvant chemoradiotherapy. NT-proBNP has been considered to be a predictor for perioperative cardiac complication risk with suggested cutoff values as 201-791 ng/mL (31,32). Pre-and perioperative NT-proBNP level also predict atrial fibrillation even though its cutoff concentrations are in debates (33). Lund's study suggested that adjuvant CRT but not CT prior to esophagus cancer surgery seems to cause an acute adverse effect on systolic and also diastolic functions of the left ventricle (18). In the present study, diastolic alterations were observed to be larger with decreased E/A levels when adjuvant therapy approaching grade I in diastolic dysfunction, which may have medical implications. Levels of NT-proBNP were upregulated after chemoradiotherapy, which may indicate a higher risk for cardiac events and atrial fibrillation post esophagectomy (34). Deficiency in systolic as well as diastolic function of the left ventricle after adjuvant chemoradiotherapy might explain the higher incidences and grades in the innately cardiovascular events after esophagectomy (35). Taken together, among the relevant dedicated studies, our results specially emphasize and further clarify the cardiotoxicity consequences caused by present chemoradiotherapy regimens.
In conclusion, our study demonstrated a higher occurrence of pneumonia with pCRT treatment in patients with esophageal cancer. With multivariate analysis, pCRT was positively and significantly associated with pneumonia risk. Moreover, pCRT instead of pCT post-surgery seemingly induces an acute adverse effect on heart function for treatment of esophageal cancer. Our results emphasize the necessity for cautious pCRT use in treating patients because of the higher risk for cardiopulmonary complications. Further investigations for esophageal cancer adjuvant treatment are suggested in order to add cardiopulmonary function measurements to limit toxicity of the heart and lungs induced by pCRT.