Surgical complications after different therapeutic approaches for locally advanced rectal cancer

Tian-Cheng Zhan, Da-Kui Zhang, Jin Gu, Ming Li

Tian-Cheng Zhan, Jin Gu, Ming Li, Key Laboratory of Carcinogenesis and Translational Research (Ministry of Education), Department of Colorectal Surgery, Peking University Cancer Hospital and Institute, Beijing Cancer Hospital, Beijing 100142, China

Da-Kui Zhang, Department of General Surgery, China-Japan Friendship Hospital, Beijing 100029, China

Abstract

Key words:Rectal cancer; Radiotherapy; Surgical complications; Total mesorectal excision; Anastomotic leakage

INTRODUCTION

Patients with locally advanced rectal cancer were recommended to receive neoadjuvant radiochemotherapy, especially for those with positive circumferential resection margin or extensive nodal involvement. It can improve local control for these patients.The rate of local recurrence has decreased significantly as a result of neoadjuvant radiochemotherapy. Some researchers believe that preoperative radiotherapy can improve survival of patients with resectable rectal cancer. It is suggested that patients who are sensitive to radiotherapy can achieve better prognosis. Approximately 15% of patients can achieve complete response after long-course neoadjuvant radiochemotherapy[1]. Different protocols for short-course radiotherapy that consists of 30 Gy in 10 fractions are recommended by the Chinese Anti-Cancer Association[2]. The biological equivalent dose of the short-course radiotherapy is similar to the commonly used regimen (5 × 5 Gy). Although the tumor regression is not as good as with longcourse radiochemotherapy, with ＜ 5% complete response rate, the advantages are no surgical delay, reduced toxicity from capecitabine, and avoidance of overtreatment of non-responders. However, some surgeons believe that preoperative radiochemotherapy increases surgical complications. Anastomotic leakage is thought to be associated with malnutrition resulting from radiotherapy[3]. Perineal wound complications after abdominoperineal resection (APR) are also considered to be associated with tissue edema and infection caused by radiotherapy. In addition,toxicity of radiochemotherapy may decrease patients’ tolerance to surgery.

The aim of this study was to evaluate surgical complications of patients with locally advanced rectal cancer following different neoadjuvant therapy and radical surgery.We compared the incidence and severity of surgical complications at 30 d after surgery in different groups and the contribution of neoadjuvant therapy to surgical complications.

MATERIALS AND METHODS

Patients

We performed a retrospective consecutive study of 1197 patients with mid-to-low rectal cancer (≤ 10 cm from anal verge) who received low anterior resection and APR at the Peking University Cancer Hospital between 2008 and 2010. Among them, 346 patients were treated with long-course chemoradiotherapy, and 259 received shortcourse radiotherapy. Radical resection was performed in all patients. The remaining 592 patients received total mesorectal excision (TME) immediately after rectal cancer was diagnosed. Surgical complications were evaluated for up to 30 d after discharge from hospital according to Clavien-Dindo classification. The median duration of admission for patients who underwent resection was 19 (range 5-81) d. Among them,197 patients (16.3%) were hospitalized for ＞ 30 d.

Radiotherapy

Two different neoadjuvant radiotherapy regimens were applied. Three hundred and forty-six patients received long-course preoperative radiochemotherapy that consisted of 50 Gy in 25 fractions with capecitabine (825 mg/m2, twice daily) as radiosensitizer.The other 259 patients were treated with short-course radiotherapy that consisted of 30 Gy in 10 fractions. Its biological equivalent dose was 36 Gy, which was close to the dose of 5 × 5 Gy radiation (37.5 Gy).

Surgery

TME was the standard approach for surgical treatment of rectal cancer. All patients underwent laparotomy at Beijing Cancer Hospital at 6-8 wk after long-course radiochemotherapy or 7-10 d after short-course radiotherapy. Low anterior resection(LAR) was performed in 894 patients. Temporary ileostomy was performed based on the pathological conditions during the operation. APR was performed in 303 patients.

Surgical complications

Surgical complications were evaluated using predetermined conditions of common complications (Table 1). The main complications are anastomotic leakage and perineal wound complications. The definition of anastomotic leakage was different from those in the literature. It was confirmed by detection of fluid collection through the drainage tubes. Digital rectal examinations were used to evaluate the size of the leakage.Computed tomography was not routinely performed unless puncture drainage or surgical reintervention was needed. The severity of these complications was evaluated by Clavien-Dindo classification (Table 2).

Statistical analysis

The association between neoadjuvant radiotherapy and surgical complications was analyzed using two-sidedχ2or Fisher’s exact test. The two key complications,anastomotic leakage and perineal wound complications, were also evaluated. The clinical variables included general information about the patients and tumor characteristics, as well treatment-related variables such as diverting ileostomy.Logistic regression was performed to investigate the independent factors associated with anastomotic leakage and perineal would complications.P＜ 0.05 was considered as statistically significant.

RESULTS

Groups and patient characteristics

A total of 1197 patients with locally advanced rectal cancer who received LAR and APR were analyzed. They all underwent laparotomy. The patients were divided into three groups according to different preoperative therapy (Figure 1). Group 1：346 patients treated with preoperative long-course chemoradiotherapy followed by TME 6-8 wk after. Group 2：259 patients were treated with short-course radiotherapy (10 ×3 Gy) followed by TME 7-10 d after. Group 3：592 patients received radical surgery only. Patient and tumor characteristics are summarized in Table 3. The medianduration of admission for patients who underwent resection was 19 (range 5-81) d.One hundred and ninety-seven patients (16.3%) stayed in the hospital for ＞ 30 d.

Table 1 Definition of postoperative surgical complications (during admission and 30 d thereafter)

Treatment-related postoperative complications

Eight hundred and ninety-four patients underwent LAR, and 303 patients received APR. There were no deaths within 30 d after surgery. Forty-three patients required surgical reintervention. We analyzed 12 different complications, including anastomotic leakage, perineal wound complications, ileus, bleeding, intra-abdominal abscess, abdominal wound complications, urological complications, intestinal necrosis, gastrointestinal perforation, intravenous line infection, stoma complications,and general complications. Anastomotic leakage and perineal wound complications were the two major complications after resection. The severity of postoperative complications is summarized in Table 4. There were no significant differences in the grade of treatment-related complications except for perineal wound complications.Higher grade of perineal wound complication was observed in patients following short-course radiotherapy.

In 894 patients who received LAR, anastomotic leakage was the most obvious complication. Anastomotic leakage developed in 48 (5.4%) patients. Nineteen (2.1%)patients who required surgical reintervention were classified as Grade 3b according to the Clavien-Dindo classification. Our data suggested that preoperative radiotherapy(P= 0.001) and diverting ileostomy (P＜ 0.001) were significant independent factors(Table 5). Both long-course [odds ratio (OR) = 3.624, 95% confidence interval (CI)：1.689-7.775,P= 0.001] and short-course (OR = 5.150, 95%CI：1.828-14.515,P= 0.002)neoadjuvant radiotherapy increased the incidence of anastomotic leakage (6.78%,5.96%, and 4.54% in Groups 1, 2, and 3, respectively), but neither was associated with the severity of the complication (P= 0.411) (Table 4). Temporary diverting ileostomy was a protective factor to reduce the incidence of anastomotic leakage (OR = 6.211,95%CI：2.525-15.385,P＜ 0.001). The majority of patients with neoadjuvant radiotherapy underwent additional surgery of temporary ileostomy, especially in those with short-course radiotherapy (69.4%, 83.1%, and 7.4% in Groups 1, 2, and 3,respectively,P＜ 0.001).

Three hundred and three patients received APR. More than 16.5% of patients suffered from perineal wound complications. The incidence of perineal wound complications in the three groups was 11.8%, 26.8%, and 9.4%, respectively. Shortcourse chemoradiotherapy was closely associated with perineal wound complications(OR = 5.565, 95%CI：2.203-14.057,P＜ 0.001). In contrast, long-course radiochemotherapy did not significantly influence development of perineal wound complications (OR = 1.692, 95%CI：0.651-4.394,P= 0.280) (Table 6). The grade of these complications differed significantly among the three groups (P＜ 0.001) (Table 4).Patients receiving short-course radiotherapy had higher-grade perineal wound complications. However, there were no Grade 3b perineal wound complications, andnone of these patients required surgical reintervention.

Table 2 Clavien-Dindo classification of surgical complications

Reintervention

Among all the studied patients, only 43 with Grade 3b complications required reintervention. The reasons included anastomotic leakage, ileus, bleeding, intraabdominal abscess, abdominal wound complications, urological complications, and intestinal necrosis. Some patients with anastomotic leakage required reintervention.The reintervention rate for anastomotic leakage repair in in all three groups did not differ significantly (37.5%, 66.7%, and 30.4% in Groups 1, 2, and 3, respectively,P=0.411), indicating that neither long-course nor short-course radiotherapy increased the need for reintervention. The increase in perineal wound complications caused by short-course radiotherapy was mild. None of the patients with perineal wound complications required reintervention.

DISCUSSION

Neoadjuvant radiochemotherapy has become important to reduce local recurrence of locally advanced rectal cancer. In China, an increasing number of patients have been recommended to receive radiotherapy before surgery, but an increase in postoperative complications if patients receive preoperative radiation has been a major concern[4]. The present study compared the major postoperative complications associated with long-course and short-course radiotherapy followed by TME in a large series of patients with locally advanced rectal cancer. Compared to those without radiotherapy, the increase in surgical complications caused by two different preoperative radiotherapy regimens was acceptable, and postoperative mortality did not increase. No patients died within 30 d after surgery, although preoperative radiotherapy might have been associated with anastomotic leakage and perineal wound complications. Temporary ileostomy prevented the occurrence and severity of anastomotic leakage. The grade of surgical complications did not differ significantly,except for perineal wounds, which did not always require surgical reintervention.

Several studies have investigated whether preoperative radiotherapy increases surgical complications[5]. The conclusions were not in agreement. Most of these studies have suggested that preoperative radiotherapy does not increase postoperative morbidity[6]. However, the complications in patients with neoadjuvant radiotherapy seem to be more severe, as demonstrated by the need for more surgical reintervention to treat the complications[7]. Our study indicated that long-course and short-course chemoradiotherapy were not associated with increased incidence or grade of complications. Conservative measures do not have any benefit after radiotherapy[8].Reintervention is more often used for patients who receive radiotherapy if anastomotic leakage cannot be healed. This was one of the major setbacks ofradiotherapy.

Figure 1 Flow diagram of treatment.

Anastomotic leakage is the most serious surgical complication after LAR for rectal cancer. It occurs in 3.5%-25.0% of patients after surgery[9,10]. The rates reported varied according to different definitions being used. In some studies, anastomotic leakage was diagnosed by computed tomography, magnetic resonance imaging, or radiography. The incidence rates were usually higher by imaging diagnosis than clinical observation. In this study, we defined leakage by the presence of gas or feces around the anastomosis in the drainage tubes. Radiology was used only when surgical reintervention was required. There were many risk factors believed to be associated with anastomotic leakage, such as male gender, lower location of the tumor, and preoperative radiotherapy[11,12]. Neoadjuvant radiotherapy has been implicated as a causative factor for the increased rate of anastomotic leakage[13]. It might be due to the local effect of radiation, and subsequently increased technical difficulty during the operation. Radiotherapy may also decrease the oxygen supply to the local tissue around the anastomosis. It can slow down the healing process and cause leakage. In our study, the rate for all these patients was 5.4%. It was a low incidence and within the acceptable range compared with other complications.

It is agreed that preoperative radiotherapy is associated with an increase in anastomotic leakage. Patients with radiotherapy may need a longer time for healing of leakage. As a result, a temporary defunctioning ileostomy was constructed during rectal surgery for patients who underwent preoperative radiotherapy, and it was reversed after 6 mo[14]. It is believed that ileostomy can reduce anastomotic leakage[15].In the present study, more patients with radiotherapy had a defunctioning ileostomy than those who underwent surgery alone. Defunctioning ileostomy also decreases the grade of leakage as the feces are diverted[16]. This may be an effective approach to avoid surgical reintervention caused by anastomotic leakage.

Perineal wound complications are commonly seen after APR. Previously reported rates varied between 5.9% and 31.0%[17,18]. Most perineal wound complications, such as wound gaping, are mild and do not require a prolonged stay in hospital or surgical reintervention. However, severe wound complications can impair quality of life[19]. For example, patients with open wounds are usually accompanied with pain and movement limitation. They might also delay subsequent adjuvant chemotherapy,which may result in worse prognosis. Previous studies have shown that the rate of perineal wound complications increases in patients who receive preoperative radiotherapy[4,20]. Our study also showed a significant increase in perineal wound complications after short-course radiochemotherapy compared with patients who received long-course raidochemotherapy and those who did not receive neoadjuvant treatment. In addition, the grade of the complications was also higher in the shortcourse radiotherapy group. The grade of these complications was always 3a.However, no reintervention was required to manage these complications.

Table 3 Patient and treatment characteristics

Although our study had a large patient cohort, it was limited by its retrospective nature. It is difficult to fully evaluate late complications for ＞ 30 d after hospital discharge, and the incidence might have been underestimated.

The use of preoperative radiochemotherapy for rectal cancer has been debated for decades, including indications, methods, TRG, and so on. Preoperative chemoradiotherapy may be affected by several factors, such as carcinoembryonic antigen and histological regression score. Acellular mucin pools are also thought to be a useful predictor for complete response in several studies[21,22], but it is controversial. The association between postoperative complications and mucin pools is worth study.From 2008-2010, which is the recruitment period, we had not begun to detect routinely regression rate and mucin pool in our hospital. Therefore, these data were not collected for this study. It is one of the limitations of our study.

Since the sample includes more 1000 patients, we can compare the local control,survival, and quality of life among different groups. We can compare differences of clinical effect between different preoperative therapies, and additional studies are necessary in the future.

In conclusion, preoperative radiotherapy was associated with two major surgical complications：anastomotic leakage and perineal wound complications. There were no significant differences in other complications. Both long-course and short-courseradiochemotherapy increased the incidence of anastomotic leakage, but the grade remained close to that in patients treated with surgery alone. A temporary defunctioning ileostomy seemed to be an effective method to reduce the risk of anastomotic leakage for patients who received radiotherapy. Short-course radiochemotherapy increased the incidence and grade of perineal wound complications. Reintervention may not be necessary to ameliorate the perineal wound complications as the damage is usually low grade.

Table 4 Postoperative complications (events during admission and 30 d thereafter)

Table 5 Logistic regression analysis of anastomotic leakage

Table 6 Logistic regression analysis of perineal wound complications

ARTICLE HIGHLIGHTS

Research background

Preoperative radiochemotherapy can improve local control of rectal cancer. However, some researchers believe it increases the incidence of surgical complications. Patients with locally advanced rectal cancer receive three different treatments in our hospital, including long-course radiochemotherapy, short-course radiotherapy, and surgery directly. We can compare differences in their postoperative complications

Research motivation

Some surgeons suspect that preoperative radiochemotherapy increases surgical complications,such as anastomotic leakage. As a result, surgeons are more likely to do additional diverting ileostomy for these patients. Our motivation is to determine if radiochemotherapy increases the incidence of complications or only increases the severity of complications. These findings can guide our treatment strategies.

Research objectives

To investigate surgical complications caused by three different preoperative radiotherapy regimens. It includes the incidence and severity of complications.

Research methods

This is a retrospective study. We analyzed 1197 patients with locally advanced rectal cancer between 2008 and 2010. Three hundred and forty-six patients were treated with preoperative long-course radiochemotherapy, and 259 patients received short-course radiotherapy (10 × 3 Gy)before surgery. The remaining 592 patients underwent total mesorectal excision (TME) alone without neoadjuvant therapy. The incidence of surgical complications was evaluated for up to 30 d after discharge from hospital. Severity was also studied according to Clavien-Dindo classification.

Research results

The major complications were anastomotic leakage and perineal wound complications. Both long-course and short-course radiotherapy were associated with incidence of anastomotic leakage, but the severity of anastomotic leakage did not increase in patients following preoperative radiotherapy. Temporary ileostomy can reduce incidence of anastomotic leakage.Compared with TME alone, short-course radiotherapy was associated with an increase in incidence and severity of perineal wound complications. Long-course radiotherapy seemed safe regarding this complication.

Research conclusions

Radiotherapy increased incidence but not severity of anastomotic leakage. Short-course radiotherapy was also accompanied with perineal wound complications. However, intervention appeared unnecessary to ameliorate the complications. The increase of complications seems to be acceptable. Our surgeons are more likely to use diverting ileostomy for patients with preoperative radiotherapy.

Research perspectives

We determined the advantages and disadvantages of preoperative radiotherapy, and this knowledge will inform our selection of different preoperative treatments. Our study is a retrospective study with a large sample size. In our opinion, a prospective randomized controlled study needs to be designed and performed.