Efficacy and safety of chemotherapy after endoscopic double stenting for malignant duodenal and biliary obstructions in patients with advanced pancreatic cancer: a single-institution retrospective analysis

Patients

Fifty consecutive patients with pancreatic cancer who had undergone endoscopic double stenting, comprising DuS and BS, from January 1, 2007 to October 31, 2015 at our institution, were retrospectively analyzed. The records of these patients were reviewed, and the following data were analyzed: patient characteristics, duodenal stenosis sites, timings of double stenting, selected BD methods, stent patency periods, technical and clinical success rates for reintervention, chemotherapy after double stenting, adverse events associated with chemotherapy after double stenting, time from initial diagnosis to double stenting, survival times following double stenting, and overall survival times. The hospital’s institutional review board for human research approved this study.

The duodenal stenosis types were classified according to the location of the stenosis relative to the ampulla of Vater, as follows: type I: proximal to and no involvement of the ampulla of Vater; type II: affecting the second part of the duodenum and the ampulla of Vater; and type III: affecting the third part of the duodenum without ampulla of Vater involvement [5]. In simultaneous type II stenosis cases, if passage of the scope was possible, we deployed the biliary metal stent (MS) or plastic stent (PS) through ERCP-BD, followed by the DuS. If passage of the scope was impossible, we evaluated the duodenal bulb invasion of the tumor to determine whether EUS-choledochoduodenostomy (EUS-CDS) was possible. If possible, we performed EUS-CDS first, followed by the DuS. Otherwise, we selected EUS-hepaticogastrostomy (EUS-HGS) for biliary drainage. In cases of biliary stenosis after DuS for metachronous type II stenosis, we attempted a transpapillary approach through the MS mesh following insertion of the scope into the lumen of the duodenal stent. If the transpapillary approach failed, we performed EUS-BD. EUS-CDS was adopted for cases without duodenal bulb invasion, while EUS-HGS was adopted for duodenal bulb invasion.

The method of reintervention for ERCP-BDs was following, a duodenal scope was inserted through the duodenal stent. PS exchange involved placing a guidewire into the target branch before PS removal, then inserting a 7–10 Fr PS. When tumor ingrowth or overgrowth obstructed the MS, we inserted a 5–7 Fr PS into the lumen of the previously deployed MS. The methods of reintervention for EUS-CDS was following, a duodenal scope or a forward-viewing endoscope was used. PS exchange involved occluded stent removal. Then, a seeking guidewire was passed through the fistula and a new PS was placed. When the MS obstructed with sludge, a stone retrieval balloon was used and removed the sludge.

The stent patency periods for DuS and BS were measured from the day on which double stenting was undertaken to the day on which the stent became dysfunctional, the day the patient died, or the day of the last follow-up appointment. Stent dysfunction included stent obstruction, stent migration, and cholangitis. Obstructive jaundice recurrence, which was based on laboratory examinations and biliary dilation on computed tomography (CT) images, was considered to result from biliary stent obstruction. Symptom recurrence associated with gastroduodenal obstruction, including nausea, vomiting, and difficulties with oral intakes, and a grossly dilated stomach on CT images was considered to be caused by duodenal stent obstruction. Cholangitis was defined as elevations of liver enzyme level and the presence of typical symptoms, including fever. For reintervention, clinical success of biliary stent was defined as a decrease in bilirubin level to < 75% of the pre-drainage levels within 30 days. Clinical success of duodenal stent was based on oral intakes before and after stent placement using the Gastric Outlet Obstruction Scoring System (GOOSS) [16]. Overall survival time was defined as the period between pathological diagnosis and patient death or lost to follow-up.

The basic chemotherapy regimen after double stenting is as follows: gemcitabine (Gemzar, intravenous 1000 mg/m² on days 1, 8, and 15 on a 28-day cycle; Eli Lilly and Company, Indianapolis, IN, USA), S-1 (TS-1, 80–100 mg/day per oral from days 1 to 14 on a 21-day cycle; Taiho Pharmaceutical Co., Ltd., Tokyo, Japan), GEM/S-1 (Gemzar, 1000 mg/m² on days 1 and 8; TS-1, 80–100 mg/day from days 1 to 14 on a 21-day cycle), irinotecan (Topotecin, intravenous 180 mg/m² on a 14-day cycle; Daiichi-Sankyo Co., Ltd., Tokyo, Japan), and modified FOLFIRINOX [mFOLFIRINOX; oxaliplatin 85 mg/m² (Elplat; Yakult Honsha Co., Ltd., Tokyo, Japan), irinotecan 150 mg/m², leucovorin 400 mg/m² (Isovorin injection; Pfizer Inc., New York, NY, USA), and continuous infusion of fluorouracil 2400 mg/m² over 46 h (5-FU injection; Kyowa Hakko Kirin Co., Ltd., Tokyo, Japan), with routine subcutaneous infusion port on a 14-day cycle]. The toxicities were assessed using the National Cancer Institute Common Terminology Criteria for Adverse Events, version 4 (CTCAE v4). Chemotherapy was continued until disease progression, intolerable adverse events, or patient refusal.

Statistical analysis

Continuous variables were expressed as median and range or interquartile range (IQR) or mean ± standard deviation (SD). Chi-square test was performed to analyze categorical variables. Student’s t-test was used to compare continuous variables. The biliary stent patency periods were estimated using the Kaplan-Meier method, and they were compared using the log-rank test. Factors with P values < .05 were considered survival factors after double stenting using the forward method, and they were analyzed in a multiple logistic regression model. Odds ratios (ORs) and 95% confidence intervals (CIs) were calculated. Two-tailed P values < .05 were considered statistically significant. All analyses were performed using JMP Pro 12 (SAS Institute, Cary, NC, USA).

Patient characteristics

Stent dysfunction after double stenting

The mean duodenal stent patency after double stenting was 106 ± 136 days (median not reached). During follow-up (median 73 days, IQR 44–131 days), duodenal stent dysfunction occurred in 6 patients (12%). All patients who experienced stent dysfunction underwent endoscopic reintervention, and all of the procedures (2 patients for overgrowth, 2 patients for food impaction, 1 patient for stent migration) were successful. The GOOSS score increased from 0.6 (SD 0.5) to 2.2 (SD 0.8) points following reintervention (P < .001).

The median biliary stent patency period after double stenting was 230 days (range 17–375 days). During follow-up (median 54 days, IQR 43–118 days), biliary stent dysfunction occurred in 12 patients (24%). The median biliary stent patency period for MS and PS was 270 days (range 17–337) and 81 days (19–375), respectively (P < .005). Of the 12 patients who experienced stent dysfunction (6 patients with MS, 6 patients with PS), 6 had undergone ERCP-BD and 6 had EUS-CDS. The median time of stent dysfunction after double stenting was 87 days (range 25–270 days). All patients who experienced stent dysfunction underwent endoscopic reintervention, and all of the procedures (7 patients for sludge, 3 patients for ingrowth, 1 patient for migration, 1 patient for kinking) were successful. However, one patient experienced clinical failure following ERCP-BD, thus the patient underwent an additional intervention using EUS-HGS. Therefore the clinical success rate was 92% (11/12).

Post double stenting chemotherapy and adverse event

Prior to double stenting, 26 (52%) patients received chemotherapy and 10 (20%) patients were treated with best supportive care (BSC). Both bile duct and duodenal obstruction developed simultaneously in the remaining 14 (28%) patients at the initial diagnosis. After double stenting, 12 (46%) out of 26 patients with chemotherapy could continue another chemotherapy, and 9 (64%) out of 14 patients with simultaneous onset could start initial chemotherapy (Fig. 1). Results showed that 21 (42%) patients received chemotherapy post double stenting; 9 patients received chemotherapy as a first-line treatment, 9 as a second-line treatment, and 3 as a third-line treatment. The regimens were GEM in 8 patients, S-1 in 8 patients, CPT-11 in 2 patients, mFORFIRINOX in 2 patients, and GEM+S-1 in one patient. During chemotherapy, 8 (38%) patients had grade 3–4 adverse events: neutropenia in 4, appetite loss in 2, and diarrhea in 2. However, these adverse events were manageable by dose reduction, drug holiday, or temporary drip. No fatal adverse events associated with chemotherapy were noted.

On the other hand, 29 (58%) out of 50 patients who underwent double stenting received BSC because of contraindications to systemic chemotherapy. These contraindications include poor general condition in 22, patient refusal in 5, and refractory to the standard treatment (GEM, TS-1, or mFOLFIRINOX) in 2 patients.

Prognostic factors after double stenting

Comparison of survival times and prognostic factors

The median survival time after double stenting was significantly longer in patients with reintervention for BS (189 days: IQR 115–325) than that in patients without reintervention (52 days: IQR 40–119) (P = 0.0025) (Fig. 2). It was also longer in patients with PS <  2 (155 days: IQR 77–325) than that in patients with PS ≥ 2 (42 days: IQR 32–53) (P = <.0001). Moreover, the median survival time was longer in patients with chemotherapy post double stenting (175 days: IQR 149–337) than that in patients without chemotherapy (46 days: IQR 34–60) (P = <.0001) (Fig. 3). Considering the patients with PS <  2, the median survival time after double stenting was significantly longer in patients with chemotherapy post double stenting (175 days: IQR 149–337) than that in patients without chemotherapy (77 days: IQR 53–87) (P = 0.0029) (Fig. 4).

Median survival time after double stenting for patients who had any chemotherapy (n = 35) and never had any chemotherapy (n = 15) was 101 days (IQR 51–224) and 43 days (IQR 34–87), respectively (P = 0.0029). Median survival time after double stenting for patients who had upfront double stenting, and then received frontline chemotherapy (n = 9) and double stenting during chemotherapy, and then continued second or third line chemotherapy (n = 12) was 325 days (IQR 149–375) and 158 days (IQR 89–337), respectively (P = 0.63) (Fig. 5).

Median overall survival time for patients who had metastasis (n = 33) and without metastasis (n = 22) at the initial diagnosis was 264 days (IQR 197–489) and 410 days (IQR 223–791), respectively (P = 0.40). Median survival time after double stenting for patients who had metastasis and without metastasis at the initial diagnosis was 53 days (IQR 44–149) and 119 days (IQR 53–175), respectively (P = 0.46). Median overall survival time for patients who had simultaneous and metachronous double stenting was 264 days (IQR 123–410) and 367 days (IQR 223–647), respectively (P = 0.13). Median survival time after double stenting for patients who had simultaneous and metachronous was 149 days (IQR 49–255) and 53 days (IQR 40–101), respectively (P = 0.14).