This study protocol received ethical approval from the Regional Ethics Committee of Guangzhou Fuda Cancer Hospital. Written informed consent was obtained from each participant in accordance with the Declaration of Helsinki.

This was a retrospective study of patients treated for metastatic pancreatic tumors in our tumor hospital from November 2008 through to August 2010. Before hospitalization, 145 patients received a comprehensive evaluation, with their tumors being considered unresectable. These evaluations were multidisciplinary decisions incorporating a radiologist, a gastrointestinal pancreatic surgeon, and an oncologist in our hospital. Their diagnoses were principally based on computed tomography (CT) imaging and CT-guided needle biopsy to obtain a definitive diagnosis as pancreatic adenocarcinoma histologically. An 18-gauge Tru-Cut (Baxter, Deerfield, IL, United States) biopsy needle was used percutaneously to obtain one to two cores of tissue from the solid pancreatic tumor. Patients received their final treatments in our hospital and almost 2.5 years of follow-ups were reviewed.

A patient was deemed unsuitable for surgery and chemotherapy due to any of the following reasons: (1) multifocal disease; (2) an unresectable primary tumor; (3) the patient refusing surgery and chemotherapy, or seeking further treatment after previous chemotherapy; (4) severe complications (e.g., hypertension, hydrothorax, and ascites); or (5) advanced age. Written informed consent was obtained from all patients. The inclusion criteria were as follows: (1) the Karnofsky performance status (KPS) score was ≥ 70; (2) the platelet count was ≥ 80 × 10⁹/L, white blood cell count was ≥ 3 × 10⁹/L, neutrophil granulocyte count was ≥ 2 × 10⁹/L, or hemoglobin ≥ 90 g/L; (3) the prothrombin time international normalized ratio was ≥ 1.5; (4) the largest primary or metastatic tumor diameter was < 6 cm, as measured by preoperative CT; (5) the pancreatic tumor did not obviously invade the main pancreatic duct, postcava, duodenum or colon; (6) absence of level 3 hypertension, severe coronary disease, myelosuppression, respiratory system disease, acute and chronic infection; (7) the basic normal liver function and puncture release ascites < 1L; and (8) the patient was deemed incapable of cooperating during the procedure.

When the patient had primary or metastatic tumors with a diameter ≥ 6 cm, or obviously invading the main pancreatic duct, postcava, duodenum or colon, we preferentially performed the treatment in other ways[11,12], and so these patients were not enrolled in this study.

According to the patients’ own selection, 51 patients were under radical treatment and 31 patients received palliative treatment. Their treatments had been reported by us[9,10]. In the radical treatment group, the CandS treatment of the pancreatic tumor was performed under double row helical CT (Somatom Emotion Duo; Siemens, Germany) and color ultrasound device (ALOKA SSD-5500SA; Aloka, Japan) guidance. Before the cryosurgery, the patients were administrated general or local anesthesia, and positioned for an upper abdominal incision. All cryosurgery were performed by Dr. Niu LZ and assistants (Zhou L and Zhang C). Based on the location of the pancreatic tumor, the cryoprobes were inserted percutaneously via the retroperitoneal, transhepatic, or transgastric approach. For tumors greater than 3 cm in length, more than two 1.7 mm cryoprobes (CRYO-42; Endocare, Irvine, CA, United State) were used, in an attempt to avoid puncturing the main pancreatic duct and duodenum. A 1-3 cycle freeze/thaw procedure was used with an argon gas-based cryosurgical unit (Endocare, United State)[9,10]. The iodine-125 seed (Syncor Pharmaceuticals, Shanghai, China) implantation was performed by PTC needle, either at the time of cryosurgery or after cryosurgery through the percutaneous approach under a 3D treatment planning system. The seeds (activity of a single seed 0.7 mCi, half-life period 1-6 mCi) were implanted at the tumor borderline. The number of seeds deployed depended on the tumor size (matching dose around 120Gy, usually ≤ 20 particles), with the seeds implanted at intervals of 0.5 cm. In the palliative treatment group, the iodine-125 seeds were implanted in all parts of the pancreatic tumor under ultrasound or CT guidance, and the planting density and quantity were both more than the radical treatment group. For extrapancreatic metastases in the two groups, the CandS treatment were performed percutaneously via the retroperitoneal or transabdominal approach at the same time, or after the treatment of the intrapancreatic tumor.

Once cryoablation was completed, 1 mL of both fibrinogen and thrombin for each probe were injected into the sheath simultaneously. The patients were then observed in the intensive care unit for at least 6 h, and fasted for at least 24 h. Therapies of anti-infection and inhibition of pancreatic juice secretion were given for a few days. Patients under radical treatment received some special treatment: for patients under the transgastric approach of cryosurgery, antacid and stomach mucosa-protecting drugs were delivered for a few days; for patients under the transhepatic approach of cryosurgery, oppressing hemostasis, bellyband, and liver-protecting drugs were all administered for a few days.

Complications were recorded and classified in accordance with the Common Terminology Criteria of Adverse Events v4.0. Local tumor control and IS-IV were also evaluated. Radiographic local tumor control was assessed by image-guided tumor ablation criteria[13]. A post-operative plain abdominal CT was performed immediately after the removal of the cryoprobes for verification as to whether any major complications, such as pancreatic fistula, bile leakage, or intestinal fistula, had occurred. Abdominal ultrasound was performed at both 1 d and 1 wk after the cryoablation procedure. Follow-up dynamic CT abdominal scans of patients were carried out at 1 mo, and then at 3 to 4 mo intervals. The revised RECIST criteria (version 1.1) were used to assess the basic response of the intrapancreatic and extrapancreatic cancer[14]. Three diagnostic radiologists (Piao XH, Zhou Q, and Tang J) with 17, 20, and 13 years of clinical experience, respectively, determined whether progression or recurrence had occurred; reviewing CT scans in every case. Diagnoses were made independently, and the radiologists discussed with each other when they the results were different. The IS-IV was calculated from the date when a patient was first diagnosed as suffering stage IV pancreatic cancer, and compared using the Kaplan-Meier test with long-rank analysis. A significant difference was indicated by P < 0.05. All analyses were performed using GraphPad Software (San Diego, CA, United States).

Percutaneous cryoablation was performed on 81 patients (42-84 years of age, median age: 65 years; 43 male patients, 38 female patients). The patients of each treatment half were from both China (38 patients) and abroad (43 patients). Two-thirds of patients (54 patients) were treated in our hospital when diagnosed as having stage IV pancreatic cancer; one-third of patients were diagnosed and treated with chemotherapy (27 patients, 95 sessions) and/or radiation (9 patients, 24 sessions) in other hospitals first, and came to our hospital 2-14 mo later for further treatment. Liver metastases (75 lesions) were found in 47 patients, peritoneum and liver metastases (76 lesions) were found in 27 patients, and all other metastases (26 lesions) were found in seven patients. Diabetes (16 patients), hydrothorax/ascites (15 patients), and hypertension/coronary disease (8 patients) were common complications in these patients. Radical treatment was performed on 50 patients (79 sessions), and 31 patients (48 sessions) were put under palliative treatment. The therapeutic option of cryoablation was chosen according to the previously-stated criteria.

The mean intrapancreatic tumor diameter in the radical treatment group was 4.3 ± 0.9 cm (range: 3.2-5.7 cm). For the first treatment of the 50 lesions, 13 were treated by using 2 cryoprobes; 16 by 3 cryoprobes; and 21 by four cryoprobes. The mean intrapancreatic tumor diameter in the palliative treatment group was 4.5 ± 0.8 cm (range: 3.5-5.8 cm). For the first treatment of the 31 lesions, 6 were implanted by using about 20 iodine-125 seeds; 9 by about 30 seeds; and 16 by about 40 seeds.

The mean extrapancreatic tumor diameter in the radical treatment group was 3.5 ± 0.8 cm (range: 2.1-4.5 cm). For the first treatment of the 110 lesions, 46 were treated by using 2 cryoprobes; 43 by 3 cryoprobes; and 21 by four cryoprobes. The mean extrapancreatic tumor diameter in the palliative treatment group was 3.7 ± 0.9 cm (range, 2.3-5.0 cm). For the first treatment of the 67 lesions, 31 were treated by using 2 cryoprobes; 27 by 3 cryoprobes; and 9 by 4 cryoprobes.

All percutaneous cryoablations of primary and metastatic pancreatic tumors by ultrasound and CT monitoring were performed successfully. No severe complications, such as pancreatic fistula, bile leakage, and intestinal fistula were discovered post-cryoablation. In the radical treatment group, some common adverse effects were found and are shown below. Serum amylase, an important index of acute pancreatitis, increased in 25 sessions (31%) for 16 patients (32%) on the first day after the procedure, unaccompanied by ascites or leukocytosis, but returned to normal levels in the following 5 d after symptomatic treatment. Seven patients (14%) with diabetes after 12 sessions (15%) experienced a rise in fasting blood glucose levels up to 20-25 mmol/L on the first day post-cryoablation, which were well-controlled with insulin injections. A mild decrease in the platelet count occurred after 14 sessions (18%) for 10 patients (20%), which returned to normal within 8-13 d without any treatment. Abdominal distension and nausea occurred after 23 sessions (29%) for 14 patients (28%) on the first day post-cryoablation, and disappeared automatically the following day. Eight patients (16%) in 11 sessions (14%) complained of poor appetite after the cryosurgery and were found to have ascites by ultrasound, which improved the following 3-5 d without any treatment. Eleven patients (22%) in 17 sessions (22%) were found to have abdominal bleeding, 8 patients (16%) of whom developed fever and a mild increase in white blood cells, and neutrophil granulocytes; they improved in the following 7 d after symptomatic treatment. Similar adverse effects were also found in the palliative treatment group, and disappeared in 2 wk after symptomatic treatment. There were no treatment-related deaths or any conversions to chemotherapy.

Within 1 wk after the first CandS treatment, 64 patients (79%) experienced a ≥ 50% reduction in pain score, 57 patients (70%) experienced a 50% decrease in analgesic consumption, and 43 patients (53%) experienced a ≥ 20 increase in KPS score.

The median follow-up period was 8 mo. Up to the date of the last follow-up of each patient, the median IS-IV of all patients was 7.12 mo (95%CI: 3.79-13.37). The median IS-IV of patients with radical treatment was 8 mo, and those with palliative treatment was 4 mo. The IS-IV of the two groups differed significantly by long-rank test (P < 0.001, Figure 1).

Open in New Tab   Full Size Figure   Download Figure

Figure 1 IS-IV of patients under radical or palliative treatment. All 81 patients had suffered from stage IV pancreatic cancer and died before October of 2011. There were 51 patients in the radical treatment group, and 30 patients in the palliative treatment group. The median follow-up period was 8 mo. The IS-IV of patients was accumulated from the early diagnosis of stage IV pancreatic cancer in our own or other hospitals.

Two-third of patients received CandS treatment 2 mo after the diagnosis of stage IV pancreatic cancer, and one-third of patients received CandS treatment 2-14 mo after diagnosis. The influences of treatment timing on IS-IV of patients were detected: for the radical treatment group, the median IS-IV of patients under timely treatment was 6 mo, and 13 mo for those under delayed treatment (P = 0.0034, Figure 2A); for the palliative treatment group, the median IS-IV of patients under timely treatment was 3 mo, and 5 mo for those under delayed treatment (P = 0.0415, Figure 2B). In the two groups, the patients who received CandS treatment ≥ 2 mo after diagnosis were associated with a longer IS-IV, which was more obvious in the radical treatment group.

Open in New Tab   Full Size Figure   Download Figure

Figure 2 IS-IV of patients under timely or delayed treatment. A: Comparison between patients in the radical treatment group: 32 patients under timely treatment and 18 patients under delayed treatment; B: Comparison between patients in the palliative treatment group: 22 patients under timely treatment and 9 patients under delayed treatment.

According to disease progression, tumor recurrence and individual intents of patients, 31 patients (39%) received repeated CandS treatment when re-examined. For the radical treatment group, the median IS-IV of patients under single treatment was 7 mo, and 11 mo for those under multiple treatments (P = 0.0389, Figure 3A); for the palliative treatment group, the median IS-IV of patients under single treatment was 3 mo, and 4 mo for those under multiple treatments (P = 0.99, Figure 3B). Only in the radical treatment group did multiple treatments show a significant superiority in prolonging the IS-IV of patients.

Open in New Tab   Full Size Figure   Download Figure

Figure 3 IS-IV of patients under different treatment frequency. A: Comparison among patients in radical treatment group, 17 patients under multiple radical treatment (2 times, by 10 patients; 3 times, by 4 patients; 4, 5 and 9 times, by 1 patient each) and 33 patients were under single radical treatment; B: Comparison among patients in palliative treatment group, 14 patients under multiple palliative treatment (2 times, by 11 patients; 3 times, by 3 patients) and 17 patients under single palliative treatment.

Pancreatic cancer is the fourth leading cause of cancer-related death clinically. Less than 20% of pancreatic cancers are curable on diagnosis. Without effective treatment, the overall survival for patients with advanced pancreatic cancer is usually less than 1 year, and for those with metastatic stage IV pancreatic cancer, the median survival is only 3-4 mo. Treatment effects of radiation and chemotherapy are poor, and radiofrequency ablation can not be applied for patients percutaneously.

This research concerns the field of ablation for pancreatic cancer. Currently, cryosurgery is the only useful method for in situ ablation of pancreatic cancer.

Three methods were used to treat different parts of the pancreatic cancer by percutaneous ablation, which increased the safety and efficiency of cryosurgery for patients. The relationship between treatment timing, treatment frequency, and survival time of patients with metastatic pancreatic cancer were analyzed.

Depending on the retrospective analysis of radical and palliative treatment on patients with stage IV pancreatic cancer, clinical evidence was provided concerning survival time, which will benefit the future application of this technology.

Percutaneous cryosurgery: Under the guidance of B ultrasound or computed tomography (CT), cryoprobes are directly stuck into the tumor through skin, and cryoablation is performed in situ by the ultra-low temperature of the probe tip; Seed implantation: under the guidance of B ultrasound or CT, radioactive seeds are implanted into the tumor to act as brachytherapy; IS-IV: since the patients enrolled in this study were all suffered from metastatic tumors (stage IV pancreatic cancer), and many treatments had been carried out in other countries and hospitals before, the ideal evaluating method for survival time for our treatment is interval survival for stage IV pancreatic cancer.

This study is the first retrospective analysis for the radical and palliative treatment of patients with stage IV pancreatic cancer, which clarified the benefits of percutaneous cryosurgery for this disease, along with the relationship between treatment timing, treatment frequency, and survival time of patients. These findings can provide important evidence for the clinical treatment of stage IV pancreatic cancer.