Background

Pancreatic ductal adenocarcinoma (PDAC) is a lethal disease with poor prognosis, even in patients who have undergone resection with curative intent [Statistical analysis

Data are expressed as median values and ranges. Clinical background and clinical course were compared according to metastatic site in Table 1. Continuous or categorical variables were compared by Mann–Whitney U test, chi-square test, or Fisher’s exact test as appropriate. Overall survival curve, defined as the time from initial treatments to death or the last follow-up date, were compared using the log-rank test according to metastatic site. In addition, profound factors identified by the univariate analysis were further examined by multivariate Cox proportional-hazard models to determine independent significant factors for survival. The hazard ratio and 95 % confidence intervals were calculated for all estimates. A two-tailed p value less than 0.05 was considered to be statistically significant. Statistical analyses were performed using SPSS version 18.0 for Windows (Chicago, IL, USA).

Table 1 Patient characteristics and clinical course in each group
Full size table

Results

Outcome measures of staging laparoscopy

Median operative time was 67 min (34–168) and median extent of blood loss was 0 ml (0–168). Past history of previous surgery was found in 12 out of 67 patients (18 %). Three patients needed open mini-laparotomy converted from staging laparoscopy because of severe adhesion due to previous surgeries. Biopsy specimens were obtained from 43 patients (64 %) for evaluating if occult distant organ metastasis was present. One patient incidentally had intestinal perforation during staging laparoscopy, and primary closure was immediately done. Her postoperative course was uneventful and underwent chemotherapy without delay.

Exploration using staging laparoscopy clearly demonstrated the presence of occult distant organ metastasis in 39 out of 67 patients (58 %). There was positive cytology in 16 patients (CY, 24 %), peritoneal dissemination in 13 patients (P, 19 %), liver metastasis only or concomitant with peritoneal metastasis in 10 patients (L including 3 patients with concomitant peritoneal metastasis, 15 %), and locally advanced disease in 28 patients (LA, 42 %), respectively.

Comparison of clinical backgrounds and clinical course according to metastatic site

With regard to the clinical backgrounds and the clinical course, as shown in Table 1, the L or P groups had a higher frequency of pancreatic body and tail cancer, compared with the LA or CY groups. Addition of radiation therapy during chemotherapy was frequently performed in the LA and CY groups according to the departmental policy. The clinical response rate, as based on the RECIST [7], was higher in the LA or CY groups than in either the P or L groups. When the rate of emergence of ascites according to the site of occult distant organ metastasis was compared, it was significantly lower in the LA group than in CY or P group (Fig. 1, p < 0.05). Development of ascites within 1 year after initial treatment was found most frequently in patients in the P group (11 out of 13 patients, 85 %), which was higher than in the other 3 groups. Second-line chemotherapy was administered less frequently in P group (1 out of 13 patients, 8 %), relative to the other three groups (Table 1). Moreover, duration of chemotherapy in P group was shorter than in the other groups.

Fig. 1
figure 1

The rate of emergence of ascites according to metastatic site diagnosed by staging laparoscopy. Figure 1 shows the curve of emergence of ascites in patients with locally advanced PDAC (LA, solid black line), with positive cytology (CY, solid grey line), with peritoneal metastasis (P, dotted grey line), and with liver metastasis (L, dotted black line). There were significant differences in the rate of emergence of ascites between LA and P (p < 0.001) and LA and CY (p = 0.033) groups

Full size image

Median survival time was 13 months in the CY group (including 2 patients who underwent surgical resection), 11 months in the LA group (including 1 patient who underwent surgical resection), and 7 months in the P and L groups, respectively. Actual 1-year survival rates in the 4 groups were as follows: 46 % (LA), 63 % (CY), 23 % (P), and 20 % (L). No patient in the P and L groups was alive beyond the 2-year follow-up. Thus, as shown in Fig. 2a, the overall survival curve in P group was significantly worse than in either the LA or CY groups (p = 0.04 and p = 0.007, respectively). No significant differences in the survival curves were seen between CY and LA, P and L, or LA and L groups. As shown in Fig. 2b, the survival curve in the LA + CY groups was significantly better than in the P + L groups (p = 0.011). In the LA or CY groups, additional radiation therapy was administered to approximately 30–50 % of patients, PR according to RECIST [7] was observed in approximately 35 % of patients, and second-line chemotherapy was administered to around 60 % of patients. In this study, out of 44 patients in LA and CY groups, conversion surgery was performed in two patients in the CY group (one alive after 60 months; the other died at 18 months, respectively) and one patient in the LA group (alive after 17 months). The multivariate analysis showed that the presence of partial response and administration of second-line chemotherapy were significantly independent favorable factors for prognosis (Table 2).

Fig. 2
figure 2

Survival curves according to metastatic site diagnosed by staging laparoscopy. a shows survival curve of patients with locally advanced PDAC (LA, solid black line), with positive cytology (CY, solid grey line), with peritoneal metastasis (P, dotted grey line), and with liver metastasis (L, dotted black line). There were significant differences between the survival curves of P and CY (p = 0.007) and P and LA (p = 0.04) groups. b shows survival curves of patients with locally advanced PDAC and positive cytology (LA + CY, solid black line) and with peritoneal metastasis and liver metastasis (P + L, dotted black line). The survival curve in LA + CY groups was significantly better than in P + L groups (p = 0.011)

Full size image
Table 2 Uni- and multivariate analyses for prognosis
Full size table

Discussion

The majority of patients with PDAC have unresectable disease at initial presentation, and their prognosis is extremely poor. The median survival time of these patients ranges from 6 to 12 months in spite of developed, advanced, and newly provided regimens of chemotherapy [8–11]. Accurate staging is necessary for selecting patients who should be curatively resected, for introducing new regimens of chemotherapy or chemo(radio)therapy, or for conducting a clinical trial. Several articles have reported that the use of staging laparoscopy to further explore locally advanced PDAC, initially diagnosed by CT scan, has identified the presence of occult distant organ metastasis [4, 5, 12–17]. However, less information is available on clinical background and clinical course according to occult distant metastatic site diagnosed by staging laparoscopy in patients with RD-LA PDAC. Understanding the clinical features according to the site of occult distant organ metastasis is important for the development of new strategies and improving prognosis in patients with locally advanced PDAC.

We have focused on RD-LA PDAC defined by the NCCN guidelines [6], excluding resectable, borderline resectable, and metastatic diseases. The results of staging laparoscopy in this study showed occult distant organ metastasis in 58 % of patients including positive cytology in 24 %, peritoneal dissemination in 19 %, and liver metastasis in 15 %, respectively. Finally, pure locally advanced disease without minute distant organ metastasis was found in 42 % of patients only. Clark et al. [12] reported that diagnostic laparoscopy for borderline resectable and unresectable locally advanced diseases showed occult distant organ metastasis in 58 of 202 patients (29 %) including positive peritoneal lavage cytology in 20 % (n = 41), liver metastasis in 13 % (n = 26), and peritoneal metastases in 3 % (n = 5) (Table 3). Relative to Clark et al. [12], our study consisted of a more limited population of unresectable, locally advanced PDAC. As shown in Table 3, compared with other studies, our study consisted of high frequency of pancreatic body and tail cancer, which was associated with unresectability and the presence of occult distant organ metastasis [4, 5]. In our experience, we had an aggressive attitude towards surgical resection, even in patients with portal or superior mesenteric vein invasion, common hepatic artery invasion, and celiac axis invasion for pancreatic body cancer. The definition of local unresectability for our patients was strictly limited. Moreover, we also recruited patients with suspicion of peritoneal metastasis less than 10-mm diameter on MDCT or ascites limited to the Pouch of Douglas. Therefore, this could explain why a high incidence of occult distant organ metastasis (58 %) and peritoneal metastasis (19 %) was found in our study.

Table 3 Review of the recently published articles on staging laparoscopy in patients with locally advanced pancreatic cancer
Full size table

We have introduced palliative chemotherapy in patients with distant organ metastasis including peritoneal metastasis, and chemo(radio)therapy for tumor remission in patients with LA or CY. The clinical findings by staging laparoscopy can provide to change the paradigm of the treatment in patients with RD-LA PDAC. Actually, peritoneal and/or liver metastasis was observed in 34 % of patients, who were treated with palliative chemotherapy. Thus, the results of staging laparoscopy changed the treatment according to distant organ metastatic sites. Considering the high frequency of patients with occult distant organ metastasis, staging laparoscopy should be routinely performed in patients with RD-LA PDAC.

With regard to the comparison of survival curves according to the site of occult distant organ metastasis, P or L groups had a worse prognosis than either the CY or LA groups in our study. Actual 1-year survival rate was 46 % in LA, 63 % in CY, 23 % in P, and 20 % in L groups. In the LA or CY groups, high frequency of additional radiation, PR according to RECIST, second-line chemotherapy and conversion surgery might result in better prognosis. Appearance of partial or complete response after chemo(radio)therapy can sometimes lead to subsequent surgical resection. It has been reported that the overall survival rate in patients with initially unresectable PDAC who underwent surgical resection after favorable response to anti-cancer treatments over a period of time was similar to that in patients with resectable PDAC [18, 19]. Even in patients with initially unresectable disease, long-term survival can be expected if an unresectable tumor was surgically resected after tumor shrinkage. Advanced chemotherapy or chemoradiation therapy may have a role in downsizing an unresectable tumor sufficiently to render it resectable. Thus, a new strategy to shrink the tumor in LA or CY group should be introduced where possible to allow for subsequent surgical resection, which is associated with long-term survival.

The multivariate analysis showed that the presence of partial response and administration of second-line chemotherapy were significantly independent favorable factors for prognosis (Table 2). The curve of emergence of ascites in CY or P group was significantly worse than in LA group (Fig. 1, p < 0.05). The P group in particular had ascites in 85 % of patients within 1 year after initial treatment. Moreover, no patient survived longer than 2 years after initial treatment in P group. A lower frequency of second-line chemotherapy and clinical response and shorter duration of chemotherapy in P group were found in this study, probably due to emergence of ascites and poor performance status. Thus, conventional chemotherapy may not be effective for patients with peritoneal metastasis, and therefore, a new regimen to control or prevent ascites will be required for achieving better prognosis. In this regard, intraperitoneal chemotherapy may be one option for patients with peritoneal metastasis that should be investigated. A Japanese case report described the success of intraperitoneal paclitaxel combined with systemic chemotherapy of S-l and gemcitabine for locally advanced PDAC with peritoneal carcinomatosis [20]. After 20 months, all peritoneal deposits had disappeared and a radical resection of the tumor was performed. Sugarbaker et al. [21] have conducted a phase II study of hyperthermic intraoperative gemcitabine and long-term intraperitoneal gemcitabine in the adjuvant setting for resectable PDAC patients. Further protocol-based studies should be performed to investigate whether prognosis in patients with occult peritoneal metastasis can be improved or not.

The limitation of the study was the small number of patients recruited retrospectively and the presence of undetectable liver metastasis deep within the liver and/or other organ metastasis.

Conclusions

In conclusion, staging laparoscopy is necessary for selecting patients with occult distant organ metastasis which is the majority population of patients with RD-LA PDAC. Clinical features and survival curves were different depending on the site of occult distant organ metastasis. Administration of second-line chemotherapy and responsiveness to chemotherapy were associated with favorable prognosis. Staging laparoscopy should be routinely performed in PDAC patients with RD-LA.