- Email: [email protected]
Follow-up after curative surgery for pancreatic ductal adenocarcinoma: Asymptomatic recurrence is associated with improved survival
Available online at www.sciencedirect.com
EJSO 39 (2013) 559e566
www.ejso.com
Follow-up after curative surgery for pancreatic ductal adenocarcinoma: Asymptomatic recurrence is associated with improved survival T. Nordby a,d, H. Hugenschmidt a, M.W. Fagerland b, T. Ikdahl c, T. Buanes a,d, K.J. Labori a,* a
Department for Hepato-Pancreato-Biliary Surgery, Division of Cancer Medicine, Surgery and Transplantation, Oslo University Hospital, Sognsvannsveien 20, N-0027 Oslo, Norway b Unit of Biostatistics and Epidemiology, Oslo University Hospital, Norway c Department of Oncology, Oslo University Hospital, Norway d Institute of Clinical Medicine, Faculty of Medicine, University of Oslo, Norway Accepted 20 February 2013 Available online 14 March 2013
Abstract Aim: There is no consensus on the optimal follow-up schedule of patients after surgery for pancreatic cancer. In this retrospective study, recurrence and survival were investigated for patients presenting with either symptomatic or asymptomatic recurrence. Patient, tumor and treatment characteristics that predicted the length of postrecurrence survival were identified. Methods: Clinical records of 164 patients who underwent a pancreatic resection (R0/R1) for pancreatic ductal adenocarcinoma from January 2000 to December 2010 were retrieved. Patients underwent a systematic follow-up program. Patient, tumor and treatment characteristics were compared between patients with asymptomatic and symptomatic recurrence. Results: Of 164 consecutive patients, 144 patients (88%) had recurrence (29 asymptomatic, 115 symptomatic). The most frequent reported symptoms were abdominal pain, fatigue/weakness, back pain, weight loss, nausea/loss of appetite and jaundice. Median time to recurrence was 12.0 months for asymptomatic and 7.0 months for symptomatic patients (P ¼ 0.036). Median postrecurrence survival was 10.0 months for asymptomatic and 4.0 months for symptomatic patients (P < 0.0001). Median overall survival was 24.5 months for asymptomatic and 11.0 months for symptomatic patients (P < 0.0001). Symptomatic recurrence, disease free survival <12 months, and no adjuvant chemotherapy were the only independent predictors of poor postrecurrence survival. 72% of asymptomatic and 37% of symptomatic patients received oncological treatment. Conclusions: Patients with asymptomatic pancreatic cancer recurrence have improved recurrence-free, postrecurrence and overall survival. Symptoms when recurrence is diagnosed are a good surrogate marker of biological aggressiveness. Detection of asymptomatic recurrence may facilitate patient eligibility for investigational studies or other forms of treatment. Ó 2013 Elsevier Ltd. All rights reserved. Keywords: Pancreatic neoplasms; Pancreaticoduodenectomy; Recurrence; Follow-up studies; Survival
Background Pancreatic cancer is the fourth leading cause of cancerrelated deaths in Europe and the United States.1,2 Surgical resection provides the only chance of cure. The median survival of resected patients is 11e20 months and the 5-year survival rate 10e20% in recent reports.3,4 There is no consensus on the optimal follow-up schedule of patients after surgery for pancreatic cancer. All of the recommendations * Corresponding author. Tel.: þ47 23070000; fax: þ47 22117470. E-mail address: [email protected] (K.J. Labori). 0748-7983/$ - see front matter Ó 2013 Elsevier Ltd. All rights reserved. http://dx.doi.org/10.1016/j.ejso.2013.02.020
on surveillance are based on low level evidence or no evidence.5 Recently, a lack of established guidelines for surveillance for pancreatic cancer patients has been observed in the United States.6 The inconsistency of follow-ups is revealed by the fact that the leading societies propose different guidelines.7e10 As a consequence, follow-up strategies may differ between hospitals depending on preference of physicians. It is not clear whether survival and quality of life of resected pancreatic cancer patients are improved by follow-up and early detection of recurrence. The prognosis of recurrent pancreatic cancer is dismal with a median
560
T. Nordby et al. / EJSO 39 (2013) 559e566
postrecurrence survival of 3e7 months.11,12 Given the poor survival of patients with recurrent pancreatic cancer, the prognostic effect of early detection seems doubtful. However, with surgery being the only option with a potential to cure, it has been hypothesized that reresection for localized recurrence might open a small window of opportunity for some patients.13 In the context of a systematic follow-up program by medical history, physical examination and computed tomography (CT) every 6 months after a curative pancreatic resection, recurrence and survival were investigated for patients presenting with either symptomatic or asymptomatic recurrence. In addition, we sought to identify patient, tumor and treatment characteristics that predicted the length of postrecurrence survival. Patients and methods Patient selection and data collection From January 2000 to December 2010, 187 patients with pancreatic ductal adenocarcinoma (PDAC) underwent surgical resection with curative intent. Final date of data collection was December 31, 2011. The study was approved by the Data Protection Officer for Research. Information about the patients was recorded retrospectively using hospital records. In one patient the only available information came from the patient’s family physician. PDAC was confirmed histologically in all cases. Patients with other histological variants were excluded. Preoperative workup included chest CT or X-ray and abdominal CT to evaluate primary and metastatic tumor sites. Until 2008 the pathologist reported a margin positive (R1) only if tumor cells were present at the surface (clearance equals 0 mm). In 2008 the definition was changed to a 1 mm clearance. From 2005 adjuvant fluorouracil and leucovorin was administered every second week for six months to patients <75 years old, ECOG 0-1. Follow-up All patients were followed regularly with history and physical examination to pursuit postoperative complications and symptoms. Chest and abdominal CT were performed every 6 months or if the patients had symptoms suspect of a recurrence. Recurrence was defined as radiological evidence of intra-abdominal soft tissue around the surgical site or of distant metastasis. All patients had a CT performed to evaluate site of first recurrence, except two patients who had an abdominal ultrasound performed. If possible a recurrence was confirmed histologically. Symptomatic patients had a symptom or finding that was new and had increased in either severity or frequency since the last follow up. Asymptomatic patients had no new symptoms or findings at a scheduled visit. Patients with recurrence were referred to the Department of Oncology to be
considered for gemcitabine or radiotherapy with or without concurrent capecitabine, and to the Palliative Care Unit for assessment of symptoms and to receive the best palliation of symptoms, as previously described.14 Statistics Patient, tumor and treatment characteristics were compared between patients with asymptomatic and symptomatic recurrence using the WilcoxoneManneWhitney (age, tumor size, and lymph node ratio) tests for continuous variables and the Pearson chi-squared (sex, nodal status, resection margins, site of first recurrence, and adjuvant chemotherapy), Fisher-Freeman-Halton (type of resection), and WilcoxoneManneWhitney (tumor stage, tumor grade, and AJCC stage) tests for categorical variables. Cox regression analyses were used to assess the prognostic capacity of patient and tumor characteristics on time from recurrence to death. First, univariable models were fitted. Second, all variables with P < 0.20 in univariable analysis were included in a multivariable model. Third, the final model was obtained after removing, one at a time, variables with P > 0.05. The proportional hazard assumption was investigated using a test based on Schoenfeld residuals. Estimates of the median overall survival time and disease free survival time from resection, and survival time after detection of recurrence were obtained from the KaplaneMeier estimator. The log-rank test was used to compare survival functions. The statistical analyses were done using Stata 12.0 (StataCorp, College Station, Texas). All P-values are two-tailed. Results Patient characteristics A total of 187 patients underwent surgical resection for PDAC. Patients who died within 3 months after surgery (n ¼ 9) or had non-cancer related death (n ¼ 4) were excluded from the study. Patients with macroscopic residual tumor (n ¼ 7) or distant metastasis at the time of resection (n ¼ 3) were also excluded from further analysis. Accordingly, 164 patients were included in the study. There were 72 women and 92 men. Median age at surgery was 68 years (range 34e84 years). The procedures performed included pylorus-preserving pancreatoduodenectomy (n ¼ 85), pancreatoduodenectomy with antrectomy (n ¼ 63), distal pancreatectomy with splenectomy (n ¼ 10), and total pancreatectomy with splenectomy (n ¼ 6). Recurrence Of the 164 patients, 144 had recurrences, whereas 20 patients were free of recurrence. Patterns of recurrence and patient demographics at the time of initial surgery are listed in Table 1. Twenty-nine patients were asymptomatic when recurrence was detected, whereas 115 patients either
T. Nordby et al. / EJSO 39 (2013) 559e566
561
Table 1 Patient characteristics based on the presence or absence of symptoms at the detection of pancreatic cancer recurrence in 144 patients. Variable Sex, no. Male Female Age, median (range) years Type of resection, no. Whipple Distal resection Total pancreatectomy Tumor stage, no. T1 T2 T3 T4 Tumor grade, no. G1 G2 G3/4 Tumor size, median (range), cm Nodal status, no. N0 N1 Lymph node ratio, median (range) AJCC stage, no. IA IB IIA IIB III Resection margins, no. R0 R1 Site of first recurrence, no. Locoregional Distant Locoregional þ distant Adjuvant chemotherapy, no. Yes No
Asymptomatic recurrence n ¼ 29
Symptomatic recurrence n ¼ 115
P-value
19 10 68 (50e79)
64 51 68 (34e84)
P ¼ 0.34
26 3 0
104 6 5
P ¼ 0.35
0 7 22 0
7 21 85 2
P ¼ 0.97
1 23 5 3.0 (1.8e8.0)
3 86 26 3.0 (1.0e7.0)
P ¼ 0.51
14 15 0.06 (0.0e0.8)
51 64 0.10 (0.0e1.0)
0 4 10 15 0
7 12 31 63 2
P ¼ 0.79
14 15
58 57
P ¼ 0.84
7 18 4
40 40 35
P ¼ 0.024
14 15
38 72
P ¼ 0.13
consulted a physician (n ¼ 77) because of symptoms suggesting recurrence or had such symptoms at the time they came to the hospital for a regular follow-up (n ¼ 38). Except for site of first recurrence, no differences in clinicopathologic variables were evident between patients with asymptomatic or symptomatic recurrence (Table 1). Distant recurrence only was more common in asymptomatic patients (62% vs 35%; P ¼ 0.007). Locoregional recurrence only was similarly frequent in the two groups (24% vs 35%; P ¼ 0.28). Combined locoregional and distant recurrence was slightly more common in symptomatic patients (14% vs 30%; P ¼ 0.072). Distant recurrences were found in liver (n ¼ 72), intraabdominal lymph nodes (n ¼ 37), peritoneum (n ¼ 20), and lung (n ¼ 16). There were no significant differences between the two groups except for peritoneal recurrence that was only observed in symptomatic patients. Eight patients with recurrence (four in each group) were alive at the time of final data collection. The reported
P ¼ 0.96
P ¼ 0.43 P ¼ 0.70 P ¼ 0.87
symptoms were abdominal pain (n ¼ 85), fatigue/weakness (n ¼ 40), back pain (n ¼ 24), weight loss (n ¼ 21), nausea/ loss of appetite (n ¼ 21), jaundice (n ¼ 11), cachexia (n ¼ 5), diarrhea (n ¼ 4), and miscellaneous (n ¼ 12). Patient survival Median disease-specific survival was 24.5 (95%CI: 16.0, 31.0) months for asymptomatic and 11.0 (95%CI: 10.0, 13.5) months for symptomatic patients (P < 0.0001) (Fig. 1a). Median time to recurrence was 12.0 (95%CI: 6.0, 14.0) months for asymptomatic and 7.0 (95%CI: 6.0, 8.0) months for symptomatic patients (P ¼ 0.036) (Fig. 1b). Median postrecurrence survival was 10.0 (95% CI: 7.0, 15.0) months for asymptomatic and 4.0 (95%CI: 3.0, 5.0) months for symptomatic patients (P < 0.0001) (Fig. 1c). Overall, an isolated locoregional recurrence was found in 47 patients (33%), a combined locoregional and
562
T. Nordby et al. / EJSO 39 (2013) 559e566
distant recurrence in 39 patients (27%), and an isolated distant recurrence in 58 patients (40%) (Table 1). The type of cancer recurrence did not significantly influence survival, with a median postrecurrence survival of 5.0 (95%CI: 3.5, 6.0) months in locoregional recurrence, 5.0 (95%CI: 3.5, 6.0) months in combined locoregional/distant recurrence, and 4.0 (95%CI: 3.0, 6.0) months in distant recurrence (P ¼ 0.92), with overall survival of 15.0 (95%CI: 10.0, 18.0), 12.0 (95%CI: 8.0, 15.0) and 13.0 (95%CI: 10.5, 16.5) months (P ¼ 0.83), respectively. Predictors of post-recurrence survival Treatments given after detection of recurrent disease are listed in Table 2. Eight of 29 asymptomatic patients and 69 of 115 symptomatic patients received best supportive care only. Symptomatic recurrence (HR ¼ 2.75 [1.71, 4.42]; P < 0.001), disease free survival less than 12 months (HR ¼ 1.95 [1.33, 2.86]; P ¼ 0.001), and no adjuvant chemotherapy (HR ¼ 1.72 [1.19, 2.49]; P ¼ 0.004) were the only independent predictors of poor postrecurrence survival (Table 3). Discussion To improve the prognosis of resected pancreatic cancer it is necessary to focus on the situation when tumor recurrence takes place. Two recent studies from large population-based data sources have evaluated the posttreatment surveillance patterns in the United States.6,15 Sheffield found a lack of established guidelines for surveillance after curative resection for pancreatic cancer, implying a need for evaluation and standardization of surveillance protocols.6 However, Witkowski found no significant survival benefit among patients who received routinely CT.15 The present study evaluates a surveillance strategy with regular CT every 6 months after a curative pancreatic resection. Our study shows that patients with asymptomatic pancreatic cancer recurrence have improved recurrence-free, postrecurrence and overall survival. In Table 2 Treatment after detection of pancreatic cancer recurrence.
Figure 1. a: Overall survival in 144 patients with pancreatic cancer who had asymptomatic (n ¼ 29) or symptomatic (n ¼ 115) recurrent disease after resection (P < 0.0001). b: Disease free survival in 144 patients with pancreatic cancer who had asymptomatic (n ¼ 29) or symptomatic (n ¼ 115) recurrent disease after resection (P ¼ 0.036). c: Survival after detection of recurrence in 144 patients with pancreatic cancer who had asymptomatic (n ¼ 29) or symptomatic (n ¼ 115) recurrent disease after resection (P < 0.0001).
Best supportive care only Oncological treatment Chemotherapy only Radiotherapy only Chemoradiotherapy Liver resection or RFA Chemotherapy þ Vaccination Radiotherapy þ Vaccination Vaccination only
Asymptomatic recurrence n ¼ 29
Symptomatic recurrence n ¼ 115
8 21 12 1 4 2 2 0 0
72 43 29 4 6 0 1 2 1
RFA ¼ radiofrequency ablation (liver).
T. Nordby et al. / EJSO 39 (2013) 559e566
563
Table 3 Univariable and multivariable analyses of the prognostic variables associated with early death after pancreatic cancer recurrence in 144 patients with complete follow up data. Variable Recurrence Asymptomatic Symptomatic Gender Female Male Age Tumor stage T1 T2 T3 T4 Nodal status N0 N1 Lymph node ratio AJCC stage IA IB IAA IIB III Tumor grade G1 G2 G3/G4 Tumor size Resection margins R0 R1 Site of first recurrence Distant only Locoregional only Locoregional þ distant Disease free interval 12 months <12 months Adjuvant chemotherapy Yes No a
No.
Univariable HRa (95%CI); P-value
Multivariable HRa (95%CI); P-value
29 115
Ref. 2.90 (1.81, 4.64); P < 0.001
Ref. 2.75 (1.71, 4.42); P < 0.001
61 83 144
Ref. 0.79 (0.56, 1.13); P ¼ 0.20 1.01 (0.99, 1.02); P ¼ 0.44
7 28 107 2
Ref. 0.67 (0.29, 1.55); P ¼ 0.35 0.68 (0.32, 1.48); P ¼ 0.34 0.91 (0.19, 4.37); P ¼ 0.90
65 79 144
Ref. 1.07 (0.76, 1.52); P ¼ 0.69 0.88 (0.43, 1.82); P ¼ 0.74
7 16 41 78 2
Ref. 0.76 0.69 0.78 0.99
(0.31, (0.30, (0.36, (0.21,
1.87); 1.55); 1.72); 4.79);
P P P P
¼ ¼ ¼ ¼
0.55 0.37 0.54 0.99
4 109 31 144
Ref. 2.31 (0.71, 7.47); P ¼ 0.16 2.58 (0.76, 8.79); P ¼ 0.13 0.92 (0.80, 1.07); P ¼ 0.29
72 72
Ref. 0.86 (0.61, 1.21); P ¼ 0.39
58 47 39
Ref. 0.94 (0.63, 1.40); P ¼ 0.75 1.02 (0.66, 1.56); P ¼ 0.94
46 96
Ref. 1.87 (1.29, 2.72); P ¼ 0.001
Ref. 1.95 (1.33, 2.86); P ¼ 0.001
52 92
Ref. 1.63 (1.14, 2.35); P ¼ 0.008
Ref. 1.72 (1.19, 2.49); P ¼ 0.004
HR ¼ Hazard ratio obtained from Cox regression.
a similar study from MD Anderson Cancer Center, a surveillance strategy with CT every 3e4 months the first 2 years after PDAC surgery was evaluated.16 In Tzeng’ study, 77.7% of the patients underwent neoadjuvant therapy, and probably excluded from surgery up to 25% of patients initially found clinically resectable.17 In contrast, our centre followed patients every 6 months and no patients had neoadjuvant therapy. Tzeng found no difference in median time to recurrence, whereas survival after detection was 5.1 months in symptomatic patients and 13.0 months in asymptomatic patients (P < 0.001), compared to 4 months and 10 months in our study.16 A more intensive follow up with CT every 3e4 months identified 54.6% of recurrences prior to the onset to symptoms, whereas our surveillance program with CT every 6 months identified 20.1% of patients.
Treatment for recurrence Some authors advocate that it is essential to identify tumor recurrence early in order to offer patients further disease controlling measures or potentially curative options.18 The only measure taken after the detection of recurrence was best supportive care in 63% of symptomatic patients and 28% of asymptomatic patients. Symptomatic recurrence was an independent predictor of poor postrecurrence survival. We suggest that symptoms are a good surrogate marker of biological aggressiveness and more disease burden. Management of symptomatic patients is probably focused upon symptom control rather than initiating more aggressive maneuvers with chemotherapy. Only two patients with asymptomatic liver metastases had a potentially
564
T. Nordby et al. / EJSO 39 (2013) 559e566
curative therapeutic intervention (RFA and liver resection). Both patients had a relatively short posttreatment survival. Three of 47 patients with locoregional recurrence alone and three of 58 patients with liver metastases alone had postrecurrence survival >20 months and >30 months, respectively. Reresection was never discussed in these patients. Only a few studies address the issue of resection of metastatic or recurrent disease in pancreatic cancer.13,19e21 Simultaneous pancreatic and liver resections are not generally recommended.19,20 Adam found that hepatic resection may be warranted in those patients who demonstrate good tumor biology, and reported a 5-year survival of 20% in selected patients with liver metastases after resection for pancreatic adenocarcinoma.21 Kleeff published their results of re-resection of localized recurrence for pancreatic adenocarcinoma in a group of eight patients.13 Compared to a matched control group of patients who did not undergo re-resection, median survival was significantly longer for patients undergoing re-resection (29 vs 14.5 months; P < 0.0001). According to these studies, resection for liver metastases and locoregional recurrence might open a small window of opportunity for a small subgroup of patients. Location of recurrence Patients with pancreatic cancer recurrence are managed similarly to nonresectable patients.22 Chemotherapy and/or chemoradiation intend to control tumor progression, improve survival and quality of life, but the survival benefit is limited.8 The prognosis of recurrent pancreatic cancer is dismal. Sperti found a median survival of 7 months in local recurrence and 3 months in hepatic recurrence.12 As shown in other studies, the type of cancer recurrence did not significantly influence the survival in our study.11 Van den Broeck reported the first location of recurrence to be locoregional in 17%, combined locoregional and distant in 23% and distant in 60%, compared with rates of 33%, 27%, and 40%, respectively, in our study.11 Overall, distant metastasis was found in 83% and 67% of patients in these two studies. Even though local recurrence occurs frequently, it is rarely a direct cause of death since most patients die from metastases.23 Accordingly, it is likely that treatment focused on local control cannot improve the survival of patients with recurrent pancreatic cancer. Thus, treatment regimens that are effective against systemic metastasis are needed. Recommendations for surveillance There are significant differences in the recommendations for surveillance of pancreatic cancer patients following resection.7e9 National Comprehensive Cancer Network (NCCN) recommend a close surveillance with history, physical examination, CA 19-9 and CT every 3e6 months the first two years and then annually.8 European Society of Medical Oncology find no advantage in an earlier detection
of recurrences, but recommend that in the case of elevated preoperative CA19-9 levels, CA 19-9 assessment can be performed every 3 months for 2 years and abdominal CT every 6 months.7 However, all of the recommendations on surveillance are based on low level evidence or no evidence.5,8 Ultimately, the decision regarding the intensiveness of follow-up is left to the preferences of the treating physician after discussion with the patient. The surgical philosophy may vary widely from total nihilism to optimism. A striking underuse of pancreatectomy for localized pancreatic cancer in the USA has been identified.24 Furthermore, CT testing for patients undergoing resection for pancreatic cancer seems to be obtained to evaluate symptoms rather than for routine surveillance.6 These facts may reflect a nihilistic attitude toward pancreatic cancer. Physicians may need to approach surveillance on a case-by-case basis according to estimated recurrence risk and likelihood for therapy benefit should recurrent disease be identified.6 In a recent study, patients were stratified into 4-risk-groups depending on patient, tumor and resection characteristics at the time of surgery, with 5-year survival varying between 0% and 54.5%.25 However, Katz has recently shown that the survival expectancy of all patients with PDAC improves over time regardless of patient characteristics or the pathologic variables associated with the primary cancer.26 Some may argue that follow-up should be able to detect at least some recurrent disease amenable to curative treatment, and not only bring forward the time of diagnosis of incurable disease.27 However, there are several reasons for follow-up.27 Follow-up may be needed to manage post-surgical complications, address feeding and nutritional issues associated with a pancreatic resection per se and to perform audit and quality control of surgical outcomes. In addition follow-up may reassure and provide psychological support to the patients. D’Angelica found that surgeons need to develop effective long-term support systems for their patients after surgery.28 Fear of cancer recurrence is pervasive following resection of pancreatic neoplasms.29 Regular follow-up and appreciation of improvements in conditional survival over time may gradually reduce the anxiety and fear of disease progression that exert a negative impact on the quality of life of survivors of cancer and their caregivers.26,29 Limitations The findings in our study must be interpreted with some caution. The key limitation of our study is the retrospective nature of our analysis. The methodology rely on a retrospective audit of documentation of patient symptoms. However, our institution has had a focus on symptom profiles in pancreatic cancer patients during the study period, and the patients were assessed for gastrointestinal or cancer related symptoms at all consultations.14 Furthermore, an important end point of cancer treatment, apart from cure rate and survival, concerns quality of life. This issue was not evaluated
T. Nordby et al. / EJSO 39 (2013) 559e566
with established instruments for assessment. How detection of recurrent disease affect quality of life of asymptomatic pancreatic cancer patients by confronting them with their dismal prognosis, remains to be established. Lastly, the number of patients with asymptomatic recurrence is small. The power to detect small and moderate differences between asymptomatic and symptomatic patients thereby is low. Non-significant between-group differences should be interpreted with caution, and definite conclusions with respect to these differences are not warranted. The large differences in overall and postrecurrence survival, on the other hand, are highly significant (P < 0.0001), and we are confident that these results represent true differences. Surveillance strategy The important question is whether CT should be performed only in case of signs and symptoms of recurrent cancer or at regular intervals. Symptoms suggesting recurrence in our study were abdominal pain, fatigue/weakness, back pain, weight loss and nausea/loss of appetite. It may be difficult to avoid performing CT scans in patients after surgery for pancreatic cancer because they may frequently experience symptoms that would prompt diagnostic imaging.6 High volume centres both in Europe and the United States recommend regular follow up with CT after surgery for PDAC.6,13,16,18 Heidelberg University Hospital point out that it is essential to identify tumor recurrence early in order to offer patients further disease controlling measures or potentially curative options.13,18 Approximately 3/4 of asymptomatic patients in our study underwent subsequent oncological treatment compared to only 1/3 of symptomatic patients. One might argue that this alone could explain the better post-recurrence survival in the asymptomatic group, and this should be kept in mind when interpreting the results. However, both disease-free and postrecurrence survival were significantly better in the asymptomatic group. We find that our data support the results from MD Anderson Cancer Center that CT surveillance might identify patients with good performance status and tumor biology that are most likely to benefit from subsequent therapy.16 Recently, a chemotherapy regimen consisting of fluorouracil, leucovirin, irinotecan and oxaliplatin has shown to be an effective treatment for patients with metastatic PDAC, and may represent a treatment opportunity for patients with asymptomatic recurrence with general good health.16,30 In the absence of evidence-based guidelines, we recommend surveillance after potentially curative surgery for PDAC with regular CT. Most patients develop recurrence within 2 years after surgery, suggesting a more intensive follow-up the first 2 years after surgery.6,8 In asymptomatic patients, detection of disease through CT may facilitate patient eligibility for investigational studies or other forms of treatment. However, the patients have to be aware of the limitations of detection of an asymptomatic recurrence and that no curative treatment is currently
565
available. Larger prospective trials are needed to determine whether early detection of asymptomatic recurrence can lead to new therapy options and, ultimately, improved survival rates and quality of life in patients with resected PDAC. Conflict of interest The authors have no conflict of interest. References 1. Malvezzi M, Arfe A, Bertuccio P, Levi F, La VC, Negri E. European cancer mortality predictions for the year 2011. Ann Oncol 2011;22(4): 947–56. 2. Siegel R, Ward E, Brawley O, Jemal A. Cancer statistics, 2011: the impact of eliminating socioeconomic and racial disparities on premature cancer deaths. CA Cancer J Clin 2011;61(4):212–36. 3. Hoem D, Viste A. Improving survival following surgery for pancreatic ductal adenocarcinomaea ten-year experience. Eur J Surg Oncol 2012;38(3):245–51. 4. Winter JM, Brennan MF, Tang LH, et al. Survival after resection of pancreatic adenocarcinoma: results from a single institution over three decades. Ann Surg Oncol 2012;19(1):169–75. 5. Poonacha TK, Go RS. Level of scientific evidence underlying recommendations arising from the national comprehensive cancer network clinical practice guidelines. J Clin Oncol 2011;29(2):186–91. 6. Sheffield KM, Crowell KT, Lin YL, Djukom C, Goodwin JS, Riall TS. Surveillance of pancreatic cancer patients after surgical resection. Ann Surg Oncol 2012;19(5):1670–7. 7. Cascinu S, Falconi M, Valentini V, Jelic S. Pancreatic cancer: ESMO clinical practice guidelines for diagnosis, treatment and follow-up. Ann Oncol 2010;21(Suppl. 5):v55–8. 8. Pancreatic adenocarcinoma. National comprehensive cancer network clinical practice guidelines in oncology. http://www.nccn.org/professionals/ physician_gls/f_guidelines.asp#pancreatic [accessed 01.10.12]. 9. Verslype C, Van CE, Dicato M, et al. The management of pancreatic cancer. Current expert opinion and recommendations derived from the 8th World Congress on gastrointestinal cancer, Barcelona, 2006. Ann Oncol 2007;18(Suppl. 7):vii1–vii10. 10. Guidelines for the management of patients with pancreatic cancer periampullary and ampullary carcinomas. Gut 2005;54(Suppl. 5):v1–16. 11. Van den BA, Sergeant G, Ectors N, Van SW, Aerts R, Topal B. Patterns of recurrence after curative resection of pancreatic ductal adenocarcinoma. Eur J Surg Oncol 2009;35(6):600–4. 12. Sperti C, Pasquali C, Piccoli A, Pedrazzoli S. Recurrence after resection for ductal adenocarcinoma of the pancreas. World J Surg 1997; 21(2):195–200. 13. Kleeff J, Reiser C, Hinz U, et al. Surgery for recurrent pancreatic ductal adenocarcinoma. Ann Surg 2007;245(4):566–72. 14. Labori KJ, Hjermstad MJ, Wester T, Buanes T, Loge JH. Symptom profiles and palliative care in advanced pancreatic cancer: a prospective study. Support Care Cancer 2006;14(11):1126–33. 15. Witkowski ER, Smith JK, Ragulin-Coyne E, Ng SC, Shah SA, Tseng JF. Is it worth looking? Abdominal imaging after pancreatic cancer resection: a national study. J Gastrointest Surg 2012;16(1):121–8. 16. Tzeng CW, Fleming JB, Lee JE, et al. Yield of clinical and radiographic surveillance in patients with resected pancreatic adenocarcinoma following multimodal therapy. HPB (Oxford) 2012;14(6): 365–72. 17. Tzeng CW, Fleming JB, Lee JE, et al. Defined clinical classifications are associated with outcome of patients with anatomically resectable pancreatic adenocarcinoma treated with neoadjuvant therapy. Ann Surg Oncol 2012;19(6):2045–53.
566
T. Nordby et al. / EJSO 39 (2013) 559e566
18. Heye T, Zausig N, Klauss M, et al. CT diagnosis of recurrence after pancreatic cancer: is there a pattern? World J Gastroenterol 2011; 17(9):1126–34. 19. Shrikhande SV, Kleeff J, Reiser C, et al. Pancreatic resection for M1 pancreatic ductal adenocarcinoma. Ann Surg Oncol 2007; 14(1):118–27. 20. Gleisner AL, Assumpcao L, Cameron JL, et al. Is resection of periampullary or pancreatic adenocarcinoma with synchronous hepatic metastasis justified? Cancer 2007;110(11):2484–92. 21. Adam R, Chiche L, Aloia T, et al. Hepatic resection for noncolorectal nonendocrine liver metastases: analysis of 1,452 patients and development of a prognostic model. Ann Surg 2006;244(4):524–35. 22. Friess H, Kleeff J, Buchler M. Management of cancer recurrence. In: Beger HG, Warshaw AL, Buchler MW, et al, editorss. The pancreas. Blackwell Publishing Ltd.; 2008, p. 772–5. 23. Hishinuma S, Ogata Y, Tomikawa M, Ozawa I, Hirabayashi K, Igarashi S. Patterns of recurrence after curative resection of pancreatic cancer, based on autopsy findings. J Gastrointest Surg 2006;10(4): 511–8.
24. Bilimoria KY, Bentrem DJ, Ko CY, Stewart AK, Winchester DP, Talamonti MS. National failure to operate on early stage pancreatic cancer. Ann Surg 2007;246(2):173–80. 25. Hartwig W, Hackert T, Hinz U, et al. Pancreatic cancer surgery in the new millennium: better prediction of outcome. Ann Surg 2011;254(2):311–9. 26. Katz MH, Hu CY, Fleming JB, Pisters PW, Lee JE, Chang GJ. Clinical calculator of conditional survival estimates for resected and unresected survivors of pancreatic cancer. Arch Surg 2012;147(6):513–9. 27. Northover J. Realism or nihilism in bowel cancer follow-up? Lancet 1998;351(9109):1074–6. 28. D’Angelica M, Hirsch K, Ross H, Passik S, Brennan MF. Surgeon-patient communication in the treatment of pancreatic cancer. Arch Surg 1998;133(9):962–6. 29. Petzel MQ, Parker NH, Valentine AD, et al. Fear of cancer recurrence after curative pancreatectomy: a cross-sectional study in survivors of pancreatic and periampullary tumors. Ann Surg Oncol 2012;19(13):4078–84. 30. Conroy T, Desseigne F, Ychou M, et al. FOLFIRINOX versus gemcitabine for metastatic pancreatic cancer. N Engl J Med 2011;364(19): 1817–25.
EJSO 39 (2013) 559e566
www.ejso.com
Follow-up after curative surgery for pancreatic ductal adenocarcinoma: Asymptomatic recurrence is associated with improved survival T. Nordby a,d, H. Hugenschmidt a, M.W. Fagerland b, T. Ikdahl c, T. Buanes a,d, K.J. Labori a,* a
Department for Hepato-Pancreato-Biliary Surgery, Division of Cancer Medicine, Surgery and Transplantation, Oslo University Hospital, Sognsvannsveien 20, N-0027 Oslo, Norway b Unit of Biostatistics and Epidemiology, Oslo University Hospital, Norway c Department of Oncology, Oslo University Hospital, Norway d Institute of Clinical Medicine, Faculty of Medicine, University of Oslo, Norway Accepted 20 February 2013 Available online 14 March 2013
Abstract Aim: There is no consensus on the optimal follow-up schedule of patients after surgery for pancreatic cancer. In this retrospective study, recurrence and survival were investigated for patients presenting with either symptomatic or asymptomatic recurrence. Patient, tumor and treatment characteristics that predicted the length of postrecurrence survival were identified. Methods: Clinical records of 164 patients who underwent a pancreatic resection (R0/R1) for pancreatic ductal adenocarcinoma from January 2000 to December 2010 were retrieved. Patients underwent a systematic follow-up program. Patient, tumor and treatment characteristics were compared between patients with asymptomatic and symptomatic recurrence. Results: Of 164 consecutive patients, 144 patients (88%) had recurrence (29 asymptomatic, 115 symptomatic). The most frequent reported symptoms were abdominal pain, fatigue/weakness, back pain, weight loss, nausea/loss of appetite and jaundice. Median time to recurrence was 12.0 months for asymptomatic and 7.0 months for symptomatic patients (P ¼ 0.036). Median postrecurrence survival was 10.0 months for asymptomatic and 4.0 months for symptomatic patients (P < 0.0001). Median overall survival was 24.5 months for asymptomatic and 11.0 months for symptomatic patients (P < 0.0001). Symptomatic recurrence, disease free survival <12 months, and no adjuvant chemotherapy were the only independent predictors of poor postrecurrence survival. 72% of asymptomatic and 37% of symptomatic patients received oncological treatment. Conclusions: Patients with asymptomatic pancreatic cancer recurrence have improved recurrence-free, postrecurrence and overall survival. Symptoms when recurrence is diagnosed are a good surrogate marker of biological aggressiveness. Detection of asymptomatic recurrence may facilitate patient eligibility for investigational studies or other forms of treatment. Ó 2013 Elsevier Ltd. All rights reserved. Keywords: Pancreatic neoplasms; Pancreaticoduodenectomy; Recurrence; Follow-up studies; Survival
Background Pancreatic cancer is the fourth leading cause of cancerrelated deaths in Europe and the United States.1,2 Surgical resection provides the only chance of cure. The median survival of resected patients is 11e20 months and the 5-year survival rate 10e20% in recent reports.3,4 There is no consensus on the optimal follow-up schedule of patients after surgery for pancreatic cancer. All of the recommendations * Corresponding author. Tel.: þ47 23070000; fax: þ47 22117470. E-mail address: [email protected] (K.J. Labori). 0748-7983/$ - see front matter Ó 2013 Elsevier Ltd. All rights reserved. http://dx.doi.org/10.1016/j.ejso.2013.02.020
on surveillance are based on low level evidence or no evidence.5 Recently, a lack of established guidelines for surveillance for pancreatic cancer patients has been observed in the United States.6 The inconsistency of follow-ups is revealed by the fact that the leading societies propose different guidelines.7e10 As a consequence, follow-up strategies may differ between hospitals depending on preference of physicians. It is not clear whether survival and quality of life of resected pancreatic cancer patients are improved by follow-up and early detection of recurrence. The prognosis of recurrent pancreatic cancer is dismal with a median
560
T. Nordby et al. / EJSO 39 (2013) 559e566
postrecurrence survival of 3e7 months.11,12 Given the poor survival of patients with recurrent pancreatic cancer, the prognostic effect of early detection seems doubtful. However, with surgery being the only option with a potential to cure, it has been hypothesized that reresection for localized recurrence might open a small window of opportunity for some patients.13 In the context of a systematic follow-up program by medical history, physical examination and computed tomography (CT) every 6 months after a curative pancreatic resection, recurrence and survival were investigated for patients presenting with either symptomatic or asymptomatic recurrence. In addition, we sought to identify patient, tumor and treatment characteristics that predicted the length of postrecurrence survival. Patients and methods Patient selection and data collection From January 2000 to December 2010, 187 patients with pancreatic ductal adenocarcinoma (PDAC) underwent surgical resection with curative intent. Final date of data collection was December 31, 2011. The study was approved by the Data Protection Officer for Research. Information about the patients was recorded retrospectively using hospital records. In one patient the only available information came from the patient’s family physician. PDAC was confirmed histologically in all cases. Patients with other histological variants were excluded. Preoperative workup included chest CT or X-ray and abdominal CT to evaluate primary and metastatic tumor sites. Until 2008 the pathologist reported a margin positive (R1) only if tumor cells were present at the surface (clearance equals 0 mm). In 2008 the definition was changed to a 1 mm clearance. From 2005 adjuvant fluorouracil and leucovorin was administered every second week for six months to patients <75 years old, ECOG 0-1. Follow-up All patients were followed regularly with history and physical examination to pursuit postoperative complications and symptoms. Chest and abdominal CT were performed every 6 months or if the patients had symptoms suspect of a recurrence. Recurrence was defined as radiological evidence of intra-abdominal soft tissue around the surgical site or of distant metastasis. All patients had a CT performed to evaluate site of first recurrence, except two patients who had an abdominal ultrasound performed. If possible a recurrence was confirmed histologically. Symptomatic patients had a symptom or finding that was new and had increased in either severity or frequency since the last follow up. Asymptomatic patients had no new symptoms or findings at a scheduled visit. Patients with recurrence were referred to the Department of Oncology to be
considered for gemcitabine or radiotherapy with or without concurrent capecitabine, and to the Palliative Care Unit for assessment of symptoms and to receive the best palliation of symptoms, as previously described.14 Statistics Patient, tumor and treatment characteristics were compared between patients with asymptomatic and symptomatic recurrence using the WilcoxoneManneWhitney (age, tumor size, and lymph node ratio) tests for continuous variables and the Pearson chi-squared (sex, nodal status, resection margins, site of first recurrence, and adjuvant chemotherapy), Fisher-Freeman-Halton (type of resection), and WilcoxoneManneWhitney (tumor stage, tumor grade, and AJCC stage) tests for categorical variables. Cox regression analyses were used to assess the prognostic capacity of patient and tumor characteristics on time from recurrence to death. First, univariable models were fitted. Second, all variables with P < 0.20 in univariable analysis were included in a multivariable model. Third, the final model was obtained after removing, one at a time, variables with P > 0.05. The proportional hazard assumption was investigated using a test based on Schoenfeld residuals. Estimates of the median overall survival time and disease free survival time from resection, and survival time after detection of recurrence were obtained from the KaplaneMeier estimator. The log-rank test was used to compare survival functions. The statistical analyses were done using Stata 12.0 (StataCorp, College Station, Texas). All P-values are two-tailed. Results Patient characteristics A total of 187 patients underwent surgical resection for PDAC. Patients who died within 3 months after surgery (n ¼ 9) or had non-cancer related death (n ¼ 4) were excluded from the study. Patients with macroscopic residual tumor (n ¼ 7) or distant metastasis at the time of resection (n ¼ 3) were also excluded from further analysis. Accordingly, 164 patients were included in the study. There were 72 women and 92 men. Median age at surgery was 68 years (range 34e84 years). The procedures performed included pylorus-preserving pancreatoduodenectomy (n ¼ 85), pancreatoduodenectomy with antrectomy (n ¼ 63), distal pancreatectomy with splenectomy (n ¼ 10), and total pancreatectomy with splenectomy (n ¼ 6). Recurrence Of the 164 patients, 144 had recurrences, whereas 20 patients were free of recurrence. Patterns of recurrence and patient demographics at the time of initial surgery are listed in Table 1. Twenty-nine patients were asymptomatic when recurrence was detected, whereas 115 patients either
T. Nordby et al. / EJSO 39 (2013) 559e566
561
Table 1 Patient characteristics based on the presence or absence of symptoms at the detection of pancreatic cancer recurrence in 144 patients. Variable Sex, no. Male Female Age, median (range) years Type of resection, no. Whipple Distal resection Total pancreatectomy Tumor stage, no. T1 T2 T3 T4 Tumor grade, no. G1 G2 G3/4 Tumor size, median (range), cm Nodal status, no. N0 N1 Lymph node ratio, median (range) AJCC stage, no. IA IB IIA IIB III Resection margins, no. R0 R1 Site of first recurrence, no. Locoregional Distant Locoregional þ distant Adjuvant chemotherapy, no. Yes No
Asymptomatic recurrence n ¼ 29
Symptomatic recurrence n ¼ 115
P-value
19 10 68 (50e79)
64 51 68 (34e84)
P ¼ 0.34
26 3 0
104 6 5
P ¼ 0.35
0 7 22 0
7 21 85 2
P ¼ 0.97
1 23 5 3.0 (1.8e8.0)
3 86 26 3.0 (1.0e7.0)
P ¼ 0.51
14 15 0.06 (0.0e0.8)
51 64 0.10 (0.0e1.0)
0 4 10 15 0
7 12 31 63 2
P ¼ 0.79
14 15
58 57
P ¼ 0.84
7 18 4
40 40 35
P ¼ 0.024
14 15
38 72
P ¼ 0.13
consulted a physician (n ¼ 77) because of symptoms suggesting recurrence or had such symptoms at the time they came to the hospital for a regular follow-up (n ¼ 38). Except for site of first recurrence, no differences in clinicopathologic variables were evident between patients with asymptomatic or symptomatic recurrence (Table 1). Distant recurrence only was more common in asymptomatic patients (62% vs 35%; P ¼ 0.007). Locoregional recurrence only was similarly frequent in the two groups (24% vs 35%; P ¼ 0.28). Combined locoregional and distant recurrence was slightly more common in symptomatic patients (14% vs 30%; P ¼ 0.072). Distant recurrences were found in liver (n ¼ 72), intraabdominal lymph nodes (n ¼ 37), peritoneum (n ¼ 20), and lung (n ¼ 16). There were no significant differences between the two groups except for peritoneal recurrence that was only observed in symptomatic patients. Eight patients with recurrence (four in each group) were alive at the time of final data collection. The reported
P ¼ 0.96
P ¼ 0.43 P ¼ 0.70 P ¼ 0.87
symptoms were abdominal pain (n ¼ 85), fatigue/weakness (n ¼ 40), back pain (n ¼ 24), weight loss (n ¼ 21), nausea/ loss of appetite (n ¼ 21), jaundice (n ¼ 11), cachexia (n ¼ 5), diarrhea (n ¼ 4), and miscellaneous (n ¼ 12). Patient survival Median disease-specific survival was 24.5 (95%CI: 16.0, 31.0) months for asymptomatic and 11.0 (95%CI: 10.0, 13.5) months for symptomatic patients (P < 0.0001) (Fig. 1a). Median time to recurrence was 12.0 (95%CI: 6.0, 14.0) months for asymptomatic and 7.0 (95%CI: 6.0, 8.0) months for symptomatic patients (P ¼ 0.036) (Fig. 1b). Median postrecurrence survival was 10.0 (95% CI: 7.0, 15.0) months for asymptomatic and 4.0 (95%CI: 3.0, 5.0) months for symptomatic patients (P < 0.0001) (Fig. 1c). Overall, an isolated locoregional recurrence was found in 47 patients (33%), a combined locoregional and
562
T. Nordby et al. / EJSO 39 (2013) 559e566
distant recurrence in 39 patients (27%), and an isolated distant recurrence in 58 patients (40%) (Table 1). The type of cancer recurrence did not significantly influence survival, with a median postrecurrence survival of 5.0 (95%CI: 3.5, 6.0) months in locoregional recurrence, 5.0 (95%CI: 3.5, 6.0) months in combined locoregional/distant recurrence, and 4.0 (95%CI: 3.0, 6.0) months in distant recurrence (P ¼ 0.92), with overall survival of 15.0 (95%CI: 10.0, 18.0), 12.0 (95%CI: 8.0, 15.0) and 13.0 (95%CI: 10.5, 16.5) months (P ¼ 0.83), respectively. Predictors of post-recurrence survival Treatments given after detection of recurrent disease are listed in Table 2. Eight of 29 asymptomatic patients and 69 of 115 symptomatic patients received best supportive care only. Symptomatic recurrence (HR ¼ 2.75 [1.71, 4.42]; P < 0.001), disease free survival less than 12 months (HR ¼ 1.95 [1.33, 2.86]; P ¼ 0.001), and no adjuvant chemotherapy (HR ¼ 1.72 [1.19, 2.49]; P ¼ 0.004) were the only independent predictors of poor postrecurrence survival (Table 3). Discussion To improve the prognosis of resected pancreatic cancer it is necessary to focus on the situation when tumor recurrence takes place. Two recent studies from large population-based data sources have evaluated the posttreatment surveillance patterns in the United States.6,15 Sheffield found a lack of established guidelines for surveillance after curative resection for pancreatic cancer, implying a need for evaluation and standardization of surveillance protocols.6 However, Witkowski found no significant survival benefit among patients who received routinely CT.15 The present study evaluates a surveillance strategy with regular CT every 6 months after a curative pancreatic resection. Our study shows that patients with asymptomatic pancreatic cancer recurrence have improved recurrence-free, postrecurrence and overall survival. In Table 2 Treatment after detection of pancreatic cancer recurrence.
Figure 1. a: Overall survival in 144 patients with pancreatic cancer who had asymptomatic (n ¼ 29) or symptomatic (n ¼ 115) recurrent disease after resection (P < 0.0001). b: Disease free survival in 144 patients with pancreatic cancer who had asymptomatic (n ¼ 29) or symptomatic (n ¼ 115) recurrent disease after resection (P ¼ 0.036). c: Survival after detection of recurrence in 144 patients with pancreatic cancer who had asymptomatic (n ¼ 29) or symptomatic (n ¼ 115) recurrent disease after resection (P < 0.0001).
Best supportive care only Oncological treatment Chemotherapy only Radiotherapy only Chemoradiotherapy Liver resection or RFA Chemotherapy þ Vaccination Radiotherapy þ Vaccination Vaccination only
Asymptomatic recurrence n ¼ 29
Symptomatic recurrence n ¼ 115
8 21 12 1 4 2 2 0 0
72 43 29 4 6 0 1 2 1
RFA ¼ radiofrequency ablation (liver).
T. Nordby et al. / EJSO 39 (2013) 559e566
563
Table 3 Univariable and multivariable analyses of the prognostic variables associated with early death after pancreatic cancer recurrence in 144 patients with complete follow up data. Variable Recurrence Asymptomatic Symptomatic Gender Female Male Age Tumor stage T1 T2 T3 T4 Nodal status N0 N1 Lymph node ratio AJCC stage IA IB IAA IIB III Tumor grade G1 G2 G3/G4 Tumor size Resection margins R0 R1 Site of first recurrence Distant only Locoregional only Locoregional þ distant Disease free interval 12 months <12 months Adjuvant chemotherapy Yes No a
No.
Univariable HRa (95%CI); P-value
Multivariable HRa (95%CI); P-value
29 115
Ref. 2.90 (1.81, 4.64); P < 0.001
Ref. 2.75 (1.71, 4.42); P < 0.001
61 83 144
Ref. 0.79 (0.56, 1.13); P ¼ 0.20 1.01 (0.99, 1.02); P ¼ 0.44
7 28 107 2
Ref. 0.67 (0.29, 1.55); P ¼ 0.35 0.68 (0.32, 1.48); P ¼ 0.34 0.91 (0.19, 4.37); P ¼ 0.90
65 79 144
Ref. 1.07 (0.76, 1.52); P ¼ 0.69 0.88 (0.43, 1.82); P ¼ 0.74
7 16 41 78 2
Ref. 0.76 0.69 0.78 0.99
(0.31, (0.30, (0.36, (0.21,
1.87); 1.55); 1.72); 4.79);
P P P P
¼ ¼ ¼ ¼
0.55 0.37 0.54 0.99
4 109 31 144
Ref. 2.31 (0.71, 7.47); P ¼ 0.16 2.58 (0.76, 8.79); P ¼ 0.13 0.92 (0.80, 1.07); P ¼ 0.29
72 72
Ref. 0.86 (0.61, 1.21); P ¼ 0.39
58 47 39
Ref. 0.94 (0.63, 1.40); P ¼ 0.75 1.02 (0.66, 1.56); P ¼ 0.94
46 96
Ref. 1.87 (1.29, 2.72); P ¼ 0.001
Ref. 1.95 (1.33, 2.86); P ¼ 0.001
52 92
Ref. 1.63 (1.14, 2.35); P ¼ 0.008
Ref. 1.72 (1.19, 2.49); P ¼ 0.004
HR ¼ Hazard ratio obtained from Cox regression.
a similar study from MD Anderson Cancer Center, a surveillance strategy with CT every 3e4 months the first 2 years after PDAC surgery was evaluated.16 In Tzeng’ study, 77.7% of the patients underwent neoadjuvant therapy, and probably excluded from surgery up to 25% of patients initially found clinically resectable.17 In contrast, our centre followed patients every 6 months and no patients had neoadjuvant therapy. Tzeng found no difference in median time to recurrence, whereas survival after detection was 5.1 months in symptomatic patients and 13.0 months in asymptomatic patients (P < 0.001), compared to 4 months and 10 months in our study.16 A more intensive follow up with CT every 3e4 months identified 54.6% of recurrences prior to the onset to symptoms, whereas our surveillance program with CT every 6 months identified 20.1% of patients.
Treatment for recurrence Some authors advocate that it is essential to identify tumor recurrence early in order to offer patients further disease controlling measures or potentially curative options.18 The only measure taken after the detection of recurrence was best supportive care in 63% of symptomatic patients and 28% of asymptomatic patients. Symptomatic recurrence was an independent predictor of poor postrecurrence survival. We suggest that symptoms are a good surrogate marker of biological aggressiveness and more disease burden. Management of symptomatic patients is probably focused upon symptom control rather than initiating more aggressive maneuvers with chemotherapy. Only two patients with asymptomatic liver metastases had a potentially
564
T. Nordby et al. / EJSO 39 (2013) 559e566
curative therapeutic intervention (RFA and liver resection). Both patients had a relatively short posttreatment survival. Three of 47 patients with locoregional recurrence alone and three of 58 patients with liver metastases alone had postrecurrence survival >20 months and >30 months, respectively. Reresection was never discussed in these patients. Only a few studies address the issue of resection of metastatic or recurrent disease in pancreatic cancer.13,19e21 Simultaneous pancreatic and liver resections are not generally recommended.19,20 Adam found that hepatic resection may be warranted in those patients who demonstrate good tumor biology, and reported a 5-year survival of 20% in selected patients with liver metastases after resection for pancreatic adenocarcinoma.21 Kleeff published their results of re-resection of localized recurrence for pancreatic adenocarcinoma in a group of eight patients.13 Compared to a matched control group of patients who did not undergo re-resection, median survival was significantly longer for patients undergoing re-resection (29 vs 14.5 months; P < 0.0001). According to these studies, resection for liver metastases and locoregional recurrence might open a small window of opportunity for a small subgroup of patients. Location of recurrence Patients with pancreatic cancer recurrence are managed similarly to nonresectable patients.22 Chemotherapy and/or chemoradiation intend to control tumor progression, improve survival and quality of life, but the survival benefit is limited.8 The prognosis of recurrent pancreatic cancer is dismal. Sperti found a median survival of 7 months in local recurrence and 3 months in hepatic recurrence.12 As shown in other studies, the type of cancer recurrence did not significantly influence the survival in our study.11 Van den Broeck reported the first location of recurrence to be locoregional in 17%, combined locoregional and distant in 23% and distant in 60%, compared with rates of 33%, 27%, and 40%, respectively, in our study.11 Overall, distant metastasis was found in 83% and 67% of patients in these two studies. Even though local recurrence occurs frequently, it is rarely a direct cause of death since most patients die from metastases.23 Accordingly, it is likely that treatment focused on local control cannot improve the survival of patients with recurrent pancreatic cancer. Thus, treatment regimens that are effective against systemic metastasis are needed. Recommendations for surveillance There are significant differences in the recommendations for surveillance of pancreatic cancer patients following resection.7e9 National Comprehensive Cancer Network (NCCN) recommend a close surveillance with history, physical examination, CA 19-9 and CT every 3e6 months the first two years and then annually.8 European Society of Medical Oncology find no advantage in an earlier detection
of recurrences, but recommend that in the case of elevated preoperative CA19-9 levels, CA 19-9 assessment can be performed every 3 months for 2 years and abdominal CT every 6 months.7 However, all of the recommendations on surveillance are based on low level evidence or no evidence.5,8 Ultimately, the decision regarding the intensiveness of follow-up is left to the preferences of the treating physician after discussion with the patient. The surgical philosophy may vary widely from total nihilism to optimism. A striking underuse of pancreatectomy for localized pancreatic cancer in the USA has been identified.24 Furthermore, CT testing for patients undergoing resection for pancreatic cancer seems to be obtained to evaluate symptoms rather than for routine surveillance.6 These facts may reflect a nihilistic attitude toward pancreatic cancer. Physicians may need to approach surveillance on a case-by-case basis according to estimated recurrence risk and likelihood for therapy benefit should recurrent disease be identified.6 In a recent study, patients were stratified into 4-risk-groups depending on patient, tumor and resection characteristics at the time of surgery, with 5-year survival varying between 0% and 54.5%.25 However, Katz has recently shown that the survival expectancy of all patients with PDAC improves over time regardless of patient characteristics or the pathologic variables associated with the primary cancer.26 Some may argue that follow-up should be able to detect at least some recurrent disease amenable to curative treatment, and not only bring forward the time of diagnosis of incurable disease.27 However, there are several reasons for follow-up.27 Follow-up may be needed to manage post-surgical complications, address feeding and nutritional issues associated with a pancreatic resection per se and to perform audit and quality control of surgical outcomes. In addition follow-up may reassure and provide psychological support to the patients. D’Angelica found that surgeons need to develop effective long-term support systems for their patients after surgery.28 Fear of cancer recurrence is pervasive following resection of pancreatic neoplasms.29 Regular follow-up and appreciation of improvements in conditional survival over time may gradually reduce the anxiety and fear of disease progression that exert a negative impact on the quality of life of survivors of cancer and their caregivers.26,29 Limitations The findings in our study must be interpreted with some caution. The key limitation of our study is the retrospective nature of our analysis. The methodology rely on a retrospective audit of documentation of patient symptoms. However, our institution has had a focus on symptom profiles in pancreatic cancer patients during the study period, and the patients were assessed for gastrointestinal or cancer related symptoms at all consultations.14 Furthermore, an important end point of cancer treatment, apart from cure rate and survival, concerns quality of life. This issue was not evaluated
T. Nordby et al. / EJSO 39 (2013) 559e566
with established instruments for assessment. How detection of recurrent disease affect quality of life of asymptomatic pancreatic cancer patients by confronting them with their dismal prognosis, remains to be established. Lastly, the number of patients with asymptomatic recurrence is small. The power to detect small and moderate differences between asymptomatic and symptomatic patients thereby is low. Non-significant between-group differences should be interpreted with caution, and definite conclusions with respect to these differences are not warranted. The large differences in overall and postrecurrence survival, on the other hand, are highly significant (P < 0.0001), and we are confident that these results represent true differences. Surveillance strategy The important question is whether CT should be performed only in case of signs and symptoms of recurrent cancer or at regular intervals. Symptoms suggesting recurrence in our study were abdominal pain, fatigue/weakness, back pain, weight loss and nausea/loss of appetite. It may be difficult to avoid performing CT scans in patients after surgery for pancreatic cancer because they may frequently experience symptoms that would prompt diagnostic imaging.6 High volume centres both in Europe and the United States recommend regular follow up with CT after surgery for PDAC.6,13,16,18 Heidelberg University Hospital point out that it is essential to identify tumor recurrence early in order to offer patients further disease controlling measures or potentially curative options.13,18 Approximately 3/4 of asymptomatic patients in our study underwent subsequent oncological treatment compared to only 1/3 of symptomatic patients. One might argue that this alone could explain the better post-recurrence survival in the asymptomatic group, and this should be kept in mind when interpreting the results. However, both disease-free and postrecurrence survival were significantly better in the asymptomatic group. We find that our data support the results from MD Anderson Cancer Center that CT surveillance might identify patients with good performance status and tumor biology that are most likely to benefit from subsequent therapy.16 Recently, a chemotherapy regimen consisting of fluorouracil, leucovirin, irinotecan and oxaliplatin has shown to be an effective treatment for patients with metastatic PDAC, and may represent a treatment opportunity for patients with asymptomatic recurrence with general good health.16,30 In the absence of evidence-based guidelines, we recommend surveillance after potentially curative surgery for PDAC with regular CT. Most patients develop recurrence within 2 years after surgery, suggesting a more intensive follow-up the first 2 years after surgery.6,8 In asymptomatic patients, detection of disease through CT may facilitate patient eligibility for investigational studies or other forms of treatment. However, the patients have to be aware of the limitations of detection of an asymptomatic recurrence and that no curative treatment is currently
565
available. Larger prospective trials are needed to determine whether early detection of asymptomatic recurrence can lead to new therapy options and, ultimately, improved survival rates and quality of life in patients with resected PDAC. Conflict of interest The authors have no conflict of interest. References 1. Malvezzi M, Arfe A, Bertuccio P, Levi F, La VC, Negri E. European cancer mortality predictions for the year 2011. Ann Oncol 2011;22(4): 947–56. 2. Siegel R, Ward E, Brawley O, Jemal A. Cancer statistics, 2011: the impact of eliminating socioeconomic and racial disparities on premature cancer deaths. CA Cancer J Clin 2011;61(4):212–36. 3. Hoem D, Viste A. Improving survival following surgery for pancreatic ductal adenocarcinomaea ten-year experience. Eur J Surg Oncol 2012;38(3):245–51. 4. Winter JM, Brennan MF, Tang LH, et al. Survival after resection of pancreatic adenocarcinoma: results from a single institution over three decades. Ann Surg Oncol 2012;19(1):169–75. 5. Poonacha TK, Go RS. Level of scientific evidence underlying recommendations arising from the national comprehensive cancer network clinical practice guidelines. J Clin Oncol 2011;29(2):186–91. 6. Sheffield KM, Crowell KT, Lin YL, Djukom C, Goodwin JS, Riall TS. Surveillance of pancreatic cancer patients after surgical resection. Ann Surg Oncol 2012;19(5):1670–7. 7. Cascinu S, Falconi M, Valentini V, Jelic S. Pancreatic cancer: ESMO clinical practice guidelines for diagnosis, treatment and follow-up. Ann Oncol 2010;21(Suppl. 5):v55–8. 8. Pancreatic adenocarcinoma. National comprehensive cancer network clinical practice guidelines in oncology. http://www.nccn.org/professionals/ physician_gls/f_guidelines.asp#pancreatic [accessed 01.10.12]. 9. Verslype C, Van CE, Dicato M, et al. The management of pancreatic cancer. Current expert opinion and recommendations derived from the 8th World Congress on gastrointestinal cancer, Barcelona, 2006. Ann Oncol 2007;18(Suppl. 7):vii1–vii10. 10. Guidelines for the management of patients with pancreatic cancer periampullary and ampullary carcinomas. Gut 2005;54(Suppl. 5):v1–16. 11. Van den BA, Sergeant G, Ectors N, Van SW, Aerts R, Topal B. Patterns of recurrence after curative resection of pancreatic ductal adenocarcinoma. Eur J Surg Oncol 2009;35(6):600–4. 12. Sperti C, Pasquali C, Piccoli A, Pedrazzoli S. Recurrence after resection for ductal adenocarcinoma of the pancreas. World J Surg 1997; 21(2):195–200. 13. Kleeff J, Reiser C, Hinz U, et al. Surgery for recurrent pancreatic ductal adenocarcinoma. Ann Surg 2007;245(4):566–72. 14. Labori KJ, Hjermstad MJ, Wester T, Buanes T, Loge JH. Symptom profiles and palliative care in advanced pancreatic cancer: a prospective study. Support Care Cancer 2006;14(11):1126–33. 15. Witkowski ER, Smith JK, Ragulin-Coyne E, Ng SC, Shah SA, Tseng JF. Is it worth looking? Abdominal imaging after pancreatic cancer resection: a national study. J Gastrointest Surg 2012;16(1):121–8. 16. Tzeng CW, Fleming JB, Lee JE, et al. Yield of clinical and radiographic surveillance in patients with resected pancreatic adenocarcinoma following multimodal therapy. HPB (Oxford) 2012;14(6): 365–72. 17. Tzeng CW, Fleming JB, Lee JE, et al. Defined clinical classifications are associated with outcome of patients with anatomically resectable pancreatic adenocarcinoma treated with neoadjuvant therapy. Ann Surg Oncol 2012;19(6):2045–53.
566
T. Nordby et al. / EJSO 39 (2013) 559e566
18. Heye T, Zausig N, Klauss M, et al. CT diagnosis of recurrence after pancreatic cancer: is there a pattern? World J Gastroenterol 2011; 17(9):1126–34. 19. Shrikhande SV, Kleeff J, Reiser C, et al. Pancreatic resection for M1 pancreatic ductal adenocarcinoma. Ann Surg Oncol 2007; 14(1):118–27. 20. Gleisner AL, Assumpcao L, Cameron JL, et al. Is resection of periampullary or pancreatic adenocarcinoma with synchronous hepatic metastasis justified? Cancer 2007;110(11):2484–92. 21. Adam R, Chiche L, Aloia T, et al. Hepatic resection for noncolorectal nonendocrine liver metastases: analysis of 1,452 patients and development of a prognostic model. Ann Surg 2006;244(4):524–35. 22. Friess H, Kleeff J, Buchler M. Management of cancer recurrence. In: Beger HG, Warshaw AL, Buchler MW, et al, editorss. The pancreas. Blackwell Publishing Ltd.; 2008, p. 772–5. 23. Hishinuma S, Ogata Y, Tomikawa M, Ozawa I, Hirabayashi K, Igarashi S. Patterns of recurrence after curative resection of pancreatic cancer, based on autopsy findings. J Gastrointest Surg 2006;10(4): 511–8.
24. Bilimoria KY, Bentrem DJ, Ko CY, Stewart AK, Winchester DP, Talamonti MS. National failure to operate on early stage pancreatic cancer. Ann Surg 2007;246(2):173–80. 25. Hartwig W, Hackert T, Hinz U, et al. Pancreatic cancer surgery in the new millennium: better prediction of outcome. Ann Surg 2011;254(2):311–9. 26. Katz MH, Hu CY, Fleming JB, Pisters PW, Lee JE, Chang GJ. Clinical calculator of conditional survival estimates for resected and unresected survivors of pancreatic cancer. Arch Surg 2012;147(6):513–9. 27. Northover J. Realism or nihilism in bowel cancer follow-up? Lancet 1998;351(9109):1074–6. 28. D’Angelica M, Hirsch K, Ross H, Passik S, Brennan MF. Surgeon-patient communication in the treatment of pancreatic cancer. Arch Surg 1998;133(9):962–6. 29. Petzel MQ, Parker NH, Valentine AD, et al. Fear of cancer recurrence after curative pancreatectomy: a cross-sectional study in survivors of pancreatic and periampullary tumors. Ann Surg Oncol 2012;19(13):4078–84. 30. Conroy T, Desseigne F, Ychou M, et al. FOLFIRINOX versus gemcitabine for metastatic pancreatic cancer. N Engl J Med 2011;364(19): 1817–25.