j o u r n a l o f s u r g i c a l r e s e a r c h j a n u a r y 2 0 1 8 ( 2 2 1 ) 1 5 2 e1 6 0
Available online at www.sciencedirect.com
ScienceDirect journal homepage: www.JournalofSurgicalResearch.com
Use of imaging during symptomatic follow-up after resection of pancreatic ductal adenocarcinoma Vincent P. Groot, MD,a Lois A. Daamen, MD,a Jeroen Hagendoorn, MD, PhD,a Inne H.M. Borel Rinkes, MD, PhD,a Hjalmar C. van Santvoort, MD, PhD,a,b and I. Quintus Molenaar, MD, PhDa,* a b
Department of Surgery, UMC Utrecht Cancer Center University Medical Center Utrecht, Utrecht, The Netherlands Department of Surgery, St. Antonius Hospital Nieuwegein, Nieuwegein, The Netherlands
article info
abstract
Article history:
Background: Controversy exists whether follow-up after resection of pancreatic ductal
Received 24 April 2017
adenocarcinoma (PDAC) should include standardized imaging for the detection of disease
Received in revised form
recurrence. The purpose of this study was to evaluate how often patients undergo imaging
3 July 2017
in a setting where routine imaging is not performed. Secondly, the pattern, timing, and
Accepted 11 August 2017
treatment of recurrent PDAC were assessed.
Available online xxx
Materials and methods: This was a post hoc analysis of a prospective database of all consecutive patients undergoing pancreatic resection of PDAC between January 2011 and
Keywords:
January 2015. Data on imaging procedures during follow-up, recurrence location, and
Pancreatic neoplasms
treatment for recurrence were extracted and analyzed. Associations between clinical
Pancreatectomy
characteristics and post-recurrence survival were assessed with the log-rank test and Cox
Recurrence
univariable and multivariable proportional hazards models.
Follow-up studies
Results: A total of 85 patients were included. Seventy-four patients (87%) underwent
Survival
imaging procedures during follow-up at least once, with a mean amount of 3.1 1.9 imaging procedures during the entire follow-up period. Sixty-eight patients (80%) were diagnosed with recurrence, 58 (85%) of whom after the manifestation of clinical symptoms. Additional tumor-specific treatment was administered in 17 of 68 patients (25%) with recurrence. Patients with isolated local recurrence, treatment after recurrence, and a recurrence-free survival >10 mo had longer post-recurrence survival. Conclusions: Even though a symptomatic follow-up strategy does not include routine imaging, the majority of patients with resected PDAC underwent additional imaging procedures during their follow-up period. Further prospective studies are needed to determine the actual clinical value, psychosocial implications, and cost-effectiveness of different forms of follow-up after resection of PDAC. ª 2017 Elsevier Inc. All rights reserved.
* Corresponding author. UMC Utrecht Cancer Center, Department of Surgery, HP G04.228, PO Box 85500, 3508 GA Utrecht, The Netherlands. Tel.: þ31 (88) 7558074; fax: þ31 (30) 2541944. E-mail address:
[email protected] (I.Q. Molenaar). 0022-4804/$ e see front matter ª 2017 Elsevier Inc. All rights reserved. http://dx.doi.org/10.1016/j.jss.2017.08.023
groot et al imaging during follow-up after resection of pdac
Introduction Pancreatic ductal adenocarcinoma (PDAC) is the fourth leading cause of cancer-related mortality in Europe, with more than 43,000 deaths predicted for 2017.1,2 Due to late onset of symptoms, up to 80% of patients present with locally advanced or metastatic disease.3 Most prospective studies for advanced pancreatic cancer involve locally advanced disease, primary metastatic disease, and recurrent disease after surgery.4 It has been recently acknowledged, however, that survival outcomes for patients with recurrent disease can be superior when compared to locally advanced and primary metastatic disease, possibly warranting additional treatment for these patients.4,5 Management of recurrent PDAC is less well established as it is for other stages of PDAC. Furthermore, controversy exists as to whether follow-up after pancreatic resection should include imaging testing for the diagnosis of recurrence.6-8 The European Society for Medical Oncology, for instance, only recommends symptomatic follow-up that concentrates on symptoms, nutrition, and psychosocial support without additional diagnostics for diagnosing recurrence.9 Adhering to the European guidelines, follow-up at our institution consists of regular checkups focusing on postoperative symptoms without imaging for the detection of recurrence. The purpose of this study was to evaluate how often patients undergo imaging in daily practice based on clinical suspicion of disease recurrence in a setting where routine imaging is not performed. Secondly, the pattern, timing, and treatment of recurrent PDAC resulting from symptomatic follow-up were assessed. Finally, potential predictive factors for extended survival after recurrence were evaluated.
Materials and methods Patient selection The primary source of information for this study was the prospective pancreatic surgery database of the UMC Utrecht Cancer Center Department of Surgery, which has been approved by our Institutional Review Board for data acquisition and query. Only patients undergoing pancreatic resection of PDAC between January 2011 and January 2015 were included for analysis. Patients with 30-d perioperative mortality were excluded.
Primary treatment Resectability and staging were established using pancreatic protocol computed tomography (CT) and were discussed in a multidisciplinary setting. A team specialized in hepatobiliary and pancreatic surgeries performed all pancreatic resections. If their performance status allowed it, patients were offered adjuvant chemotherapy in the form of gemcitabine (1000 mg/m2) every 4 wk for six cycles following Dutch national guidelines based on the “ESPAC” and “CONKO” trials.10 All patients included in this study were considered resectable or borderline resectable and underwent upfront
153
surgery. Neoadjuvant therapy for resectable and/or borderline and/or locally advanced PDAC is a relatively recent development in the Netherlands and is generally only performed in the setting of current multi-institutional and national trials. Cancer stage was defined according to the seventh edition of the American Joint Committee on Cancer staging system.11
Follow-up Patient follow-up occurred at the outpatient clinic of our institution to a standardized schedule of visits. Following completion of all therapy, regular follow-up visits were scheduled at 3, 6, 9, 12, 18, and 24 mo, followed by yearly check-ups up to 5 y. After 5 y, patients were discharged from follow-up. The follow-up was performed by either an hepatopancreatobiliary surgeon or a medical oncologist and consisted mainly of an inquiry concerning postoperative symptoms and signs of endocrine and exocrine pancreatic insufficiency, followed by a physical examination and weight measurement. Imaging was explicitly not part of our standard follow-up protocol. For this study, the actual use of imaging procedures was assessed to evaluate the results and implications of this follow-up protocol. Only those examinations ordered with a clear inquiry regarding tumor recurrence were included. Radiological or histological evidence was required for the diagnosis of recurrent disease. Only the first site of recurrence was documented, using four mutually exclusive categories. Local recurrence was defined as recurrence in the remnant pancreas or in the surgical bed, such as soft tissue along the celiac or superior mesenteric artery, aorta or around the pancreatojejunostomy site. Distant recurrence was defined as recurrence restricted to a single organ or site. When both isolated local recurrence and isolated distant recurrence were revealed simultaneously, recurrence was defined as “local þ distant” recurrence. When multiple distant sites and/or carcinomatosis peritonei were revealed, recurrence was recorded as multiple. Records were further analyzed to determine the type of treatment received after detection of recurrence. Symptomatic recurrences were defined as those discovered due to a significant patient-initiated complaint that was new or had increased in severity or frequency.
Statistical analysis Summary statistics were obtained using established methods. Patients were dichotomized based on whether recurrence occurred fewer or greater than 10 mo following surgery. This used cutoff point of 10 mo recurrence-free survival (RFS) was found to be the most significant value for showing differences in post-recurrence survival when using a minimum P-value approach to analyze our data and was similar to proposed cutoff in previously published surgical data.12,13 The proportion of patients receiving imaging procedures for each month of follow-up was calculated by dividing the amount of patients receiving one or more imaging procedure(s) for the detection of recurrence by the amount of uncensored patients. Censoring occurred at date of recurrence, death, or last followup if recurrence did not occur. Also, the median interval between surveillance imaging procedures was calculated for
154
j o u r n a l o f s u r g i c a l r e s e a r c h j a n u a r y 2 0 1 8 ( 2 2 1 ) 1 5 2 e1 6 0
each patient from day of surgery to the date of recurrence by dividing this interval with the number of imaging procedures actually performed in that period. Overall survival (OS) and post-recurrence survival were defined as the time from the date of operation (OS) or the date of recurrence (post-recurrence survival) to either death from any cause or last follow-up. RFS was defined as the time interval between the date of the operation and either date of recurrence or last follow-up if recurrence was not observed. KaplaneMeier curves were used to estimate median OS, RFS, and post-recurrence survival with corresponding 95% confidence intervals (CIs). The log-rank test was performed to compare between subgroups. Univariable and multivariable analyses were performed using Cox proportional hazard models to identify potential predictive factors for post-recurrence survival. A two-tailed P value of 5
9 (12%)
3 (4%)
0 (0%)
0 (0%)
0 (0%)
0 (0%)
2.5 (1.5)
2.1 (1.2)
1.8 (1.1)
1.0 (1.4)
0.7 (1.0)
4.9 (1-30)
5.1 (1-30)
3.1 (1-16)
3.8 (0-20)
Mean no. of procedures (SD) Median time interval, months (range)
3.1 (1.9) 5 (1-30)
6 (1-30)
SD ¼ standard deviation.
Table 2. Dichotomization showed that initial patient, tumor, and treatment characteristics did not differ significantly between patients with a RFS 10 mo, except for the administration of adjuvant gemcitabine (P ¼ 0.024). The majority of recurrences (n ¼ 58, 85%) were diagnosed after the manifestation of clinical symptoms. In the 10 asymptomatic patients, recurrence was detected by imaging performed at the explicit wish of the patient. Symptoms leading to detection of recurrence included abdominal and/or back pain in 40 (69%) and/or weight loss in 10 of 58 (17%) symptomatic patients. Further symptoms, including overall malaise and obstipation among others, were seen in 19 patients (33%). At the time of diagnosis of recurrence, most patients had either recurrence at an isolated distant site (n ¼ 22, 32%) or at
multiple sites (n ¼ 23, 34%) (Table 2). Isolated distant sites included the liver (n ¼ 12), lung (n ¼ 4), lymph nodes (n ¼ 3), and bone (n ¼ 3). Isolated local recurrence occurred after a significantly longer median RFS (20 mo, 95% CI ¼ 12-27 mo) when compared to isolated distant recurrence (8 mo, 95% CI ¼ 3-14 mo; P ¼ 0.021) and recurrence at multiple sites (10 mo, 95% CI ¼ 4-15 mo; P ¼ 0.009). Median RFS before local þ distant recurrence was 11 mo (95% CI ¼ 0-22 mo) and did not differ significantly from local recurrence (P ¼ 0.076). Median post-recurrence survival was longer after isolated local recurrence (11 mo, 95% CI ¼ 5-17 mo) when compared to local þ distant recurrence (3 mo, 95% CI ¼ 2-3 mo; P < 0.001) or recurrence at multiple sites (4 mo, 95% CI ¼ 1-6 mo; P ¼ 0.003), but did not differ significantly from isolated distant recurrence (7 mo, 95% CI ¼ 3-10 mo; P ¼ 0.066).
Fig. 2 e Percentage of patients who received CT, MRI, US or fluorodeoxyglucose positron emission tomography for the detection of recurrence per month of follow-up. Patients who recurred or died were censored at the date of event. N indicates the number of patients who survived uncensored at the end of the month. US [ ultrasound. (Color version of figure is available online.)
156
j o u r n a l o f s u r g i c a l r e s e a r c h j a n u a r y 2 0 1 8 ( 2 2 1 ) 1 5 2 e1 6 0
Table 2 e Clinicopathological and treatment characteristics of patients with recurrence. Variable Recurrence, n (%) Male, n (%) Age (y), mean (SD) CA 19-9 (U/mL)
All patients with recurrence
RFS 10 mo
68
35 (51%)
33 (49%)
42 (62%)
23 (66%)
66.6 (8.9)
67.1 (10.3)
19 (58%) 66.0 (7.2)
P value
0.490 0.625
*
37 U/mL, n (%)
36 (80%)
19 (83%)
17 (78%)
Median (IQR)
170 (41-450)
170 (38-540)
175 (42-440)
Operation procedure, n (%) PPPD
52 (77%)
24 (69%)
28 (85%)
7 (10%)
5 (14%)
2 (6%)
Distal pancreatectomy
5 (7%)
4 (11%)
1 (3%)
Total pancreatectomy
3 (4%)
2 (6%)
1 (3%)
Modified Appleby
1 (2%)
0 (0%)
1 (3%)
40 (58%)
16 (46%)
24 (73%)
R0
22 (32%)
10 (29%)
12 (36%)
R1
29 (43%)
13 (37%)
16 (49%)
R2
17 (25%)
12 (34%)
5 (15%)
55 (81%)
26 (74%)
29 (88%)
Resection margin, n (%)
Poor Tumor size (cm), median (IQR)
0.024 0.190
Tumor differentiation, n (%) Well-moderate
0.716 0.322
Whipple
Adjuvant GEM, n (%)
0.722
0.220
13 (19%) 3.7 (2.8-4.5)
9 (26%)
4 (12%)
4.0 (3.0-4.5)
3.5 (2.4-4.2)
T-stage, n (%)
0.052 0.140
2
13 (19%)
4 (11%)
9 (27%)
3
44 (65%)
23 (66%)
21 (64%)
4
11 (16%)
8 (23%)
3 (9%)
Positive lymph nodes, n (%)
53 (78%)
28 (80%)
25 (76%)
0.673
Microscopic perineural invasion, n (%)
62 (91%)
32 (91%)
30 (91%)
1.000
Microscopic perilymphatic invasion, n (%)
28 (41%)
15 (43%)
13 (39%)
0.772
Microscopic vascular invasion, n (%)
41 (60%)
20 (57%)
21 (64%)
0.584
Symptomatic recurrence, n (%)
58 (85%)
29 (83%)
29 (88%)
0.735
Local
10 (15%)
4 (11%)
6 (18%)
Local þ distant
13 (19%)
6 (17%)
7 (21%)
Distant
22 (32%)
13 (37%)
9 (27%)
Multiple
23 (34%)
12 (34%)
11 (33%)
17 (25%)
8 (23%)
9 (27%)
Recurrence site, n (%)
Treatment after recurrence, n (%)
0.755
0.674
P values that are 10 mo (hazard ratio [HR] ¼ 0.42, 95% CI ¼ 0.23-0.76, P ¼ 0.004) and administration of further treatment for recurrence (HR ¼ 0.32, 95% CI ¼ 0.16-0.65, P ¼ 0.002) were independently associated with a longer post-recurrence survival. Furthermore, isolated local recurrence was associated with longer survival than local þ distant recurrence (HR ¼ 8.20, 95% CI ¼ 2.37-28.33, P ¼ 0.001), isolated distant recurrence (HR ¼ 3.77, 95% CI ¼ 1.34-10.54, P ¼ 0.012), or recurrence at multiple sites (HR ¼ 8.98, 95% CI ¼ 2.99-26.97, P < 0.001).
Discussion
B
Fig. 3 e KaplaneMeier plots showing that (A) postrecurrence survival was longer (P [ 0.016) for patients receiving treatment when compared to BSC and (B) postrecurrence survival was longer (P [ 0.013) for patients with RFS of more than 10 mo. (Color version of figure is available online.)
(95% CI ¼ 2-4 mo) found for the group receiving BSC (n ¼ 51) (Fig. 3A). Median RFS did not differ (P ¼ 0.142) between the treatment and BSC group (11 mo versus 10 mo, respectively). Patients with recurrence after more than 10 mo following surgery had a longer median post-recurrence survival (5 mo, 95% CI ¼ 2-8 mo) when compared to patients with recurrence within 10 mo (3 mo, 95% CI ¼ 1-4 mo; P ¼ 0.013) (Fig. 3B). Nine patients out of 68 patients with recurrence (13%) were alive after a median follow-up of 26 mo (14-38 mo).
Although imaging was explicitly not part of our symptomatic follow-up protocol, the majority of patients (87%) in this study underwent additional imaging procedures for the detection of recurrence at least once. Furthermore, patients underwent an average of three imaging procedures during the entire followup period. Most of the recurrences (85%) were diagnosed at a late stage and after the manifestation of clinical symptoms, resulting in only 17 patients (25%) receiving additional treatment for their recurrence. The primary objective of this study was to evaluate how often patients undergo imaging during a symptomatic followup approach based on European guidelines.9 Although our single-institutional, nonrandomized study did not allow for direct comparison of symptomatic surveillance and follow-up with routine imaging, our results did reveal interesting results that differ from studies reporting on diagnostic follow-up. For instance, in our cohort, 85% of recurrences were diagnosed after the manifestation of clinical symptoms. Plausibly, studies employing more intensive follow-up have reported lower rates of symptomatic recurrence (20%-74%).14-16 Furthermore, only 25% (n ¼ 17) of patients in our study received additional treatment for their recurrence, primarily due to advanced disease and poor performance status upon diagnosis of recurrence. This was considerably less than the 44%-84% reported by other studies using regular imaging testing for the detection of recurrence.14-17 A credible explanation for this difference is provided by Tzeng et al., as their study found that additional treatment was less frequently administered to symptomatic patients in whom recurrence was detected at a later stage (91% versus 61%; P < 0.001).14 However, these findings could also be partly explained by the differences in the approach to treatment of recurrence between different institutions.18 When considering the more intensive follow-up of some major North American, European, and Asian institutions compared to those who follow current European guidelines, it is reasonable to assume they will also promote additional treatment to their patients in a more active manner. Above mentioned differences between institutions might further explain some results presented in our study. Firstly, we found a relatively high R2 resection rate of 25% in our
158
j o u r n a l o f s u r g i c a l r e s e a r c h j a n u a r y 2 0 1 8 ( 2 2 1 ) 1 5 2 e1 6 0
Table 3 e Factors associated with post-recurrence survival. Clinical characteristics
Univariable analysis HR
95% CI
R1 versus R0
0.54
R2 versus R0
1.19
Multivariable analysis P value
HR
95% CI
P value
0.29-1.02
0.059
0.50
0.23-1.06
0.071
0.61-2.31
0.610
1.39
0.66-2.96
0.387
Resection margin
T-stage 3 versus 2
2.03
0.94-4.35
0.070
2.34
1.01-5.41
0.048
4 versus 2
2.22
0.86-5.78
0.101
1.73
0.63-4.76
0.287 0.001
Recurrence site Local þ distant versus local
5.31
1.79-15.71
0.003
8.20
2.37-28.33
Distant versus local
2.47
0.92-6.66
0.073
3.77
1.34-10.54
0.012
Multiple versus local
3.57
1.35-9.44
0.010
8.98
2.99-26.97
10 mo versus
