Abstract
Background.
Although IPMN are thought to represent a whole-gland disease, segmental resection remains the most frequently performed treatment. We sought to determine the rates, patterns, and predictors of IPMN progression in the pancreatic remnant following segmental resection of noninvasive or microinvasive IPMN.
Methods.
A prospectively maintained database was queried to identify all patients who underwent resection of noninvasive or microinvasive IPMN (≤ 10 mm of invasive component) between 1989 and 2015. Progression (recurrence) was defined as either the development of cancer, a new IPMN cystic lesion >1 cm or ≥ 50% increase in the diameter of residual IPMN lesions in the remnant. Univariate and multivariate cox regression models were created to determine predictors of progression.
Results.
A total of 319 patients underwent resection for noninvasive and microinvasive IPMN. The median age was 68, 53% had branch-duct (BD) IPMN, and 6% had microinvasive disease. After a median follow-up of 42 months, 71 patients (22%) experienced IPMN progression. Within this group of 71 patients, 11 (16% of recurrence) developed invasive cancer in the pancreatic remnant after a median of 28 months. Twelve patients (17%) experienced progression > 5 years following initial resection. On multivariate analysis, a distal location of the initial lesion was associated with an increased risk of progression (multivariate hazards ratio = 2.43, confidence interval 1.47–4.0, p < 0.001).
Conclusions.
In this study, 22% of patients had disease progression following resection of noninvasive or microinvasive IPMN; 16% of these progressions represented invasive disease. These patients represent a high-risk group and should undergo long-term radiographic surveillance.
Intraductal papillary mucinous neoplasms (IPMN) of the pancreas are defined as grossly visible, mucin-producing neoplasms that arise in the main pancreatic duct, or its branches, and lack the ovarian stroma characteristically seen in mucinous cystic neoplasms1 These tumors are the most common identifiable precursor of pancreatic ductal adenocarcinoma (PDAC) and may exhibit a wide spectrum of dysplasia that ranges from low-grade dysplasia, to high-grade dysplasia, to invasive carcinoma.2–5
Surgical management of IPMN has been a topic of controversy and management has evolved over the past two decades. According to published consensus guidelines in 2012, surgical resection is generally indicated for all main duct IPMN as well as any branch duct IPMN with high-risk features (enhanced solid component or mural nodule).6 Despite the general agreement that IPMN represent a whole-gland “field defect,” segmental resection remains the most commonly performed procedure.6–9 This risk-reduction strategy may result in resection of the most radiographically concerning lesion(s); however, it does not prevent the development of further IPMN or invasive cancer in the pancreatic remnant.
Previous studies on IPMN “recurrence” have been relatively small in size and have reported a wide range of remnant progression rates (10–50%).10–15 This inconsistency in the rate of recurrence, or progression, is likely due to the lack of a consistent definition for progression across these studies, variable follow-up duration, as well as limitations associated with combining both invasive and noninvasive IPMN in the analyses. Inclusion of patients with invasive IPMN may underestimate remnant recurrence, because the majority of these patients will experience recurrence and death from metastatic disease within 2 years of resection.16 Additionally, data on the risk factors for IPMN remnant progression have been inconsistent, especially with respect to the role of margin status, presence of residual cystic disease in the remnant, and the association between grade of dysplasia and recurrence.17–20 We performed an updated analysis that defined “progression” as a broader representation of remnant recurrence and included a larger number of patients over a longer period to discern the rate and risk factors for IPMN progression.
METHODS
Study Population
After institutional review board approval was obtained, the prospectively maintained pancreatic database at Memorial Sloan Kettering (MSK) was queried for patients who had undergone resection for IPMN between January 1989 and June 2015. We narrowed our analysis to a subset of patients with noninvasive or microinvasive disease (defined as ≤ 10 mm of invasive component) and had at least 6 months of postoperative follow-up imaging with CT or MRI.
Exclusion criteria included: (1) patients with invasive IPMN (>10 mm); (2) concomitant other neoplasm on final pathology (e.g., cholangiocarcinoma, neuroendocrine tumor); (3) patients who had their initial resection at an outside institution; (4) patients who underwent total pancreatectomy as the initial procedure; and (5) patients with a postoperative follow-up period < 6 months or unavailable imaging.
All patients had their initial disease evaluated both radiographically and through histopathological examination. Patient demographics, clinical history, laboratory markers, radiologic features of the initial lesion, surgical procedure, detailed pathologic characteristics, margin status, progression details, and follow-up details were extracted from the electronic medical records. For survival and progression analysis, the last follow-up was defined as the last patient encounter with adequate abdominal imaging or patient death.
Histopathological Assessment
All surgical specimens were entirely submitted and reviewed by gastrointestinal pathologists with extensive experience in pancreatic diseases. A pathologic diagnosis of IPMN was made if the specimen contained grossly identifiable cystic structures at least 1.0 cm in greatest diameter, dilated ducts connected to the remainder of the ductal system, and lined by flat, papillary mucinous, or oncocytic epithelium exhibiting variable degrees of cytoarchitectural atypia, with a component of invasive carcinoma ≤ 10 mm in diameter. The final surgical resection margin was evaluated using a section of ductal epithelium at the transection margin, and margin was considered positive if any degree of dysplasia, PanIN, or carcinoma was identified on final pathologic assessment.
Lesions were classified in accordance with the recently proposed classification system into low-grade or high-grade dysplasia depending on the highest level of dysplasia identified within the specimen.4 Both radiologic and pathological analyses were used to assign an IPMN subtype as a branch duct (BD-IPMN), main duct (MD-IPMN), or mixed-type IPMN depending on main pancreatic duct size and the presence of cystic lesions. All specimens of mucinous cystic tumors resected before the WHO recognition of IPMN in 1996 were reviewed to ensure uniform assessment and identification of all IPMN cases.
Disease Progression
Following resection, patients were typically followed with CT or MRI imaging every 6–12 months. The presence of baseline residual lesions in the remnant pancreas was determined by comparing the preoperative and the first postoperative CT or MRI for patients with a known residual lesion following the initial resection. The maximum cyst diameter was recorded at each scan and compared with the initial diameter to calculate cyst growth.
Disease progression was defined as one of the following: (1) Radiologic evidence of new cystic lesions in the remnant pancreas > 1 cm; (2) Increase in the diameter of a residual cystic lesion in the remnant by more than 50% during postoperative surveillance; (3) Development of histologically proven localized or metastatic pancreatic cancer. MPD dilation was not included in the definition given the difficulty of determining whether the observed dilation was secondary to a recurrent IPMN or rather a manifestation of postoperative changes (e.g., anastomotic stricture). However, some patients in the study cohort underwent repeat resection for suspicious increased postoperative MPD diameter and were indeed found to have recurrent disease. This group of patients (n = 3) were counted as progression, whereas those with postoperative MPD dilation who did not undergo repeat resection were not included in the definition of progression, unless they met one of the other criteria for progression.
Progression-free survival was measured from the date of initial resection to the date in which the IPMN recurrence was diagnosed. This was the date in which the new lesion was identified on imaging, the date in which the old residual lesion met the growth criteria, or for those who underwent repeat resection for carcinoma or biopsy with cytology examination, the date of positive biopsy for malignancy.
Statistical Analysis
Continuous variables were summarized using median and range; categorical variables were summarized using frequency and percent. The cumulative incidence of progression and cancer development were estimated using competing risks methods, with death without recurrence and death without cancer as competing events, respectively. For the Cox regression and Kaplan–Meier curves, deaths were censored. Univariate Cox regression was used to assess the association between variables of interest and progression. Multivariate Cox regression models were created using selected significant variables on univariate analysis (p < 0.05) and variables previously found to be associated with recurrence in other studies. Kaplan–Meier curves stratified by risk group were created to determine differences in cancer development and compared by the log-rank test. All statistical analysis was done in R 3.1.1, and p values < 0.05 were considered significant.
RESULTS
Demographic and Clinical Data
Between January 1989 and June 2015, 454 patients underwent resection for IPMN at our institution. Invasive disease > 10 mm was identified in 89 patients, and these were excluded. The findings of concomitant malignancy, total pancreatectomy as initial procedure, or initial resection at an outside hospital (OSH) also were exclusionary criteria and were present in 46 patients. The remaining 319 patients comprised the population.
Table 1 presents the demographic data of the 319 patients. The median age at operation was 68 years, and 53% of patients were female. Mean serum CEA and CA 19–9 levels were 2.4 ng/ml (range 0–1820) and 19 units/ml (0–55), respectively. Endoscopic-guided ultrasound with cyst fluid aspiration was performed in 38% of patients with a median cyst fluid CEA of 442.7 ng/ml (3–497,830). Symptoms were present in 49% of patients, whereas 51% had their disease incidentally discovered during imaging obtained for other reasons. Most of initially resected lesions were in the head (42%) and uncinate process (21%) of the pancreas, whereas the neck, body, and tail harbored the culprit lesion in 6, 20, and 12%, respectively. The mean size of the index cyst was 2.9 (range 0.7–9.2) cm, the MPD was dilated (> 5 mm) in 42% of patients, and 33% had either a solid component, thickened cyst, or a mural nodule. Operative procedures included pancreaticoduodenectomy (n = 203), distal pancreatectomy with or without splenectomy (n = 89), central pancreatectomy (n = 16), and enucleation (n = 11).
TABLE 1.
Demographic, clinical, radiologic and pathologic data of study population
| Variable | Overall (n = 319) |
|---|---|
| Age, yr (median, range) | 68 (33, 89) |
| BMI | 27 (18, 47) |
| Race | |
| Asian | 17 (5.3) |
| African American | 8 (2.5) |
| Unknown/others | 8 (2.5) |
| White | 286 (89.7) |
| Preoperative symptoms | 155 (48.6) |
| History of diabetes | 68 (21.3) |
| History of pancreatitis | 50 (15.7) |
| Prior malignancy | 80 (25.1) |
| Family history of PDAC | 47 (14.8) |
| History of smoking | 150 (48.1) |
| History of alcohol use | 136 (43.7) |
| Serum CA 19–9 (units/ml) | 19 (0, 1820) |
| Serum CEA (ng/ml) | 2.4 (0, 55) |
| Largest cyst size | 2.9 (0.7, 9.2) |
| Number of cysts | 1 (0, 5) |
| Disease risk | |
| High-risk | 113 (35.8) |
| Low-risk | 203 (64.2) |
| Microinvasion | 20 (6.3) |
| Margin status | |
| Negative | 143 (46.9) |
| LGD | 70 (23) |
| Noninvasive IPMN | 59 (19.3) |
| PanIN | 19 (6.2) |
| In situ carcinoma | 14 (4.6) |
| MPD size | |
| MPD ≤ 0.5 cm | 183 (58.1) |
| 0.5 cm < MPD ≤ 1.0 cm | 97 (30.8) |
| MPD >1.0 cm | 35 (11.1) |
| Location | |
| Head | 131 (41.9) |
| Uncinate | 64 (20.4) |
| Neck | 18 (5.8) |
| Body | 63 (20.1) |
| Tail | 37 (11.8) |
| Solid component | 28 (8.8) |
| Thickened cyst | 42 (13.2) |
| Mural nodule | 36 (11.4) |
| Concurrent lesion | 29 (9.1) |
| Duct type | |
| Branch duct | 166 (52.5) |
| Main duct* | 150 (47.5) |
| Procedure | |
| Central pancreatectomy | 16 (5) |
| Distal pancreatectomy +− splenectomy | 89 (27.9) |
| Enucleation | 11 (3.4) |
| Pancreaticoduodenectomy | 203 (63.6) |
| Residual lesion | 49 (16.2) |
| Epithelial subtype | |
| Gastric | 80 (50) |
| Intestinal | 51 (31.9) |
| Mixed | 15 (9.4) |
| Oncocytic | 4 (2.5) |
| Pancreatobiliary | 10 (6.2) |
Main duct includes mixed duct type
Continuous variables are summarized using median and range
Continuous variables are summarized using frequency and percentage
CEA carcinoembryonic antigen, PDAC pancreatic ductal adenocarcinoma, MPD main pancreatic duct, BMI body mass index, LGD low-grade dysplasia
Pathologic Evaluation
Pathological examination demonstrated high-risk lesions (high-grade dysplasia or microinvasive disease) in 36% of patients (n = 113), and low-risk disease (low and intermediate-grade dysplasia) in 64% (n = 203). Microinvasive disease (≤ 10 mm) was identified in 6% of patients (n = 19) with a median invasive size of 3 mm. BD-IPMN was present in 52% of patients (n = 166), 37% (n = 116) had MD-IPMN, and 11% (n = 34) had mixed disease. The predominant histologic subtype was gastric (50%), followed by intestinal (32%), mixed (more than one dominant morphologic subtype, 9%), pancreatobiliary (6%), and oncocytic (3%). Final surgical margin assessment was negative (free of dysplasia, PanIN, or IPMN) in 47% patients, low-grade dysplasia in 42%, high-grade dysplasia in 5% of patients, and some degree of PanIN in 6%.
IPMN Progression
After a median follow-up of 42 months, 71 patients (22%) had experienced disease progression. Within this group of 71 patients, a new cystic lesion > 1 cm developed in 56 patients (79%), 10 patients (14%) experienced a doubling in the diameter of preexisting cysts, and 11 patients (16% of progression, 3% of all patients) developed invasive cancer in the pancreatic remnant after a median of 28 months from initial resection. Metastatic disease was present in 2 of these 11 patients at the time of diagnosis.
Treatment of Progression
Repeat operative intervention was performed in 11 of the 71 patients (15%) with IPMN progression. Indications for resection were as follows: new lesion suspicious for IPMN in six patients, progression of old lesion in the remnant in two patients, and a dilated MPD in three patients. Final pathology revealed invasive carcinoma in four patients, high-grade dysplasia in four patients, and low-grade IPMN in three patients. In addition to these 11 patients, there were 2 other patients who underwent repeat resection, for presumed IPMN progression based on dilated MPD, but were found to have no IPMN or dysplasia on final pathology, so they were not considered to have IPMN progression.
For the 11 patients who developed invasive disease during the study period, four underwent salvage pancreatectomy and three of them were disease-free at the time of last follow-up, whereas the fourth died due to disease progression. Within the group of patients who developed invasive disease and did not undergo a salvage procedure due to various reasons (n = 7), two had metastatic disease at the time of diagnosis and died from their disease, two are undergoing salvage chemotherapy for cancer in the remnant, two died due to cancer progression in the remnant, and the last one died due to an unrelated disease.
Patient Outcomes
Two- and five-year estimates for cumulative incidence of IPMN progression were 10% and 26%, respectively. Cumulative incidence curves based on competing risks analyses are shown in Fig. 1.
FIG. 1.
Cumulative incidence of IPMN progression (recurrence) based on competing risks analysis
A total of 41 patients died after a median of 43 months, 36 (88%) of which were due to other disease or an unknown cause. The other five were due to the development of pancreatic cancer. One-year overall survival and 5-year overall survival rates were 96.7 and 86%, respectively.
Factors Associated with Progression
Univariate Cox models for progression (with death censored) identified increased age at operation, increased body mass index (BMI), increasing number of identified cysts, a family history of pancreatic cancer, and location of initial lesion in the tail of the pancreas with a subsequent distal pancreatectomy to be significantly associated with an increased risk of progression (Table 2). On multivariate analysis, the only variable significantly associated with progression was the location of the initial lesion, as proximal lesions in the head, neck, or uncinate (those who initially underwent a pancreaticoduodenectomy) were less likely to recur [hazard ratio (HR) = 0.41, confidence interval (CI) (0.25–0.68), p < 0.001]. None of the other tested variables, including degree of dysplasia on initial pathology, residual lesion in the remnant pancreas, or margin status, were significantly associated with increased risk of progression (Table 3).
TABLE 2.
Clinicopathological factors associated with progression by univariate analysis
| Hazard ratio (95% confidence interval) | p value | |
|---|---|---|
| Age at operationa | 1.34 (1.04, 1.74) | 0.025 |
| BMIb | 1.36 (1.07, 1.73) | 0.013 |
| Serum CA19–9 (units/ml) | 1.00 (0.99, 1.01) | 0.431 |
| Serum CEA (ng/ml) | 0.90 (0.78, 1.05) | 0.189 |
| Size of largest cyst size (cm) | 0.95 (0.79, 1.15) | 0.616 |
| Number of cysts | 1.24 (1.03, 1.49) | 0.023 |
| Size of microinvasion (mm) | 1.21 (0.75, 1.94) | 0.440 |
| Disease risk | 0.319 | |
| High-risk | Referent | |
| Low-risk | 1.29 (0.78, 2.14) | |
| Presence of microinvasion | 1.21 (0.44, 3.34) | 0.708 |
| Margin status | 0.756 | |
| Negative | Referent | |
| LGD | 1.17 (0.65, 2.1) | |
| Noninvasive IPMN | 0.88 (0.45, 1.69) | |
| PanIN | 1.09 (0.42, 2.78) | |
| In situ carcinoma | 0.33 (0.04, 2.41) | |
| Race | 0.208 | |
| Asian | Referent | |
| African American | 0.23 (0.03, 1.84) | |
| Unknown/others | 0.22 (0.03, 1.81) | |
| White | 0.49 (0.23, 1.08) | |
| Symptoms | 0.74 (0.46, 1.19) | 0.212 |
| Diabetes | 1.23 (0.71, 2.12) | 0.46 |
| Pancreatitis | 0.92 (0.45, 1.84) | 0.804 |
| Prior malignancy | 1.15 (0.68, 1.97) | 0.601 |
| Family history of PDAC | 2.49 (1.46, 4.22) | < 0.001 |
| History of smoking | 0.76 (0.47, 1.21) | 0.249 |
| History of alcohol use | 0.75 (0.46, 1.21) | 0.234 |
| MPD size | 0.528 | |
| 0.5 cm < MPD ≤ 1.0 cm | Referent | |
| MPD ≤ 0.5 cm | 1.19 (0.7, 2) | |
| MPD > 1.0 cm | 0.70 (0.24, 2.04) | |
| Residual lesion | 0.89 (0.48, 1.66) | 0.717 |
| Location | 0.006 | |
| Body | Referent | |
| Head | 0.49 (0.27, 0.88) | |
| Neck | 0.48 (0.14, 1.61) | |
| Tail | 1.32 (0.65, 2.68) | |
| Uncinate | 0.40 (0.19, 0.85) | |
| Location (grouped) | < 0.001 | |
| Body, tail | Referent | |
| Head, neck, uncinate | 0.41 (0.26, 0.66) | |
| Solid component | 0.63 (0.25, 1.56) | 0.314 |
| Thickened cyst | 0.79 (0.39, 1.6) | 0.513 |
| Mural nodule | 1.66 (0.82, 3.35) | 0.161 |
| Concurrent lesion | 1.99 (0.99, 4.03) | 0.054 |
| Duct subtype | 0.443 | |
| Branch duct | Referent | |
| Main duct | 0.83 (0.51, 1.34) | |
| Procedure | < 0.001 | |
| Central Pancreatectomy | Referent | |
| Distal pancreatectomy ± splenectomy | 0.68 (0.3, 1.56) | |
| Enucleation | 0.54 (0.16, 1.87) | |
| Pancreaticoduodenectomy | 0.22 (0.09, 0.50) | |
| Epithelial subtype | 0.628 | |
| Gastric | Referent | |
| Intestinal | 0.52 (0.22, 1.21) | |
| Mixed | 0.88 (0.29, 2.64) | |
| Oncocytic | 0 (0, Inf) | |
| Pancreatobiliary | 0.50 (0.07, 3.73) | |
Hazard ratio associated with a 10-year increase in age
Hazard ratio associated with a 5 unit increase in BMI
CEA carcinoembryonic antigen; PDAC pancreatic ductal adenocarcinoma; MPD main pancreatic duct; BMI body mass index; LGD low-grade dysplasia
TABLE 3.
Multivariable cox model for predictors of progression
| Variable | Hazard ratio (95% confidence interval) | p value |
|---|---|---|
| Location | < 0.001 | |
| Body, tail | Referent | |
| Head, neck, uncinate | 0.41 (0.25, 0.68) | |
| Disease risk | 0.865 | |
| High-risk | Referent | |
| Low-risk | 1.05 (0.62, 1.77) | |
| Margin status* | 0.403 | |
| Negative | Referent | |
| LGD/IPMN/PanIN | 0.86 (0.53, 1.41) | |
| In situ carcinoma | 0.26 (0.03, 2.03) | |
Margin status was insignificant when separate, rather than combined, margin groups were used
LGD low-grade dysplasia
Univariate analysis for development of invasive cancer identified family history of pancreatic cancer as the only factor associated with development of PDAC [HR = 4.9, CI (1.31–18.27), p = 0.018]. Degree of dysplasia approached significance with p value of 0.067. The event occurrence was too small to allow for multivariate analysis (Fig. 2).
FIG. 2.
Kaplan-Meier curves for progression into cancer, stratified by disease risk
DISCUSSION
In this study, we describe our institutional experience with patients who underwent resection for noninvasive and microinvasive IPMN between 1989 and 2015 (n = 319). We found an overall disease progression rate of 22% (n = 71), with a cumulative incidence of progression of 10% at 2 years and 26% at 5 years. Distal location of the original IPMN was the only factor associated with progression (HR 2.43, CI 1.47–4.0, p < 0.001). Cancer development (invasive IPMN and PDAC) occurred in 16% of these cases (n = 11, 3% of total). A family history of pancreatic cancer was associated with the development of carcinoma in the remnant pancreas after the initial resection of benign or microinvasive IPMN.
The literature regarding “recurrence” rates after resection of noninvasive IPMN is conflicting. In a recent study from Massachusetts General Hospital (MGH) on resected invasive and noninvasive IPMN, the overall risk of recurrence in the remnant gland was 17%. Columbia University, Mayo Clinic, and Seoul University have reported lower rates of IPMN recurrence, with rates of 5.3, 10, and 9.7%, respectively.11,19,21 Similar to our findings, the Johns Hopkins group and the Indiana University group reported recurrence rates of 17% and 20%, respectively.15,20
The slightly higher rate of recurrence in this current study, compared with data from other institutions, may be explained by the focus on disease progression rather than recurrence. Our definition of progression incorporated radiologic progression as well as development of malignancy. This definition was chosen, although arbitrarily, because these three features of progression (new lesions > 1 cm, doubling size of preexisting lesions, and new PDAC) are not only radiographically identifiable but also clinically relevant. Because IPMN is considered a whole gland process, the development of a “new” lesion, or the growth of a previous lesion, reflect progression of disease that was present at the time of initial operation, rather than development of de novo disease.
In contrast, the MGH study did not consider the growth of IPMN lesions that were left in the remnant at the time of operation to represent recurrence.17 Similarly, in the aforementioned study from Japan only lesions confirmed by resection or fine needle aspiration were considered as “recurrence.” More importantly, benign lesions were not regarded as intrapancreatic recurrence and were excluded from their definition. Conversely, the study from Columbia had a somewhat similar definition to ours, which included histologic confirmation of pancreatic neoplasia, radiographic appearance of a new cyst or development of MPD dilatation clinically confirmed as IPMN. However, they did not include the growth of an old remnant lesion in the criteria for recurrence.
We did not find an association between margin status and disease progression, and this is consistent with multiple studies from the United States and Japan.19,20,22,23 Interestingly, in our previous experience at MSK, dysplasia of any degree at the resection margin was an independent predictor of recurrent disease in the remnant gland but not at the resection margin itself (odds ratio 2.9, p = 0.02).10,18 The discrepancy between the current series and our previous report may be due to the broader definition of progression, the inclusion of microinvasive IPMN, and perhaps the larger number of patients included in this study (n = 319 vs. 192).
In the present series, distal lesions were associated with an increased risk of progression, which is consistent with our previous data and data from Japan but is contrary to the data from South Korea.12,18,22 Data regarding the association between a residual lesion in the remnant and risk of recurrence is not any less confusing. For instance, the Indiana group found that patients with residual IPMN after segmental pancreatectomy are not at an increased risk for the development of invasive disease or reduced survival.15 However, in the study from Japan as well as the MGH series, IPMN in the remnant pancreas was associated with an increased risk of recurrence on multivariate analysis. These conflicting findings can be probably explained by the inconsistent definition of residual IPMN across different reports.17,22
Knowing the rate, patterns, and predictors of progression is an invaluable asset for clinicians when deciding on the extent of follow-up and help clear some of the confusion in surveillance guidelines for resected IPMN patients. For example, the most recent consensus guidelines categorize the frequency of follow-up based on margin status. However, we have found that margin status is not associated with increased risk of progression, that not only patients with MD-IPMN but also those with BD-IPMN are at increased risk of developing invasive IPMN or PDAC, and that 17% of progression in our series occurred > 5 years following the initial resection. These findings should be considered when planning future surveillance strategies.
Limitations of this study include its retrospective nature, which makes selection bias highly likely. Also, our cohort included surgical patients over the past two decades, which inevitably come with heterogeneous indications for resection, criteria for histopathologic diagnosis, management, and a variable follow-up approach. However, all pathology slides were reviewed, and the pathologic diagnosis per the most recent consensus classification was utilized. Moreover, most patients who were categorized as having progression based on radiologic criteria did not have a pathologic confirmation of their presumed progression, due to the lack of surgical indication for resection, which may overestimate the true incidence of progression. Finally, the small number of events limited the number of variables in the multivariate model and the statistical power to detect factors associated with progression.
CONCLUSIONS
In our entire population of resected noninvasive and microinvasive IPMN, the cumulative incidence of recurrence at 2 years was 10 and 26% by 5 years; 16% of these were invasive IPMN or PDAC. On multivariate analysis, the location of the original IPMN lesion in the body or tail was the only factor significantly associated with increased risk of progression. Thus, these patients represent a high-risk group for cancer development, and we currently recommend life-long surveillance for these patients.
Acknowledgments
FUNDING Supported in part by Marshall and Therese Sonenshine Foundation
Footnotes
DISCLOSURE Authors have nothing to disclose.
REFERENCES
- 1.Shi C, Hruban RH. Intraductal papillary mucinous neoplasm. Human Pathol. 2012;43(1):1–16. [DOI] [PubMed] [Google Scholar]
- 2.Lafemina J, Katabi N, Klimstra D, Correa-Gallego C, Gaujoux S, Kingham TP, et al. Malignant progression in IPMN: a cohort analysis of patients initially selected for resection or observation. Ann Surg Oncol. 2013;20(2):440–7. [DOI] [PubMed] [Google Scholar]
- 3.Dudeja V, Allen PJ. Premalignant cystic neoplasms of the pancreas. Sem Oncol. 2015;42(1):70–85. [DOI] [PubMed] [Google Scholar]
- 4.Basturk O, Hong SM, Wood LD, Adsay NV, Albores-Saavedra J, Biankin AV, et al. A revised classification system and recommendations from the Baltimore consensus meeting for neoplastic precursor lesions in the pancreas. Am J Surg Pathol.2015;39(12):1730–41. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Adsay V, Mino-Kenudson M, Furukawa T, Basturk O, Zamboni G, Marchegiani G, et al. Pathologic evaluation and reporting of intraductal papillary mucinous neoplasms of the pancreas and other tumoral intraepithelial neoplasms of pancreatobiliary tract: recommendations of Verona consensus meeting. Ann Surg. 2016;263(1):162–77. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Tanaka M, Fernandez-del Castillo C, Adsay V, Chari S, Falconi M, Jang JY, et al. International consensus guidelines 2012 for the management of IPMN and MCN of the pancreas. Pancreatology. 2012;12(3):183–97. [DOI] [PubMed] [Google Scholar]
- 7.Fong ZV, Ferrone CR, Lillemoe KD, Fernandez-Del Castillo C. Intraductal papillary mucinous neoplasm of the pancreas: current state of the art and ongoing controversies. Ann Surg. 2016;263(5):908–17. [DOI] [PubMed] [Google Scholar]
- 8.Hruban RH, Takaori K, Canto M, Fishman EK, Campbell K, Brune K, et al. Clinical importance of precursor lesions in the pancreas. J Hepato-biliary-pancreat Surg. 2007;14(3):255–63. [DOI] [PubMed] [Google Scholar]
- 9.Al Efishat M, Allen PJ. Therapeutic approach to cystic neoplasms of the pancreas. Surg Oncol Clin N Am. 2016;25(2):351–61. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.White R, D’Angelica M, Katabi N, Tang L, Klimstra D, Fong Y, et al. Fate of the remnant pancreas after resection of noninvasive intraductal papillary mucinous neoplasm. J Am Coll Surg. 2007;204(5):987–93. [DOI] [PubMed] [Google Scholar]
- 11.Kang MJ, Jang JY, Lee KB, Chang YR, Kwon W, Kim SW. Long-term prospective cohort study of patients undergoing pancreatectomy for intraductal papillary mucinous neoplasm of the pancreas: implications for postoperative surveillance. Ann Surg. 2014;260(2):356–63. [DOI] [PubMed] [Google Scholar]
- 12.Park J, Lee KT, Jang TH, Seo YW, Lee KH, Lee JK, et al. Risk factors associated with the postoperative recurrence of intraductal papillary mucinous neoplasms of the pancreas. Pancreas. 2011;40(1):46–51. [DOI] [PubMed] [Google Scholar]
- 13.Niedergethmann M, Grutzmann R, Hildenbrand R, Dittert D, Aramin N, Franz M, et al. Outcome of invasive and noninvasive intraductal papillary-mucinous neoplasms of the pancreas (IPMN): a 10-year experience. World J Surg. 2008;32(10):2253–60. [DOI] [PubMed] [Google Scholar]
- 14.Nagai K, Doi R, Kida A, Kami K, Kawaguchi Y, Ito T, et al. Intraductal papillary mucinous neoplasms of the pancreas: clinicopathologic characteristics and long-term follow-up after resection. World J Surg. 2008;32(2):271–8. [DOI] [PubMed] [Google Scholar]
- 15.Miller JR, Meyer JE, Waters JA, Al-Haddad M, Dewitt J, Sherman S, et al. Outcome of the pancreatic remnant following segmental pancreatectomy for non-invasive intraductal papillary mucinous neoplasm. HPB. 2011;13(11):759–66. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Raut CP, Cleary KR, Staerkel GA, Abbruzzese JL, Wolff RA, Lee JH, et al. Intraductal papillary mucinous neoplasms of the pancreas: effect of invasion and pancreatic margin status on recurrence and survival. Ann Surg Oncol. 2006;13(4):582–94. [DOI] [PubMed] [Google Scholar]
- 17.Marchegiani G, Mino-Kenudson M, Ferrone CR, Morales-Oyarvide V, Warshaw AL, Lillemoe KD, et al. Patterns of recurrence after resection of IPMN: who, when, and how? Ann Surg. 2015;262(6):1108–14. [DOI] [PubMed] [Google Scholar]
- 18.Frankel TL, LaFemina J, Bamboat ZM, D’Angelica MI, DeMatteo RP, Fong Y, et al. Dysplasia at the surgical margin is associated with recurrence after resection of non-invasive intraductal papillary mucinous neoplasms. HPB. 2013;15(10):814–21. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19.Winner M, Epelboym I, Remotti H, Lee JL, Schrope BA, Chabot JA, et al. Predictors of recurrence in intraductal papillary mucinous neoplasm: experience with 183 pancreatic resections. J Gastrointest Surg. 2013;17(9):1618–26. [DOI] [PubMed] [Google Scholar]
- 20.He J, Cameron JL, Ahuja N, Makary MA, Hirose K, Choti MA, et al. Is it necessary to follow patients after resection of a benign pancreatic intraductal papillary mucinous neoplasm? J Am Coll Surg. 2013;216(4):657–65. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Schnelldorfer T, Sarr MG, Nagorney DM, Zhang L, Smyrk TC, Qin R, et al. Experience with 208 resections for intraductal papillary mucinous neoplasm of the pancreas. Arch Surg.2008;143(7):639–46. [DOI] [PubMed] [Google Scholar]
- 22.Yogi T, Hijioka S, Imaoka H, Mizuno N, Hara K, Tajika M, et al. Risk factors for postoperative recurrence of intraductal papillary mucinous neoplasms of the pancreas based on a long-term follow-up study: proposals for follow-up strategies. J Hepatobiliary-pancreatic Sci. 2015;22(10):757–65. [DOI] [PubMed] [Google Scholar]
- 23.Bhardwaj N, Dennison AR, Maddern GJ, Garcea G. Management implications of resection margin histology in patients undergoing resection for IPMN: a meta-analysis. Pancreatology.2016;16(3):309–17. [DOI] [PubMed] [Google Scholar]