State-of-the-Art Update of Pancreatic Cysts

Pancreatic cystic lesions (PCLs) represent a heterogenous group of lesions with varying degrees of malignant potential and a wide clinicopathologic spectrum that mandates careful diagnostic and therapeutic consideration. Numerous guidelines have been developed to assist in the management of these lesions (Table 1) [1‐7]. However, the diagnosis and management of PCLs remain problematic due to the variable biological behavior of these lesions, cost of surveillance, morbidity of surgical resection, lack of reliable diagnostics, and need for improved guidelines. Thus, advances in pancreatic cyst diagnostics and therapeutics with endoscopic ablation techniques require further study. Our review aims to shed light on these advances while highlighting the management strategies for PCLs.

With the advent of higher resolution cross-sectional imaging techniques, incidental PCLs have been increasingly discovered. The prevalence in the asymptomatic adult population ranges from 2.4 to 24.3% [6, 8, 9]. Different imaging modalities have reported variable detection rates ranging from 0.2% on ultrasound to over 20% with MRI [10]. De Jong et al. examined 2803 abdominal magnetic resonance imaging (MRI) performed for preventive screening in asymptomatic people in Europe that uncovered 66 (or 2.4%) PCLs, which increased in frequency with age [8]. Individuals younger than 40 years of age, 70 to 79 years of age, and older than 80 years of age had prevalence rates of 0.5%, 11%, and 37%, respectively [8, 9]. Similar findings were seen in an autopsy series by Kimura et al. [11]. While pancreatic cysts are increasingly being discovered, pancreatic cancer-related mortality has not improved. Hence, managing these PCLs has gained considerable attention as clinicians focus on stratifying the malignant potential of these cysts to prevent the mortality associated with progression to pancreatic cancer.

These PCLs may be broadly categorized as mucin producing or nonmucinous lesions. Serous cystadenomas (SCA) are one of the most common nonmucinous cysts that are considered benign neoplasms which occur more commonly in women between the fifth and seventh decades of life along the body–tail region of the pancreas (Fig. 1) [12]. Reassuringly, a large retrospective multinational study of 2622 patients found only three serous cystadenocarcinomas in the entire cohort [13]. Unless a patient is symptomatic or the lesion is large, SCAs should be managed conservatively. If confident the lesion is a SCA, no further imaging surveillance is necessary by most guidelines.

On other end of the spectrum, mucin producing lesions, such as mucinous cystic neoplasms (MCNs) and intraductal papillary mucinous neoplasms (IPMNs), have malignant potential [6]. MCNs are seen mostly in middle aged women as unilocular lesions in the tail of the pancreas without pancreatic duct communication (Fig. 2), and typically surgical resection is recommended [6, 14]. In a study of 90 surgically resected MCNs, high-grade dysplasia (HGD) and pancreatic cancer were seen in 5.5% and 4.4%, respectively [15]. MCNs carry a 10–39% risk of malignancy [16] although less than 0.4% of cysts less than 3 cm in size without a nodule contain HGD or invasive cancer [17, 18]. Therefore, the European Study Group refined surgical resection criteria for MCN to those with size at least 4 cm, nodule, and/or symptoms [3]. Surveillance is not necessary following surgical resection of MCN unless pathology shows malignancy as these patients have a nearly 25% risk of recurrence (Table 2) [1, 6]. The American College of Gastroenterology (ACG) guideline suggests stopping surveillance after 5 years while the other guidelines do not explicitly recommend stopping surveillance.

Intraductal papillary mucinous neoplasms (IPMN) have an equal gender distribution and often occur in the head of the pancreas as solitary or multifocal lesions [16]. IPMNs are radiologically categorized based on their site of involvement within the pancreatic duct: main duct (MD), branch duct (BD), or a combination of both known as mixed (Figs. 3, 4). Malignant potential varies across these different types of IPMNs. MD-IPMNs carry a high risk of malignant transformation of 40 to 90% with 38 to 68% shown to harbor HGD or cancer at the time of resection [6, 10]. Mixed-type IPMNs are thought to have similarly high rates of malignancy (19–68%) although a subtype with only microscopic involvement of the main duct may have favorable rates of malignancy and survival compared with more extensive involvement of the main duct [19]. These high rates of malignancy lead to recommendations for surgical resection of MD-IPMN and mixed-type IPMNs. On the other hand, BD-IPMNs carry a far lower risk (6 to 46%) of malignancy at time of resection with a 22% risk of malignant transformation [1, 12]. Therefore, surgical resection is only recommended in patients with at least one worrisome or high-risk feature. It is important to recognize that not all risk factors carry the same likelihood of malignancy. Five-year survival was 96% for IPMN patients with worrisome features (size > 3 cm, main pancreatic duct 5–9 mm, non-enhancing nodules, cytology with atypical cells, pancreatitis) compared with 60% survival for patients with high-risk features (enhancing nodule, main pancreatic duct ≥ 1 cm, cytology with HGD or cancer, jaundice) [20]. Various guidelines recognize different malignant implications of various features by discussing absolute and relative indications for surgery. The absolute indications for resection by the Fukuoka guideline are main duct dilation ≥ 10 mm, jaundice, enhancing nodule ≥ 5 mm, cytology suspicious or positive for malignancy. The European guideline also includes presence of a solid mass and cytology with HGD or malignancy as absolute indications for surgery [3]. Diabetes may be associated with increased risk of main duct involvement, HGD, and invasive cancer [21].

There remains a lifelong risk from IPMNs [2] as even after resection there is a 0 to 65% risk of recurrence depending on the grade of dysplasia of the resected IPMN [6, 22]. Recurrence includes development of new IPMNs in the remnant pancreas, progression of previously present IPMNs, or occurrence of pancreatic cancer in the remnant pancreas unrelated to an IPMN. Therefore, postoperative surveillance is critical (Table 2).

Despite these advances in diagnostic techniques, managing PCLs is a constantly evolving topic that revolves around stratifying the malignant potential of the cyst. The authors’ suggested algorithm for approaching incidental pancreatic cysts provides a framework for approaching these lesions (Fig. 7). However, before proceeding with this management paradigm, it is important to determine if the patient even requires further evaluation—Are they medically fit for surgery or is the cyst an asymptomatic non-neoplastic (pseudocyst) or benign lesion (SCA) that requires no follow up? Patients with IPMNs and high Charlson Comorbidity Index (≥ 7) had significantly shorter survival (43 months) compared to 180 months for patients with lower scores and had an 11-fold higher risk of dying from non-IPMN-related causes within 3 years [99]. A similar study using the Charlson Comorbidity Index noted that high-risk patients with low-risk cysts were more likely to die from non-pancreatic cancer [100]. Another scoring system (Adult Comorbidity Evaluation 27) was used in a study of 793 patients with IPMNs and reached similar conclusions that patients with higher scores were more likely to die from non-IPMN-related causes [101]. Thus, these scoring systems may assist clinical decision-making when significant comorbidities are present.

The AGA guideline with its controversial recommendation to stop surveillance after 5 years of stability engendered a flurry of long-term studies of pancreatic cysts. A long-term retrospective study found a very low rate of malignant progression (0.9%) for 108 BD-IPMN ≤ 1.5 cm that were followed for more than 5 years [102]. However, another study following BD-IPMNs without worrisome or high-risk features for a minimum of 5 years showed that worrisome or high-risk features developed in 18% of BD-IPMNs well beyond 5 years of observation [103]. Similar findings were observed in a cohort of 1,036 BD-IPMNs without worrisome features in which 4% and 1% of cysts developed worrisome features and pancreatic cancer, respectively, after a median 62 months [104]. In light of these studies, surveillance cannot end in all patients after 5 years, and whether there is a subset of patients in whom surveillance may be safely stopped requires further investigation.

While no study has yet proved reduction in mortality with surveillance, the risk of malignancy cannot be ignored, and it is critical that shared decision-making with the patient is pursued when offering surveillance. Issues to consider include (1) the psychological distress associated with annual surveillance and (2) the significant financial burden to the health-care system. One study estimated that if all patients with cysts (aged 40–79) underwent MRI surveillance, the median cost would be $9.3 billion per year [105]. These calculations did not factor in the cost of surgery or EUS. There is also the concern that patients can be lost to follow up with recent studies finding that 47% to 71.5% of patients received follow-up imaging for incidental cysts [106, 107]. In a large urban safety net hospital, access to gastroenterology care was associated with surveillance completion [106]. More specifically, referral to a multidisciplinary pancreas group is recommended for cysts with any concerning features and when considering surgical resection as evaluation by such a center changed management in 30% of patients away from surgery and toward surveillance [108]. Pancreatic resection should be performed by experienced surgeons at these high-volume centers due to significantly lower post-surgical mortality rates (1–5%) compared with low-volume centers (11–15%) [4, 6].

Patient education about their cysts is important as a survey found that 96% of patients were unaware of their specific cyst type and 58% did not realize some cysts had malignant potential [109]. Physicians need to help patients understand the risks and benefits of pursuing imaging surveillance.

The need for minimally invasive treatment of PCLs has led to the development of EUS-guided cyst ablation, designed to destroy the neoplastic epithelial lining of the cyst in patients who decline or are not operative candidates. Cyst ablation can be achieved by the injection of ethanol, antitumor agents (paclitaxel), or radiofrequency ablation (RFA). Although in its infancy, EUS-guided ablation has been shown to resolve cysts in 33% to 79% of cases while also improving quality of life by avoiding surgery [10]. Moreover, propensity score matching analysis found that compared to PCLs undergoing surveillance initially (n = 428), the EUS alcohol ablation group (n = 118) had significantly lower surgical resection rates (4.8% vs 26.2%) with similar survival over mean follow-up of over 6 years [110].

Patients eligible for cyst ablation by injection of ethanol or antitumor agents include those with a mucinous cyst larger than 3 cm or over 2 cm and growing while malignancy and short life expectancies are contraindications [10, 111‐113]. Ideal cysts for EUS-guided ablation have the following features: unilocular or oligolocular (less than 7 locules) to ensure each locule is injected appropriately [10]; cyst size 2–6 cm; and no open communication with the main pancreatic duct.

Response to ablation is defined by volume reduction in the cyst, which is susceptible to variations in cyst measurement even when using the same modality pre- and post-procedure. Complete response is defined as at least 95% reduction in volume, partial response as 75% to 95%, and non-response as 0% to 74% [113].

Pancreatic cysts remain a diagnostic and management challenge to clinicians and require shared decision-making with patients and occasionally a multidisciplinary approach. While the majority of cysts may be followed with MRI likely without gadolinium, careful appreciation for when to stop surveillance in higher-risk patients, when to refer for EUS in the presence of worrisome or high-risk features or indeterminate cysts, and when to consider surgical resection is essential. The exciting potential of newer diagnostic tools including through the needle biopsy and needle-based confocal laser endomicroscopy as well as novel therapeutics with EUS-guided ablation techniques requires further study.