Kempeneers Rens

Chronic pancreatitis is still for a large part shrouded in mists. Although chronic pancreatitis was first described around 1800 and since then over 20.000 research articles are published on this topic, there are still many knowledge gaps and misunderstandings regarding diagnostics, course of disease and treatment of chronic pancreatitis. Comprehensive understanding of chronic pancreatitis is needed to understand and successfully treat a patient with chronic pancreatitis. As such, this thesis aims to fulfill several knowledge gaps in chronic pancreatitis. The first part of the thesis explores the diagnostics and natural course of chronic pancreatitis, whereas the second part focusses on the surgical treatment of chronic pancreatitis.

Part I. Diagnosis and Clinical Course of Chronic Pancreatitis

Pancreatic imaging plays a key role in the diagnosis of chronic pancreatitis. There is, however, lack of international consensus regarding choice of imaging for the diagnosis of chronic pancreatitis. In Chapter I, we performed a systematic review and meta-analysis to determine the diagnostic accuracy of imaging modalities for the initial diagnostic assessment of chronic pancreatitis. In total, 43 studies evaluating 3460 patients were included on the five most frequently used imaging modalities: endoscopic ultrasonography (EUS), endoscopic retrograde cholangiopancreatography (ERCP), magnetic resonance imaging (MRI), computed tomography (CT) and ultrasonography (US). For the initial diagnosis of chronic pancreatitis, EUS, ERCP, MRI and CT all had comparable high diagnostic accuracy. EUS and ERCP were outperformers and US had the lowest accuracy. The choice of imaging modality can therefore be made based on invasiveness, local availability, experience and costs.

Investigating the natural course of chronic pancreatitis started in this thesis with Chapter II, where it evaluated the natural course and treatment of pancreatic exocrine insufficiency (PEI), which is a common complication of chronic pancreatitis. This study is based on the nationwide chronic pancreatitis registry (CARE) and included 987 patients with chronic pancreatitis from 29 Dutch hospitals. Patients were classified in three groups: definite PEI (304 patients, 31%), potential PEI (451 patients, 46%), and no PEI (232 patients, 24%). Patients with definite PEI had significantly more malabsorption symptoms, a lower body mass index, and aberrant defecation. Of the PEI patients using pancreatic enzyme replacement therapy, 47% still reported steatorrhea. We concluded therefore that some form of pancreatic enzyme replacement therapy is reasonably effective in alleviating malabsorption symptoms, but that improvement of treatment is needed.

Chapter III is also based on CARE and evaluates diabetes mellitus, which frequently complicates chronic pancreatitis and is then categorized as diabetes mellitus type 3c. Since there are no pathophysiological criteria to distinguish diabetes mellitus type 3c from other types of diabetes mellitus, we developed and applied an epidemiological classification algorithm to 1130 patients with chronic pancreatitis in CARE. Of them, 368 (33%) had diabetes mellitus and 762 (67%) had no diabetes mellitus. Among the patients with diabetes mellitus, 11 were classified as having diabetes type 1 (3%), 159 as diabetes type 2 (43%) and 191 as diabetes type 3c (52%). During longitudinal follow-up of median 47 months, 120 (20%) patients developed diabetes mellitus after study inclusion of which 99 (16%) patients were classified as diabetes type 3c. Development of diabetes mellitus type 3c was associated with longer duration of chronic pancreatitis and more severe chronic pancreatitis sequelae.

Pain is the most frequent and dominant symptom in chronic pancreatitis. In clinical practice and international guidelines, the presentation of pain is often subdivided in specific patterns, each proposed as being potentially the result of different pathophysiological entities of chronic pancreatitis. In Chapter IV, we studied pain patterns in a longitudinal manner in patients with chronic pancreatitis from CARE. Patients with continuous pain had more severe pain, used more opioids and neuropathic pain medication, and had a lower quality of life, as compared with patients with intermittent pain. During a median follow-up of 47 months, 61% of patients alternated at least once between pain patterns. These alternations were not associated with endoscopic or surgical interventions, but only with the VAS pain intensity score and opioid use. These results suggest that continuous and intermittent pain patterns in chronic pancreatitis are not the result of distinctly different pathophysiological or clinical entities. The subjectively reported character of pain is not related to imaging findings or duration of disease. Pain patterns often change over time and are a feature that plays a role in how the severity of pain is experienced in patients with chronic pancreatitis.

Part II. Endoscopic and Surgical Treatment of Chronic Pancreatitis

In Chapter V, a pro-con debate is presented on the current interventional management of patients with chronic pancreatitis, in which we authored the opinion in favor (pro) of the current interventional treatment recommendations. We described what currently is known and what the knowledge gaps are about the interventional therapy in chronic pancreatitis. Based on the available evidence, both surgery and endoscopy are durable (and not time dependent as suggested by the con group), although surgery provides the superior outcome. After the onset of pain, a (long) delay before invasive treatment increases the risk of opioid dependence and may lead to sensitization of central pain pathways and changes in pain perception, which increases the chance of treatment failure. On this basis, we (and the con group actually) favor earlier intervention in painful, obstructive chronic pancreatitis. To resolve this longstanding debate about the timing of intervention, the Dutch Pancreatitis Study Group conducted the ESCAPE trial, which results are presented in another chapter of this thesis. Although all endoscopic and surgical intervention trials in chronic pancreatitis have shown pain relief in a clinically relevant proportion of patients, none of them used sham interventions. Long-term sham-controlled studies, however, are unethical and have several practical problems, and are therefore difficult to perform. Overall, this debate showed that there is an urgent need for well-designed, prospective randomized trials in this field of research.

Chapter VI contains a part of the first truly international consensus guidelines on chronic pancreatitis, which focused on surgery and timing of intervention. It is meant to guide clinical practitioners and surgeons in the treatment of patients with chronic pancreatitis. An international working group with 15 experts on chronic pancreatitis surgery from the major international pancreas societies (IAP, APA, JPS, and EPC) evaluated 20 statements generated from evidence on 5 questions deemed to be the most clinically relevant in chronic pancreatitis. Strong consensus was obtained for the following statements: Surgery in chronic pancreatitis is indicated as treatment of intractable pain and local complications of adjacent organs, and in case of suspicion of malignant (cystic) lesion; Early surgery is favored over surgery in a more advanced stage of disease to achieve optimal long-term pain relief; In patients with an enlarged pancreatic head, a combined drainage and resection procedure, such as the Frey, Beger, and Berne procedure, may be the treatment of choice; Pancreaticoduodenectomy is the most suitable surgical option for patients with groove pancreatitis; The risk of pancreatic carcinoma in patients with chronic pancreatitis is too low (2% in 10 year) to recommend active screening or prophylactic surgery; Patients with hereditary chronic pancreatitis have such a high risk of pancreatic cancer that prophylactic resection can be considered (lifetime risk of 40-55%).

There is debate whether to perform partial resection of the pancreatic head in patients with chronic pancreatitis and pancreatic duct dilatation. Therefore, in Chapter VII, we report long-term outcomes of a surgical strategy comprising selective use of either Frey procedure (with partial pancreatic head resection) or extended lateral pancreaticojejunectomy (without partial pancreatic head resection) in patients with symptomatic chronic pancreatitis and a dilated main pancreatic duct as advised by recent guidelines. In total, 140 patients were included that underwent an extended lateral pancreaticojejunostomy (70) or a Frey procedure (70) for confirmed chronic pancreatitis. Results of this study showed that in patients with a pancreatic head smaller than 4 cm an extended lateral pancreaticojejunostomy is indicated, whereas a Frey procedure is indicated in case of pancreatic head enlargement or when large stones are located in the pancreatic head. This surgical strategy of selective use of either an extended lateral pancreaticojejunostomy or Frey procedure had low morbidity, excellent long term pain relief and a low-rate of new onset diabetes.

One of the most prominent longstanding debates in chronic pancreatitis is about the timing of surgical intervention in chronic pancreatitis. In Chapter VIII, a randomized clinical trial is performed to determine whether early surgery is more effective than the endoscopy-first approach in terms of clinical outcomes. This ESCAPE trial was an unblinded, multicenter, randomized clinical superiority trial involving 30 Dutch hospitals. From April 2011 until September 2016, a total of 88 patients with chronic pancreatitis, a dilated main pancreatic duct, and who only recently started using prescribed opioids for severe pain were included. There were 44 patients randomized to the early surgery group who underwent pancreatic drainage surgery within 6 weeks after randomization and 44 patients randomized to the endoscopy-first approach group who underwent medical treatment, endoscopy including lithotripsy if needed, and surgery if needed. During 18 months of follow-up, patients in the early surgery group had a lower Izbicki pain score than patients in the endoscopy-first approach group, but pancreatic function and quality of life were not significantly different. It was concluded that, among patients with chronic pancreatitis, early surgery compared with an endoscopy-first approach resulted in lower pain scores when integrated over 18 months. However, further research is needed to assess persistence of differences over time and to replicate the study findings.

The economic evaluation of the ESCAPE trial is presented in Chapter IX. In all 88 included patients, use of healthcare was prospectively recorded during the 18 months of follow-up. Unit costs of resources were determined, and cost-effectiveness and cost–utility analyses were performed from societal and healthcare perspectives. Early surgery had a probability percentage of 88.4% of being more cost-effective than the endoscopy-first approach at a willingness-to-pay threshold of €0 per day per unit decrease on the Izbicki pain score. The probability percentage per additional QALY was 75.7% at a willingness-to-pay threshold of €50,000. It was therefore concluded that early surgery was more cost-effective than the endoscopy-first approach in terms of Izbicki pain score and quality-adjusted life-years.

Total pancreatectomy with islet autotransplantation (TP-IAT) can be the last resort option for chronic pancreatitis patients with refractory pain despite previous surgical treatment. In Chapter X, the impact of TP-IAT on pain and quality of life was systematically examined. In total, 15 observational studies evaluating 1,255 patients were included. Meta-analyses showed that one year after TP-IAT, the opioid-free rate had improved from 0-15% to 63% and the insulin-free rate had decreased from 89.5-100% to 30%. Quality of life improved statistically after TP-IAT. Our conclusion was that, in selected patients with painful treatment refractory chronic pancreatitis, evidence shows that TP-IAT is effective for pain control in almost two-thirds of patients, whereas the insulin-free rate is relatively low.

General Discussion and Future Perspectives

Early diagnosis of CP

Chronic pancreatitis is a fibro-inflammatory disease which is already irreversible when diagnosed. This thesis starts with a study on the diagnostic accuracy of different imaging modalities for the diagnosis of chronic pancreatitis. These imaging modalities search for morphological abnormalities that are present due to the chronic inflammation, like calcifications and lesions of the pancreatic duct. Abnormalities on imaging are one of the main criteria of the often used M-ANNHEIM criteria to diagnose chronic pancreatitis.1 However, these criteria are features of an already end-stage chronic pancreatitis, when the inflammation caused severe irreversible damage to the pancreas.

Diagnosing an earlier stage of chronic pancreatitis or even a precursor stage of chronic pancreatitis could make this disease potentially resolvable. It is known that a large part of the patients with chronic pancreatitis starts with acute pancreatitis attacks, years before they eventually develop chronic symptoms.2 Even after developing (unclear) abdominal pain, it takes often months to years before chronic pancreatitis is diagnosed. Identifying and treating patients in this early development stage is very important to be able to change the natural disease course and prevent for complications. Research on early characteristics and biomarkers of chronic pancreatitis is needed to be able to diagnose chronic pancreatitis in an earlier, more resolvable stage.3-5

Exocrine and endocrine insufficiency in chronic pancreatitis

During the course of chronic pancreatitis, most of the patients develop pancreatic exocrine insufficiency, for which they are treated with pancreatic enzyme replacement therapy. In chapter II of this thesis, it is showed that 47% of the patients still have malabsorption complaints despite therapy, and therefore, more research is needed on how to improve the efficacy of pancreatic enzyme replacement therapy. Furthermore, results of this chapter suggest that just increasing the dosage of the therapy without appropriate dietary counseling and use of necessary adjuvant therapy does not lead to fewer malabsorption-related symptoms. It seems that patients with pancreatic exocrine insufficiency are not treated according to the guidelines.3, 6 In clinical practice, efficacy of pancreatic enzyme replacement therapy should be frequently evaluated and tailored to complaints, the diet, and residual pancreatic function. Thereby, a patient has to be educated about diet and enzyme administration to help to reduce malabsorption complaints.

After median 15 years of chronic pancreatitis, more than half of the patients have developed diabetes mellitus and this is mostly diabetes type 3c. Regarding the pathophysiological genesis of diabetes, type 3c seem more similar to type 1 than to type 2 since there is an irreversible destruction of the pancreatic beta cells in both diabetes type 1 and 3c. In chapter III, we found that patients with diabetes type 3c were often treated as diabetes type 2, starting with oral antidiabetic medication with metformin. Therefore, future research should focus on the specific treatment of diabetes mellitus type 3c in chronic pancreatitis.

Pain in chronic pancreatitis

In total, 80-90% of patients with chronic pancreatitis have pain.7 It is thought that this pain is often caused by morphological abnormalities, such as inflammatory masses and obstruction of the pancreatic duct. As a result, these abnormalities are treated by endoscopy and surgery to relieve the pain, with moderate to high success rates. The contrast is, however, that the several studies who investigated the correlation between pain symptoms and morphological abnormalities (including chapter IV in this thesis) failed to find a clear correlation.8, 9 Potentially there is limited or even no correlation between pain and morphology in chronic pancreatitis, but good quality, longitudinal studies focused on morphology are needed.

It is known that there is also a neuropathic basis for pain in patients with chronic pancreatitis, for which invasive treatments may be no effective option. This neuropathic pain develops because of changes in the nervous system due to the consistent pain stimuli and inflammation in chronic pancreatitis. This constant input can eventually result in an autonomous, self-perpetuating pain, which is independent of the stimulus. While there are good randomized studies on invasive therapy, there is a paucity of high-quality clinical studies investigating neuropathic pain and the treatment with neuropathic pain medication in chronic pancreatitis.

Treatment of pain

Analgesics are the cornerstone of pain management in chronic pancreatitis. Half of the patients with chronic pancreatitis require daily pain medication and the vast majority of them take opioids. There is, however, lack of evidence that they are effective for long-term pain management and thereby, there is a risk of developing opioid dependence and opioid-induced hyperalgesia. However, in current clinical practice, a reasonable part of the patients with chronic pancreatitis still unnecessarily use opioids, often for years. Clinical guidelines should be pursued in which is stated that the lowest effective dose of opioids should be used and that there is an indication to start invasive treatment when there are morphological abnormalities on which the treatment can be targeted. 6, 10 In patients with morphological abnormalities, this first medical step should only be used as a short bridging period to interventional therapy. Interventional therapy currently often starts with endoscopy and surgery is postponed until the pain becomes unmanageable despite multiple endoscopic interventions. In chapter VIII, we showed that early surgery outreached the current endoscopy-first approach in terms of pain control in patients with an obstructed pancreatic duct. Based on this important study in chronic pancreatitis, we should strive from now on to early surgical treatment instead of starting with endoscopy and postponing the surgery to the end stage of treatment.

The question now is if there remains any role for endoscopic treatment in a patient with an obstructed pancreatic duct. In the ESCAPE trial in chapter VIII, post hoc analysis showed that in patients who received endoscopy, short-term complete duct clearance was associated with a lower Izbicki pain score that was almost as low as in the early surgery group. This might leave the option open for endoscopy to be tried first during a predefined short period of time to reach a successful treatment in a subgroup of patients, for example in patients with a single stone located in the pancreatic head without any other obstructions or strictures. First, further research is needed to identify this subgroup and, if endoscopy is nonetheless tried first, complete duct clearance and pain reduction should be obtained and confirmed at short-term follow-up.

Pain measurement methods

In the literature on chronic pancreatitis, many methods are used for pain measurement. In chapter VII, we showed the subjectivity of pain measurement in chronic pancreatitis by measuring pain relief in two different ways and this led to very different results. It is almost impossible to generalize and compare studies on pain treatment in chronic pancreatitis when different pain measurement methods are used. We should therefore strive to use one specific pain measurement method for all studies on pain treatment in chronic pancreatitis. Pain measurement should not only involve the intensity of pain, but also the frequency of pain and use of pain medication. This is all involved in the Izbicki pain score, which makes it an well-considered pain measurement method.11 However, over the last years, progression is made in medical pain management. Although Cahen et al. made it applicable to current pain medication, many medication types like neuropathic pain medication are not included in this score. Another methodological limitation of the Izbicki pain score, that was observed in chapter VIII, is that the distribution of the outcome of this score is often not linear, but bi- or multimodal in a group of patients. This is because of local dependence of the items of the Izbicki pain score. For example, a higher intensity of pain leads also to stronger pain medication, which makes these two items dependent of each other. Therefore, a new or updated pain score is needed in which these limitations are taken into account. But most important is that future studies on pain treatment in chronic pancreatitis should use one standardized method of pain measurement.

Importance of (inter)national research collaborations

Many studies in this thesis are the result of (inter)national collaborations of both clinicians and researchers from many disciplines. The nationwide Dutch Pancreatitis Study Group is worldwide noted for high quality, randomized trials in the field of pancreatitis, which were only possible because of the collaboration of many hospitals and specialties. The ESCAPE trial is the first randomized trial of the Dutch Pancreatitis Study Group in the field of chronic pancreatitis and is widely regarded as a very important, high-quality study. The Dutch Chronic Pancreatitis Registry (CARE) is also the result of a nationwide collaboration of hospitals in the Dutch Pancreatitis Study Group. This large registry currently contains data from over 1500 patients from 30 hospitals, and has allowed us to perform several studies which revealed new important insights in the natural course of chronic pancreatitis. (Inter)national research collaborations as the Dutch Pancreatitis Study Group showed their advantage and effectiveness and should therefore be the basis for future research, to fulfil the required sample sizes and to remain the high-quality research standard.

Conclusions

Chronic pancreatitis is a disease with an unpredictable course. It is therefore important to frequently monitor these patients and offer them an individually tailored treatment based on the current clinical guidelines. Early surgery outperforms the endoscopy-first approach and should therefore become part of standard first choice interventional treatment in patients with painful obstructed chronic pancreatitis. Endoscopy-first approach can be attempted in single stone or stricture obstruction in the pancreatic head, but over a fixed short-term period to achieve success. Need for repeated pancreatic duct stent exchanges or unsuccessful endoscopic treatment in terms of pain is reason for surgical treatment. Results of such proactive and targeted treatment strategies need to be evaluated in large clinical trials. Future research should focus on early diagnosis of chronic pancreatitis, the correlation between pain symptoms and morphology, and the development of a standardized pain measurement method for pain treatment studies in chronic pancreatitis.