Pancreatic Enzyme Replacement Therapy in Pancreatic Exocrine Insufficiency—Real-World’s Dosing and Effectiveness: A Systematic Review

Abstract

Background

Pancreatic exocrine insufficiency (PEI) results from impaired exocrine function of the pancreas, causing distressing symptoms like steatorrhea, diarrhea, abdominal distention, bloating, and pain. Treatment involves pancreatic enzyme replacement therapy (PERT), yet compliance with clinical guidelines and its real-world effectiveness are unclear. This systematic review aimed to assess the concordance between real-world PERT doses and clinical guidelines and to evaluate the effectiveness of PERT doses, considering their average administration.

Methods

A systematic search of MEDLINE and EMBASE up to June 2023 identified observational studies reporting PERT doses and their effects on diarrhea and nutrition in PEI patients. Studies were classified based on adherence to current clinical guidelines for PERT dosing.

Results

Twenty-five observational studies involving 3818 patients met the inclusion criteria. In 40% of the studies, average PERT doses were lower than the recommended 40,000–50,000 lipase units (LU) per meal. Significant alleviation of diarrhea was observed in nearly all studies with lower-than-recommended doses, but none showed benefits in nutritional status. PERT doses compliant with guidelines helped reduce diarrhea in most studies and improved or maintained nutritional status.

Conclusions

This review revealed that real-world PERT doses were lower than European guidelines in 40% of studies. While lower doses alleviated gastrointestinal symptoms, they were insufficient for maintaining normal nutritional status. Therefore, PERT dosing should aim for nutritional improvement by adhering to guideline-recommended doses. Individualized dosing, considering both symptom management and nutritional status, is essential due to varying responses to treatment.

Supplementary Information

The online version contains supplementary material available at 10.1007/s10620-025-09011-0.

Introduction

Pancreatic exocrine insufficiency (PEI) is a medical complication of pancreatic diseases, which results in maldigestion and malabsorption of nutrients due to reduced quantity or activity of secreted pancreatic enzymes [1]. Several conditions may be the primary cause of PEI, and the most probable are cystic fibrosis, acute/chronic pancreatitis (AP/CP), pancreatic cancer (PC), pancreatic duct obstruction, and pancreatic resection or surgery [2–4]. Less common etiologies of PEI include intestinal diseases (celiac or inflammatory bowel disease), states after gastrointestinal operations, genetic syndromes, and metabolic disorders (type 1 and 2 diabetes mellitus) [5]. The estimated prevalence of predisposition to PEI varies between conditions, from about 5% in patients with symptoms suggesting irritable bowel syndrome to nearly 100% in patients with inoperable pancreatic cancer [6, 7]. The prevalence of PEI appears to increase with the severity of the primary disease—PEI is observed in 30% of patients with mild changes of CP and 85% with severe changes of CP [3, 5, 8].

Due to nonspecific and varied symptoms, often confused with other gastrointestinal conditions, PEI remains underdiagnosed, especially in patients with mild manifestations or early stages of the disease [5]. In order to diagnose exocrine insufficiency in chronic pancreatitis, it is recommended that fecal elastase-1 (FE-1) activity be measured. The determination is performed on a single stool sample, with normal concentrations of the enzyme exceeding 200 μg/g of stool. An elastase-1 activity of less than 200 μg/g is indicative of mild exocrine pancreatic insufficiency, whereas a value of less than 100 μg/g, especially less than 50 μg/g, is indicative of severe exocrine pancreatic insufficiency [9, 10]. However, this test has limited specificity and sensitivity, particularly in cases of mild to moderate disease [11].

Diarrhea, steatorrhea, and weight loss are the hallmarks of PEI. Other common gastrointestinal symptoms of PEI include abdominal pain, flatulence, and distention [1, 6, 12]. Qualitative assessment of PEI symptoms can be performed using patient-reported outcome instruments [13].

In addition to clinical symptoms, several nutritional deficiencies have been described in PEI patients, including fat-soluble vitamins (A, D, E, K), minerals and trace elements (zinc, selenium, magnesium), and serum proteins (lipoproteins, prealbumin, retinol-binding protein) [14–16]. As a result of persistent malnutrition, PEI may lead to rare but severe complications, including osteoporosis, sarcopenia, impaired night vision, infections due to decreased immunity, coagulation problems, anemia, peripheral neuropathy, and ataxia [1, 6]. These malnutrition-related manifestations have been demonstrated to negatively impact a patient’s quality of life [12, 15]. Furthermore, malnourished PEI patients are believed to suffer more pain episodes and require more hospitalization than those with normal nutritional status [17]. Finally, severe malnutrition, including apolipoprotein A1 and lipoprotein A or C alterations, is a well-known risk factor for cardiovascular events [18, 19]. Therefore, PEI therapy should be focused not only on avoiding clinically relevant gastrointestinal symptoms but also on preventing weight loss and malnutrition.

The only known treatment for PEI is pancreatic enzyme replacement therapy (PERT), which is prandial, oral supplementation of exogenous pancreatic enzymes. The primary goal of PERT is to provide exocrine enzymes at a level considered sufficient to relieve clinical symptoms and improve the nutritional status of patients [2]. Modern formulations of PERT consist of pancreatic enzymes encapsulated in mini-microspheres or micro-pellets with an acid-protective coating designed to prevent enzymes from releasing in the stomach’s acid environment [10]. Until now, several medicinal products have been approved for marketing in PEI conditions (specific brand names differ in region/countries). Formulations differ in the composition of amylase, protease, and lipase amounts [20]. As lipase levels are believed to be the most crucial for normal nutritional status maintenance, the issue of proper PERT administration mainly focuses on lipase intake [2]. In the last decades, there has been no consensus on appropriate PERT dosing—doses recommended by clinical practice guidelines range from 10,000 lipase units (LU) for snacks up to 80,000 LU per main meal [21, 22]. In accordance with the European guidelines on chronic pancreatitis, the recommended enzyme dose is a minimum of 40,000–50,000 LU [23, 24]. Conversely, the Polish Society of Gastroenterology permits the commencement of enzyme therapy with 30,000–40,000 LU, which can be safely augmented up to 50,000 or even 75,000–80,000 LU per meal, contingent on the clinical response [10]. It is unexplored whether PERT administration in actual clinical routines has complied with these recommendations, as only a few studies have been published on this topic [25–27]. Secondarily, whether the recommended doses are clinically effective in resolving gastrointestinal and nutritional problems in PEI patients remains unknown.

To answer these questions, we performed a systematic review of real-world data to assess the compliance of real-world PERT dosing with the guidelines and evaluate its effects on improving gastrointestinal symptoms and nutritional status in PEI patients.

Methods

Search Strategy

The systematic review followed the PRISMA guidelines [28], and the review protocol was registered in PROSPERO (CRD42023434776). A systematic search of major medical databases, including MEDLINE (via Pubmed) and EMBASE, was conducted to identify real-world evidence published until June 2023. Keywords for the MEDLINE and EMBASE search included queries for PEI and related diseases, PERT, and observational studies combined with appropriate Boolean operators. Detailed search strategies are presented in the supplementary appendix. Additionally, references of the included studies were also screened.

Inclusion and Exclusion Criteria

Observational studies published in English were eligible if they reported data on PEI patients treated with PERT in a real-world setting, provided dosage information expressed in LU per meal or day (regardless of the type of pharmacopeia, as 1 FIP/PhEur = 1 USP for lipase [29]), and provided data on patients’ nutritional status, diarrhea, or other clinical outcomes. Studies performed exclusively in children were not included. Similarly, studies focusing on patients with cystic fibrosis were also excluded, as they primarily involve pediatric patients and require a separate analysis due to disease-specific guidelines and significantly different PERT dosing recommendations. Additionally, studies published only as conference proceedings were not included.

Study Selection and Quality Assessment

Two independent analysts screened database records and full-text articles at each stage of the selection process. Two reviewers also independently extracted information about PERT doses and treatment effects from included studies. Any discrepancies between reviewers on the selection and extraction process were resolved by consensus. The quality of the included studies was assessed according to the criteria proposed by NICE for case series or by the Newcastle–Ottawa Scale (NOS) for cohort and cross-sectional studies [30, 31].

Outcomes of Interest and Statistical Analysis

The dosage of PERT expressed in LU was the primary endpoint of this analysis. An average dose of PERT was defined as the mean or the median dose reported by authors or calculated from the number of PERT capsules received. If PERT doses were reported for several subgroups, an average dose for the entire study population was calculated by aggregating data from these subgroups, if possible.

For the assessment of concordance between real-world lipase intake and clinical practice guidelines, the United European Gastroenterology (HaPanEU/UEG) guidelines were used [24]. Although this document was specifically developed for the diagnosis and management of chronic pancreatitis, its recommendations were extrapolated by the authors of this review to all analyzed PEI indications. This approach is justified for two reasons: firstly, no other relevant European clinical guidelines for all PEI conditions had been published in recent years, and secondly, these recommendations were based on several previous world clinical guidelines and clinical trials dedicated to a broad spectrum of PEI disorders. After the completion of the systematic review and manuscript preparation, updated UEG guidelines for PEI were published (December 2024) [32]. While these recent guidelines did not specify recommended dosing range stating that the initial doses of PERT vary mainly depending on the patient’s age [32], the doses we adopted (as outlined below) were confirmed by the guidelines as effective in adult patients, further supporting our approach.

The HaPanEU/UEG guideline recommends a minimum PERT dose of 40,000–50,000 LU for main meals, and these doses may be doubled or even tripled in case of unsatisfactory clinical response [24]. As a result, it has been proposed that recommended PERT dosing ranges between 40,000 and 150,000 LU/meal, corresponding with 120,000–450,000 LU/day, assuming that PERT is supplemented three times daily with main meals. Compliance of average PERT doses with HaPanEU/UEG recommendations was analyzed as a dichotomous variable (yes/no) [24]. A four-grade scale was developed to assess the compatibility between real-world and recommended lipase intake (Table 1).

Table 1.

PERT dose compatibility scale

Compliance	Available measurements in studies
	Mean or median	Range
YES
Complete	The mean/median dose of PERT in a study is within the ranges of recommended doses by clinical guidelines	Ranges of doses used in a study are within ranges recommended by guidelines
Partial	NA	Ranges for real-world doses and recommended doses overlap
NO
Lower	The mean/median dose of PERT in a study is below the minimum recommended by guidelines	Ranges for real-world doses in a study are lower and do not overlap with ranges for recommended doses
Higher	The mean/median dose of PERT in a study is above the maximum recommended by guidelines	Ranges for real-world doses are higher and do not overlap with ranges for recommended doses

NA not applicable, PERT pancreatic enzyme replacement therapy

Diarrhea and nutritional status were analyzed as secondary endpoints. Other outcomes, including body weight, various gastrointestinal symptoms, quality of life, and overall survival, were also extracted, if available. Unless otherwise specified, PERT doses and continuous effect measurements were presented as mean and range. Dichotomous endpoints were analyzed as a percentage of patients with an event. A meta-analysis of clinical outcomes was initially planned; however, due to the limited number of studies for each outcome and their heterogeneity, it was not feasible to conduct. The summary of PERT effectiveness was based on p values reported in studies and the authors’ conclusions.

Results

Study Selection and Characteristics

The search of MEDLINE and EMBASE retrieved 7509 records, and after removing duplicates, 7144 records were qualified for further assessment. Based on titles and abstracts, 7008 papers were discarded because they did not match the predefined inclusion criteria, and 136 full-text articles were assessed for eligibility. Finally, 30 studies were considered relevant for this review [33–62]. The main reasons for exclusion are presented in the PRISMA flow diagram (Fig. 1), while the list of excluded studies is shown in the supplementary appendix.

Fig. 1.

PRISMA flow diagram

A total number of 3818 patients with PEI from 30 studies (30 publications; one publication described two original studies, and one study was described in two papers) were included in this analysis (Table 2). Eight studies were performed in chronic pancreatitis patients, three in pancreatic cancer, and eight in patients after pancreatectomy. The remaining nine studies were conducted in patients with various underlying causes of PEI, including HIV infection, hepatocellular carcinoma, irritable bowel syndrome, celiac disease, and states after esophagectomy, among others. Most of the included studies were prospective (20 studies), single-center (24 studies), and conducted in Europe (19 studies). The remaining studies were performed in the United States (n = 4), Japan (n = 2), India (n = 3), Argentina (n = 1), and South Korea (n = 1). Sample sizes were relatively small, and only nine studies reported data on more than 100 patients. Most studies did not provide the trade name and type of PERT formulation. Detailed characteristics of included studies are described in the supplementary appendix.

Table 2.

Summary of included studies [31–60]

Indication	No. of studies	P/R/CS	S/M	No. of patients	Study size [pts]
Pancreatic exocrine insufficiency (PEI)	11	7/4/0	11/0	693	20–210
Chronic pancreatitis (CP)	8	6/0/2	6/2	1889	29–987
Pancreatectomy (PY)	8	6/1/1	5/3	875	25–355
Pancreatic cancer (PC)	3	1/2/0	2/1	361	91–160
All	30	20/7/3	24/6	3818	20–987

CS cross-sectional, M multi-center, P prospective, R retrospective, S single center

The methodological quality of the included studies was highly heterogeneous. Scores for case series ranged from 4 to 7 points, according to the NICE questionnaire. The most common reasons for lower quality scores included retrospective design, single-center origin, and lack of consecutiveness. The quality of the cohort and cross-sectional studies assessed by NOS was good. None of the studies was excluded from this analysis due to their low quality.

Average Doses of PERT

Average PERT doses varied broadly between individual studies, ranging from 15,000 to 96,000 LU/meal. An average dose of PERT in chronic pancreatitis patients was between 20,000 and 96,000 LU/meal. In patients with pancreatectomy, an average PERT dose was 25,000–66,000 LU/meal; however, in some post-pancreatectomy individuals, lipase intake was escalated even to 216 000 LU/meal [43]. Patients with pancreatic cancer and various PEI conditions received an average of about 25,000–75,000 LU/meal (Table 3).

Table 3.

Average doses of PERT—compliance with HaPanEU/UEG

Study	Indication	N	PERT doses		Compliance with HaPanEU/UEG	References
			Mean or median*	Range or IQR**
LU/meal
Naffouje 2018	PEI	9	NDA	20,000–40,000	YES (partial)	[33]
Yilmaz 2018	PEI	12	50,000	NDA	YES (complete)	[34]
Huddy 2013	PEI	22	25,000	NDA	NO (lower)	[36]
Kiefer 2018	PEI	23	NDA	25,000–50,000	YES (partial)	[37]
Rinzivillo 2020	PEI	7	NDA	40,000–50,000	YES (complete)	[38]
Saif 2020	PEI	19	72,000	72,000–72,000	YES (complete)	[39]
Shandro 2020	PEI	103	50,000*	10,000–100,000	YES (complete)	[40]
Diaz-Gonzalez 2021	PEI	6	25,000	25 000–25,000	NO (lower)	[56]
Desai 2023	PEI	50	15,000	NDA	NO (lower)	[58]
Olmos 2022	PEI	7	50,000	NA	YES (complete)	[59]
Struyvenberg 2017	PY	39	66,000*	9000–216,000	YES (complete)	[43]
Saluja 2019	PY	16	25,000	NDA	NO (lower)	[47]
Kim 2018	PY	47	NDA	40,000–80,000	YES (complete)	[46]
Latenstein 2021	PY	62	NDA	25,000–56,250	YES (partial)	[60]
Dominguez-Munoz 2010	CP	21	20,000*	20,000–60,000	NO (lower)	[49]
Dominguez-Munoz 2007	CP	29	40,000	NDA	YES (complete)	[50]
Min 2018	CP	77	96,000	NDA	YES (complete)	[51]
Dieguez-Castillo 2020	CP	28	50,000	50,000–50,000	YES (complete)	[52]
Arutla 2021	CP	201	24,694a	10,000–40,000	NO (lower)	[57]
Dominguez-Munoz 2007	CP
	1st year	30	20,000*	0–40,000	NO (lower)	[50]
	2nd year	20	40,000*	40,000–60,000	YES (complete)	[50]
Saito 2017	PC	46	48,000	NDA	YES (complete)	[54]
Dominguez-Munoz 2018	PC	49	75,000*	NDA	YES (complete)	[55]
LU/day
Yilmaz 2018	PEI	12	150,000	NDA	YES (complete)	[34]
Evans 2010	PEI	20	45,000	10,000–120,000	NO (lower)	[35]
Diaz-Gonzalez 2021	PEI	6	75,000	75,000–75,000	NO (lower)	[56]
Desai 2023	PEI	50	45,000	NDA	NO (lower)	[58]
Barbier 2013	PY	25	150,000*	75,000–450,000	YES (complete)	[42]
Crippa 2011	PY	65	80,000*	30,000–160,000	NO (lower)	[44]
Suzuki 2016	PY	41	125,145	NDA	YES (complete)	[45]
Latenstein 2021	PY	62	NDA	75,000–181,250	YES (partial)	[60]
Kroon 2022	PY	25	150,000*	NDA	YES (partial)	[61]
		21	112,500*	NDA
D’Haese 2014	CP	294	94,630a	NDA	NO (lower)	[48]
Dominguez-Munoz 2007	CP	29	120,000	NDA	YES (complete)	[50]
Kempeneers 2020	CP	266	100,000*	75,000–175,000**	NO (lower)	[53]
Dominguez-Munoz 2018	PC	49	325,000*	200,000–400,000	YES (complete)	[55]
Trestini 2021	PC	55	80,000*	NDA	NO (lower)	[62]

CP chronic pancreatitis, IQR Interquartile Range, NDA no data available, PEI pancreatic enzyme insufficiency, PY pancreatectomy

^aOwn calculations

*Median, **IQR

Daily Doses of PERT

In most PEI studies, patients received between 80,000 and 150,000 LU/day, which translates to 26,000–50,000 LU/meal, assuming daily doses correspond with lipase supplementation for three main meals per day. A significant deviation from the pattern was observed in three studies, in which the median PERT doses were 45 000 or 325,000 LU/day [23, 35, 58]. In some individuals, daily lipase supplementation did not exceed 10,000 LU, while in others, even 450,000 LU needed to be administered (Table 3) [35, 42].

Compliance of Average Doses of PERT with Clinical Practice Guidelines

Compliance of average PERT doses with HaPanEU/UEG guidelines is presented in Table 3. Complete and partial compliance were confirmed in thirteen and three studies in which lipase intake was expressed in LU/meal. Average PERT doses in seven studies were distinctly lower than recommended. A similar distribution of compliance was observed in studies where PERT doses were expressed in LU/day; however, this conclusion needs to be interpreted with caution due to methodological simplification. Average PERT doses did not exceed those recommended by guidelines in any study.

Clinical Efficacy of Average Doses of PERT

Doses Lower Than Recommended

Diarrhea was evaluated in 6 out of 12 studies with average PERT doses lower than recommended by European guidelines. In five studies, lower PERT doses reduced diarrhea [35, 36, 44, 48, 56]. In the Arutla 2021 study, no differences were observed in the diarrhea rates between chronic pancreatitis patients who were compliant with their prescribed PERT dose, those who were non-compliant, and those who were not receiving PERT [57]. However, as the study was designed as cross-sectional, it is unclear if prescribed PERT doses had no therapeutic effect or if other factors influenced these results.

The benefits of PERT use were also reported on other gastrointestinal symptoms, including lowered levels of FE1 and reduced steatorrhea [35, 36, 44, 47, 48, 53, 58, 62]. Lower doses of PERT could stabilize weight or lead to weight gain in patients from 6 of 10 studies.

Weight results were mixed in the remaining four studies, with about half of the patients experiencing weight gain or stabilization and the other half experiencing weight loss or no significant difference in weight between PERT and no enzyme supplementation [44, 47, 53, 57]. In one study, lower doses of PERT improved patients’ quality of life, which can be explained by reduced abdominal pain [48]. However, no difference was observed in the quality of life in studies comparing lower-than-recommended doses of PERT with no-PERT [53, 57].

Regardless of the positive impact on gastrointestinal symptoms, lower doses of PERT were insufficient to improve patients’ nutritional status (Table 4) [48–50].

Table 4.

Summary of PERT average doses and their efficacy

Study	PERT	Compl.	N	Diarrhea	Nutrition	General	Weight	Steatorr.	Other GI	QoL	Survival	Conclusions	References
D’Haese 2014	94 630 LU/day	Lower	294	✓	X	NR	✓	✓	✓ (pain)	✓	NR	PERT was effective in reducing GI symptoms and weight loss as well as in improving QoL; however, it was not enough in reducing malabsorption	[48]
Evans 2010	45 000 LU/day	Lower	20	✓	NR	✓	✓	NR	✓ (FE1)	NR	NR	PERT was effective in weight stabilization, normalizing FE1, reducing stool frequency as well as in resolving all symptoms in some patients	[35]
Huddy 2013	25 000 LU/meal	Lower	10	✓	NR	NR	✓	✓	NR	NR	NR	PERT was effective in increasing weight and providing symptomatic benefits in diarrhea/steatorrhea	[36]
Dominguez-Munoz 2007	20 000 LU/meal	Lower	30	NR	X	NR	NR	NR	NR	NR	NR	PERT was too low to improve nutritional status measured by a breath test	[50]
Dominguez-Munoz 2010	20 000 LU/meal	Lower	21	NR	X	NR	NR	NR	NR	NR	NR	PERT was not enough to prevent malnutrition in about half of the patients	[49]
Crippa 2011	80 000 LU/day	Lower	65	✓	NR	NR	=	✓	=  (pain)	NR	NR	PERT was effective in weight gain/stabilization in about half of the patients and in reducing steatorrhea and diarrhea in most of the patients; abdominal pain was still present in 22% of the patients	[44]
Saluja 2019	25 000 LU/meal	Lower	16	NR	NR	NR	=	✓	NR	NR	NR	PERT was effective in weight gain/stabilization in about 44% of the patients and reducing steatorrhea in all of them	[47]
Kempeneers 2020	100 000 LU/day	Lower	266	NR	NR	NR	=	NR	✓	=	NR	PERT was effective in reducing GI symptoms compared to lack of PERT, but groups were similar in terms of weight and QoL	[53]
Diaz-Gonzalez 2021	25 000 LU/meal	Lower	6	✓	NR	NR	✓	NR	✓ (FE1)	NR	NR	PERT resolved diarrhea/weight loss in all of the patients and restored FE1 levels in 83% of them	[56]
Arutla 2021	24 694 LU/meal	Lower	201	=	=	NR	=	=	=  (pain)	= a	NR	No differences between PERT and no-PERT groups were observed, as well as between compliant and non-compliant PERT groups	[57]
Desai 2023	15 000 LU/meal	Lower	50	NR	=	NR	✓	NR	✓	NR	NR	PERT was effective in stabilizing weight and improving GI symptoms in some of the patients. Mean values of albumin and hemoglobin also increased; however, they are below population norms	[58]
Trestini 2021	80,000 LU/day	Lower	55	NR	NR	NR	✓	NR	✓	NR	✓	PERT was effective in improving GI symptoms in PEI patients; survival of patients with PERT also improved compared with lack of PERT	[62]
Naffouje 2018	20–40,000 LU/meal	Partial	9	✓	NR	NR	NR	NR	NR	NR	NR	PERT was effective in resolving diarrhea in nearly all of the patients	[33]
Kiefer 2018	25–50,000 LU/meal	Partial	23	NR	NR	NR	✓	✓	NR	NR	NR	PERT was effective in resolving steatorrhea as well as helped to gain weight in some patients	[37]
Latenstein 2021	25–56 250 LU/meal	Partial	62	NR	NR	✓	NR	NR	NR	NR	NR	PERT resolved or decreased PEI complaints in 81% of patients	[60]
Kroon 2022	112,5–150 000 LU/day	Partial	46	NR	NR	✓	NR	NR	NR	NR	NR	PERT was effective in resolving symptoms in 44–70% of patients	[61]
Barbier 2013	150 000 LU/day	Complete	25	=	NR	NR	X	NR	NR	X	NR	PERT did not prevent weight loss in 60% of the patients and diarrhea in 20%; QoL was decreased due to fatigue and diarrhea	[42]
Suzuki 2016	125 145 LU/day	Complete	41	✓	✓	NR	✓	NR	NR	NR	NR	PERT was effective in stabilizing BMI, reducing diarrhea, and normalizing albumin level	[45]
Saito 2017	48 000 LU/meal	Complete	46	NR	✓	NR	✓	NR	NR	NR	✓	PERT was effective in stabilizing body weight, maintaining nutritional status, and improving overall survival compared to a lack of PERT	[54]
Yilmaz 2018	50 000 LU/meal	Complete	12	✓	NR	✓	NR	NR	NR	NR	NR	Nearly all of the patients receiving PERT became asymptomatic or experienced a reduction of diarrheas	[34]
Dominguez-Munoz 2007	40,000 LU/meal	Complete	29	NR	✓	NR	NR	NR	NR	NR	NR	PERT was able to improve nutritional status (expressed as an increase in CFA and breath test results) as well as to normalize these outcomes at a numerically high rate in the patients	[50]
Dominguez-Munoz 2007	40 000 LU/meal	Complete	20	NR	✓	NR	✓	NR	NR	NR	NR	PERT was effective in gaining weight and improving nutritional status in all of the patients	[50]
Min 2018	96 000 LU/meal	Complete	77	NR	NR	NR	NR	✓	NR	NR	NR	PERT was effective in resolving steatorrhea in 90% of CP patients	[51]
Kim 2018	40–80 000 LU/meal	Complete	47	NR	✓	NR	NR	NR	NR	NR	NR	PERT was effective in improving or maintaining nutritional status in 74% of patients	[46]
Dieguez-Castillo 2020	50 000 LU/meal	Complete	28	NR	=	NR	=	NR	=  (FE1)	NR	NR	No differences were observed between the PERT and no-PERT groups in terms of weight, vitamin deficiency, or FE-1 levels	[52]
Rinzivillo 2020	40–50 000 LU/meal	Complete	4	NR	NR	NR	NR	NR	✓	NR	NR	PERT was effective in resolving abdominal symptoms in all of the patients	[38]
Dominguez-Munoz 2018	325 000 LU/day	Complete	40	NR	NR	NR	NR	NR	NR	NR	✓	PERT was effective in improving overall survival in patients with baseline weight loss compared to no-PERT	[55]
Saif 2020	72 000 LU/meal	Complete	19	✓	X	NR	✓	NR	✓	NR	NR	PERT was effective in improving GI symptoms and reversing weight loss, but vitamin/mineral deficiencies were still observed in individuals	[39]
Shandro 2020	50 000 LU/meal	Complete	81	NR	NR	✓	NR	NR	✓	NR	NR	PERT was effective in improving GI symptoms, and 65% responded to treatment	40]
Olmos 2022	50 000 LU/meal	Complete	7	NR	NR	NR	NR	NR	✓	NR	NR	PERT was effective in improving GI symptoms, including stool scores, distention, and pain	[59]

✓ PERT dose enough to improve or stabilize symptoms, X PERT dose not enough to improve symptoms; = mixed results, NR not reported.

Not shown above: results from Struyvenberg 2017 study—clinical efficacy of PERT was dose-dependent.

^aAccording to the authors, there is no difference in the score between patients on PERT and those without (no numerical data was presented).

Doses Compliant with Guidelines

Ten of eighteen studies assessed gastrointestinal symptoms with complete or partial compliance of PERT doses with the recommended guidelines. In four studies, average PERT doses effectively improved diarrhea, and in two studies, they improved steatorrhea in PEI patients. The recommended 150,000 LU/day was insufficient to relieve gastrointestinal symptoms, including diarrhea, and to prevent weight loss in two studies [42, 52]. Nutritional status was evaluated in seven studies, where doses equal to or higher than 40,000 LU/meal led to maintenance or improvement of nutritional status, expressed as albumin level or retinol-binding protein level. Improvements were also observed in fat digestion measured by CFA and breath tests [39, 45, 46, 49, 50, 52, 54].

In five studies, PERT doses recommended by the HaPanEU/UEG were also effective in preventing weight loss and in four studies in improving general PEI symptoms [34, 37, 39, 41, 45, 50, 54, 58, 62]. In two studies, PERT therapy improved patients’ overall survival compared to a lack of enzyme supplementation [54, 55]. Detailed results are presented in the supplementary appendix.

Another study (Struyvenberg 2017) showed that PERT efficacy increases with higher doses [43]. Rates of diarrhea, steatorrhea, abdominal distention, gas, and bloating were much lower in patients with doses > 96,000 LU/meal compared with those who received doses ≤ 36,000 LU/meal (data shown in the supplementary appendix) [43].

Discussion

This study is the first systematic review investigating PERT dosing and its real-world effectiveness in resolving gastrointestinal symptoms and improving nutritional status. Adequate lipase supplementation is necessary to improve the quality of a patient’s life by reducing gastrointestinal symptoms. Confirmation of response to PERT is mainly obtained by reducing PEI-specific symptoms like diarrhea, steatorrhea, abdominal pain, and flatulence and avoiding weight loss. However, according to the extensive literature, optimal therapy of PEI should not only relieve gastrointestinal symptoms but also normalize digestion and ensure normal nutritional status on a biochemical level to prevent secondary complications of malnutrition [49, 50, 63–65].

There is no consensus on the appropriate dose of PERT for patients with PEI. Some gastrointestinal societies recommend administering 25,000 LU/meal, while others consider 72,000 LU/meal the minimum [4, 10, 21, 66]. For example, Australasian guidelines suggest initiating treatment with the lowest recommended dose of 25,000–40,000 LU per meal for patients with PEI [66]. In contrast, British guidelines recommend starting with at least 50,000 LU per meal and 25,000 LU per snack with patients encouraged to adjust their dose if ineffective [67]. The most recent UEG guidelines state that initial PERT dosing depends primarily on patient age (adult or child), PEI severity, and meal fat content. The most recent UEG guidelines state that initial PERT dosing depends primarily on patient age (adult or child), PEI severity, and meal fat content. However, doses of 40,000–50,000 LU per meal have been recognized as clinically effective [32]. The guidelines used in our analysis (HaPanEU/UEG 2018) recommended a wide range of lipase intake, 40,000–150,000 LU/meal, which translates to about 120,000–450,000 LU/day [23]. Average lipase intake was lower than the recommended 40,000 LU/meal in 40% of the studies included in this analysis. This observation is supported by recent analyses of PERT administration patterns in the US, showing that about 31% of chronic pancreatitis and 27.5% of pancreatic cancer patients who initiated PERT were prescribed doses ≥ 40,000 LU/meal [27, 68]. Another analysis performed by Barkin et al. revealed that 36% of members of the Inspire Pancreatitis or Pancreatic Cancer Support communities were prescribed suboptimal doses of PERT < 40,000 LU per meal [69].

Adherence to PERT treatment seems to be higher in Europe compared to the USA. Recent studies showed that 85.6% of Swedish patients were given PERT according to the HaPanEU/UEG guidelines, while in the Netherlands, 76.3% of patients with PEI initiated PERT [25, 26]. However, there is still an issue with adjusting the PERT dose—available data showed that the dose is not increased even in cases of enzyme ineffectiveness in 37.9% of patients [25].

Several reasons may explain the observed non-compliance with guidelines in real-world practice. Firstly, previous practice guidelines, particularly local and regionally limited, recommended slightly lower lipase doses than European guidelines [21, 22]. For example, the guidelines of the Polish Society of Gastroenterology and the Polish Pancreatic Club, developed by the co-authors of this systematic review, recommended 30,000–40,000 LU per meal in the treatment of chronic pancreatitis [10], which, in light of the results of this systematic review, may be insufficient. Therefore, in the case of many local guidelines, it may be necessary to revise the current recommendations toward increasing the recommended initial PERT dose. Secondly, physicians might initiate lower doses and gradually increase them if there is no improvement in gastrointestinal symptoms. A third potential explanation may be the cost of PERT. According to the analysis by Prescott et al., in 2016, 37% of prescriptions for PERT in the USA were never filled [70]. Physicians may prescribe lower doses of PERT to reduce costs and not discourage patients from taking therapy for financial reasons.

Notably, the latest data indicate that adequate PERT doses were more commonly prescribed by pancreas specialists (82.8%) compared with primary care physicians (29.2%) and gastroenterologists (55.0%) [68]. Better dosage adjustment by more qualified specialists may be associated with greater knowledge and consideration of gastrointestinal symptoms and nutritional status in PEI patients.

This systematic review confirmed that various PERT doses have been used in the last two decades. Nonetheless, prandial doses of PERT rarely exceeded 50,000 LU. Average enzyme supplementation in included studies ranged between 20,000 and 96,000 LU/meal and 45,000–325,000 LU/day. In most of the included studies, lower-than-recommended doses (< 40 000 and < 120 000 LU/day) were sufficient to improve PEI symptoms, including defecation urgency and consistency and weight loss, but could not prevent malabsorption and malnutrition.

However, it is uncertain if deficient lipase intake helps resolve gastrointestinal symptoms. In the Evans 2010 study, 45,000 LU/day was sufficient to reduce diarrhea in 90% of celiac patients, almost three times less than the recommended dose [35]. It has to be noted that PEI diagnosis in these patients was obtained solely by lowered FE1 concentration, which often gives a false-positive result [71]. According to the authors, low concentrations of FE1 in celiac patients may be an artifact resulting from stool dilution characteristic of this condition [35]. Then, it is possible that part of the study population was misdiagnosed with PEI, and their symptoms improved not due to lipase supplementation but because of diet modification leading to stool concentration. Additionally, adherence to a gluten-free diet promotes villous recovery [72], which may further contribute to symptom resolution. Regarding the uncertainties mentioned above, a minimal dose of PERT effective in resolving gastrointestinal symptoms seems to be 25,000 LU/meal. Unfortunately, this dose appears insufficient to improve patients’ nutritional status.

The utilization of PERT plays a crucial role in maintaining optimal nutritional status in PEI patients. The latest research showed that patients with chronic pancreatitis who did not receive PERT had a significantly higher prevalence of vitamin E deficiency (75 vs. 0%, p = 0,012) and osteoporosis (43 vs. 5.6%, p = 0,013) than those receiving PERT. Notably, 75% of osteoporosis cases in the PERT group occurred in patients with inadequate dosing [73]. Our systematic review showed that despite the gastrointestinal symptom benefits obtained by lower-than-recommended doses of PERT, abnormally low nutritional parameters are detected in up to 70% of patients. Doses recommended by the European guidelines, 40,000–50,000 LU/meal, were generally effective in symptomatic benefit and improving or maintaining nutritional status.

For example, in the study by Yilmaz et al., 75% of patients treated with 50,000 LU/meal observed an improvement in diarrhea, with 25% becoming asymptomatic. Unfortunately, nutritional status was not evaluated in this study [34]. Results from other studies showed that a lipase intake of 40,000 LU/meal led to CFA normalization in 66% of patients, while a median dose 20 000 LU/meal was linked with persistent malnutrition in 67% of patients, which was reversed in all patients by doubling the PERT dose [49, 50].

It was necessary to increase the PERT dose during the study course from the initial dose in other included studies (e.g., [42, 45]). However, in one study, even doses recommended by guidelines (150 000 LU/day) were suboptimal, as 68–74% of patients complained about steatorrhea, and weight loss and fat restriction were present in 39–60% and 40–58% of patients, respectively [42, 45]. These findings confirmed that the initial recommended doses are insufficient in some PEI patients. Therefore, the European guidelines allow for an increase of up to 150,000 LU/meal [23]. Nonetheless, according to the results of this systematic review, this does not seem to be a common practice.

In only four of the included studies, an average lipase intake exceeded the minimum recommended doses (40,000–50,000 LU/meal). In the remaining studies, despite suboptimal treatment response, the average lipase intake did not exceed the minimum recommended doses. As we know, enzyme doses should be individually adjusted based on the severity of symptoms, residual pancreatic function, enteric physiology, patient’s age, primary condition, and dietary fat intake [43, 74].

In the included studies, there was a tendency for higher doses of PERT to provide better control of PEI symptoms than lower doses. These outcomes are supported by results from Struyvenberg 2017 study, in which steatorrhea, bloating, abdominal distention, and gas decreased with high doses of pancreatic enzymes (> 36 000, > 96 000 LU/meal) compared with low doses (≤ 36 000 LU/meal), and in which low lipase intake was significantly related to unintended weight loss [43].

Many reasons other than suboptimal dose may also be responsible for the insufficient response to PERT, including those associated with patient compliance, such as an inadequate diet with high-fat content, alcohol abuse, sporadic lipase intake, or incorrect timing of administration [48]. Other factors include acidic intestinal pH, other intestinal diseases, and using acid-sensitive or macrosphere preparations that do not mix effectively with chyme [74]. According to recent guidelines, recommended enzyme formulations should contain mini-microspheres and micro-pellets with dimensions less than 2 mm [23]. It has been shown that formulations with microspheres or pellets with a diameter of 1.0–1.2 have 25% higher therapeutic efficacy than those with a diameter of 1.8–2.0 mm [10]. Also, enteric-coated formulations of PERT are generally preferred to avoid acid-mediated inactivation lipase in the stomach [10, 23]. Hence, patients with inadequate PERT benefits should first ensure that none of the above factors apply to the observed lack of efficacy and then consider increasing the lipase dose. Some authors also conclude that the efficacy of PERT is higher when it is self-dosed flexibly rather than fixed by physicians [75]. It is possible that patients with suboptimal responses were not instructed to adjust their lipase intake.

This systematic review has some limitations, mainly due to the heterogeneity of studies and sample size. No meta-analysis could be performed due to patient characteristics, study design, and follow-up discrepancies. Many of the included studies were not primarily designed to evaluate or determine the efficacy of the lipase dosing. Notably, the trade name of the enzyme formulation was not mentioned in most studies, so it is also unknown if macrosphere or microsphere formulations were used. Moreover, values of average PERT doses were obtained mainly from medical records, physicians, and investigators, not directly from patients. How many patients self-modified their prescribed dose to achieve the desired response is uncertain.

The limitations of this systematic review also extend to the parameters used for assessment and the lack of comprehensive information in the included studies regarding patients’ dietary habits. Nutritional status assessment presents challenges, particularly due to the use of albumin as a marker, which is not a reliable indicator in conditions with chronic inflammation, as its levels may be influenced by disease-related factors rather than actual malnutrition [76].

Moreover, the included studies often lacked detailed data on key factors influencing nutritional status, such as inadequate caloric intake, uncontrolled diabetes, and micronutrient deficiencies. Variations in dietary intake—such as frequent small meals, skipped meals, or reliance on fortified foods and nutritional supplements—can significantly impact enzyme requirements, making standardized dosing approaches less applicable to all patients. However, despite these limitations, this systematic review provided essential insight into real-world PERT dosing and its effectiveness.

To conclude, our systematic review revealed that the real-world doses of PERT were lower than recommended by European guidelines in 40% of identified studies. While lower doses (< 40,000 LU/meal) were found to alleviate gastrointestinal symptoms in most studies, they were insufficient to ensure the normal nutritional status of PEI patients. These findings suggest that relying solely on gastrointestinal symptoms may not be optimal when determining PERT dosing. Instead, the goal should be to achieve nutritional improvement, which requires adhering to guideline-recommended doses. Due to various factors affecting response to treatment, appropriate PERT doses should be established individually, considering symptom management and nutritional status.

Supplementary Information

Below is the link to the electronic supplementary material.

Abbreviations

AP

Acute pancreatitis

CP

Chronic pancreatitis

CFA

Coefficient of fat absorption

FE1

Fecal elastase-1

LU

Lipase units

NOS

Newcastle–Ottawa Scale

PEI

Pancreatic enzyme insufficiency

PERT

Pancreatic enzyme replacement therapy

Author Contributions

Conceptualization: Roland Kadaj-Lipka, Grażyna Rydzewska, Magdalena Monica, Anita Stożek-Tutro, and Przemysław Ryś; Methodology: Magdalena Monica, Anita Stożek-Tutro, and Przemysław Ryś; Database searches and evidence selection: Magdalena Monica and Anita Stożek-Tutro; Formal analysis and investigation: Magdalena Monica, Anita Stożek-Tutro, and Przemysław Ryś; Writing—original draft preparation: Magdalena Monica, Anita Stożek-Tutro, and Przemysław Ryś; Writing—review and editing: Roland Kadaj-Lipka and Grażyna Rydzewska; Funding acquisition: Przemysław Ryś; Resources: Magdalena Monica, Anita Stożek-Tutro, and Przemysław Ryś; Supervision: Roland Kadaj-Lipka, Grażyna Rydzewska, and Przemysław Ryś. All authors read and approved the final manuscript.

Funding

The study was sponsored by Viatris.

Data Availability

Data Sharing Statement: All data that support the findings of this study have been included in the article and its supplementary materials.

Declarations

Competing interest

The authors declare no competing interests.