Clinical Efficacy and Safety of Treatments for Exocrine Pancreatic Insufficiency: A Systematic Literature Review

Exocrine pancreatic insufficiency (EPI) – a condition characterized by deficiency of the exocrine pancreatic enzymes – can lead to clinical manifestations such as steatorrhea, weight loss, severe malnutrition, and maldigestion [1]. Typically, EPI is caused by a range of pancreatic disorders and clinical conditions, including chronic pancreatitis (CP), pancreatic cancer (PC), pancreatic surgery, and cystic fibrosis (CF), in addition to extra-pancreatic disorders such as diabetes mellitus and inflammatory bowel disease [1, 2]. If untreated, EPI can negatively impact patients’ quality of life (QoL), preventing them from completing everyday tasks. Untreated EPI is also associated with an increased risk of mortality [1, 3]. Therefore, early diagnosis and treatment of EPI is crucial to prevent negative clinical manifestations.

Diagnosis of EPI can be challenging, given that it is often associated with other diseases with similar symptoms, especially in the early stages. The current diagnostic gold standard involves quantification of the coefficient of fat absorption (CFA) after 72-h fecal fat determination, but this test is not practical nor typically used in a clinical setting [1]. Instead, noninvasive indirect tests for exocrine pancreatic function are often conducted, including the determination of fecal elastase-1 (FE-1) levels [1, 3, 4]. The FE-1 assay measures pancreatic elastase-1, an enzyme secreted only by the pancreas, which has been shown to be stable during intestinal transit. The assay is often used given its diagnostic accuracy, together with it being patient-friendly, affordable and practicable [5]. This was highlighted in a meta-analysis by Vanga et al. [5] which reported that FE-1 had a pooled sensitivity of 0.96 (95% confidence interval [CI]: 0.79–0.99) and a specificity of 0.88 (95% CI: 0.59–0.97) for identifying patients with EPI compared to quantitative fecal fat estimation. In clinical practice, an FE-1 value of <200 μg/g is used as a conventional cutoff for helping to diagnose EPI [3].

The standard management of EPI includes pancreatic enzyme replacement therapy (PERT) along with a standard high calorie intake nutrition (oral or tube-feeding) and consultation with a dietitian [6]. PERT therapy is composed of lipase, amylase, and protease to compensate for the deficiency of pancreatic enzymes in patients with EPI. Pancreatic lipase hydrolyses triglycerides into monoglycerides, fatty acids and glycerol, while pancreatic amylase hydrolyses the alpha 1–4 linkages in the polysaccharides of three or more linked glucose units [7]. Pancreatic protease comprises trypsin and chymotrypsin, enzymes only produced by the pancreas, responsible for the regulation of other key digestive enzymes and the hydrolyzation of amino acids phenylalanine, tyrosine and tryptophan [8]. PERTs, such as Creon^® and Zenpep^®, are typically administered as enteric-coated Minimicrospheres, with commercially available PERT therapies typically exhibiting minor differences in pH-related release mechanisms and by the amounts of lipase, amylase and protease [9]. However, international guidelines regarding the optimal dosing of PERT do not always align due to the impact of patients’ residual pancreatic function and the fat content on the diet.

While PERTs have been available for several decades, there have been a multitude of recent studies investigating their clinical efficacy and safety in EPI. Therefore, these newer publications necessitate a comprehensive review and summary of the most recent findings. This systematic literature review (SLR) aimed to examine all available information on treatments for EPI, with specific regard to the efficacy and safety of current treatments, and to provide a useful resource on the current treatment landscape.

This SLR was conducted in line with the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines [10], and the Cochrane Collaboration guidance [11]. Relevant articles were identified using the following electronic databases: Embase^®, Medline^®, and Evidence-Based Medicine (EBM) Reviews via the Ovid^® search platform. The searches were designed to capture publications relating to efficacy and safety of PERT. The SLR included a 12-year time horizon (2010 to September 2022) to ensure publications captured were recently published and were of high relevance. In addition to the database searches, a gray literature review and a review of conference proceedings (Digestive Disease Week [DDW], The Professional Society for Health Economics and Outcomes Research (ISPOR), The European Society of Neurogastroenterology and Motility [NeuroGASTRO], American Pancreatic Association [APA], European Pancreatic Club [EPC]) from 2021 to October 2022 were conducted to identify evidence not reported in the published literature. The search strategies used are provided in the supplementary material (for all online suppl. material, see https://doi.org/10.1159/000541326).

Two independent reviewers screened titles, abstracts, and full-text publications, with a consensus reached over any disagreements by a third reviewer. The Population, Intervention, Comparison, Outcome, Time, and Study (PICOTS) design criteria are provided in the online supplementary material (online suppl. Table 1). Key outcomes of interest included CFA, coefficient of nitrogen absorption (CNA), stool characteristics, body mass index (BMI), body weight, nutritional parameters, adverse events, and health-related QoL (HRQoL). Additionally, records that were not published in English were excluded. All relevant data were extracted from the included publications by a single reviewer and a full quality check was subsequently conducted by a senior reviewer.

Included records were also subjected to a quality assessment to determine the strength and applicability of their findings. The critical appraisal tools published in the Joanna Briggs Institute handbook for evidence synthesis were used.

A total of 28 studies that investigated the clinical efficacy and safety of treatments for patients with EPI were included. Figure 1 presents the full PRISMA flow diagram highlighting the number of articles included and excluded at each stage of the review. Twenty-two studies reported data from randomized controlled trials (RCTs). Of the eight that reported the study phase, four were phase II (one phase IIa), two phase III, one phase I, and one was a phase IV trial. The study design was reported for 13 studies: seven double-blinded trials, two prospective interventional studies, and four observational studies. The reported follow-up periods ranged from 4 days to 51 weeks [6, 12].

This SLR identified patients with EPI across a range of comorbid population groups, including patients with CF (n = 14 publications), CP (n = 8), PC (n = 3), and concurrent DM (n = 1). Additionally, two studies were in patients who developed EPI post-pancreatic surgery. The majority of studies (n = 19) reported outcomes in adult populations (≥18 years); overall, the mean age of patients ranged from 7.5 years (patients with CF treated with Creon^®) to 66 years (patients with CP treated with Creon^®) [13, 14]. The sex demographics within patient populations varied, with the percentage of women ranging from 0% to 75% [14, 15].

All studies reported the use of PERT as an intervention: Creon^® was the most frequently reported (n = 13) and included 12 studies comparing various doses of Creon^® with placebo. Other reported PERTs of interest included NM-BL (burlulipase), Zenpep^®, Norzyme^®, Sollpura^®, Ultrase MT, Pancrease MT, and MS1819 (liprotamase). There was no standard approach to reporting dosing, with studies using the US pharmacopoeia lipase units (Ph. U), European Ph.U, international lipase units (IU), and milligrams (mg). Additionally, doses used varied, ranging from 5,000–75,000 in studies reporting US Ph.U and 500 to 75,000 lipase units in those reporting EU Ph.U. Dosing also varied by disease and PERT used; for example, for patients with PC, dosing ranged from 25,000 units of lipase of Norzyme^® to 48,000 lipase units of pancrelipase per meal for a duration of 8 weeks in all three studies. Whereas, for patients with CF, dosing ranged from 1,833 units of Creon^® to 40,000 units of Pancrease Microtablets (MT) per meal.

In the majority of clinical studies investigating PERT, CFA is the primary outcome measure to evaluate improvements in steatorrhea, as a reflection of treatment effect in fat malabsorption [16, 17]. Fourteen studies reported CFA results (Table 1); CFA definitions were consistent in 11 of those studies (see Fig. 2), and three studies did not provide a clear definition [18‒20].

Improvements in CFA levels in patients with EPI were observed in all studies, irrespective of the PERT used. In all studies where interventions were compared to a placebo, the intervention was more effective at increasing CFA levels in patients with EPI, with PERT arms reporting greater levels of CFA in comparison to the placebo arm. For example, Konstan et al. [19] reported that patients with CF and EPI treated with pancrelipase had significantly higher mean CFA percentage (88.6% [standard deviation, SD: 5.0%]) compared with patients treated with placebo (53.9% [SD: 25.5%]; p < 0.0001). Likewise, the Trapnell et al. [26] study reported a statistically significant improvement in mean CFA for patients with CF treated with Pancrease. This improvement in CFA was also reported when stratified by pediatric and adult patients, highlighting that PERTs are efficacious at improving the CFA across EPI patient groups [26]. Among the four studies that investigated the use of Creon^® compared to a placebo arm, improvements in CFA levels were reported. Whitcomb et al. [28] stated that the mean change from baseline in CFA was significantly greater for patients treated with Creon^® versus placebo: 32.1% (SD: 18.5%) versus 8.8% (SD: 12.5%; p ≤ 0.0001). Similarly, Graff et al. [29] reported that, in a two-period crossover study comparing Creon^® (12,000-lipase unit capsules) with a placebo in pediatric populations, CFA levels were significantly greater for patients with EPI treated with Creon^®.

Additionally, studies that investigated the use of Creon^® or Zenpep^® found that the improvement in CFA levels was statistically similar for patients with EPI treated with either intervention. For instance, Taylor et al. [21] compared Creon^® and Zenpep^® (both containing 25,000 lipase units) in an RCT conducted in patients with CF. Overall, the least square (LS) mean CNA-72 h values were similar for Zenpep^® (84.1% [standard error, SE: 1.1%]) and Creon^® (85.3% [SE: 1.1%]), with a difference in LS means of −1.3 (95% CI: −3.6–1.1 [p = 0.297]) [21]. Together, these data display that PERTs are efficacious in improving CFA levels, irrespective of the PERT used and the population treated.

Whereas CFA is considered the primary outcome used to measure steatorrhea, CNA is considered the primary outcome to measure changes in azotorrhea (excess discharge of nitrogen in feces and urine). Ten studies reported CNA results (Table 2). Six of the ten studies presented consistent definitions for CNA (see Fig. 3), whereas four studies did not provide a definition [6, 19‒21]. Where interventions were compared to placebo, the intervention was more effective at increasing CNA levels. Konstan et al. [19] reported that patients treated with pancrelipase had a significantly higher mean CNA (84.0% [SD: 7.4%]) compared with placebo (58.3% [SD: 20.6%]. Equally, Trapnell et al. reported that patients with CF and EPI treated with pancrelipase improved significantly compared with placebo. This observation was true for both adults and pediatric patients.

There was little difference between Zenpep^® and Creon^® regarding their efficacy in improving CNA levels. Zenpep^® and Creon^® (both containing 25,000 lipase units) were compared by Taylor et al. [21] in an RCT conducted in patients with CF. Overall, the LS mean CNA-72 h values were similar for Zenpep^® (80.9% [SE: 1.2%]) and Creon^® (82.0% [SE: 1.2%]; p = 0.334), showing that either PERT was an efficacious treatment for raising CNA levels [21].

These data display a similar trend to that observed for CFA: PERTs are efficacious at improving CNA levels, irrespective of the PERT used and the population treated. For a full list of abbreviations, please see Table 3.

Due to the variation in fat and nitrogen absorption, stool frequency and specific characteristics play an important role in assessing improvements in digestion and absorption derived from treatment with PERTs. However, the reporting of stool characteristics varied across studies, with stool frequency (n = 11) [6, 18, 19, 21, 23, 24, 28, 30‒33], stool consistency/fat content (n = 6) [13, 25, 26, 28, 34, 35], stool weight (n = 5) [6, 18, 23, 24, 29], and stool nitrogen content (n = 4) [6, 23, 24, 29] being reported (online suppl. Tables 2, 3, 4, 5).

Treatment with PERTs reduced stool frequency. For example, Kim et al. [31] investigating patients following pancreatoduodenectomy found that the mean defecation frequency decreased more for the PERT group compared to the placebo group, although the results were not significant (p = 0.073). These results were consistent with Konstan et al. [19] that reported patients treated with pancrelipase having fewer mean bowel movements per day (1.7 [SD: 0.6]) when compared to placebo (2.9 [SD: 1.1]). Improvement in stool characteristics following PERT was not only limited to stool frequency, with all stool outcomes reporting improvements with PERT treatment. Graff et al. [29] reported that the mean stool fat, stool nitrogen, and stool weight were significantly lower during pancrelipase treatment compared with placebo (p ≤ 0.001). Additionally, as observed in data for other efficacy outcomes, Creon^® and Zenpep^® were reported to similarly improve stool characteristics; Taylor et al. [21] reported that both treatments had similar effects on several characteristics including stool frequency, stool consistency, and visible fat/grease in the stool. While these data are more limited, as a consequence of varied reporting of stool characteristics, they suggest that PERT may improve stool characteristics across patients with EPI as a result of improved digestion and absorption.

All publications reporting body weight (n = 10) [21, 24, 25, 27, 30‒33, 35, 36] and BMI (n = 6) [24, 25, 30, 31, 37, 38] as efficacy outcomes concluded that PERTs had a positive effect: either increasing or limiting the decline of body weight or BMI in patients with EPI (online suppl. Table 6). For example, the range of BMI values reported in studies for patients receiving PERTs (16.4 kg/m² to 23.6 kg/m²) was greater than in placebo comparator groups (12.3 kg/m² to 21.2 kg/m²) [29, 38]. In comparison, a study in adolescent patients showed little difference between intervention and placebo groups [29]. Given adolescence is associated with changes in BMI and body weight during puberty, changes in body weight and BMI as a consequence of PERT may be difficult to measure.

Creon^® was the most frequently reported PERT used in studies examining body weight and BMI. Across all populations investigated, Creon^® was successful in increasing body weight and BMI, with Littlewood et al. [35] reporting that patients with EPI due to CF experienced an increase in body weight ranging from 0.1 to 6.1 kg. Other PERTs investigated in relation to BMI and weight loss included Pancrease, Zenpep^®, Norzyme^®, and unspecified pancrelipase and were successful in increasing body weight for patients with CF and CP.

Four studies examined body weight (n = 2) and BMI (n = 2) outcomes in patients with PC treated with PERTs. The efficacy of PERTs to maintain or improve weight/BMI within this population varied; however, results generally indicated that PERTs were unable to maintain or improve both body weight and BMI. For instance, in one publication comparing weight changes in patients with PC who received Norzyme^® or placebo, weight loss was observed irrespective of Norzyme^® use, with no significant differences recorded between the treatment arms [36]. Additionally, in a population of patients with PC undergoing chemotherapy, no significant difference in improvement of BMI was reported between patients treated with pancrelipase and placebo. However, a significant weight gain was observed in one publication reporting on patients who had PC or have undergone total/partial pancreaticoduodenectomy treated with pancrelipase, when compared to placebo (+2.7 kg [SD: 3.4, p < 0.0001]) [32].

Malnutrition is a commonly reported complication of EPI; thus, understanding nutritional parameters is an important aspect of clinical management of EPI when treating patients with PERTs [39]. Nutritional parameters were assessed in six studies (online suppl. Table 7) [15, 31, 32, 37, 38, 40], highlighting a significant gap in the evidence base. The most commonly reported nutritional parameters were cholesterol and transferrin, with median cholesterol levels ranging from 162 mg/dL to 174 mg/dL posttreatment with PERTs [37, 38]. Fat-soluble vitamins were assessed in one RCT investigating the use of PERT; however, there was no significant change in levels of fat-soluble vitamins (vitamins A, E, and K) from baseline following PERT treatment [25]. Two studies reported on the change in nutritional parameters in patients with PC treated with pancrelipase. However, the median levels of albumin, prealbumin and hemoglobin did not increase or decrease significantly with PERT therapy in this population [37, 38].

Eight studies reported an association between improved HRQoL outcomes and PERT therapy (online suppl. Table 8) [13, 15, 21, 26, 31, 32, 36, 41]. In patients with EPI treated with Creon^®, the QoL scores were positive, with just 1–2% of patients reporting “very mild” to “very severe” QoL problems [32, 41]. Creon^® was also shown to provide a positive increase in SF-36 scores in 12% of patients with EPI regardless of the underlying condition, demonstrating that PERT did not negatively affect HRQoL and could cause improvements. Despite the positive impact of Creon^®, the intervention was shown to be less effective than Zenpep^®. For example, in a multinational RCT by Taylor et al. [21] which compared patients with CF treated with either Creon^® or Zenpep^®, the reported CFQ-R Teen scores were 6.8 and 3.7, respectively. However, there was no significant difference in the reported CFQ-R Adult score [21]. While a range of recognized tests, including the gastrointestinal QoL index (GILQI) and the general adaptive composite (GAC) score, was investigated, only a few validated surveys were used, highlighting the lack of standardized questionnaires such as PROMIS and EQ-5D measurement tools currently being used in the literature.

Safety outcomes were reported in 18 publications, with 17 studies reporting treatment-emergent adverse events (TEAEs; online suppl. Table 9, 10) [6, 12, 13, 18, 21, 22, 24‒30, 32, 33, 35, 40]. The rate of TEAEs varied by population group and PERT treatment. Based on the wide variations of reported TEAEs with different branded PERTs, no conclusions can be made regarding differences among them. The incidence of TEAEs in studies investigating the safety of Creon^® was 0–50% [26, 29], while the incidence of TEAEs in populations treated with Zenpep^® and burlulipase were 11.7–19.6% and 65.5%, respectively [12, 21, 40]. Treatment discontinuation due to TEAEs was infrequent; the highest level of TEAE-related discontinuation was 13.8% of patients who received burlulipase [12]. The most common individual TEAEs were headaches, vomiting, and diarrhea. Additionally, ten studies reported no treatment discontinuation for the PERT being investigated. Therefore, while the reported number of TEAEs varies, the low rates of treatment discontinuation indicate that the TEAEs are manageable or not severe enough to warrant discontinuation.

This review of the literature highlights that PERTs are an effective treatment option for patients with EPI, indicated by the consistent improvement of almost all efficacy outcomes following PERT treatment, but especially CFA and CNA levels. Improvements in CFA and CNA are consistent with existing EPI clinical consensus, which recommends the use of PERT to improve fat and nitrogen absorption and reduce EPI-related symptoms [42]. We have identified evidence gaps relating to nutritional parameters and HRQoL, highlighting the need for additional research relating to EPI treatment. This SLR provides a complete overview of the available evidence base across a number of comorbid conditions including CF, CP, and PC, which can be expanded upon in future research.

The effectiveness of PERT-based interventions over placebo is well documented in patients with EPI in a variety of populations. In particular, PERT therapy was considerably more effective at increasing CFA and CNA compared to placebo. In most investigations, the majority of patients experienced a considerable increase in CFA post-initiation of treatment, regardless of intervention or timepoint. These improvements in CFA and CNA following PERT are consistent with results from Gan et al. [43] (2017) who conducted a meta-analysis of seven clinical trials in patients with CP. This study reported that PERT was effective in improving CFA and CNA in contrast to baseline and placebo populations [43]. In addition, the efficacy of Creon^® and other PERTs were highlighted to be similar when compared in an individual study. In a single study in which Creon^® was compared to Zenpep^®, there was minimal difference reported between both treatments for CFA. Frequently, CFA and CNA were reported together highlighting the reliability of these measures in clinical trials when assessing efficacy of PERTs, such as Creon^®. However, given the difficulty of assessing these outcomes in clinical practice due to the time required to conduct the test and patient compliance, further emphasis needs to be placed on identifying other assessments that may be more easily applied in a real-world setting.

The most commonly reported outcomes were CFA and CNA, while the reporting of other outcomes of interest were sporadic, highlighting a gap in the evidence base. Fewer studies reported on outcomes such as stool characteristics, BMI, weight management, nutritional parameters, and HRQoL. Stool characteristics were reported by a limited number of studies but were stated as a good indication of the function of PERT to reduce the inability of patients to process nutrients. In addition to the limited number of studies, the stool characteristics identified varied by definition, specific outcome measure, population, and PERT used. However, these results are aligned with those from Garcia et al. [44] which reported the finding of an SLR and meta-analysis on the clinical efficacy of PERT use in PC and found that PERT use reduced fecal fat excretion, fecal nitrogen excretion, fecal weight, and abdominal pain. Nonetheless, given these limitations, it is difficult to substantiate any key trends relating to the efficacy of PERTs in relation to stool characteristics. Therefore, these findings should be interpreted with care.

Improvement in body weight and BMI was observed with PERTs for all populations. Creon^® was shown to be effective at helping patients gain weight and BMI in CF and CP populations. As poor nutrition is associated with increased mortality in patients with EPI, PERT therapy in combination with a nutritional management plan is imperative to mitigate this risk and reduce the number of complications [45]. Given the importance of nutritional parameters in disease management and the assessment of improvement in patients with EPI, further emphasis should be placed on retrieving and assessing the effect of PERTs on fat-soluble vitamins, blood minerals, and trace elements. While limited, there is some evidence that nutritional parameters may have use in diagnosing and predicting the probability of EPI [46]. For instance, in a study by Lindkvist et al. [46], magnesium below 2.05 mg/dL, hemoglobin, albumin, prealbumin, and retinol binding protein below standard thresholds were associated with EPI in patients with CP in univariate analysis (positive predictive value of 0.88 [95% CI: 0.66–0.97]). A greater understanding of the diagnostic significance of these parameters, through validatory research to confirm their diagnostic accuracy, may lead to the development of more accurate and easily applicable diagnostic modalities.

Nutritional management plans are typically individualized for each patient, depending on disease severity and the level of insufficiency [45, 47]. Therefore, monitoring an individual’s levels of nutritional parameters in response to PERT and nutritional intervention may allow clinicians to tailor interventions to more appropriately address malnourishment in patients with EPI. Patients with EPI may receive nutritional supplementation in addition to PERT therapy. However, across the studies identified in this review comparing nutritional parameters, it was not explicitly apparent if patients with EPI were provided nutritional supplementation, which could potentially introduce heterogeneity and influence results regarding the efficacy of PERTs on nutritional parameters. Additionally, the use of nutritional supplementation may lead to nutritional parameter outcomes being overlooked and underreported within the literature, as individualized nutritional supplementation may vary and prevent accurately analyzing pooled nutritional outcomes in clinical trials. Nevertheless, the lack of available evidence highlights a need for further research to gain a better understanding of the effect of PERT therapy and their relationship with nutritional parameters to reduce symptoms and malnourishment.

HRQoL data were reported infrequently with only eight studies providing results on patients’ QoL. The tools used to assess QoL also varied, with SF-36, SF-12, GILQI, GAC score, global health status, CFQ-R Teen/Adult all being reported. Across studies reporting QoL, minor improvements across treatments were reported. No studies reporting on the EPI-specific patient-reported outcome measure, the PEI-Q, were identified [48, 49]. Given this is an EPI specific measure, it would be beneficial if future studies incorporate this assessment into clinical study outcomes.

In relation to safety, discontinuation of treatments was fractionable in relation to overall incidence of TEAEs. Many of the studies reporting TEAEs included fewer than 100 patients, which were expected due to the low number of patients with these specific underlying conditions. TEAE incidence was relatively high in some of the studies: as high as 65.5% in patients aged ≥12 years with EPI and CF receiving burlulipase [12]. Individual TEAEs were similar throughout studies, and few TEAEs led to discontinuation of treatment.

While patients with EPI attributed to PC were included in this SLR, caution must be taken when interpreting and comparing the results within this population. Treatment guidelines recommend adjuvant chemotherapy treatment following pancreatic resection, which can greatly influence EPI relevant outcomes [50]. For instance, chemotherapy can cause malnutrition and therefore may further impair the effectiveness of PERTs, when compared to those patients who did not receive chemotherapy [50]. Furthermore, the lack of publications identified within this SLR reporting on the clinical efficacy and safety of PERTs in patients with EPI due to PC highlights the need for further research within this patient population.

To our knowledge, this is the first SLR that reviewed published data on the efficacy and safety of PERT in patients with EPI across a number of comorbid conditions (such as CF or PC) in both pediatric and adult populations. As such, this SLR provides a comprehensive account of the clinical evidence for PERT-based treatments for EPI. The results of this review highlight that PERTs are an effective treatment option for improving CFA and CNA values in EPI, as well as some other outcomes. The review identified a number of evidence gaps including a lack of data relating to nutritional parameters, weight management outcomes and stool characteristics. Additionally, safety data and QoL data were not commonly reported despite the importance of these outcomes from a patient-perspective, all of which require further investigation in future trials.

Overall, this SLR aligns with the existing literature on the use of PERT in EPI, highlighting that treatment with PERT is efficacious in alleviating the symptoms of EPI, and the impact of malnutrition in these patients. Future studies should look at using a more consistent approach to the reporting of clinical and PRO endpoints. In particular, outcomes that are more easily assessed in real-world clinical practice (e.g., nutritional parameters) should be included in future studies. This should aid in addressing some of the gaps in evidence and unmet needs for patients with EPI.

The authors would like to thank Louise Heron and James Cochrane from Adelphi Values PROVE™ for their support in conducting the literature review and drafting and revising content.

An ethics statement was not required for this study type since no human or animal subjects or materials were used.

Paula Chu and Jasmina Mioc are employees and stockholders of Organon. Peter O’Donovan and Owen Henry are employees of Adelphi Values PROVE™ and have received funding from Organon to conduct this research.

This study was funded by Organon.

Peter O’Donovan and Owen Henry performed the literature search, selected the relevant studies, completed data extraction, and wrote the manuscript. Paula Chu supervised the development of the SLR, reviewed all the documents, and provided guidance. Paula Chu and Jasmina Mioc reviewed and edited the manuscript. All authors approved the final version for submission.

Data for this manuscript were obtained from published studies, all of which have been cited appropriately in text and tables. Further enquiries can be directed to the corresponding author.