ALV003, also known as latiglutenase, is the most commonly studied glutenase. ALV003 is a glutenase composed of two gluten-specific proteases: ALV001 and ALV002. ALV001 is a genetically engineered form of cystine endoprotease B, isoform 2 sourced from barley (Hordeum vulgare) while ALV002 is a modified form of prolyl endopeptidase extracted from the bacterium Sphingomonas capsulata. As of today, ALV003 is being studied through phase 2 trials as the molecule has progressed beyond phase 1 trials, demonstrating tolerability and safety (20).

In a randomised, double-blind study, the efficacy of ALV003 was evaluated by pre-treating food with this glutenase and assessing the T-cell response in 20 patients (including 10 treated with a placebo). Unlike the food pre-treated with ALV003, 6 out of 10 patients in the placebo-treated gluten group exhibited gluten-specific T-cell response in peripheral blood. Both groups, however, experienced gastrointestinal symptoms after ingestion. It is worth noting that, in contrast to other studies investigating ALV003 where a dose of 2 g of gluten for 6 weeks was administered (22, 24), this one involved the administration of 16 g (21).

In a randomised, placebo-controlled, double-blind clinical phase 2 trial by Lähdeaho et al., ALV003 appeared to mitigate gluten-induced damage to the small intestinal mucosa in patients with CeD, within the context of a daily gluten-free diet containing up to 2 g of gluten for 6 weeks, although a statistically significant difference in the presence of any symptoms was not found (22). Conversely, Murray et al. in a multicenter, randomised, double-blind, placebo controlled, dose-ranging study enrolled 494 symptomatic coeliac patients on a GFD for at least 1 year with duodenal mucosal atrophy to assess the efficacy and safety of ALV003. The primary endpoint evaluated any histological changes in the mucosa, while the secondary endpoints included the number of intraepithelial lymphocytes (IELs), antibody positivity, symptom frequency, and drug safety. Although the drug was well tolerated by all participants, the study’s endpoints were not achieved, as ALV003 failed to improve villous atrophy or reduce the severity and frequency of symptoms (23).

In a subsequent study including 50 patients receiving 2 g/day of gluten for 6 weeks and 1,200 mg of latiglutenase a reduction in both mucosal damage and symptom severity compared to placebo, was demonstrated (24).

In the ALV003–1221 clinical trial, a multi-center, randomised, double-blind, placebo-controlled study, although the primary endpoint to achieve histological improvement was not met, treated subjects experienced significant improvement in symptoms and quality of life (QOL). There was a statistically significant, dose-dependent reduction in the severity and frequency of symptoms (abdominal pain, bloating, tiredness, and constipation) in subjects treated with ALV003. Interestingly, Diarrhoea and nausea were the only symptoms which did not improve after receiving the glutenase (25).

Overall, ALV003 (latiglutenase) shows mixed prospects. While it demonstrated some efficacy in symptom reduction, inconsistent results across trials and failure to meet primary endpoints in larger studies suggest limited future development potential.

An endoprotease derived from Aspergillus Niger named AN-PEP is able to degrade both whole gluten and gluten peptides into non-immunogenic residues within minutes (26). A total of 2 studies have evaluated AN-PEP so far (27, 28). Both these studies had limitations due to the small sample size and the short duration of gluten intake (2 weeks). In a randomised double-blind placebo-controlled pilot study, the safety and efficacy of AN-PEP were evaluated. However, prevention of histological damage after receiving gluten and AN-PEP, i.e., the primary endpoint, was not met, despite the overall good tolerability by all participants (27).

A recent RCT investigated the role of AN-PEP in reducing stool gluten immunogenic peptides (GIP). While the use of AN-PEP has been associated with a lower incidence of severe GI symptoms, it failed to meet the primary endpoint, as a significant decrease of stool GIP was not found in patients receiving AN-PEP when compared with the placebo group (28).

The presence of nausea, bloating and abdominal pain were the most commonly reported adverse events (AEs) during the administration of gluten-digestive peptidases; however, their incidence rates did not statistically differ from the placebo group.

To summarise, future development for AN-PEP appears limited due to its failure to significantly reduce GIP or prevent histological damage without major modifications to improve its efficacy.

Four phase-2 studies have examined larazotide acetate, also known as AT-1001, an 8-amino-acid synthetic peptide able to decrease the intestinal permeability, by acting as an antagonist of the zonulin, a key protein in regulation of the gut’s tight junctions. Larazotide acetate is a paracellular permeability inhibitor derived from a protein produced by Vibrio Cholerae and it regulates tight junctions, preventing the passage of gluten into the mucosal lamina propria and the subsequent trigger of the inflammatory response. This drug has no effect on the transcellular passage of gluten (31, 32). Table 3 summarise the results of these studies.

A recent meta-analysis of RCTs on larazotide acetate including 626 CeD patients who underwent ingestion of gluten ranging from 2.5 grams for 3 days up to 2.7 grams for 6 weeks, of which 456 receiving Larazotide acetate and 161 receiving a placebo, showed that the drug reduced the weekly number of symptomatic days and improved symptom severity scores compared to the placebo in patients undergoing gluten challenge. Unfortunately, it failed to demonstrate a reduction in intestinal permeability compared to placebo (33).

During treatment with larazotide acetate, no severe AEs were reported. As previously mentioned, it was able to significantly reduce the incidence of Diarrhoea, abdominal pain, and bloating, which were the most frequently AEs in both the intervention and control groups.

Despite Larazotide acetate’s ability to alleviate gastrointestinal symptoms, it seems unlikely to be a definitive cure for coeliac patients. Instead, it may be considered a complementary option to a GFD in patients with persistent symptoms rather than a substitute of GFD itself.

So far, larazotide investigations have been discontinued despite showing some promise in symptom management.

Nexvax2, was the first therapeutic vaccine created to treat CeD. It consists of synthetic peptides recognised by gluten specific CD4+ T-lymphocytes, leading to their non-reactivity to further gluten stimuli (35).

A phase 1 randomised placebo-controlled trial was initially conducted to evaluate the safety and tolerability of Nexvax2, highlighting that the vaccine did not cause changes in circulating lymphocyte subgroups and no significant changes in the villus-crypt ratio. A subsequent phase 2 study was conducted to evaluate the efficacy of Nexvax2, but it did not demonstrate any beneficial effect in lowering the levels of circulating coeliac antibodies (anti-tTG, anti-DGP), improving duodenal histology and reducing gastrointestinal symptoms (35–37). Regarding AEs, nausea and bloating were usually more frequent in Nexvax2 group (35, 36).

So far, trials on Nexvax2 have been discontinued after unsatisfactory results of phase 2 studies.

Recently, another drug used to induce immunotolerance was engineered, its name is TAK-101 and it consists of gliadin encapsulated in nanoparticles to induce tolerogenic effects (38).

By administering the drug TAK-101 intravenously instead of subcutaneously as Nexvax2, antigen-presenting cells (APCs) with tolerogenic properties in the liver and spleen are activated instead of APCs with immunogenic properties in the skin and lymph nodes. This different approach allows for the induction of an anergic state in gluten-specific T-lymphocytes, while simultaneously activating regulatory T-lymphocytes, which are crucial for achieving the desired tolerogenic effect.

In a phase 2a trial, 33 patients were randomised to TAK-101 and placebo. The number of circulating gliadin-specific IFN-gamma spot-forming T-cells in response to oral gluten challenge was reduced in TAK-101 group compared to placebo. Furthermore, this drug prevented the deterioration of villous-crypt ratio compared to placebo, even if this did not reach the statistical significance (p = 0.1). On the contrary, TAK-101 did not induce clinical changes and did not decrease the percentage of IELs (38).

TAK-101 was generally well tolerated, and no serious AEs occurred. Flushing, headache, back pain, were the most commonly reported AEs, with significant differences between treatment and placebo groups.

Regarding future perspectives, TAK-101 seems promising with its unique mechanism of action, but larger trials are needed to confirm preliminary results.

The monoclonal antibody AMG 714 administered by intravenous infusion exploits a different mechanism of action, namely the inhibition of IL-15 production by APCs and epithelial cells (39).

IL-15 plays a fundamental role in the activation and proliferation of lymphocytes, making CD4+ T-lymphocytes insensitive to the inhibition of regulatory T-lymphocytes and promoting the loss of tolerance to food antigens. AMG 714 did not induce statistically significant changes in the villous-crypt ratio compared with placebo, but only an improvement in lymphocyte density and clinical picture was observed. The authors therefore concluded that AMG 714 may be used beneficially in coeliac patients with persistent symptoms despite a GFD (39).

The effect of AMG 714 was also investigated for type 2 refractory coeliac disease (RCD) in a RCT, given its pathophysiological link with IL-15. After 10 weeks of AMG 714 or placebo, there was no difference between the groups in terms of histological endpoints; nevertheless, patients in the AMG 714 group showed improvement of symptoms compared to the placebo group (40). Serious AEs were reported in 5 patients (26.3%) in the AMG 714 group (pneumococcal infection, elevated transaminases, balance disorder, tuberculosis, and cerebellar syndrome). Safety profile was considered acceptable by the authors considering the severity of RCD type 2. Nasopharyngitis was also commonly reported in AMG 714 group compared to the placebo group (42%vs. 11%).

Overall, the results obtained for AMG714 were poorly satisfactory. Future developments are unlikely due to poor efficacy and concerns regarding its safety profile.

ZED1227 is an orally administered small molecule tissue transglutaminase (TG2) inhibitor that selectively binds to the active form of TG2, thus preventing the formation of deamidated gliadin, its antigenic presentation resulting in gluten-induced T-cell activation (41).

In a double-blind, placebo-controlled study, ZED1227 demonstrated efficacy compared to placebo in reducing mucosal injury and preserving the villous-crypt ratio (p-value <0.001) in CeD patients undergoing a moderate-dose gluten challenge (3 g/daily for 6 weeks), in all proposed dosage (10, 50, 100 mg). Moreover, the effectiveness of ZED1227 has been shown to be dose dependent, with doses ranging from 50 to 100 mg exhibiting greater efficacy in preventing intestinal villous atrophy compared to 10 mg. Furthermore, 100 mg of ZED1227, was effective in inhibiting the increase of IELs consequently to gluten ingestion.

Regarding AEs, headache, nausea, vomiting were the most commonly reported, but there were no differences between groups, except for rash, which occurred in 3 patients (8%) in the 100 mg treatment group.

ZED1227 appears to be the most promising candidate drug with demonstrated dose-dependent efficacy and a good safety profile. It is likely to progress to further development and there is ongoing recruitment for a phase II, double-blind, randomised, placebo-controlled trial in coeliac patients with persistent symptoms despite a GFD (EudraCT/CTIS number 2023–506150-21).

There is significant evidence that gut microbiota can influence and alter the immune system, playing an important role in maintaining a healthy state. Consequently, it is plausible that in genetically susceptible host, imbalances between microbiota and immunity could lead to the onset of a major immune-mediated inflammatory disease, including CeD (42, 43). Three randomised placebo-controlled trials have investigated the role of probiotics as an alternative to a GFD.

In a three-month double-blind, placebo-controlled randomised study, Bifidobacterium longum CECT7347 was found to attenuate the inflammatory effects of dysbiotic intestinal microbiota, decreasing peripheral CD3+ T lymphocytes (p = 0.004), slightly reducing TNF-α concentration (even though it was not statistically significant, p = 0.067), reducing the numbers of the Bacteroides fragilis group (p = 0.02) and the content of fecal IgA (p = 0.011) (44).

Others examined the role of VSL#3™, a well-known probiotic mixture used in inflammatory bowel disease, on patients with CeD. Harnett et al. randomised 42 CeD patients with only partial symptom improvement despite strict adherence to a GFD, in a group treated with VSL#3™ and a placebo group for 12 weeks (45). Unfortunately, no significant differences were found between the two groups at the end of the treatment in bacteria, mycotoxins, or parasites composition, nor for blood urea levels or urinary organic acids.

Rifaximin is a non-absorbable, broad spectrum antibiotic, which acts as an inhibitor of bacterial RNA synthesis and it is mainly used to treat travelers’ Diarrhoea and irritable bowel syndrome. Chang et al. conducted a single-center, double-blind, randomised, placebo-controlled study involving 50 patients to evaluate the improvement of gastrointestinal symptoms in patients with non-responsive CeD with a dose of 1,200 mg per day for 10 days of rifaximin. After randomisation, authors concluded that rifaximin did not improve symptoms in CeD patients with persistent gastrointestinal symptoms following a GFD (46).

The efficacy of budesonide for RCD is well known, on the contrary fewer were the studies about its role in acute reactions to gluten or as alternative of GFD (47, 48).

The impact of budesonide was assessed, in an in vivo and in vitro pilot study, in 20 patients randomised to GFD with or without 6 mg/day of budesonide (49).

Individuals receiving both a GFD and budesonide reported higher well-being scores, increased body weight, reduced frequency of evacuations, and decreased stool weight compared to those on a gluten-free diet. Duodenal biopsies in CeD patients and non-CeD patients were exposed in vitro to gliadin (0.5 mg/mL) and budesonide (10–30 μg/mL) for 3 and 24 h. In vitro budesonide led to a decrease in epithelial tyrosine phosphorylation and histocompatibility leucocyte antigen complex DR (HLA-DR) expression induced by gliadin-derived peptides and in cyclo-oxygenase (COX)-2 and intercellular adhesion molecule (ICAM)-1 in the lamina propria compared to those treated with gliadin alone (49).

Budesonide was also assessed to evaluate its effect on histological response, but no statistically significant differences were observed regarding Marsh grading and villous-height in the studies (48, 50). No major AEs occurred during the therapy with budesonide.

Parasitic helminths may potentially regulate gut microbiota and alter the progression of inflammatory disease.

A successful small trial (12 patients) was conducted by an Australian team by inoculating subcutaneously Necator americanus larvae in CeD patients undergoing gluten challenge (GC), which prevented the worsening of villous trophism and symptoms (51). However, a subsequent larger (54 patients) randomised, placebo-controlled trial failed to reproduce the previous results but confirmed the protective effects on symptoms (52). However, the inoculation of Necator americanus larvae appeared safe, with no severe AEs occurred.

Cross-contamination and inadvertent gluten intake have always been a significant fear for coeliac patients to cope with. However, to define the concept of cross contamination to a GFD is very difficult, as currently no precise definition exists in the literature. It is well known that 50 mg of gluten/day for 90 days represents the minimal toxic dose for coeliac patients; on the other hand, 10 mg of gluten/day is the maximum non-toxic amount of gluten for coeliac patients (53, 54). With regard to these doses of gluten, it has been previously shown that 50 mg of gluten (equivalent to 0.05 grams of gluten) are contained in food samples that a well instructed and conscientious coeliac patient is not likely to eat by mistake. In practical terms, 50 mg of gluten are contained in a large breadcrumb, with a size of approximately 1–2 cm, if we consider that gluten is 75% of the whole protein content of wheat (55). In the trials conducted so far where gluten was administered to patients, the dose varied between 2 g (roughly equivalent to a slice of bread or a packet of crackers) and 16 g per day (roughly equivalent to a large serving of Italian pasta) (20–39, 56, 57), which is definitely a toxic dose of gluten that is very unlikely to be eaten by mistake. A recent international consensus on outcomes measures for CeD trials established that 9 g of gluten/die is the maximum amount tolerated for clinical trials to simulate normal ingestion (58), which is more or less the equivalent of 90 gr of common Italian pasta (a medium portion).

Another relevant aspect to consider is that, although inadvertent gluten intake has been repeatedly reported as a leading cause for persistent symptoms in CeD (11–13), particularly in those patients who may be supersensitive (59), it is very difficult to properly ascertain its causative role in clinical practice. In this regard, a recent study by our group showed that minimal and inadvertent ingestion of gluten in coeliac patients who had been correctly instructed on how to follow a GFD is likely to have no role on triggering intestinal symptoms (60).

A major factor to consider in the evaluation of drug efficacy is the administration of a gluten challenge. In fact, dose, vehicle of gluten administration and duration of gluten-challenge in trials have not been standardised so far. As previously mentioned, GC dose varied between 2 g and 16 g per day (20–39, 56, 57). This aspect is even more challenging if we consider that also in clinical practice diagnostic gluten-challenge is complex to perform. A recent ESPGHAN position paper (64) provides guidance on how to perform GC in children, although this is mainly based on expert opinion, whereas in adults no guidelines provide guidance for gluten challenge (65–67). Table 5 summarises the main concerns related to GC in clinical trials.

Run-in periods are useful to reduce drop-outs from a trial (68); however, run-in periods were introduced only in 5 trials (24, 27, 32, 36, 37). Recently, run-in periods have been suggested for trials contemplating gluten-challenge in order to increase compliance and reduce confounding in the evaluation of symptomatic response (58).

The length of GC varied greatly among studies, ranging from a single bolus dose to up to 10 weeks in one case (31, 37, 39). Furthermore, the type of gluten vehicle was not standardised and several different methods of administration were used, such as capsules but also baked products such as cookies, bread and biscuits that may be rich in fermentable oligo-, di-, and monosaccharides and polyols (FODMAP) (32, 35, 39, 40, 57), which are known to trigger symptoms in IBS and also in coeliac patients on a GFD (69, 70).

Additionally, the adherence to GC in the trials was not extensively evaluated, neither with specific questionnaires nor with objective tools such as GIP (24, 28). Lastly, the influence of the so called ‘trial effect’ on patients enrolled in trials should also be considered, as this may lead patients to improve their adherence to the GFD, potentially confounding the beneficial effect of the drug compared to controls (23).

The precise definition of histological recovery is a mandatory outcome to establish before starting a trial. This concept is challenging also in clinical practice, as many parameters should be considered such as the patchiness of duodenal lesions, the amount of time required for healing, the histological criteria adopted. So far, histology has been the primary endpoint of 5 trials (22–24, 39, 41) and this was effectively met only in one (41). Moreover, different methods (villous height to crypt depth (Vh:Cd) ratio, lactulose-to-mannitol (LAMA) ratio and densities of IELs) have been used to evaluate histological changes, which makes it difficult to compare the results and inevitably introduces an observer variability. Indeed, LAMA is not specific for CeD, but it only provides an indirect and less appropriate measure of histological damage by assessing intestinal permeability (31, 32, 56, 57). According to a recent consensus, a Vh:Cd ratio ≥ 2.5 or ≥ 3 or Marsh 1 lesions were considered necessary criteria to enter a trial where gluten challenge is performed in order to avoid the ethical concerns related to offering gluten to patients with persistent villous atrophy (58).

The use of PROs as trial endpoints has been gaining importance over the last decade, due to their extensive application in pharmacological trials, particular those related to inflammatory bowel disease and functional gastro-intestinal disorders (71–73).

PROs provide measures of patients’ QOL and assess how objective clinical effects alter the subjective sphere and viceversa. Indeed, patients’ clinical characteristics such as anxiety, resilience and hyper-vigilance could potentially skew the results, contaminate trial’s endpoints and change symptoms perception (74–76). Furthermore, PROs promote a more patient-centered evaluation and regulatory agencies such as the European Medicine Agency and the Food and Drugs Administration have also recognised their significance.

Few CeD trials have investigated PRO (37, 56), but their inclusion is desirable in future trials as suggested by a recent international consensus (58).