The Cautious Comeback of Oral Peptides

Converting approved injectable peptides and proteins into non-injected formats would represent a major advance for both the biopharma industry and patients alike. Taking insulin (molecular weight of approximately 6000 Da and 50 amino acids) as the upper limit of the molecular definition of a peptide, there were over 100 injectable peptides in pharma’s clinical pipelines in 2013 (1) – with a market estimated at $23.5 billion by 2020.

Therapeutic peptides have highly specific targets on cell surfaces, exhibit high potency and efficacy, and tend to have fewer side effects than traditional organic small molecules. However, almost all of the currently available agents are designed for injection. Why? Following costly but unyielding investment in the 1990s, few companies seem prepared to take on the financial and scientific risk of developing a new peptide and a non-injected route of delivery in the same formulation. Drug makers are much more likely to consider a non-injected formulation for an established marketed peptide or protein only (2). A consequence of this approach is that it has limited oral peptide formulation scientists to working with just a handful of established marketed peptides – those which were originally designed by medicinal chemists for injection. Important examples include insulin- and glucagon-like peptide-1 (GLP-1)-analogs to treat diabetes (3). It’s a rather poor starting point, with missed opportunities to use novel chemistry to select peptides specifically for oral delivery.

Overcoming oral obstacles

The upper gastrointestinal tract has evolved to degrade and digest proteins and peptides. We can provide the peptide with a safe passage through the stomach using enteric-coated tablets or capsules, so the main challenges lie in managing both the pancreatic and brush-border enzymes (proteases), as well as the permeability of the intestinal epithelial layer. One approach to achieve these two separate goals is to construct a dosage form comprising the peptide of interest, an enzyme inhibitor(s), combined with a permeation enhancer. The aim is to  co-release all three components in proximity to the gut wall to achieve high concentration gradients at the actual site of absorption. Examples of such composite formulations are already employed in over a dozen technologies that have reached clinical trials.

A second approach is to develop a nanoparticle-based construct in which the peptide of interest is entrapped, stabilized with a hydrophilic neutral or negatively charged (anionic) coating to permeate mucus overlying the epithelium, thus delivering nanoparticles in close proximity to the absorptive epithelium. These nanoparticles can be delivered in, for example, a polymeric-coated tablet or capsule. Established permeation enhancers, including stable cell-permeating peptides, can also be used as components of nanocarrier systems, but there is no consensus on whether it is best to release the peptide before or after nanocarrier uptake by epithelial cells. Neither is there agreement on whether nanoparticle uptake across the small intestine in vivo is appreciable or sufficient. Although the nanoparticle approach is elegant, it is also complex and most constructs are far from clinical evaluation.

Regardless of the technology, further significant challenges to safe and effective oral delivery of peptides also exist, such as individual patient variation, not only in their underlying disease conditions but also in respect of variability in gastric emptying, dilution, intestinal transit time, regional luminal pH, and formulation interaction with intestinal contents, including mucus.

To compound the issues above, pharma companies, at least in their published works, appear to restrict oral peptide studies to a limited range of established formulation components, which is  understandable given the regulatory frameworks that exist. Thus, formulation components already classified as non-active ingredients (excipients) or recognized as food-grade materials are highly attractive. In spite of technology improvements in peptide synthesis and design, aversion to creating new chemical entities (NCEs) may be one reason established peptides are used and why the chemical structures of peptides are not normally changed for an oral program. Only in cases where the investment will likely pay off commercially (for example, for long-acting injectable insulins or GLP-1 analogs) are the additional clinical trial and regulatory costs worth the risk. Encouragingly, new, more lipophilic and stable peptide analogues with longer half-lives are starting to filter across to the oral programs of big pharma, and some of these have not yet been approved as injectables. One such GLP-1 analog is claimed to have returned positive Phase II data in February 2015 when formulated with a long-established permeation enhancer (4).

Enhanced controversy

For over 50 years there has been research interest in permeation enhancers in oral drug delivery. The most studied include medium-chain fatty acids, bile salts, acyl carnitines, and calcium chelators. Most have a mild detergent-like effect on intestinal epithelial membranes, but their mechanisms vary – and there is a gap in knowledge about what occurs in vivo in the intestinal lumen environment compared with in vitro permeability studies in a well-defined media in closed systems.

Although most of these agents have a history of fairly safe application to humans for other uses, their use for intestinal permeation enhancement typically involves relatively high concentrations with attendant concerns regarding possible toxicity. For example, bile salts, the medium chain fatty acids, and salicylic acid derivatives all damage the epithelium to some extent. We and others have investigated the damage–repair cycle in the intestine and found that the disturbance to the epithelium caused by sodium caprate is similar to the reaction to many foodstuffs. Transient changes to epithelial cells resolve within 30–60 minutes (5) and the mucosa completely regenerates within 4–5 days. These results are encouraging but they are not a reason for complacency; repeat-dose studies in man may still reveal problems in using these types of agents. In clinical trials, enhancers used to promote the absorption of relatively low-molecular-weight peptides (< 6000 Da) also appear to have a limited effect on promoting systemic circulation access of microorganisms from the intestine (6). However, recent research suggests that even low-grade damage to the epithelium, such as that which accompanies acute binges of alcohol, can promote the permeability of other bystander molecules, including bacterial exo- and endotoxins such as lipopolysaccharides (7).

Doubts regarding the safety of enhancers in drug delivery science have encouraged academic research into NCEs. It is understandable, but a little ironic, that some of these were originally derived from bacterial toxin motifs with precisely targeted effects on specific tight junction proteins (for example, claudin-4) (8). Others are attempting to mimic nature by creating viral-like cell-penetrating peptide analogue NCEs based on a poly-arginine motif to influence the transcellular pathway when co-administered with insulin (9).

Many of these NCEs may have toxicology issues of their own. Those which act by opening tight junctions may be unstable, and there is little evidence in animal models to date that either type of NCE are any more effective than ‘traditional’ non-specific surfactant-like enhancers. So far, none of these NCEs has gone beyond preclinical studies and they have had some difficulty gaining traction with industry due to the early stage of research and the high risks involved. To date, the many clinical trials involving early-generation enhancers have not flagged any particular safety issues associated with such formulations. In fact, the most common issues emerging from the trials are the bug-bears of relatively low bioavailability and high intra-subject variability – in other words, the danger of under-dosing some patients or over-dosing others. In fact, insulin is, prima facie, probably a poor candidate for oral delivery given the necessarily complex plasma profile required to match post-absorptive metabolism of nutrients. The high doses required could lead to hypoglycaemia in some patients, while low bioavailability could allow hyperglycaemia in others. Consequently, it may make better sense to select an oral peptide candidate that is potent and efficacious, and has a wide safety margin in order to generate proof of principle for a therapeutically useful peptide in a large patient cohort.

Success at last?

In recent years, two Phase III oral peptide clinical trials achieved their primary end points, sparking renewed interest in the field. The first was reported in 2012 by Tarsa Therapeutics (USA) with a daily-administered oral formulation of salmon calcitonin (sCT, MW 3420 Da) in postmenopausal osteoporotic women for 48 weeks (10). In a controlled study, patients received enteric-coated tablets containing 0.2mg recombinant sCT, plus several hundred mg of citric acid in the core to prevent attack by serine proteases. The key pharmacodynamic data from the ORACAL trial was a rather weak 1.5-percent increase in bone mineral density over the period, accompanied by reductions in serum cartilage breakdown biomarkers. Importantly, this data still compared well against 33 µg doses of an approved nasal sCT product and was well tolerated by patients. No pharmacokinetic analysis has yet been published, but because the nasal products have an absolute bioavailability of 1 percent, and have no permeation enhancers, we can reasonably assume a similar pharmacokinetic profile to that of the nasal comparator. Whether this oral peptide formulation (OSTORA) will eventually be approved is hard to say; recent regulatory concerns in the US and EU about a possible cancer risk associated with long-term use of marketed sCT in post-menopausal women, makes the benefit–risk for a new oral formulation of sCT debatable.

The second technology for which positive data emerged from a Phase III trial was an oral formulation of the somatostatin peptide analogue, octreotide (MW 1019 Da), from Chiasma (Israel). Octreotide is administered monthly by a painful, long-acting intramuscular injection to patients with acromegaly, an orphan disease caused by overproduction of growth hormone by the pituitary gland. The oral technology is based on the company’s “Transient Permeability Enhancer (TPE)” system and the formulation contains an oily suspension of hydrophilic particles containing sodium caprylate, polyvinyl pyrrolidone, and octreotide, entrapped in an enteric-coated capsule. In the Phase III study, approximately 150 patients responding to injectable somatostatin were then switched to oral octreotide in a complex protocol. The read-outs for the oral daily formulation were reductions in the biomarkers of insulin growth factor and growth hormone over a period of up to 13 months in some cases (11). Chiasma’s oral octreotide formulation was submitted as a New Drug Application to the FDA in June 2015 using the 505(b)(2) regulatory pathway, a route that benefits from previous approvals of individual components and actives therein. However, when comparing the plasma octreotide level achieved in Phase I studies of a 0.1 mg sub-cutaneous injection of octreotide with those seen using a 20 mg octreotide tablet, the relative oral bioavailability was just 0.5 percent by comparison with injection of the much lower dose (12). This result is still promising because the pharmacodynamics end-points were acceptable. Octreotide can be made relatively cheaply, so the loss of over 99 percent of material in the gut might be commercially-viable, in addition to achieving a currently unmet clinical need for patients receiving a highly painful injection.

The uncertain road ahead

The oral peptide field has gone through several cycles. In the 1990s, companies anticipated clinically-useful delivery of even large proteins, such as erythropoietin, using platform technologies. But they failed to deliver commercial products in clinical trials. As a result, big pharma exited the field, but scientists are now revisiting the development of oral formulations for established, highly potent, low-molecular-weight peptides. However, even the most advanced oral clinical trials report consistently low bioavailability (~1 percent).

Few prototype candidates are available and so the race is on to create oral insulin and GLP-1 analogs. Insulin has a narrow therapeutic index, so proposed platforms would need to address patient variability issues to avoid potential risk to the patient. But on a more optimistic note, discovery programs are yielding many interesting and potent peptide molecules for pain, cancer, and cardiovascular disease management. Several are cyclic, low-molecular-weight compounds, and therefore already have more oral delivery potential than many established injectable peptides. Some of these do not require sustained pharmacokinetic profiles – perhaps needing only a short period of peak plasma concentration.

One thing is clear: if we are to fully explore the potential of new peptide structures for oral delivery, we must re-examine the default mind-set of trying to convert an already-marketed injectable peptide to an oral form – and be prepared to create more peptides with structures more amenable for oral delivery as a starting point. While this would involve both scientific and commercial risk, the pay-off could eventually be worth it.