3.1 Liquid spray (LS): Traditional LNS Delivery (STOP 101 Study)

Appreciating the potential benefits outlined above, a liquid nasal spray (LS) DHE formulation was developed and approved in the 1990s. As bioanalytical methods have evolved since the product was approved, and to provide a current comparison against UNS delivery (and IV delivery), the PK parameters, for this “traditional” LS formulation at a dose of 2.0 mg were recently investigated in the STOP 101 study (NCT03401346) [15], which was a randomized, open label, 3-arm, 3-period, 3-way crossover study in health volunteers, dosed with consecutive single doses of 1.0 mg of IV DHE, 2.0 mg of the approved LS formulation to the LNS using a traditional spray and 1.45 mg of the same formulation by Precision Olfactory Delivery (POD®) to the upper nasal space in 6 treatment sequences [15]. A lower dose (of the same formulation) was delivered by POD compared to the LS, as it was predicted that it would lead to greater absorption through the UNS mucosa – as was shown to be the case. The device is actuated by pressing up with the thumb under the assembled device while holding the actuator between 2nd and 3rd fingers in a scissor grip. For further information about the technology, and how it is administered, see Cooper et al. [16].

Subjects were domiciled for 48 h after each dose for safety assessments and PK blood draws and waited 7 days until the next dose [15].

The product was self-administered by the 34 healthy volunteers (the safety population), after reading the instructions for use, by spraying a broad plume of spray once into each nostril, waiting 15 min and then spraying a second spray into each nostril. The PK results showed a Tmax (median) of 0.78 h, with a Cmax of 299.6 pg/mL, and a coefficient of variation (CV) of 91.8%. The AUC0–inf was 2199 h*pg/mL (with CV 74.7%) and T1/2 10.4 h (Table 1). One of the recognized problems of delivery using the “traditional” LS device is that much of the spray plume coats the non-absorbing squamous epithelium of the nasal vestibule and does not land on the absorptive epithelium of the LNS, with very little penetrating through the nasal valve to enter the UNS. This leads to the loss of drug via gravity both out on to the upper lip and down the back of the throat causing complaints of unpleasant taste. Both of these situations, which reduce the amount of drug product on the LNS epithelium available for absorption, were observed and reported, by 77% (dripping) and 56% (running down back of throat) in the STOP 101 study with the LS [15]. In addition, the epithelium of the LNS has a strong rostral-caudal mucociliary clearance (MCC) towards the nasopharynx, which sweeps the mucus and entrapped drug to the back of the nasopharynx with a clearance time of ~ 10 min, in rodents [17], and estimated to be a mean 8 mm/min in humans, ranging from 1 to > 20 mm/min [18].

Table 1 Pharmacokinetics of various DHE products from STOP101 study [15] and STS101-006 study [22]3.2 Liquid Spray: Upper Nasal Space (UNS) Delivery (STOP 101 Study)

After a 7-day washout, the same formulation was again self-administered using the novel POD device [15]. The device was assembled in much the same way as the traditional spray but the POD is activated by pushing the bottle into the body of the device, which triggers a release of propellant [16], which gently pushes the liquid dose through narrow channels in the tip, creating a narrow plume that passes through the nasal valve and into the UNS [16]. That is then followed by the release (from the single dose) of more propellant that pushes the delivered dose further into the UNS and spreads the dose over the epithelium [16]. The 3-period, 3-way crossover study was conducted in 6 sequences, 2 of which received the POD DHE first. Thirty-one subjects provided data showing Tmax 0.5 h with Cmax 1301 pg/mL (CV 51.4%), AUC0–inf was 6275 h*pg/mL (with CV 41.8%) and T1/2 11.8 h [15] (Table 1). In total, 32% reported nasal dripping and 32% reported product running down the back of the throat with POD DHE [15]. The olfactory epithelium of the UNS has cilia but these are non-motile, making MCC much slower from the UNS, estimated at several days [19].

3.3 IV Administration (STOP 101 Study)

In a third period of the same study, 31 volunteers received 1.0 mg of DHE by IV infusion over 1 min [15] giving Tmax 0.08 h with Cmax 14,190 pg/mL (CV 37.0%), AUC0–inf was 7490 h*pg/mL (with CV 16.6%) and T1/2 14.2 h (Table 1). The absolute bioavailability of IV DHE is assumed to be 100% and using the PK population (i.e., results from the volunteers who received and provided data from all 3 treatments), the absolute bioavailability of POD DHE was 58.9% and for traditional LS DHE 15.2% in this study, a 4-fold improvement with the same formulation when delivered to the UNS, despite the lower dose [15].

3.4 DHE Metabolites

Most drugs are metabolized in the liver giving rise to one or more metabolites some of which may be bioactive, while others are not. Dihydroergotamine was first approved in 1946 but it was not until 1984 that its five metabolites were characterized [20], with one, the 8′hydroxy DHE (8′OH-DHE), being the major metabolite. The other metabolites are: 8′,10′-dihydroxy-dihydroergotamine (8′,10′-OH-DHE), 2,3seco,N(1)formyl,3-keto,8′-hydroxy-dihydroergotamine (8′-OH,N(1)formyl-DHE), dihydrolysergic acid amide (DH-LSA) and dihydrolysergic acid (DH-LS) all present at much lower concentrations and not considered to contribute to efficacy (or safety) of the administered DHE product.

It had been believed that the high levels of metabolite were responsible for the recognized long duration of action of DHE. However, as reported, even the levels of the main metabolite 8′OH-DHE are low, of the order of 5%–10% of the parent molecule [15]. More recently, it has been suggested that the long duration of DHE action was due to slow dissociation kinetics from the serotonin receptors (5-HT1B and 5-HT1D) [21], being only one of the classes of receptors to which DHE binds. Notwithstanding that mechanistic explanation for DHE’s well known long duration of action, there has been continued interest (and a regulatory requirement) to characterize the PK profile of the 8′OH-DHE metabolite, if not for the potential efficacy of the compound, then certainly from a safety perspective. This was done in the STOP 101 trial [15]. As can be seen, the plasma concentrations for 8′OH-DHE (Table 2) were of an order of magnitude lower than for the parent molecule (Table 1), which was an important result considering the new route of administration, with the Tmax for both nasal administrations being longer than the parent (as would be expected to give the liver time to metabolize the parent).

Table 2 Pharmacokinetics of 8’OH-DHE from the STOP 101 study (safety population) [15]3.5 Powder Spray: Lower Nasal Space (STS101-006 Study)

Satsuma conducted a separate DHE program with two PK studies with slightly different formulations of DHE from slightly different devices [22]. Both STS101 powder formulations incorporated a proprietary bio-adhesive drug carrier and specially engineered drug particles in an attempt to increase residence time on the mucosa and hence facilitate a greater degree of absorption.

The second program, which incorporates a slightly different plastic bottle delivery device, with thinner walls making it easier to squeeze and expel the powder and incorporated revised training and updated instructions for use (IFU) for the device. Both of the STS101 plastic bottle devices come fully assembled and before use require the plastic tab to be removed from the top of the nose tip, device inserted into a nostril and then a forceful squeeze between 2nd finger and thumb given to the plastic bottle. The revised program with the improved device and IFU included a PK study with a dose of dihydroergotamine base of 5.2 mg (equivalent to a 6.0-mg dose of the mesylate salt), which provided the following results in 35 healthy volunteers: Tmax 0.50 h with Cmax 2090 pg/mL (CV 37.8%), AUC0–inf was 10,100 h*pg/mL (with CV 74.7%). The T1/2 was not reported [23] (Table 1). Although assessment of the 8’OH-DHE levels, specifically AUC, were an important secondary endpoint in both PK studies, as stated on ClinicalTrials.gov (NCT03874832 and NCT05337254) and confirmed in the primary publication from the first PK study [22], that (8’OH-DHE) data have not been abstracted or otherwise reported to date.

3.6 Liquid Spray: Lower Nasal Space Delivery (STS101-006 Study)

In the same study the “traditional” LS formulation at a dose of 2.0 mg in 33 healthy volunteers provided the following results: Tmax (median) was 1.00 h, with a Cmax then of 417 pg/mL, with a coefficient of variation (CV) of 155%. The AUC0–inf was 3450 h*pg/mL (with CV 74.7%), T1/2 10.4 h [23] (Table 1). The product was self-administered by the volunteers after reading the instructions for use; however, it is not clear what degree of device training patients received prior to administration. It is also unclear if the same bioanalytical methodology was used in the STS101-006 study, or laboratory conducting it, as the STOP 101 study, which highlights an important point when conducting a PK study. It is necessary to compare results with different products from within the same study using the same subjects, the same bioanalytical methods run by the same laboratory and analyzed by the same scientists with the same methods at the same time. Despite that the results for the same standard dose of LS DHE across the two programs are not too dissimilar given the high degree of variability observed with the LS DHE in both studies. The bioavailability of STS101 is 2- to 6-fold greater than the LS based on the ratios of the geometric means for Cmax (5-fold) or AUC (AUC0–inf 2.9-fold) [23], but that does not allow for the difference in dose, 6.0 mg vs 2.0 mg and in neither of the STS101 PK studies was an IV arm included to give the 100% absolute bioavailability results. However, within each of the above two studies, comparing different nasal products show the following:

STOP 101: despite delivering less than 75% of the dose, when DHE was delivered by POD to the upper nasal space, it increased the Cmax 4-fold and the AUC0–inf 3-fold, with much reduced variability compared to the LS DHE [15]. Across all 3 DHE products, the major bioactive metabolite, 8’OH-DHE, generated both Cmax and AUC0–last data that were an order of magnitude lower than the parent molecule.

STS101-006: delivering 6.0 mg powder increased the Cmax 5-fold compared to 2.0 mg of the traditional nasal spray (but with much reduced variability). Adjusting for the greater dose of STS101 (6.0 mg), the Cmax of the powder may represent a 1.7-fold improvement in Cmax [23]. The AUC0–inf of the 6.0-mg powder product increased 2.9-fold, compared to the 2.0-mg traditional nasal spray. Adjusting for the 3-fold increase in dose, there appears to be little difference in the amount of DHE absorbed over time between the two formulations when both were delivered, presumably, to the LNS. The data for 8′OH-DHE have not been reported.

There were high hopes that STS101 would be effective; however, two large Phase 3 single migraine attack efficacy studies failed to show statistical superiority versus placebo in the standard co-primary endpoints of pain freedom and most bothersome symptom freedom both at 2 h; EMERGE in 2020 [24] (NCT03901482) with 1065 patients in the modified intent-to-treat (mITT) population and SUMMIT in 2022 [25] (NCT04940390) with 1424 patients in the mITT population. Neither study has been fully published. In the initial study it was reported that patients failed to squeeze the plastic bottle device adequately, with patients only getting a mean of 73% of the intended dose out of the device and into their nose. It is not clear why the SUMMIT trial failed to demonstrate efficacy.

A previous orally inhaled formulation of DHE (MAP0004) was developed but never gained regulatory approval. In a Phase 2 study, the highest dose tested, 2.0 mg nominal (1.0 mg systemic equivalent), was not effective, whereas the lower 1.0 mg nominal (0.5 mg systemic equivalent dose) was [26] and was taken forward to a successful, large Phase 3, FREEDOM 301 study [27]. It was speculated during the development program that Cmax was not the best predictor of subsequent clinical efficacy, but that there might be a “sweet spot” of AUC0-2 [28] that would be a better predictor, based on the results with MAP0004. The AUC0–2 of MAP0004 was reported as 1513 h*pg/mL [28], which was closely matched by POD DHE in the STOP 101 study at 1595 h*pg/mL [15] (Table 1).

In contrast (to the STS101 development), POD DHE performed an open-label safety study of 360 patients (as requested by FDA) (NCT03557333) for Phase 3 [29]. The satisfactory PK results showed that the Cmax and AUC with POD DHE lay between the two approved DHE products (IV DHE as D.H.E.45® and LS DHE as Migranal®) [15], which had both previously shown sufficient safety and efficacy, and these two studies (STOP101 and STOP 301) were sufficient to gain regulatory approval of POD DHE NDA in September 2021.