Strategy to develop a MAO-A-resistant 5-hydroxy-l-[β-¹¹C]tryptophan isotopologue based on deuterium kinetic isotope effects

Non-invasive assessment of the serotonergic biosynthesis by 5-hydroxy-l-[β-¹¹C]tryptophan ([¹¹C]HTP) is an established PET marker for many forms of neuroendocrine tumors (NETs) [1] and has recently been suggested as a surrogate marker for the neuroendocrine pancreas [2]-[4].

Di Gialleonardo et al. showed that [¹¹C]HTP accumulation in a pancreatic neuroendocrine cell line was significantly increased by blocking the major route of degradation, monoamine oxidase-A (MAO-A) [2]. Furthermore, this mode of degradation was not present in a ductal cell line. The authors therefore posit that an HTP analogue, following the formation of 5-hydroxy-[¹¹C]tryptamine ([¹¹C]5-HT) by aromatic amino acid decarboxylase (AADC), resistant to MAO-A-mediated degradation will have an increased retention in neuroendocrine tissues containing the amine precursor uptake and decarboxylation (APUD) mechanism. APUD cells include not only the islets of Langerhans in the pancreas but in a broader scope also many forms of NETs. In fact, [¹¹C]HTP-PET was originally developed for assessment of the AADC step in serotonin biosynthesis but has found extensive use as a highly sensitive diagnostic tool for primary or metastatic NETs [1]. However, the contrast to surrounding tissues and hence the sensitivity are affected by the rapid MAO-A-mediated conversion of [¹¹C]5-HT into 5-hydroxy-[¹¹C]indoleacetic acid ([¹¹C]5-HIAA). Current strategies to improve contrast involve peroral administration of the AADC inhibitor carbidopa, which increases the biological half-life of the tracer by reducing peripheral metabolism, e.g., in the liver. However, this method induces large individual variations in tracer metabolism and precludes comparison between examinations and therefore therapy efficacy monitoring [5]. Increased retention of [¹¹C]HTP in NETs could improve the sensitivity of examination in addition to potentially visualizing smaller NETs currently below detection threshold.

Development of new HTP analogues with improved resistance to MAO-A degradation and with retained affinity to AADC is a difficult challenge [6]. A straightforward approach to potentially alter metabolic stability while retaining [¹¹C]HTP structurally congruent to biogenic HTP is to isotopically substitute the compound with deuterium in a position that would induce a kinetic isotope effect (KIE) large enough to deter MAO-A interaction. Similar approaches have previously been employed to improve pharmacodynamics properties of radiotracers and to increase metabolic stability against unwanted degradation which impairs visualization of a target tissue [7]-[9].

Deuterium substitution at the β-carbon adjacent to the reaction center involved in the deamination pathway would typically be expected to give a minor KIE of 1.1 to 1.2. While this effect would not be expected to be particularly strong, the β-position was predicted to be readily accessible for ¹¹C- and ²H-labeling by only minor changes to the original synthesis of [¹¹C]HTP, and thus, it would serve as a convenient test case to explore the means of deuteration to improve MAO-A resistance. A more potent position could be deuteration at the α-carbon, which is expected to yield a primary isotope effect that is significantly higher than the secondary KIE. However, this would require that the deuterium label in α-position remain intact also after conversion by AADC to 5-HT, which is uncertain. [1-N-¹¹C, α,α-²H₂]l-deprenyl, a selective MAO-B inhibitor, has previously been reported to induce a KIE of 3.8 ± 1.1 in baboon brain [9].

Here, we present the radiosynthesis of 5-hydroxy-l-[β-¹¹C²H]tryptophan ([¹¹C]DHTP) and the preclinical evaluation of its stability in regard to MAO-A in vitro and in vivo, in a direct comparison with [¹¹C]HTP.

Improved resistance to MAO-A will potentially yield markedly increased intracellular retention of [¹¹C]HTP in neuroendocrine tissues. [¹¹C]HTP is currently used as a universal diagnostic imaging biomarker for benign and malignant NETs [1]. However, a higher amount of intact [¹¹C]HTP in plasma has been shown to improve the tracer uptake in NETs, and patients therefore are pre-medicated with the AADC inhibitor 100 to 200 mg carbidopa orally 1 h prior to the PET examination [5]. This procedure not only improves sensitivity but also introduces a new parameter that directly influences absolute uptake in lesions: the individual time for gastrointestinal absorption of carbiodopa. Additionally, the time from ingestion of carbidopa until PET examination can be variable in case of delays in radiochemistry. This taken together makes direct quantitative comparison between examinations - even repeated scans within an individual - impossible or difficult. Importantly, it precludes the possibility of quantitatively monitoring treatment efficacy by using SUVs in individual lesions.

One could consider to induce an increase in retention of [¹¹C]HTP in NETs as seen in the endocrine pancreas in Figures 3 and 4 by pre-treating patients with a clinically approved MAO-A inhibitor such as moclobemide. This would likely improve sensitivity substantially, but peroral administration would introduce similar individual variation to that in carbidopa. Intravenous administration on a routine basis is counter-indicated due to the risk of causing serotonergic syndrome due to elevated 5-HT levels in the CNS. Theoretically, development of non-brain-penetrant MAO-A inhibitors could potentially provide an alternative way forward, but we instead suggest minimal re-design of the biomarker itself to achieve the same goal in an elegant manner.

The modification to the previously published [¹¹C]HTP synthesis method to obtain [¹¹C]DHTP was straightforward and involved only change in the [¹¹C]methyl iodide synthesis where LiAlH₄ was replaced with LiAl²H₄. Despite possible KIEs throughout the different synthesis steps, there was no significant difference in radiochemical yields between [¹¹C]HTP and [¹¹C]DHTP (740 ± 240 vs. 620 ± 190 MBq, Scheme 1). LC-MS analysis showed that the incorporation yield of ²H was moderate for [¹¹C]DHTP with 22% containing either one or two ²H. Generally, labeling using deuterated methyl iodide is expected to give higher incorporation of deuterium in the final product [12], but in this case, the reversible enzymatic reaction from pyruvate to hydroxytryptophan is suspected to induce loss of deuterium label. The degradative process that transforms 5-hydroxy-l-tryptophan back to pyruvate proceeds over α-aminoacrylate, the same enzyme-bound intermediate as in the synthetic process [13]. The reversibility allows addition of ¹H at the deuterated β-position in pyruvate and consequently also loss of ²H when converted back again to hydroxytryptophan.

Cell studies showed that [¹¹C]DHTP had increased retention compared to [¹¹C]HTP (Scheme 2). Furthermore, inhibition of MAO-A by clorgyline had less effect on [¹¹C]DHTP as opposed to [¹¹C]HTP. The effect still seen (for example, at the 15 min time point) can still be expected given that only a partial deuteration was achieved. Importantly, inhibition of MAO-A increased the retention of [¹¹C]HTP to the levels seen for [¹¹C]DHTP under baseline conditions. This suggests that [¹¹C]DHTP has improved stability in regard to MAO-A-mediated degradation in vitro, while retaining the affinity for AADC which is crucial for preserving high sensitivity for imaging of neuroendocrine tissues.

After these promising results in cells, we investigated the biodistribution of [¹¹C]DHTP in rodents and especially the effect of inhibition of MAO-A. Increased resistance to MAO-A should in theory result in increased retention in tissues, which facilitates serotonin to a high degree (i.e., the neuroendocrine pancreas in this study). We found a tendency for increased pancreatic uptake of the deuterated isotopologue, but this was not a significant change. [¹¹C]DHTP retention in the pancreas was still markedly increased by inhibition of MAO-A, but to a lesser degree than [¹¹C]HTP. These results show that the current formulation of [¹¹C]DHTP is less sensitive to, but still to a considerable extent metabolized by, MAO-A in rodents. Given that only 25% of the [¹¹C]DHTP was deuterated and still increased retention by 17% to 40%, we estimate that a formulation of 100% deuterated [¹¹C]DHTP would improve the retention in rodent pancreas by approximately 70% to 160%. More substantial resistance against MAO-A, which may be reached by deuteration at the α-position, potentially increases the absolute uptake by up to threefold after 30 min and almost tenfold after 60 min (as seen when inducing complete inhibition of MAO-A by clorgyline, Figure 3).

In non-human primates, however, there was no measurable difference between [¹¹C]DHTP and [¹¹C]HTP regarding baseline accumulation and retention or following MAO-A inhibition. Here, as in the rodent, we can see the potential of reducing degradation against MAO-A, as retention of [¹¹C]HTP in the endocrine pancreas is increased by tenfold by inhibition by MAO-A [3]. A similar increase in retention in APUD tissues will clearly yield a significant improvement in diagnostic sensitivity in the oncological setting.

PET imaging measurement data will invariably have lower sensitivity and more noise than data acquired by well counter measurements (in this case, ex vivo organ distribution studies and in vitro cell studies). This is conceivably one of the reasons for the lack of measurable effect in non-human primates despite promising in vitro and small-animal in vivo results.

We conclude that the present formulation of [¹¹C]DHTP (22.1% ±4.7% deuterated) is still to a major extent degraded by MAO-A in non-human primates and that the secondary KIE introduced by a partial deuteration of the β-position has only a minor effect on the stability of [¹¹C]DHTP against this enzyme despite promising results in cells and in rodents. The small effect is likely due to the low proportion of deuterated molecules, and it follows that increasing the deuteration efficiency at the β-position, or direct labeling of the α-position, may yield improved MAO-A resistance.

The current study shows the feasibility of improving the metabolic in vivo stability of [¹¹C]HTP by minimal re-design. Partial deuteration of the β-position yields improved resistance to MAO-A-mediated degradation in INS-1 cells and in rodent pancreas in vivo, but no appreciable increased retention in non-human primate neuroendocrine tissue. In future studies, we aim to improve labeling efficiency with respect to deuteration of the β-position and attempt deuteration of the α-position to further improve the MAO-A resistance of [¹¹C]HTP and increase its sensitivity as a neuroendocrine marker.