This week, important developments for two of the most advanced cancer RNAi Therapeutic candidates were announced. Arrowhead reported that it has now completed enrolment of the phase I CALAA-01 study and is initiating a dose-optimizing phase Ib study, while Tekmira disclosed that it has received the green light from the FDA to go ahead with an additional phase I study of TKM-PLK1, this time using Hepatic Arterial Infusion to complement an ongoing effort involving intravenous administration. This study will be conducted in collaboration with the National Cancer Institute (NCI).
CALAA-01: Optimizing dosing schedule
CALAA-01 is a targeted cyclodextrin-based nanoparticle formulation (RONDEL) of unmodified siRNA against RRM2, a gene involved in DNA synthesis (more background on CALAA-01 here). Enrolment in the phase I clinical trial commenced 3 years ago. Interim data were presented last year in a Nature paper and the ASCO annual meeting. At that time, about 15 patients with various forms of advanced solid cancers had been treated. Encouraging qualitative biopsy results showed that at least some of the RNAi Therapeutic reached its physical target and was able to mediate RNAi target cleavage there.
Until then, no dose-limiting toxicities had been seen. The ASCO presentation, however, revealed that there were fairly widespread, albeit low-grade (g1-2) immune stimulations (‘flu-like symptoms’). The reason why immune stimulation with this candidate was fairly common is likely related to the fact that unmodified siRNAs were used. Judging from Arrowhead’s quarterly conference call this week, it seems as if the immune stimulation warranted a discontinuation of further dose escalation. The company stated, and I certainly agree, that it should use modified siRNAs in the future to mitigate this side effect.
Interestingly, what the company apparently found in the study is that in some patients the immune stimulation subsided following repeat-administration of the drug, as if they became desensitized to the immunostimulatory potential. It is therefore the goal of the newly announced phase Ib study to find dosing regimes that take advantage of this preliminary finding to achieve higher doses, e.g. by first giving a lower dose in order to desensitize before continuing with the predicted higher, hopefully efficacious dose. If this hypothesis can be confirmed, it would not only be positive for CALAA-01, but a valuable insight for the entire field of RNAi Therapeutics.
Having said that, in light of the clinical findings so far and reservations about the chosen gene target, I do not find CALAA-01 to be a compelling candidate. While I can agree with the phase Ib study, it is more because it will allow the company to gather further information about the RONDEL delivery system without much further delay as it continues its efforts to partner and monetize the delivery technology.
Situations like these highlight the risk of RNAi Therapeutics platform companies like Arrowhead/Calando and Benitec that refuse to have in-house R&D. In this case, in-house R&D could have already readied a modified RRM2 siRNA formulation, or even a formulation targeting a more exciting cancer gene target, without too much additional effort. Instead, more needs to be invested in a candidate that is difficult to partner already and may never make it into phase II, leaving Arrowhead to hope that their argument that the CALAA-01 studies are a way of de-risking the RONDEL delivery platform will convince potential partners and investors.
In my opinion, virtual drug development may be possible for one-in-a-kind therapeutic candidates, but is not compatible with building a platform.
Tekmira and NCI to Begin Hepatic Arterial Infusion (HAI) Study of PLK-1
The clinical development of TKM-PLK1 began earlier this year with a phase I dose escalation study using intravenous administration. Similar to ALN-VSP02, TKM-PLK1 is a SNALP-based formulation, but should be more widely applicable because its pharmacology was tailored so that it can also reach tumor tissues outside the liver. What makes TKM-PLK1 particularly exciting is that its gene target is probably the most potent one in the cancer field, having emerged many times as a top hit in genome- and kinome-wide RNAi screens for rapidly killing cancer cells without affecting normal cells.
The new phase I study of TKM-PLK1 that was announced this week will not only allow the company to more quickly gather important pharmacologic information on the drug, but should also be a viable alternative route of administering TKM-PLK1 for treating primary liver cancer and cancers with liver involvement by using Hepatic Artery Infusion to expose tumors to very high concentrations of the drug which would otherwise not be possible by simple intravenous administration.
Black box less black
One challenge in the development of novel cancer therapeutics is that it can take a long time before their true potential becomes clear. Biomarkers are therefore highly sought after, with RECIST tumor response criteria being a popular way of determining anti-tumor efficacy. Unfortunately, merely measuring immediate tumor responses is not always predictive of therapeutic success as sometimes you see such responses, but without concomitant increases in life expectancies (e.g. Avastin in advanced breast cancer, and many chemotherapeutic settings), and sometimes you don’t see them, yet you do get increases in life expectancies, an issue particularly relevant for targeted therapeutics (e.g. Provenge in prostate cancer).
Access to tumor biopsies is therefore very valuable in determining at an early stage whether your drug functions as expected on a molecular level and what tissue concentrations you need to achieve to trigger these responses. It is therefore an attraction of the new study that it will allow for biopsy collection. Through this, Tekmira should further learn more about which tumors types may be most suitable (e.g. primary/metastatic, genotype, pH).
Aside from the scientific value of the study, this study has the added benefit in that it could help validate Tekmira's platform technology for cancer in a controlled and timely manner, at a time when Alnylam evaluates Tekmira’s technology mainly for knocking down genes in ‘normal’ liver. Clinical data from both these approaches are expected in coming months and should serve as catalysts.
A treatment in its own right
Liver tumors almost exclusively derive their blood supply from the arterial circulation, whereas 75% of the blood flow in normal liver comes from the portal vein. Because of SNALP pharmacokinetics, this means that directly infusing TKM-PLK1 should have the benefit of increasing the drug exposure of liver tumors while at the same time greatly reducing potentially dose limiting exposures of the normal liver(***see correction and clarification below), the predicted dose-limiting organ for SNALP technology. This therefore should make it possible to achieve higher maximally tolerated doses, an important goal in most cancer drug developments.
It is for this reason that Hepatic Artery Infusion (HAI) has become an accepted and frequently used drug delivery modality in the treatment of unresectable liver tumors. Therefore, should the results suggest therapeutic efficacy in this setting, it could well pave the way towards studying HAI of TKM-PLK1 directly in phase III studies.
****Correction and clarification: I have received justified criticism of my statement that HAI would ‘greatly reduce potentially dose limiting exposures of the normal liver’.
The criticism was based on the notion that the normal liver in patients with liver tumors would still go on and be supplied by blood from the hepatic artery in addition to the blood supplied by the portal vein. Consequently, a good chunk of what you provide by HAI would also end up in the normal liver parenchyma. On the other hand, the critique was based on the assumption that the liver tumors would not compete for blood from the hepatic artery. My take on that is that this is probably a good assumption especially for smaller tumors, but in cases where liver tumors make up the bulk of the total liver mass, at least based on high-school physics (pressure and diameter of a tube determining the flow of liquids), there should be at least some quantitative shunting of hepatic arterial blood away from the normal liver.
Moreover, the criticism raised another factor determining whether direct HAI of SNALP will enhance SNALP uptake by cancer cells relative to intravenous infusion, namely whether the SNALP particles get taken up efficiently during the first pass of SNALP through the liver tumor. SNALP uptake is determined by various factors such as time, ionization behavior, active targeting, LDL-receptor status, so that’s probably a more complicated discussion.
Overall, the point is well taken, and while SNALP administration by HAI should have a positive effect on the ratio of drug concentration in Tumor relative to Normal Liver and therefore should increase the therapeutic index, I will have to partially retract my statement in that this is mainly due to optimized exposure of liver tumors during first pass and may vary significantly from tumor to tumor, and that another reason for choosing the hepatic artery vs portal vein may be just the relative surgical ease of accessing these structures.