Rational Design and Animal Studies Remain Gold Standard in Liposomal Delivery Development

In a recent paper (Chen et al 2012), the ‘Lipidoid Group’ from MIT admitted that their library screening method to find useful lipids for therapeutically relevant in vivo applications was flawed.  Essentially, the lipids in the much-heralded 2008 paper with over 200 citations (Akinc et al: 'A combinatorial library of lipid-like materials for delivery of RNAi therapeutics') were tested as crude lipoplexes in tissue culture, whereas successful in vivo delivery has very different and additional demands (stable circulation, extravasation, tissue penetration, biodegradability etc etc)- as is known from decades of liposomal delivery research.  To improve upon the predictive value of the lipid library screening approach, the group now reports, in a solid academic paper, a microfluidics-based approach for formulating lipids into more SNALP-like particles.

I’ve been following with a mix of amusement and annoyance that the Alnylam-funded MIT LNP effort has been getting so much attention to the point that many, if not most, observers believe that the next-generation LNPs are based on that research.  Needless to say, if one bothered to study the respective research coming from Vancouver versus those from MIT, it should be clear which effort alone is the therapeutically relevant one (the Vancouver one to remove any doubt).  The clinical SNALP pipeline says it all.

Who is to blame for this confusion and mis-attribution of credit?  Much of it comes down to scientific journalism looking for eye-grabbing headlines, best if high-profile institutions like the MIT are involved.  Of course, Alnylam does not always bother to point out the relative therapeutic relevance of company-funded research when it issues a press release to go along with a scientific publication.  At least in that regard does their inordinate investment in 10 MIT post-docs pay off.  A bit of a luxury in these times of lay-offs, if you ask me, and I don't expect it to be renewed as the original contract is running out any day now.  I respect Nature Biotechnology as a journal, but having such high visibility, it also has a responsibility to make sure that its prime criteria for selecting papers for publication is originality and research quality, and not renown as it sometimes seems to be the case.

This brings me to Prof. Rob Langer from the MIT and a talk by him at an OTS meeting a few years ago on his exciting ‘lipidoid’ research.  As most readers will be aware, Prof. Langer is famous, amongst other achievements, for fathering what must be dozens of biotech companies (and probably making a buck or two in the process) in various areas of nanotechnology.  Indeed, the apparent frequency of the spin-outs and the different technologies involved makes me wonder at times how a single person, no matter how genius, can be expert in all of them.  In any case, if VCs are keen to throw money after him that's up to them and their investors.  What bothered me about that talk though was that Prof. Langer played along the theme of ‘lipidoids’ having all these unique and interesting properties, including structural differentiation, without really showing proof for them and which, according to the latest paper, do not seem to exist: a SNALP formulation takes on the shape of a SNALP (in this case one with constitutive positive charge at physiological pH).

Looking at the positive side of things, the latest research shows that microfluidics promises to be a technology that allows you to test various SNALP formulations without it costing you an arm-and-a-leg as is current practice.  However, the research also shows that there is no substitute for testing them in vivo as a first indication of therapeutic utility (a lot of mice were used in that study to arrive at that conclusion).  The research parenthetically also strongly supported my impression that microfluidics is unlikely to yield the 20nm-sized SNALP LNPs as Alnylam said their microfluidics-based collaboration with Vancouver start-up Precision Nanosystem would yield (interesting to see Alnylam-supported LNP microfluidics in both Vancouver and Cambridge).  A recent publication involving Precision Nanosystem by the way suggests the same (Zhigaltsev et al 2012) .  In my opinion, 40nm is going to be a magic number for SNALP LNPs.

With the exception of perhaps 1^st-generation LNP delivery to phagocytes, and as had been known in the liposomal research community (including Tekmira, of course) for a long time, rational design approaches and animal experimentation are most likely to rule the delivery space for the foreseeable future.