Skip to Content

Nanoparticles Mix It Up With Reality

Nanoparticles (well, papers about nanoparticles) have been impossible to avoid for. . .what, ten years now, would you say? There’s so much potential there in so many fields, and there are so many things to try, that the literature is a gigantic pile that gets more deliveries dumped on it every week. And how many times have you heard about some great new nanoparticle drug delivery idea, especially targeting cancer? What exactly has happened to all of them?

This paper in Nature Reviews Materials has the answer, or answers. Working out that delivery and PK aspects of these things was already known to be a challenge, but it’s proven to be even more of one than anybody thought:

In this article, we explore the influence of delivery on the efficacy of nanoparticles for cancer-targeting applications. Delivery is important because systemically administered nanoparticle carriers cannot function as designed if they do not access the diseased cells and tissues at a sufficiently high dosage. Nanoparticles, once injected into the body, face both physical and biological barriers (for example, diffusion, flow and shear forces, aggregation, protein adsorption, phagocytic sequestration and renal clearance) that affect the percentage of administered nanoparticles reaching target diseased tissue and cells. We provide a quantification of the delivery efficiency of nanoparticles, review the fundamental principles and the current state and misconception of the biological mechanisms of the nanoparticle delivery process, and describes research strategies that may enhance the delivery efficiency.

Looking over the last ten years of publications, for example, the authors find that the efficiency of nanoparticle delivery to tumors has not improved at all over this period. There are over two hundred papers to work with, but only about half of them gave enough data to be useful (“More time points might have allowed a more precise calculation of the AUC, but very few researchers presented data with more than three time points“). They even contacted authors looking for more data, which is real dedication. But overall, about 0.7% of a systemic dose of nanoparticles actually reaches tumor tissue, it seems, so if you’re looking to do better, there’s the mark to shoot for. Only four papers report values over 5%, and (for what it’s worth) they’re all using particles under 100 nm diameter that are electrically neutral. But working in that space is still no guarantee of success, by any means. It’s also noted that all of these numbers may prove to be overestimates, because it can be difficult to tell if the nanoparticles were actually hitting the malignant cells, or just going into the tumor matrix, etc.

The authors go on to show that typical nanoparticle loadings in these papers would, based on mouse studies, extrapolate to unfeasible human doses – 90 to 200 mL of nanoparticle suspension, which is a rather unlikely clinical strategy. Even synthesizing the nanoparticles on that scale would currently be a challenge. Then you’re faced with that dosing problem, and then you have to worry about the off-target effects of the glass full of nanoparticles that don’t make it to the tissue of interest. The big point here is that the state of the art for cancer nanoparticles is still well short of what it needs to be: dosing efficiencies are going to have to go up by at least an order of magnitude to have any chance.

That makes a nice lead-in to the news that Bind Therapeutics this week filed for bankruptcy protection. Bind has been one of the big names in cancer nanoparticle therapy, but it hasn’t been easy for them, to say the least. I last wrote about them here, after some disappointing clinical results, and those have clearly been a problem for the company since then. The sorts of problems described in this new review article are surely just the sort of things that have been slowing their progress (and are, by the way, the same issues facing the various nanoparticle diagnostic ideas out there as well). In all of these, you’re seeing a very promising concept (several, actually) running up against some severe bioengineering constraints.

The current paper also has a number of interesting things to say about how nanoparticles behave, specifically, how they get out of blood vessels (extravasate) into tumor tissue. The dominant view is that this happens through leaks in the tumor vasculature, gaps between endothelial cells, but the authors suggest that this may well be mistaken, and that new approaches need to be tried:

Distinguishing between intercellular gaps and transendothelial cell pores is difficult, and intercellular gaps can only be demonstrated definitively by labelling the margins with selective junctional markers such as vascular endothelial cadherin; to our knowledge, this has never been attempted. However, the mechanisms and pathways that mediate nanoparticle transport to the tumour are important. If the extravasation is mediated primarily by the transcellular route, nanoparticles that will actively target this transport pathway can be designed. At present, the nanotechnology community has not investigated this transport mechanism. Instead, a heavy emphasis has been placed on studying nanoparticle transport through intercellular gaps via the EPR mechanism — an approach that has thus far yielded poor delivery efficiency. Therefore, there is a need to probe the tumour endothelium transport mechanism to guide the future design of nanoparticles.

EPR is “enhanced permeability and retention”, and the problem is that there doesn’t seem to be very much enhancement going on. There are also a lot of fluid-flow issues inside the tumor tissue itself, even once your nanoparticles have made it there, with plenty of issues yet to be worked out. Overlaying all of these, of course, are the traditional PK issues of having the therapeutic agent sluicing through high-blood-flow organs like the liver and kidneys, as well as partitioning into other tissues you don’t want. Macrophages are especially problematic, since they’re ready and able to go after unidentified particles in this size range. None of these issues are (in theory) unsolvable, but it’s equally clear that no one has solved them yet. As it is, the few approvals in this area have been less targeted and more outcomes-based:

Nanoparticles can be approved for human use by health agencies if they offer an improvement in diagnosis, therapeutic index and/or a reduction of toxicity. At present, only a few non-targeted nanoparticle formulations (for example, Abraxane and Doxil) have been clinically approved because they alter the toxicological profile of drugs in patients. Interestingly, these formulations did not yield significant improvements in therapeutic index or diagnostics. The approval process was based on end-outcome measurements, which are strongly related to how the nanoparticles alter the transportation and function of the small molecule anticancer agent in the body. This suggests the importance of understanding and controlling the delivery of nanoparticles in vivo.

So keep this in mind next time a breathless press release comes out in the field. Those are a bit less common than they used to be (ten or fifteen years of grinding away at these problems has eroded enthusiasm somewhat), but there’s still plenty of room for hype. The good news is that as the hype recedes, and the real problems become more apparent, people are getting to work on them. Nanoparticles may yet be a wonderful delivery vehicle; the promise is still there. But it’s already been a slower road than people were hoping for, and the good results are (once again) probably going to sneak up on us at their own pace.


38 comments on “Nanoparticles Mix It Up With Reality”

  1. ksr15 says:

    Trouble is, everyone wants to get results fast once they see ideas with great promise. Very rarely does this happen because there are usually a myriad of barriers between testing and usable material. Often, the money goes to the low-hanging fruit which isn’t as advanced but can deliver much sooner. (at least, this has been my limited experience)

  2. bhip says:

    Next time a VP/CEO/VC asks an unanswerable hypothetical that requires actual data not yet in existence, reply “I have no nano-idea”” or ” Not a microbiome clue”- it will distract them & their eyes will sparkle with wonder…..

  3. Sleepless in SSF says:

    I’m a little surprised that you’d write so much about nanoparticles without at least mentioning that Harry Kroto died over the weekend. Close enough for a segue at least.

  4. milkshake says:

    I just spent almost 5 years working for a nanoparticle company. Our system based on stabilized micelle was reasonably simple (these nanoparticles self-assembled from detergent-like ambiphilic PEGylated aminoacid oligomer, the encapsulated drug partitioned itself into hydrophobic core, the micelle was stabilized by crosslinking afterwards) and the chemistry part was sound but there were some mysteries about the behavior in vivo, and our management did not encourage taking a closer look… Without going into specifics, I got a strong impression that the entire field of nanoparticle-based drug delivery is plagued by wishful thinking, sales pitch and shoddy biology. To paraphrase an old joke about advertising, half of nanoparticle literature is rubbish and it is not even clear which half.

  5. Anchor says:

    Nanoparticle also means nano cellular toxicity. I wonder if there is any good review on that subject? Thanks.

    1. J Tyson says:

      BIND’s nanoparticle degrade to lactic acid– toxicity wasn’t a problem. Indeed, even in the small so-called “disappointing” trials of BIND-014 nanoparticle docetaxel, they clearly demonstrated a massive reduction in some side effects.

      Last week I heard the CEO of T2 biosystems claim that 25% of cancer deaths are actually from sepsis– maybe it’s about time to take chemotherapy’s ravages on the immune system seriously.

  6. Dr. Manhattan says:

    @milkshake “Without going into specifics, I got a strong impression that the entire field of nanoparticle-based drug delivery is plagued by wishful thinking, sales pitch and shoddy biology. To paraphrase an old joke about advertising, half of nanoparticle literature is rubbish and it is not even clear which half.”

    We do see a lot of wishful (and even magical) thinking in the biotech worlds. However, I would lean more toward Sturgeon’s Law rather than the 50% figure you cite…

  7. mikeb says:

    And isn’t there still a huge question mark about the genotoxicity of nanoparticles (highlighted by a few reviews)? They’re supposed to degrade upon intracellular delivery, but very, very few papers screen for genotoxicity at all.

    1. J Tyson says:

      Nature is already full of natural nanoparticles, some of which attack cancer. Indeed a number of companies are developing so-called “oncolytic viruses”, such as Oncolytics Biotech’s wild-type reovirus and Amgen’s existing approved oncolytic herpes virus T-vec.

      Bind-014 degrades to lactic acid, so the risk to environment was probably entirely due to the payload chemotherapy drug docetaxel. Indeed, the patients administered the drug had markedly reduced side effects.

  8. I think that maybe one should not judge the potential of a technology based on the limitations of not very smart humans looking for a miracle, either in nanomedicine, gentherapy, head transplant or whatever… the low quality of researchers, the poor discussions, the simplistic conclusions, the faking of the results to make them look nice and get into newspapers and get funded, vanity, ignorance, etc, etc… I think that many of the papers on marvelous nanoparticles are simply pieces of low scintific value, low intelectual quality and very poor insight and seriousness… what recreates a field of ignorance that unfortunately impedes a proper development of a truly challenging discipline… I think is not the fault of nanomedicine but the failure of the scientific community as a whole (including funding agencies, companies and scientific journals)… fortunately, time will put things in place and the nonsense will move to the next hype, hot, cool, whatever speculative million grant super journal cover subject…

  9. I may be biased (this group is at UTSW and we collaborate), but I think this group is actively trying to address the main issues being brought up in regards to nanoparticle delivery. Problems are most definitely not solved, but smart people (without the flash) are working on it.

  10. Anon says:

    I am not as pessimistic on nanoparticles as Derek and others here. We are talking about 10-15 years of research here and the fundamental things about nanoparticles are still not understood. You can’t even bring a drug into market in a decade. In my opinion, people should focus on the fundamental aspects of nanoparticle research before jumping into drug delivery.

    PS: Science Magazine please hire a good web developer who can fix this site!

  11. Ex Expat says:

    There are a lot of small companies with nanoparticles, but very few of them will admit to problems with delivery, either because they’re too young or they don’t want to scare off potential funding or partnerships.

  12. LiqC says:

    Great critical review among a lot of noise serving to please friends in the field.

    Check out this review (link in my username) from 2001 about one of the most successful nanoparticles ever conceived. 125 pages and almost 2000 citations: PMID 11749396, Oxygen carriers (“blood substitutes”)–raison d’etre, chemistry, and some physiology.

    Few nanoparticles have ever been characterized in such detail, perhaps short for the clinically approved ones. As the HIV scare faded, PFOB-based artifical blood remained too expensive for general use. They say it was tested in the field and saved a number of lives during Soviet-Afghan conflict.

  13. LiqC says:

    Since we’re on this topic, I’d like to take this opportunity to promote Gordon Research Conference and Seminar on Cancer Nanotechnology. It will take place in June 2017 in Mount Snow, VT. GRC pays for biannual ads in Science, so this announcement is not completely out of place here.

    Confirmed speakers thus far include Chad Mirkin (Northwestern), Omid Farokhzad (MIT), Darrell Irvine (MIT), Sam Gambhir (Stanford).

    Please feel free to contact me with any questions: akscrps at gmail dot com

  14. Raphael Levy says:

    Indeed this review (and your review of it) are an important and very necessary wake up call. I would say that we have reached a point where entire articles, research projects, public perceptions of both risks and potential, as well as personal prizes, professorships and distinctions are based on fictions and fantasies. The one that exercises me most is translocation of nanoparticles across the cell membrane (see e.g. this post at my blog:, but what Warren Chan’s review shows is that this gulf between reality and peer reviewed science extends to these other biological barriers that nanoparticles were supposed to cross. Victor Puntes above thinks that time will sort things out, but I hope that we, scientists interested and active in this field, can take responsibility to accelerate that process to give our science a more solid basis by systematically sharing raw data, being more critical of each others and engaging in scientific discussions, publishing negative results, etc.

    1. banonymous says:

      Was hoping to see Raphael’s comment here 🙂

    2. matt says:

      Hear, hear…
      Going back to a theme from a few days back, it seems science needs a robust number of short-sellers, to balance out all the longs.

      Maybe that should be the theme of a science publication. Rather than purely negative results in a vacuum, they should frankly target papers like this, intending to show specific difficulties with popularly cited or high-impact-published papers/buzzwords/fads. And yes, that means Cell/Nature/Science papers would be scrutinized more heavily, as they should be, and as some people erroneously assume is somehow being done by comments and corrections. (Too often, rather than air both sides of an argument, the journal seems prepared to defend the original published work to the death or to the courts.) The arsenic-based life discussion would be a natural, as might some of the PAINS work.

      An important goal, I think, is to establish or highlight quantitative measurements of the weakness of the field or article, so that starry-eyed beginners can quickly get up to speed on the real problems they need to tackle. Or, as in the case of the PAINS, highlight quantitative measurements that should NOT be used or should be used with caution and properly corroborating evidence.

  15. anon says:

    A similar commentary can be extended to cell-penetrating peptides. Lots of seemingly cool science and papers, no real results.

  16. cancer_man says:

    All I know is that if Uncle Ray says that we will have swarms of nanobots in our brains by 2030, then there will be swarms of nanobots in our brains by 2030, dammit. It is up to you scientists to “make it so.”

  17. Running Comment says:

    I’m a clinical pharmacologist and have been subjected to these things (nanoparticles, liposome-encapsulations, polymers with ‘warheads’, what-have you) about a dozen times in oncology…in a clinical setting, with real patients. I have been meaning to sit down and write a withering piece on the theoretical basis on why these approaches will not…cannot work in terms of delivering improved clinical outocmes, but can never find the time (or the guts to take on an entire industry and some taken-for granted-truths).

    In ‘short’, the drug concentration in a tissue (brain, tumour, liver, muscle, whatever) is largely dependent on 1) the tissue partitioning coefficient of the drug molecule, 2) the drug rate in and 3) the drug rate out. In the basic case, Cb is the bound concentration in the input compartment (e.g. blood), Cf is the free (and therefore diffusible) concentration in the input compartment, Ct is the tissue concentration and Ceff is the effective concentration at the site of action (enzyme, receptor, ion channel etc). What we are really interested in is the concentration of drug at the site of action (Ceff), but since this is not really measureable in vivo, we usually state that the concentration of drug at the site of action is a differential of the concentration of drug in tissue less the effective concentration, Ceff (with their respective rate constants). Rather than the fluid dynamics, the major contributor to the Ceff is the input function, or the extraction of drug to the tissue compartment, which is kinetically dependent of perfusion rates, but the equilibrium is not. This is the micro-manifestation of the volume of distribution, which depends on the physicochemical properties of the compound. Juggling the equations, the equilibrium partitioning (and therefore Ceff) depends on the distributive properties of the drug in the local environment – usually, we are not in a position to change the composition of the tissue, and we can contend ourselves with measuring the rate in and rate out of compound. Changing the properties of the tissue, seems difficult…

    So far, I have not seen a single example of ‘tumour targeting’ achieving the goal of improving the therapeutic index of an established oncology drug (we should bear in mind that, for the most part, improving safety at the expense of efficacy is a no-go in oncology, no matter how many life-cycle management dollars we throw at it).

    1. Shasqi says:

      “Running comment” I really like the way you have analyzed this problem and would love to hear more about it. Can you send me an email ( )

  18. Dan Smithey says:

    Having worked with nanotechnology for drug delivery for the better part of 2 decades, I can say that these problems have been well known and documented, and is why many pharma companies abandoned programs in this area long ago.

    The problem is really one of physics as much as anything. To get anything near a reasonable concentration of drug at the cell surface requires a set of conditions that are simultaneously nearly impossible to achieve using present materials engineering. For example, one must really optimize for two separate PK profiles, one for the nanoparticle, and one for the drug. It is immensely difficult to generally target multiple cell types such that a targeted amount of nanoparticle is achieved at the desired tissue. Then, the particle needs to magically go away, but only after releasing the drug once it reaches its target. And, of course, one must keep the drug within the particle during the transit of the nanoparticle to the target tissue, and then, release at just the desired rate. And if the particle must enter the cell, then endo/phago cytosis must occur, with very high rates of transport, which in itself is exceedingly difficult. Of course, the nanoparticle must have surface structure/chemistry that also fools the immune system as well.

    One can calculate all this using a set of coupled differential equations, and this yields a very narrow set of conditions for true efficacious drug concentrations at the targeted tissue.

    There are better ways to affect disease with drugs.

  19. Odd Physicist says:

    Just an minor note: In describing the nanoparticles that made it to 5% delivery to tumors, you wrote that they were less than 100 nM in diameter. I think you meant nm not nM, as I am pretty sure that nM is nanoMolar, not nanometers

  20. Daniel Barkalow says:

    I have the sneaking suspicion that, in 40 years, people will be talking about why that new nanoparticle drug was so cheap to develop, and someone will be pointing out that it’s because it’s a trivial combination of a 50-year-old nanoparticle and a 5-year-old targetting mechanism, and bringing up all the money spent back in the 2010s by companies that went bankrupt because they were working on a treatment type that would only become viable in 2050.

    1. cancer_man says:

      I was going to write almost the same comment except I would say 2025 to 2030, not 2050. We are still on the exponentially increasing computer power curve and aren’t leaving that anytime soon. Of course, other approaches may be better.

    2. banonymous says:

      In 40 years, you will conveniently never have to worry about the consequences of being wrong 🙂

  21. milkshake says:

    I should emphasize that nanoparticles are expensive, for example the carefully tailored polymers that I was working with cost on the order hundred USD per gram to manufacture (and you would dose a patient with a formulation that contained perhaps ten grams of the stuff). One has to make sure that the nanoparticles do not unravel the moment they hit the plasma albumin, or that the drug does not leak out of these nanoparticles within few minutes and partitions itself into hydrophobic pockets of plasma proteins. I would like to see more unambiguous studies that can distinguish a “free drug” apart from the drug contained within the nanoparticle. Using co-encapsulated FRET pairs would be one convincing way to do this. But not too many groups are doing this, so it could be that all the nanoparticles are just not stable in vivo and the nanoparticle formulation works just as a fancy IV excipient.

  22. farmer says:

    In fairness to Bind, they declared chapter 11 because their lender demanded immediate repayment of the entire loan, not because they are under water based on assets.

    I don’t think there was much hope to improve the efficacy of docetaxel and it seems like they are starting to realize that. With their astrazeneca partnership, they’re improving the tox profile of the asset.

    Sad to see that the field has not been able to make a huge impact. Nano success has pretty much made doxorubicin safer and paclitaxel more tolerable. At least some scientists were trained.

  23. brian says:

    A quick correction, it should be “100 nm”, not “100 nM”.

  24. MattF says:

    One thing to note, from a physicist’s point of view, is that nanoparticles are going to experience fluid flow at -very- low Reynolds numbers. This, in turn, means that the flow is dominated by viscosity and diffusive forces rather than by inertial forces. It’s a common error– anyone who talks about microorganisms ‘swimming’ around is making the same mistake.

  25. Anonymous says:

    I haven’t read the originally cited review yet, but I have worked with carbon nanoparticles (CNP) and I’m familiar with a lot of the literature. I consider a lot of the literature to be highly suspect.

    Non-chemists mix things together and claim that they get the Intro to Organic Chemistry textbook reactions to occur. Non-chemists throw products into analytical instruments that they don’t understand and interpret various blips as proof of textbook structures. It is very difficult to purify CNP. Raw materials (crude CNP) contain a lot of amorphous carbon. Even partially purified CNP contain a lot of amorphous carbon (sometimes as much as 60% by electron microscopy). Reactions and purification methods are often partially destructive. In many cases, you destroy or remove amorphous carbon but the methods also destroy CNPs and convert them into more amorphous carbon: two steps forward, one step back. CNPs form aggregates of various sizes and CNP product mixtures can contain particle sizes spanning several orders of magnitude.

    In a couple of papers, it was shown that different labs studying the same CNP materials (shared between the labs) got different analytical results. Even intralab results varied with the experimentalist and were not even always reproducible by the same technician. (Small changes in sonication efficiency, pipet technique, settling (aggregation) times, etc.)

    Just dose me with around 35 grams of C60 fullerenes in olive oil so I can outlive all the balderdash and see how things shake out.

    Buckyball Longevity – There’s A Problem By Derek Lowe, April 20, 2012

  26. David says:

    “The main advantages of using nanoparticles are that they can be engineered with precise functional properties …” This statement is highly doubtful!

  27. Frank Claire says:

    This kind of paper is very influential because now people in this area will try to do one thing: exceed the previous “records” for systemic dose and publish big time articles — whether or not this was a better therapy or not.

    This is akin to battles for valuing solar cell efficiency over all other factors such as cost scalability or durability, etc.

  28. Raphael Levy says:

    Today saw Nanosphere, founded by Chad Mirkin, being bought for $58M. That company has never made a profit but has raised several hundred millions of dollars over its 17 years of existence ( This is not completely relevant to this thread since Nanosphere is/was a diagnostic company. However, another Mirkin company Aurasense (now Exicure) based on the same nanoparticle constructs, the “spherical nucleic acids”, has also raised huge sums ( and has now announced a first in man (with, as far as I can see very little published preclinical data)
    It will be interesting to see if Exicure does better than NanoInk and NanoSphere.

  29. anon says:

    C&EN just published an article that responds to this paper:

  30. loupgarous says:

    I just got daunorubicin + mitomycin delivered by transluminal catheter to inoperable mets from malignant metastatic paraganglioma in my liver, then embolization of the blood vessel serving the tumors. Side effects from the chemo were actually mild; I was able to suppress the urge to vomit just after the procedure (a big dose of ondansetron helped).

    Perhaps the near future of nanoparticles in oncology will be delivery directly to the tumors (so that it won’t be a good systemic therapy). You’d have a lower body burden of the nanoparticles, and fewer potential sequelae.

Leave a Reply

Your email address will not be published. Required fields are marked *

Solve the math problem. * Time limit is exhausted. Please reload CAPTCHA.