Published on: Mar 3, 2016
Transcripts - Nanoscale_Constrained_Delivery_A_Novel_Technology_for_Subdermal_Implants_2014
Beyond general dislike of needles, standard injection therapy has
several limitations. Poor pharmacodynamic responses, including
side eﬀects deriving from high concentrations immediately
after injection and reduced eﬃcacy owing to subtherapeutic
concentrations between injections, can limit tolerability and
eﬃcacy of treatment. Pressure from payers for improved
compliance and, thus, convenience is driving the need to
innovate novel solutions in drug delivery. Extended-release,
constant-rate implants address all of these issues and may even
improve patient outcomes.1 In addition to producing constant-
rate delivery, an optimal device would be small, easy to implant,
and most importantly, safe over the duration of the implantation.
Recent developments have enabled the creation of a titania nano-
porous membrane that can enable a small implant to provide
passive, long-term, constant-rate drug delivery (Figure 1). Under
standard conditions, the diﬀusion rate is proportionate to the
concentration gradient, according to Fick’s laws. Pore size reduc-
tion decreases the eﬀective coeﬃcient of diﬀusivity, extending
the release while maintaining a correlation between rate of diﬀu-
sion and concentration gradient. Polymer implants, such as Nex-
planon, utilize the aforementioned phenomena to extend the re-
lease of many drugs, including synthetic hormones, for which the
Nanoscale Constrained Delivery: A Novel
Technology for Subdermal Implants
Kathleen Fischer, Krista Degenkolb, William Fischer, and Adam Mendelsohn
Nano Precision Medical, Inc., Emeryville, CA, U.S.A.
therapeutic window is large enough that ﬁrst-order release kinet-
ics do not adversely impact side eﬀects or eﬃcacy. When the pore
size approaches the size of a molecule, the concentration-driven
diﬀusion regime is no longer relevant, and molecules diﬀuse
through the membrane pores at a linear rate that no longer cor-
relates with the concentration gradient across the membrane.2–4
Numerous materials have been used to constrain diﬀusion,
including aluminum, aluminum oxide (alumina), silicon/silicon
oxide (silica), and titanium/titanium oxide (titania). Because the
desired duration for an implant represents a period of months
to years, it must be both stable and biocompatible (nontoxic,
noncarcinogenic, nonantigenic, and nonmutagenic). Additionally,
nanoporous membranes, because of their extensive surface area,
are more susceptible to degradation processes. Furthermore, the
materials must not adsorb suﬃcient material to clog membranes.
Unmodiﬁed alumina membranes can be prone to surface fouling
and pore clogging and can induce mild to moderate inﬂamma-
tion.5 Silicon/silica has not shown signiﬁcant toxicity in vitro or
in animal studies; however, there are only limited studies of its
interaction with human tissue. Nanoporous polymer systems are
in development in the academic laboratory and show signiﬁcant
promise for degradable delivery systems,6 but long-term safety
still needs to be demonstrated.
In addition to the porous membrane materials, one must
consider the material of a reservoir and any material used to
attach the membrane to the reservoir. Although the reservoir
may comprise several types of biocompatible materials,
nanoporous membranes often require adhesives to seal the
membrane with the reservoir, increasing potential toxicity and
stability issues.Titanium has been used extensively in humans
and is well accepted as an implantable biomaterial; furthermore,
a titanium/titania membrane may be sealed to a titanium
reservoir with no other materials added.
Vertically aligned titania nanotubes were grown from titanium
following a protocol similar to that of Paulose et al.7 To produce
the NanoPortal™ membrane, a portion of the titanium structure
and the closed bottoms of the nanotubes were opened.8 Pore
sizes were determined with high-resolution scanning electron
microscopy (FEI Nova NanoSEM 650). Membranes were
attached to reservoirs temporarily using a screw-cap prototype
To test diﬀusion kinetics in vitro, assembled capsules were loaded
with ﬂuorescein isothiocyanate IgG Fab2
Immunochemicals) or ﬂuorescein labeled dextran 3000 (Dextran
3000, Invitrogen) in phosphate buﬀered saline (PBS). Loaded
Figure 1. (A) Molecules are approximately the same size as the pores for
non-Fickian, constrained, linear diﬀusion (middle), whereas they are
signiﬁcantly smaller than the pore size in Fickian, nonlinear diﬀusion
(right). (B) Schematic showing diﬀerences in diﬀusion curves for the two
types of diﬀusion. (C) Scanning electron micrograph of titania nanotubes.
capsules were immersed in PBS and incubated in closed vials at
37°C, with agitation. Samples were read on a ﬂuorescent plate
reader (Cytoﬂuor 4000).
To test in vivo release, assembled capsules were loaded with
polyethylene glycol (PEG, MW 40 kDa) and implanted
subcutaneously in rats. PEG was chosen as a model molecule
because of its stability and ease of detection in blood plasma.
Identical capsules were tested both in vitro (following the
described protocol) and in vivo. To test biocompatibility, animals
with implants were euthanized at 12 months for histopathology
(hematoxylin and eosin stain).
The NanoPortal membrane can be manufactured with pore sizes
ranging from just a few nanometers to 100 nm, and it is made
exclusively of titanium and titania (Figure 2). It can be made as
small as 2 mm, thus ﬁtting inside a 12 gauge needle for
implantation, while still accommodating appropriate amounts of
drug for months of release. By adjusting pore size and the
number of exposed nanotubes, the NanoPortal membrane can be
tailored to ﬁt numerous molecules and applications.
In vitro studies with identical membranes but diﬀerently sized
molecules demonstrated the impact of nanotube diameter on
release rate (Figure 3). Although the FITC-Dextran 3000 had
roughly linear release to 42 days, the residual errors were not
random, and the overall curve was more consistent with extended
Fickian release.The FITC-Fab2
remained linear to 56 days, and
the curve was consistent with non-Fickian, constrained release
kinetics (R2 = 0.99).
As expected with a molecule of 40 kDa, the release rate in vitro
was consistent with that of extended Fickian kinetics (Figure 4).
In vitro delivery was completed around day 14, without any
signiﬁcant subsequent release. The plasma concentration
immediately increased from the time of implant until day 3, as
the PEG released from the capsule in the subcutaneous space
diﬀused into the bloodstream. As the capsule continued to
release PEG, plasma concentrations approached equilibrium
until delivery completed around day 14, at which point the PEG
When implanted in rats, the NanoPortal devices produced no
signiﬁcant diﬀerences in immune reaction compared with control
titanium devices. At 12 months, mature ﬁbrous capsules with
relatively thin walls (10–20 layers thick) and rare mononuclear
cells surrounded both membranes and sham implants.
Figure 2. Schematic of NanoPortal implant. (A) Overview of device
assembly. (B) Constant-rate delivery for a variety of molecules can be
achieved by adjusting the size of the nanopores; target delivery rates can be
achieved by adjusting the number of accessible nanotubes.
Figure 3. In vitro data: with a membrane of the same pore size, FITC-
diﬀuses at a constant rate, whereas the smaller FITC-Dextran diﬀuses
in an extended Fickian manner.
Figure 4. (A) In vitro (top) and in vivo (bottom) experiments using
identical devices tested in parallel (n = 5; error bars are standard
deviation). Delivery is complete at day 14, when around 80–85% of
the loaded mass has been released. (B) In situ (top) and standard
(bottom) histopathology at 12 months conﬁrms no remaining immune
response and a thin ﬁbrous capsule.
Scientiﬁcally Speaking Fischer continued on page 16
Constrained nanoscale diﬀusion oﬀers a promising approach for
extended, constant-rate delivery of a variety of molecules. By
crafting a membrane from titanium and titania, it is possible to
have a high degree of stability and biocompatibility, produce
zero-order release in vitro, generate stable plasma concentrations
in vivo, and easily integrate the membrane with a reservoir to
create a functional, biocompatible system. Furthermore, the
NanoPortal membrane has a range of pore sizes and device
conﬁgurations, permitting its use with numerous molecules while
maintaining a device that could be implanted subcutaneously in
minutes with a syringe needle in an outpatient setting.
Although it has not been addressed here, signiﬁcant, though not
insurmountable (as demonstrated by Intarcia9), formulation chal-
lenges are associated with stabilizing a therapeutic for extended
times in vivo. Nonetheless, because the devices allow a high per-
centage of their volume to be loaded with drug and may be used
with a wide variety of molecules, nanoporous implants provide an
excellent opportunity to reduce side eﬀects and improve eﬃcacy
in the treatment of chronic disease around the world.
1. Henry, RR, Rosenstock, J, Baron, M, Luskey, K, and Alessi,T.
Optimising phase 3 dose selection with ITCA 650: Correlation
between baseline HbA1c and reduction of HbA1c after 24 weeks of
treatment, European Association for the Study of Diabetes Annual
Meeting, Berlin (2012).
2. Fine, D, Grattoni, A, Hosali, S, Ziemys, A, De Rosa, E, Gill, J,
Medema, R, Hudson, L, Kojic, M, Milosevic, M, Brousseau, L, III,
Goodall, R, Ferrari, M, Liu, X. A robust nanoﬂuidic membrane with
tunable zero-order release for implantable dose speciﬁc drug delivery,
Lab Chip 10: 3074-3083 (2010).
3. Martin, F, Walczak, R, Boiarski, A, Cohen, M, West,T, Cosentino,
C, Shapiro, J, Ferrari, M. Corrigendum to “Tailoring width of
microfabricated nanochannels to solute size can be used to control
diﬀusion kinetics,” J. Controlled Release 107: 123-133 (2005).
4. Leoni, L, Boiarski, A, Desai,TA. Characterization of nanoporous
membranes for immunoisolation: Diﬀusion properties and tissue
eﬀects, Biomed. Microdevices 4: 131-139 (2002).
5. Gultepe, E, Nagesha, D, Sridhar, S, Amiji, M. Nanoporous inorganic
membranes or coatings for sustained drug delivery in implantable
devices, Adv. Drug Del. Rev. 62: 305-315 (2010).
6. Bernards, D, Desai,T. Nanoscale porosity in polymer ﬁlms:
Fabrication and therapeutic applications, Soft Matter 6: 1621-1631
7. Paulose, M, Prakasam, HE, Varghese, OK, Peng, L, Popat, KC, Mor,
GK, Desai,TA, Grimes, CA.TiO2
nanotube arrays of 1000 µm
length by anodization of titanium foil: Phenol red diﬀusion, J. Phys.
Chem. C 111: 14992-14997 (2007).
8. Parker, ER,Thibeault, BJ, Aimi, MF, Rao, MP, MacDonald, NC.
Inductively coupled plasma etching of bulk titanium for MEMS
applications, J. Electrochem. Soc. 152: C675-683 (2005).
9. Yang, B, D’Vaz, R, Negulescu, C, Carr, J, Horwege, K, Eftimie, C,
Weeks, E, Gumucio, JC, Gou, M, Alessi,T. ITCA 650 for zero-
order continuous delivery of exenatide at therapeutic levels using 3-,
6-, and 12-month mini-pumps, Diabetes Technology Meeting,
Bethesda, MD (2012).
Scientiﬁcally Speaking Fischer continued from page 15
CRS JOB CENTER
Are you ready to take the
next step in your delivery
Check out the CRS Job Center
Are you looking for qualiﬁed
candidates in delivery
science and technology?
The CRS Job Center can help