We are Nitinol.™
`
`(cid:1) (cid:1) (cid:1) (cid:1) (cid:1) (cid:1) (cid:1) (cid:1) (cid:1)
`
`(cid:2)(cid:37)(cid:1)(cid:14)(cid:44)(cid:29)(cid:40)(cid:44)(cid:33)(cid:29)(cid:45)(cid:1)(cid:38)(cid:30)(cid:1)(cid:13)(cid:33)(cid:42)(cid:33)(cid:37)(cid:38)(cid:35)(cid:1)(cid:12)(cid:29)(cid:28)(cid:33)(cid:27)(cid:25)(cid:35)(cid:1)(cid:2)(cid:39)(cid:39)(cid:35)(cid:33)(cid:27)(cid:25)(cid:42)(cid:33)(cid:38)(cid:37)(cid:41)(cid:1)
`
`(cid:1)
`
`(cid:5)(cid:47)(cid:31)(cid:44)(cid:35)(cid:33)(cid:52)(cid:1)(cid:17)(cid:31)(cid:38)(cid:46)(cid:41)(cid:40)(cid:52)(cid:1)(cid:19)(cid:46)(cid:41)(cid:31)(cid:29)(cid:37)(cid:31)(cid:38)(cid:1)
`
`(cid:1)
`
`(cid:14)(cid:27)(cid:46)(cid:31)(cid:44)(cid:35)(cid:27)(cid:38)(cid:45)(cid:1)(cid:19)(cid:29)(cid:35)(cid:31)(cid:40)(cid:29)(cid:31)(cid:1)(cid:27)(cid:40)(cid:30)(cid:1)(cid:6)(cid:40)(cid:33)(cid:35)(cid:40)(cid:31)(cid:31)(cid:44)(cid:35)(cid:40)(cid:33)(cid:1)(cid:2)(cid:68)(cid:73)(cid:69)(cid:60)(cid:68)(cid:73)(cid:71)(cid:1)
`(cid:42)(cid:42)(cid:54)(cid:67)(cid:70)(cid:75)(cid:60)(cid:67)(cid:72)(cid:66)(cid:1)
`
`(cid:1)
`
`(cid:67)(cid:75)(cid:75)(cid:75)(cid:1)
`
`www.nitinol.com
`
`47533 Westinghouse Drive Fremont, California 94539 t 510.683.2000 f 510.683.2001
`
`MEDTRONIC 1012
`
`

`

` Materials Science and Engineering A273—275 (1999) 149-160
`
`
`
`MATERIALS
`SCIENCE &
`ENGINEERING
`
`
`
`www.elsevier.com/locate/msea
`
`An overview of nitinol medical applications
`T. Duerig *, A. Pelton, D. Stéckel
`a Cordis Company, 47533 Westinghouse Drive, Fremont, CA 94539, USA
`Nitinal Devices and Components,
`
`Abstract
`
`Superelastic nitinol is now a common and well-known engineering material in the medical industry. While the greater flexibility
`a
`of the alloy drives many of the applications, there are
`large numberof lesser-known advantages of nitinol in medical
`actually
`reasons for nitinol’s success, both past and future. Several
`devices. This paper reviews 10 of these less-obvious, but very important,
`new medical applications will be used to exemplify these points, including the quickly growing and technologically demanding
`a
`particularly interesting in that they involve new and complex manufacturing techniques, present
`applications. Stents are
`stent
`demanding and interesting fatigue environment, and most
`interestingly, take advantage of the thermoelastic hysteresis of nitinol.
`© 1999 Elsevier Science $.A. All rights reserved.
`
`Aevwerds; Nitinol; Mechanical applications; Superplasticity; Shape memory behavior; Stents
`
`
`1. Introduction
`
`The commercialization of shape memory alloys, and
`is a
`success
`story. The
`specifically nitinol,
`truly unique
`discovery of shape memory in Au—Cd and Cu-Zn
`occurred with little fanfare in somewhat obscure techni-
`cal papers withlittle, if any, follow-on work. However,
`when the shape memory effect was rediscovered in
`a
`equiatomic Ni—Ti in 1962, there was
`great
`suddenly
`deal of commercial
`interest. Early commercialization
`by applications such as
`activities, fueled
`rivets, heat
`engines, couplings, circuit breakers and automobile ac-
`tuators, Were intense and often highly secretive. Metal-
`were
`to solve
`the microstructural
`lurgists
`quick
`mysteries of shape memory, and by the early 1970’s
`even the minutes details of the
`could
`explain
`shape
`memory process. Unfortunately, industry’s understand-
`ing of engineering lagged well behind: non-linear tensile
`fatigue, and adiabatic
`properties. hysteresis,
`heating
`and cooling effects were
`a few of the new
`problems
`just
`baffling product designers during these years [1]. More-
`over,
`alloy melting and
`processing remained expensive
`and unreliable, with
`relatively few product forms avail-
`able.
`By the early 1980’s it became clear that shape
`memory was not a financial panacea. When the
`picture
`*
`
`Corresponding author.
`E-mail address: tduerig@mail.nitinol.com (T. Duerig)
`
`had not
`improved much by the early 1990’s, many
`companies threw in the towel and returned to core
`technologies.
`As we
`the end of the century the
`picture
`approach
`looks very different. Nitinol has become a ‘household’
`word in the medical engineering world. Nitinol produc-
`ers have experienced explosive growth. In general,
`it
`has not been the original
`large companies that have
`survived, but a core of entrepreneurs whose belief in the
`was so
`strong that they started small com-
`technology
`to
`panies when the efforts of larger companies began
`flag. Attention turned towards
`superelasticity instead of
`the more
`complicated shape memory effect and towards
`medical applications, particularly implants. The human
`body offered an isothermal environment that
`seemingly
`Indeed
`solved many of the design complexities.
`body
`temperature turns out
`to be perfectly tuned to the
`superelastic “sweet spot’ of the basic binary Nitinol
`alloys, and does not
`require the cryogenic alloys of
`and fasteners or the very high-M, alloys of
`couplings
`actuators.
`What
`
`in fortune?
`the sudden change
`really triggered
`was
`certainly well known since the early
`Superelasticity
`1970’s. Material costs have come
`down, but not
`enough
`to drive the market and, in fact, material costs make up
`a small fraction of the total cost of a medical
`only
`device and costs seldom enable or disable an
`applica-
`reasons for the
`tion. There appear to be three
`
`primary
`
`

`

`150
`
`T. Duerig et al.
`
`Materials Science and Engineering A273-275 (1999) 149-160
`
`sudden success.
`the medical
`most
`Perhaps
`importantly,
`industry itself has been driven towards less and less
`invasive medical procedures. This, in turn has created a
`can not be
`demand for new medical devices, that really
`made with conventional materials. Other factors were
`to laser
`and the ability
`the availability of microtubing
`we should
`cut tubing with very high precision. Finally,
`not underestimate the importance of the ‘release’ of
`technology from materials science technologists and
`to
`product designers and doctors.
`companies
`illustration of all
`these points is
`Probably the best
`also the most celebrated superelastic
`medical device: the
`stent. The word stent derives from a
`self-expanding
`dentist, Dr C.T. Stent, who in the late 1800°s developed
`an
`a dental device to assist
`impression of
`in forming
`teeth. Nowadays, the term stent is reserved for devices
`used to scaffold or brace the inside circumference of
`tubular passages or lumens, such as the esophagus,
`a host of blood
`biliary duct, and most
`importantly,
`iliac, aorta and
`vessels
`including coronary, carotid,
`femoral arteries (Fig. 1), Stenting in the cardiovascular
`to balloon
`system is most often used as a
`follow-up
`procedure in which a balloon as
`a
`placed
`angioplasty,
`in order reopen a
`in the diseased vessel and expanded
`a stenosis). These balloons are
`clogged lumen (called
`most often
`introduced
`percutaneously (non-surgically),
`imme-
`through the femoral artery. Ballooning provides
`in blood flow, but 30%of the
`diate improvement
`patients have restenosed within a year and need further
`placement of a stent
`treatment. The
`immediately after
`angioplasty has been shown to
`significantly decrease the
`propensity for restenosis [2].
`Stents are also used to support grafts, e.g.
`in the
`treatment of aneurysms (Fig. 2). An aneurysm is caused
`by the weakening of an arterial wall, that then balloons
`are
`a risk of rupture. Surgical repairs
`out and presents
`often difficult. With the endovascular approach,a
`graft
`is placed through the aneurysm and anchored in the
`
` Fig. 1. A stent is portrayed in a cut-away viewof the internal carotid
`
`artery, maintaining vessel patency and blood flow to the brain. The
`stent is also pinning loose debris against the vessel wall, reducing risk
`of embolization and stroke.
`
` Fig. 2. The Hemobahn product
`
`a
`superelastic Nitinol
`incorporates
`wire into a PTFE. Stent-grafts such as these are used to exclude
`an artificial replacement for injured vessels, or
`aneurisms, to provide
`to prevent restenosis after angioplasty.
`
`healthy part of the artery at least at the proximal neck
`of the aneurysm. Thus, blood is excluded from the
`are
`and
`aneurysm sack. The grafts
`typically supported
`stent structures.
`anchored by self-expanding
`are 316L stainless steel. and are
`Most stents
`today
`deformation
`expanded against the vessel wall by plastic
`the inflation of a balloon placed inside the
`caused by
`stent, Nitinol stents, on the other hand, are
`self-expand-
`—
`are
`to the open configuration,
`ing
`shape-set
`they
`into a catheter,
`out of the
`then pushed
`compressed
`catheter and allowed to
`expand against the vessel wall.
`the manufactured stent OD is about 10%
`Typically,
`in order to assure the stent
`greater than the vessel
`Nitinol stents are nowavailable
`in
`anchorsfirmly
`place.
`are made from
`in both Europe and in the USA. They
`knitted or welded wire. laser cut or
`photoetched sheet,
`devices are
`and laser cut
`tubing. Clearly the preferred
`laser cut from tubing,
`thus avoiding overlaps and welds
`(Fig. 3).
`stents and other devices?
`Why usenitinol for making
`The most apparent feature of superelastic (SE)nitinol is
`
` Fig. 3. Stents are often made from laser cut tubing. In this case, the
`
`laser beam cuts a kerf of less than 25 jim, through
`200 pm,
`
`a thickness of over
`
`

`

`T. Duerig et al. / Materials Science and Engineering A273—275 (1999) 149-160
`
`151
`
`*
`
`Fluid Exit
`Port
`
`
`
`
`"
`
`Nemperature
`Montloring
`
` \
`
`— —— lem Markings
`
`
`
`Trecar
`
` Insulated —_
`
`
`Fig. 4. The RITA tissue ablation device uses four sharply curved
`tubular needles which are
`deployed from a
`straight needle after
`a trocar.
`insertion through
`
`ployed until the position is verified to be marked cor-
`rectly for the surgeon. A stainless steel device would
`require the wire diameter to be much smaller, assuming
`the hook geometry is to remain fixed. Such a fine wire
`would be too
`to anchor the hook effectively. It
`flimsy
`would also allow inadvertent transsection, leaving bits
`of wire behind in the breast.
`a curved device
`The concept of
`elastically deploying
`a
`straight needle or cannula is probably
`the
`through
`most common use of nitinol for medical instrumenta-
`tion. Among the newer devices are the TUNAprostrate
`ablation device, the Daum defiectable puncture needle,
`and the RITAtissue ablation device (Fig. 4). While the
`latter two devices deliver curved needles, other devices
`such as suture passers, retractors, deflectable graspers
`and scissors have been in use since the early 1990’s in
`endoscopic surgery.
`The atrial
`defect occlusion system (ASDOS)is
`septal
`a more
`complex device [4]. This system allows non-sur-
`gical occlusion of holes in the atrial wall of the heart,
`ranging from 20 to 35 mm in diameter. The entire
`two low profile
`is conducted
`procedure
`through
`catheters. The actual device is comprised of two small
`of five nitinol wire loops support-
`umbrellas consisting
`ing webs of
`microporous polyurethane (Fig. 5). The
`two devices are
`while folded one
`passed into the
`body
`on either
`each in two catheters, and then positioned
`side of the defect area. A
`guide wire passing directly
`the two
`through the hole is used to ensure that
`catheters and umbrella devices are
`aligned correctly.
`the umbrellas are advanced from their
`Once
`positioned,
`a
`catheters, and screwed together using
`special torquing
`resulting ‘sandwich’ forms a
`oc-
`catheter. The
`patch,
`to
`is too
`cluding the atrial defect.
`Although it
`early
`
`that its flexibility is 10-20 times greater than stainless
`steel; that is to say, one can observe devices to
`‘spring-
`back’ with strains as
`as 11%. This in situ flexibility
`high
`a role in some
`superficial stent
`applications such
`plays
`as the carotid and femoral arteries, where the vessels
`to outside pressures that would cause
`may be subject
`conventional stents to crush. Such deformations have
`been observed in stainless steel stents, and can lead to
`serious consequences. Although superficial applications
`are rare,
`there are many
`in situ flexibility
`requiring
`subtle aspects of
`superelasticity that actually drive the
`selection of nitinol for all stent
`even those
`applications,
`not
`to deformations. The purpose of this
`subjected
`some of these more subtle aspects of
`paperis to
`explore
`and to discuss them with respect
`nitinol
`performance,
`to
`medical
`11
`applications. More specifically,
`specific
`will be introduced:
`Elastic
`deployment
`Therma! deployment
`Kink resistance
`
`Biocompatibility
`stresses
`Constant unloading
`Biomechanical compatibility
`Dynamic interference
`Hysteresis
`MR compatibility
`Fatigue resistance
`Uniform
`plastic deformation
`
`2, Elastic deployment
`
`One of the most common reasons to use nitinol is to
`allow the efficient deployment of a medical device.
`Modern medicine has been
`steadily driving towards less
`and less invasive procedures. Entire operations of in-
`are
`performed through small, leak-
`creasing complexity
`into the body called trocars; vascular
`tight portals
`diseases are
`repaired by passing wires and instruments
`percutaneously through needles into the femoral artery
`re-
`and on to the heart, brain, etc. These procedures
`quire instruments and devices that can pass through
`and then
`very small openings
`elastically spring back
`into the desired shapes. Many of these devices could be
`made with any flexible material, but clearly nitinol
`allows the greatest freedom in
`design.
`to be marketed was
`Probably the first such product
`use
`to
`the Homer Mammalok, which radiologists
`‘mark’ the location of a breast tumor. It consists of a
`nitinol wire hook and a stainless steel cannulated needle
`[3]. The wire hook is withdrawn into the needle can-
`nula, and then the cannula is inserted into the breast
`and adjusted until its tip is verified to be at the site of
`a
`the tumor. The hook is then advanced, reforming
`If necessary, the device can be
`tight hook configuration.
`withdrawn into the needle,
`and re-de-
`repositioned,
`
`

`

`152
`
`T. Duerig et al,, Materialy Science and Engineering A273-275 (1999) 149-160
`
`convincingly evaluate the success of this particular
`product, it well illustrates the concept ofelastic deploy-
`are now
`ment. Several companies
`marketing similar
`devices.
`
`3. Thermal deployment
`
`An additional unique attribute of nitinol devices is
`can be deployed using the shape memory
`that
`they
`effect. One example is the Simon venacave filter (Fig.
`em-
`6). The device is intended to filter rather large
`bolized blood clots in the vena cave vein [5],
`as blood
`returns to the heart from the lower body. The clots are
`the filter's
`trapped by the legs of the filter and by
`‘flower’, then dissolved over time. Such clots are com-
`mon in bed-ridden patients and pose a serious hazard if
`they reach the heart or
`lungs. The device itself is
`preloaded into a catheter while in the martensitic state.
`Flushing chilled saline solution through the catheter
`keepsthefilter in the martensitic phase while position-
`to the deployment site. When released from the
`ing
`the
`catheter and the flow of chilled saline stopped,
`device is warmed by the surrounding blood, and recov-
`ers its
`‘pre-programmed’ shape.
`Like the vena cave
`filter, stents also have an
`A,
`temperature only slightly above room
`temperature.
`
` Fig. 5. The atrial septal occlusion device incorporates nitinol wires in
`
`polyurethane film in order to repair defects in the wall of the heart.
`
`a
`
` Fig. 6. The Simonvena cavafilter expands
`
`once in the vena cava, and
`left permanently in place to prevent emoblized thrombus from
`is
`reaching the heart and lungs.
`
`are
`superelastic in the body yet martensitic
`Thus they
`at
`when constrained into the sheath. When deployed
`adopt their ex-
`room
`the stents will not
`temperature,
`panded shape; that occurs
`only when body temperature
`to 3-8 times the
`is reached. Stents typically expand
`catheter diameter.
`
`4. Kink resistance
`
`To some extent this design property stents from the
`increased elasticity of superelastic nitinol, but it is also
`a result of the shape of the stress—strain curve. When
`strains are
`locally increased beyond the
`plateau strain,
`stresses increase markedly. This causes strain to
`parti-
`tion to the areas of lower strain, instead of increasing
`or strain localiza-
`the peak strain itself. Thus kinking,
`a more uniform strain
`is prevented by creating
`tion,
`than could be realized with a conventional material.
`to take advantage of this fea-
`The first applications
`ture were
`angioplasty guidewires, which must be passed
`tortuous
`without kinking [6]. The wires,
`paths
`through
`place, form a
`over which other devices
`once in
`guide
`are advanced,
`including angioplasty balloons, stents,
`filters, etc. The wires must be very long when accessing
`in the more distal parts of the body, such as the
`places
`Paths can also be
`brain from a femoral access
`point.
`very tortuous and full of side branches, making it very
`that the wires are steerable and torquable.
`important
`Even very small permanent deformations will cause the
`to steer the wire
`wire to
`and destroy the ability
`whip
`through side branches or around sharp bends. In order
`improvelubricity, the wires are
`to
`generally coated with
`a
`Teflon or a
`hydrophilic coating, and often employ
`to
`at the
`helical wrap of fine Pt
`improve radiopacity
`tip. There can be
`littke doubt
`that nitinol
`distal
`important role in the success
`an
`guidewires have played
`of angioplastic medicine.
`
`

`

`T. Duerig
`
`et al. / Materials Seience and Engineering A273-275 (1999) 149-160
`
`153
`
`Another early application is retrieval baskets with
`as well as a
`nitinol kink-resistant shafts,
`superelastic
`basket to retrieve stones from kidneys, bladders, bile
`with an
`ducts. etc. Still another
`ingenious
`example,
`is a line of laparoscopic instruments made by
`twist.
`Surgical Innovations (Fig. 7). The shaft of the instru-
`ofa series of hollow knuckles strung
`ments is
`composed
`on a stainless steel cable. When the cable is relaxed, the
`knuckles are loose and shapeless; when the actuatoris
`the knuckles are
`and the
`tensioned,
`pulled together,
`The
`shaft is forced into specific, predetermined shapes.
`a means to pass
`object of the innovation is to
`provide
`a shaft
`a trocar, then reform a
`rigid complex
`through
`once the instrument end is
`the trocar
`shape
`through
`and into the body. The actuator was
`originally braided
`stainless steel so that
`it would not kink when the
`knuckles locked into their formed shape. The problem
`the device became so
`with braided cable was that
`flaccid when the tension was released that
`it was
`difficult to advance through the trocar. The solution
`was to
`replace the actuator cables with a nitinol rod,
`some
`to the relaxed state yet
`which provides
`rigidity
`can be bent with
`the
`knuckles without plastic
`deformation.
`superelastic tubing became available in the
`Since
`a
`to mid 1990's,
`variety of catheter products and
`
`early
`
` Fig. 7. Surgical Innovations endoscopic instruments use nitinol rods
`
`to actuate scissors, graspers, etc. A braided stainless steel cable
`(visible in the center grasper) is used to tension the articulating joints,
`locking the instrument in a
`once past the trocar.
`predetermined shape
`
`
`
`4\
`
` Fig. 8, The arrow interaortic balloon pump uses a nitinol tube to
`
`pressurize the balloon. Stainless steel would not be flexible enough,
`and a
`polymer tube would have to be a muchlarger diameter.
`
`other endovascular devices using nitinol as the inner
`on the market. An interesting example
`lumen appeared
`is the intraaortic balloon pump (IABP) used in cardiac
`procedures (Fig. 8). The use of nitinol allowed a
`assist
`reduction in the size of the device compared with
`polymer tube based designs, and increased theflexibility
`and kink resistance compared with stainless steel tube
`designs.
`tubing limits the use of
`Kinking of thin-wall steel
`many interventional devices. Biopsy forceps made from
`are very delicate instru-
`stainless steel, for example,
`ments, which can be destroyed by
`even very slight
`mishandling. Nitinol instruments on the other hand can
`or
`handle serious bending without buckling, kinking
`permanent deformation. Fig. 9 shows a 1.5 mm
`biopsy
`forceps that consists of a thin wall nitinol tubing with a
`are able to
`nitinol actuator wire inside. Together they
`
` Fig. 9. Nitinol
`
`tubing with an internal actuation wire allows this
`0.8mm diameter grasper to operate while tied in a knot.
`
`

`

`154
`
`T. Duerig et al.; Materials Science and Engineering A273-275 (1999) 149-160
`
`be bent around radii of less than 3 cm without kinking,
`and
`and still allow
`closing of the distal grasper
`opening
`without increased resistance. This instrument con-
`jaws
`even while bent around
`tinues to operate smoothly
`tortuous
`paths. It should be pointed out, however, that
`the wall thickness of a nitinol tube stressed in bending
`should be at least 10% of the outer diameter to with-
`stand buckling [8].
`
`5. Biocompatibility
`
`‘biocompatibility’ may be simply defined as
`The term
`the ability of a material to be accepted by the body.
`a
`Since all materials generate
`‘foreign body reaction’
`in the body,
`the degree of biocompati-
`when implanted
`bility is related to the extent of this reaction. Therefore,
`is directly related to the corrosion
`biocompatibility
`behavior of the material in a
`solution and the
`specified
`to release potential toxic ions.
`tendency for the alloy
`Literature reviews generally indicate that nitinol has
`extremely good biocompatibility, due to the formation
`of a
`passive titanium-oxide layer (TiO,) [9-12] similar
`to that found on Ti alloys [13]. This finding corrobo-
`rates basic thermodynamics data that specify that the
`free energy of formation of TiO, is favored over other
`nickel or titanium oxides [12]. This oxide layer
`serves
`two purposes:
`e Increases the stability of the surface layers by pro-
`tecting the bulk material from corrosion.
`e@ Creates a
`and chemical barrier against Ni
`physical
`oxidation and modifies the oxidation pathways of Ni
`{14}.
`Several comparative studies have shown that in simu-
`lated physiological solutions NiTi is more resistant to
`chemical breakdown than 316L stainless steel, but less
`so than Ti-6Al—4V. Fig. 10 compares typical
`anodic
`curves for the three materials (from Ref.
`polarization
`More recently, Trepanier et al.
`[10] investigated the
`[9]).
`treat-
`effects of electropolishing and
`heat
`subsequent
`ments of NiTi stents on their corrosion resistance in
`Hank’s physiological solution at 37°C. Fig.
`11 shows
`curves of these stents and illustrates
`anodic polarization
`or
`that the
`electropolished and chemi-
`electropolished
`stents have the
`highest corrosion resis-
`cally passivated
`tance. These two treatments were shown to promote
`a
`thin and very uniform Ti-based oxide layer. Therefore,
`it appears that uniformity of, rather than thickness of,
`the oxide is most
`to protect
`the material
`important
`from corrosion.
`During in vitro dissolution studies in saliva, Barret et
`[15] and Bishara et al.
`al.
`found that NiTi
`[16]
`appli-
`ances released an average of 13.05 ug day~' Ni, which
`is significantly below the estimated average dietary in-
`In addition, orthodontic
`take of 200-300 tg day~'.
`patients with NiTi appliances had the Ni-concentration
`
`
`
`
`TiGAI4V
`
`’
`
`ét'1
`1t1
`t'
`
`(V)
`Potential
`
`0.2
`
`
`
`“0.2
`Current Density (10*A/em?)
`ial
`
`compared for three common
`Fig. 10. Potentiodynamic results are
`to be
`medical materials in 37°C Hanks solution, showing NiTi
`roughly between Ti-6-4 and 316L. These results have been confirmed
`by many independent researchers,
`
`a
`
`of 5 months
`in their blood measured during
`period
`The results show no
`increase in the
`significant
`[16].
`nickel blood level throughout this study. Furthermore,
`on
`an in vivo study
`et al.
`[17] performed
`Trepanier
`NiTi stents.
`Implantation of the material in
`passivated
`muscles and study of the inflam-
`rabbit paravertebral
`periods ranging from 3 to 12 weeks
`matory reaction for
`demonstrated good biological response to NiTi. Analy-
`sis of the fibrous capsule surrounding NiTi stents re-
`vealed a decrease of the thickness as a function of time.
`Fig. 12 illustrates the typical fibrous capsule surround-
`ing the implants after 12 weeks.
`These laboratory investigations confirm the clinical
`reports from Japan, Germany, China and Russia dating
`
`(mVvsSCE)
`Potential
`
`1000
`
`800
`
`600
`
`S
`
` 0.001
`
`0.1
`
`10
`
`1000
`
`Current Density (mA/cm2)
`Fig. 11. Anodic polarization testing in 37°C Hank’s solution com-
`pares several different surface treatments of nitinol: NT, natural
`processing oxide; EP, electropolished; AA, electropolished and aged
`in air to return
`light blue oxide; and PA, electropolished and passi-
`vated,
`
`

`

`T. Duerig et al. / Materials Science and Engineering A273-275 (1999) 149-160
`
`155
`
`Nitinol archwires were introduced tn the late
`less
`pain.
`1970's. We estimate that over 30%of the archwires
`are nitinol.
`used today
`Superelastic eyeglass frames provide another example
`These are nowavailable at
`of this property [21].
`nearly
`and are the most
`of all frames
`every optician,
`popular
`sold in the USA and Europe, despite the fact that they
`in the top 5%, These frames can be twisted a
`are
`priced
`full 180° without permanent deformation, but more
`the head with a
`importantly the frames press against
`constant and comfortable stress. Not only is ‘fit’
`less
`but small bends and twists that may develop
`important,
`do not cause discomfort to the wearer. It should be
`noted that this particular product is a very demanding
`product in terms of manufacturing technology, requir-
`or
`ing dissimilar material welding
`brazing, and sophisti-
`cated plating technologies. These technologies did not
`to the need of the
`exist prior
`eyeglass frame application,
`and were all developed by the frame industry.
`
`7. Biomechanical compatibility
`
`titanium and other metals are very
`Stainless steel,
`biological materials, yielding little if at
`stiff relative to
`all in response to pressure from the surrounding tissue.
`The extraordinary compliance of nitinol clearly makes
`the metal most
`mechanically similar to
`it
`biological
`even the stress—strain hys-
`materials (Fig. 13). In fact,
`teresis that is so
`to
`metallurgy is commonplace
`foreign
`with
`materials [22]. Though nitinol
`is
`the
`biological
`to the world of metallurgy,
`with respect
`exception
`stainless steel is the misfit in the world of biology. This
`improved physiological similarity promotes bone in-
`
`Stress Strain
`
`Fig, 13. The stress—strain curves of several natural biological materi-
`on those of stainless steel and nitinol. The close
`als are
`superimiposed
`similarity of nitinol to natural materials leads to more
`rapid healing
`times and less trauma to surrounding tissue.
`
` Fig. 12. A nitinol stent strut is shown in cross section 12 weeks after
`
`back to the early 1980’s. Perhaps the longest and most
`in use
`to the dental implant,
`extensive history pertains
`the
`in Japan since the early 1980’s [18]. In the USA,
`FDA has approved several nitinol class III implants,
`among them the Simon vena cava filter and the
`stents. The FDA has
`SMART, radius and
`symphony
`also
`the Mitek bone anchor system, another
`approved
`nitinol device
`permanently implanted
`[19].
`
`implantation.
`
`6. Constant stress
`
`Another
`feature of superelastic materials is
`important
`stresses over
`that
`they exhibit constant unloading
`large
`a
`device
`strains. Thus, the force applied by
`superelastic
`not strain as in conven-
`is determined by temperature,
`tional materials. Since body temperature is substantially
`constant, one can
`a device that applies
`a con-
`design
`stant stress over a wide range of shapes.
`to use
`The orthodontic archwire was the first product
`this property [20]. Stainless steel and other conventional
`or-
`appliances require adjustment by the attending
`thodontist, often to the point of causing pain. As
`the teeth move and the forces
`treatment continues,
`applied by the appliances quickly relax, which retards
`the correction process. Re-tightening by the orthodon-
`a narrow
`list recycles the process, with only
`optimum-
`period. In contrast, Nitinol wires are able to
`treatment
`a constant force over a
`move with the teeth, applying
`very broad treatment time and tooth position.
`Different
`grades of wire stiffness are available allowing the or-
`the treatment stress and be sure
`thodontist to
`‘program’
`treatment will continue properly with fewer visits and
`
`

`

`156
`
`T. Duerig et al, Materials Science and Engineering A273-275 (1999) 149-160
`
`growth and proper healing by sharing loads with the
`surrounding tissue. A large number of orthopedic
`devices take vantage of this property,
`including hip
`and
`implants, bone spacers, bone
`staples, skull plates
`the like.
`in
`This latter application is
`particularly interesting
`[23], which further lever-
`that it utilizes porous nitinol
`ages the above advantages, particularly bone in-growth,
`or
`using the heat of fusion to
`Combustion synthesis,
`the formation of the NiTi compound from
`‘ignite’
`nickel and titanium, has been shown to be an effective
`a
`way to
`porous ‘sponge’ ofnitinol, with
`produce
`densities from 40—90%. The sponge maintains supere-
`lastic and shape memory properties, has a reduced
`modulus of elasticity, accelerates bone in-growth and
`adhesion to
`has
`surrounding tissue. The ap-
`improved
`plication of these particular devices was
`in
`pioneered
`Russia, and warrants a
`good deal more attention than
`it has thus far received in the USA.
`The concept of physiological compatibility also plays
`important role in stents. Vessels are
`generally rather
`on the other hand,
`are
`tortuous;
`angioplasty balloons,
`and straight when fully inflated,
`hard, noncompliant
`often to pressures in excess of 15 atmospheres. Stainless
`steel stents are thus invariably deployed in a
`straight
`to be straight. This
`configuration, forcing the vessel
`leads to
`high bending stresses, and potential restenosis
`problems, Nitinol stents, on the other hand, are much
`more
`compliant and will contour themselves to the
`stresses. To
`vessel wall while minimizing these bending
`we should say that
`this stent
`be
`property,
`thorough,
`called ‘contourability’ is largely design related, butstill,
`an
`important role.
`materials do play
`
`an
`
`8.
`
`Dynamic interference
`
`nature
`a
`
`Dynamic interference refers to the long-range
`of nitinol stresses. To illustrate this, we
`compare
`self-expanding nitinol stent with a balloon
`expanded
`stainless steel stent.
`the
`Following balloon
`expansion,
`balloon is deflated, causing the stent to
`back’
`‘spring
`towards its smaller, undeformed
`shape. This spring-
`back, or
`loosening, is called acute recoil and is a
`highly
`undesirable feature. In order to fill a 5 mm lumen, a
`stainless steel stent
`to 6.0
`might have to be expanded
`mm so thatit is certain it will
`spring back to at least the
`5 mm lumen diameter. This
`over-expansion may dam-
`age the vessel and cause restenosis. Moreover,
`if the
`vessel diameter relaxes with time, or
`a tem-
`undergoes
`a stainless steel stent will not follow the
`porary spasm,
`vessel wall. The interference stresses will be reduced and
`the stent could even embolize.
`expansion forces in a nitinol stent are
`In contrast, the
`of a
`nature. The stent is oversized in the
`long-range
`an outward force until it
`vessel and continues to
`
`apply
`
` Fig. 14. The product performance cycle of a stent is portrayed on a
`
`typical superelastic stress—strain curve. The x-axis represents the
`the uncon-
`interference of the stent diameter with the vessel
`(e.g.
`set diameter of 10 mm,
`strained state, to the left, represents the shape
`1 mm ofinterference a diameter of 9 mm, and so on), The stent is
`constrained in a catheter, following the curve from A to B,
`then
`unloaded until a stress equilibrium is achieved at pomt C. Loading
`and unloading (D and E) exhibit dramatically different compliances.
`
`diameter. The nitinol
`reaches its
`fully
`preprogrammed
`cross
`will also try to fill an
`or
`oblong
`irregularly shaped
`section, and dynamically apply force during changes in
`cross sectional shape.
`
`9, Hysteresis
`
`superelastic hysteresis of nitinol has long been
`The
`considered a drawback because it reduces the energy
`a device
`5 J for deforma-
`storage efficiency:
`requiring
`return 2 J of mechanical energy upon
`tion, may only
`is a desirable feature in stent
`unloading. This hysteresis
`a very
`stent should provide only
`design. A superelastic
`a vessel wall,
`light chronic outward force (COF) against
`—
`and at the same time be highly resistant to
`crushing
`in one direction, and stiff in the other. This
`compliant
`is a very important feature in stent
`Nitinol
`design.
`a
`offers both a very low, dynamic outward force, yet
`very high radial resistive force (RRF).
`This is exemplified in Fig. 14. The stent is manufac-
`tured in the expanded diameter (A), then compressed
`into a catheter (B). Note this is simplified in that it is
`unlikely that the actual compression would be done at
`body temperature; the actual path from A to B would
`a martensitic deforma-
`follow a different curve, possibly
`tion. Once in the body, the stent would be deployed,
`the unloading path from B to C, at which
`following
`the stent makes contact with the vessel wall. Most
`point
`interesting is that point C exhibits a biased stiffness:
`to D),
`while the COFsfollow the unloading plateau (C