Accepted Manuscript Long-term in vivo corrosion behavior, biocompatibility and bioresorption mechanism of a bioresorbable nitrided iron scaffold Wenjiao Lin, Li Qin, Haiping Qi, Deyuan Zhang, Gui Zhang, Runlin Gao, Hong Qiu, Ying Xia, Ping Cao, Xiang Wang, Wei Zheng PII: DOI: Reference:
S1742-7061(17)30182-4 http://dx.doi.org/10.1016/j.actbio.2017.03.020 ACTBIO 4788
To appear in:
Received Date: Revised Date: Accepted Date:
25 November 2016 11 February 2017 13 March 2017
Please cite this article as: Lin, W., Qin, L., Qi, H., Zhang, D., Zhang, G., Gao, R., Qiu, H., Xia, Y., Cao, P., Wang, X., Zheng, W., Long-term in vivo corrosion behavior, biocompatibility and bioresorption mechanism of a bioresorbable nitrided iron scaffold, Acta Biomaterialia (2017), doi: http://dx.doi.org/10.1016/j.actbio.2017.03.020
This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
Long-term in vivo corrosion behavior, biocompatibility and bioresorption mechanism of a bioresorbable nitrided iron scaffold Wenjiao Lina, b, Li Qinb, Haiping Qib, Deyuan Zhangb*, Gui Zhangb, Runlin Gaoc, Hong Qiuc, Ying Xiab, Ping Caod, Xiang Wanga, Wei Zhenga* a
College of Materials Science and Chemical Engineering, Harbin Engineering University, Harbin
150001, China b
R&D Center, Lifetech Scientific (Shenzhen) Co Ltd, Shenzhen 518057, China
Department of Cardiology, Fuwai Hospital, Beijing 100037, China
Shenzhen Testing Center of Medical Devices, Shenzhen 518057, China
*Correpsonding authors, Please send all correspondance to Prof. W. Zheng, Center for Biomedical Materials and Engineering Harbin Engineering University No.145, Na-Tong-Da Street, Nan-Gang District Harbin 150001 China Tel&Fax: 0086-451-8251 8644 E-mail: [email protected]
Dr Prof Deyuan Zhang, Research & Development Center Lifetech Scientific (Shenzhen) Co Ltd Cybio Electronic Building, Langshan 2nd Street, North Area of High-tech Park, Nanshan District Shenzhen 518057 China Tel: 0086-755-8602 6224; Fax: 0086-755-86026251 E-mail: [email protected]
Abstract Pure iron as a potential bioresorbable material for bioresorbable coronary scaffold has major disadvantages of slow corrosion and bioresorption. However, so far, there are neither quantitative data of long-term in vivo corrosion nor direct experimental evidence for bioresorption of pure iron and its alloys, which are fundamental and vital for developing novel Fe-based alloys overcoming the intrinsic drawbacks of pure iron. This work systemically investigated scaffold performance, long-term in vivo corrosion behavior and biocompatibility of a nitrided iron coronary scaffold and explored its bioresorption mechanism. It was found that the 70µm Fe-based scaffold was superior to a state of the art Co-Cr alloy stent (Xience PrimeTM) in terms of crossing profile, recoil and radial strength. Mass loss was 76.0 ± 8.5 wt.% for the nitrided iron scaffold and 44.2 ± 11.4 wt.% for the pure iron scaffold after 36 months implantation in rabbit abdominal aorta (p<0.05). The Fe-based scaffold showed good long-term biocompatibility in both rabbit and porcine model. Its insoluble corrosion products were demonstrated biosafe and could be cleared away by macrophages from in situ to adventitia to be indiscernible by Micro Computed Tomography and probably finally enter the lymphatics and travel to lymph nodes after 53 months implantion in porcine coronary artery. The results indicate that the nitrided iron scaffold with further improvements shall be promising for coronary application. Keywords: Fe-based alloy; Scaffold performance; In vivo corrosion profile; Biocompatibility; Bioresorption mechanism.
1. Introduction Bioresorbable scaffold (BRS) is undoubtedly the latest invention after percutaneous transluminal coronary angioplasty (PTCA), bare metallic stent (BMS) and drug-eluting stent (DES) in endovascular intervention. Several devices have been CE marked for sale, including the DESolve 100 (Elixir Medical) in 2014, Absorb GT1 Bioresorbable Vascular Scaffold System (Abbott Vascular) in 2015 and Magmaris (Biotronik) formerly known as DREAMS-2 in 2016. The former two are poly-L-lactide (PLLA) bioresorbable scaffold while the latter is the first clinically proven magnesium scaffold. Absorb GT1 also won US Food and Drug Administration (FDA) approval in 2016. At current, an ideal BRS should have comparable or superior device performance to the state of the art drug-eluting stent (DES), good handleability, reasonable corrosion and bioresorption period, safety and efficacy demonstrated to be non-inferior to the current DES in large-scale randomized clinical trials [1-2]. Despite some initial encouraging results on the performance of BRS especially the Absorb GT1, there are many limitations for these representatives of the first-generation BRS. Poor trackability and deliverability of BRS due to thick struts and high profile discount their handleability and efficacy in complex lesions [3-5]. Post dilation is needed for BRS, however, over-dilatation might lead to strut fracture for the current generation BRS [6-7], which are vulnerable to develop in-scaffold restenosis by reports from ABSORB B trial and the GHOST-EU registry [8-9]. Recent registries have challenged the initial claim that Absorb scaffold is immune from scaffold thrombosis (ST) , both early scaffold thrombosis [10-11] and very late scaffold thrombosis [12-13] have been reported. Degradation and bioresorption period of polymer-based BRS (approximately 4 years) [14-15] is not as short as expected. For Mg-based BRS, amorphous calcium phosphate is left in the strut footprint after complete resorption of magnesium in 3
about 2 years, and whether it will be ultimately resorbed or not is still pending long-term results [2,16-17]. Based on the above, these novel devices are still immature to some extent, further refinements need to be performed and new devices need to be developed before BRS truly become a game-changer in the everyday practice in the catheter lab . Pure iron, as candidate material for BRS, has mechanical strength inferior to 316L stainless steel and Co-Cr alloy (traditional materials for fabricating permanent stents), long corrosion period and slow clearance of corrosion products [18-20]. However, iron has been demonstrated safe in vascular applications [21-24] up to 18 months. Theoretically, it is easy to strengthen pure iron, and corrosion period of pure iron scaffold can be shortened by increasing corrosion rate of material itself and/or decreasing scaffold mass or volume , but it is more important to obtain adequate biodegradation-strength-ductility balance . Results of the past studies show that it is difficult to achieve high corrosion rate and good comprehensive mechanical performance without deteriorating biocompatibility. For example, corrosion rate and material strength of pure iron can be increased by alloying [27-29], but cytocompatibility deteriorates due to toxicity of introduced metallic elements, especially Mn [30-33]. Fe-Mn alloys with very low Mn concentrations exhibit good mechanical features and biocompatibility but show no significant corrosion after 9 months implantation . Electroformed or cross-rolling iron exhibited good biocompatbility, increased corrosion rates and/or high strength but reduced plasticity [35-37]. There are numerous considerations from material to device . However, new hope has been lightened by our recently published study , which demonstrated that a newly developed sirolimus-eluting nitrided iron scaffold with a protection layer of electrodeposited pure zinc had high strength and plasticity and completely corroded after 13 months implantation in a rabbit model. 4
Unlike Fe alloying with Mn, this Fe-0.07N alloy would not deteriorate the good biocompatibility of pure iron. A lot of studies have investigated in vitro corrosion behaviors of newly designed or modified Fe-based materials [40-43], however, few studies have investigated in vivo corrosion behaviors and mechanical performance evolution with corrosion in a systematical and quantitative way. The reported longest follow-up was 18 months by Peuster M et al  in a qualitative way. Materials developed for BRS should be tested in vivo because in vitro tests do not completely simulate the in vivo physiological environment. A recent report demonstrated differences between in vivo and in vitro corrosion behaviors of Fe-Mn alloys with low Mn concentrations (maximum Mn 6.9 wt %) . Moreover, no in vitro and in vivo correlation (IVIVC) concerning corrosion is established to facilitate a quick corrosion assessment of newly designed or modified Fe-based alloys. IVIVC plays a very important role in screening newly developed materials and quality monitoring of devices in vitro for purposes of animal protection and saving time since in vivo investigation generally lasts for months or years. More importantly, although iron is not favored for extremely slow bioresorpion, actually so far, there are no any data concerning in vivo bioresorption of iron. Besides the clinically invasive and non-invasive methods , more methods for evaluating corrosion and bioresorption of Fe-based alloys remain to be established. All these are fundamental and vital for developing novel bioresorbable Fe-based alloy scaffolds. In this study, device performance of a nitrided iron scaffold (Fe alloyed with 0.074wt.%N)  was evaluated and compared with a pure iron scaffold, a state of the art Co-Cr stent (Xience PrimeTM, Abbott), PLLA-based Absorb GT1 (Abbott) and Mg-based Magmaris (Biotronik) bioresorbable 5
scaffolds. Corrosion test performed under simulated blood flow condition was designed to assess the in vitro corrosion profile of the nitrided iron scaffold. The in vivo corrosion profile of the nitrided iron scaffold in terms of mass loss and radial strength evolution was investigated up to 36 months implantation in a rabbit model with pure iron scaffold as control. The composition and distribution of corrosion products of the nitrided iron were characterized by scanning electron microscope (SEM) and Energy dispersive spectroscopy (EDS). In vivo thrombus risk of the nitrided iron scaffold was evaluated by observing endothelialization and acute thrombosis in rabbit model. Moreover, histopathological observation was conducted to investigate the local tissue response to the implanted nitrided iron scaffold with 316L SS stent having the same design as control. Finally, in a porcine coronary artery model, long-term biosafty and bioresorption of the insoluble corrosion products of the nitrided iron scaffold were evaluated after 33 and 53 months implantation by micro-computed tomography (Micro-CT), transimission electronic microscope (TEM) and histopathological analyses. This porcine model study was initiated in the beginning of year 2012 to investigate the long-term biosafety and bioresorption of nirided iron. 2. Materials and Methods 2.1. Materials The composition and microstructure of pure iron and nitrided iron (Fe alloyed with 0.074wt.%N) scaffolds were reported in our previous study . Pure iron scaffolds, nitrided iron scaffolds (Φ3.0 ×18 mm, thickness ~70 µm, weight ~12 mg), 316L SS stents (Φ3.0 ×18 mm, the same design except for strut thickness of 90 µm) were all manufactured by Lifetech Scientific Co., Ltd. (Shenzhen, China). Vacuum plasma nitriding for 2h at 500℃ with 50Pa pressure (N2:H2=1:3) was applied to obtain the nitrided iron 6
scaffolds and plates using an in-house designed vacuum nitriding furnace. All the scaffolds/stents used in tests and implantation experiment were electrochemically polished, then crimped onto balloon of rapid exchange catheter (Φ3.0 ×18 mm) by automatic crimping machine and finally ethylene oxide sterilized. All the plates were electrochemically polished and then ethylene oxide sterilized. A clinically widely-used polymer-coated drug-eluting Co-Cr alloy stent (Xience PrimeTM, Abbott Vascular, Santa Clara, CA, USA) was chosen as the control for device performance bench testing. 2.2. Device parameters and performance 2.2.1. Percent surface area and crossing profile Scaffold strut thickness and percent surface area were reported as design parameters in this work. Scaffold percent surface area was the percentage of the surface area of the expanded scaffold in contact with vessel to full cylindrical surface area at the expanded scaffold diameter and length @ nominal pressure. Scaffold percent surface area was calculated using Computer Aided Design (CAD) 2013 software from the design drawings or obtained from the open Instructions for Use of Xience PrimeTM stent. The crossing profile for the scaffold system was defined as the maximum diameter found in the scaffold/stent crimpped segment and tested using digital microscope (VHX-700F, KEYENCE, Japan). 2.2.2. Foreshortening and recoil Foreshortening was calculated from equation (1) as follows: Foreshortening (%) = [1-( Lengthinflated / Lengthoriginal)]
where Lengthinflated indicates the length of the scaffold/stent after inflation and Lengthoriginal indicates the original length of the scaffold/stent before inflation. Recoil was calculated from equation (2), as follows: 7
Recoil (%) = [1-(Diameterfinal/ Diameterinflated)]
where Diameterfinal indicated the outer diameter of the scaffold/stent after deflation of the delivery balloon and Diameterinflated indicated the outer diameter of the scaffold/stent while the delivery balloon was inflated. Scaffolds were deployed into a mock vessel (inner diameter 2.8±0.2mm, radial compliance 5–7% per [email protected]
bpm, Dynatec Labs Inc., Galena, MO, USA) at an inflation speed of 2atm/5s in 37 ± 2℃ water bath. The profile of the mock vessel in the scaffolded region was measured with the inflated balloon inside (under nominal pressure) after 30 s dwell time and removal of the balloon (for determination of acute recoil). The length of the implanted scaffold was measured before balloon inflation and with the pressurized balloon inside (under nominal pressure) after 30 s dwell time. All dimension measurements were carried out with digital microscope (VHX-700F, KEYENCE, Japan). 2.2.3. Radial strength and stiffness Curves of radial load vs. compressive diameter were measured at a compression rate of 0.1 mm/s using a radial strength tester (RX550-100, Machine Solution Inc., USA). The radial strength (kPa) was defined as the strength at 10% compression of the zero compression diameter (D0) and stent stiffness (kPa/mm) as the strength needed to produce unit change in the scaffold diameter during the radial compression process.
2.2.4 Maximal expansion diameter and side-branch accessability Maximal expansion diameter of a scaffold was defined as the maximal scaffold inner diameter or inflation balloon outer diameter, to which the scaffold after expansion could maintain structure integrity 8
and had no fractures. Side-branch accessability was defined as the maximal scaffold cell inner diameter or inflation balloon outer diameter, to which the scaffold cell after expansion could maintain structure integrity and had no fractures. At least five samples in each group were tested for the above items. 2.3. In vitro corrosion test The in vitro corrosion test was performed in an in-house designed apparatus with a circulating system as illustrated in Fig.1 (simulated blood flow condition). A number of silicone mock vessels (ID2.75*OD3.25mm) could be attached to this apparatus simultaneously to obtain a circular flow of phosphate buffered saline (PBS) inside, with temperature controlled at 37℃, speed at 25±5cm/s, pH at 7.4 and oxygen content at 4±0.5mg/L (simulating the coronary artery inner environment). Oxygen content must be strictly controlled, because our pre-test showed iron corrosion speed in PBS is proportional to local oxygen concentration, which is a crucial influencing factor for the corrosion of Fe-based materials in neutral liquids. The polished nitrided iron scaffolds and pure iron scaffolds (ID3.0×18mm) were weighed using an electronic balance with an accuracy of 0.01mg (ME5, Sartorius, Germany), crimped onto the balloon (OD3.0×18mm) and then deployed into the silicone mock arteries with nominal pressure (8atm) for immersion of 6,12,24 and 48h, respectively. At each time point, five samples for each group were removed after slitting the mock arteries with a blade, and ultrasonically cleaned in tartaric acid (3~5 wt.%), NaOH solution (1mol/L), deionised water, and absolute ethyl alcohol in sequence to eliminate the corrosion products, and then weighed for calculation of absolute mass losses and relative mass losses. 2.4. Implantation experiment 9
As shown in Table 1, 98 nitrided iron scaffolds, 42 pure iron scaffolds, and 16 316L SS stents (Φ3.0 ×18 mm) were deployed in the abdominal aortas of 78 adult New Zealand white rabbits, with 2 scaffolds/stents of the same material implanted in each rabbit. Initial mass was recorded for every nitrided/pure iron scaffold before implantation. The right femoral artery was surgically exposed and a 5F guide catheter was introduced over a 0.014 inch guidewire. Then, scaffold/stent was introduced and positioned in abdominal aorta of rabbit. RX balloon catheters (Φ3.0 ×18 mm, Lifetech Scientific, Shenzhen, China) were inflated with 8 atm (nominal pressure)-10 atm for 30s to deploy the scaffolds/stents. Placing the scaffolds/stents across the orifice of major branches of the descending aorta was avoided. Selectively, 39 male rabbits and 39 female rabbits (mean weight 2.0kg, range 1.6-2.4kg) purchased from Pearl Laboratory Animal Science & Technology Co., Ltd were fed with a standard diet without cholesterol or lipid supplementation throughout the experiment. Besides, 8 nitrided iron scaffolds automatically crimped on the rapid exchange balloon catheters (Φ3.0 ×18 mm, Lifetech Scientific, Shenzhen, China) were implanted into 4 healthy Tibet minipigs (mean weight 30kg, range 25-35kg) through the right femoral artery using 5F guiding catherter. Every minipig had one scaffold implanted in the left anterior descending (LAD) coronary artery and the other in the right coronary artery (RCA). The left circumflex artery (LCX) was used when the LAD or RCA was not suitable for scaffold implantation. The balloon inflation pressure during implantation was selected based on the measurement of vessel diameter using quantitative coronary angiographic (QCA) to control the scaffold/artery (diameter) ratio within 1.1~1.2:1. Two minipigs with 4 scaffolds were sacrificed at the follow-up of 33 months and the other two minipigs with 4 scaffolds at 53months. The use of all experimental animals in the study was in accordance with accepted institutional 10
policies, with the New Zealand rabbits under the approval of the Ethics Committee of the Shenzhen Testing Centre of Medical Devices and the Tibet minipigs under the approval of the Institutional Animal Ethics Committee of Fuwai Hospital (Beijing). 2.4.1 In vivo corrosion test In rabbit implantation experiment, the scaffolds with vessel tissues were explanted and tested their radial strength according to Table 1, . Subsequently,
the scaffolds were immersed in NaOH
solution (1mol/L) for 12h to dissolve tissues and ultrasonically cleaned in tartaric acid (3~5 wt.%), NaOH solution (1mol/L), deionised water, and absolute ethyl alcohol in sequence to eliminate the corrosion products, and then weighed for calculation of mass loss. The pretreatment method to remove tissue and rusts had been verified to be effective and has no adverse influenc on the corrosion and mechnical performance of the Fe-based scaffolds. Another 24 nitrided iron scaffolds with tissue dissolved were also used for radial strength testing at follow up times within 12 months. According to Table 1, explanted nitrided iron scaffolds with vessle tissue were prepared to be resin-embedded sections of scaffold crossing-section for time-dependent analysis of element composition and distribution around the scaffold struts using scanning electron microscope (SEM, JSM-6510, JEOL, Japan) equipped with an energy-dispersive spectrometer (EDS, Oxford Inca Energy 350; Oxford Instruments, United Kingdom). To prevent in vitro corrosion products formed in the conventional section preparation procedure, the scaffolded vessel segments were fixed for 24-48 h at room temperature immediately after explantation. Then dehydration, vitrification, embedding, sectioning (IsoMet 5000; Buehler, USA), staining were conducted in a conventional way (Saw and Grinding method) as described by Rippstein et al . The in vivo iron corrosion products of the nitrided iron 11
scaffolds were collected after 12 months implantation in rabbit and analyzed using X-ray photoelectron spectroscopy (XPS, AXIS ULTRA DLD, SHIMADZU, Japan), optical microscope (VHX-700F, KEYENCE, Japan), magnet and Raman microscope (inVia Reflex, Renishaw, UK) for phase identification. 2.4.2. Endothelialization and histopathologic observation Sixteen finished nitrided iron scaffolds with sixteen 316L SS stents as control were implanted into the abdominal aorta of 16 adult New Zealand rabbits in a hybrid mode for evaluation of endothelialization and histopathologic observation. According to Table 1, four nitrided iron scaffolds and four 316L SS stents were explanted with vessel tissue 7 days after implantation and slit longitudinally for endothelialization observation under SEM. Gross dissection observation of the explanted scaffolded vascular segments and organs (heart, liver, spleen, lung, kidney) was conducted at 1, 4, 6 9, 12, 24 and 36 months follow-up while histopathologic observation was only conducted at 1, 6 and 12 months follow-up for evaluation of the local tissue response. Resin-embedded sections of the 316L SS stented vessel segments were prepared by conventional fixing, dehydration, embedding, sectioning and staining with haematoxylin and eosin (HE). For better observation of the local tissue response, the nitrided iron scaffolded vessel segments were placed into a solution consisting of 80 mL 100% ethanol, 10 mL 40% formaldehyde, 5 mL 99.5% glacial acetic acid, and 5 mL 65% concentrated nitric acid for 3 ~ 5 hours at room temperature. The solution volume was 5 ~ 10 times larger than the sample volume and completely immersed the samples, which were observed at 30 min intervals after 3 h immersion. Slight blanching of the vessel wall, and elastic recovery after compression and absence of metallic barrier when stabbing with a needle indicated 12
complete dissolution of the iron-based scaffold. This dissolving pretreatment method had been verified to be effective in removing the residual nitrided iron struts and the rusts around the scaffold struts and have no adverse influence on the fixed tissue/cells. Then the samples underwent dehydration with 100% ethanol for 30 min to 1 h, vitrification with xylol for 20–30 min and paraffin embedding for 30 min. Finally, 4–5 µ m paraffin-embedded sections were prepared using a rotary microtome (LEICA RM2235, Germany), followed by xylene dewaxing, gradient rehydration, neutralization in a 0.05 M sodium hydroxide solution for 5 min, rinse and HE staining. Early study showed that the media and intima are destroyed when preparing the paraffin-embedded sections after mechanically removing struts of the pure iron stent . However, using the struts dissolving method, the media was not injured or fractured. 2.4.3 Histopathology and bioresorption observation after long-term implantation Gross observation and histopathology analysis on the organs (at least including heart, liver, spleen, lung, kidney) of two minipigs at follow-up of 33 months and the other two minipigs at follow-up of 53 months were conducted and compared with those of healthy minipig. Four explanted nitrided iron scaffolds with vessel tissues at follow-up of 33 months and the other Four explanted nitrided iron scaffolds with vessel tissues at follow-up of 53 months all underwent 3D reconstruction using Micro-Computed Tomography (Micro-CT, Skyscan1172, Bruker, Germany). Then they were prepared for transmission electron microscope (TEM, H-7500, Hitachi, Japan) observation (resin-embedded sections) and histopathologic observation on bioresorption of the corrosion products (resin-embedded and paraffin-embedded sections). 2.5. Statistical methods
Normality and homogeneity of variance of the experimental data in the current work were examined firstly before the authors chose suitable statistical methods to analyze and compare data of various groups. Results were expressed as mean ± standard deviation. An ANOVA with LSD post-hoc tests were conducted for the mechanical performance tests. A two factor ANOVA with LSD post-hoc tests were conducted for the corrosion performance tests. Minitab 16 software was used and statistical significance was defined as P≤0.05. 3. Results 3.1. Device parameters and performances Table 2 presents the device parameters and performance of a nitrided iron scaffold compared with a pure iron scaffold, a Co-Cr alloy stent - Xience PrimeTM, Absorb GT1 poly(l-lactide)-based (PLLA-based) and Magmaris Mg-based bioresorbable scaffolds (BRSs). The performance outcomes for Absorb GT1 and Magmaris were cited from reference, which were obtained under the same testing methods. It could be known from Table 2 that the nitrided iron scaffold has a percent surface area nearly the same as Xience Prime, much lower than those of Absorb GT1 and Magmaris, and has struts much thinner than those of Xience Prime, Absorb GT1 and Magmaris. A significantly lower crossing profile of 0.99±0.02 mm for the nitrided iron scaffold system means better crossability in small arteries and narrowed lesion locations when compared with Xience PrimeTM stent (p<0.05), Absorb GT1 BRS (p<0.05) and Magmaris BRS (p<0.05). According to equations (1) and (2), the calculated foreshortening and recoil values are listed in Table 1. There is no significant difference in foreshortening and recoil between the nitrided iron scaffold 14
and the pure iron scaffold (p>0.05), however, elastic recoil of the nitrided iron scaffold are significantly lower than that of Xience PrimeTM stent (p<0.05), Absorb (p<0.05) and Magmaris (p<0.05). Fig.2 shows typical curves of radial strength vs. scaffold diameter, where the linear segment indicates recoverable deformation. Scaffold stiffness (kPa/mm), defined as the strength needed to produce unit change in the scaffold diameter during the radial compression process, is exactly the slope of the linear segment. The intersection of the linear segment and horizontal axis is defined as the scaffold’s original diameter D0. The nitrided iron scaffold with 70 µm strut shows significantly elevated radial strength and stiffness due to nitriding, compared with both pure iron scaffold (p<0.01) and Xience PrimeTM stent (p<0.01). It can be seen from Fig. 3 that the nitrided iron scaffold maintains high plasticity to tolerate severe plastic deformation in stress concentration positions, e.g. the bottom areas of the omega-shaped links and the inner sides of the peaks of the rings with many obvious slip bands. The maximal expansion diameter and side-branch accessability illustrated in Fig.4 for the 3.0×18mm nitrided iron scaffold is 4.4mm and 2.4mm, respectively, as presented in Table 1, the same as those of pure iron scaffold and Xience PrimeTM stent and higher than that of Absorb GT1 scaffold. To ensure a side branch accessible, a cell of the scaffold/stent is commonly expanded to design limit in clinical application. 3.2. In vitro and in vivo corrosion profiles In vitro corrosion profiles of the nitrided iron scaffold and the pure iron scaffold deployed in mock vessel are shown in Fig. 5(a). The in vitro corrosion rate of the scaffold has been significantly increased by nitriding (p<0.05). The mass loss of the nitrided iron scaffold is up to about 45wt.% after approximately 20h immersion and 80wt.% or so after only 48h corrosion in mock vessel with circular 15
flow of PBS. Moreover, under a temperature, flow speed, oxygen content and pH controlled environment, it was found that considerable solid corrosion products covered the surfaces of both scaffolds tightly and were not easily be flushed away by the circular flowing of PBS. In vivo corrosion causes decrease in the radial strength and mass loss for both scaffolds. Both of the nitrided iron scaffold and the pure iron scaffold featured non-uniform corrosion by macrography. The in vivo corrosion profile of the nitrided iron scaffold is shown in Fig. 5(b) up to 36 months after implantation in the rabbit abdominal aorta with the pure iron scaffold as control. Both of the two mass loss curves show gradually decreasing slope, indicating gradually slowing corrosion with time. The nitrided iron scaffold has a significantly higher mass loss than the pure iron stent (p<0.05) (44.5 ± 6.4 wt.% vs 24.0± 5.6 wt.% at 12 months; 62.9 ± 9.0 wt.% vs 35.0± 8.3 wt.% at 24 months; 76.0 ± 8.5 wt.% vs 44.2± 11.4 wt.% at 36 months;), indicating a higher corrosion rate of the nitrided iron scaffold. As shown in Fig. 5(c), the radial strength of the nitrided iron scaffold with tissue is higher than that of the sample without tissue, which is significant (p<0.05) after implantation of 6 months or longer. The scaffolding strength of the nitrided iron scaffold with tissue is still around 150kPa after 6 months implantation and above 120kPa after 9 months implantation. After disintegration of scaffold, it is not suitable to continue to test the radial strength of the scaffolds with tissue, because the radial strength will not drop to zero due to the existence of tissue. 3.3. In vivo corrosion products identification Fig. 6(a) reveals the wide scanning X-ray Photoelectron Spectroscopy (XPS) spectrum of in vivo corrosion products around the struts of the nitrided iron scaffold after 12 months implantation in rabbit abdominal aorta. It indicates that there are at least elements of Fe, O, Ca and P in the corrosion products 16
of the nitrided iron scaffold. The strong signals of elements C and O are greatly contributed by tissues and resins of slice. Gaussian fitting of the Fe2p peak shown in Fig. 6(b) demonstrates that iron in the corrosion products exists in the valence state of Fe2+ and Fe3+, whose binding energy are 709.8eV and 712.31eV, respectively. Fig. 6(c) presents the representative photo of the nitrided iron scaffold after 12 months implantation in rabbit abdominal aorta. The innermost black corrosion products near the scaffold struts are magnetic while the outer yellowish-brown corrosion products is nonmagnetic as shown in Fig. 6(d), (e). Raman spectrum of the separated in vivo corrosion products is shown in Fig. 6(f). The characteristic peaks indicate existence of γ-FeOOH (1298, 384 cm-1), Fe3O4 (675, 534 cm-1) and α-Fe2O3 (238, 296, 496 and 1324cm-1). The in vivo corrosion evolution, corrosion products composition and distribution were illustrated in Fig. 7 by the strut cross sections and its corresponding element distributions of the nitrided iron scaffold after implantation of 1, 4, 6, 9 and 12 months in rabbit abdominal aorta, respectively. The struts are found to be surrounded by a layer of corrosion product (iron oxides or hydroxides) whereas the outermost is a layer of corrosion products containing Ca and P. The corrosion products increase and accumulate extensively in the vicinity of the struts with ongoing corrosion. However, there is a tendency of corrosion products deaggregation and dispersion, and tissue regeneration within the original strut footprint could be observed 9-12 months after implantation, since the carbon content is high in the gap among the scattered corrosion products. 3.4. Endothelialization In rabbit abdominal aorta model, the explanted scaffoled vessel segments with the nitrided iron scaffold and the 316L SS stent after 7 days implantation were examined to determine the extent of 17
neointimal coverage. As is shown in Fig. 8, squamous endothelial cells (ECs) align on the surface of the nitrided iron scaffold and complete and homogeneous endothelial coverage has been observed 7 days after implantation, suggesting very low thrombosis risk. Whereas for the 316L SS stent, incomplete endothelial coverage is observed, and fibrin clusters or giant cells could be found over struts not covered by endothelial cells. 3.5. Local tissue response All rabbits survived the follow-up period (the longest one - 36 months) without any adverse events. Gross dissection observation for the rabbits revealed that there were no abnormalities found for the organs and no pathologic changes. None of the animals showed changes in the abdominal aorta indicative of aneurysm formation by gross observation. As shown in Fig. 9, the local tissue shows consistent slight inflammatory response to the nitrided iron scaffold after 1 month, 6 month and 12 months implantation, similar to that of 316L SS stent. In addition, no tissue necrosis of the neointima or its underlying layers was found in the follow-ups by histopathological observation. Macrophages engulfing corrosion particles around the strut have been found for the nitrided iron scaffold after 12 months implantation. All porcines survived the follow-up period (the longest one - 53 months) without any adverse events. Gross dissection observation revealed no abnormalities found for the organs and no pathologic changes. None of the porcines showed coronary aneurysm formation by gross observation. As shown Fig. 11, histomorphology is normal for the five main organs including (a) heart, (b) liver, (c) spleen, (d) lung and (e) kidney of porcine with nitrided iron scaffold after 53 months implantation in the coronary artery. And there are no pathological changes when compared to the histomorphology of (f) heart, (g) 18
liver, (h) spleen, (i) lung, (j) kidney of healthy porcine without scaffold implanted. Fig. 12(a) shows the Micro-CT 2D image of one of the nitrided iron scaffold after 53 months implantation in a porcine coronary, revealing non-uniform corrosion and bioresorption. Determining from the radiopacity contrast, Fig. 12(b) represents areas with most majority of struts corroded and corrosion products remaining in situ, while Fig. 12(c) represents areas with large area of disappearing struts and corrosion products and few incompletely corroded struts. Histopathological observation of resin-embedded sections from area B (Fig. 12(d)) and area similar to A (Fig. 12(e)) demonstrates that corrosion products disappearing under Micro-CT are actually existent in histopathology however in a different morphology - numerous small aggregations which are cleared away from in situ. The expansive corrosion products in situ cause the dimension increase in the direction of strut thickness from the initial 70 µm to about 200 µm after complete corrosion. Moreover, observation of paraffin-embedded sections of the nitrided iron scaffolded vessel segment after 53 months implantation shows a moving tendency of the corrosion products from in situ and peri-strut areas (Fig. 12(f)) to adventitia (Fig. 12(g)), and finally tissue regeneration within the original strut footprints has been observed (Fig. 12(g) and Fig. 12(h)). The vessel indicated by red arrow in Fig. 12(f) is speculated to be a lymphatic or a capillary vein and the one indicated by asterisk may be a lymph trunk or a capillary artery. It is worth noting that the paraffin-embedded sections were prepared after dissolving the metallic strut in acid solution, which means that the corrosion products of the nitrided iron scaffold would also be dissolved if they were dispersed between the cells instead of within the cells. However, the amount and location of the numerous yellow-brown corrosion products indicated by the paraffin-embedded sections are comparable to those indicated by the resin-embedded sections, suggesting that the corrosion products are in cells. It 19
has been further found from the enlargement of the corrosion products in Fig. 12(i) that the aggregation of the corrosion products has a well-defined outline, a size of about 20µm similar to that of a macrophage and seems to have obscure cell nucleus within it. As shown in Fig. 12(d-h), there are no adverse local tissue responses (e.g. severe inflammation with neutrophils, monocytes, lymphcytes; hypersensitivity with eosinophils; tissue necrosis) throughout the implantation in the Tibet minipig coronary artery up to 53 months. TEM images of the nitrided iron scaffolded vessel segment after 33 months implantation in a porcine coronary show that the insoluble corrosion particles (with size level of micron or submicron) as indicated by the red triangles could move in interstitial fluid between smooth muscle cells (Fig. 12(j)) or be found in a somatic cell (Fig. 12(k)). 4. Discussion The composition and microstructure of the nitrided iron material were reported in our previous work . The mechanism of strengthening brought by nitriding is dispersion strengthening of the second phase of fine iron nitrides, which would not significantly decrease plasticity. And the mechanism of accelerating corrosion is micro-galvanic corrosion between the pure iron matrix and the dispersed fine iron nitrides . Although the alloying of N element is extremely low (0.074 wt.%), it has been confirmed to be an effective way to obtain biocorrosion-strength-plasticity balance without deteriorating biocompatibility. 4.1. Device performance New stent platforms with thinner struts and more conformable designs has been driven by clinicians' desire for more consistently deliverable stents that cause less injury, restenosis and lower thrombosis risk (due to better strut apposition and less disturbance on blood flow) . However, the 20
currently available bioresorbable scaffolds require thick strut (resulting in high crossing profile) to ensure sufficient radial support. In addition, high plasticity is indispensible for device fabrication, e.g. tube drawing (with severe plastic deformation), and also for clinical uses. For example, aggressive scaffold expansion is required clinically for optimal strut apposition. Chacko Y et al investigated how frequently stents were over-expanded and found that stents were expanded under above nominal in 99% of cases and above rated burst pressure in 52%. Stents were expanded > 20% above nominal diameter in 12% cases and in some cases even >25% above nominal diameter . Sometimes, certain scaffold cell/unit has to be expanded to its design limit to ensure side branch accessible after scaffold is deployed in the main branch vessel. In such cases, risk of scaffold fracture and subsequent in-stent restenosis is potentially great with such severe deformation . Material plasticity and scaffold design determine the over-expansion and side branch capabilities. Acute scaffold recoil after deflation of the balloon reflects the capability of the scaffold to maintain acute luminal gain achieved after adequate expansion. A study has demonstrated that the magnitude of this in vivo acute recoil is significantly greater than that reported from bench testing and varies with scaffold design . The acute scaffold recoil consists of two parts. One part is the elastic recoil after scaffold expansion, which depends on material elastic performance and scaffold structural design, and could be tested in lab. High elastic recoil is adverse for scaffold to be crimped onto the balloon, this is the reason why annealing should be applied to 316L SS and Co-Cr alloy stents, to reduce the elastic limit, however at the expense of decreasing strength. The other part is the acquired recoil to resist the compressive force from the vessel vasomotion and calcified plaque in lesions. Actually, the factors that influence this acquired recoil are scaffold radial strength and stiffness which depends on 21
material strength, elastic modulus and scaffold design. For scaffold with low radial strength and stiffness, once the external force exceeds its elastic limit (the linear segment in a typical curve of radial strength vs. diameter), unrecoverable deformation, namely acquired recoil, occurs. Consequently, bioabsorbable scaffolds made of polymer or magnesium-based alloy with inferior mechanical strength have to be designed with thick struts of 120-200 µm , to reduce acute recoil post-operation and guarantee effective scaffolding. For coronary artery application, the crossing profile, foreshortening and recoil of the scaffold/stent are desired to be as low as possible; the maximal expansion diameter and side-branch accessability of the scaffold/stent are desired to be as high as possible. Radial strength and stiffness should be adequate to support various lesion vessels and should not be too high to sacrifice flexibility, ranging from approximately 110~170 kPa using the testing method listed in the current work for those permanent stents clinically-demonstrated to be effective (Φ3.0 ×18 mm). The ideal platform therefore comprises a highly deliverable, thin-strut, low-profile design with high radiopacity, sufficient radial strength, high over-expansion diameter and minimal recoil and foreshortening . The present work has demonstrated that the 70µm nitrided iron scaffold has good comprehensive device performance superior to the state of the art Co-Cr alloy stent (Xience PrimeTM) and might also be superior to the currently available polymer-based (Absorb GT1) and Mg-based (Magmaris) bioresorbable scaffolds. 4.2. Corrosion behavior and corrosion products Our in vivo corrosion results reveal that the nitrided iron scaffold with vessel tissue shows higher radial strength when compared to those with tissue removed, which becomes statistically significant (p<0.05) especially after implantation of 6 months or longer. It is speculated that corrosion products 22
might contribute to radial strength when they are surrounded or connected by tissues. When tissue is removed, corrosion products are broken and struts become disintegrated. Another reason might be that the tissue could act as a connector for the fractured or disintegrated struts after corrosion to maintain structural integrity. Unlike in porcine coronary artery, the thickening of neointima with implantation time is rare in rabbit aorta abdominal, so it would not continuingly increase the radial strength. Considering this, it is suggested to test the in vivo scaffold samples with vessel tissue, avoiding underestimation of the radial strength. Our result has shown promise for establishing an accurate in vitro and in vivo correlation (IVIVC) (IVIVC) concerning corrosion that would be of use in rapid development and testing of newly Fe-based alloys as well as experimental animal protection. However, further work has to be conducted and the established IVIVC need to be validated and verified through repeated tests by other groups. If Fe-based materials feature uniform corrosion, the 70×90 um struts should decrease to 50×70 um at 12 months and 30×50 um at 36 months on average, corresponding to the mass losses of 44.5 wt% @12 months and 76.0 wt.% @36 months for the nitrided iron scaffold in the rabbit model,. However, metallic materials all have an intrinsic drawback of localized corrosion, therefore there were struts corroded completely into corrosion products after 12 months implantation in a rabbit model as shown in Fig. 7 while nearly intact struts also exist after 53 months implantation in a porcine model as shown in Fig. 12(a). Considering this, individual strut corrosion extent could not indicate the corrosion of the whole scaffold whereas mass or volume loss may better indicate it. It is worth noting that, the striking finding of completely uncorroded nitrided iron scaffold struts after 53 months implantation might raise concern about long-term safety. The most possible reasons for this might be that (1) the corrosion 23
resistant white layer (γ′-Fe4N) formed on the scaffold surface was not removed completely by polishing and (2) the nitriding modification processing is not uniform. Considering the adjustabale nitriding processing, the small size of these integrated struts and the good local tissue responses after 53 months implantation, the risks of these uncorroded stent struts should be controllable. Extrapolating from the in vivo corrosion profile concerning mass loss, the nitrided iron scaffold would probably completely corrode in 4~5 years, which has been confirmed by the implantation results in the porcine conronary artery model up to 53 months. Although the nitrided iron scaffold has a significantly higher corrosion rate than that of pure iron scaffold, it is not ideal when compared with polymer-based and Mg-based bioresorbable scaffolds. And under such a corrosion rate, the nitrided iron scaffold with 70 µm strut could still maintain effective scaffolding with 150kPa radial strength after 6 months implantation (vascular healing stage), which leaves room for us to further shorten the corrosion period. Based on this, the authors have successfully designed a new drug-eluting nitrided iron scaffold recently with much thinner strut (50 µm) and poly-DL-lactide (PDLLA) to shorten its corrosion period to about 1 year with significantly improved corrosion uniformity . Combining the results of iron valence state analysis, color and magnetism, phase constitution, and components distribution for the in vivo nitrided iron corrosion products, it could be speculated that the typical distribution of the in vivo iron corrosion products is magnetic Fe3O4 near remaining iron struts, unmagnetic Fe(OH)3 and/or its dehydreation products (FeOOH and Fe2O3) and/or Fe3(PO4)2 in middle area, and Ca3(PO4)2 in outmost area. This experimental conclusion could also be supported by a theoretical analysis in our previous work . Based on the calculation of Pourbaix diagrams of iron corrosion in a phosphate physiological environment (pH=7.4), among various iron corrosion products, 24
Fe(OH)3, FeOOH and Fe2O3 are relatively stable existence. Fe3(PO4)2 and Fe3O4 are formed under local anoxic condition and could also exist stably owing to their slow reaction kinetics to transform to more stable Fe(OH)3, FeOOH and Fe2O3. The formation of a Ca/P layer or “passive layer” in the outermost layer of the corrosion products area has been reported by many studies, so as in our work. Daniel Pierson et al.  reported that a Ca/P layer formed at the outer layer of pure iron wire corrosion products in the rat arterial wall. Peuster et al reported a passive layer around the corrosion products of Fe-Mn alloys implanted in NMRI mice . One concern is that the corrosion product layer containing Ca/P or passive layer may prevent the residual iron matrix from corrosion. However, a tendency of disaggregation and dispersion of corrosion products can be observed in this work, which is speculated to be the result of volume expansion with accumulation of the corrosion products. The volume of the iron corrosion products was reported to be several times larger than that of iron itself . Moreover, tissue regeneration within the original strut footprint has also been found in this work as shown in Fig. 6 and Fig. 10(d)(j). This is beneficial not only for continuing corrosion of the residual struts but also for the clearance of the corrosion products. 4.3. Bioresorption and biocompatibility In order to investigate the bioresorption of the insoluble corrosion products, porcine coronary artery was chosen for long-term implantation since it is closer to human coronary artery and minipigs could live longer than New Zealand rabbit. The determined in vivo iron corrosion products in the present work (Fe3O4, Fe2O3, FeOOH, Fe(OH)3 and Fe3(PO4)2) are hard to dissolve in neutral body fluid. All these insoluble corrosion products have solubility equilibrium in liquids and the solubility is pH-related. Iron ions could be easily absorbed 25
by organism, because of reduced concentration of iron ions, the solubility equilibrium shifts in the direction of forming iron ions. In this way, the insoluble iron corrosion products could also dissolve gradually and be absorbed by the organism. However, iron ion concentration determined by the solubility product is extremely low, especially for ferric oxides and hydroxides, e.g. the solubility product constant of Fe(OH)3 (pH 7.0, solubility at 10-17M) decreases by 16 orders of magnitude than that of the Fe(OH)2 (pH 7.0, solubility at 10-1M) . Therefore, bioresorption of ferric oxides and hydroxides by means of solubility equilibrium shift is indiscernible in short time and takes years. This is the reason of slow bioresorption for Fe-based alloys. Many studies have shown that foreign bodies could be eliminated by phagocytosis of macrophages [56-57], which are abundant in the body and could move by amoeboid motion in local tissues. Lysosome pH in macrophages were reported to be around 5, and this low pH enables the macrophages to dissolve metal particles like cobalt oxide particles [58-59]. In the present study, macrophages were found in an increasing number around the struts to engulf the insoluble iron corrosion particles. After 53 months implantation, numerous refractive yellow-brown particles could be indentified from the rounded or oval aggregations in Fig. 12(i), where the former are speculated to be hemosiderin formed after the insoluble iron corrosion particles were engulfed by the latter - macrophages. The present work has demonstrated that the hemosiderin-laden macrophages could move from the strut positions to the adventita to be indiscernible by Micro Computed Tomography (Micro-CT). It is known that coronary arteries normally run in the adipose tissue in subepicardium or in the cardiac muscle fibers, where there are many lymphatics. It is difficult to differentiate these lymph vessels from the blood vessels only by morphology. It is speculated that the hemosiderin-laden 26
macrophages after moving from the strut positions to the adventita, could then enter the lymphatics and travel to the adjacent lymph nodes to fulfill clearance of the insoluble iron corrosion products. Although small insoluble corrosion particle were found to be able to be engulfed by some somatic cells or move in interstitial fluid (intercellular path), phagocytosis by macrophages and then being carried to lymph nodes is the major bioresorption mechanism for insoluble iron corrosion products. The non-uniform corrosion might lead to complete bioresoption in 5~6 years after implantation for the nitrided iron scaffold. Tissue could regenerate within the original strut footprints after clearance of the insoluble corrosion products from in situ without significant calcium phosphate remaining 53 months after implantation in a porcine model. However, there is still calcification area for the Absorb scaffold after 4 years implantation  and complete replacement of magnesium phosphate by armorphous calcium phosphate 1 year after implantation  in a porcine model. This might be because the insoluble iron corrosion products tend to expand and disperse in tissue to facilitate complete clearance by phagocytosis of macrophages. Our newly designed 50µm PDLLA-coated nitrided iron scaffold could completely corrode in around 1 year with apparently more uniform corrosion. Based on the present work results, it could probably be completely bioresorbed in around 3 years after implantation. Endothelial coverage on the nitrided iron scaffold inner surface was more rapid and complete than that on the 316L SS stent with no local or downstream thrombus found for both scaffolds 7 days after implantation in the rabbit abdominal aorta, which indicates low in vivo thrombosis risk. There are studies revealing that the extent of endothelial coverage is dependent on stent thickness , and endothelial coverage is the greatest in stents with the thinnest strut . These results are consistent with our finding. In addition, thin strut has lower disturbance on the blood flow, therefore futher contributes 27
to lower risk of thrombosis . From another perspective, rapid and complete endothelialization after 7 days implantation in the rabbit abdominal aorta indicates low or no cytotoxicity for the nitrided iron material. Two previous studies [47,63] of the authors also find complete endothelialization for the nitrided iron scaffold after one month implantation in the porcine artery, demonstrating good in vivo cytocompatibility. Consistent slight inflammation in local tissues without any abnormalities in organs and scaffolded vessel segments up to 36 months implantation in rabbit, indicates good biocompatibility of the nitrided iron scaffold. Potential biotoxicity of insoluble iron corrosion products include irritation and toxicity to the surrounding local tissue and systemic toxicity to the individual animal. Based on the previous discussion, the insoluble corrosin products (mainly Fe3O4, Fe2O3, FeOOH, Fe(OH)3 and Fe3(PO4)2) are biostable in the neutral physiological environment by thermodynamics and kinetics analysis. A previous study of the authors  has demonstrated that the insoluble iron corrosion products in nature have no in vitro toxicity to the cultured endothelial cells (ECs) and smooth muscle cells (SMCs). Only when the dimensions of the corrosion products are as small as could be engulfed by the common somatic cells, in vitro toxicity were observed due to size effect. More importantly, after 53 months implantation in the porcine model when the nitrided iron scaffold has corroded severely with a large amount of insoluble corrosion products, there was only slight inflammation without tissue necrosis, hypersensitivity (with eosinophils) or any other abnormalities, indicating good long-term biocompatibility of the insoluble iron corrosion products to the local tissues. There was reporting that in a healthy adult, 1 to 2 mg of iron enters and leaves the body each day through metabolism . It is speculated from this data that the risk of systemic toxicity of the insoluble 28
corrosion products is probably also very low, since the mass of the whole nitrided iron scaffold is approximately 12mg which will be resorbed by the organism in years after complete corrosion. Gross observation and histopathologic analysis on the porcine organs (heart, liver, spleen, liver, lung) after 53 months implantation with the nitrided iron scaffold indicated no abnormalities, which might support the above speculation. Even if the insoluble iron corrosion products exist in vivo for a very long time, they are believed to be tolerated by the local tissue and the individual body, which will be investigated in the follow-ups of 6 years or longer in our future work. In addition, our future work will also focus on further improving the corrosion non-uniformity of the nitrided iron material and reducing the amount of insoluble corrosion products by way of chelating a portion of iron to form soluble substances to facilitate bioresorption. Before the nitrided iron scaffolds could be tested in human subjects, comprehensive performance should be verified by bench testing, biological evaluation should be systemically conducted in vitro and in vivo, and safety and effectiveness should be validated by porcine experiment, with statistically sufficient sample size. Study Limitations In this work, the authors only studied some crucial properties of a limited number of scaffolds/stents in bench testing. Meanwhile, bench testing does not always predict clinical performance of devices. The small sample size of the unsystematic histopathology analysis in this work could only provide a preliminary evaluation on biocompatibility and bioresorption. In addition, effectiveness of the nitrided iron scaffold evaluated by imaging methods was not included in the present work. 5. Conclusions 29
(1) Nitriding technology is an effective modification method to obtain biocorrosion-strength-plasticity balance without deteriorating biocompatibility. The 70µm nitrided iron scaffold has good comprehensive device performance superior to the state of the art Co-Cr alloy stent (Xience PrimeTM). (2) Although the nitrided iron scaffold corrodes significantly faster than pure iron scaffold, it is not ideal when compared with polymer-based and Mg-based bioresorbable scaffolds. Typical in vivo corrosion products distribution for this scaffold is magnetic Fe3O4 near the remaining iron struts, unmagnetic Fe(OH)3 and/or its dehydreation products (FeOOH and Fe2O3) and/or Fe3(PO4)2 in middle area, and Ca3(PO4)2 in outmost area. The 70 µm nitrided iron scaffold is probable to completely corrode in 4~5 years and completely bioresorbed in 5~6 years in the porcine conronary artery model with further improvement in corrosion uniformity. (3) The nitrided iron scaffold showed good long-term biocompatibility in both rabbit and porcine model. Its insoluble corrosion products were demonstrated biosafe and could be cleared away by macrophages from in situ to adventitia to be indiscernible by Micro Computed Tomography after 53 months implantation in porcine coronary artery, and probably enter the lymphatics and travel to lymph nodes finally. Tissue could regenerate within the original strut footprints after clearance of the insoluble corrosion products from in situ. In conclusion, this study demonstrates that the nitrided iron scaffold with further improvements shall be promising for coronary application as an alternative to permanent stents.
Acknowledgements This study was supported by the China National High-tech Research and Development Program 30
Research Project No. 2011AA030103 and the Municipal Science and Technology Plans of Shenzhen No. JSGG20140701161153656. The authors thank Renu Virmani for her expert assistance on histopathology analysis.
Disclosures It is declared that De-Yuan Zhang, Li Qin, Gui Zhang, Ying Xia are employees of Lifetech Scientific Corporation. References  Oberhauser JP, Hossainy S, Papoza RJ. Design principles and performance of bioresorbable polymeric vascular scaffolds. EuroIntervention 2009; 5 (Supplement F):F15-F22.  Tenekecioglu E, Farooq V, Bourantas CV, Silva RC, Onuma Y, Yılmaz M, Serruys PW. Bioresorbable scaffolds: a new paradigm in percutaneous coronary intervention. BMC Cardiovasc Disord 2016, 16:38; doi: 10.1186/s12872-016-0207-5.  Kawamoto H, Ruparelia N, Tanaka A, Chieffo A, Latib A, Colombo A. Bioresorbable scaffolds for the management of coronary bifurcation lesions. JACC Cardiovasc Interv 2016; 9:989-1000.  Azzalini L, L'Allier PL, Tanguay JF. Bioresorbable Scaffolds: The Revolution in Coronary Stenting? AIMS Med Sci 2016; 3:126-146.  Schmidt W, Behrens P, Brandt-Wunderlich C, Siewert S, Grabow N, Schmitz KP. In vitro performance investigation of bioresorbable scaffolds - standard tests for vascular stents and beyond. Cardiovasc Revasc Med 2016; 17:375-383.  Caiazzo G, Kilic ID, Fabris E, Serdoz R, Mattesini A, Foin N, De Rosa S, Indolfi C, Di Mario C. Absorb bioresorbable vascular scaffold: what have we learned after 5 years of clinical experience? Int J Cardiol 2015; 201:129–36. 31
 Ormiston JA, Webber B, Ubod B, Darremont O, Webster MW. An independent bench comparison of two bioresorbable drug-eluting coronary scaffolds (Absorb and DESolve) with a durable metallic drug-eluting stent (ML8/Xpedition).EuroIntervention 2015; 11:60-7.  Nakatani S, Onuma Y, Ishibashi Y, Muramatsu T, Iqbal J, Zhang YJ, van Geuns RJ, Ormiston JA, Serruys PW. Early (before 6 months), late (6-12 months) and very late (after 12 months) angiographic scaffold restenosis in the ABSORB Cohort B trial. EuroIntervention 2015; 10: 1288–1298.  Longo G, Granata F, Capodanno D, Ohno Y, Tamburino CI, Capranzano P, La Manna A, Francaviglia B, Gargiulo G, Tamburino C. Anatomical features and management of bioresorbable vascular scaffolds failure: A case series from the GHOST registry. Catheter Cardiovasc Interv 2015; 85: 1150–1161.  Cassese S, Byrne RA, Ndrepepa G, Kufner S, Wiebe J, Repp J, Schunkert H, Fusaro M, Kimura T, Kastrati A. Everolimus-eluting bioresorbable vascular scaffolds versus everolimus-eluting metallic stents: a meta-analysis of randomised controlled trials. The Lancet 2015; 6736: 1–8.  Lipinski MJ, Escarcega RO, Baker NC, Benn HA, Gaglia MA Jr, Torguson R, Waksman R. Scaffold thrombosis after percutaneous coronary intervention with ABSORB bioresorbable vascular scaffold: a systematic review and meta-analysis. JACC Cardiovasc Interv 2016; 9:12-24.  Azzalini L, Al-Hawwas M, L’Allier PL. Very late bioresorbable vascular scaffold thrombosis: a new clinical entity. Eurointervention 2015; 11: e1–2.  Karanasos A, van Geuns R-J, Zijlstra F, Regar E. Very late bioresorbable scaffold thrombosis after discontinuation of dual antiplatelet therapy. Eur Heart J 2014; 35: 1781.  Onuma Y, Serruys PW, Perkins LE, Okamura T, Gonzalo N, García-García HM, Regar E, Kamberi M, Powers JC, Papoza R, van Beusekom H, van der Giessen W, Virmani R. Intracoronary optical
coherence tomography and histology at 1 month and 2, 3, and 4 years after implantation of everolimus-eluting bioresorbable vascular scaffolds in a porcine coronary artery model. Circulation 2010; 122:2288-2300.  Vorpahl M, Finn AV, Nakano M, Virmani R. The bioabsorption process tissue and cellular mechanisms and outcomes. EuroIntervention 2009; 5 Suppl F:F28-35.  Haude M, Ince H, Abizaid A, Toelg R, Lemos PA, von Birgelen C, Christiansen EH, Wijns W, Neumann FJ, Kaiser C, Eeckhout E, Lim ST, Escaned J, Onuma Y, Garcia-Garcia HM, Waksman R. Sustained safety and performance of the second-generation drug-eluting absorbable metal scaffold in patients with de novo coronary lesions 12-month clinical results and angiographic findings of the BIOSOLVE-II first-in-man trial. Eur Heart J 2016; 37:2701-2709.  Wittchow E, Adden N, Riedmüller J, Savard C, Waksman R, Braune M. Bioresorbable drug-eluting magnesium-alloy scaffold- design and feasibility in a porcine coronary model. EuroIntervention 2013; 8:1441-1450.  Foin N, Lee RD, Torii R, Guitierrez-Chico JL, Mattesini A, Nijjer S, Sen S, Petraco R, Davies JE, Di Mario C, Joner M, Virmani R, Wong P. Impact of stent strut design in metallic stents and biodegradable scaffolds. Int J Cardiol 2014; 177:800-808.  Zheng YF, Gu XN, Witte F. Biodegradable metals. Mater Sci Eng R 2014; 77:1-34.  Kraus T, Moszner F, Fischerauer S, Fiedler M, Martinelli E, Eichler J, Witte F, Willbold E, Schinhammer M, Meischel M, Uggowitzer PJ, Loffler JF, Weinberg A. Biodegradable Fe-based alloys for use in osteosynthesis: outcome of an in vivo study after 52 weeks. Acta Biomater 2014; 10:3346-3353.  Peuster M, Wohlsein P, Brügmann M, Ehlerding M, Seidler K, Fink C, Brauer H, Fischer A, Hausdorf G. A novel approach to temporary stenting: degradable cardiovascular stents produced
from corrodible metal-results 6-18 months after implantation into New Zealand white rabbits. Heart 2001; 86:563-569.  Peuster M, Hesse C, Schloo T, Fink C, Beerbaum P, von Schnakenburg C. Long-term biocompatibility of a corrodible peripheral iron stent in the porcine decending aorta. Biomaterials 2006; 27:4955-4962.  Waksman R, Pakala R, Baffour R, Seabron R, Hellinga D, Tio FO. Short-term effects of biocorrodible iron stents in porcine coronary arteries. J Interv Cardiol 2008; 21:15-20.  Mueller PP, Arnold S, Badar M, Bormann D, Bach FW, Drynda A, Meyer-Lindenberg A, Hauser H, Peuster M. Histological and molecular evaluation of iron as degradable medical implant material in a murine animal model. J Biomed Mater Res A 2012; 100:2881-2889.  Grogan JA, Leen SB, McHugh PE. Comparing coronary stent material performance on a common geometric platform through simulated bench testing. J Mech Behav Biomed Mater 2012; 12: 129-138.  Obayi CS, Tolouei R, Mostavan A, Paternoster C, Turgeon S, Okorie BA, Obikwelu DO, Mantovani D. Effect of grain sizes on mechanical properties and biodegradation behavior of pure iron for cardiovascular stent application. Biomatter 2016; DOI: 10.4161/21592527.2014.959874.  Liu B, Zheng YF, Ruan L. In vitro investigation of Fe30Mn6Si shape memory alloy as potential bioresorbable metallic material. Mater Lett 2011; 65:540-543.  Schinhammer M, Pecnik CM, Rechberger F, Hanzi AC, Loffler JF, Uggowitzer PJ. Recrystallization behavior, microstructure evolution and mechanical properties of bioresorbable Fe-Mn-C(-Pd) TWIP alloys. Acta Mater 2012; 60:2764-2756.  Schinhammer M, Steiger P, Moszner F, Loffler JF, Uggowitzer PJ. Degradation performance of bioresorbable Fe-Mn-C(-Pd) alloys. Mater Sci Eng C 2012; 33:1882-1893.
 Liu B, Zheng YF. Effects of alloying elements (Mn, Co, Al, W, Sn, B, C and S) on biodegradability and in vitro biocompatibility of pure iron. Acta Biomater 2011; 7:1407-1420.  Hermawan H, Purnama A, Dube D, Couet J, Mantovani D. Fe-Mn alloys for metallic bioresorbable stents: degradation and cell viability studies. Acta Biomater 2010; 6:1852-1860.  Schinhammer M, Gerber I, Hanzi AC, Uggowitzer PJ. On the cytocompatibility of bioresorbable Fe-based alloys. Mater Sci Eng C 2013; 33:782-789.  Purnama A, Hermawan H, Champetier S, Mantovani D, Couet J. Gene expression profile of mouse fibroblasts exposed to a biodegradable iron alloy for stents. Acta Biomater 2013; 9:8746-8753.  Drynda A, Hassel T, Wilhelm Bach F, Peuster M. In vitro and in vivo corrosion properties of new iron-manganese alloys designed for cardiovascular applications. J Biomed Mater Res B Appl Biomater 2015; 103:649-660.  Moravej M, Prima F, Fiset M, Mantovani D. Electroformed iron as new biomaterial for degradable stents: development process and structure-properties relationship. Acta Biomater 2010; 6:1726-1735.  Moravej M, Amira S, Prima F, Rahem A, Fiset M, Mantovani D. Effect of electrodeposition current dentsity on the microstructure and the degradation of electroformed iron for degradable stents. Mater Sci Eng B 2011; 176:1812-1822.  Obayi CS, Tolouei R, Paternoster C, Turgeon S, Okorie BA, Obikwelu DO, Cassar G, Buhagiar J, Mantovani D. Influence of cross-rolling on the micro-texture and biodegradation of pure iron as biodegradable material for medical implants. Acta Biomater 2015; 17:68-77.  Hermawan H, Mantovani D. Process of prototyping coronary stents from biodegradable Fe-Mn alloys. Acta Biomater 2013; 9:8585-8592.  Lin WJ, Zhang DY, Zhang G, Sun HT, Qi HP, Chen LP, Liu ZQ, Gao RL, Zheng W. Design and
characterization of a novel biocorrodible iron-based drug-eluting coronary scaffold. Mater Design 2016; 91:72-29.  Moravej M, Purnama A, Fiset M, Couet J, Mantovani D. Electroformed pure iron as a new biomaterial for degradable stents: in vitro degradation and preliminary cell viability studies. Acta Biomater 2010; 6:1843-1851.  Zhu SF, Huang N, Xu L, Zhang Y, Liu HQ, Sun H, Leng YX. Biocompatibility of pure iron: in vitro assessment of degradation kinetics and cytotocity on endothelial cells. Mater Sci Eng C 2009; 29:1589-1592.  Chen CZ, Li Q, Leng YX, Chen JY, Zhang PC, Bai B, Huang N. Improved hardness and corrosion resistance of iron by Ti/TiN multilayer coating and plasma nitriding duplex treatment. Surf Coat Tech 2010; 204:3082-3086.  Nie FL, Zheng YF. Surface chemistry of bulk nanocrystalline pure iron and electrochemistry study in gas-flow physiological saline. J Biomed Mater Res B Appl Biomater 2012; 100:1404-1410.  García-García HM,Serruys PW, Campos CM, Muramatsu T, Nakatani S, Zhang YJ, Onuma Y, Stone GW. Assessing Bioresorbable Coronary Devices: Methods and Parameters. JACC Cardiovasc Imag 2014; 7:1130-1148.  Lin WJ, Zhang G, Cao P, Zhang DY, Zheng YF, Wu RX, Qin L, Wang GQ, Wen TY. Cytotoxicity and its test methodology for a bioabsorbable nitrided iron stent. J Biomed Mater Res B Appl Biomater 2015; 103:764-776.  Rippstein P, Black MK, Boivin M, Veinot JP, Ma X, Chen YX, Human P, Zilla P, O’Brien ER. Comparison of processing and sectioning methodologies for arteries containing metallic stents. J Histochem Cytochem 2006; 54:673-681.  Feng QM, Zhang DY, Xin CH, Liu XD, Lin WJ, Zhang WQ, Chen S, Sun K. Characterization and
in vivo evaluation of a bio-corrodible nitrided iron stent. J Mater Sci: Mater Med 2013; 24:713-724.  Chacko Y, Chan R, Haladyn JK, Lim R. Overaggressive stent expansion without intravascular imaging: impact on restenosis. Heart Asia 2014; 6:32-35.  Foin N, Sen S, Allegria E, Petraco R, Nijjer S, Francis DP, Mario CD, Davies JE. Maximal expansion capacity with current DES platforms: a critical factor for stent selection in the treatment of left main bifurcations. EuroIntervention 2013; 8:1315-1325.  Carrozza JP Jr, Hosley SE, Cohen DJ, Baim DS. In vivo assessment of stent expansion and recoil in normal porcine coronary arteries. Circulation 1999; 100:756-760.  Bourantas CV, Onuma Yoshinobu, Farooq V, Zhang Y, Garcia HM, Serruys PW. Bioresorbable scaffolds: current knowledge, potentialities and limitations experienced during their first clinical applications. Int J Cardiol 2013; 167:11-21.  Serruys PW, Onuma Y. Bioresorbable scaffolds: from basic concept to clinical applications. CRC Press, Boca Raton, 2017.  Pierson D, Edick J, Tauscher A, Pokorney E, Bowen P, Gelbaugh J, Stinson J, Getty H, Lee CH, Drelich J, Goldman J. A simplified in vivo approach for evaluating the bioabsorbable behavior of candidate stent materials. J Biomed Mater Res Part B Appl Biomater 2012; 100B:58-67.  Jaffer SJ, Hansson CM. Chloride-induced corrosion products of steel in cracked-concrete subjected to different loading conditions. Cem Concr Res 2009; 39: 116-125.  Ibrahim TA, Ramatu I. Iron in body metabolism. Chem Res J 2016; 1:38-42.  Anderson JM, Rodriguez A, Chang DT. Foreign body reaction to biomaterials. Semin Immunol 2008; 20:86-100.  Haas A. The phagosome: compartment with a license to kill. Traffic 2007; 8: 311-330.  Lundborg M, Falk R, Johansson A, Kreyling W, Camner P. Phagolysosomal pH and dissolution of
cobalt oxide particles by alveolar macrophages. Environ Health Perspect 1992; 97:153-157.  Kurz T, Eaton JW, Brunk UT. The role of lysosomes in iron metabolism and recycling. Int Biochem Cell Biol 2011; 43:1686-1697.  Simon C, Palmaz JC, and Sprague EA. Influence of topography on endothelialization of stents: clues for new designs. J Long Term Eff Med Implants, 2000; 10:143–151.  Joner M, Nakazawa G, Finn AV, Quee SC, Coleman L, Acampado E, Wilson PS, Skorija K, Cheng Q, Xu X, Gold HK, Kolodgie FD, Virmani R. Endothelial cell recovery between comparator polymer-based drug-eluting stents. J Am Coll Cardiol, 2008; 52: 333–342.  Sudhir K, Hermiller JB, Ferguson JM, Simonton CA. Risk factors for coronary drug-eluting stent thrombosis: influence of procedural, patient, lesion, and stent related factors and dual antiplatelet therapy. ISRN Cardiol 2013; 2013:1-8.  Wu C, Qiu H, Hu XY, Ruan YM, Tian Y, Chu Y, Xu XL, Xu L, Tang Y, Gao RL. Short-term safety and efficacy of the bioresorbable iron stent in mini-swine coronary arteries. Chin Med J, 2013; 126:4752-4757.  Andrews NC. Disorders of iron metabolism. New Engl J Med 1999; 341:1986-1995.
Figure Captions Fig.1 Schematic drawing of the in-house designed apparatus for in vitro corrosion test. Fig.2 Radial compression curves of the nitrided iron scaffold, pure iron scaffold and Xience PrimeTM stent. Fig.3 Scanning electron microscope images of the nitrided iron scaffold at pre-expansion, normal and maximal expansion state. Fig.4 Illustration of the side-branch accessablity of a Φ3.0×18mm nitrided iron scaffold. Into (a) a fenestrated mock vessel, the (b) nitrided iron scaffold was deployed with a cell aligned to the opening on the mock vessel, and (c) the cell was then expanded with a balloon of Φ2.25 mm, (d) measuring of the inner diameter of the expanded cell on the nitrided iron scaffold. Fig.5 In vitro and in vivo corrosion profiles of the nitrided iron scaffold and the pure iron scaffold. Fig. 6 (a) X-ray Photoelectron Spectroscopy (XPS) spectrum and (b) high-resolution spectrum of Fe2p peak in the XPS spectrum of the in vivo corrosion products, (c) representative optical photo, (d) color and (e) magnetism of the in vivo corrosion products, (f) Raman spectrum of the in vivo corrosion products; of the nitrided iron scaffold after 12 months implantation in the rabbit abdominal aorta. Fig.7 The corrosion evolution, corrosion products composition and distribution illustrated by the strut cross sections and its corresponding element distributions of the nitrided iron scaffold after implantation of 1,4,6,9 and 12 months in rabbit abdominal aorta, respectively. Fig. 8 Endothelialization comparison for the nitrided iron scaffold and the 316L SS stent after 7 days implantation in rabbit abdominal aorta. Fig. 9 Pathological tissue slices with hematoxylin and eosin (HE) staining of (a, b, c) 316L SS stent and (d, e, f) the nitrided iron scaffold after implantation in rabbit abdominal aorta for 1 month, 6 months and 12 months, respectively. (Resin-embedded sections for the 316L SS stent and paraffin-embedded sections for the nitrided iron scaffold) Fig. 10 (a) Paraffin-embedded sections with hematoxylin and eosin (HE) staining of the nitrided iron scaffold after implantation in rabbit abdominal aorta for 12 months, (b) an enlargement of the rectangular area in Fig. 10(a); (c) an enlargement of the rectangular area in Fig. 10(b), indicating the 39
macrophages engulfing the insoluble corrosion particles around the struts. Fig. 11 Histopathological observations on the tissues of (a) heart, (b) liver, (c) spleen, (d) lung, (e) kidney of porcine with nitrided iron scaffolds after 53 months implantation in the coronary artery; and histopathological observations on the tissues of (f) heart, (g) liver, (h) spleen, (i) lung, (j) kidney of healthy porcine without scaffold implanted. Fig. 12 (a) Micro-CT 2D images of one of the nitrided iron scaffold after 53 months implantaion in porcine coronary; (b) representative enlargement of area similar to A; (c) representative enlargement of area B; (d) resin-embedded histopathological section from area B (expanded corrosion products in situ, indicated by the black arrows); (e) resin-embedded histopathological section from area A (corrosion products in media or adventitia, indicated by the blue arrows); TEM images of the nitrided iron scaffolded vessel segment tissue after 33 months implantaion in a porcine coronary, showing insoluble corrosion particles of the nitrided iron scaffold as indicated by the red triangles in (f) interstitial fluid between smooth muscle cells and (g) somatic cell; paraffin-embedded sections of the nitrided iron scaffolded vessel segment of area B, after 53 months implantation in porcine coronary, show (h) corrosion products in situ and peri-strut areas, (i) corrosion products migrating to adventitia, (j) tissue regeneration within the original strut footprint and (k) enlargement of the indiscernible corrosion products under Micro-CT.
Table 1 Testing groups and sample size in rabbit experiment Testing groups - Testing items
(1) Radial strength (without tissue) A-(2) Radial strength (with tissue) A-(3) Mass loss B-(4) Corrosion product characterization (XPS, color and magnetism, Raman) B-(5) Corrosion product distribution (resin-embedded section for EDS) C-(6) Endothelialization (SEM)
Follow-up time 1 7days month / □ 6/3 □ 6/3 / ■ 6/3 □ 6/3 / ■ 6/3
4 months □ 6/3 □ 6/3 ■ 6/3 □ 6/3 ■ 6/3
6 months □ 6/3 □ 6/3 ■ 6/3 □ 6/3 ■ 6/3
9 months □ 6/3 □ 6/3 ■ 6/3 □ 6/3 ■ 6/3
12 months / □ 6/3 ■ 6/3 □ 6/3 ■ 6/3
24 months / □ 6/3 ■ 6/3 □ 6/3 ■ 6/3
36 months /
□ 4/2 ▲ 4/2
Rabbit number 12
/ □ 6/3 ■ 6/3
C-(7) Histopathological observation / / / / / 6 □ 4/2 □ 4/2 □ 4/2 (resin-embedded section) C-(8) Histopathological observation / / / / / 6 ▲ 4/2 ▲ 4/2 ▲ 4/2 (paraffin-embedded section) Rabbit number Total:78 4 15 10 15 10 12 6 6 Description was in the format of scaffolds/stents type and number (rabbit number), e.g. □ 6/3 meant 6 nitrided iron scaffolds from a total of 3 rabbits. Scaffolds/stents types included □ :Nitrided iron scaffold; ■ :Pure iron scaffold; ▲ :316L SS stent. The highlighted samples were reused ones of their previous testing item.
Table 2 Device parameters and performance of a nitrided iron scaffold compared with a pure iron scaffold, a drug eluting stent (DES) and two bioresorbable scaffolds. Φ3.0×18
Bare strut Thickness (µm)
Percent Surface Area @ NP (%)
Crossing Profile (mm)
Recoil @ NP (%)
Foreshortening @ NP (%)
Radial Strength (kPa)
Radial Stiffness (kPa/mm)
Maximal expansion diameter (mm)
Side-branch accessability (mm)
Nitrided iron Scaffold
Pure iron Scaffold
Absorb GT1 TM
Format of displayed values: mean value ± standard deviation(sample size)[reference]. Bare metallic stent version for Xience Prime is MultiLink 8 (Abbott Vascular). *: data per AbsorbTM Instructions for User (IFU); —: No data available.
Statement of Significance Pure iron as a potential bioresorbable material has major disadvantages of slow corrosion and bioresorption. However, So far, there are neither quantitative data of long-term in vivo corrosion nor direct experimental evidence for bioresorption of pure iron and its alloys. Only this work systemically investigated long-term in vivo corrosion behavior and biocompatibility of a nitrided iron (Fe-0.07N) coronary scaffold up to 53 months after implantation and explored its bioresorption mechanism. These are fundamental and vital for developing novel Fe-based alloys overcoming the intrinsic drawbacks of pure iron. Novel testing and section-preparing methods were also provided in this work to facilitate future research and development of novel Fe-based alloy scaffolds.