To answer the question if implantation of newly developed iron alloys as bone substitutes causes pathologic deposition, liver, spleen, kidneys, heart, brain, supraregional lymph nodes and implant site were tested. In order to monitor systematic responses laboratory tests were conducted screening for inflammatory markers and iron metabolism parameters.
Statistical analysis of inflammatory organ changes did not reveal any significant implant-related changes compared to controls. In contrast to these findings, iron deposition around the implant site was significantly elevated in all implant groups. Even popliteal lymph nodes were affected in the 1.4 g / cm³ group, though no other organs were impaired.
A number of studies, particularly in context of hemochromatosis, showed deposition of iron in liver, pancreas, heart muscle, kidneys, skin and joints according to excessive storage [31]. Against this background, an absence of more severe iron deposits in the above-mentioned organs is a clear evidence for a minor deposition during degradation of the implants.
There are, however, few studies on biodegradable porous iron-based implant materials in general [25–27] and only one in-vivo study known to us [28]. The latter does not observe matters of tissue reaction and inflammation in ex-vivo investigations. However, a study performed by Ulum et al. focusses on the cellular response and iron ion levels in the vicinity of implanted Fe, 316L and Fe-HA, Fe-TCP and Fe-BCP composites implanted into the periosteum of the medio proximal region of sheep radial bones [32]. A less pronounced swelling of the implant site indicates a less pronounced inflammatory response for pure Fe compared to 316L. This is confirmed by the blood cell count performed for pure Fe implants that shows normal numbers for erythrocytes, indicating low toxicity, lower number of leucocytes compared to 316L, indicating lower rejection and a lower neutrophil to lymphocyte ration than for 316L, indicating a lower cellular stress level. Giant cells were found in the vicinity of all investigated materials, however, there was less granular tissue found around pure iron compared to 316L implants. Finally, more iron ions were released from pure Fe in the first 2 weeks than from all other investigated materials, but the Fe blood plasma level remained in a normal range in all sheep.
The question of local toxic or inflammatory reactions and possible iron deposition after implantation of a biodegradable iron implant in form of a stent has been investigated by Peuster et al. who come to similar results. Within this study, biocorrodible stents consisting of 99.5 wt.% iron were implanted into the descending aorta of minipigs [33], while an additional non-corrodible commercial stent was positioned by catheter intervention. Throughout the experiment, leukocytes and serum iron levels were constantly normal. Animals were euthanized at various time points after stent implantation to perform histological evaluation of lung, heart, spleen, liver and kidneys. Comparable to our own results, no increased iron deposition was shown in these organs, but a slight increase in the number of extracellular iron deposits and iron- laden macrophages in the local lymph nodes has been found [33]. This effect was observed for the first time in the histological studies conducted a month after stent implantation and was seen to be replicable until the end of the trial.
In a paper published in 2001 by the same group, similar results were presented for an iron-based degradable stent implanted in the aorta descendens of rabbits. In addition to regular angiography to assess the patency of the aorta, histopathological examination of spleen, liver, kidneys, lungs and heart, and the implant-related para-aortic tissue were carried out [16]. Within these tests, few iron deposits and small lymphocytic infiltrations in the para-aortic tissue and the kidneys were detected, while no changes were found in the spleen, liver, lung or heart. Analysis of the vessels supplied with the stent matched our own results as iron-laden macrophages and multinucleated giant cells could be seen.
A study with nitrided iron stents implanted in the right liliac arteries of pigs by Feng et al. showed a constantly increased inflammatory response (ranging between moderate and abundant) in the first 6 months of implantation [34]. With advancing degradation of the iron struts, a decreased number of inflammatory cells was found at 12 months post-operation despite a severe accumulation of corrosion products in the tissue. Lin et al. produced similar results for similar pure and nitrided iron stents coated with drug-eluting PDLLA implanted in the abdominal aorta of New Zealand whit rabbits [35]. Constant slight inflammatory responses were found throughout the 13-month implantation despite the almost complete conversion of the iron to solid products. There was, however, no infiltration of inflammatory cells like neutrophils, lymphocytes or eosinophils was detectable at all time points investigated. There were also no signs for severe and prolonged inflammatory response observed for PLA-coated pure iron stents implanted into the abdominal aorta of rabbits by Qi et al. [36].
While known benchmark values for laboratory data within veterinary medicine are based on only a small number of studies, many parameters have not even been analyzed yet. Respectively current literature does not provide any reference values for serum ferritin. Therefore, comparison with the control group and assessment of ferritin during the experiment are extremely important and helpful. For the first time, a publication in 2011 illustrated the association between systemic inflammation and decrease in serum iron concentration in sheep [37].
Literature of veterinary medicine quote different reference values for serum iron levels in sheep ranging from 18.8 to 34.3 µmol/L [38], 18–48 µmol/L [39] or 29–40 µmol/L [40]. Assuming of the most accepted reference value 18–48 µmol/L, none of the measured serum iron values is out of the defined threshold, which suggests that no significant iron burden occurred in the context of the experiments [39]. An animal experiment on pigs demonstrated no significant differences in iron levels initially as well as one day or 360 days after implantation of a biodegradable stent [33]. Matching our own results, the reference range for serum iron was not exceeded in any of the animals. This is even more important as the implant used by Peuster et al. consisted of a much smaller amount of iron (200 mg) compared to the approximately 1.2 g or 1.6 g used in our study [33]. As 1.6 % phosphorus is used in the iron alloys, serum levels of phosphorus were determined by chemical analysis supplementally. The level of phosphorus, which acts as an essential trace element in the sheep organism, was evaluated in various publications. Serum standard values of 1.0–2.6 mmol/L [41], 1.2 to 2.3 mmol/L [38], 1.23 to 1.98 mmol/L [42] or 1.2 to 2.5 mmol/L were defined for sheep [40]. Assuming acceptance of the reference range of Pernthaner et al. with 1.0–2.6 mmol/L only one animal tested at the beginning of the trial showed phosphorous amounts outside the reference range [41].
Standard values of alkaline phosphatase in sheep range between 46–395 U/L [38]; 44–355 U/L [41], 45–208 U/L [42] or 4–175 U/L [43]. None of the measured values within the study period exceeded Dias´ et al. reference range of 4–175 U/L [43].
Acute phase proteins (APP) show a change in serum levels in the course of infection, inflammation, tissue injury, neoplastic processes or stress [44, 45] stimulated by proinflammatory cytokines such as IL-6 or TNF-α [46].
Various studies have shown that acute-phase proteins differ greatly in their reactions in different species [45, 47, 48]. Thereby, the largest increases in acute-phase response were determined for haptoglobin, while c-reactive protein is of no importance in small ruminants [45, 47, 48].This is confirmed by a fast and significant increase in haptoglobin levels subsequent to sheep`s infection with corynebacterium pseudotuberculosis. Within four weeks values hit reference range, while serum peak values (factor 17) were seen until 1–2 weeks after infection [48].Therefore, haptoglobin appears to be suitable to detect acute infections or to assess acute systemic immune responses in sheep. Reference haptoglobin value for mature ruminants seems to be less than 2 mg/dL [48], while in young sheep a range of 6–12 mg/dL was shown [49].
Upon acceptance of the reference value by Eckersall et al. only three measurements exceeded 2 mg/dL [48]. In summary, there is no evidence of a significant acute phase response in the context of degradable bone substitute material implantation.
Summarizing blood count analyses, no changes were seen in the number of leukocytes or differential counts and therefore chronic inflammatory changes in the experimental animals are highly unlikely.
Limitations
In our recently published study we were able to show that the planned degradation of the implant could not be achieved within a period of about 2 years. There may simply not have been enough degradation products to detect toxicity. In addition, the effect of ionised iron species must be questioned.
Initial loss of laboratory values due to hemolysis in the control group and the short-term implant group 1.0 g / cm3 is a systematic problem and imparts an even higher significance to asset longitudinal laboratory values in each group. By developing a statistical workaround, this problem could be compensated. Thereby, all available preoperative laboratory values were defined as a population by positive statistical testing and could therefore be used for comparison within the experimental groups.
To enable a more accurate evaluation of iron balance, additional transferrin values and transferrin saturation should be determined next to the serum iron and ferritin values in a possible following project. Potentially even liver biopsy may be performed in order to detect iron overload to rule out implant independent increases in serum iron and ferritin values [50].
Baumgartner 1994 kept 44 1-year old sheep on a farm and measured phosphorus levels at regular intervals over a year to determine possible seasonal variations [51]. In this experiment, a significant difference in phosphorus levels was described showing higher values in summer and lower values in winter. However, all values lay within the reference range identified by Pernthaner et al. [41]. Therefore, dietary seasonal variations for phosphorus content must be considered as a possible cause for the increase in value in our study.
After 6 and 12 months we could identify many solid degradation products in the microscopical evaluation. Particles were identified inside the implant and in the surrounding soft tissue. Also, macrophages with phagocyted products were detected. There are only few in-vivo studies on the degradation behavior and the corresponding tissue reaction so far and all of them only consider blood vessel wall tissues [34–36, 52]. All of the studies observed a certain degree of prolonged inflammatory response, but none addresses a possible correlation between solid degradation products and inflammatory reaction in detail. This matter should thus be of intensified interest for both vascular and bone tissues in future studies.