Keywords sterilization - printing, three-dimensional - ethylene oxide - biodegradable plastics
Introduction
The use of three-dimensional (3D) technology for the printing of objects by additive
manufacturing (AM) or 3D printing (prototyping) has been growing exponentially in
the health area (orthopedics, bucomaxilofacial surgery, neurosurgery, and cardiac
surgery, among others).[1 ] It can be applied for educational purposes (printing of anatomical parts, for example),
surgical planning, creation of customized implants, orthotics, and external fixers
and surgical reparators.[2 ]
[3 ]
[4 ]
[5 ] Specifically in the orthopedic area, surgeons and patients have benefited from this
technology in the creation of surgical guides and in the prior planning for the intraoperative
use of printed parts, guiding the correct position during osteotomies, bone perforations,
and placement of various types of implant materials (Kirschner wires, drills and screws,
etc.), reducing surgical time and improving accuracy.[6 ]
[7 ]
[8 ]
[9 ]
[10 ] With the popularization and greater accessibility of home 3D printers, surgeons
have planned and created their guides in a homemade mode, sterilizing them in their
institutions for use during surgery, discarding them after their application. The
most used materials in mold prototyping are plastic filaments in polylactic acid (PLA)
or acrylonitrile butadiene styrene (ABS) polymer, due to their cost-effectiveness
and handling, but both still have difficulties for sterilization, mainly because they
are thermosensitive. Some countries have rules for the specific processing of these
types of 3D printed materials, but we have not found them in our environment so far.[11 ] The objective of the present work is to compare the efficacy and reliability of
the autoclave and ethylene oxide (EO) methods for sterilization of objects printed
in PLA, enabling their safe use in surgeries.
Material and Methods
Objects were designed in 3D format, creating standard STL files for prototyping (stereolithography),
using the computer-assisted design (CAD) software Rhinoceros, version 5.5.4, licensed.
After their creation, the files were prepared for 3D printing with the software Simplify3D,
EULA, version 4.0.0, licensed, and were forwarded to printing on PLA plastic material.
The printer used was the home model (desktop) Minibot 120. In the printing process,
different percentages of object filling (infill) were chosen, creating totally solid
("massive") models or with empty space inside (hollow with "dead space"). ([Figure 1 ]) Thus, 20 objects were printed in square format (1 cm2 ), named solids (S), and 20 in rectangular (5.0 × 2.0 × 0.5cm), hollow objects, named
"nonsolid" (NS). Two study groups were separated, the first with 10 objects type S
and 10 type NS (G1), and the second (G2) in the same way, totaling 2 groups with 20
objects each. Objects from G1 were sent for sterilization by the steam method with
autoclave (Sercon model), being processed by the "fast cycle" method at 121°, preventing
the melting of the part. Objects from G2 were sterilized by the EO method ("cold")
in a specialized center contracted by the institution. Each object was sterilized
and packed separately in a standardized manner with double plastic protection, keeping
it sterile and stored in an appropriate environment for 1 week ([Figure 2 ]). In the 2nd week, the objects were referred to culture in the microbiology laboratory of the
institution. The procedures were performed by a specialized professional, duly attired,
with the samples manipulated in a standard environment (laminar flow chapel for the
protection of products handled inside, avoiding external contamination), after sterilization
of the flow with 70% alcohol and with continuously lit fire. All samples from groups
G1 and G2 were placed in sterile vials with Brian Heart Infusion (BHI) broth, which
is an enrichment medium used in the recovery of fastidious or nonfastidious microorganisms,
including aerobic and anaerobic bacteria and fungi) and maintained for 48 hours in
an oven (34° to 37°C). At this stage, the NS type objects of the 2 groups were broken
immediately before being introduced into the BHI culture medium, communicating the
internal space ("dead space") with the exterior in order to also analyze the effectiveness
in sterilization inside the hollow parts. For this reason, NS-type objects were printed
in rectangular format, making them easier to break. ([Figure 3 ]) After 48 hours, the samples were sowed in Blood Agar-MacConkey (using a rich base
that provides growth conditions for most microorganisms) and in MacConkey Agar (a
culture medium intended for the growth of Gram-negative bacteria and indication of
lactose fermentation). After sowing, the cultures were kept in a greenhouse for 24 hours
(for bacterial growth and subsequent reading) and the samples in broth were returned
to the greenhouse (34° to 37°C) for incubation for another 15 days. After this period,
they were sowed again in the same way, being submitted to a new reading. The collected
data were analyzed with the aid of IBM SPSS Statistics for Windows, version 22.0 (IBM
Corp., Armonk, NY, USA) software and of the Fisher exact test, followed by residue
analysis when statistical significance was observed.
Fig. 1 Computer images demonstrating the creation of study objects with 3D technology. A)
square and rectangle drawing in CAD software; B) preparation of the object for 3D
printing with internal filling (infill) of 100%; C) preparation of the object with
partial internal filling (hollow); D) photography showing printed objects with different
filling percentages. Source: authors' file.
Fig. 2 Photograph showing storage mode of objects printed in double plastic after sterilization.
Source: authors' file.
Fig. 3 Photography demonstrating detail in the process of sowing hollow objects (NS), which
were broken immediately before placement in Brian Heart Infusion (BHI) broth. Source:
authors' file.
Results
The results after 48 hours and 15 days of incubation were similar. In group G1 (sterilized
in autoclave), there was bacterial growth in 50% of the samples of S objects (50%
negative) and in 30% of NS objects (70% negative). In group G2 (sterilized in EO),
there was no growth in 100% of the samples of S objects, but growth was observed in
20% of the NS objects (80% negative). These data, including the statistical calculations
performed, are shown in [Table 1 ] and in [Figures 4 ] and [5 ]. The bacteria isolated in all cases of contamination was non-coagulase-producing
Staphylococcus Gram positive.
Fig. 4 Graphic demonstration of the statistical analysis comparing positive (growth) and
negative (sterile) results after reading the crop samples with solid pieces (S). Abbreviation:
OE, ethylene oxide; * Statistically significant value after residue analysis. Source:
research data.
Fig. 5 Graphic demonstration of the statistical analysis comparing positive (growth) and
negative (sterile) results after reading the samples of cultures with hollow pieces
(NS). Abbreviation: OE:, ethylene oxide. Source: research data.
Table 1
Objects, n (%)
Autoclave
EO
p-value[† ]
n = 10
n = 10
Solid parts (S)
Negative
5 (50.0)
10 (100.0)b
0.033
Positive
5 (50.0)b
0 (0.0)
Hollow parts (NS)
Negative
7 (70.0)
8 (80.0)
0.999
Positive
3 (30.0)
2 (20.0)
Discussion
The use of 3D technology in medicine has grown rapidly, benefiting several areas with
its application, including orthopedics,[2 ] which is demonstrated by the growing number of publications on the subject. In a
systematic review, Tack et al.[1 ] initially collected 7,482 papers for analysis. Among these, 60% were studies with
applications of printed surgical guides or surgical planning. Despite the ease in
manufacturing domestically these objects, the type of material and its sterilization
remain the greatest difficulties. Among the available materials, PLA is the most used
synthetic because it is biocompatible, nonpolluting (biodegradable and from renewable
resources), low-cost, and is easy to handle, being also the material of preference
by the authors.[12 ]
[13 ] For medical use, its main disadvantage is being thermosensitive, with the beginning
of its melting occurring from 120°C, which can cause deformation in the part during
the processes of steam sterilization and high temperature (autoclave), making its
use unfeasible.[14 ] Since autoclave is the most accessible sterilization option available in most hospitals,
it can be used by being programmed to run in "fast cycle" mode as an alternative for
thermosensitive objects, subjecting the material to 121°C for a shorter period. This
has demonstrated effective preservation of the original PLA.[12 ]
[15 ]
[16 ] The alternative method viable in our environment for "cold" sterilization of thermosensitive
materials is EO.[17 ]
[18 ]
[19 ]
[20 ] Other "cold" methods, such as plasma gas and gamma rays, among others, are also
effective, but are costly and may become unfeasible in some institutions. In a recent
systematic review, Davila et al. concluded that the most universally used methods
for this type of material are EO and gamma rays. Other methods, such as hydrogen peroxide/plasma
gas, peracetic acid, and ozone have been explored as alternatives, but there is no
defined standardization yet.[21 ] Materials more resistant to autoclave, such as the resin used in the dental environment,
also require more expensive printers and raw material. Regulatory mechanisms standardize
the use of autoclave and EO in the processing of the most common surgical materials,
but this has not yet been clearly established for the objects obtained with 3D printing
in our environment. For materials considered thermosensitive (punch batteries, endoscope
plastic parts, etc.), EO remains the most recommended to prevent possible melting.[14 ]
[20 ] A concern in our study was regarding the efficacy in complete sterilization, including
the internal space created in rectangular parts (NS), differentiating from the efficacy
observed in solid parts (S). Printing with partial internal filling (% infill ) is common in household printings because the process is faster and more economical
by using less raw material. Neches et al.[22 ] and Skelley et al.[23 ] demonstrated efficient sterilization of PLA printed objects automatically by the
high temperature generated for the melting of the material during the printing of
the objects, including the interior of the parts (∼ 200°C), requiring no further processing.
Aguardo-Maestro et al.[24 ] compared autoclave, OE, and plasma gas methods in the sterilization of hollow printed
objects after inoculating a bacteria suspension inside them, finding efficacy only
in the first two methods. The plasma gas method was recommended by the authors only
for objects without internal space (solids).[24 ] Our results demonstrated failures in the efficacy of the sterilization of hollow
parts (NS) both by autoclave (G1) and by EO (G2), with bacterial growth in 30 and
in 20% of the samples, respectively, suggesting that the "dead space" was not properly
sterilized by neither method. Autoclave sterilization was also not proven safe by
the "fast cycle" method, with contamination observed, in addition to the 30% of contamination
observed in NS type parts and to the 50% of contamination observed solid parts (S).
Therefore, we do not recommend autoclave for PLA sterilization. The Type S parts sterilized
by EO were the only ones that did not have bacterial growth. The use of EO, in addition
to being effective in this type of printing (S), has the advantage of not deforming
PLA due to the the risk of its melting because it is a "cold" method. Therefore, we
recommend, for objects printed with PLA material, full-fill printing (100% infill ) and sterilization in EO as an alternative to autoclave. As limitations of the present
study, we can include the nonblinding and nonrandomization of objects, the possibility
of contamination during preparation and sowing, the absence of a control group and
of a comparison with other types of material. The small number of samples decreases
the statistical relevance of our results, but does not invalidate it, since the sample
test performed prior to the application of the statistical test showed a confidence
of 95%, with a sampling error of 5% (or 0.05). Thus, future studies are necessary
to define the most effective method for the sterilization of these objects, standardization,
and control by regulatory mechanisms.
Conclusion
Sterilization by both autoclave and OE was not effective for hollow printed objects.
Solid objects (printed with 100% internal filling) sterilized by autoclave did not
demonstrate 100% of negative samples and were not safe in the present assay. Complete
absence of contamination occurred only with solid objects sterilized by EO, with this
being the combination recommended by the authors.
