See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/326240451

Association between metal hypersensitivity and implant failure in patients

who underwent titanium cranioplasty

Article in Journal of Neurosurgery · July 2018

DOI: 10.3171/2018.1.JNS171804

CITATIONS READS

19 109

8 authors, including:

Yirui Sun Xing Wu

Fudan University Fudan University
48 PUBLICATIONS 2,011 CITATIONS 68 PUBLICATIONS 1,466 CITATIONS

SEE PROFILE SEE PROFILE

Zhuoying Du Xuehai Wu
Fudan University Fudan University
19 PUBLICATIONS 246 CITATIONS 48 PUBLICATIONS 964 CITATIONS

SEE PROFILE SEE PROFILE

Some of the authors of this publication are also working on these related projects:

Spine cord injury View project

skin bioprinting View project

All content following this page was uploaded by Yirui Sun on 21 April 2020.

The user has requested enhancement of the downloaded file.

 CLINICAL ARTICLE

Association between metal hypersensitivity and implant

failure in patients who underwent titanium cranioplasty
*Yirui Sun, MD, PhD,1 Yue Hu, MD,2 Qiang Yuan, MD, PhD,1 Jian Yu, MD, PhD,1 Xing Wu, MD, PhD,1
Zhuoying Du, MD, PhD,1 Xuehai Wu, MD, PhD,1 and Jin Hu, MD, PhD1
Departments of 1Neurosurgery and 2Dermatology, Huashan Hospital, Fudan University, Shanghai, People’s Republic of China

OBJECTIVE Digitally designed titanium plates are commonly used for the reconstruction of craniofacial defects, al-
though implant exposure (referred to as failure) is one of the major complications. Metal hypersensitivities have been
suggested as possible causes of implant failure of orthopedic, intravascular, gynecological, and dental devices, yet there
has been no consensus on the requirement for allergy screening before cranioplasty.
METHODS In this study, the authors prospectively investigated the prevalence of metal hypersensitivity in patients for
whom cranioplasty is planned and assess its relationship with titanium implant failure (exposure).
RESULTS Based on records from 207 included patients, 39.61% of patients showed hypersensitivity to at least one kind
of metal. Approximately one-quarter (25.12%) of patients had multiple metal allergies. Co, Cd, and Zn were the 3 most
frequently identified metal hypersensitivities. No allergy to titanium was detected in this study. The overall incidence of
cranioplasty implant failure was 5.31% (11 of 207). Patients showing hypersensitivities to more than 3 kinds of metal had
higher risks of titanium plate exposure.
CONCLUSIONS Based on their findings, the authors suggest that routine allergy screening be performed before tita-
nium plate cranioplasty. For patients with hypersensitivities to more than 3 metals, alternative materials, such as poly-
etheretherketone, should be considered for cranioplasty.
https://thejns.org/doi/abs/10.3171/2018.1.JNS171804
KEYWORDS cranioplasty; metal hypersensitivity; implant failure; patch test; surgical technique

C
ranioplasty is performed for skull defects result- expensive, particularly for uninsured patients in developing
ing from neurotrauma, decompressive craniectomy, countries.15,19 Therefore, metal implants, such as titanium
and other etiologies. The purposes of cranioplasty plates, are commonly used for cranioplasty in many coun-
usually include restoring the contour of the skull, pre- tries, including mainland China. Previous epidemiological
serving the rigid and protective integrity of the calvaria, investigations have suggested that over 95% of patients re-
and minimizing morbidity and potential complications. ceive CT-modeled titanium implants during cranioplasty,
Recently, studies have also shown that cranioplasty may which has been proven to be a precise, time-saving, and
also accelerate and improve neurological recovery.1,5,10 affordable technique.2,16,23
Cranioplasty with an autogenous bone flap carries the risk One significant complication of titanium cranioplasty
of resorption and requires additional storage strategies is implant failure, which often leads to exposure of the
after decompressive craniectomies, such as freezer cryo- metal plates or screws, requiring additional operations for
preservation or subcutaneous abdominal implantation.3,9,12 removal. The incidence of cranioplasty failure has been re-
Computer-designed polyetheretherketone (PEEK) recon- ported to be between 4% and 25%.13,19,21,22 Infections were
struction has recently been proven as an effective and safe often presumed as a major cause of cranioplasty failure,
cranioplasty technique, but the price of each PEEK unit is but recent studies have suggested that such complications

ABBREVIATIONS PEEK = polyetheretherketone.

SUBMITTED July 26, 2017. ACCEPTED January 19, 2018.
INCLUDE WHEN CITING Published online July 6, 2018; DOI: 10.3171/2018.1.JNS171804.
* Y.S., Y.H., and Q.Y. contributed equally to this work.

©AANS 2018, except where prohibited by US copyright law J Neurosurg July 6, 2018 1
Sun et al.

may also be caused or accelerated by host immune sensi- at the third appointment, usually 72–96 hours after initial
tivity to metal substances within the implant. Metal hyper- application. The dermatologists recorded findings at the
sensitivities may lead to cutaneous eczematous eruptions, second and third appointments. The results for each test
chronic inflammation, pain, and a range of other adverse site were recorded as follows: negative (–); weak erythema
reactions, such as device breakage or exposure, which have only/uncertain (±); erythema and edema (+); erythema
been noted in orthopedic, endovascular, obstetric, and den- and edema with papules and vesicles confined within the
tal surgeries.18 Although consensus has not been achieved, chamber (++); and erythema with vesicles, papules, or
it is theorized that skin patch testing may be an accurate bullae extending beyond the chamber (+++). Metals that
way to identify this immune sensitivity and that this test are most frequently seen in metallic alloy implants were
can be used to guide implant material selection and the tested as allergens, including 0.5% potassium dichromate,
potential need for implant removal.4,17 However, the inci- 5% nickel nitrate, 1% cobalt chloride, 10% titanium chlo-
dence of metal hypersensitivities in patients requiring cra- ride, 2% palladium chloride, 5% aluminum nitrate, 2%
nioplasty has not been investigated. Metal mesh/plates for manganese chloride, 2% iron trichloride, 1% ammonium
cranioplasty are usually made of titanium alloy (91%–99% molybdate tetrahydrate, 1% cadmium chloride, 2% mer-
titanium). Hypersensitivities to titanium or other metals curic chloride, 2% silver nitrate, and 2% zinc sulfate. All
within the alloy may preclude their use as implantable ma- allergens were dispersed in Vaseline, with the exception of
terial, either as a cranioplasty plate or as screw fixation for ammonium molybdate tetrahydrate, which was dispersed
autologous bone or an alloplastic plate. The aims of this in water. Vaseline was applied as a negative control.
study were to investigate the incidence of metal hypersen-
sitivity in patients scheduled for cranioplasty and to assess Cranioplasty and Implant Failure
the relationship between metal allergy and titanium alloy Cranioplasty was performed at least 8 weeks after ini-
implant failure (exposure). tial decompressive craniectomy, with the exception of the
2 cases of osteofibrous dysplasia in which cranioplasty was
Methods performed immediately after the craniectomy. Patients
were free to choose a titanium plate or PEEK for cranio-
Patient Population plasty. The patch test results and the possibility of implant
Between September 2012 and January 2017, a prospec- failure were well explained to all patients. Informed con-
tive cohort study was conducted at 2 independent medi- sent was acquired from all patients and/or their closest di-
cal institutions: Shanghai Huashan Hospital and Shanghai rect relative. To achieve precision-fitting prostheses, cra-
Minhang Hospital, both of which are university clinical nial defects were modeled from CT scanning data using
institutions. Approval was obtained from the institutional stacked slices, computer-driven milling, and selective laser
review boards. Informed consent was acquired from all sintering. Cranioplasty with autologous bone was not per-
patients and/or their closest direct relative. The study ap- formed at our institution. Follow-up visits were conducted
proaches were approved by the ethics committee and the for 12 months after cranioplasty.
Chinese Clinical Trial Registry. Patients ≥ 14 years with Cranioplasty failure was defined as removal of a pa-
skull defects who were scheduled for titanium alloy cra- tient’s implant due to any of the following reasons: 1) su-
nioplasty were recruited. Patients with a history of met- perficial infection, involving skin and subcutaneous tissues
al dermatitis were excluded and were advised to receive of the incision site; 2) deep infection, involving deep soft
PEEK cranioplasty. Patients who underwent decompres- tissues, meninges, or brain; and 3) exposed implant, de-
sive craniectomy due to malignant intracranial tumors, fined as exposure of implant due to impaired wound heal-
such as glioma, were not included, since tumor recurrence, ing or erosion of the skin.
chemotherapy, and potential secondary operations could
affect the follow-up observations. Patients with the follow- Statistical Analysis
ing conditions were also excluded from this study: preg- Statistical analysis was carried out using Stata 13 (Stata-
nancy, rheumatoid arthritis currently under corticosteroid Corp LP) and Excel 2013 (Microsoft Corp.) software.
or other immunosuppressive therapy, and other contraindi- Quantitative data are expressed as the mean ± standard de-
cations to either cranioplasty or patch testing. viation (SD). The Shapiro-Wilk W test and F test were per-
formed for normality and homogeneity of variance. The
Patch Testing independent-samples t-test, Satterthwaite t-test, Wilcoxon
All included patients underwent skin patch tests at 3 rank-sum test, and 1-way ANOVA were used for the quan-
appointments, which were carried out by 2 independent titative data of independent groups. The chi-square test
dermatologists. Application of the patch tests takes about was used for qualitative data. Fisher’s exact test was used
15 minutes at the first appointment. Very small quantities in the analysis of contingency tables where sample sizes
of allergens in individual square plastic are applied to the are small (n < 40). Statistical inference was conducted at a
upper back and are kept in place with hypoallergenic ad- significance level of 0.05.
hesive tape. The patches stay in place for at least 48 hours,
during which period patients were advised to avoid vigor-
ous exercise, stretching, and showering. At the second ap- Results
pointment, usually 48 hours later, the patches are removed. Patient Demographics
The patient’s back is marked with an indelible black felt-tip During the period between September 2012 and Janu-
pen to identify the test sites. These marks must be visible ary 2017, 223 patients were scheduled for cranioplasty. Re-

2 J Neurosurg July 6, 2018

 Sun et al.

TABLE 1. Patient demographics

Demographic Characteristics Value (%)
No. of patients 207
Male/female ratio 155:52
Mean age, yrs 44.60 ± 11.60
Cause of skull defect
Trauma & decompression 160 (77.29)
Intracranial hemorrhage & decompression 41 (19.81)
Ischemic stroke 4 (1.94)
Osteofibrous dysplasia 2 (0.97)
Type of skull deficiency
Unilateral 165 (79.71)
Bilateral 28 (13.53)
Bifrontal 14 (6.76)
Other metal implants
Orthopedic implants 25 (12.08)
Intravascular devices 14 (6.76)
Dental 36 (17.39)
Cranioplasty implant failure 11 (5.31)
Implant exposure 11 (5.31)
Superficial infection 4 (1.93)
Deep infection* 3 (1.45)
Mortality 0 (0)
* Detected deep infections were found restricted to soft tissues or meninges.
No intracranial infection was detected.

FIG. 1. Flowchart showing patient inclusion. (25.12%) patients showing hypersensitivity to more than
1 kind of metal. Among patients with multiple metal hy-
persensitivities, allergies to Cd+Zn (10.15%) and Cd+Cr
cords for 6 patients were incomplete due to loss of contact. (7.25%) were most commonly identified, followed by hy-
Two patients refused cranioplasty for personal reasons. persensitivities to Co+Cd (7.25%), Co+Zn (6.28%), and
Six patients chose PEEK cranioplasty before patch testing. Cd+Cr+Zn (5.80%; Table 4).
Two patients decided to receive PEEK cranioplasty after
patch testing, but neither showed metal hypersensitivity in Cranioplasty and Implant Failure
the tests. A total of 207 patients underwent titanium cra- Cranioplasty was performed in 215 patients. Eight pa-
nioplasty in this prospective study. Twenty-one patients tients chose PEEK reconstruction, and 207 patients re-
received ventriculoperitoneal shunts due to the presence ceived computer-designed titanium plate cranioplasty.
of hydrocephalus either at the same time of or within 12 Two patients with osteofibrous dysplasia underwent cra-
months of cranioplasty. The flow chart for patient inclusion nioplasty during the craniectomy. The remaining 205
and demographic data are summarized in Fig. 1 and Table patients underwent cranioplasty between 8 weeks and 9
1. No patient died during the study period. months after the initial craniectomy (mean 21.54 ± 3.17
weeks). Significant serious adverse events, such as death,
Patch Testing CSF leakage, intracranial hemorrhage, intracranial infec-
Skin patch tests were conducted before cranioplasty. tion, skin flap necrosis, or refractory epilepsy, were not de-
Test results from the 207 patients are summarized in Ta- tected in this study. Subcutaneous hydrops was detected in
ble 2; 125 patients (60.39%) showed no hypersensitivity 11 (5.31%) patients after cranioplasty, and all cases were
to any of the tested metals. Only one patient showed an cured after conservative treatment. There was no impaired
uncertain allergy to titanium (±) in this study. The most wound healing or prolonged fever during the perioperative
frequently identified metal hypersensitivities (skin reac- periods of cranioplasty. Stitches were removed within 2
tions showing “+”, “++”, or “+++”) were: Co (19.32%), Cd weeks after cranioplasty, and the patients were discharged
(15.94%), Zn (14.49%), and Cr (10.63%). Fewer than 3% of for further follow-up visits. There were 27 (13.04%) pa-
patients showed allergies to Sn, Mo, or Al, of which the tients who required prolonged antiseizure treatment due to
frequencies were 2.42%, 2.42%, and 0.97%, respectively. the presence of epilepsy after either craniectomy or cra-
Hypersensitivities to multiple metals were identified in a nioplasty.
significant proportion of patients (Table 3). There were 52 During the 12-month follow-up, the 8 patients who re-

J Neurosurg July 6, 2018 3

Sun et al.

TABLE 2. Results of patch testing before cranioplasty in 207 patients

Results
Tested Metals – ± + ++ +++ Confirmed Allergies*
Co 153 (73.91) 14 (6.76) 36 (17.39) 3 (1.45) 1 (0.48) 40 (19.32)
Cd 162 (78.26) 12 (5.80) 32 (15.46) 1 (0.48) 0 (0) 33 (15.94)
Zn 174 (84.06) 3 (1.45) 16 (7.73) 5 (2.42) 9 (4.35) 30 (14.49)
Cr 178 (85.99) 7 (3.38) 22 (10.63) 0 (0) 0 (0) 22 (10.63)
Hg 180 (86.96) 11 (5.31) 11 (5.31) 2 (0.97) 3 (1.45) 16 (7.73)
Ag 183 (88.41) 10 (4.83) 11 (5.31) 2 (0.97) 1 (0.48) 14 (6.76)
Mn 185 (89.37) 12 (5.80) 9 (4.35) 1 (0.48) 0 (0) 10 (4.83)
Ni 190 (91.79) 7 (3.38) 4 (1.93) 4 (1.93) 2 (0.97) 10 (4.83)
Pd 191 (92.27) 8 (3.86) 7 (3.38) 1 (0.48) 0 (0) 8 (3.86)
Fe 194 (93.72) 3 (1.45) 8 (3.86) 1 (0.48) 1 (0.48) 10 (4.83)
Sn 196 (94.69) 6 (2.90) 4 (1.93) 1 (0.48) 0 (0) 5 (2.42)
Mo 198 (95.65) 4 (1.93) 5 (2.42) 0 (0) 0 (0) 5 (2.42)
Al 201 (97.10) 4 (1.93) 2 (0.97) 0 (0) 0 (0) 2 (0.97)
Ti 206 (99.51) 1 (0.48) 0 (0) 0 (0) 0 (0) 0 (0)
Vaseline (control) 207 (100.00) 0 (0) 0 (0) 0 (0) 0 (0) 0 (0)
– = negative; ± = weak erythema only/uncertain; + = erythema and edema; ++ = erythema and edema with papules and vesicles confined
within the chamber; +++ = erythema with vesicles, papules, or bullae extending beyond the chamber.
Values presented as no. (%).
* Reactions of +, ++, or +++ in patch testing were considered confirmed allergies.

ceived PEEK cranioplasty all showed good outcomes, al- examinations. Staphylococcus epidermidis and Staphylo-
though they were not included in the final statistics. Eleven coccus aureus were found in 5 and 2 cases, respectively.
(5.31%) of 207 patients who received titanium cranioplasty The infections involved the skin, subcutaneous tissues,
developed delayed scalp erosion and implant exposure deep soft tissues, or meninges of the incision site. Brain
(Table 1). None of the 11 patients showed titanium hyper- tissue or any other kind of intracranial infection was not
sensitivity in patch testing. The first implant failure oc- detected in this study.
curred at 13 weeks after cranioplasty, while the last case The association between metal hypersensitivities and
was identified at 11 months of follow-up. The mean time titanium implant failure is shown in Tables 3 and 4. Pa-
of such complication occurred at 8.27 ± 2.53 months af- tients who showed negative (–) results to all tested metals
ter cranioplasty (median 8.75 months). All 11 patients had a significantly lower risk of titanium implant failure (p
underwent reoperations and titanium plate removal. Nine < 0.001), while any confirmed metal hypersensitivities (+,
patients underwent reoperations at the same institutions ++, or +++) could be considered a risk factor (p < 0.001,
where the cranioplasty was performed, and the other 2 pa- Table 3). Patients having multiple metal hypersensitivities
tients underwent reoperation at one of our joint facilities. in patch tests, particularly for those who had confirmed
The removed titanium plates and adjacent tissue samples allergies to 4 or 5 metals, had a statistically significantly
were collected for bacteriological examinations. Bacte- higher chance of developing implant failure (Table 3). The
rial infections were confirmed in 7 of 11 cases by culture odds ratio also indicated a potential association between

TABLE 3. Results of patch testing in association with titanium implant failure

No. of Confirmed Allergies* No. of Pts (%) No. of Implant Failures (%) OR (95% CI) p Value
Null 125 (60.39) 1 (0.48) 0.058 (0.0073–0.46) <0.001
≥1 positive reaction 82 (39.61) 10 (4.83) 17.22 (2.16–137.31) <0.001
1 30 (14.49) 1 (0.48) 0.57 (0.071–4.67) >0.99
2 21 (10.14) 1 (0.48) 0.88 (0.11–7.24) >0.99
3 17 (8.21) 2 (0.97) 2.68 (0.53–13.55) 0.23
4 7 (3.86) 3 (1.45) 18.00 (3.44–94.25) 0.0035
5 4 (1.93) 2 (0.97) 43.33 (3.58–523.56) 0.0075
≥6 3 (1.45) 1 (0.48) 9.70 (0.81–116.20) 0.15
Pts = patients.
Boldface type indicates statistical significance.
* Reactions of +, ++, or +++ in patch testing were considered confirmed allergies.

4 J Neurosurg July 6, 2018

 Sun et al.

TABLE 4. The most frequently identified metallic hypersensitivities associated with titanium implant failure
Metal* No. of Pts (%) No. of Implant Failures (%) OR (95% CI) p Value
None 125 (60.39) 1 (0.48) 0.058 (0.0073–0.46) <0.001
Co 40 (19.32) 5 (2.42) 3.83 (1.11–13.27) 0.039
Cd 33 (15.94) 4 (1.93) 3.29 (0.91–11.96) 0.078
Zn 30 (14.49) 3 (1.45) 2.34 (0.59–9.40) 0.20
Cr 22 (10.62) 3 (1.45) 3.49 (0.85–14.29) 0.09
Cd+Zn 21 (10.15) 2 (0.97) 2.07 (0.41–10.29) 0.31
Cd+Cr 15 (7.25) 2 (0.97) 3.12 (0.61–16.00) 0.18
Co+Cd 14 (6.76) 2 (0.97) 3.40 (0.66–17.56) 0.16
Co+Zn 13 (6.28) 1 (0.48) 1.53 (0.18–12.99) 0.54
Cd+Cr+Zn 12 (5.80) 1 (0.48) 1.68 (0.20–14.35) 0.49
Boldface type indicates statistical significance.
* Patients may not be allergic to the listed metals only (e.g., patients showing hypersensitivity to Co could also be allergic to Cd, Zn, or Cr).

titanium implant failure and patients with more than 6 plant failure was not statistically significant. The possible
positive readings in patch tests, although the relation- association of other demographic characteristics, such as
ship was not statistically significant, probably due to the sex, age, the timing of cranioplasty, side of craniectomy,
limited sample size (Table 3). Allergies to Co were more and history of seizure, was also analyzed, but none showed
frequently detected in patch testing than allergies to other a significant relationship to titanium implant failure (Table
metals. Table 4 also indicates that patients showing hyper- 5). However, the powers of these statistical calculations are
sensitivities to Co had a higher risk of implant failure. The less than 0.80, suggesting that further investigations with
association between other metal allergies and titanium im- larger sample sizes are required.

TABLE 5. Confounding factors in association with titanium implant failure

Variable No. of Pts (%) No. of Implant Failures (%) OR (95% CI) p Value
Sex
Male 155 (74.88) 9 (4.35) 1.54 (0.32–7.37) 0.73
Female 52 (25.12) 2 (0.97) 1.95 (0.32–11.99) 0.61
Age, yrs
16–40 68 (32.85) 4 (1.93) 1.18 (0.33–4.17) 0.75
41–60 103 (49.76) 4 (1.93) 0.57 (0.16–2.01) 0.54
≥61 36 (17.39) 3 (1.45) 1.63 (0.42–6.37) 0.44
Indication for craniectomy
Trauma 160 (77.29) 10 3.07 (0.38–24.60) 0.46
Intracranial hemorrhage 41 (19.81) 1 0.39 (0.048–3.14) 0.70
Ischemic stroke 4 (1.94) 0 (0) 1.86 (0.094–36.68) 1.00
Osteofibrous dysplasia 2 (0.97) 0 (0) 3.38 (0.15–74.64) 1.00
Laterality of skull defect
Unilateral 165 (79.71) 8 (3.86) 0.66 (01.7–2.61) 0.70
Bilateral 28 (13.53) 2 (0.97) 1.45 (0.30–7.10) 0.64
Bifrontal 14 (6.76) 1 (0.48) 1.41 (0.17–11.86) 0.55
Time btwn cranioplasty & cranioplasty, mos
2–3 67 (32.37) 2 (0.97) 0.45 (0.094–2.13) 0.51
4–6 97 (46.86) 7 (3.38) 2.06 (0.58–7.27) 0.35
7–9 43 (20.77) 2 (0.97) 0.84 (0.17–4.04) 1.00
Metal implants other than cranioplasty 77 (37.20) 3 (1.45) 0.62 (0.16–2.40) 0.74
Diabetes 19 (9.18) 0 (0) 0.40 (0.22–6.98) 0.60
VP shunt 11 (5.31) 1 (0.48) 1.86 (0.21–15.99) 0.46
Seizure 27 (13.04) 0 (0) 0.27 (0.015–4.68) 0.37
Subcutaneous hydrops 11 (5.31) 0 (0) 0.70 (0.039–12.66) 1.00

J Neurosurg July 6, 2018 5

Sun et al.

Discussion posure occurred at week 13 and the last at 11 months af-

Computer-modeled titanium prostheses are important ter cranioplasty, the timing of which did not fit the usual
materials for cranioplasty, although implant failure, par- pattern of surgical infections. In addition, although bac-
ticularly exposure of the plates, remains one of the major terial infections were confirmed in 7 cases after implant
complications. Metal hypersensitivities and their associa- removal, a factor to consider is that dermatitis, effusions,
tions with implant failure have been noted in orthopedic, skin erosions, and consequent exposure of the implant due
endovascular, obstetric, and dental devices. Previous stud- to hypersensitivities are easily accompanied by secondary
ies have indicated that dermal hypersensitivity to metal is bacterial infections.
common and can affect up to 15% of the population. The In the current study, preoperative skin patch testing was
hypersensitivity reactions are primarily T cell–mediated applied to evaluate metal hypersensitivities. Patch testing is
type IV delayed-type reactions that have been established currently the gold standard for diagnosing allergic contact
in a multitude of cohort studies. The link between eczema- dermatitis on the skin, although its utility in diagnosing al-
tous dermatitis and nickel was first reported by Foussereau lergic reactions in a peri-implant environment is much less
and Laugier in 1966.6 Later studies conducted in the 1970s clear. Some authors have argued that the results of patch
and 1980s indicated that the prevalence of hypersensitiv- testing may not correctly reveal the immune response to
ity to nickel, cobalt, or chromium in patients with hip im- implants, as cutaneous exposure to an allergen is not the
plant failure was approximately 4-fold higher than that in same as the constant exposure in the closed environment
the general population.8 Systematic reviews also showed of a metal implant. In the skin, Langerhans cells are the an-
that the risk of developing a metal allergy was higher in tigen-presenting cells, while in the closed in vivo environ-
patients with failed implants (OR 2.76, 95% CI 1.14–6.70).7 ment, macrophages and dendritic cells may take the role.
Although the existing literature generally recognizes the In some studies, lymphocyte transformation testing (LTT)
value, the necessity of preimplantation diagnosis of metal or leukocyte migration inhibition testing has been applied
hypersensitivity remains controversial due to a lack of ev- to assess delayed hypersensitivity.11,14 The LTT measures
idence-based support. The metal implants for cranioplasty the proliferation of lymphocytes from peripheral blood
are usually made of titanium alloys, which may comprise in the presence of a potential allergen after incubation for
titanium, aluminum, nickel, zinc, and other metals based 7 days. The leukocyte migration inhibition test measures
on different manufacturing technologies. Therefore, im- the migration activity of mixed-population leukocytes, as
plant failure is still possible for patients with no allergy to the presence of a potential allergen slows migration down.
titanium, although the association has not been assessed However, the procedures of LTT or leukocyte migration
in the field of neurosurgery. In this study, a prospective inhibition testing are more complicated than patch testing,
multicenter investigation was conducted that involved 207 and the results sometimes are questionably reproducible
patients underwent both titanium cranioplasty and preim- and not clearly relevant.20 Therefore, although patch testing
plantation patch testing. Our study indicates that titanium may not be the ultimate gold standard for internal metal
prostheses are overall stable and safe for cranioplasty, and hypersensitivity reactions, it remains the most clinically
none of the recruited patients showed allergies to titanium. used test.
Eleven (5.31%) patients experienced delayed scalp erosion Implant failure is an important complication of cranio-
and implant exposure, and these patients underwent addi- plasty but with a relatively low incidence. During the pe-
tional operations for implant removal. Further investiga- riod of this study, only 11 of 207 patients developed unsat-
tions indicated that positive reactions in patch testing, par- isfactory outcomes. This leads to a limitation of the study
ticularly multiple metal hypersensitivities, were associated that some powers of the statistical calculations were less
with titanium implant failure. than 0.80. Randomized controlled trials with larger sample
There have been ethics concerns regarding whether sizes have been proposed to better evaluate how patch test-
patients with metal allergies should receive titanium cra- ing could sensitize individuals to react to a metal implant.
nioplasty. Although metal hypersensitivity has previously However, based on the knowledge that patients with metal
been noted in orthopedic, intravascular, gynecological, and allergies are at higher risk of implant failure, ethics issues
dental devices, there has been no established guideline on may arise when conducting further randomized controlled
whether implantation should be avoided. In the field of neu- trials. Instead, comparative effectiveness research may be
rosurgery, this issue has never been investigated. For many an option.
years, titanium plates have been routinely used for cranio-
plasty around the world, and preoperational patch testing is Conclusions
not performed at most institutions. In this study, allergies
to titanium were not detected. The results of patch tests and Based on our findings, we suggest routine patch test-
the possibility of implant failure were well explained to all ing for all patients who will undergo titanium cranioplasty.
included patients. Patients were free to choose titanium Attention shall be paid to those with allergies to multiple
plates or other prostheses, such as PEEK, for cranioplasty. metals. For these patients, PEEK prostheses are an option
Another key issue to consider is whether the implant for cranioplasty, although these implants are much more
failures are caused by infections rather than hypersensitiv- expansive than titanium plates.
ity. In this study, signs of infections or impaired wound
healing were not detected during the perioperative peri- Acknowledgments
ods of cranioplasty. Stitches were removed within 2 weeks This work was supported by the National Natural Science
after cranioplasty in all patients. The earliest implant ex- Foundation of China (NSFC grants 81471241, 81301339, 81671200

6 J Neurosurg July 6, 2018

 Sun et al.

and 81271374), and The Science and Technology Commission of zadeh K, et al: Outcomes following polyetheretherketone
Shanghai Municipality Project (16411955300, 18441903300). (PEEK) cranioplasty: systematic review and meta-analysis. J
Clin Neurosci 41:30–35, 2017
16. Reddy S, Khalifian S, Flores JM, Bellamy J, Manson PN,
References Rodriguez ED, et al: Clinical outcomes in cranioplasty: risk
1. Agner C, Dujovny M, Gaviria M: Neurocognitive assess- factors and choice of reconstructive material. Plast Reconstr
ment before and after cranioplasty. Acta Neurochir (Wien) Surg 133:864–873, 2014
144:1033–1040, 2002 17. Schalock PC, Thyssen JP: Patch testers’ opinions regarding
2. Al-Tamimi YZ, Sinha P, Trivedi M, Robson C, Al-Musawi diagnostic criteria for metal hypersensitivity reactions to me-
TA, Hossain N, et al: Comparison of acrylic and titanium tallic implants. Dermatitis 24:183–185, 2013
cranioplasty. Br J Neurosurg 26:510–513, 2012 18. Teo ZWW, Schalock PC: Hypersensitivity reactions to im-
3. Corliss B, Gooldy T, Vaziri S, Kubilis P, Murad G, Fargen K: planted metal devices: facts and fictions. J Investig Allergol
Complications after in vivo and ex vivo autologous bone flap Clin Immunol 26:279–294, 2016
storage for cranioplasty: a comparative analysis of the litera- 19. Thien A, King NK, Ang BT, Wang E, Ng I: Comparison of
ture. World Neurosurg 96:510–515, 2016 polyetheretherketone and titanium cranioplasty after decom-
4. Crawford GH: The role of patch testing in the evaluation of pressive craniectomy. World Neurosurg 83:176–180, 2015
orthopedic implant-related adverse effects: current evidence 20. Thyssen JP, Menné T, Schalock PC, Taylor JS, Maibach HI:
does not support broad use. Dermatitis 24:99–103, 2013 Pragmatic approach to the clinical work-up of patients with
5. Di Stefano C, Sturiale C, Trentini P, Bonora R, Rossi D, putative allergic disease to metallic orthopaedic implants
Cervigni G, et al: Unexpected neuropsychological improve- before and after surgery. Br J Dermatol 164:473–478, 2011
ment after cranioplasty: a case series study. Br J Neurosurg 21. Wiggins A, Austerberry R, Morrison D, Ho KM, Honeybul
26:827–831, 2012 S: Cranioplasty with custom-made titanium plates—14 years
6. Foussereau J, Laugier P: Allergic eczemas from metallic for- experience. Neurosurgery 72:248–256, 2013
eign bodies. Trans St Johns Hosp Dermatol Soc 52:220– 22. Williams LR, Fan KF, Bentley RP: Custom-made titanium
225, 1966 cranioplasty: early and late complications of 151 cranioplas-
7. Granchi D, Cenni E, Giunti A, Baldini N: Metal hypersensi- ties and review of the literature. Int J Oral Maxillofac Surg
tivity testing in patients undergoing joint replacement: a sys- 44:599–608, 2015
tematic review. J Bone Joint Surg Br 94:1126–1134, 2012 23. Wu X, Hu J, Zhuo L, Fu C, Hui G, Wang Y, et al: Epidemiol-
8. Hallab N, Merritt K, Jacobs JJ: Metal sensitivity in patients ogy of traumatic brain injury in eastern China, 2004: a pro-
with orthopaedic implants. J Bone Joint Surg Am 83- spective large case study. J Trauma 64:1313–1319, 2008
A:428–436, 2001
9. Iwama T, Yamada J, Imai S, Shinoda J, Funakoshi T, Sakai
N: The use of frozen autogenous bone flaps in delayed cra- Disclosures
nioplasty revisited. Neurosurgery 52:591–596, 2003 The authors report no conflict of interest concerning the materi-
10. Jelcic N, De Pellegrin S, Cecchin D, Della Puppa A, Cagnin als or methods used in this study or the findings specified in this
A: Cognitive improvement after cranioplasty: a possible paper.
volume transmission-related effect. Acta Neurochir (Wien)
155:1597–1599, 2013 Author Contributions
11. Luque I, Leyva L, José Torres M, Rosal M, Mayorga C, Se-
gura JM, et al: In vitro T-cell responses to beta-lactam drugs Conception and design: J Hu, Sun. Acquisition of data: J Hu,
in immediate and nonimmediate allergic reactions. Allergy Sun, Y Hu, Yuan. Analysis and interpretation of data: Sun, Y Hu,
56:611–618, 2001 Yuan, J Yu. Drafting the article: Sun, Yuan, J Yu. Critically revis-
12. Morton RP, Abecassis IJ, Hanson JF, Barber J, Nerva JD, ing the article: J Hu. Reviewed submitted version of manuscript:
Emerson SN, et al: Predictors of infection after 754 cranio- Sun, Y Hu, Yuan, J Yu, Xing Wu. Statistical analysis: Sun, Y
plasty operations and the value of intraoperative cultures for Hu, Yuan, J Yu. Administrative/technical/material support: Sun,
cryopreserved bone flaps. J Neurosurg 125:766–770, 2016 Y Hu, Xing Wu, Du, Xuehai Wu. Study supervision: Sun, Y Hu,
13. Ng ZY, Ang WJ, Nawaz I: Computer-designed polyether- Xing Wu, Du, Xuehai Wu.
etherketone implants versus titanium mesh (± acrylic cement)
in alloplastic cranioplasty: a retrospective single-surgeon, Correspondence
single-center study. J Craniofac Surg 25:e185–e189, 2014 Jin Hu: Huashan Hospital, Fudan University, Shanghai, China.
14. Nyfeler B, Pichler WJ: The lymphocyte transformation test hujindn@126.com.
for the diagnosis of drug allergy: sensitivity and specificity.
Clin Exp Allergy 27:175–181, 1997
15. Punchak M, Chung LK, Lagman C, Bui TT, Lazareff J, Rez-

J Neurosurg July 6, 2018 7

View publication stats