Content uploaded by Gianluca Piatelli
Author content
All content in this area was uploaded by Gianluca Piatelli on Feb 14, 2016
Content may be subject to copyright.
TECHNICAL NOTE
Cranioplasty is well established as a reconstructive
procedure in restoring skull anatomy and repairing
skull defects. The causes of skull defects are well
known and may be acquired or congenital. Skull defects
may produce disturbing deformity, lack of brain protec-
tion, and a variety of symptoms such as chronic headaches
and mild developmental delay in young children. In ad-
dition, the repair of cranial defects could provide relief
to psychological drawbacks and increase social perfor-
mance. Most commonly, cranioplasty in children younger
than 3 years of age is performed to correct growing skull
fractures and congenital anomalies, while in all age groups
tumor removal and previous decompressive craniectomies
are the most frequent reason for the defect.1
ABBREVIATIONS PMMA = polymethylmethacrylate.
SUBMITTED August 18, 2015. ACCEPTED October 23, 2015.
INCLUDE WHEN CITING Published online January 29, 2016; DOI: 10.3171/2015.10.PEDS15489.
* Drs. Fiaschi and Pavanello contributed equally to this work.
Surgical results of cranioplasty with a
polymethylmethacrylate customized cranial implant in
pediatric patients: a single-center experience
*Pietro Fiaschi, MD,2,5 Marco Pavanello, MD,1 Alessia Imperato, MD,1 Villiam Dallolio, MD,3
Andrea Accogli, MD,1,4 Valeria Capra, MD,1 Alessandro Consales, MD,1 Armando Cama, MD,1 and
Gianluca Piatelli, MD1
1Department of Pediatric Neurosurgery, Istituto Giannina Gaslini, Genoa; 2Department of Neurosurgery, San Martino - IST
University Hospital (IRCCS), Genoa; 3Department of Neurosurgery, ICCS Hospital, Milan; 4University of Genoa; and 5Department
of Neuroscience, University of Turin, Italy
OBJECTIVE Cranioplasty is a reconstructive procedure used to restore skull anatomy and repair skull defects. Optimal
skull reconstruction is a challenge for neurosurgeons, and the strategy used to achieve the best result remains a topic of
debate, especially in pediatric patients for whom the continuing skull growth makes the choice of material more difcult.
When the native bone ap, which is universally accepted as the preferred option in pediatric patients, is unavailable, the
authors’ choice of prosthetic material is a polymethylmethacrylate (PMMA) implant designed using a custom-made tech-
nique. In this paper the authors present the results of their clinical series of 12 custom-made PMMA implants in pediatric
patients.
METHODS A retrospective study of the patients who had undergone cranioplasty at Gaslini Children’s Hospital between
2006 and 2013 was conducted. A total of 12 consecutive cranioplasties in 12 patients was reviewed, in which a patient-
specic PMMA implant was manufactured using a virtual 3D model and then transformed into a physical model using
selective laser sintering or 3D printing. All patients or parents were administered a questionnaire to assess how the pa-
tient/parent judged the aesthetic result.
RESULTS Patient age at craniectomy ranged from 5 months to 12.5 years, with a mean age of 84.33 months at cra-
nioplasty. The mean extension of the custom-made plastic was 56.83 cm2. The mean time between craniectomy and
cranioplasty was 9.25 months. The mean follow-up duration was 55.7 months. No major complications were recorded; 3
patients experienced minor/moderate complications (prosthesis dislocation, granuloma formation, and uid collection).
CONCLUSIONS In this patient series, PMMA resulted in an extremely low complication rate and the custom-made
technique was associated with an excellent grade of patient or parent satisfaction on long-term follow up.
http://thejns.org/doi/abs/10.3171/2015.10.PEDS15489
KEY WORDS cranioplasty; polymethylmethacrylate; custom-made technique; pediatric reconstructive surgery
©AANS, 2016 J Neurosurg Pediatr January 29, 2016 1
P. Fiaschi et al.
J Neurosurg Pediatr January 29, 20162
The optimal skull reconstruction remains a challenge
for neurosurgeons, especially in pediatric patients, in
whom the strategy choice, commonly based on the fac-
tors of biocompatibility and cosmetic result, is made even
more difcult by the continuing skull growth with age. Re-
garding biocompatibility, many autograft, xenograft, and
allograft materials have now been used for cranioplasty.
Many characteristics have been suggested to describe the
ideal material for cranioplasty: biocompatibility features
such as tissue tolerance, simplicity of manufacture, ease of
sterilization, low thermal conductivity, radiolucency, light
weight and biomechanical reliability, resistance to infec-
tions, no dilatability with heat, low cost, ready to use, etc.,
but there is no perfect material that ts all of these criteria.
The various implant materials in use today are either au-
tografts or allografts.
Autologous bone grafts offer superior resistance to in-
fection and a decreased likelihood of extrusion but suffer
from variable resorption, difculties in reshaping, and do-
nor-site morbidities.3,9,10,20 While autologous bone is widely
used and favored in reconstructive procedures, synthetic
alternatives are indicated in select cases, such those with
severe bone graft resorption, bone comminution, infection,
and limited donor site options. Over time, metals, ceram-
ics, plastics, and recently, resorbable polymers and bioma-
terials have been used in craniofacial reconstructions.
Polymethylmethacrylate (PMMA) is a thermoplastic
and transparent plastic and is the most frequently used
material for allogenic cranial reconstructions with long-
term results. Zander, in 1940, was the rst physician to
implant a methylmethacrylate prosthetic into a patient.17
Chemically, PMMA is commonly called acrylic glass or
plexiglass. Methylmethacrylate proved to be superior to
metals because of its light weight, low cost, malleability,
radiolucency, and lack of thermoconduction. A disadvan-
tage of PMMA is that it behaves in a brittle manner when
loaded, especially under a strong impact force. Another
disadvantage of this cement is that during polymerization
it heats up to approximately 70°C, therefore it is manda-
tory to remove the modeled PMMA cranioplastic in a de-
formable status with a small but residual risk of modifying
the exact tting of the implant. If the polymerization pro-
cess takes place in situ, the surrounding tissue (dura mater
or cortex) could be severely damaged due to the heat.2 The
major advantage of PMMA is the exible intraoperative
application and the unlimited possibilities of adaptation
to individual anatomy. In the majority of all cases PMMA
cranioplasties are performed freehand. Solid PMMA is
nonporous and does not allow for the ingrowth of native
tissue; consequently, the integration with the surrounding
tissue often necessitates the use of hardware to immobi-
lize the implant. Porous PMMA is a modication of tra-
ditional PMMA that improved this aspect. It is composed
of acrylic bone cement and an aqueous carboxymethyl
cellulose gel. Histological analysis demonstrated that both
hard and soft tissues are able to “grow” into the pores of
the cement, thereby anchoring the implant to native struc-
tures, with evidence of bone ingrowth up to 4.5 mm into
the prosthesis.5 In addition, porous PMMA, despite having
30%–40% less material than nonporous PMMA, offered
no less impact resistance.7
Methods
A retrospective study was conducted of the hospital re-
cords of patients who had undergone cranioplasty at Gasli-
ni Children’s Hospital between 2006 and 2013 (Table 1).
A total of 12 consecutive cranioplasties using nonporous
PMMA (SIAD Healthcare Tecres Cranos) in 12 patients
were reviewed. The custom-made implants cost approxi-
mately 5 times as much as a noncustomized implant with
PMMA. With the purpose of manufacturing a PMMA
prosthesis using computer-aided design/computer-aided
manufacturing, a digital subtraction mirror image tech-
nique with 1-mm-thick spiral CT scan data (using infor-
mation from the contralateral side of the defect) was used.
We considered the following contraindications for cra-
nioplasty: the presence of hydrocephalus, infection, and
brain swelling. The DICOM data were then downloaded
to medical imaging visualization software for editing and
3D reconstruction to obtain a virtual 3D model of the skull.
Starting from a segmentation process, data on cranial
curvature are obtained. With information about the curves
of the shape of the affected side and the contralateral side
of the skull, a model of closure of the defect is proposed.
The curves obtained are then used as a framework on
which the virtual solid model is produced. The virtual
solid model is fused with the virtual model of the skull, to
verify the correct coupling.
The virtual model is then read from a prototyping ma-
chine that builds the solid model section by section. This
model is used to create the mold for the nal step of the
process. The PMMA is injected into the mold to obtain
(after the polymerization) the nal result ready to be ster-
ilized (Figs. 1–3). All patients or parents provided written
informed consent for this study.
The surgical procedure consists of reopening and fol-
lowing the previous incision and separating the epicranial
tissues to the dura mater and to the osseous contour. In
cases in which the prosthesis does not t perfectly on the
skull defect, it can be modied by drilling. The xation is
performed with mini plates or silk thread. To avoid epi-
dural uid collection the rm suspension of the dura mater
is crucial. All patients were strictly followed-up with se-
rial clinical evaluations at 7, 14, 21, and 28 days, a CT scan
at 6 months after the surgery, and clinical evaluation 12
months later. All families were administered a question-
naire (proposed by Fischer et al.)8 to assess how the pa-
tient (or parent in very young patients) judged the aesthetic
result (Table 2). The study was performed with approval
from the Giannina Gaslini Institutional Review Board.
Results
Patient age at craniectomy ranged from 5 months to
12.5 years (mean 75.08 months) with a mean age of 84.33
months at cranioplasty. The mean time between craniecto-
my and cranioplasty was 9.25 months. There were 6 boys
and 6 girls. All data concerning sex, age, etiology, and de-
fect dimension were recorded (Table 1). The mean follow-
up time was 55.7 months, ranging between 24 months (in
one of the oldest patients of the series, almost 11 years old)
and 96 months (in one of our youngest patients, 14 months
old). Two patients were previously treated elsewhere with
Cranioplasty with PMMA customized cranial implant
J Neurosurg Pediatr January 29, 2016 3
2 different techniques to cover the bone defect: 1 patient
(Case 8) had parietal bone splitting that underwent resorp-
tion after 8 months, and the other patient (Case 9) had an
initial acrylic resin cranioplasty that cracked after minor
head trauma. All complications are reported in Table 1; all
but 2 patients were treated conservatively.
One patient incurred prosthesis dislocation after a head
trauma (Case 6). He underwent a new surgical procedure
to replace the prosthesis. He was 39 months old at the time
of the trauma, and only 5 months had passed since the cra-
nioplasty, thus remanufacturing on the preexisting model
was performed.
In another case (Case 9) the prosthesis underwent spon-
taneous dislocation; in the previous procedure, the pros-
thesis was xed with absorbable sutures. At the second
intervention, the same prosthesis was used and xed with
resorbable plates and screws. The same patient experi-
enced a cutaneous foreign body granuloma, as a reaction
to intradermal absorbable suture. He consequently under-
went a cutaneous granuloma removal procedure and skin
closure with silk sutures. Case 4 showed an extradural and
subcutaneous uid collection (probably due to the lack of
surgical drain) with spontaneous recovery after a period
of compressive bandaging.
The mean extension of the custom-made plastic was
56.83 cm2, with the smallest extension only 6 cm2 and the
largest reaching 98 cm2. The aesthetic results, assessed
through the questionnaire noted above, were judged as
satisfying to very satisfying by all but 1 patient (Case 9,
wound revision due to a granuloma), with the implant size
estimated as medium in 7 cases, small in 4 cases, and
large in 1 case; the grade of satisfaction did not change
over time except for Cases 6 and 9, whose implants were
removed and replaced after dislocation. Medical compli-
cations related to cranioplasty are reported in Table 1.
Discussion
Cranioplasty is a reconstructive procedure used to
restore skull anatomy and repair skull defects. Optimal
skull reconstruction is a challenge for neurosurgeons and
the strategy to achieve the best result remains a topic of
debate. There are 3 main determinants that inuence the
choice. The rst is biocompatibility: many features should
be evaluated, such as tissue tolerance, allergic reaction,
early vascularization, possibility of intraoperative ther-
mic damage, thermal conductivity, and resistance to in-
fections. The second determinant is the cosmetic result.
Regarding the aesthetic outcome, an important innovation
in the cranioplasty technique had been achieved with the
novel method of rapid prototyping, developed in the past
few decades. This method offers the possibility of pre-
operatively forming various materials into custom-made
implants to precisely t each individual’s cranial defect.
This procedure has been demonstrated to not signi-
cantly increase the overall treatment cost, together with
TABLE 1. Patients who underwent cranioplasty at Gaslini Children’s Hospital (2006–2013)
Case
No. Sex Diagnosis Location
Age at
Craniectomy
(mos)
Age at
Cranioplasty
(mos) Other
Cranioplasty
Defect
Dimension
(cm2)Complications FU
(mos)
1 F Meningioma Frontoparietal
(bilat) 140 148 —82 None 84
2 F Osteoma Lt parietooccipital 14 34 —48 None 96
3 F TBI Lt parietal 10 20 —49 None 80
4 F Plagiocephaly Lt frontoparietal 70 75 —35 Intracranial hypotension,
uid collections
73
5 M TBI Rt parietal 100 109 —55 None 55
6 M Trigonocephaly Frontal (bilat) 534 Later acrylic resin 80 After 6 mos: head injury,
prosthesis dislocation,
re moval & new acrylic
implant
58
7 F Fibrous dysplasia Lt parietal 150 155 —61 None 53
8 M TBI Lt frontotemporal 68 79 Previous parietal
bone splitting
(resorption)
87 EEG alterations, anticonvul-
sant therapy, no crisis 47
9 M Temporal AVM
+ aneurysm
(intracerebral
hematoma)
Lt frontotemporo-
occipital 54 58 Previous acrylic
resin (rupture) 50 After 2 mos prothesis
dislocation, xation w/
resorbable plates &
screws; after 23 mos
granuloma formation,
wound revision
45
10 MFibrous dysplasia Lt occipital 59 62 —31 None 28
11 FFibrous dysplasia Rt hemispheric 130 130 —98 None 24
12 MHistiocytosis Rt pterional 101 108 — 6 None 25
AVM = arteriovenous malformation; EEG = electroencephalography; FU = follow-up; TBI = traumatic brain injury.
P. Fiaschi et al.
J Neurosurg Pediatr January 29, 20164
a signicant patient benet, compared with those patients
who underwent autologous bone cranioplasty11 and con-
sequently can be reasonably considered the gold standard
cranial defect repair method when the native bone ap
is unavailable. To produce a custom-made prefabricated
PMMA prosthesis (such as in our patients) manufactured
using CAD/CAM, a preoperative, transversal, 1-mm spiral
CT scan with 3D reconstruction of the cranium has to be
performed before surgery. Each implant then undergoes
a sterilization process.16 In addition to the signicant ad-
vantage of improving the cosmetic result (mostly in large
skull defects), custom-made cranioplasty implantation im-
plies a shorter operative time, a positive effect on the heal-
ing process, less invasiveness and low bleeding risk, fewer
infectious complications, no donor site morbidity from the
use of allografts, and faster recuperation compared with
intraoperatively molded cranioplasty surgery. In addition,
in the case of infection or early trauma, the customized
implant may be easily remanufactured on the preexisting
model.6,15,18,19,21 Application of the custom-made technique
does, however, have a few disadvantages. First, it is rela-
tively expensive: it costs approximately 5 times more than
a noncustomized implant with the same material. Second,
the problem of temporal muscle atrophy after a crani-
ectomy also persists using a custom-made cranioplasty
implant, which causes some asymmetry in the temporal
regions. Third, custom manufacturing of a prefabricated
PMMA prosthesis is very time consuming.12 While in
adult patients the main determinants of material choice
are biocompatibility and cosmetic results, in pediatric pa-
tients a third concern is the continuing skull growth with
age, especially in young children. The major skull growth
occurs in the rst 2 years, achieving about 84% of the
adult size, with a strong deceleration after that age. Even
FIG. 1. Case 1. Preoperative CT (upper row) and MR images (lower row) of a 140-month-old patient affected by a falcine menin-
gioma with hyperostotic deformation of the skull.
TABLE 2. Five questions asked to assess how the patient (or parent, in very young patients) judged the aesthetic result of the procedure*
1. Please estimate the size of your implant (small/medium/large)†
2. Choose one of the following statements, which best describes your satisfaction with the aesthetic result of cranioplasty:
a. I do not accept the aesthetic result after cranioplasty & would like to improve the appearance with another surgical intervention
b. I am not satised with the aesthetic result
c. I am satised with the aesthetic result
d. I am very satised with the aesthetic result & think that the cranioplasty does not impair my appearance at all
3. If you are dissatised, please indicate the reason for your dissatisfaction: e.g., dents, bulges, scars, or bulging of bone edges
4. Did your grade of satisfaction change over time after cranioplasty?
5. Did you have any medical complications after cranioplasty?
* As proposed by Fischer et al.8
† Implant size was assessed objectively from radiological images and postoperative treatment from patient records. The slice of greatest cross-sectional area of the
defect was chosen for size analysis: small, medium, and large defects were dened as ≤ 50 cm2, 50–100 cm2, and ≥ 100 cm2, respectively.
Cranioplasty with PMMA customized cranial implant
J Neurosurg Pediatr January 29, 2016 5
in the youngest patients we did not encounter a lack of
ability of the prosthesis (made of nonporous PMMA) to
become integrated over time, consistent with the growth
of the pediatric skeleton. Porous PMMA is a modication
of traditional PMMA that further improves this aspect,
with its particular architecture allowing ingrowth of na-
tive tissue, resulting in a better xation of the prosthesis.
The 3D pore structure serves as an effective “scaffold”
for the attachment of osteoprogenitor cells, and as a sub-
strate for deposition of new extracellular matrix material.13
Acrylic cranioplasty is generally well tolerated in the adult
population, despite an overall complication rate ranging
between 5% and 25% and an infection rate of 5%–20%
in different series.4,5,14 Conversely, hydroxyapatite cranio-
plasty was found to have the highest complication rate,
mostly due to infection and material exposure.
When the native bone ap (universally accepted as the
preferred option in pediatric patients) is unavailable, our
choice of prosthetic material is PMMA; in addition to the
above-mentioned explanation, we prefer it for many other
reasons: it is one of the most biocompatible alloplastic ma-
terials currently available, with a low rate of foreign body
reaction, and provides adequate protection for the under-
lying neural tissues, comparable to that of native osseous
tissue. Furthermore, the custom-made technique offers
important advantages, such as an implant that perfectly
ts to the bone defect, the avoidance of exposing intra-
cranial tissue to the heat of polymerization, the absence
of monomer residue and dust produced during intraopera-
tive molding, and easy remanufacturing on the preexisting
model in cases of infection or early trauma. In addition,
we determined the already mentioned shorter operative
time to be of particular importance in children due to de-
creased intraoperative blood loss.
In our patient series the choice of implant material was
evaluated on a case-by-case basis and discussed with the
parents, who were informed about the features, benets,
and risks of the available materials. None of the proce-
dures we performed caused symptomatic foreign body
reactions, nor an infection rate. In addition to the intrinsic
porus architecture of PMMA, we believe that the surgical
lling of the prosthesis with periosteal tissue may help fu-
ture bone growth as a “scaffold” for the osseous margins
of the cranial gap. The mean follow-up duration time was
55.7 months, with two of the longest follow-up durations in
the second- and third-youngest patients (10 and 14 months
old, respectively). As reported in the questionnaire, these
2 patients did not experience any complications or residual
skull defects, with a high level of cosmetic satisfaction.
Conclusions
A large skull defect mandates subsequent cranioplasty.
This procedure is full of possible complications. Our choice
of implanting a custom-made PMMA prosthesis was accu-
rately discussed with the parents when the native bone ap
was unavailable. In our series, use of PMMA resulted in
an extremely low complication rate and the custom-made
technique was associated with an excellent grade of patient
or parent satisfaction on long-term follow-up.
Acknowledgements
We are grateful to Dr. A. Rossi for his contribution and the
images provided.
FIG. 2. Case 1. Postoperative CT scan (A) with 3D reconstruction (B)
and MR images (C) 3 months after custom-made prosthesis implanta-
tion.
FIG. 3. Case 1. Intraoperative images at the age of 148 months, includ-
ing the custom-made prosthesis before implantation (right) and how it
perfectly ts to the skull defect (left). Figure is available in color online
only.
P. Fiaschi et al.
J Neurosurg Pediatr January 29, 20166
References
1. Aydin S, Kucukyuruk B, Abuzayed B, Aydin S, Sanus GZ:
Cranioplasty: Review of materials and techniques. J Neuro-
sci Rural Pract 2:162–167, 2011
2. Azmi A, Latiff AZ, Johari A: Methyl methacrylate cranio-
plast y. Med J Malaysia 59:418–421, 2004
3. Baumeister S, Peek A, Friedman A, Levin LS, Marcus JR:
Management of postneurosurgical bone ap loss caused by
infection. Plast Reconstr Surg 122:195e–208e, 2008
4. Blum KS, Schneider SJ, Rosenthal AD: Methyl methacrylate
cranioplasty in children: long-term results. Pediatr Neuro-
surg 26:33–35, 1997
5. Bruens ML, Pieterman H, de Wijn JR, Vaandrager JM: Po-
rous polymethylmethacrylate as bone substitute in the cranio-
facial area. J Craniofac Surg 14:63–68, 2003
6. Dean D, Min KJ, Bond A: Computer aided design of large-
format prefabricated cranial plates. J Craniofac Surg
14:819–832, 2003
7. Eppley BL: Biomechanical testing of alloplastic PMMA cra-
nioplasty materials. J Craniofac Surg 16:140 –143, 2005
8. Fischer CM, Burkhardt JK, Sarnthein J, Bernays RL,
Bozinov O: Aesthetic outcome in patients after polymethyl-
methacrylate (PMMA) cranioplasty—a questionnaire-based
single-centre study. Neurol Res 34:281–285, 2012
9. Flannery T, McConnell RS: Cranioplasty: why throw the
bone ap out? Br J Neurosurg 15:518–520, 2001
10. Gao LL, Rogers GF, Clune JE, Proctor MR, Meara JG, Mul-
liken JB, et al: Autologous cranial particulate bone grafting
reduces the frequency of osseous defects after cranial expan-
sion. J Craniofac Surg 21:318–322, 2010
11. Gilardino MS, Karunanayake M, Al-Humsi T, Izadpanah A,
Al-Ajmi H, Marcoux J, et al: A comparison and cost analysis
of cranioplasty techniques: autologous bone versus custom
computer-generated implants. J Craniofac Surg 26:113–117,
2015
12. Lee SC, Wu CT, Lee ST, Chen PJ: Cranioplasty using poly-
methyl methacrylate prostheses. J Clin Neurosci 16:56– 63,
2009
13. Menderes A, Baytekin C, Topcu A, Yilmaz M, Barutcu A:
Craniofacial reconstruction with high-density porous poly-
ethylene implants. J Craniofac Surg 15:719–724, 2004
14. Moreira-Gonzalez A, Jackson IT, Miyawaki T, Barakat K,
DiNick V: Clinical outcome in cranioplasty: critical review
in long-term follow-up. J Craniofac Surg 14:144 –153, 2003
15. Rotaru H, Baciut M, Stan H, Bran S, Chezan H, Iosif A, et al:
Silicone rubber mould cast polyethylmethacrylate-hydroxy-
apatite plate used for repairing a large skull defect. J Cranio-
maxillofac Surg 34:242–246, 2006
16. Rotaru H, Stan H, Florian IS, Schumacher R, Park YT,
Kim SG, et al: Cranioplasty with custom-made implants:
analyzing the cases of 10 patients. J Oral Maxillofac Surg
70:e169– e176, 2012
17. Sanan A, Haines SJ: Repairing holes in the head: a histor y of
cranioplasty. Neurosurgery 40:588–603, 1997
18. Solaro P, Pierangeli E, Pizzoni C, Bof P, Scalese G: From
computerized tomography data processing to rapid manu-
facturing of custom-made prostheses for cranioplasty. Case
report. J Neurosurg Sci 52:113–116, 2008
19. van Putten MC Jr, Yamada S: Alloplastic cranial implants
made from computed tomographic scan-generated casts. J
Prosthet Dent 68:103–108, 1992
20. Weber RS, Kearns DB, Smith RJ: Split calvarium cranio-
plast y. Arch Otolaryngol Head Neck Surg 113:84 –89, 1987
21. Wulf J, Busch LC, Golz T, Knopp U, Giese A, Ssenyonjo H,
et al: CAD generated mold for preoperative implant fabrica-
tion in cranioplasty. Stud Health Technol Inform 111:608–
610, 2005
Disclosures
The authors report no conflict of interest concerning the materi-
als or methods used in this study or the findings specified in this
paper.
Author Contributions
Conception and design: Fiaschi, Pavanello, Imperato, Dallolio,
Piatelli. Acquisition of data: Consales. Analysis and interpretation
of data: Consales. Drafting the article: Accogli, Capra, Cama.
Study supervision: Capra.
Correspondence
Pietro Fiaschi, Department of Neurosurgery, San Martino - IST
University Hospital (IRCCS), Largo Rosanna Benzi 10, Genoa
16132, Italy. email: pietro.fiaschi@alice.it.