Articles
www.thelancet.com Published online March 8, 2011 DOI:10.1016/S0140-6736(10)62354-9 1
Published Online
March 8, 2011
DOI:10.1016/S0140-
6736(10)62354-9
See Online/Comment
DOI:10.1016/S0140-
6736(11)60312-7
Wake Forest Institute for
Regenerative Medicine and
Department of Urology, Wake
Forest University School of
Medicine, Winston-Salem, NC,
USA (A Raya-Rivera MD,
J J Yoo MD, D R Esquiliano MD,
S Soker PhD, Prof A Atala MD);
and HIMFG Tissue Engineering
Laboratory, Metropolitan
Autonomous University and
Centro de Investigación y de
Estudios Avanzados del
Instituto Politécnico Nacional
(CINVESTAV-IPN), Mexico City,
Mexico (A Raya-Rivera,
D R Esquiliano,
Prof E Lopez-Bayghen PhD)
Correspondence to:
Prof Anthony Atala, Wake Forest
Institute for Regenerative
Medicine, Wake Forest University
School of Medicine, Winston
Salem, NC 27157, USA
aatala@wfubmc.edu
Introduction
Complex urethral problems can be caused by injury, which
can lead to an inability to void. Catheters might need to be
inserted below the abdomen in the suprapubic region for
adequate bladder emptying, because often the eff ects of
the trauma on the involved tissues need to be minimised.
Some patients with urethral strictures repeatedly have
infections, experience straining and major discomfort,
and have blood in their urine. Urethral functional
inadequacy can occur in patients with pelvic fractures or
straddle trauma; previous manipulation with indwelling
catheters, endoscopy, or surgery; and congenital or
acquired disease.
Reconstructive techniques depend on the urethral
defect location, length, and the surgeon’s preference
and previous experience. Short, non-complex urethral
defects can be repaired with an end-to-end anastomosis
by aligning and joining the normal urethral ends.1 For
long defects, surgeons might need to do a pubectomy,
to gain better access to the damaged tissue and to help
shorten the urethral gap.2 An onlay repair, in which
about half the strictured circumference portion of the
tubular urethra is replaced with a tissue graft (eg, skin
or buccal mucosa), is often used for damaged urethras
that have a healthy urethral bed.3,4 Tubularised tissue
grafts might be needed for complex or long urethral
defects, but have a high proportion of failures
(sometimes over 50%).5
Regenerative medicine might help to overcome some
of the drawbacks associated with the native tissues that
are used for urologic reconstruction.6,7 However,
engineered tubularised constructs (eg, urethras or small
blood vessels) tend to stricture over time.5,8–10 We aimed to
assess whether engineered autologous tubularised
urethral tissue could be used as an alternative method for
the treatment of complex posterior urethral defects.
Methods
Patients
Five boys with urethral defects were invited to participate
in this study of engineered urethras at the Federico
Gomez Children’s Hospital in Mexico City, Mexico.
Three patients presented with a complete posterior
urethral disruption caused by pelvic trauma and had
substantial widespread injury and two had previous
failed posterior urethral repairs, one with a buccal
mucosa graft and one with a skin graft, both of which
were tubularised.
The study protocol was approved by the hospital’s
investigational review board (protocol number 2002055)
Tissue-engineered autologous urethras for patients who
need reconstruction: an observational study
Atlantida Raya-Rivera, Diego R Esquiliano, James J Yoo, Esther Lopez-Bayghen, Shay Soker, Anthony Atala
Summary
Background Complex urethral problems can occur as a result of injury, disease, or congenital defects and treatment
options are often limited. Urethras, similar to other long tubularised tissues, can stricture after reconstruction. We
aimed to assess the eff ectiveness of tissue-engineered urethras using patients’ own cells in patients who needed
urethral reconstruction.
Methods Five boys who had urethral defects were included in the study. A tissue biopsy was taken from each patient,
and the muscle and epithelial cells were expanded and seeded onto tubularised polyglycolic acid:poly(lactide-co-
glycolide acid) scaff olds. Patients then underwent urethral reconstruction with the tissue-engineered tubularised
urethras. We took patient history, asked patients to complete questionnaires from the International Continence
Society (ICS), and did urine analyses, cystourethroscopy, cystourethrography, and fl ow measurements at 3, 6, 12, 24,
36, 48, 60, and 72 months after surgery. We did serial endoscopic cup biopsies at 3, 12, and 36 months, each time in a
diff erent area of the engineered urethras.
Findings Patients had surgery between March 19, 2004, and July 20, 2007. Follow-up was completed by July 31, 2010.
Median age was 11 years (range 10–14) at time of surgery and median follow-up was 71 months (range 36–76 months).
AE1/AE3, α actin, desmin, and myosin antibodies confi rmed the presence of cells of epithelial and muscle lineages
on all cultures. The median end maximum urinary fl ow rate was 27·1 mL/s (range 16–28), and serial radiographic
and endoscopic studies showed the maintenance of wide urethral calibres without strictures. Urethral biopsies
showed that the engineered grafts had developed a normal appearing architecture by 3 months after implantation.
Interpretation Tubularised urethras can be engineered and remain functional in a clinical setting for up to 6 years.
These engineered urethras can be used in patients who need complex urethral reconstruction.
Funding National Institute of Diabetes and Digestive and Kidney Diseases, National Institutes of Health.
Articles
2 www.thelancet.com Published online March 8, 2011 DOI:10.1016/S0140-6736(10)62354-9
and ethics board. Written informed parental consent was
obtained before acquirement of the tissue biopsy and
4–7 weeks before surgery.
Procedures
All fi ve boys had an engineered urethra created from
autologous cells and underwent urethral tubularised
posterior urethroplasty by the same surgical method
and procedure. Before surgery, the urethral defect length
and characteristics such as location and associated
pelvic path ology were measured by retrograde
urethrography, voiding cystourethrograms, and ureth-
roscopy. We took a detailed history from all patients and
comp leted symptom question naires from the Inter-
national Continence Society (ICS).
We took tissue samples from every child and used
these to engineer the urethras. All tissue samples were
taken by the same person (A Raya-Rivera). A 3 cm
suprapubic transverse incision was made and a bladder
biopsy (1×1 cm) was taken and sent to the approved
manu facturing facility at the Metropolitan Autonomous
University, Mexico City, Mexico, where the urethras
were engineered under regulatory guidelines
(HIM87120BSO) from the Federal Commission on
Safety and Health Protection; Investigation, Ethics, and
Biosafety Division (COFEPRIS). Primary cultures of
smooth muscle and urothelial cells were collected as
described previously.6 Briefl y, cells were col lected by
explanting pieces of muscle tissue and scraping the
uroepithelial tissue onto 10 cm culture plates. The
muscle tissue was placed in Dulbecco’s modifi ed Eagle’s
medium (Invitrogen; Carlsbad, CA, USA) and the
epithelial tissue was placed in keratinocyte growth
medium, both with epidermal growth factor. The cells
were expanded to a density of 1·0–3·0×10⁷ per cm² for
3–6 weeks before seeding.
Muscle and urothelial cells were characterised by
histology and immunohistochemistry before seeding
onto the scaff olds. 5 μm sections of formalin-
fi xed, paraffi n-embedded tissues were processed and
stained with haematoxylin and eosin. We used represen-
tative sections for immunohistochemistry. Uroepithelial
cell layers were identifi ed with broadly reacting
monoclonal mouse anti-human cytokeratins AE1/AE3,
desmin, and myosin (Dako, CA, USA; product codes
M3515, D33, and M3558) antibodies at 1:100, 1:50, and
1:50 dilutions, respectively, for 24 h at 4°C. Smooth
muscle fi bres were labelled with monoclonal α smooth
muscle anti-actin antibodies (Novocastra, Newcastle,
UK) at 1:100 dilution for 24 h at 4°C. Sections were incu-
bated with the Dako Real EnVision Peroxidase/DAB
detec tion system (Dako). All sections were counter-
stained with hematoxylin. Images were taken with a
light microscope.
We seeded cells onto scaff olds as described
previously.7,11,12 Epithelial cells were seeded onto the
luminal surface and muscle cells onto the outer surface
of the tubular scaff olds. Briefl y, the cells were removed
from the culture plate with a 0·05% trypsin solution.
Cells were centrifuged, rinsed with medium with no
additives, and the pellet was re-suspended to a
concentration of 1×10⁷ cells per mL. A biodegradable
mesh made of polyglycolic acid (PGA; Sherwood
Medical, St Louis, MO, USA) was tubularised with
running 5-0 PGA sutures (Ethicon; Piscataway, NJ,
USA) and sized for every individual patient, according
to measurements of urethral defect from radiography,
urethral calibration, and endoscopy. The size of the
engineered urethras ranged from 4 to 6 cm (median
5 cm), with a 16 French diameter. The scaff olds were
constructed in the same manner as used previously,7,12
and the biomaterials were processed with the same
techniques as used previously for engineered bladders
implanted in people.12,13 Briefl y, the mesh was sprayed
with a 50:50 liquid PGA:poly(lactide-co-glycolide acid)
(Sigma; St Louis, MO, USA). The scaff olds were placed
in a vacuum desiccator; samples confi rmed scaff old
purity.13 The scaff olds were further sterilised with
ultraviolet light and sequentially seeded with epithelial
cells within the lumen and smooth muscle cells on the
outer surface. The seeded scaff olds were incubated in
culture for 7 days with Dulbecco’s modifi ed Eagle’s
medium. Overall, construction of the neo-urethras took
4–7 weeks.
We measured cell viability on the seeded grafts with an
MTT (3-[4, 5-dimethylthiazolyl-2]-2, 5-diphenyltetrazolium
bromide) assay.14 After cells were incubated with 1 mg/mL
MTT reagent for about 4 h, an isopropanol:hydrochloric
acid (ratio 1:1) solution was added. The samples were
read with a GENios (Tecan, Durham, NC, USA) ELISA
plate reader (630 nm wavelength). Measure ments from
non-seeded grafts were used as controls.
We did scanning electron microscopy for seeded and
non-seeded scaff olds to assess whether cells had grown
on the scaff olds. Random samples were taken on days 3
and 6 after seeding. Fragments were washed with a
0·1 M sodium cacodylate buff er and fi xed for 2 h in a
gluta raldehyde (2·5%) solution. Samples were post-
fi xed in a 1:100 sodium cacodylate osmium tetroxide
solution and dehydrated in a series of ethanol
concentrations. We dried the preparations, covered
Age
(years)
Primary diagnosis Previous
urethroplasty
Defect site Defect
length
(cm)
Follow-up
(months)
1 10 Motor vehicle accident No Membranous urethra 5 76
2 14 Straddle trauma Buccal mucosa Membranous urethra 6 73
3 11 Motor vehicle accident No Membranous urethra 4 71
4 11 Motor vehicle accident Foreskin Membranous urethra 4 65
5 12 Straddle trauma No Membranous urethra 5 36*
*Patient followed up for 36 months because he was the last patient to enter the study.
Table: Characteristics of patients
Articles
www.thelancet.com Published online March 8, 2011 DOI:10.1016/S0140-6736(10)62354-9 3
them with ionised gold, and examined them with a
JEOL JSM-5300 microscope.
Once the engineered urethras had been made, the
patients underwent surgery. The urethras were
catheterised and exposed through a perineal inverted
semicircular incision. Defects were identifi ed, the
stricture and scar tissue was removed, and the urethral
ends were dissected free from surrounding scar tissue.
The tubularised engineered urethral constructs were
surgically implanted with absorbable sutures. All
engineered urethras were trimmed in the overlap
region with the native urethras. A urethral catheter was
left in place for 2 weeks in the fi rst patient, and for
4 weeks in the other patients. The catheter size ranged
from 12 to 14 French. We took patient history, asked
patients to complete the ICS questionnaires, and did
urine analyses, cystoure throscopy, cystourethrography,
and fl ow measurements at 3, 6, 12, 24, 36, 48, 60, and
Figure 1: Cellular expression of epithelial and muscular markers, and viability assay and scanning electron microscopy assessments
(A) Labelling of smooth muscle and uroepithelial cells in cultures incubated with primary antibodies (positive cells) and without (control cells). All these cell cultures
came from patient 3. (B) Assessment of cell viability on the seeded grafts from every patient. Vaginal cells were assessed as positive controls. Bars represent SD.
(C) Scanning electron microscopy (15 Ku) of seeded scaff olds 6 days after seeding.
Cytokeratins AE1/AE3
α actin
Desmin
Myosin
Positive cells Control cells
A
C
B
PGA/PLGA
non-seeded
grafts
PGA/PLGA
vaginal
cells
Patient 1 Patient 2 Patient 3 Patient 4 Patient 5
0
0·2
0·4
0·6
0·8
1·0
1·2
1·4
O
pt
ica
l d
en
sit
y
at
5
95
n
m
Urothelium
Patient 1 Patient 2 Patient 3 Patient 4 Patient 5
Muscle
Scaffold without cells
10 μm 10 μm 10 μm 10 μm 10 μm
10 μm 10 μm 10 μm 10 μm 10 μm
50 μm
Articles
4 www.thelancet.com Published online March 8, 2011 DOI:10.1016/S0140-6736(10)62354-9
72 months after surgery. We did serial endoscopic cup
biopsies at 3, 12, and 36 months, each time in a diff erent
area of the engineered urethras.
Role of the funding source
The sponsor funded the basic research and its analysis
but had no role in the study design (preparing the
Figure 2: Neo-urethra implantation and clinical outcome
(A) A cell-seeded graft sutured to the normal urethral margins from the fi rst patient. (B) Voiding cystourethrograms of all fi ve patients before surgery (arrows show
the abnormal margins), 12 months after surgery (arrows show margins of tissue engineered urethras), and at last follow-up (arrows show margins of tissue
engineered urethras).
Before surgery
12 months
after surgery
At last follow-up
A
B
Patient 1 Patient 2 Patient 3 Patient 4 Patient 5
Articles
www.thelancet.com Published online March 8, 2011 DOI:10.1016/S0140-6736(10)62354-9 5
protocol), data collection, data analy ses, data inter-
pretation, or writing of the report. All authors had full
access to all the data and had fi nal responsibility for the
decision to submit for publication.
Results
Patients had surgery between March 19, 2004, and
July 20, 2007. Follow-up was completed by July 31, 2010.
The table shows the characteristics of each patient.
Median age was 11 years (range 10–14 years) at the time of
surgery and patients were followed up for a median of
71 months (range 36–76 months) after surgery.
AE1/AE3, α actin, desmin, and myosin antibodies
confi rmed the presence of cells of epithelial and muscle
lineages on all cultures (fi gure 1). There was a substantial
increase in cell growth on the seeded scaff olds after
7 days in standard culture conditions. Scanning electron
microscopy on day 6 of culture showed all scaff old
surfaces were covered with cells.
The engineered urethras were implanted without
intraoperative complications (fi gure 2). None of the
patients had fi stulae or urinary tract infections during
follow-up. Patient history and serial symptom
questionnaires completed by the patients with their Figure 3: Urofl ow analyses
0 3 6 12 24 36 48 60 72
0
5
10
15
20
25
30
35
Fl
ow
ra
te
m
L/
s
Patient 1
Patient 2
Patient 3
Patient 4
Patient 5
Follow-up (months)
Figure 4: Histological and immunocytochemical analyses
Results of urethral biopsies of engineered segments from every patient at 1 year after surgery and from a control sample of healthy urethral tissue. All scale bars
indicate 25 μm.
Patient 1
Patient 2
Patient 3
Patient 4
Patient 5
Control
Haematoxylin and eosin α actin Desmin Mvosin AE1/AE3
Articles
6 www.thelancet.com Published online March 8, 2011 DOI:10.1016/S0140-6736(10)62354-9
parents confi rmed patients’ satisfaction with the results
of the operation, as measured by diff erent parameters
including the absence of dysuria, voiding frequency,
straining, or dribbling.
The fi rst patient, who had their catheter removed
2 weeks after surgery, presented with a decreased
urinary voiding stream 2 weeks later. Cystourethroscopy
showed a narrowing at the proximal superior graft
anastomotic site that needed a transurethral incision
4 weeks after surgery. The patient was able to void well
thereafter, and no further interventions were needed.
Cystourethroscopy confi rmed the radiographic fi ndings
of a patent urethra in all patients (fi gure 2). The
fourth patient had pelvic disruption that involved the
sphincter, and he needed a pubovesical sling after
surgery for his daytime stress incontinence. All patients
are now continent.
Voiding cystourethrograms, which showed the extent
of the defects before surgery, showed the maintenance
of wide urethral calibres without diverticula 12 months
after surgery (fi gure 2). The absence of strictures and
diverticula was confi rmed with serial cystourethroscopy.
Urofl ow studies showed adequate fl ow rates, with a
mean end maximum fl ow rate of 25·1 mL/s, and a
median of 27·1 mL/s (range 16–28) at the last follow-up
(fi gure 3).
Histology of the serial biopsies, confi rmed with
immunohistochemistry, showed that the engineered
urethras appeared to have a normal architecture by
3 months after implantation, and consisted of distin-
guishable layers of epithelia and smooth muscle
(fi gure 4). There were no aberrant histological changes
over time.
Discussion
We have successfully constructed engineered urethras
with autologous cells and implanted them into patients
with urethral defects. All fi ve boys were continent at last
follow-up.
The results of this study are consistent with the results
from experimental studies we did in the early 1990s,
when we seeded biodegradable matrices with autologous
cells.15 In an experimental model in rabbits, autologous
bladder epithelial and smooth muscle cells were grown
and seeded onto preconfi gured tubular matrices.11 Entire
urethral segments were resected and anterior
urethroplasties were done with tubularised matrices,
either seeded with cells or without cells. The tubularised
matrices that were seeded with autologous cells formed
new tissue that was histologically similar to native
urethra. There was poor tissue development, fi brosis,
and stricture formation from those matrices that were
not seeded. The study showed that the use of engineered
urethras with matrices of cells was crucial to the success
of tubularised repairs in the long term. Although the
scaff old alone was not suffi cient for a tubularised repair
and there was fi brosis and scarring present, we noted
that there was limited tissue regeneration at the edges of
the construct. In animal experiments, we noted that the
maximum distance at which native cells from the normal
urethra migrated onto the scaff old and contributed to
adequate tissue regeneration was 0·5 cm.16 This fi nding
led us to assess the use of matrices without cells to
replace only the top part of the tubular urethra, in an on-
lay manner, instead of replacing the entire tubular
urethral circumference.17 Thereafter, we showed that if a
portion of the circumferential urethral tissue is healthy
and can be preserved, off -the-shelf matrices without cells
can be used as on-lay grafts with adequate long-term
outcomes.18–20 Buccal mucosal cell-seeded matrices have
been used as on-lay grafts for urethral disease in patients
with lichen sclerosus.21 Lich