Biomaterials 22 (2001) 1869}1874

Self-expandable chitosan stent: design and preparation

A. Lauto*, M. Ohebshalom, M. Esposito, J. Mingin, P.S. Li, D. Felsen,
M. Goldstein, D.P. Poppas
Center for Pediatric Urology and Minimal Invasive Urologic Surgery, Department of Urology, New York-Presbyterian Hospital-Weill
Medical College of Cornell University, New York, NY 10021, USA
Received 6 June 2000; received in revised form 22 August 2000; accepted 18 October 2000

Abstract

Stents are largely used in surgical procedures to relieve pathological obstructions. The purpose of the present study was to design
and prepare a biocompatible stent with a self-expandable mechanism. Thin "lms were prepared from deacetylated chitosan (4% w/v)
dissolved in acetic acid solution (2% v/v). The chitosan "lms were tested by a calibrated tensiometer to measure the Young's module
(E). The "lms were used to manufacture stents by pulling and winding them around a cylindrical rod in a helical fashion. Thirteen
stents (diameter"0.5$0.05 mm, length &4 mm) were inserted into the vas deferens of wistar rats. Upon stent insertion, the vasal
anastomosis was achieved with a laser-soldering technique. The animals were sacri"cied 8 weeks later. The stress test showed that the
chitosan "lm was elastic (maximum strain"105%$6%, E"0.7655$0.0288 Mpa). The stents self-expanded by releasing their
elastic energy. All the stents but one remained open inside the vasa despite high incidence of sperm granuloma. A biocompatible and
self-expandable stent with a helical design is proposed.  2001 Elsevier Science Ltd. All rights reserved.

Keywords: Elasticity; Cylindrical helix; Biopolymers; Laser solder

1. Introduction implanting rigid stents [1,3,10]. Detrimental complica-

A large variety of stents is currently used in clinical
practice to relieve pathological obstruction of tubular
structures in vascular, urologic and gastroenterologic
systems. Stents can be manufactured with metals or syn-
thetic polymers or biomaterials; they can also have
a rigid or self-expandable structure [1}7]. Temporary
stents, which are made of synthetic materials or bio-
polymers, usually degrade over time once inside the
body. Therefore, they typically do not require surgical
removal after their implantation in patients [6,8]. As
compared to metals and some synthetic materials, bio-
polymers are more biocompatible, less traumatic to sur-
rounding tissue and possibly bene"cial to wound healing
[6,9]. In recent years, self-expandable stents have been
introduced because of their unique property to enlarge
following tissue insertion, which results easier than
* Correspondence address: Department of Medical Physics, Memor-
ial Hospital, room F1134, 1275 York Avenue, New York, NY 10021,
USA. Tel.: 212-639-8835, fax 212-717-3676.
E-mail address: a}lauto@hotmail.com (A. Lauto).
tions secondary to stent insertion have been clinically
described, such as migration, tissue trauma, fragmenta-
tion and encrustation [3,11}14]. Improvements in both
quality and performance of stents are necessary to reduce
clinical complications and failure. In this study, we de-
scribe the design and preparation of a chitosan stent,
which has a helical structure and is self-expandable.
2. Materials and methods
2.1. Chitosan xlm
Deacetylated chitosan from crab shells (Sigma, St.
Louis, MO) was dissolved to a concentration of 4.3%
(w/v) in a water solution containing 2% (v/v) of acetic
acid. The jelly-like chitosan solution was spread evenly
(thickness &2}4 mm) over a plastic plate and let dry
overnight at room temperature [15]. The resulting
chitosan "lm was carefully detached from the plate and
immersed in a 10 N NaOH solution for 15 min. The latter
treatment made the "lm water insoluble. The basic "lm
was rinsed several times with phosphate bu!er solution

0142-9612/01/$ - see front matter  2001 Elsevier Science Ltd. All rights reserved.
PII: S 0 1 4 2 - 9 6 1 2 ( 0 0 ) 0 0 3 7 1 - 9
1870 A. Lauto et al. / Biomaterials 22 (2001) 1869}1874

a surgical blade and stored in a plastic container at room

temperature. The structure of the stents was analyzed in
detail by scanning electron microscopy (SEM). Speci-
mens were sputter coated with AuPd before being viewed
at 20 Kv in the microscope (Jeol 100cx-II with ASID
attachment for SEM).

2.3. Elasticity test

A calibrated tensiometer (Instron Mini 55, Instron,

Fig. 1. Schematic drawing of the stent preparation. The stent is manu-
factured by pulling and winding the chitosan strip around the metallic
rod. There is a partial superimposition of the strip at each turn to close
the helical structure.
Boston, Ma), controlled by a personal computer, mea-
sured the Young's module of chitosan "lms. Merlin IX
software recorded and analyzed the data. The ten-
siometer consisted of a single-column system with pneu-
matic grips and a static load cell (maximum load 50 N).
The "lms were cut into rectangular strips, which were
immersed in PBS and clamped to the system grips. The
tensiometer stressed each strip as the upper grip moved
away from the other grip at a constant speed of
25 mm/min. The test stopped when the strip broke in
two. The test parameters are summarized in Table 1.

(PBS) until its pH was neutralized. A digital caliper was
used to measure the thickness of the "lms (0.3}0.6 mm)
after they were dried again.
2.2. Stent preparation
The chitosan "lm was cut into rectangular strips with
approximate surface dimensions of &0.4;5 cm and
a thickness of 0.055$0.005 mm. These strips were im-
mersed in PBS and wound around a metallic cylindrical
rod (diameter"0.5 mm) in a helical fashion. The strips
were elongated about 30% while wound around the rod
(Fig. 1). The "lm was partially superimposed at each turn
of the helix to ensure its closure. A `circular helixa tube
(stent) was "nally obtained and it was coated externally
with a thin layer of chitosan solution to prevent its
expansion and opening. The stent was let to dry, removed
from the rod, shortened to the desired length (3}5 mm) by
2.4. Stent self-expansion
Twenty vas deferens were freshly harvested from adult
male wistar rats. A chitosan stent was inserted in each vas
(diameter &0.5 mm) to observe and test its expansion.
The vas was "rst enlarged by metallic dilators of bigger
diameters (&0.6}0.8 mm). The stent was placed on the tip
of a thin toe#on rod (diameter&0.3 mm) and then it was
inserted into a vas deferens. The procedure was per-
formed with microforceps to precisely handle stent and
tissue, under an operative microscope (;20).
2.5. In vivo study
The stability of the stent self-expansion was tested
in vivo using a vasovasostomy model, as the vas deferens
has strong muscular walls capable of contracting and
exerting inward pressure. Eleven adult male wistar rats

Table 1
The parameters of the elastic test are summarized

Parameters N Max stress Max strain E Max load Thickness Width Length
(Mpa) (%) (Mpa) (N) (mm) (mm) (mm)

Mean 33 0.5922 105.3 0.7655 0.9255 0.034 4.55 7.61

Std. error 33 0.0292 6.4 0.0288 0.0572 0.001 0.12 0.30
Normality test
KS distance 33 0.1509 0.09040 0.1497 0.09841 0.1335 0.09229 0.3787
P value 33 '0.10 '0.10 '0.10 '0.10 '0.10 '0.10 0.0002
Passed normality test 33 Yes Yes Yes Yes Yes Yes No

N, number of tested strips. Max stress, maximum stress. Max strain, Maximum strain. E, Young's module. Max load, maximum load applied on the
specimen. Thickness, width, length, dimensions of the specimen. Std. error, standard error. All the test parameters passed the normality test but the
specimen length, since the distance between the grips was intentionally kept constant.
 A. Lauto et al. / Biomaterials 22 (2001) 1869}1874
Fig. 2. Diagram of the surgical procedure. (a) The chitosan stent is
inserted in both vasal ends after vasectomy. (b) The solder collar is fused
to the tissue by the laser to complete the vas anastomosis.
(weight &270 g) were anaesthetized with 0.37 cm of so-
dium pentabarbital based upon the rodents'weight. The
vas deferens of the animals was obstructed bilaterally by
a titanium clamp under an operative microscope, as
described previously [16]. The vas was re-exposed after
2 weeks to simulate the clinical scenario of reversal vasec-
tomy. No sperm granuloma was developed in 16 obstruc-
tions. Therefore, the clipped sections were resected and
the vasa stumps were approximated end to end with
a surgical clamp. Sperm were retrieved from the proximal
end of the vas and motility was observed in all cases
under an optical microscope. A chitosan stent was par-
tially inserted in the proximal vasal end, following the
procedure described in the previous section. The distal
vasal end was also enlarged by dilators and gently
pushed against the stent until complete insertion of the
stent was achieved (Fig. 2a). Sperm were often observed
in the stents because of capillary e!ect. The vas anasto-
mosis was completed without suture, but rather with
a laser welding technique [17}19]. A thin "lm of solid
albumin solder (collar) was wrapped around the contigu-
ous ends of the vas and a "ber-coupled diode laser
(
"810 nm) fused the solder to the tissue (Fig. 2b). The
solder collar contained bovine serum albumin (65%
w/w), carbon black (0.25% w/w) and distilled water. The
laser was absorbed by the carbon black and the gener-
ated heat welded the protein collar around the vas anas-
tomosis. The laser welding parameters are summarized
Table 2
The laser soldering parameters are given as mean and standard deviation
N Collar surface (mm) Collar thickness (mm) Power (mW)
13 +3;6 0.13$0.02 250$20
N, number of vasal anastomosis. Collar surface, dimensions of the solder collar surface. Spot size, laser spot size on the collar. Power, laser power.
Time, laser irradiation time. Weight, collar weight. Dose, laser energy per mg of solder.
1871
in Table 2. Upon completing the laser anastomoses
(n"13), the abdominal muscle layers and skin were
closed with a continuous 3}0 chromic catgut suture.
Three vasal anastomoses were performed as control, by
using the same procedure described above without the
chitosan stent. The animals were sacri"ced 8 weeks later
to examine the patency of the chitosan stents inside the
vasa. Some specimens were harvested and stained with
Masson's trichrome for hystological examination.
2.6. Statistical analysis
Merlin IX software generated the strain}stress plot.
Data were also analyzed by Prism Software (GraphPad
Software Inc., San Diego, Ca), which calculated the
Kologorov}Smirnov (KS) distance to test the normal
distribution of the collected measures. The signi"cant
level for the p value was 0.05.
3. Results
3.1. Stent preparation
A typical stent is shown in Fig. 3a}c. The stent geo-
metry is a circular helix with constant pitch and inner
diameter. Strip overlaps are noticeable at each turn of the
helix and the rigid tubular structure appears completely
sealed. No appreciable irregularities were detected at the
openings of the stents.
3.2. Elasticity test
The wet chitosan strips behaved elastically, as shown
by the strain}stress diagram of Fig. 4. The E value was
0.7655$0.0283 Mpa (mean$standard error); the max-
imum strain and stress were 105% and 0.5922 Mpa,
respectively. The test results are shown in Table 1.
3.3. Stent expansion
About 3 min after insertion in the vas, the stents were
observed to self-expand to about 50% over their initial
diameter. This dilation did not regress but remained
stable, suggesting that the expansion process was irre-
versible. The insertion procedure was done safely without
crushing or damaging the stent or tissue.
Spot size (mm) Time (s) Weight (mg) Dose (J/mg)
&0.6 171$36 4.1$0.9 10.5$1.9
1872 A. Lauto et al. / Biomaterials 22 (2001) 1869}1874
Fig. 3. (a) SEM view of a 3 mm long chitosan stent (;50). The cylindrical
helix shape is revealed in details. The helix pitch is&1.4 mm and the strip
overlaps at each turn to close the tubular structure. (b) SEM view of the
stent aperture (diameter"0.5 mm, ;100). The aperture pro"le appears
regular without appreciable deformations. (c) Typical helix turn of the
chitosan stent. The tubular structure is sealed along the turn line (;50).
3.4. In vivo study
Eight weeks after the operations all the anastomoses
were in continuity at the anastomosis site (n"16) and
the solder collars were completely degraded. In the ex-
perimental group, sperm granulomas (mean dimansions
&0.9;0.4 mm) developed proximally (n"2) and at the
anastomotic site (n"7). The other 4 anastomoses were
stenotic. The anastomotic site was harvested and sperm
retrieved from the vas proximally and distally for micro-
scopic analysis. In all cases, sperm debris were found
distally and no motile sperm was observed. Sperm motil-
ity was observed proximally in only 4 specimens. Twelve
chitosan stents were rescued upon careful dissection of
the vasa. A rod (diameter"0.5 mm) was inserted inside
the stents to detect bending of their structure or chitosan
obstructing debris inside them. The stents were larger
than the rod and they appeared intact and free of
chitosan debris under the operative microscope (Fig. 5).
One stent could not be found in a sperm granuloma. Two
control animals developed a sperm granuloma. The his-
tology showed in#ammatory cells (macrophages, giant
cells and lymphocytes) inside the granulomas of the ex-
perimental and control groups. Others described similar
"ndings inside rat granulomas after vasectomy [20].
4. Discussion
Many metallic, synthetic or biodegradable stents are
currently used to relieve pathological obstruction of
ducts in urology and gastroenterology. Their common
side e!ects are stent migration, fragmentation and en-
crustation [3,11}14]. The search for a new design and
improved performance of stents is therefore necessary.
In this study, we developed and manufactured a tem-
porary chitosan stent with a self-expanding mechanism
and a helical design. The stents expand in a few minutes,
by releasing their elastic energy when in contact with
tissue moisture. Upon insertion in rat vasa, they were
observed to expand &50% over their initial diameter.
Self-expansion occurs when the thin layer of chitosan jell
coating the stent is dissolved by tissue moisture. Conse-
quently, the elongated chitosan strip contracts and re-
leases its elastic energy by opening the helical structure.
Chitosan stents can either o!er a rigid support or exert
a controlled pressure on the tissue walls adjacent to
them. It is possible to estimate the average pressure ( p)
exerted by the stent upon surrounding tissue, using the
Young's module (E). The work (L) done by the tube while
opening causes an increase in volume (<) given by
¸"p<, hence p"¸/<. The work is proportional to
the elastic energy accumulated by the chitosan strip dur-
ing its elongation to prepare the stent. Therefore,
¸J(k/2)x, where x is the chitosan strip elongation
and k is the elastic constant for the chitosan strip. We
also know that k"ES/l , where S is the strip cross-

sectional area and l is the initial strip length. Finally, we

obtain pJ(Sl /<)(E/2)(x/l ). The pressure exerted
 
by the stent on the tissue can be controlled by varying the
 A. Lauto et al. / Biomaterials 22 (2001) 1869}1874 1873

Fig. 4. Strain}stress diagram of a chitosan strip, tested by the tensiometer. The linear behavior of the strain proved the elastic nature of the chitoasan
"lm. The strip doubled its initial length before breaking under stress.

relative elongation and cross-sectional area of the chitosan

strip. Upon substitution of the typical values of our
stents: <"3.9 mm (50% expansion), E"0.7655 Mpa,
S"0.25 mm, l "9 mm, x"3 mm, and assuming

50% of the elastic energy was used to expand the vas
deferens; we "nd p &90 mm Hg.
In our experiment, the stents enlarged the vas until the
tissue walls equilibrated with the expansion pressure. The
stents were stable for 8 weeks inside the animals and did
not shrink back to their initial dimensions, suggesting the
self-expanding process was irreversible. The stents re-
mained dilated because tangential frictional forces were
probably created between the chitosan strip at the helix
turn. Such friction prevented narrowing of the stent
diameter.
It is remarkable that 7 chitosan stents remained en-
larged in sperm granulomas, where the pressure can be
higher than the normal physiological level (Fig. 5). The
stents did not migrate in 6 cases, where granuloma did
not form at the operation site. No conclusion could be
drawn about the stability of the stent position when
granulomas were present at the anastomotic site (n"7).

䉳&&&&&&&&&&&&&&&&&&&&&&&&&&&&&
Fig. 5. Longitudinal section of a sperm granuloma at the anastomotic
site, 8 weeks post surgery. The chitosan stent is open despite the in#am-
matory response of the granuloma. Slide preparation partially disrupt-
ed the stent (Masson's trichrome, ;40).
1874 A. Lauto et al. / Biomaterials 22 (2001) 1869}1874
The development of sperm granulomas may be facilitated
by the spermicidal properties of chitosan, already re-
ported by Sanford [21]. The laser energy could also
induce stenosis secondary to thermal damage, which may
promote granuloma formation [22].
For clinical applications, requirements for stent use are
variable. Chitosan stents can be prepared with di!erent
dimensions. The diameter can vary between 0.3 mm and
a few centimeters, and the length can range from a few
millimeters to several centimeters.
We chose chitosan among others biomaterials because
of its tissue biocompatibility, elasticity and antimicrobial
activity, as well as homeostatic and wound healing prop-
erties [9,23}28]. The stent preparation presented in this
study can be extended to other suitable biomaterials with
elastic properties. Our chitosan stents are not indicated
for vasovasostomy because of their spermicidal proper-
ties. Nevertheless, they may be implanted in the prostate,
bile duct or ureter during open surgery. Problems may
arise following endoscopic insertion of chitosan stents if
the stents are not implanted quickly, as bodily #uids may
prematurely trigger the self-expanding mechanism. Fur-
ther studies are required to test and evaluate the chitosan
stents in other animal models. Also, delay mechanisms
for the stent self-expansion are currently being investi-
gated to facilitate their endoscopic insertion.
Acknowledgements
The authors wish to thank Drs. M. Hyman and S. Shin
for their suggestions.
References
[1] Lee BH, Do YS, Lee JH, Kim KH, Chin SY. New self-expandable
spiral metallic stent: preliminary evaluation in malignant biliary
obstruction. J Vasc Interv Radiol 1995;6:635}40.
[2] Tschada RK, Henkel TO, Junemann KP, Rassweiler J, Alken P.
Spiral reinforced ureteral stent: an alternative to internal urinary
diversion. J Endourol 1994;8:119}23.
[3] Braf, Z Chen J, Sofer M, Matzkin H. Intraprostatic metal stents:
more than 6 years' clinical experience with 110 patients. J En-
dourol 1996;10:555}8.
[4] Cozzi G, Colnago MF, Bellomi M, Giovannardi G, Salvetti M,
Saverini A. Protesi metalliche autoespandibili nel trattamento
delle stenosi neoplastiche esofagee. Radiol Med 1994;88:272}6.
[5] van Sonnenberg E, D'Agostino HB, O'Laoide R, Donaldson J,
Sanchez RB, Hoyt A, Pittman CC. Malignant ureteral obstruc-
tion: treatment with metal stents techniques, results, and observa-
tions with percutaneous intraluminal US. Radiology 1994;191:765}8.
[6] Petas A, Talja M, Tammela TLJ, Taari T, Valimaa T, Tormala P.
The biodegradable self-reinforced poly-DL-lactic acid spiral stent
compared with a suprapubic catheter in the treatment of post-
operative urinary retention after visual laser ablation of the pros-
tate. Br J Urol 1997;80:439}43.
[7] Petas A, Karkkaainen P, Talja M, Taari K, Laato M, Valimaa T,
Tormala P. E!ects of biodegradable self-reinforced polyglicolic
acid, poly-DL-lactic acid and stainless-steel spiral stents on
uroepithelium after Nd : YAG laser irradiation on the canine
prostate. Br J Urol 1997;80:903}7.
[8] Schlick RW, Planz K. Potentially usefull materials for biodegrad-
able ureteric stents. Br J Urol 1997;80:908}10.
[9] Shigemasa Y, Shibazaki K, Minami S, Matsuhashi A, Tanioka S,
Shigemasa Y. Evaluation of chitin and chitosan for biomaterials.
Biotechnol Genet Engng Rev 1996;13:383}420.
[10] Smits M, Huibregtse K, Tytgat G. Results of the new nitinol self-
expandable stents for distal biliary structures. Endoscopy
1995;27:505}8.
[11] Marsman JW, Hoedmaker HP. Necrotizing fasciitis: fatal com-
plication of migrated biliary stent. Aust Radiol 1996;40:80}3.
[12] Benoit G, Blanchet P, Eschwege P, Alexandre L, Bensadoun H,
Charpentier B. Insertion of a double pigtail ureteral stents for the
prevention of uroligical complications in renal transplantation:
a prospective randomized study. J Urol 1996;156:881}4.
[13] Sauter ER, Ho!man JP, Hartz WH, Barber LW, Eisemberg BL.
Atraumatic method of intraoperative retrograde transhepatic bili-
ary stent insertion. J Surg Oncol 1996;62:10}4.
[14] Holmes SAV, Miller PD, Crocker PR, Kirby RS. Encrustation of
intraprostatic stents: a comparative study. Br J Urol 1992;69:
383}7.
[15] Costain DJ, Kennedy R, Ciona C, McAlister VC, Lee TDJ.
Prevention of postsurgical adhesion with N,O-carboxymethyl
chitosan. Surgery 1997;121:314}9.
[16] Young GPH, Li PS, Gardner TA, Goldstein M. Animal models
for microsurgical training and research. In: Goldstein M, editor.
Surgery of male infertility. London: Saunders Company, 1995.
p. 297}320.
[17] Lauto A, Kerman I, Ohebshalon M, Felsen D, Poppas DP.
Two-layer "lm as a laser soldering biomaterial. Lasers Surg Med
1999;25:250}6.
[18] Lauto A, Dawes JM, Piper JA, Owen ER. Laser nerve
repair by solid protein band technique (II). Microsurgery 1998;
18:60}4.
[19] Lauto A. Repair strength dependence on solder concentra-
tion: a study in laser tissue welding. Lasers Surg Med 1998;22:
120}5.
[20] Caldwell JC, McGadey J, Kerr R, Bennett NK, Mc Donald SW.
Cell recruitment to the sperm granuloma, which follow vasectomy
in the rat. Clin Anat 1996;9:302}8.
[21] Sanford PA. Chitosan and alginate: new forms of commercial
interest. Am Chem Soc Div Polym Chem 1990;31(1):628.
[22] Trickett RI, Wang D, Maitz P, Lanfetta M, Owep ER. Laser
welding of vas deferens in rodens: initial experience with #uid
solders. Microsurgery 1998;18:414}8.
[23] Okamoto Y, Shibazaki K, Minami S, Tanioka S, Shigemasa Y.
Evaluation of chitin and chitosan on open wound healing in dogs.
J Vet Med Sci 1995;57:851}4.
[24] Okamoto Y, Tomita T, Minami S, Matsuhashi A, Kumazawa
NH, Tanioka S, Shigemasu S. E!ects of chitosan on experimental
abscess with Staphylococcus aureus in dogs. J Vet Med Sci
1995;57:765}7.
[25] Muzzarelli R, Tarsi R, Filippini O, Giovanetti E, Biagini G,
Varaldo PE. Antimicrobial properties of N-carboxylbutyl
chitosan. Antimicro Ag Chemotherap 1990;34:2019}23.
[26] Suzuki K, Okawa Y, Hashimoto K, Suzuki S, Suzuki M. Protect-
ing e!ect of chitin and chitosan on experimentally induced murin
candidiasis. Microbiol immunol 1984;28:909}12.
[27] Okamura T, Masui T, Taylor RJ. Evaluation of e!ect of chitosan
in preventing hemorrhagic cystitis in rats induced by cyclophos-
phamide. Acta Urol Jpn 1995;41(4):289}96.
[28] Rao SB, Shorma CP. Use of chitosan as a biomaterial: studies on
its safety and homeostatic potential. J Biomed Mater Res
1997;34:21}8.