The Tensi Ie Strength of Uterosacral Ligament Sutures

Document Sample
The Tensi Ie Strength of Uterosacral Ligament Sutures Powered By Docstoc
					The Tensi Ie Strength of Uterosacral Ligament
Sutures: A Comparison of Vaginal and Laparoscopic

PatrickJ. Culligan, MD,John R. Miklos, MD,Miles Murphy, MD, Roger Goldberg, MD, MPH,
Carol Graham, MD, Robert D. Moore, DO,Meg Hainer, MD,and Michael H. Heit, MD, MSPH

OBJECTIVE:To compare the tensile strength of two ap-                             Recently, use of tlle uterosacralligamentsas tlle apical
proaches for uterosacral ligament suturing using a cadaver                       attachmentpoints during reconstructivepelvic surgery
model.                                                                           has gained widespread acceptance.       The intermediate
METHODS: 12 unembalmed cadavers, four polytetra.fluo-
            In                                                                   portion! of tlle uterosacralligamenthas been identified
roethylene sutures were placed through the uterosacral                           astlle optimal fixation point for safetyand strength.The
ligaments. In each cadaver, two sutures were placed lapa-                                                          of
                                                                                 long-termclinical effectiveness tlle vaginal approach
roscopically, and two more were placed vaginally. A single,                      for suturing tllese ligamentsusing standard,nondispos-
experienced surgeon placed alliaparoscopic      sutures (n =                     able needle holders has been established.2-4 suchNo
23), and another experienced surgeon placed all vaginal                                             tlle                         or
                                                                                 evidencesupports useof oilier approaches surgical
sutures (n = 22). A blinded team of investigators measured                       devices when using tllese ligamentsin reconstructive
the distance from each suture to the ipsilateral ischial spine;
                                                                                 pelvic surgery.
determined whether any sutures incorporated ureters, vis-                                    tlle
                                                                                    Because ergonomics,lighting, and surgical points
cera, or large vessels; and then passed the sutures through
                                                                                 of view of laparoscopic,   vaginal, and open abdominal
an apical vaginal incision. Using a hand-held tensiometer,
                                                                                 surgeryare quite different, one cannotassumetllat tllese
progressive tensile load was then applied to these sutures
                                                                                 approachesto suturing tlle uterosacral ligaments are
along the axis of the vagina until they either broke or were
                                                                                 equivalent. Therefore, new techniques or devices for
completely dislodged from the ligaments.
                                                                                 suturing tlle uterosacralligaments   shouldbe subjectedto
RESULTS:The average peak tension required to break or                            feasibilitystudiesbefore beingusedclinically. Cadaveric
dislodge the sutures was 26.2 :t 8.8 psi (Iaparoscopic) and                      studiesrepresentone way to compare new surgical ap-
22.5 :t 7.4 psi (vaginal) (P = .14,95% confidence interval
                                                                                 proacheswitll tlle established  vaginal2-4techniques.
[CI] -1.2, 8.6). The average force required for suture                               Our primary objective for this study was to compare
breakage (n = 28) was 28 :t 7 psi, and the average force
                                                                                 tlle tensile strength of laparoscopicallyand vaginally
applied when ligament failure occurred (n = 17) was 18.5
                                                                                 placeduterosacralligamentsuturesin a cadaveric     model.
 :t 6 psi (P < .001, 95% CI -13.8, -5.2). The average
                                                                                 Our secondaryobjective was to qualitatively compare
distance from a laparoscopic or vaginal suture to the ipsi-
                                                                                 tlle two techniquesfor suture position along tlle liga-
lateral ischial spine was 19.1 :t 7 mm and 17.4 :t 6 mm,
 respectively (P = .46,95% CI -3.0, 6.4). None of the sutures                    mentsand whetller tlley incorporated oilier vital struc-
from either technique were found to incorporate              a visceral                                              a
                                                                                 tures.We also soughtto establish cadavericmodel that
structure, ureter, or great vessel.                                              could be used for future studies involving uterosacral
                                                                                 ligament suturing devicesor techniques.
CONCLUSION:  These suturing techniques appear to be equal
in tensile strength.   (Obstet Gynecol 2003;101:500-3.
@ 2003 by The American College of Obstetricians and                                            AND
                                                                                 MATERIALS METHODS
                                                                                 Twelve unembalmedfemalecadavers       were used for this
From theDepartmentofObstelricr, Gynecology Woman'sHealth, Uniuersiiy             project.These cadavers    were obtained through the Uni-
                            Center,Louisuille, Kentucky; Atlonf1l Centerfor
of Louisuil/e Health Sciences                                                    versity of Texas Southwestern    Medical SchoolWilled
LaparoscopicUrogynecology, Atlanta, Georgia;andEuanston Conti1I£nC£ Center,      Body Program. Through that program, all cadaversare
NorthwesternUniuersiiyMedi£al ScIwol,Euanston,  1llinoir.                        screenedfor human immunodeficiencyvirus and hepa-
Supported mI intramural "Research Women"grant throughthe Uniuersiiyof
                                on                                                                                              for
                                                                                 titis B virus before they are deemedacceptable usein
Louisuille.                                                                      research.No other inclusion or exclusioncriteria were

500       YOLo 101, NO.3, MARCH 2003                         ';:j                                                            0029-7844/03/$30.00
          @ 2003 by The American College of Obstetricians   and Gynecologists.   Published   by Elsevier.   doi: 10.1016/S0029-7844(02)03123-X
applied.They had all beenfrozen shordy after deathand             For the vaginal suturing, all cadaversremained in the
were shipped to our study site to be thawed 72 hours          dorsal lithotomy and Trendelenburgposition. The peri-
before our investigation. The only demographicinfor-          toneal cavity was entered through a transverseincision
 mation available for thesecadaverswas their age at the        at the vaginalapex, and a surgical towel was insertedto
time of death.To maximizethe clinical relevanceof our          retractthe bowel out of the pelvis.The posterioredgeof
 results,we simulated actual operating room conditions        the vaginawas graspedwith two 12-inAllis clamps,and
and techniquesas closelyas possible.This entire study         the uterosacralligamentswere visualized.The sutures
was performed on a singleday, and the ergonomicsand            that had been placed laparoscopicallywere visualized
lighting did not differ betweencadavers.                      but not touchedby the vaginal surgeon.On theopposite
   We did not seekformal approval for this study proto-       side, the ischial spine and uterosacralligamentwas pal-
col because University of Louisville Human Studies            pated. Two retractors (Briesky-Navratil,Marina Medi-
Committee doesnot require approval for cadavericstud-          cal,Hollywood, FL) were positioned at3 or 9 o'clock (to
ies. However, the study was funded (ie, approved)by a         protect the structuresof the pelvic wall) and at 5 or 7
competitive intramural grant through the University of         o'clock (to protect the rectum). A curved needleholder
Louisville Intramural ResearchIncentive Grant Com-             (14-inNolan CVD tip, Marina Medical,Hollywood, FL)
mittee.                                                       was used to place two sutures through the uterosacral
   In each cadaver,four 2.0 polytetrafluoroethylenesu-        ligament in figure-of-eight fashion.As was the casewith
tures (GoreTex, W.L. Gore and Associates     Inc., Flag-      the laparoscopicsutures,an attemptwas made to place
staff, AZ) were placed through the uterosacralligaments       all vaginal sutures in the intermediate1portion of the
in a "figure-of-eight"fashion.Two of thesesutureswere         ligament.
placed laparoscopically,and two were placedvaginally.             A separate,  blinded team of investigators then per-
We attemptedto place alliaparoscopic and vaginal su-          formed laparotomieson all of the cadavers.This team
tures in the intermediatesection(ie, 1~2cm posteriorto        wasblinded asto which sutureshad beenplacedlaparo-
the ischial spine) of the uterosacralligamentperpendic-        scopicallyor vaginally. The distancefrom eachsuture to
ular to the fibers of the ligament, as recommendedby          the ipsilateral ischial spine was measuredin mill. The
                                                              sutures were then passedthrough the vaginal incision
   Block randomization (using blocks of four) was per-                      to
                                                              and fastened a hand-helddigital tensiometer(DFG51,
formed to determine whether the laparoscopicand vag-          Omega Engineering Inc., Stanford, CT). Progressive
inal sutureswere to be placed on the left or right side for   tensileload wasappliedto the suturesindividually along
 eachcadaverand to determinewhether the proximal or           the axis of the vagina until they broke or were dislodged
 distal sutureswere the first to be testedin eachcadaver.     from the ligament.Before and after suture pullout, this
Therefore, both the side of suturing and the order of         team also determined whether any of the sutures had
pullout were randomly assigned.                               compromisedthe ureters,viscera,or any large vessels.
   Two laparoscopicsutures were placed first, on one              A power calculation was performed, based on an
side in eachcadaver.The laparoscopic      sutureshad to be    assumptionthat the vaginallyplacedsutureswould have
placed first, becausedoing otherwisewould have com-           a meantensilestrengthof 24 psiand a standarddeviation
promisedthe pneumoperitoneum.A singlesurgeonaM)               of3 psi.There is no publisheddataregardingthe "acute"
placed all of thesesuturesusing a nondisposable,5-mm          tensilestrengthof uterosacralligament sutures,so these
needledriver (Y-Handle Snap-A-Part,      Elmed Co., Addi-     estimateswere made based on our previous, unpub-
son, IL). Three laparoscopicports (DuraGold, Applied          lished experiencemeasuringtensile strength of various
Medical Co., Rancho SantaMargarita, CA) were used:            pelvic ligaments in female cadavers.We first decided
one 12-mm epigastricport (2 cm caudadto the umbili-           that a 20% differencein tensilestrengthbetweenthe two
cus)and two 5-mm ports (onesuprapubicand one right            techniqueswould be clinically significant.We then de-
paramedian). A 10-mm, 00 laparoscope was placed               termined that in order to have 99%power of detectinga
through the epigastricport. SteepTrendelenburg posi-          20% differencebetweenthe suturing techniqueswith an
tion was used, and the bowel was retracted away from          a of 0.05, we needed20 data points in eacharm of the
the sidewall. The uterosacral ligaments were held on          study. We did not power the study to detecta significant
stretch with a vaginal probe, and they were tracedback-       difference between suturing techniques for position
ward to their most proximal point of origin. Two figure-      along the ligament relative to the ischial spine. For the
of-eightsutureswere placed through the designated      lig-   purposesof statistical           we
                                                                                     analysis, considered   eachutero-
amentin the intermediate1portion. No knots were tied          sacral suture "pull-out" as an independent event. As
in thesesutures.The two free ends of eachsuture were          such, there were four potential tensile strength data
left lying along the ipsilateralpelvic sidewall.              points per cadaver.

VOL. 101, NO.3,   MARCH 2003                                    Culligan et al   Uterosacral   Ligament Suture Tensile Streneth   501
  The data was assessed normality to determine                            strength of vaginal sutures placed on the left side and
wheilier mean or median values were compared.The                          right sidewere 23 :t 9 psi and 22 :t 6 psi, respectively(P
separate  variance t test was used to compareilie mean                    = .9,95% CI -6.5,7.4).
tensile strength of sutures placed vaginally or laparo-
scopically.Mean distances   from ilie suturesto ilie ischial
spineswere comparedusing ilie separate     variance t test                DISCUSSION
as well. Separate analyses  were performed to determine                   Our objectivesfor this study were to comparethe tensile
wheilier eiilier surgical techniqueresulted in a "favor-                                            of
                                                                          strengthand placement two popular uterosacralsutur-
able" side for tensilestrength and to determinewheilier                                     By
                                                                          ing techniques. doing so,we createda scientificmodel
presence a uterusmadea differencein tensilestrength.                      that could be used for studying feasibility of other de-
Statistical analysiswas performed willi SPSSfor Win-                      vices and techniquesdesignedfor suturing the uterosa-
dows 10.0 (SPSS   Inc., Chicago,IL).                                      cralligaments.
                                                                              We recognizedthe liniitations of this study design,
                                                                          and we attemptedto accountfor all of the confounding
RESULTS                                                                   factors. Our inability to fully standardizethe surgical
The averageageof the cadaversat the time of deathwas                      approaches     was both a strength and weaknessof our
72 ::!:9.8 years. A total of 48 sutureswere placed in the                 study design.The surgeonsused identical sutures and
twelve cadavers, as planned. During tensile strength                      nondisposable     needleholders, and they eachattempted
testing,one set of laparoscopicsutures and two setsof                     to placesuturesthrough the intermediate1portion of the
vaginal suturesdislodged togetherwhen the first suture                    ligament.However, no further standardization      was pos-
waspulled. That left 22vaginal and 23laparoscopic     data                sible. In an effort to maximize the clinical relevanceof
points availablefor analysis.A uteruswas presentin 6 of                   our results,both surgeonssimply tried to duplicate the
the 12 cadavers.                                                          techniquesused in their clinical practices.Because     we
    The averagedistancefrom a laparoscopicor vaginal                      could not account for surgical skill as a confounding
suture to the ipsilateral ischial spine was 19.1 :!: 7 mm                 factor in our statisticalanalysis,we used only one expe-
and 17.4::!:6 mill, respectively(P = .46,95%confidence                    rienced surgeonfor each technique.Both surgeonsare
interval [CI] ~3.0, 6.4). However, according to a post                    fellowship-trained urogynecologistsand have refined
hoc power calculation, our chanceof making a type n                        their techniquesduring more than 200 uterosacralvagi-
error in this assessment    was 86% (power to detect a                    nal vault suspensions.   Thus, we attemptedto minimize
difference: .14). None of the sutures from either tech-                   the effectof differing surgicalability by having only very
nique were fou.nd to incorporate a visceral structure,                    experiencedsurgeonsplace the sutures. That decision
ureter, or greatvessel.                                                    may liniit the external validity of our results,which is a
    When all laparoscopicand vaginal tensile strength                      liniitation that existswith any surgicalstudy.
values were consideredtogether, the averageforce re-                          The cadavers                 age,and medicalhistory,
                                                                                             varied in size,
quired for suture breakagewas significantlyhigher than                    and therewasa possibilityof varied techniqueduring the
the force required for ligamentfailure. The presenceor                    tensilestrengthtesting.We usedunembalmedcadavers,
absence a uterus madeno difference.The meanvalue                          becausethe integrity of their tissueplanes is similar to
 for suture breakage (n = 28) was 28 ::!: 7 psi, and the                   living tissue.We attemptedto deal with the confounding
mean force applied when ligamentfailure occurred (n =                                                   by
                                                                          factorswithin the cadavers randomizingboth the side
 17) was 18.5 ::!:6 psi (P< .001,95% CI -13.8, -5.2).                     used for eachtechniqueand the order of suture pullout.
The meantensilestrength of all suturesplacedin cadav-                     We liniited the bias associatedwith tensile strength
ers with and without uteri were 24 ::!:8 psi and 24 ::!:9                 testing by blinding the investigatorsas to the suturing
psi, respectively(P = .83,95% CI -5.6,4.5).                               techniques,and we attemptedto reducethe varianceof
    We found no significantdifferencebetween tensile                      tensilestrengthtestingby standardizingthe pullout tech-
strength of vaginal or laparoscopic sutures, and both                     ruque.
techniquesresulted in similar tensile strength between                        Our study designhinges on the assumptionthat the
the left and right sides.The averagepeak tension re-                      tensilestrength of the suturesshould correlatewith their
quired to break or dislodgethe laparoscopic   and vaginal                 clinical utility in reconstructive surgery. We assumed
sutures was 26.2 ::!:8.8 psi and 22.5 ::!: 7.4 psi, respec-                that a 20%differencein tensilestrengthbetweenthe two
 tively (P = .14, 95% CI -1.2, 8.6). The mean tensile                     techniqueswould be clinically relevant. Although the
strength of laparoscopicsuturesplaced on the left side                    clinical relevanceof this outcomemeasureis unknown,
and right side were 25 ::!:9 psi and 27 ::!:9 psi, respec-                we do know that the vaginal approach to uterosacral
 tively (P = .46, 95% CI -10.5, 5.0). The mean tensile                     ligament suturing is clinically effective.2-4 Becausethe

502     Culligan et al   Uterosacral   Ligament Suture Tensile Strength                                        OBSTETRICS& GYNECOLOGY
threshold of tensile strength necessaryto achieve this                                    for             as
                                                            tion hasno suchrequirements newdevices, long as
clinical success unknown, we assumedthat any new            they have "substantialequivalence"to existing technol-
uterosacralligamentsuturing deviceor technique(in this      ogy.6 In other words, United StatesFood and Drug
case the laparoscopic technique) should be similar in       AdIninistration approval for a surgical device does not
tensilestrengthto the vaginal technique.                    necessarilyimply that the device will perform as ex-
   Our decisionto considereachsuture pull-out testan        pected.By insisting to seefeasibilitydata prior to using
"independent" event could be questioned. Within a           new surgical devices for reconstructivepelvic surgery,
given cadaver,the strengthand integrity of the right and    clinicians could be more confident about the clinical
left uterosacralligamentsare largely independentof each     outcomesassociated   with thesedevices.
other. Although tissuequality within a given cadaveris
relatively unifonn, the right and left uterosacral liga-
ments could have other differences (breakagepoints,
stretching, etc) related to a patient's medical history.    1. BullerJL, ThompsonJR, CundiffGW, Sullivan LK, Schon
                                                               Ybarra MA, BentAE. Uterosacralligament: Descriptionof
Therefore, pulling suturesfrom one ligamentshould not
                                                               anatomic relationshipsto optimize surgical safety. Obstet
affectthe resultsfound on the oppositeside.We recog-
nized that within a givenuterosacralligament,the tensile       GynecoI2001;97:873-9.
                                                            2. ShullBL, BachofenC, Coates KW, Kuehl1J. A transvag-
strength testing of one suture would certainly influence
                                                               inal approachto repair of apicaland otherassociated
the results for the other suture. ill other words, we
                                                               pelvic organ prolapse. Am] Obstet Gynecol 2000;183:
expectedthe tensile strength of the first suture in any        1365-74.
ligament to be greater than that of the secondsuture.       3. Barber MD, Visco AG, Weidner AC, Amundsen CL,
However, we choseto considerthe two sutureswithin a            Bump RC. Bilateral uterosacralligamentvaginal vault sus-
given ligamentas independent,because randomized                                        endopelvicfasciadefectrepair for
                                                               pensionwith site-specific
the order of suture pull-out. Thus we designed our             treatment of pelvic organ prolapse. Am] Obstet Gynecol
randomizationscheme(ie, randomizing both the side of           2000;183:1403-11.
placementand order of pullout for each technique)as a                                   S,                   A,
                                                            4. Karram M, Goldwasser Kleeman S,Steele VassalloB,
way of dealingwith the fact that determinationsof tensile      Walsh P. High uterosacralvaginal vault suspension     with
strength within a single cadaver (ie, subject)may not,         fascial reconstruction for vaginal repair of enteroceleand
statistically speaking, be considered "independent             vaginal vault prolpase. Am] Obstet Gynecol 2001;185:
events."                                                       1339-43.
   These limitations not withstanding,we demonstrated       5. Luber KM, Boero S, Choe]. The demographicsof pelvic
similar tensilestrength betweenuterosacralligamentsu-           Hoor disorders: Current observationsand future projec-
tures placed vaginally and laparoscopicallyby experi-          tions. Am] Obstet Gynecol2001;184:1496-503.
enced surgeons using nondisposable needle holders.          6. U.S. Departmentof Health and Human Services.Determi-
This study designcould be usedto evaluatethe feasibil-          nation of intendeduse for 510(k)devices.Document 41'897,
ity of other devices and techniques for suturing the           ]anuary30, 1998.Rockvi1le,  Maryland: U.S. Departmentof
                                                                Health and Human ServicesFood and Drug Administra-
                                                                tion Center for Devicesand RadiologicalHealth, 1998.
   As the proportion of elderly women in the United
Statesrises,the demand for urogynecologicsurgerywill        Address reprint requests to: Patrick]. Culligan, MD, University
increasedramatically as well.5 This demand is likely to     of Louisville Healfu SciencesCenter, Department of Obstetrics,
inspire the developmentof products designedto make          Gynecology and Woman's Healfu, 315 East Broadway M-18,
prolapsesurgery fasterand easier.Ideally, the feasibility   Louisville, KY 40202; E-mail:
of such surgical deviceswould be scrutinized in cadav-
eric and/or animal models prior to their clinical use.             July IS, 2002. Received reviredformSeptember
                                                            Received                      in                  23,2002.
However, the United States   Food and Drug Administra-             Odober9, 2002.

YOLo101, NO.3, MARCH 2003                                     Culligan et al   Uterosacral   Ligament Suture Tensile Strength   503

Shared By: