and cleavage stages compares favourably with day 5 blastocyst transfer.Hum Reprod 2002;17:1852–5.

5. Fisch JD, Rodriguez H, Ross R, Overby G, Sher G. The graduatedembryo score (GES) predicts blastocyst formation and pregnancy ratefrom cleavage-stage embryos. Hum Reprod 2001;16:1970–5.

6. Milki AA, Hinckley MD, Gebhardt J, Dasig D, Westpal LM, Behr B.Accuracy of day 3 criteria for selecting the best embryos. Fertil Steril2002;77:1191–5.

doi:10.1016/S0015-0282(03)01130-0

Reply of the Authors:

We appreciate Drs. Correa-Perez and Fernandez-Pelegri-na’s interest in our work. We do not disagree with the spiritof their comments, but we wish to clarify some of the details.

A prospective study, in which transfers are randomlyassigned to day 3 or day 5, would be a better way ofcomparing these two techniques. Several investigators (1–3)have done this, but not specifically for patients older than 40years of age. We believe that our study provides valuablepreliminary data, that may encourage a randomized prospec-tive study in this age group.

During the study, we had no good feel for what techniqueworked best and offered either day 3 or day 5 transfer topatients older than 40 with four or more 8-cell embryos. Thisprocess may have introduced an unintended selection bias.The suggestion to assign patients with a minimum number ofzygotes to either day 3 or day 5 transfer would be a reason-able approach for a prospective study. In clinical practice,however, we have found it more relevant to assess thenumber of good embryos on day 3 before making a decision.It is not uncommon for patients with multiple zygotes, es-pecially older patients, to fail to produce 8-cell embryos andaccordingly become poor candidates for extended culture(4).

The alternative suggestion is “ to compare patients whomade their choice on day 3 but were similar in terms ofnumber and quality of embryos transferred.” We did com-pare patients who were similar in number of 8-cell embryosavailable; however, the number of embryos transferred wassignificantly higher on day 3, as one would expect, given thelower implantation rate of cleavage-stage embryos comparedwith blastocysts (1, 3–5). As we discussed, the risk of notconceiving rather than the risk of “overconceiving” is themajor gamble among women with a mean age of 41.6 years,and in most U.S. programs, the transfer of a high number ofday 3 embryos is justified. This prevailing tendency leads tosignificantly fewer cycles with excess embryos available forcryopreservation and may decrease the cumulative preg-nancy rate.

Use of stringent selection criteria that follow the embryothrough pronuclear morphology and early 2-cell cleavage onday 1, day 2 multinucleation, and day 3 morphology maysharpen our ability to select the best cleavage-stage embryo.There is no consensus, however, about the accuracy of these

criteria. In addition, their implementation requires individualembryo culture and multiple thorough examinations outsidethe incubator. We believe that if several good-quality em-bryo are available, blastocyst culture is a practical and morereliable tool for selecting the best for transfer. The informa-tion gained from extended culture may be an added benefit inolder women by shedding light on embryo quality and moreclearly guiding future therapy, including oocyte donation.

Amin A. Milki, M.D.Mary D. Hinckley, M.D.Barry Behr, Ph.D.Department of Gynecology and ObstetricsStanford University School of MedicineStanford, CaliforniaMay 30, 2003

References1. Gardner DK, Schoolcraft WB, Wagley L, Schlenker T, Stevens J, Hesla

J. A prospective randomized trial of blastocyst culture and transfer inhuman in vitro fertilization. Hum Reprod 1998;13:3434–40.

2. Coskun S, Hollanders J, Al-Hassan S, Al-Sufyan H, Al-Mayman H,Jaroudi K. Day 5 versus day 3 embryo transfer: a controlled randomizedtrial. Hum Reprod 2000;15:1947–52.

3. Karaki RZ, Samarraie SS, Younie NA, Lahloub TM, Ibrahim MH.Blastocyst culture and transfer: a step toward improved in vitro fertili-zation outcome. Fertil Steril 2002;77:114–8.

4. Racowsky C, Jackson KV, Cekleniak NA, Fox JH, Hornstein MD,Ginsburg ES. The number of eight-cell embryos is a key determinant forselecting day three or day five transfer. Fertil Steril 2000;73:558–64.

5. Milki AA, Hinckley MD, Fisch JD, Dasig D, Behr B. Comparison ofblastocyst transfer with day 3 embryo transfer in similar patient popu-lations. Fertil Steril 2000;73:126–9.

doi:10.1016/S0015-0282(03)01131-2

Diagnosis of endometriosis: utility of MRI?To the Editor:

In the study by Stratton et al. (1), magnetic resonanceimaging did not appear to offer a reliable alternative todiagnostic laparoscopy across the spectrum of endometrio-sis. However, if endometriosis on peritoneal surfaces is aconsequence of abnormal uterine contractility secondary todifferent patterns of neurologic dysfunction, perhaps someclearer relationships might be established. Advanced nullip-arous endometriosis (American Fertility Society stage IV) isregularly observed in women who have sustained expulsiveefforts to achieve defecation and can be demonstrated by bothdiagnostic techniques. Minor parous endometriosis is fre-quently observed some time after difficult obstetric episodes(e.g., after prolonged or premature maternal voluntary efforts)and may lead to reduced specificity for both techniques.

Myofascial injuries during parturition, including damageto the uterosacral and transverse cervical ligaments, areassociated with denervation and subsequent reinnervation ofthe uterus (2). Uncoordinated activity of subserosal and en-dometrial–myometrial nerve plexi may be expected to resultin abnormal uterine contractility (3). Gamete transport and

FERTILITY & STERILITY� 1071

disordered menstrual function are direct consequences, withendometrium being deposited at sites of intraperitoneal dam-age—for example, hyperplastic uterosacral ligaments in nul-liparous women who strain to evacuate their bowels or thescarred vaginal insertions of the uterosacral ligaments afterdifficult intrapartum episodes. Chronic pain is the conse-quence of progressive reinnervation of myofascial supportsover the medium term. Intrinsic damage to uterine innerva-tion, as by a single asymmetric uterine leiomyomata, orextrapelvic damage to uterine innervation, as by accidents,falls, or traffic accidents may cause different presentations.

Disordered uterine contractility associated with aberrantneural repair in and around uterine myofascial supports mayaccount for many of the manifestations of “endometriosis.” Ifthis hypothesis is confirmed, diagnostic techniques might betailored to address clinical issues that will vary with the specificetiology and subsequent natural history of the condition.

Martin Quinn, M.D., M.R.C.O.G.Richard Slade, F.R.C.S., M.R.C.O.G.Hope HospitalSalford, Manchester, United KingdomMay 15, 2003

References1. Stratton P, Winkel C, Premkumar A, Chow C, Wilson J, Hearns-Stokes

R, et al. Diagnostic accuracy of laparoscopy, magnetic resonance imag-ing, and histopathologic examination for the detection of endometriosis.Fertil Steril 2003;79:1078–85.

2. Quinn MJ, Kirk N. Differences in uterine innervation at hysterectomy.Am J Obstet Gynecol 2002;187:1515–20.

3. Kunz G, Beil D, Huppert P, Leyendecker G. Structural abnormalities ofthe uterine wall in women with endometriosis and infertility visualizedby vaginal sonography and magnetic resonance imaging. Hum Reprod2000;15:76–82.

doi:10.1016/S0015-0282(03)01132-4

Reply of the Authors:

The purpose of our study was to determine whethercurrently available MRI technology was useful in diagnosingendometriosis (1). In this context, MRI had limited utilitybecause we could not determine the extent of disease oridentify small peritoneal lesions.

Drs. Quinn and Slade present some interesting implica-tions about the potential for neurologic dysfunction or fascialdamage in the etiology of endometriosis. The issue of etiol-ogy or coincident conditions of endometriosis alluded to inthis letter, is beyond the scope of our paper. On the basis ofDrs. Quinn and Slade’s suggestion, perhaps additional stud-ies should be considered.

Pamela Stratton, M.D.Craig Winkel, M.D.Ahalya Premkumar, M.D.Catherine Chow, M.D.Jan Wilson, R.N.Rhonda Hearns-Stokes, M.D.Sun Yeong Heo, Ph.D.Maria Merino, M.D.Lynnette K. Nieman, M.D.Warren G. Magnusen Clinical CenterNational Institutes of HealthBethesda, MarylandMay 28, 2003

Reference1. Stratton P, Winkel C, Premkumar A, Chow C, Wilson J, Hearns-Stokes

R, et al. Diagnostic accuracy of laparoscopy, magnetic resonance imag-ing, and histopathologic examination for the detection of endometriosis.Fertil Steril 2003;79:1078–85.

doi:10.1016/S0015-0282(03)01133-6

1072Vol. 80, No. 4, October 2003