- Email: [email protected]
Endometrial microstructure after long-term use of a 91-day extended-cycle oral contraceptive regimen
Contraception 71 (2005) 55 – 59
Original research article
Endometrial microstructure after long-term use of a 91-day extended-cycle oral contraceptive regimenB Freedolph D. Andersona,*, Howard Haitb, Jeng Hsiua, Allison L. Thompson-Gravesc, Walter H. Wilbornd, Robert F. Williamsa a
The Jones Institute for Reproductive Medicine, Eastern Virginia Medical School, Norfolk, VA 23501, USA b Data Management and Biostatistics, Barr Research, Inc., Bala Cynwyd, PA 19004, USA c Gulf Regional Pathologists, Mobile Infirmary Medical Center, Mobile, AL 36608, USA d Structural Research Center, Mobile, AL 36608, USA Received 19 May 2004; revised 21 July 2004; accepted 22 July 2004
Abstract Objective: To assess the effect on the endometrial microstructure of an extended-cycle oral contraceptive (OC) regimen containing ethinyl estradiol (EE) and levonorgestrel (LNG). Methodology: Subjects received up to four cycles of a 91-day extended-cycle OC regimen (84 consecutive days of monophasic 30 Ag EE/ 150 Ag LNG followed by 7 days of placebo). Endometrial biopsies were performed prior to the initiation and at the completion of therapy. All endometrial samples were processed centrally and reviewed by three independent pathologists blinded to treatment groups. Results: Endometrial biopsies were performed in 50 women. In general, samples taken after completion of therapy with no further hormonal exposure demonstrated rapid return to normal endometrial cycling. In contrast, the majority of subjects still on active extended hormonal OC therapy at the time of biopsy had inactive or atrophic endometrium. No intravascular blood clots were observed in any of the specimens. Conclusion: The endometrial findings observed in this cohort of women treated with a 91-day extended-cycle OC regimen for up to 1 year showed no significant pathology. Additionally, the endometrium reverted quickly to normal cyclic changes in those subjects who, after completing therapy, elected not to continue with hormonal contraception. D 2005 Elsevier Inc. All rights reserved. Keywords: Endometrial microstructure; Extended cycle oral contraceptive; Ethinyl estradiol; Levonorgestrel
1. Introduction Use of low-dose, combined oral contraceptives (OCs) inhibits normal proliferative changes in the endometrium and is often associated with inactive or atrophic endometrium [1]. Evaluation of histological findings for OCs has been conducted in relatively small subject populations; it has also been essentially limited to 28-day regimens [2–6]. A recent multicenter four-arm study was conducted to evaluate the safety and efficacy of a new 91-day extended-
cycle OC (Seasonale1) [7]. The study was sponsored by Barr Laboratories Inc. and involved 47 participating sites across the United States. This report describes results obtained from a cohort of women who received up to 1 year of therapy with the Seasonale regimen, which is composed of 30 Ag of ethinyl estradiol (EE) and 150 Ag of levonorgestrel (LNG) administered for 84 consecutive days followed by 7 days of placebo. The new regimen was designed to reduce the number of withdrawal bleeding episodes from 13 currently experienced with conventional OCs to only four per year. As part of the overall evaluation of study results, endometrial biopsies were performed in a
B
Financial Disclosure: FD Anderson is a consultant to Barr Laboratories and Warner-Chilcott. H Hait is employed by Barr Research, Inc. J Hsiu was a consultant to Barr Research. AG Thompson-Graves has no disclosures to report. WH Wilborn has no disclosures to report. RF Williams was a consultant to Barr Laboratories, Inc. * Corresponding author. Tel.: +1 757 490 2835; fax: +1 757 490 6689. E-mail address: [email protected] (F.D. Anderson). 0010-7824/$ – see front matter D 2005 Elsevier Inc. All rights reserved. doi:10.1016/j.contraception.2004.07.013
1 Barr Laboratories is the exclusive licensee of Seasonale, a registered trademark of the Medical College of Hampton Roads.
56
F.D. Anderson et al. / Contraception 71 (2005) 55–59
randomly selected subset of subjects who participated at selected study sites. 2. Methods 2.1. Study overview To assess the effect of extended OC regimen on the endometrium, a randomly chosen subset of 50 subjects from the 464 subjects randomized to Seasonale agreed, and provided additional written informed consent to undergo endometrial biopsies at baseline and at the end of treatment. The study was performed in accordance with the Declaration of Helsinki (Republic of South Africa, 1996) and applicable guidelines for Good Clinical Practice; ethics committee approval was also obtained at each participating clinical site. 2.2. Participants Five sites were chosen for the endometrial biopsy study (SEA 301 Endometrial Biopsy Cohort Study Group). Eligible subjects from the multicenter study at those sites who had been randomly assigned to Seasonale were recruited. No additional requirements were set other than the willingness of the subjects to participate. These subjects may or may not have taken OCs during the cycle prior to enrollment. 2.3. Study medication Study medication was supplied as blister packs containing a 91-day supply of study medication (84 active pills and 7 placebo). Each active pill contained EE 30 Ag and LNG 150 Ag. The active pills were taken once daily for 84 consecutive days followed by 7 days of placebo for a 91-day cycle. This regimen continued for up to four treatment bcyclesQ (364 days). 2.4. Endometrial biopsies The initial endometrial biopsy was performed during the third week of the OC cycle or menstrual cycle prior to initiation of study therapy. This was done after randomization but prior to the initiation of study therapy. The initial biopsy was utilized to rule out preexisting endometrial pathology such as atypical hyperplasia, polyps, chronic or acute inflammation, or endometrial carcinoma. The final biopsy was to be performed within 10 days prior to taking the last active treatment pill of the study (i.e., before the first inactive placebo pill). All biopsies were scheduled to be performed after receipt of a negative urine pregnancy test result and while subjects were not menstruating. End-oftreatment biopsies were performed during a subject’s last cycle of treatment, regardless of whether that subject had completed a full year of therapy or had discontinued the study prior to completing 1 year. Biopsies were conducted using the GynoSampler (Gynetics, Somerville, NJ, USA), were collected in 10%
neutral buffered formalin, and processed centrally (RFW). Samples were stained with hematoxylin and eosin and were identified only by number. Coordination of the evaluation of biopsy findings was done at Eastern Virginia Medical School under the supervision of the coordinating investigator for the study (FDA). Three independent reviewers (JH, ALT-G, WHW) scored each biopsy as bno endometrium,Q bendometrial sample insufficient for analysisQ or, for those biopsies with adequate tissue for evaluation, the glands were described as binactive,Q batrophic,Q bmenstrual,Q bproliferative,Q bsecretory,Q or bhyperplastic,Q with appropriate subcategorizations as possible depending on tissue adequacy. The reviewers were all blinded with respect to the treatment received by the subject and to the day on which the biopsy was performed. As a result, no reviewer would know whether the biopsy report represented a pretreatment or end-of-treatment biopsy. Stromal tissue was assessed for the following if endometrial tissue was present and sufficient for evaluation: general (not dense, dense), edema (none, mild, moderate, marked), congestion (none, mild, moderate, marked), hemorrhage (focal, multifocal), necrosis (focal, multifocal), hyalinization (focal, multifocal), vascular proliferation (none, mild, moderate, marked), and decidualization (none, mild, moderate, marked). Although the same classifications were used for both the pre- and end-of-treatment biopsy results, an end-of-treatment categorization of bmenstrualQ while the subject is still on active pills is more indicative of endometrium sloughing as a result of hormone withdrawal, or endometrial breakdown, not menstruation. For both the gland and stromal evaluations, each reviewer’s classification was given equal weight in the calculation of the frequencies. Thus, each eligible subject was counted as providing three replicate evaluations. Possible disagreement in classifications among reviewers was expected, and not considered as limiting in the interpretation of findings. Identification of the hormonal status of the subjects pretherapy or the assigned regimen at the end of therapy was not revealed until all biopsy reports had been finalized, entered into a database, and audited for accuracy. Twelve of the 50 participating subjects had their end-oftreatment biopsies performed during either the placebo/ inactive pill interval or after they had completed the study and were no longer on any hormonal contraceptive. This provided a unique opportunity to examine recovery of the endometrium after long-term extended-cycle therapy in such subjects directly in contrast to those who were still on active treatment at the time of biopsy. One subject had her end-oftreatment biopsy performed while on another OC and was, therefore, not included in this analysis. 2.5. Analysis Frequency distributions for each category of response for the glandular and stromal assessments were generated as percentages using the total number of biopsy results as the
F.D. Anderson et al. / Contraception 71 (2005) 55–59
denominator. The 49 analyzed subjects were grouped according to whether the end-of-treatment biopsy had been performed while the subject was still on active therapy or was no longer on any hormonal contraception. In addition, pretreatment biopsy results were grouped according to a whether a subject was receiving OCs at the time of enrollment (bOC user Q) or was not receiving OCs (bpriorQ user: no OC use for at least 6 months, or a bfresh start Q: never used OCs) prior to the study. The bprior Q and bfresh start Q categories were combined as bnon-OC user Q for analysis. 3. Results The subjects making up the endometrial biopsy cohort represent 10.8% of the women randomized to the extended OC regimen in the large multicenter study. The methodology and results of the pivotal study have been presented elsewhere [7]. The demographic characteristics of the subjects at baseline are summarized in Table 1. Mean age was 28.2 years (range: 18–38), more than 80% of subjects were Caucasian, and most subjects (86%) were nonsmokers. OC users, that is, those who had been regular users of OCs prior to participation in the study, made up over two thirds of participants; non-OC users at the time of enrollment (i.e., prior users and fresh starts) made up an additional 29% of subjects undergoing pretreatment endometrial biopsies. The mean weight of subjects in the cohort was 155.1 lb with a mean body mass index (BMI) of 26.3. Among the cohort of subjects, 37 had the end-oftreatment biopsy performed, on average, 9.2 days prior to the last dose date for the extended-cycle OC (SD = 5.2 days; range: 1–29 days). In all, 37 subjects had their Table 1 Demographic characteristics for the extended-cycle OC biopsy cohort Age (years) Weight (lb) BMI Race
Smoking status OC use status
Timing of end-of-study biopsy in terms of last dose of study extended-cycle OC Days from last active dose to end-of-study biopsy Before On/After
Mean (SD) Range Mean (SD) Range Mean (SD) Range Caucasian Black Other Nonsmoker User Prior user Fresh start Before last dose On/after last dose
28.2 (6.0) 18 –38 155.0 (40.2) 107.5–288.0 26.3 (6.3) 18.5– 45.2 41 (83.7%) 4 (8.2%) 4 (8.2%) 42 (85.7%) 33 (67.4%) 14 (28.6%) 2 (4.0%) 37 (75.5%) 12 (24.5%)
Mean (SD) Range Mean (SD) Range
9.2 (5.2) 1–29 10.2 (16.6) 0 –58
57
biopsy performed while they were still in the active pill phase of their last extended cycle. Twelve subjects had the biopsy performed after they had completed extended-cycle OC therapy and reported no other OC use after the last extended-cycle dose. These biopsies were completed anywhere from the last dose date to up to 58 days after the last extended-cycle dose [mean (SD) = 10.2 (16.6) days]. Note that biopsies performed on the last dose date were conducted when the subject was taking inactive placebo pills at the end of the last extended cycle. A total of 150 readings were obtained for all subjects. Nineteen (12.7%) of the readings at baseline and at the end of treatment were scored as either bno endometriumQ or bendometrium insufficient for analysis.Q This is consistent with the failure rate of endometrial biopsy in other continuous hormone use endometrial biopsy studies [4,8]. No pretherapy biopsy showed any significant pathology. The classification of findings at pretherapy biopsy was consistent with the distribution of subjects defined as OC users or non-OC users (prior users or fresh starts) (Table 1). 3.1. Glandular assessment Fig. 1 summarizes the results of the glandular assessments for all subjects at baseline, whether or not the subjects were receiving OCs prior to enrollment in the study. The findings are variable, but nearly 30% of the OC users had inactive or atrophic endometrium, as opposed to only about 10% of non-OC subjects. It is interesting that more than half (55.9%) of the OC users had active (proliferative or secretory) endometrium. This may or may not be significant, as the relatively smaller numbers of non-OC users limits interpretation of those subgroups. Fig. 2 depicts the results of end-of-treatment biopsies broken down for subjects who either were, or no longer were, on active extended-cycle treatment at the time of the biopsy. End-of-treatment biopsies sampled before active medication was completed showed primarily inactive or atrophic endometrium (64.9% of the readings). End-oftreatment biopsies sampled after active medication had been completed (subjects on placebo pills, or not taking followup OC) showed changes consistent with return to cycling endometrium. However, 36.1% of the readings suggested that inactive or atrophic endometrium was still present. This may be a function of the timing of the biopsy in relation to the last active dose of medication. Six of the 12 subjects who had biopsies performed after study completion had one or more readings categorized as binactiveQ or batrophicQ; 4 of these same subjects had the biopsy performed no more than 7 days after active treatment. In addition, the subject who was excluded from the analysis because her end-oftreatment biopsy was performed while she was on another OC showed changes consistent with those subjects still on active extended-cycle pills at the time of the end-oftreatment biopsy.
58
F.D. Anderson et al. / Contraception 71 (2005) 55–59
Fig. 1. Results of histological review of endometrial biopsies at baseline for users of OCs and non-users (fresh starts/prior users).
A greater increase in proliferative activity at the end-oftreatment was seen in the baseline non-OC users compared to baseline OC users. It is unlikely that after 1 year of use of an extended-cycle OC regimen, recovery should be faster in subjects who a year earlier were not on active hormonal medication. The relatively smaller numbers of subjects who were not on OCs at baseline may limit any interpretation of this finding. 3.2. Stromal evaluation Results from end-of-treatment biopsies in subjects still on extended OC and those no longer on hormonal contraception illustrate no loss of endometrial architecture and lack of overhyalinization due to overstimulation of stromal cells. Stromal vascularization was consistent with what has been seen with 28-day OC regimens. Specifically,
there are minimal differences in density, and only slight decreases in edema in the subjects who were not on OCs at the time of the end-of-treatment biopsy. There are minimal increases in congestion in subjects who were off active treatment at the time of the biopsy, but this may indicate approaching endometrial breakdown from hormone withdrawal. Hemorrhage and necrosis are very similar in both groups at end of therapy, as are hyalinization and vascular proliferation. The rapid recovery from hormone effect may be illustrated by the decrease in hyalinization seen in the subjects who had recently discontinued active medication. 4. Discussion This combination of a large study population and endometrial biopsy evaluation after a full year of
Fig. 2. Results of histological review of endometrial biopsies at end of treatment for women on LNG 0.15 mg/EE 0.030 mg 91-day extended-cycle OC therapy (glandular).
F.D. Anderson et al. / Contraception 71 (2005) 55–59
extended-cycle therapy substantiates the validity of results and the overall safety of the 91-day extended-cycle regimen of oral contraception. Based on the results observed in subjects who elected to discontinue hormonal contraception at the end of the Seasonale study, there appears to be a rapid return of ovarian/endometrial cycling. As in any small study of a cohort subset of subjects, inferences may be somewhat limited. As a result, some caution is required when attempting to derive definitive conclusions. This study does, however, provide important, useful information not previously available. In particular, it is one of the largest endometrial biopsy studies of OCs that included a baseline biopsy followed by a biopsy after up to 1 year of use, both of which were reviewed independently by expert pathologists. This is also the first report of a direct comparison of histological findings among subjects administered an extended-cycle OC regimen who are still on active OC therapy and those who are no longer on any hormonal contraception. Finally, these data uniquely evaluated endometrial changes within a 91-day extendedcycle OC regimen providing for only four withdrawal periods per year. 5. Conclusions The 91-day extended-cycle OC regimen does not lead to hyperplasia or other endometrial pathology after up to 1 year of use. The results of this study of endometrial biopsies conducted pre- and posttherapy with a 91-day extendedcycle OC regimen provide additional assurance that the regimen is safe.
59
Acknowledgements This study was supported by Barr Research, Inc., Bala Cynwyd, PA. The editorial assistance of Glenmere Research Inc., Montebello, NY, is acknowledged. SEA 301 Endometrial Biopsy Cohort Study Group: Freedolph D. Anderson, MD, Norfolk, VA; Davis Walsh Baldwin, MD, Palo Alto, CA; Irwin Kerber, MD, Dallas, TX; Alfred Moffet, MD, Leesburg, FL; James Simon, MD, Laurel, MD.
References [1] Deligdisch L. Effects of hormone therapy on the endometrium. Mod Pathol 1993;6:94 – 106. [2] Song J, Mark MS, Lawler Jr MP. Endometrial changes in women receiving oral contraceptives. Am J Obstet Gynecol 1970;107:717 – 28. [3] El-zenieny AH, El-kabarity H, Hafiz MA, Fikry MF. A clinicopathological study of Neogynon as a new oral contraceptive steroid. Ain Shams Med J 1976;27:213 – 6. [4] Bahamondes L, Maradiegue E, Diaz J, et al. Endometrial histology in long-term users of the once-a-month injectable contraceptive Cyclofem. Adv Contracept 1999;15:1 – 7. [5] Ludicke F, Johannisson E, Helmerhorst FM, Campana A, Foidart J, Heithecker R. Effect of a combined oral contraceptive containing 3 mg of drosperinone and 30 microg of ethinyl estradiol on the human endometrium. Fertil Steril 2001;76:102 – 7. [6] Archer DF. Endometrial histology during use of a low-dose estrogen– desogestrel oral contraceptive with a reduced hormone-free interval. Contraception 1999;60:151 – 4. [7] Anderson FD, Hait H. A multicenter, randomized study of an extended cycle oral contraceptive. Contraception 2003;68:89 – 96. [8] Wells M, et al. for the UK Continuous Combined Hormone Replacement Therapy Study Investigators. Effect on endometrium of long term treatment with continuous combined oestrogen–progestogen replacement therapy: follow up study. Br J Med 2002;325:239 – 44.
Original research article
Endometrial microstructure after long-term use of a 91-day extended-cycle oral contraceptive regimenB Freedolph D. Andersona,*, Howard Haitb, Jeng Hsiua, Allison L. Thompson-Gravesc, Walter H. Wilbornd, Robert F. Williamsa a
The Jones Institute for Reproductive Medicine, Eastern Virginia Medical School, Norfolk, VA 23501, USA b Data Management and Biostatistics, Barr Research, Inc., Bala Cynwyd, PA 19004, USA c Gulf Regional Pathologists, Mobile Infirmary Medical Center, Mobile, AL 36608, USA d Structural Research Center, Mobile, AL 36608, USA Received 19 May 2004; revised 21 July 2004; accepted 22 July 2004
Abstract Objective: To assess the effect on the endometrial microstructure of an extended-cycle oral contraceptive (OC) regimen containing ethinyl estradiol (EE) and levonorgestrel (LNG). Methodology: Subjects received up to four cycles of a 91-day extended-cycle OC regimen (84 consecutive days of monophasic 30 Ag EE/ 150 Ag LNG followed by 7 days of placebo). Endometrial biopsies were performed prior to the initiation and at the completion of therapy. All endometrial samples were processed centrally and reviewed by three independent pathologists blinded to treatment groups. Results: Endometrial biopsies were performed in 50 women. In general, samples taken after completion of therapy with no further hormonal exposure demonstrated rapid return to normal endometrial cycling. In contrast, the majority of subjects still on active extended hormonal OC therapy at the time of biopsy had inactive or atrophic endometrium. No intravascular blood clots were observed in any of the specimens. Conclusion: The endometrial findings observed in this cohort of women treated with a 91-day extended-cycle OC regimen for up to 1 year showed no significant pathology. Additionally, the endometrium reverted quickly to normal cyclic changes in those subjects who, after completing therapy, elected not to continue with hormonal contraception. D 2005 Elsevier Inc. All rights reserved. Keywords: Endometrial microstructure; Extended cycle oral contraceptive; Ethinyl estradiol; Levonorgestrel
1. Introduction Use of low-dose, combined oral contraceptives (OCs) inhibits normal proliferative changes in the endometrium and is often associated with inactive or atrophic endometrium [1]. Evaluation of histological findings for OCs has been conducted in relatively small subject populations; it has also been essentially limited to 28-day regimens [2–6]. A recent multicenter four-arm study was conducted to evaluate the safety and efficacy of a new 91-day extended-
cycle OC (Seasonale1) [7]. The study was sponsored by Barr Laboratories Inc. and involved 47 participating sites across the United States. This report describes results obtained from a cohort of women who received up to 1 year of therapy with the Seasonale regimen, which is composed of 30 Ag of ethinyl estradiol (EE) and 150 Ag of levonorgestrel (LNG) administered for 84 consecutive days followed by 7 days of placebo. The new regimen was designed to reduce the number of withdrawal bleeding episodes from 13 currently experienced with conventional OCs to only four per year. As part of the overall evaluation of study results, endometrial biopsies were performed in a
B
Financial Disclosure: FD Anderson is a consultant to Barr Laboratories and Warner-Chilcott. H Hait is employed by Barr Research, Inc. J Hsiu was a consultant to Barr Research. AG Thompson-Graves has no disclosures to report. WH Wilborn has no disclosures to report. RF Williams was a consultant to Barr Laboratories, Inc. * Corresponding author. Tel.: +1 757 490 2835; fax: +1 757 490 6689. E-mail address: [email protected] (F.D. Anderson). 0010-7824/$ – see front matter D 2005 Elsevier Inc. All rights reserved. doi:10.1016/j.contraception.2004.07.013
1 Barr Laboratories is the exclusive licensee of Seasonale, a registered trademark of the Medical College of Hampton Roads.
56
F.D. Anderson et al. / Contraception 71 (2005) 55–59
randomly selected subset of subjects who participated at selected study sites. 2. Methods 2.1. Study overview To assess the effect of extended OC regimen on the endometrium, a randomly chosen subset of 50 subjects from the 464 subjects randomized to Seasonale agreed, and provided additional written informed consent to undergo endometrial biopsies at baseline and at the end of treatment. The study was performed in accordance with the Declaration of Helsinki (Republic of South Africa, 1996) and applicable guidelines for Good Clinical Practice; ethics committee approval was also obtained at each participating clinical site. 2.2. Participants Five sites were chosen for the endometrial biopsy study (SEA 301 Endometrial Biopsy Cohort Study Group). Eligible subjects from the multicenter study at those sites who had been randomly assigned to Seasonale were recruited. No additional requirements were set other than the willingness of the subjects to participate. These subjects may or may not have taken OCs during the cycle prior to enrollment. 2.3. Study medication Study medication was supplied as blister packs containing a 91-day supply of study medication (84 active pills and 7 placebo). Each active pill contained EE 30 Ag and LNG 150 Ag. The active pills were taken once daily for 84 consecutive days followed by 7 days of placebo for a 91-day cycle. This regimen continued for up to four treatment bcyclesQ (364 days). 2.4. Endometrial biopsies The initial endometrial biopsy was performed during the third week of the OC cycle or menstrual cycle prior to initiation of study therapy. This was done after randomization but prior to the initiation of study therapy. The initial biopsy was utilized to rule out preexisting endometrial pathology such as atypical hyperplasia, polyps, chronic or acute inflammation, or endometrial carcinoma. The final biopsy was to be performed within 10 days prior to taking the last active treatment pill of the study (i.e., before the first inactive placebo pill). All biopsies were scheduled to be performed after receipt of a negative urine pregnancy test result and while subjects were not menstruating. End-oftreatment biopsies were performed during a subject’s last cycle of treatment, regardless of whether that subject had completed a full year of therapy or had discontinued the study prior to completing 1 year. Biopsies were conducted using the GynoSampler (Gynetics, Somerville, NJ, USA), were collected in 10%
neutral buffered formalin, and processed centrally (RFW). Samples were stained with hematoxylin and eosin and were identified only by number. Coordination of the evaluation of biopsy findings was done at Eastern Virginia Medical School under the supervision of the coordinating investigator for the study (FDA). Three independent reviewers (JH, ALT-G, WHW) scored each biopsy as bno endometrium,Q bendometrial sample insufficient for analysisQ or, for those biopsies with adequate tissue for evaluation, the glands were described as binactive,Q batrophic,Q bmenstrual,Q bproliferative,Q bsecretory,Q or bhyperplastic,Q with appropriate subcategorizations as possible depending on tissue adequacy. The reviewers were all blinded with respect to the treatment received by the subject and to the day on which the biopsy was performed. As a result, no reviewer would know whether the biopsy report represented a pretreatment or end-of-treatment biopsy. Stromal tissue was assessed for the following if endometrial tissue was present and sufficient for evaluation: general (not dense, dense), edema (none, mild, moderate, marked), congestion (none, mild, moderate, marked), hemorrhage (focal, multifocal), necrosis (focal, multifocal), hyalinization (focal, multifocal), vascular proliferation (none, mild, moderate, marked), and decidualization (none, mild, moderate, marked). Although the same classifications were used for both the pre- and end-of-treatment biopsy results, an end-of-treatment categorization of bmenstrualQ while the subject is still on active pills is more indicative of endometrium sloughing as a result of hormone withdrawal, or endometrial breakdown, not menstruation. For both the gland and stromal evaluations, each reviewer’s classification was given equal weight in the calculation of the frequencies. Thus, each eligible subject was counted as providing three replicate evaluations. Possible disagreement in classifications among reviewers was expected, and not considered as limiting in the interpretation of findings. Identification of the hormonal status of the subjects pretherapy or the assigned regimen at the end of therapy was not revealed until all biopsy reports had been finalized, entered into a database, and audited for accuracy. Twelve of the 50 participating subjects had their end-oftreatment biopsies performed during either the placebo/ inactive pill interval or after they had completed the study and were no longer on any hormonal contraceptive. This provided a unique opportunity to examine recovery of the endometrium after long-term extended-cycle therapy in such subjects directly in contrast to those who were still on active treatment at the time of biopsy. One subject had her end-oftreatment biopsy performed while on another OC and was, therefore, not included in this analysis. 2.5. Analysis Frequency distributions for each category of response for the glandular and stromal assessments were generated as percentages using the total number of biopsy results as the
F.D. Anderson et al. / Contraception 71 (2005) 55–59
denominator. The 49 analyzed subjects were grouped according to whether the end-of-treatment biopsy had been performed while the subject was still on active therapy or was no longer on any hormonal contraception. In addition, pretreatment biopsy results were grouped according to a whether a subject was receiving OCs at the time of enrollment (bOC user Q) or was not receiving OCs (bpriorQ user: no OC use for at least 6 months, or a bfresh start Q: never used OCs) prior to the study. The bprior Q and bfresh start Q categories were combined as bnon-OC user Q for analysis. 3. Results The subjects making up the endometrial biopsy cohort represent 10.8% of the women randomized to the extended OC regimen in the large multicenter study. The methodology and results of the pivotal study have been presented elsewhere [7]. The demographic characteristics of the subjects at baseline are summarized in Table 1. Mean age was 28.2 years (range: 18–38), more than 80% of subjects were Caucasian, and most subjects (86%) were nonsmokers. OC users, that is, those who had been regular users of OCs prior to participation in the study, made up over two thirds of participants; non-OC users at the time of enrollment (i.e., prior users and fresh starts) made up an additional 29% of subjects undergoing pretreatment endometrial biopsies. The mean weight of subjects in the cohort was 155.1 lb with a mean body mass index (BMI) of 26.3. Among the cohort of subjects, 37 had the end-oftreatment biopsy performed, on average, 9.2 days prior to the last dose date for the extended-cycle OC (SD = 5.2 days; range: 1–29 days). In all, 37 subjects had their Table 1 Demographic characteristics for the extended-cycle OC biopsy cohort Age (years) Weight (lb) BMI Race
Smoking status OC use status
Timing of end-of-study biopsy in terms of last dose of study extended-cycle OC Days from last active dose to end-of-study biopsy Before On/After
Mean (SD) Range Mean (SD) Range Mean (SD) Range Caucasian Black Other Nonsmoker User Prior user Fresh start Before last dose On/after last dose
28.2 (6.0) 18 –38 155.0 (40.2) 107.5–288.0 26.3 (6.3) 18.5– 45.2 41 (83.7%) 4 (8.2%) 4 (8.2%) 42 (85.7%) 33 (67.4%) 14 (28.6%) 2 (4.0%) 37 (75.5%) 12 (24.5%)
Mean (SD) Range Mean (SD) Range
9.2 (5.2) 1–29 10.2 (16.6) 0 –58
57
biopsy performed while they were still in the active pill phase of their last extended cycle. Twelve subjects had the biopsy performed after they had completed extended-cycle OC therapy and reported no other OC use after the last extended-cycle dose. These biopsies were completed anywhere from the last dose date to up to 58 days after the last extended-cycle dose [mean (SD) = 10.2 (16.6) days]. Note that biopsies performed on the last dose date were conducted when the subject was taking inactive placebo pills at the end of the last extended cycle. A total of 150 readings were obtained for all subjects. Nineteen (12.7%) of the readings at baseline and at the end of treatment were scored as either bno endometriumQ or bendometrium insufficient for analysis.Q This is consistent with the failure rate of endometrial biopsy in other continuous hormone use endometrial biopsy studies [4,8]. No pretherapy biopsy showed any significant pathology. The classification of findings at pretherapy biopsy was consistent with the distribution of subjects defined as OC users or non-OC users (prior users or fresh starts) (Table 1). 3.1. Glandular assessment Fig. 1 summarizes the results of the glandular assessments for all subjects at baseline, whether or not the subjects were receiving OCs prior to enrollment in the study. The findings are variable, but nearly 30% of the OC users had inactive or atrophic endometrium, as opposed to only about 10% of non-OC subjects. It is interesting that more than half (55.9%) of the OC users had active (proliferative or secretory) endometrium. This may or may not be significant, as the relatively smaller numbers of non-OC users limits interpretation of those subgroups. Fig. 2 depicts the results of end-of-treatment biopsies broken down for subjects who either were, or no longer were, on active extended-cycle treatment at the time of the biopsy. End-of-treatment biopsies sampled before active medication was completed showed primarily inactive or atrophic endometrium (64.9% of the readings). End-oftreatment biopsies sampled after active medication had been completed (subjects on placebo pills, or not taking followup OC) showed changes consistent with return to cycling endometrium. However, 36.1% of the readings suggested that inactive or atrophic endometrium was still present. This may be a function of the timing of the biopsy in relation to the last active dose of medication. Six of the 12 subjects who had biopsies performed after study completion had one or more readings categorized as binactiveQ or batrophicQ; 4 of these same subjects had the biopsy performed no more than 7 days after active treatment. In addition, the subject who was excluded from the analysis because her end-oftreatment biopsy was performed while she was on another OC showed changes consistent with those subjects still on active extended-cycle pills at the time of the end-oftreatment biopsy.
58
F.D. Anderson et al. / Contraception 71 (2005) 55–59
Fig. 1. Results of histological review of endometrial biopsies at baseline for users of OCs and non-users (fresh starts/prior users).
A greater increase in proliferative activity at the end-oftreatment was seen in the baseline non-OC users compared to baseline OC users. It is unlikely that after 1 year of use of an extended-cycle OC regimen, recovery should be faster in subjects who a year earlier were not on active hormonal medication. The relatively smaller numbers of subjects who were not on OCs at baseline may limit any interpretation of this finding. 3.2. Stromal evaluation Results from end-of-treatment biopsies in subjects still on extended OC and those no longer on hormonal contraception illustrate no loss of endometrial architecture and lack of overhyalinization due to overstimulation of stromal cells. Stromal vascularization was consistent with what has been seen with 28-day OC regimens. Specifically,
there are minimal differences in density, and only slight decreases in edema in the subjects who were not on OCs at the time of the end-of-treatment biopsy. There are minimal increases in congestion in subjects who were off active treatment at the time of the biopsy, but this may indicate approaching endometrial breakdown from hormone withdrawal. Hemorrhage and necrosis are very similar in both groups at end of therapy, as are hyalinization and vascular proliferation. The rapid recovery from hormone effect may be illustrated by the decrease in hyalinization seen in the subjects who had recently discontinued active medication. 4. Discussion This combination of a large study population and endometrial biopsy evaluation after a full year of
Fig. 2. Results of histological review of endometrial biopsies at end of treatment for women on LNG 0.15 mg/EE 0.030 mg 91-day extended-cycle OC therapy (glandular).
F.D. Anderson et al. / Contraception 71 (2005) 55–59
extended-cycle therapy substantiates the validity of results and the overall safety of the 91-day extended-cycle regimen of oral contraception. Based on the results observed in subjects who elected to discontinue hormonal contraception at the end of the Seasonale study, there appears to be a rapid return of ovarian/endometrial cycling. As in any small study of a cohort subset of subjects, inferences may be somewhat limited. As a result, some caution is required when attempting to derive definitive conclusions. This study does, however, provide important, useful information not previously available. In particular, it is one of the largest endometrial biopsy studies of OCs that included a baseline biopsy followed by a biopsy after up to 1 year of use, both of which were reviewed independently by expert pathologists. This is also the first report of a direct comparison of histological findings among subjects administered an extended-cycle OC regimen who are still on active OC therapy and those who are no longer on any hormonal contraception. Finally, these data uniquely evaluated endometrial changes within a 91-day extendedcycle OC regimen providing for only four withdrawal periods per year. 5. Conclusions The 91-day extended-cycle OC regimen does not lead to hyperplasia or other endometrial pathology after up to 1 year of use. The results of this study of endometrial biopsies conducted pre- and posttherapy with a 91-day extendedcycle OC regimen provide additional assurance that the regimen is safe.
59
Acknowledgements This study was supported by Barr Research, Inc., Bala Cynwyd, PA. The editorial assistance of Glenmere Research Inc., Montebello, NY, is acknowledged. SEA 301 Endometrial Biopsy Cohort Study Group: Freedolph D. Anderson, MD, Norfolk, VA; Davis Walsh Baldwin, MD, Palo Alto, CA; Irwin Kerber, MD, Dallas, TX; Alfred Moffet, MD, Leesburg, FL; James Simon, MD, Laurel, MD.
References [1] Deligdisch L. Effects of hormone therapy on the endometrium. Mod Pathol 1993;6:94 – 106. [2] Song J, Mark MS, Lawler Jr MP. Endometrial changes in women receiving oral contraceptives. Am J Obstet Gynecol 1970;107:717 – 28. [3] El-zenieny AH, El-kabarity H, Hafiz MA, Fikry MF. A clinicopathological study of Neogynon as a new oral contraceptive steroid. Ain Shams Med J 1976;27:213 – 6. [4] Bahamondes L, Maradiegue E, Diaz J, et al. Endometrial histology in long-term users of the once-a-month injectable contraceptive Cyclofem. Adv Contracept 1999;15:1 – 7. [5] Ludicke F, Johannisson E, Helmerhorst FM, Campana A, Foidart J, Heithecker R. Effect of a combined oral contraceptive containing 3 mg of drosperinone and 30 microg of ethinyl estradiol on the human endometrium. Fertil Steril 2001;76:102 – 7. [6] Archer DF. Endometrial histology during use of a low-dose estrogen– desogestrel oral contraceptive with a reduced hormone-free interval. Contraception 1999;60:151 – 4. [7] Anderson FD, Hait H. A multicenter, randomized study of an extended cycle oral contraceptive. Contraception 2003;68:89 – 96. [8] Wells M, et al. for the UK Continuous Combined Hormone Replacement Therapy Study Investigators. Effect on endometrium of long term treatment with continuous combined oestrogen–progestogen replacement therapy: follow up study. Br J Med 2002;325:239 – 44.