- Email: [email protected]
Intercycle variability of the ovarian response in patients undergoing repeated stimulation with corifollitropin alfa in a gonadotropin-releasing hormone antagonist protocol
ORIGINAL ARTICLE: ASSISTED REPRODUCTION 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59
Intercycle variability of the ovarian response in patients undergoing repeated stimulation with corifollitropin alfa in a gonadotropin-releasing hormone antagonist protocol Q10
Luk Rombauts, Ph.D.,a,b Cornelis B. Lambalk, M.D., Ph.D.,c Askan Schultze-Mosgau, Ph.D.,d Jacqueline van Kuijk, M.Sc.,e Pierre Verweij, Ph.D.,e Davis Gates, Ph.D.,f Keith Gordon, Ph.D.,f and Georg Griesinger, Ph.D.d
a Department of Obstetrics and Gynaecology, Monash University, Clayton, Victoria, Australia; b Monash IVF, Clayton, Victoria, Australia; c Department of Obstetrics/Gynecology, VU University Medical Center, Amsterdam, the Netherlands; d Department of Reproductive Medicine and Gynecologic Endocrinology, University Clinic of Schleswig-Holstein, Campus e f Q1 Luebeck, Luebeck, Germany; MSD Oss N.V., Oss, the Netherlands; and Merck & Co., Inc., Kenilworth, New Jersey
Objective: To determine whether individual subject variation in ovarian response between repeated cycles with the same ovarian stimulation protocol can be predicted. Design: Retrospective data analysis. Setting: Not applicable Patient(s): Women aged 18–39 from a phase 3, open-label, uncontrolled trial with complete data across all cycles (n ¼ 176). Intervention(s): Up to three cycles of a single injection of 150 mg corifollitropin alfa for 7 days, then daily recombinant FSH/hMG until three follicles reached R17 mm. Gonadotropin-releasing hormone antagonist from stimulation day 5 until day of hCG administration. Main Outcome Measure(s): Numbers of follicles R11 mm on day of hCG in cycles 1–3, transition in ovarian response type between cycles from low (0–<6), normal (6–<18), and high (R18), and serum FSH concentrations and antral follicle count (AFC) at each cycle start. Result(s): The mean (SD) numbers of follicles R11 mm on day of hCG were 13.4 (6.2), 13.3 (5.4), and 13.8 (6.4) in cycles 1, 2 and 3, respectively. Between cycles 1 and 2, 11.9% switched from normal to low or high response, and 12.5% switched from low or high to normal response; 75.6% remained in the same category. Between cycles 2 and 3, 15.9% switched from normal to low or high response, and 10.2% switched from low or high to normal response; 73.9% remained in the same category. These shifts are symmetrical in nature, in that the percentage of subjects who shift from normal to low or high response is comparable to the percentage of subjects who shift from low or high to normal response. Baseline FSH and AFC did not significantly predict transition in ovarian response.
Received January 29, 2015; revised and accepted June 22, 2015. L.R. is a minority shareholder in Monash IVF and has provided expert advice to, and received unconditional research and educational grants from, MSD, Merck Serono, and Ferring. C.B.L. has received educational and research grants for his department from MSD, Merck Serono, Ferring, and Auxogyn; speaker fees from MSD, Merck Serono, Ferring, and Auxogyn; and consultant fees from Ferring. J.v.K. and P.V. were employees of MSD Oss N.V. at the time the study was conducted. D.G. and K.G. are employees of Merck & Co., Inc. and may own stock/stock options in the company. G.G. has received consultant/honorarium fees from MSD, Ferring, Glycotope, Serono, Finox, Vitrolife, and IBSA and has served on speaker bureaus for MSD, Ferring, Serono, Vitrolife, and IBSA. Support provided by Merck & Co. Inc., Kenilworth, New Jersey. Reprint requests: Luk Rombauts, Ph.D., Monash Medical Centre, 246 Clayton Road, Clayton, Victoria 3168, Australia (E-mail: [email protected]). Fertility and Sterility® Vol. -, No. -, - 2015 0015-0282/$36.00 Copyright ©2015 American Society for Reproductive Medicine, Published by Elsevier Inc. http://dx.doi.org/10.1016/j.fertnstert.2015.06.027 VOL. - NO. - / - 2015
1
FLA 5.2.0 DTD FNS29666_proof 14 July 2015 4:37 pm ce JS
60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118
ORIGINAL ARTICLE: ASSISTED REPRODUCTION 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177
Conclusion(s): The variability in ovarian responses between repeated cycles using the same protocol was not explained by baseline FSH and AFC. Clinical Trial Registration Number: NCT00696878 Protocol P05714. (Fertil SterilÒ 2015;-: -–-. Ó2015 by American Society for Reproductive Medicine.) Key Words: Corifollitropin alfa, GnRH antagonist, repeated ovarian stimulation, ovarian response, controlled ovarian stimulation Discuss: You can discuss this article with its authors and with other ASRM members at http:// fertstertforum.com/rombautsl-ovarian-response-intercycle-variability/
A
large number of studies have attempted to predict ovarian response to ovarian stimulation for IVF, and by altering the stimulation regimen, to ultimately impact on ovarian response and IVF treatment outcome (1– 6). To put such studies in perspective, it is mandatory to know the naturally occurring variability in ovarian response between subsequent treatment cycles within one subject. For such assessment, repetitive cycles should be performed, ideally within a short time frame, and stimulation (e.g., FSH dose, GnRH analogue protocol, and decisions on patient management) should be kept constant. In the Trust trial, a large population of women received ovarian stimulation uniformly with 150 mg corifollitropin alfa (a long-acting recombinant FSH [rFSH]) in a GnRH antagonist protocol for up to three cycles (7). Serum FSH concentrations and antral follicle count (AFC) on day 1 of ovarian stimulation have been identified as predictors of low or high ovarian response in GnRH antagonist cycles (8). The objectives of the present analysis were to determine the variation between treatment cycles in terms of numbers of growing follicles, to calculate the relative frequency of a switch between cycles from a normal ovarian response to a high or low ovarian response and vice versa, and to investigate whether the changes in ovarian response per cycle can be explained by changes in predictors as, for example, AFC and serum hormones at the start of each cycle of stimulation.
MATERIALS AND METHODS Study Design and Population
Q2
Trust was a multicenter, open-label, uncontrolled clinical trial, the details of which have been reported previously (7). The study was conducted in accordance with principles of good clinical practice and was approved by the appropriate institutional review boards and regulatory agencies. Written, informed consent was provided by all subjects. Women aged 18–39 years with a body weight of >60 kg and a regular menstrual cycle underwent up to three ovarian stimulation cycles. No women were taking birth control pills before beginning stimulation. Patients who conceived in a cycle were not allowed another cycle. Demographic and fertility characteristics of patients in the Trust trial have been reported previously (7, 9). In each cycle, subjects received a single injection of 150 mg corifollitropin alfa (Elonva, N.V. Organon) for the first 7 days of stimulation, followed by a daily dose of %225 IU FSH (either rFSH or hMG) starting on day 8 until the criterion
Use your smartphone to scan this QR code and connect to the discussion forum for this article now.* * Download a free QR code scanner by searching for “QR scanner” in your smartphone’s app store or app marketplace.
for oocyte maturation with hCG was reached (three follicles R17 mm). A GnRH antagonist (0.25 mg ganirelix or cetrorelix acetate) was administered once daily, starting on stimulation day 5 up to and including the day of hCG administration. Either urinary hCG (10,000 IU or 5,000 IU in case of a high ovarian response) or recombinant hCG (250 mg) was administered for final oocyte maturation. In case of too high of an ovarian response, the dose of FSH could be reduced, or withheld for a maximum of 3 days up to and including the day of hCG administration. For normal responders, the recommended daily dose of FSH was 150 IU. If the ovarian response was too high in the opinion of the investigator, the cycle could be cancelled at any time. If there was a risk of ovarian hyperstimulation syndrome (>30 follicles R11 mm on transvaginal ultrasound), hCG was withheld and the treatment cycle was canceled. The maximum total duration of stimulation was 19 days. The mean number of days (95% confidence interval [CI]) between cycles were 144.3 (131.1–157.5) and 171.3 (155.4–187.2), for cycle 1 to cycle 2 and cycle 2 to cycle 3, respectively.
Statistical Analysis Women who received hCG and who had non-missing values across all three cycles for the number of follicles R11 mm on the day of hCG were included in the analysis (n ¼ 176). The numbers of follicles R11 mm on the day of hCG for each cycle were extracted from a longitudinal data analysis on patients repeated across cycles and correcting for region. Given that each patient was measured across three cycles, an unstructured variance–covariance matrix was specified to allow for a separate estimate of variability between and within cycles, accounting for the between-cycle correlation from repeated measurements across patients. Additionally, patients were assigned into subgroups if they received the same dose of rFSH stimulation between two consecutive cycles (cycles 1 and 2; and cycles 2 and 3). No correction for multiplicity was made in this exploratory analysis. The number of follicles R11 mm on the day of hCG and the number of oocytes retrieved were also categorized as low responders (0–<6), normal responders (6–<18), and high responders (8) and summarized (frequencies and percentages) by cycle. Switches from one category to another from cycle 1 to cycle 2 and between cycle 2 and cycle 3 were also summarized. The significance of the directionality of switches was evaluated using Bowker's Test of Symmetry (i.e., test the null hypothesis that nij ¼ nji, where the number of patients are VOL. - NO. - / - 2015
2
FLA 5.2.0 DTD FNS29666_proof 14 July 2015 4:37 pm ce JS
178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236
Fertility and Sterility® 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283 284 285 286 287 288 289 290 291 292 293 294 295
TABLE 1 Demographic and stimulation characteristics: three cycle completers. Cycle 1 (n [ 176)
Baseline Age (y) Body mass index (kg/m2) Race Asian Black Caucasian Other Primary infertility Duration of infertility (y) Cause of infertilitya Male factor Tubal factor Endometriosis Cervical mucus problems Unexplained infertility Other Start cycle Duration of stimulation (d) Total dose of FSH from day 8 onward (IU) AFCb Serum E2 concentration (pmol/L)c Serum FSH concentration (IU/L)c Serum LH concentration (IU/L)c Day of hCG Serum E2 concentration (pmol/L)d Serum FSH concentration (IU/L)d Serum LH concentration (IU/L)d
Cycle 2 (n [ 176)
Cycle 3 (n [ 176)
10.2 (1.3) 455.3 (277.5) 10.8 (4.4) 125.1 (83.0) 7.2 (1.9) 5.0 (1.8)
10.3 (1.7) 478.7 (316.6) 11.0 (4.7) 117.8 (41.1) 7.0 (2.0) 5.0 (1.6)
10.4 (1.8) 504.8 (362.8) 11.1 (4.7) 117.1 (40.8) 7.5 (5.4) 5.0 (1.9)
5,442.4 (3,354.7) 13.7 (3.2) 1.6 (1.8)
5,210.2 (2,822.7) 13.6 (3.7) 1.6 (1.4)
5,743.1 (3,754.9) 13.6 (3.5) 1.8 (2.7)
33.3 (3.5) 24.3 (2.3) 2.3 0.6 96.0 1.1 55.7 3.6 (2.7) 58.0 28.4 13.6 0.6 15.9 1.7
Note: Data are presented as mean (SD) or percentage. a A patient could have multiple causes of infertility. b Ten, eight, and four missing values in cycles 1, 2, and 3, respectively. c Five, ten, and eight missing values. d Two, five, and five missing values. Rombauts. Ovarian response intercycle variability. Fertil Steril 2015.
compared in cell number numbers 12 vs. 21, 13 vs. 31, and 23 vs. 32 in a 3 3 table). In other words, if the shifts are of random nature and do not reflect a change in direction of the overall response category but rather a similar number of subjects shifting in either direction, P values will be reported as >.05. Bowker's test reduces to the more familiar McNemar's test for 2 2 tables. Correlations with age, AFC, FSH, LH, and E2 on stimulation day 1, as well as duration of stimulation and total dose of rFSH, were calculated by cycle, and a multiple regression coefficient was obtained from a linear regression model. A similar analysis was performed for changes in ovarian response (between cycles 1 and 2, between cycles 2 and 3, and between cycles 1 and 3, respectively) and changes in predictors of ovarian response.
RESULTS Baseline Demographics and Stimulation Parameters In total, 176 patients were treated and included in the analysis from 25 centers in Australia (25 patients in 4 centers), Europe (71 patients in 12 centers), and South America (80 patients in 9 centers). The median number of patients analyzed in each center was six patients in Australia, five in Europe, and nine in South America.
Total exposure to rFSH ranged from 0 to 2,250 IU across the 176 women completing the three cycles, with only 13 women reporting the same total dose across all three cycles. The number of women who received the same dose of rFSH stimulation in two consecutive cycles includes 44 (25.0%) from cycle 1 to cycle 2, and 37 (21.0%) from cycle 2 to cycle 3. Demographic and stimulation characteristics are given in Table 1. Q3
Ovarian Response as Measured by the Number of Follicles ‡11 mm on the Day of hCG A box plot of the number of follicles in cycles 1, 2, and 3 is shown in Figure 1A. Figure 1B shows a box plot of the change in follicle number from cycle 1 to 2 and from cycle 2 to 3. The mean (SD) number of follicles R11 mm on the day of hCG was 13.4 (6.2), 13.3 (5.4), and 13.8 (6.4) in cycles 1, 2, and 3, respectively (Table 2). Patients receiving the same dose of Q4 rFSH stimulation in cycles 1 and 2 report slightly higher means of 14.4 (5.7) and 14.8 (5.1), respectively, but this trend was not replicated in cycles 2 and 3, in which means of 13.3 (5.6) and 12.8 (5.8) were equal to or less than the number of follicles reported across all subjects in the analysis. Therefore, on the basis of comparing the means and SD between the same-dose rFSH subgroups and the overall analysis
VOL. - NO. - / - 2015
3
FLA 5.2.0 DTD FNS29666_proof 14 July 2015 4:37 pm ce JS
296 297 298 299 300 301 302 303 304 305 306 307 308 309 310 311 312 313 314 315 316 317 318 319 320 321 322 323 324 325 326 327 328 329 330 331 332 333 334 335 336 337 338 339 340 341 342 343 344 345 346 347 348 349 350 351 352 353 354
ORIGINAL ARTICLE: ASSISTED REPRODUCTION 355 356 357 358 359 360 361 362 363 364 365 366 367 368 369 370 371 372 373 374 375 376 377 378 379 380 381 382 383 384 385 386 387 388 389 390 391 392 393 394 395 396 397 398 399 400 401 402 403 404 405 406 407 408 409 410 411 412 413
Intercycle Switching between the Three Ovarian Response Categories (Based on the Number of Follicles ‡11 mm on the Day of hCG)
FIGURE 1
Box plots of (A) the number of and (B) the difference ‘‘delta’’ vs. the previous cycle in the number of follicles (R11 mm) on the day of hCG per cycle. Displayed are maximum (endpoint of upper whisker), third quartile (upper edge of box), median (line inside box), mean (>), first quartile (lower edge of box), and minimum (endpoint of lower whisker). Rombauts. Ovarian response intercycle variability. Fertil Steril 2015.
population, there does not seem to be a significant difference in response attributable to maintaining the same dose of rFSH. Comparing the ovarian response between the first and second cycles in all women in the analysis, the number of follicles was similar, with a reduction of 0.1 (0.7%), not statistically significant (P¼ .798). Comparing the ovarian response between the second and third cycles, the number of follicles was also similar, with an increase of 0.5 (3.9%), not statistically significant (P¼ .126). The number of follicles was also similar between the first and third cycles, with an increase of 0.4 (3.3%), also not statistically significant (P¼ .301). Standard deviations of the changes between cycles ranged from 4.7 to 5.7, indicating a measure of within-subject variability between cycles, with between-cycle correlations ranging from 0.59 to 0.70. Therefore, even though there is a measure of within-subject variability across cycles, mean responses seem to be relatively constant.
Intercycle variability can be described further by assigning the number of follicles to low, normal, and high response categories. If the percentage of women who switch from normal to a low/high category is similar to the percentage of women who switch from a low/high to a normal category, the switch in response categories are more reflective of a symmetric distribution around a mean, which can be attributed to a random distribution of follicle numbers, rather than a directional shift indicating an overall change in response between cycles. The percentages of low (0–<6 follicles), normal (6–<18 follicles), and high (R18 follicles) responders, respectively, were 5.7%, 72.7%, and 21.6% in cycle 1, 5.1%, 73.3%, and 21.6% in cycle 2, and 6.8%, 67.6%, and 25.6% in cycle 3 (Supplemental Table 6, available online). Q5 According to the number of follicles, the majority of patients (per cycle, 67.6%–73.3%) had a normal ovarian response throughout all three cycles. None of the women switched from low to high ovarian response (or vice versa) either between consecutive cycles or between cycles 1 and 3. In the second cycle, 75.6% remained in the same response category as in the first cycle, 11.9% switched from normal to low/high response, and 12.5% switched from low/high to normal ovarian response (Supplemental Table 8). The switches in response categories were not statistically significant in (P¼ .879), indicating the switches support a symmetric distribution of subjects from normal to low/high and low/high to normal response categories. In the third cycle, 73.9% remained in the same response category as in the second cycle, 15.9% switched from normal to low/high response, and 10.2% switched from low/high to normal ovarian response (Supplemental Table 8), again supporting a symmetric distribution of switch in response categories (P¼ .140). Confirming the analysis of the number of follicles, even though there is variability in response, there does not seem to be a measurable pattern of a change in follicles across the cycles.
Intercycle Switching from a Normal to a Low/high Ovarian Response and Vice Versa (according to the Number of Follicles ‡11 mm on the Day of hCG) For women with a normal ovarian response in the first cycle (n ¼ 128), the percentage (95% CI) of women switching to a low/high ovarian response in the second cycle was 16.4% (10.5%–24.0%) and the percentage (95% CI) for those with a low/high ovarian response in the first cycle (n ¼ 48) switching to a normal response in the second cycle was 45.8% (31.4%–60.8%) (Supplemental Table 8). The percentage (95% CI) of women with a normal ovarian response in the second cycle (n ¼ 129) switching to a low/high ovarian response in the third cycle was 21.7% (14.9%–29.8%), and the percentage (95% CI) of those with a low/high ovarian response in the second cycle (n ¼ 47) switching to a normal VOL. - NO. - / - 2015
4
FLA 5.2.0 DTD FNS29666_proof 14 July 2015 4:37 pm ce JS
414 415 416 417 418 419 420 421 422 423 424 425 426 427 428 429 430 431 432 433 434 435 436 437 438 439 440 441 442 443 444 445 446 447 448 449 450 451 452 453 454 455 456 457 458 459 460 461 462 463 464 465 466 467 468 469 470 471 472
Fertility and Sterility® 473 474 475 476 477 478 479 480 481 482 483 484 485 486 487 488 489 490 491 492 493 494 495 496 497 498 499 500 501 502 503 504 505 506 507 508 509 510 511 512 513 514 515 516 517 518 519 520 521 522 523 524 525 526 527 528 529 530 531
TABLE 2 Number of follicles ‡11 mm on the day of hCG. A. Overall and by subgroup of same rFSH stimulation dose between cycles Cycle 1 Patients
n
a
176 Overall Same-stimulation subgroup b 44 Cycles 1 to 2 Cycles 2 to 3c
Cycle 2
Cycle 3
Mean
SD
n
Mean
SD
n
Mean
SD
13.4
6.2
176
13.3
5.4
176
13.8
6.4
14.4
5.7
44 37
14.8 13.3
5.1 5.6
37
12.8
5.8
B. Overall differences between cycles Q8
Cycles 1 to 2 2 to 3 1 to 3
LS mean change (SD)
Percent change
Difference P value
0.1 (5.0) 0.5 (4.7) 0.4 (5.7)
0.7 3.9 3.3
.798 .126 .301
Note: Analysis output extracted from repeated-measures analysis on patients across cycles and controlled for region. LS ¼ XXX. a PR .126 for the differences between cycles (between cycle r12 ¼ 0.64, r13 ¼ 0.59, r23 ¼ 0.70). b P¼ .577 for the difference between cycles 1 and 2 (r ¼ 0.69 between cycles). c P¼ .634 for the difference between cycles 2 and 3 (r¼0.60 between cycles). Rombauts. Ovarian response intercycle variability. Fertil Steril 2015.
response in the third cycle was 38.3% (24.5%–53.6%) (Supplemental Table 8).
Correlation between Follicles ‡11 mm on the Day of hCG Administration and Number of Oocytes Retrieved As expected, there was a correlation between the number follicles R11 mm on the day of hCG and the number of oocytes retrieved: correlation coefficients were 0.60, 0.65, and 0.67 in cycles 1, 2, and 3, respectively (P< .0001; Supplemental Table 7). Data for the percentage of women switching ovarian
TABLE 3 Q9
Variability in ovarian response between cycles according to the number of follicles (‡11 mm) on the day of hCG. Cycle number (compared with cycle in left-most column) Frequency Cycle 1 Low Normal High Total Test of symmetry Cycle 2 Low Normal High Total Test of symmetry Cycle 1 Low Normal High Total Test of symmetry
Low 5 4 0 9 3 9 0 12 5 7 0 12
Normal Cycle 2 5 107 17 129 P¼ .991 Cycle 3 6 101 12 119 P¼ .536 Cycle 3 5 99 15 119 P¼ .647
Rombauts. Ovarian response intercycle variability. Fertil Steril 2015.
response between cycles based on the number of oocytes retrieved are given in the Supplemental Materials and Q6 Supplemental Figures 1 and 2.
Correlation between Predictors of Ovarian Response and Number of Follicles ‡11 mm on the Day of hCG Administration per Cycle The correlation between the number of follicles R11 mm on the day of hCG and predictors of ovarian response (age, AFC, and prestimulation E2, FSH, and LH, as well as duration of stimulation and total dose of rFSH) are given in Supplemental Table 7. Correlations with AFC and FSH were moderate (correlation coefficient magnitudes were approximately 0.3–0.5). Multiple correlation coefficients (combining all seven predictors) were 0.58, 0.66, and 0.60 in cycles 1, 2, and 3, respectively (P< .0001).
Can Baseline FSH and AFC Predict a Change in Ovarian Response per Cycle?
High
Total
0 17 21 38
10 128 38 176
0 19 26 45
9 129 38 176
0 22 23 45
10 128 138 176
The correlation between changes in ovarian response (between cycles 1 and 2 and between cycles 2 and 3) and changes in predictors of ovarian response (Supplemental Table 8) were numerically small. Multiple correlation coefficients (combining all seven predictors as in Supplemental Table 8) were 0.42 and 0.40 between cycles 1 and 2 and between cycles 2 and 3, respectively (P< .001).
DISCUSSION This is the first study, to our knowledge, to examine the intercycle variability of the ovarian response across three consecutive IVF cycles with the same protocol. Our study shows that although the percentage of low, normal, and high ovarian responders to ovarian stimulation overall is similar after up to three stimulation cycles, an individual's ovarian response with an identical treatment protocol can
VOL. - NO. - / - 2015
5
FLA 5.2.0 DTD FNS29666_proof 14 July 2015 4:37 pm ce JS
532 533 534 535 536 537 538 539 540 541 542 543 544 545 546 547 548 549 550 551 552 553 554 555 556 557 558 559 560 561 562 563 564 565 566 567 568 569 570 571 572 573 574 575 576 577 578 579 580 581 582 583 584 585 586 587 588 589 590
ORIGINAL ARTICLE: ASSISTED REPRODUCTION 591 592 593 594 595 596 597 598 599 600 601 602 603 604 605 606 607 608 609 610 611 612 613 614 615 616 617 618 619 620 621 622 623 624 625 626 627 628 629 630 631 632 633 634 635 636 637 638 639 640 641 642 643 644 645 646 647 648 649
TABLE 4 Summary of intercycle variability. From first to second cycle (n [ 176)
Variable Patients with or without a switch in ovarian response Normal to low/high response Low/high to normal response Low/high to low/high Normal to normal response Probability of a switch in ovarian response, % Switch from normal to low/high response (95% CI)a Switch from low/high to normal response (95% CI)a Test of symmetry (McNemar's test)
21 (11.9) 22 (12.5) 26 (14.8) 107 (60.8)
From second to third cycle (n [ 176) 28 (15.9) 18 (10.2) 29 (16.5) 101 (57.4)
21/128 (16.4) (10.5–24.0) 22/48 (45.8) (31.4–60.8) P¼ .879
28/129 (21.7) (14.9–29.8) 18/47 (38.3) (24.5–53.6) P¼ .140
From first to third cycle (n [ 176) 29 (16.5) 20 (11.4) 28 (15.9) 99 (56.3) 29/128 (22.7) (15.7–30.9) 20/48 (41.7) (27.6–56.8) P¼ .199
Note: Data are presented as number (percentage) except where noted otherwise. a 95% CI of percentage calculated using the Clopper-Pearson exact method. Rombauts. Ovarian response intercycle variability. Fertil Steril 2015.
Q7
vary from one cycle to a subsequent cycle. Only approximately 75% of patients remained in the same response category (low, normal, or high) compared with the previous cycle. However, mean differences between subsequent cycles were small (on average, less than half a follicle and less than one oocyte). Although population statistics demonstrated only minimal changes from cycle to cycle, patients also need to be aware that individual responses can vary significantly in a subsequent cycle with the same protocol (Fig. 2), though not in any particular direction toward an increase or decrease in follicles. The data from the Trust trial analyzed in this study also successfully illustrate the regression to the mean effect in patients with an extreme response (i.e., low response or high response) (10). In the Trust study, 45.8% of low/high responders in cycle 1 went on to have a normal response in cycle 2 without changes in the stimulation protocol (Supplemental Table 8). Similarly, 38.3% of low/high responders in cycle 2 went onto have a normal response in cycle 3 (Supplemental Table 8). This is of particular importance when interpreting poorly designed studies that compare ‘‘before’’ and ‘‘after’’ effects of interventions in selected populations with a previous poor ovarian response. Any claimed benefits of interventions in such study designs are likely the result of regression to the mean. The data presented herein are useful for the design of future clinical trials, because this study provides accurate estimates both of the variance in ovarian response in women aged <40 years and of the natural variability in ovarian response between cycles with the same treatment protocol. Accurate sample size calculations depend not only on the predicted effect size but also on the variance of the underlying measure. The number of follicles on the day of hCG administration correlated well with the number of oocytes retrieved (Supplemental Materials and Supplemental Figs. 1 and 2). The reduction in range of follicles R11 mm on the day of hCG administration from cycle 1 to cycle 3 may be due to negative selection of patients for a subsequent cycle when an extreme response occurred in a previous cycle. Accordingly, the true intercycle variation in ovarian response might have been underestimated in this study because subjects with an extreme response in a first cycle
could not contribute to the well-known regression to the mean effect if not put on the same protocol again for a subsequent cycle. Furthermore, intercycle drop-out occurred with pregnancy in the Trust trial. The AFC and FSH levels on stimulation day 1 correlated positively and negatively, respectively, with the ovarian response in each cycle, but there was a weak correlation with LH and E2 levels on stimulation day 1. The AFC and baseline FSH levels have previously been identified as predictive factors for ovarian response in ovarian stimulation protocols in a GnRH antagonist protocol (8). The intercycle variability in ovarian response to controlled ovarian stimulation was, however, not strongly linked to individual patient demographics, infertility characteristics, or baseline predictors. This is of importance because it may be assumed that every cycle is individually different, especially in the size of the growing follicular cohort, which then might explain some of the intercycle variation in response. Changes between cycles in the predictor variable AFC, a proxy for the follicular cohort becoming FSH sensitive at the beginning of each cycle, did indeed show the best correlation with variation observed in the number of growing follicles. However, this correlation is quite small and therefore potentially of limited utility when it comes to predicting and individualizing treatment according to baseline variables. The study has some weaknesses. In the Trust trial, no measurements of serum antim€ ullerian hormone were performed, and we were therefore unable to correlate this predictor to the actual ovarian response in a fixed stimulation protocol. The original study design of the Trust trial also dictated that patients dropped out after a successful cycle. There is, therefore, some selection bias in this retrospective analysis of the Trust trial data. Additionally, no adjustment was made for the different type of gonadotropins used after day 7. However, a number of important factors strengthen the quality of this study. The large sample size allowed us to estimate accurate population statistics for the ovarian response across three consecutive cycles. In addition, for each of the three IVF cycles, all women received the same protocol and, importantly, the stimulation dose was entirely fixed for the first week (150 mg corifollitropin alfa). VOL. - NO. - / - 2015
6
FLA 5.2.0 DTD FNS29666_proof 14 July 2015 4:37 pm ce JS
650 651 652 653 654 655 656 657 658 659 660 661 662 663 664 665 666 667 668 669 670 671 672 673 674 675 676 677 678 679 680 681 682 683 684 685 686 687 688 689 690 691 692 693 694 695 696 697 698 699 700 701 702 703 704 705 706 707 708
Fertility and Sterility® 709 710 711 712 713 714 715 716 717 718 719 720 721 722 723 724 725 726 727 728 729 730 731 732 733 734 735 736 737 738 739 740 741 742 743 744 745 746 747 748 749 750 751 752 753 754 755 756 757 758 759 760 761 762 763 764 765 766 767
In conclusion, the average ovarian response may seem very similar across subsequent cycles, but the underlying individual ovarian response may vary between three subsequent controlled ovarian stimulation cycles. Awareness of an individual subject's intercycle variability is essential, because it may affect the accuracy of ovarian response prediction. Individual subject serum FSH concentrations and AFC at the beginning of each cycle did not strongly predict a switch in the ovarian response. Acknowledgments: The authors thank the following for medical writing and editorial assistance: P. Milner, Ph.D., of PAREXEL, UK, and C. McCrary Sisk and K. Lewis of Merck & Co., Inc., Kenilworth, New Jersey. This assistance from PAREXEL was funded by Merck & Co., Inc.
2.
3.
4.
5.
6.
7.
SUPPLEMENTARY DATA Supplementary data related to this article can be found at http://dx.doi.org/10.1016/j.fertnstert.2015.06.027
8.
9.
REFERENCES 1.
Ahmed Ebbiary NA, Morgan C, Martin K, Afnan M, Newton JR. Ovarian response in consecutive cycles of ovarian stimulation in normally ovulating women. Hum Reprod 1995;10:536–43.
10.
Broekmans FJ, Kwee J, Hendriks DJ, Mol BW, Lambalk CB. A systematic review of tests predicting ovarian reserve and IVF outcome. Hum Reprod Update 2006;12:685–718. Broer SL, Dolleman M, Opmeer BC, Fauser BC, Mol BW, Broekmans FJ. AMH and AFC as predictors of excessive response in controlled ovarian hyperstimulation: a meta-analysis. Hum Reprod Update 2011;17:46–54. Doldi N, Persico P, De Santis L, Rabellotti E, Papaleo E, Ferrari A. Consecutive cycles in in vitro fertilization–embryo transfer. Gynecol Endocrinol 2005;20: 132–6. Hoveyda F, Engmann L, Steele J, Lopez BA, Barlow DH. Ovarian response in three consecutive in vitro fertilization cycles. Fertil Steril 2002;77:706–10. Reichman DE, Chung P, Meyer L, Greenwood E, Davis O, Rosenwaks Z. Consecutive gonadotropin-releasing hormone-antagonist in vitro fertilization cycles: does the elapsed time interval between successive treatments affect outcomes? Fertil Steril 2013;99:1277–82. Norman RJ, Zegers-Hochschild F, Salle BS, Elbers J, Heijnen E, MarintchevaPetrova M, et al. Repeated ovarian stimulation with corifollitropin alfa in patients in a GnRH antagonist protocol: no concern for immunogenicity. Hum Reprod 2011;26:2200–8. Broekmans FJ, Verweij PJ, Eijkemans MJ, Mannaerts BM, Witjes H. Prognostic models for high and low ovarian responses in controlled ovarian stimulation using a GnRH antagonist protocol. Hum Reprod 2014;29:1688–97. Tarlatzis BC, Griesinger G, Leader A, Rombauts L, Ijzerman-Boon PC, Mannaerts BM. Comparative incidence of ovarian hyperstimulation syndrome following ovarian stimulation with corifollitropin alfa or recombinant FSH. Reprod Biomed Online 2012;24:410–9. Rombauts L. Is there a recommended maximum starting dose of FSH in IVF? J Assist Reprod Genet 2007;24:34–9.
VOL. - NO. - / - 2015
7
FLA 5.2.0 DTD FNS29666_proof 14 July 2015 4:37 pm ce JS
768 769 770 771 772 773 774 775 776 777 778 779 780 781 782 783 784 785 786 787 788 789 790 791 792 793 794 795 796 797 798 799 800 801 802 803 804 805 806 807 808 809 810 811 812 813 814 815 816 817 818 819 820 821 822 823 824 825 826
ORIGINAL ARTICLE: ASSISTED REPRODUCTION 886 887 888 889 890 891 892 893 894 895 896 897 898 899 900 901 902 903 904 905 906 907 908 909 910 911 912 913 914 915 916 917 918 919 920 921 922 923 924 925 926 927 928 929 930 931 932 933 934 935 936 937 938 939 940 941 942 943 944
SUPPLEMENTAL FIGURE 1
web 4C=FPO
827 828 829 830 831 832 833 834 835 836 837 838 839 840 841 842 843 844 845 846 847 848 849 850 851 852 853 854 855 856 857 858 859 860 861 862 863 864 865 866 867 868 869 870 871 872 873 874 875 876 877 878 879 880 881 882 883 884 885
Box plot of the number of oocytes per cycle. Displayed are maximum (endpoint of upper whisker), third quartile (upper edge of box), median (line inside box), mean (þ), first quartile (lower edge of box), and minimum (endpoint of lower whisker). Rombauts. Ovarian response intercycle variability. Fertil Steril 2015.
VOL. - NO. - / - 2015
7.e1
FLA 5.2.0 DTD FNS29666_proof 14 July 2015 4:37 pm ce JS
Fertility and Sterility® 1004 1005 1006 1007 1008 1009 1010 1011 1012 1013 1014 1015 1016 1017 1018 1019 1020 1021 1022 1023 1024 1025 1026 1027 1028 1029 1030 1031 1032 1033 1034 1035 1036 1037 1038 1039 1040 1041 1042 1043 1044 1045 1046 1047 1048 1049 1050 1051 1052 1053 1054 1055 1056 1057 1058 1059 1060 1061 1062
SUPPLEMENTAL FIGURE 2
web 4C=FPO
945 946 947 948 949 950 951 952 953 954 955 956 957 958 959 960 961 962 963 964 965 966 967 968 969 970 971 972 973 974 975 976 977 978 979 980 981 982 983 984 985 986 987 988 989 990 991 992 993 994 995 996 997 998 999 1000 1001 1002 1003
Box plot of the difference ‘‘delta’’ of the number of oocytes vs. the previous cycle. Displayed are maximum (endpoint of upper whisker), third quartile (upper edge of box), median (line inside box), mean (þ), first quartile (lower edge of box), and minimum (endpoint of lower whisker). Rombauts. Ovarian response intercycle variability. Fertil Steril 2015.
VOL. - NO. - / - 2015
7.e2
FLA 5.2.0 DTD FNS29666_proof 14 July 2015 4:37 pm ce JS
Intercycle variability of the ovarian response in patients undergoing repeated stimulation with corifollitropin alfa in a gonadotropin-releasing hormone antagonist protocol Q10
Luk Rombauts, Ph.D.,a,b Cornelis B. Lambalk, M.D., Ph.D.,c Askan Schultze-Mosgau, Ph.D.,d Jacqueline van Kuijk, M.Sc.,e Pierre Verweij, Ph.D.,e Davis Gates, Ph.D.,f Keith Gordon, Ph.D.,f and Georg Griesinger, Ph.D.d
a Department of Obstetrics and Gynaecology, Monash University, Clayton, Victoria, Australia; b Monash IVF, Clayton, Victoria, Australia; c Department of Obstetrics/Gynecology, VU University Medical Center, Amsterdam, the Netherlands; d Department of Reproductive Medicine and Gynecologic Endocrinology, University Clinic of Schleswig-Holstein, Campus e f Q1 Luebeck, Luebeck, Germany; MSD Oss N.V., Oss, the Netherlands; and Merck & Co., Inc., Kenilworth, New Jersey
Objective: To determine whether individual subject variation in ovarian response between repeated cycles with the same ovarian stimulation protocol can be predicted. Design: Retrospective data analysis. Setting: Not applicable Patient(s): Women aged 18–39 from a phase 3, open-label, uncontrolled trial with complete data across all cycles (n ¼ 176). Intervention(s): Up to three cycles of a single injection of 150 mg corifollitropin alfa for 7 days, then daily recombinant FSH/hMG until three follicles reached R17 mm. Gonadotropin-releasing hormone antagonist from stimulation day 5 until day of hCG administration. Main Outcome Measure(s): Numbers of follicles R11 mm on day of hCG in cycles 1–3, transition in ovarian response type between cycles from low (0–<6), normal (6–<18), and high (R18), and serum FSH concentrations and antral follicle count (AFC) at each cycle start. Result(s): The mean (SD) numbers of follicles R11 mm on day of hCG were 13.4 (6.2), 13.3 (5.4), and 13.8 (6.4) in cycles 1, 2 and 3, respectively. Between cycles 1 and 2, 11.9% switched from normal to low or high response, and 12.5% switched from low or high to normal response; 75.6% remained in the same category. Between cycles 2 and 3, 15.9% switched from normal to low or high response, and 10.2% switched from low or high to normal response; 73.9% remained in the same category. These shifts are symmetrical in nature, in that the percentage of subjects who shift from normal to low or high response is comparable to the percentage of subjects who shift from low or high to normal response. Baseline FSH and AFC did not significantly predict transition in ovarian response.
Received January 29, 2015; revised and accepted June 22, 2015. L.R. is a minority shareholder in Monash IVF and has provided expert advice to, and received unconditional research and educational grants from, MSD, Merck Serono, and Ferring. C.B.L. has received educational and research grants for his department from MSD, Merck Serono, Ferring, and Auxogyn; speaker fees from MSD, Merck Serono, Ferring, and Auxogyn; and consultant fees from Ferring. J.v.K. and P.V. were employees of MSD Oss N.V. at the time the study was conducted. D.G. and K.G. are employees of Merck & Co., Inc. and may own stock/stock options in the company. G.G. has received consultant/honorarium fees from MSD, Ferring, Glycotope, Serono, Finox, Vitrolife, and IBSA and has served on speaker bureaus for MSD, Ferring, Serono, Vitrolife, and IBSA. Support provided by Merck & Co. Inc., Kenilworth, New Jersey. Reprint requests: Luk Rombauts, Ph.D., Monash Medical Centre, 246 Clayton Road, Clayton, Victoria 3168, Australia (E-mail: [email protected]). Fertility and Sterility® Vol. -, No. -, - 2015 0015-0282/$36.00 Copyright ©2015 American Society for Reproductive Medicine, Published by Elsevier Inc. http://dx.doi.org/10.1016/j.fertnstert.2015.06.027 VOL. - NO. - / - 2015
1
FLA 5.2.0 DTD FNS29666_proof 14 July 2015 4:37 pm ce JS
60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118
ORIGINAL ARTICLE: ASSISTED REPRODUCTION 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177
Conclusion(s): The variability in ovarian responses between repeated cycles using the same protocol was not explained by baseline FSH and AFC. Clinical Trial Registration Number: NCT00696878 Protocol P05714. (Fertil SterilÒ 2015;-: -–-. Ó2015 by American Society for Reproductive Medicine.) Key Words: Corifollitropin alfa, GnRH antagonist, repeated ovarian stimulation, ovarian response, controlled ovarian stimulation Discuss: You can discuss this article with its authors and with other ASRM members at http:// fertstertforum.com/rombautsl-ovarian-response-intercycle-variability/
A
large number of studies have attempted to predict ovarian response to ovarian stimulation for IVF, and by altering the stimulation regimen, to ultimately impact on ovarian response and IVF treatment outcome (1– 6). To put such studies in perspective, it is mandatory to know the naturally occurring variability in ovarian response between subsequent treatment cycles within one subject. For such assessment, repetitive cycles should be performed, ideally within a short time frame, and stimulation (e.g., FSH dose, GnRH analogue protocol, and decisions on patient management) should be kept constant. In the Trust trial, a large population of women received ovarian stimulation uniformly with 150 mg corifollitropin alfa (a long-acting recombinant FSH [rFSH]) in a GnRH antagonist protocol for up to three cycles (7). Serum FSH concentrations and antral follicle count (AFC) on day 1 of ovarian stimulation have been identified as predictors of low or high ovarian response in GnRH antagonist cycles (8). The objectives of the present analysis were to determine the variation between treatment cycles in terms of numbers of growing follicles, to calculate the relative frequency of a switch between cycles from a normal ovarian response to a high or low ovarian response and vice versa, and to investigate whether the changes in ovarian response per cycle can be explained by changes in predictors as, for example, AFC and serum hormones at the start of each cycle of stimulation.
MATERIALS AND METHODS Study Design and Population
Q2
Trust was a multicenter, open-label, uncontrolled clinical trial, the details of which have been reported previously (7). The study was conducted in accordance with principles of good clinical practice and was approved by the appropriate institutional review boards and regulatory agencies. Written, informed consent was provided by all subjects. Women aged 18–39 years with a body weight of >60 kg and a regular menstrual cycle underwent up to three ovarian stimulation cycles. No women were taking birth control pills before beginning stimulation. Patients who conceived in a cycle were not allowed another cycle. Demographic and fertility characteristics of patients in the Trust trial have been reported previously (7, 9). In each cycle, subjects received a single injection of 150 mg corifollitropin alfa (Elonva, N.V. Organon) for the first 7 days of stimulation, followed by a daily dose of %225 IU FSH (either rFSH or hMG) starting on day 8 until the criterion
Use your smartphone to scan this QR code and connect to the discussion forum for this article now.* * Download a free QR code scanner by searching for “QR scanner” in your smartphone’s app store or app marketplace.
for oocyte maturation with hCG was reached (three follicles R17 mm). A GnRH antagonist (0.25 mg ganirelix or cetrorelix acetate) was administered once daily, starting on stimulation day 5 up to and including the day of hCG administration. Either urinary hCG (10,000 IU or 5,000 IU in case of a high ovarian response) or recombinant hCG (250 mg) was administered for final oocyte maturation. In case of too high of an ovarian response, the dose of FSH could be reduced, or withheld for a maximum of 3 days up to and including the day of hCG administration. For normal responders, the recommended daily dose of FSH was 150 IU. If the ovarian response was too high in the opinion of the investigator, the cycle could be cancelled at any time. If there was a risk of ovarian hyperstimulation syndrome (>30 follicles R11 mm on transvaginal ultrasound), hCG was withheld and the treatment cycle was canceled. The maximum total duration of stimulation was 19 days. The mean number of days (95% confidence interval [CI]) between cycles were 144.3 (131.1–157.5) and 171.3 (155.4–187.2), for cycle 1 to cycle 2 and cycle 2 to cycle 3, respectively.
Statistical Analysis Women who received hCG and who had non-missing values across all three cycles for the number of follicles R11 mm on the day of hCG were included in the analysis (n ¼ 176). The numbers of follicles R11 mm on the day of hCG for each cycle were extracted from a longitudinal data analysis on patients repeated across cycles and correcting for region. Given that each patient was measured across three cycles, an unstructured variance–covariance matrix was specified to allow for a separate estimate of variability between and within cycles, accounting for the between-cycle correlation from repeated measurements across patients. Additionally, patients were assigned into subgroups if they received the same dose of rFSH stimulation between two consecutive cycles (cycles 1 and 2; and cycles 2 and 3). No correction for multiplicity was made in this exploratory analysis. The number of follicles R11 mm on the day of hCG and the number of oocytes retrieved were also categorized as low responders (0–<6), normal responders (6–<18), and high responders (8) and summarized (frequencies and percentages) by cycle. Switches from one category to another from cycle 1 to cycle 2 and between cycle 2 and cycle 3 were also summarized. The significance of the directionality of switches was evaluated using Bowker's Test of Symmetry (i.e., test the null hypothesis that nij ¼ nji, where the number of patients are VOL. - NO. - / - 2015
2
FLA 5.2.0 DTD FNS29666_proof 14 July 2015 4:37 pm ce JS
178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236
Fertility and Sterility® 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283 284 285 286 287 288 289 290 291 292 293 294 295
TABLE 1 Demographic and stimulation characteristics: three cycle completers. Cycle 1 (n [ 176)
Baseline Age (y) Body mass index (kg/m2) Race Asian Black Caucasian Other Primary infertility Duration of infertility (y) Cause of infertilitya Male factor Tubal factor Endometriosis Cervical mucus problems Unexplained infertility Other Start cycle Duration of stimulation (d) Total dose of FSH from day 8 onward (IU) AFCb Serum E2 concentration (pmol/L)c Serum FSH concentration (IU/L)c Serum LH concentration (IU/L)c Day of hCG Serum E2 concentration (pmol/L)d Serum FSH concentration (IU/L)d Serum LH concentration (IU/L)d
Cycle 2 (n [ 176)
Cycle 3 (n [ 176)
10.2 (1.3) 455.3 (277.5) 10.8 (4.4) 125.1 (83.0) 7.2 (1.9) 5.0 (1.8)
10.3 (1.7) 478.7 (316.6) 11.0 (4.7) 117.8 (41.1) 7.0 (2.0) 5.0 (1.6)
10.4 (1.8) 504.8 (362.8) 11.1 (4.7) 117.1 (40.8) 7.5 (5.4) 5.0 (1.9)
5,442.4 (3,354.7) 13.7 (3.2) 1.6 (1.8)
5,210.2 (2,822.7) 13.6 (3.7) 1.6 (1.4)
5,743.1 (3,754.9) 13.6 (3.5) 1.8 (2.7)
33.3 (3.5) 24.3 (2.3) 2.3 0.6 96.0 1.1 55.7 3.6 (2.7) 58.0 28.4 13.6 0.6 15.9 1.7
Note: Data are presented as mean (SD) or percentage. a A patient could have multiple causes of infertility. b Ten, eight, and four missing values in cycles 1, 2, and 3, respectively. c Five, ten, and eight missing values. d Two, five, and five missing values. Rombauts. Ovarian response intercycle variability. Fertil Steril 2015.
compared in cell number numbers 12 vs. 21, 13 vs. 31, and 23 vs. 32 in a 3 3 table). In other words, if the shifts are of random nature and do not reflect a change in direction of the overall response category but rather a similar number of subjects shifting in either direction, P values will be reported as >.05. Bowker's test reduces to the more familiar McNemar's test for 2 2 tables. Correlations with age, AFC, FSH, LH, and E2 on stimulation day 1, as well as duration of stimulation and total dose of rFSH, were calculated by cycle, and a multiple regression coefficient was obtained from a linear regression model. A similar analysis was performed for changes in ovarian response (between cycles 1 and 2, between cycles 2 and 3, and between cycles 1 and 3, respectively) and changes in predictors of ovarian response.
RESULTS Baseline Demographics and Stimulation Parameters In total, 176 patients were treated and included in the analysis from 25 centers in Australia (25 patients in 4 centers), Europe (71 patients in 12 centers), and South America (80 patients in 9 centers). The median number of patients analyzed in each center was six patients in Australia, five in Europe, and nine in South America.
Total exposure to rFSH ranged from 0 to 2,250 IU across the 176 women completing the three cycles, with only 13 women reporting the same total dose across all three cycles. The number of women who received the same dose of rFSH stimulation in two consecutive cycles includes 44 (25.0%) from cycle 1 to cycle 2, and 37 (21.0%) from cycle 2 to cycle 3. Demographic and stimulation characteristics are given in Table 1. Q3
Ovarian Response as Measured by the Number of Follicles ‡11 mm on the Day of hCG A box plot of the number of follicles in cycles 1, 2, and 3 is shown in Figure 1A. Figure 1B shows a box plot of the change in follicle number from cycle 1 to 2 and from cycle 2 to 3. The mean (SD) number of follicles R11 mm on the day of hCG was 13.4 (6.2), 13.3 (5.4), and 13.8 (6.4) in cycles 1, 2, and 3, respectively (Table 2). Patients receiving the same dose of Q4 rFSH stimulation in cycles 1 and 2 report slightly higher means of 14.4 (5.7) and 14.8 (5.1), respectively, but this trend was not replicated in cycles 2 and 3, in which means of 13.3 (5.6) and 12.8 (5.8) were equal to or less than the number of follicles reported across all subjects in the analysis. Therefore, on the basis of comparing the means and SD between the same-dose rFSH subgroups and the overall analysis
VOL. - NO. - / - 2015
3
FLA 5.2.0 DTD FNS29666_proof 14 July 2015 4:37 pm ce JS
296 297 298 299 300 301 302 303 304 305 306 307 308 309 310 311 312 313 314 315 316 317 318 319 320 321 322 323 324 325 326 327 328 329 330 331 332 333 334 335 336 337 338 339 340 341 342 343 344 345 346 347 348 349 350 351 352 353 354
ORIGINAL ARTICLE: ASSISTED REPRODUCTION 355 356 357 358 359 360 361 362 363 364 365 366 367 368 369 370 371 372 373 374 375 376 377 378 379 380 381 382 383 384 385 386 387 388 389 390 391 392 393 394 395 396 397 398 399 400 401 402 403 404 405 406 407 408 409 410 411 412 413
Intercycle Switching between the Three Ovarian Response Categories (Based on the Number of Follicles ‡11 mm on the Day of hCG)
FIGURE 1
Box plots of (A) the number of and (B) the difference ‘‘delta’’ vs. the previous cycle in the number of follicles (R11 mm) on the day of hCG per cycle. Displayed are maximum (endpoint of upper whisker), third quartile (upper edge of box), median (line inside box), mean (>), first quartile (lower edge of box), and minimum (endpoint of lower whisker). Rombauts. Ovarian response intercycle variability. Fertil Steril 2015.
population, there does not seem to be a significant difference in response attributable to maintaining the same dose of rFSH. Comparing the ovarian response between the first and second cycles in all women in the analysis, the number of follicles was similar, with a reduction of 0.1 (0.7%), not statistically significant (P¼ .798). Comparing the ovarian response between the second and third cycles, the number of follicles was also similar, with an increase of 0.5 (3.9%), not statistically significant (P¼ .126). The number of follicles was also similar between the first and third cycles, with an increase of 0.4 (3.3%), also not statistically significant (P¼ .301). Standard deviations of the changes between cycles ranged from 4.7 to 5.7, indicating a measure of within-subject variability between cycles, with between-cycle correlations ranging from 0.59 to 0.70. Therefore, even though there is a measure of within-subject variability across cycles, mean responses seem to be relatively constant.
Intercycle variability can be described further by assigning the number of follicles to low, normal, and high response categories. If the percentage of women who switch from normal to a low/high category is similar to the percentage of women who switch from a low/high to a normal category, the switch in response categories are more reflective of a symmetric distribution around a mean, which can be attributed to a random distribution of follicle numbers, rather than a directional shift indicating an overall change in response between cycles. The percentages of low (0–<6 follicles), normal (6–<18 follicles), and high (R18 follicles) responders, respectively, were 5.7%, 72.7%, and 21.6% in cycle 1, 5.1%, 73.3%, and 21.6% in cycle 2, and 6.8%, 67.6%, and 25.6% in cycle 3 (Supplemental Table 6, available online). Q5 According to the number of follicles, the majority of patients (per cycle, 67.6%–73.3%) had a normal ovarian response throughout all three cycles. None of the women switched from low to high ovarian response (or vice versa) either between consecutive cycles or between cycles 1 and 3. In the second cycle, 75.6% remained in the same response category as in the first cycle, 11.9% switched from normal to low/high response, and 12.5% switched from low/high to normal ovarian response (Supplemental Table 8). The switches in response categories were not statistically significant in (P¼ .879), indicating the switches support a symmetric distribution of subjects from normal to low/high and low/high to normal response categories. In the third cycle, 73.9% remained in the same response category as in the second cycle, 15.9% switched from normal to low/high response, and 10.2% switched from low/high to normal ovarian response (Supplemental Table 8), again supporting a symmetric distribution of switch in response categories (P¼ .140). Confirming the analysis of the number of follicles, even though there is variability in response, there does not seem to be a measurable pattern of a change in follicles across the cycles.
Intercycle Switching from a Normal to a Low/high Ovarian Response and Vice Versa (according to the Number of Follicles ‡11 mm on the Day of hCG) For women with a normal ovarian response in the first cycle (n ¼ 128), the percentage (95% CI) of women switching to a low/high ovarian response in the second cycle was 16.4% (10.5%–24.0%) and the percentage (95% CI) for those with a low/high ovarian response in the first cycle (n ¼ 48) switching to a normal response in the second cycle was 45.8% (31.4%–60.8%) (Supplemental Table 8). The percentage (95% CI) of women with a normal ovarian response in the second cycle (n ¼ 129) switching to a low/high ovarian response in the third cycle was 21.7% (14.9%–29.8%), and the percentage (95% CI) of those with a low/high ovarian response in the second cycle (n ¼ 47) switching to a normal VOL. - NO. - / - 2015
4
FLA 5.2.0 DTD FNS29666_proof 14 July 2015 4:37 pm ce JS
414 415 416 417 418 419 420 421 422 423 424 425 426 427 428 429 430 431 432 433 434 435 436 437 438 439 440 441 442 443 444 445 446 447 448 449 450 451 452 453 454 455 456 457 458 459 460 461 462 463 464 465 466 467 468 469 470 471 472
Fertility and Sterility® 473 474 475 476 477 478 479 480 481 482 483 484 485 486 487 488 489 490 491 492 493 494 495 496 497 498 499 500 501 502 503 504 505 506 507 508 509 510 511 512 513 514 515 516 517 518 519 520 521 522 523 524 525 526 527 528 529 530 531
TABLE 2 Number of follicles ‡11 mm on the day of hCG. A. Overall and by subgroup of same rFSH stimulation dose between cycles Cycle 1 Patients
n
a
176 Overall Same-stimulation subgroup b 44 Cycles 1 to 2 Cycles 2 to 3c
Cycle 2
Cycle 3
Mean
SD
n
Mean
SD
n
Mean
SD
13.4
6.2
176
13.3
5.4
176
13.8
6.4
14.4
5.7
44 37
14.8 13.3
5.1 5.6
37
12.8
5.8
B. Overall differences between cycles Q8
Cycles 1 to 2 2 to 3 1 to 3
LS mean change (SD)
Percent change
Difference P value
0.1 (5.0) 0.5 (4.7) 0.4 (5.7)
0.7 3.9 3.3
.798 .126 .301
Note: Analysis output extracted from repeated-measures analysis on patients across cycles and controlled for region. LS ¼ XXX. a PR .126 for the differences between cycles (between cycle r12 ¼ 0.64, r13 ¼ 0.59, r23 ¼ 0.70). b P¼ .577 for the difference between cycles 1 and 2 (r ¼ 0.69 between cycles). c P¼ .634 for the difference between cycles 2 and 3 (r¼0.60 between cycles). Rombauts. Ovarian response intercycle variability. Fertil Steril 2015.
response in the third cycle was 38.3% (24.5%–53.6%) (Supplemental Table 8).
Correlation between Follicles ‡11 mm on the Day of hCG Administration and Number of Oocytes Retrieved As expected, there was a correlation between the number follicles R11 mm on the day of hCG and the number of oocytes retrieved: correlation coefficients were 0.60, 0.65, and 0.67 in cycles 1, 2, and 3, respectively (P< .0001; Supplemental Table 7). Data for the percentage of women switching ovarian
TABLE 3 Q9
Variability in ovarian response between cycles according to the number of follicles (‡11 mm) on the day of hCG. Cycle number (compared with cycle in left-most column) Frequency Cycle 1 Low Normal High Total Test of symmetry Cycle 2 Low Normal High Total Test of symmetry Cycle 1 Low Normal High Total Test of symmetry
Low 5 4 0 9 3 9 0 12 5 7 0 12
Normal Cycle 2 5 107 17 129 P¼ .991 Cycle 3 6 101 12 119 P¼ .536 Cycle 3 5 99 15 119 P¼ .647
Rombauts. Ovarian response intercycle variability. Fertil Steril 2015.
response between cycles based on the number of oocytes retrieved are given in the Supplemental Materials and Q6 Supplemental Figures 1 and 2.
Correlation between Predictors of Ovarian Response and Number of Follicles ‡11 mm on the Day of hCG Administration per Cycle The correlation between the number of follicles R11 mm on the day of hCG and predictors of ovarian response (age, AFC, and prestimulation E2, FSH, and LH, as well as duration of stimulation and total dose of rFSH) are given in Supplemental Table 7. Correlations with AFC and FSH were moderate (correlation coefficient magnitudes were approximately 0.3–0.5). Multiple correlation coefficients (combining all seven predictors) were 0.58, 0.66, and 0.60 in cycles 1, 2, and 3, respectively (P< .0001).
Can Baseline FSH and AFC Predict a Change in Ovarian Response per Cycle?
High
Total
0 17 21 38
10 128 38 176
0 19 26 45
9 129 38 176
0 22 23 45
10 128 138 176
The correlation between changes in ovarian response (between cycles 1 and 2 and between cycles 2 and 3) and changes in predictors of ovarian response (Supplemental Table 8) were numerically small. Multiple correlation coefficients (combining all seven predictors as in Supplemental Table 8) were 0.42 and 0.40 between cycles 1 and 2 and between cycles 2 and 3, respectively (P< .001).
DISCUSSION This is the first study, to our knowledge, to examine the intercycle variability of the ovarian response across three consecutive IVF cycles with the same protocol. Our study shows that although the percentage of low, normal, and high ovarian responders to ovarian stimulation overall is similar after up to three stimulation cycles, an individual's ovarian response with an identical treatment protocol can
VOL. - NO. - / - 2015
5
FLA 5.2.0 DTD FNS29666_proof 14 July 2015 4:37 pm ce JS
532 533 534 535 536 537 538 539 540 541 542 543 544 545 546 547 548 549 550 551 552 553 554 555 556 557 558 559 560 561 562 563 564 565 566 567 568 569 570 571 572 573 574 575 576 577 578 579 580 581 582 583 584 585 586 587 588 589 590
ORIGINAL ARTICLE: ASSISTED REPRODUCTION 591 592 593 594 595 596 597 598 599 600 601 602 603 604 605 606 607 608 609 610 611 612 613 614 615 616 617 618 619 620 621 622 623 624 625 626 627 628 629 630 631 632 633 634 635 636 637 638 639 640 641 642 643 644 645 646 647 648 649
TABLE 4 Summary of intercycle variability. From first to second cycle (n [ 176)
Variable Patients with or without a switch in ovarian response Normal to low/high response Low/high to normal response Low/high to low/high Normal to normal response Probability of a switch in ovarian response, % Switch from normal to low/high response (95% CI)a Switch from low/high to normal response (95% CI)a Test of symmetry (McNemar's test)
21 (11.9) 22 (12.5) 26 (14.8) 107 (60.8)
From second to third cycle (n [ 176) 28 (15.9) 18 (10.2) 29 (16.5) 101 (57.4)
21/128 (16.4) (10.5–24.0) 22/48 (45.8) (31.4–60.8) P¼ .879
28/129 (21.7) (14.9–29.8) 18/47 (38.3) (24.5–53.6) P¼ .140
From first to third cycle (n [ 176) 29 (16.5) 20 (11.4) 28 (15.9) 99 (56.3) 29/128 (22.7) (15.7–30.9) 20/48 (41.7) (27.6–56.8) P¼ .199
Note: Data are presented as number (percentage) except where noted otherwise. a 95% CI of percentage calculated using the Clopper-Pearson exact method. Rombauts. Ovarian response intercycle variability. Fertil Steril 2015.
Q7
vary from one cycle to a subsequent cycle. Only approximately 75% of patients remained in the same response category (low, normal, or high) compared with the previous cycle. However, mean differences between subsequent cycles were small (on average, less than half a follicle and less than one oocyte). Although population statistics demonstrated only minimal changes from cycle to cycle, patients also need to be aware that individual responses can vary significantly in a subsequent cycle with the same protocol (Fig. 2), though not in any particular direction toward an increase or decrease in follicles. The data from the Trust trial analyzed in this study also successfully illustrate the regression to the mean effect in patients with an extreme response (i.e., low response or high response) (10). In the Trust study, 45.8% of low/high responders in cycle 1 went on to have a normal response in cycle 2 without changes in the stimulation protocol (Supplemental Table 8). Similarly, 38.3% of low/high responders in cycle 2 went onto have a normal response in cycle 3 (Supplemental Table 8). This is of particular importance when interpreting poorly designed studies that compare ‘‘before’’ and ‘‘after’’ effects of interventions in selected populations with a previous poor ovarian response. Any claimed benefits of interventions in such study designs are likely the result of regression to the mean. The data presented herein are useful for the design of future clinical trials, because this study provides accurate estimates both of the variance in ovarian response in women aged <40 years and of the natural variability in ovarian response between cycles with the same treatment protocol. Accurate sample size calculations depend not only on the predicted effect size but also on the variance of the underlying measure. The number of follicles on the day of hCG administration correlated well with the number of oocytes retrieved (Supplemental Materials and Supplemental Figs. 1 and 2). The reduction in range of follicles R11 mm on the day of hCG administration from cycle 1 to cycle 3 may be due to negative selection of patients for a subsequent cycle when an extreme response occurred in a previous cycle. Accordingly, the true intercycle variation in ovarian response might have been underestimated in this study because subjects with an extreme response in a first cycle
could not contribute to the well-known regression to the mean effect if not put on the same protocol again for a subsequent cycle. Furthermore, intercycle drop-out occurred with pregnancy in the Trust trial. The AFC and FSH levels on stimulation day 1 correlated positively and negatively, respectively, with the ovarian response in each cycle, but there was a weak correlation with LH and E2 levels on stimulation day 1. The AFC and baseline FSH levels have previously been identified as predictive factors for ovarian response in ovarian stimulation protocols in a GnRH antagonist protocol (8). The intercycle variability in ovarian response to controlled ovarian stimulation was, however, not strongly linked to individual patient demographics, infertility characteristics, or baseline predictors. This is of importance because it may be assumed that every cycle is individually different, especially in the size of the growing follicular cohort, which then might explain some of the intercycle variation in response. Changes between cycles in the predictor variable AFC, a proxy for the follicular cohort becoming FSH sensitive at the beginning of each cycle, did indeed show the best correlation with variation observed in the number of growing follicles. However, this correlation is quite small and therefore potentially of limited utility when it comes to predicting and individualizing treatment according to baseline variables. The study has some weaknesses. In the Trust trial, no measurements of serum antim€ ullerian hormone were performed, and we were therefore unable to correlate this predictor to the actual ovarian response in a fixed stimulation protocol. The original study design of the Trust trial also dictated that patients dropped out after a successful cycle. There is, therefore, some selection bias in this retrospective analysis of the Trust trial data. Additionally, no adjustment was made for the different type of gonadotropins used after day 7. However, a number of important factors strengthen the quality of this study. The large sample size allowed us to estimate accurate population statistics for the ovarian response across three consecutive cycles. In addition, for each of the three IVF cycles, all women received the same protocol and, importantly, the stimulation dose was entirely fixed for the first week (150 mg corifollitropin alfa). VOL. - NO. - / - 2015
6
FLA 5.2.0 DTD FNS29666_proof 14 July 2015 4:37 pm ce JS
650 651 652 653 654 655 656 657 658 659 660 661 662 663 664 665 666 667 668 669 670 671 672 673 674 675 676 677 678 679 680 681 682 683 684 685 686 687 688 689 690 691 692 693 694 695 696 697 698 699 700 701 702 703 704 705 706 707 708
Fertility and Sterility® 709 710 711 712 713 714 715 716 717 718 719 720 721 722 723 724 725 726 727 728 729 730 731 732 733 734 735 736 737 738 739 740 741 742 743 744 745 746 747 748 749 750 751 752 753 754 755 756 757 758 759 760 761 762 763 764 765 766 767
In conclusion, the average ovarian response may seem very similar across subsequent cycles, but the underlying individual ovarian response may vary between three subsequent controlled ovarian stimulation cycles. Awareness of an individual subject's intercycle variability is essential, because it may affect the accuracy of ovarian response prediction. Individual subject serum FSH concentrations and AFC at the beginning of each cycle did not strongly predict a switch in the ovarian response. Acknowledgments: The authors thank the following for medical writing and editorial assistance: P. Milner, Ph.D., of PAREXEL, UK, and C. McCrary Sisk and K. Lewis of Merck & Co., Inc., Kenilworth, New Jersey. This assistance from PAREXEL was funded by Merck & Co., Inc.
2.
3.
4.
5.
6.
7.
SUPPLEMENTARY DATA Supplementary data related to this article can be found at http://dx.doi.org/10.1016/j.fertnstert.2015.06.027
8.
9.
REFERENCES 1.
Ahmed Ebbiary NA, Morgan C, Martin K, Afnan M, Newton JR. Ovarian response in consecutive cycles of ovarian stimulation in normally ovulating women. Hum Reprod 1995;10:536–43.
10.
Broekmans FJ, Kwee J, Hendriks DJ, Mol BW, Lambalk CB. A systematic review of tests predicting ovarian reserve and IVF outcome. Hum Reprod Update 2006;12:685–718. Broer SL, Dolleman M, Opmeer BC, Fauser BC, Mol BW, Broekmans FJ. AMH and AFC as predictors of excessive response in controlled ovarian hyperstimulation: a meta-analysis. Hum Reprod Update 2011;17:46–54. Doldi N, Persico P, De Santis L, Rabellotti E, Papaleo E, Ferrari A. Consecutive cycles in in vitro fertilization–embryo transfer. Gynecol Endocrinol 2005;20: 132–6. Hoveyda F, Engmann L, Steele J, Lopez BA, Barlow DH. Ovarian response in three consecutive in vitro fertilization cycles. Fertil Steril 2002;77:706–10. Reichman DE, Chung P, Meyer L, Greenwood E, Davis O, Rosenwaks Z. Consecutive gonadotropin-releasing hormone-antagonist in vitro fertilization cycles: does the elapsed time interval between successive treatments affect outcomes? Fertil Steril 2013;99:1277–82. Norman RJ, Zegers-Hochschild F, Salle BS, Elbers J, Heijnen E, MarintchevaPetrova M, et al. Repeated ovarian stimulation with corifollitropin alfa in patients in a GnRH antagonist protocol: no concern for immunogenicity. Hum Reprod 2011;26:2200–8. Broekmans FJ, Verweij PJ, Eijkemans MJ, Mannaerts BM, Witjes H. Prognostic models for high and low ovarian responses in controlled ovarian stimulation using a GnRH antagonist protocol. Hum Reprod 2014;29:1688–97. Tarlatzis BC, Griesinger G, Leader A, Rombauts L, Ijzerman-Boon PC, Mannaerts BM. Comparative incidence of ovarian hyperstimulation syndrome following ovarian stimulation with corifollitropin alfa or recombinant FSH. Reprod Biomed Online 2012;24:410–9. Rombauts L. Is there a recommended maximum starting dose of FSH in IVF? J Assist Reprod Genet 2007;24:34–9.
VOL. - NO. - / - 2015
7
FLA 5.2.0 DTD FNS29666_proof 14 July 2015 4:37 pm ce JS
768 769 770 771 772 773 774 775 776 777 778 779 780 781 782 783 784 785 786 787 788 789 790 791 792 793 794 795 796 797 798 799 800 801 802 803 804 805 806 807 808 809 810 811 812 813 814 815 816 817 818 819 820 821 822 823 824 825 826
ORIGINAL ARTICLE: ASSISTED REPRODUCTION 886 887 888 889 890 891 892 893 894 895 896 897 898 899 900 901 902 903 904 905 906 907 908 909 910 911 912 913 914 915 916 917 918 919 920 921 922 923 924 925 926 927 928 929 930 931 932 933 934 935 936 937 938 939 940 941 942 943 944
SUPPLEMENTAL FIGURE 1
web 4C=FPO
827 828 829 830 831 832 833 834 835 836 837 838 839 840 841 842 843 844 845 846 847 848 849 850 851 852 853 854 855 856 857 858 859 860 861 862 863 864 865 866 867 868 869 870 871 872 873 874 875 876 877 878 879 880 881 882 883 884 885
Box plot of the number of oocytes per cycle. Displayed are maximum (endpoint of upper whisker), third quartile (upper edge of box), median (line inside box), mean (þ), first quartile (lower edge of box), and minimum (endpoint of lower whisker). Rombauts. Ovarian response intercycle variability. Fertil Steril 2015.
VOL. - NO. - / - 2015
7.e1
FLA 5.2.0 DTD FNS29666_proof 14 July 2015 4:37 pm ce JS
Fertility and Sterility® 1004 1005 1006 1007 1008 1009 1010 1011 1012 1013 1014 1015 1016 1017 1018 1019 1020 1021 1022 1023 1024 1025 1026 1027 1028 1029 1030 1031 1032 1033 1034 1035 1036 1037 1038 1039 1040 1041 1042 1043 1044 1045 1046 1047 1048 1049 1050 1051 1052 1053 1054 1055 1056 1057 1058 1059 1060 1061 1062
SUPPLEMENTAL FIGURE 2
web 4C=FPO
945 946 947 948 949 950 951 952 953 954 955 956 957 958 959 960 961 962 963 964 965 966 967 968 969 970 971 972 973 974 975 976 977 978 979 980 981 982 983 984 985 986 987 988 989 990 991 992 993 994 995 996 997 998 999 1000 1001 1002 1003
Box plot of the difference ‘‘delta’’ of the number of oocytes vs. the previous cycle. Displayed are maximum (endpoint of upper whisker), third quartile (upper edge of box), median (line inside box), mean (þ), first quartile (lower edge of box), and minimum (endpoint of lower whisker). Rombauts. Ovarian response intercycle variability. Fertil Steril 2015.
VOL. - NO. - / - 2015
7.e2
FLA 5.2.0 DTD FNS29666_proof 14 July 2015 4:37 pm ce JS