l

FERTILITY AND STERILITY Vol. 62, No. 6, December 1994

Copyright e 1994 The American Fertility Society Printed on acid-free paper in U. S. A.

Effect of follicular size on oocyte retrieval, fertilization, cleavage,

and embryo quality in in vitro fertilization cycles:
a 6-year data collection

Frank M. Wittmaack, M.D. Richard W. Tureck, M.D.*

Donald 0. Kreger, M.D. Luigi Mastroianni, Jr., M.D.
Luis Blasco, M.D. Bruce A. Lessey, M.D., Ph.D.t

Division of Human Reproduction, Department of Obstetrics and Gynecology, University of Pennsylvania, Philadelphia, Pennsylvania

Objective: To determine the effect of follicular size, including the size of the leading follicle, on
oocyte retrieval, fertilization, cleavage, and embryo quality in IVF cycles based on a large data
collection.
Design: Retrospective analysis of 1,109 IVF cycles between 1987 and 1993 at the Hospital of the
University of Pennsylvania including 606 patients ranging in age from 23 to 49 years.
Results: Follicles with a volume ~ 1 mL show a significantly lower oocyte recovery rate than
follicles with a volume of> 1 mL. The highest recovery rate (83.5%) was found in follicles with a
volume of 3 to 4 mL. Above a follicular volume of 7 mL, the oocyte recovery drops below that
observed for follicles between 1 and 7 mL. Fertilization and cleavage rates were also higher in
oocytes obtained from follicles > 1 mL compared with follicles ~1 mL. Although fertilization rates
were fairly stable above volumes of 1 mL, cleavage rates continued to rise to a peak percentage of
92% with volumes between 6 and 7 mL. Leading follicle size did not have an effect on fertilization
and cleavage rates of cohort oocytes. Embryo quality was not influenced significantly by follicular
volume.
Conclusion: Based on this evaluation of a large number of follicles, follicular size is a useful
indicator of oocyte recovery, fertilization, and cleavage in IVF cycles. For optimal results, the
follicular fluid volume in gonadotropin- and hCG-stimulated cycles should be> 1 mL, which corre-
sponds to a follicle diameter of >12 mm, and not larger than 7 mL (24 mm). For timing of hCG
administration, the number of adequate size follicles appears to be more important than the size of
the leading follicle(s). Fertil Steril1994;62:1205-1210

Key Words: Ovulation induction, follicle size, oocyte retrieval, in vitro fertilization

Despite over 10 years of global experience with
assisted reproductive technologies (ARTs), success
rates of IVF-ET remain fairly constant at 15% to
20% clinical pregnancies per cycle (1). The success
of IVF-ET has been associated with multiple vari-
Received February 28, 1994; revised and accepted June 24,
1994.
* Reprint requests: Richard W. Tureck, M.D., Department of
Obstetrics and Gynecology, Hospital of the University of Penn-
sylvania, 3400 Spruce Street, Philadelphia, Pennsylvania 19104
(FAX: 215-573-5408).
t Present address: Department of Obstetrics and Gynecology,
University of North Carolina, Chapel Hill, North Carolina.
abies including maternal age (2), underlying infertil-
ity factors, serum and follicular fluid (FF) E 2 levels
(3), morphologic characterization of the oocyte-
corona-cumulus complex (4), basal FSH (5) and
human GH (hGH) levels (6), and the ultrasonic
(US) appearance ofthe endometrial stripe (7). The
role of follicular size in the success of IVF-ET re-
mains controversial. Simonetti et al. (8) demon-
strated that the incidence of mature oocytes in-
creased with follicular size. However, larger follicles
may result in a decrement in recovery, fertilization,
and cleavage rates as shown by our group (9). Add-
ing to this controversy, small follicles have been
associated with higher pregnancy rates (PRs) in

Vol. 62, No.6, December 1994 Wittmaack et al. Follicular size in IVF cycles 1205
 1
I
some studies (10). These conflicting and sometimes Table 1 Characteristics of Follicle Categories
confusing findings could in part be due to the small
Follicular No. of Age of Male
population of follicles examined in most studies. volume follicles patients* factort
Timing of hCG administration to induce ovula-
ml %
tion is often based on the size of the leading folli-
0.1 to 1.0 1428 35.0 ± 4.2 16.2 ± 3.7
cle(s) (16 to 18 mm) (11-13). Although the leading 1.1 to 2.0 1834 34.6 ± 4.1 14.0 ± 3.5
follicles may yield mature oocytes with good fertil- 2.1 to 3.0 1557 34.9 ± 4.1 16.2 ± 3.7
ization, cleavage, and embryonic development 3.1 to 4.0 988 35.0 ± 4.3 16.6 ± 3.7
4.1 to 5.0 576 35.4 ± 4.0 16.8 ± 3.8
rates, the outcome is less clear for the smaller folli- 5.1 to 6.0 427 35.8 ± 3.9 15.9 ± 3.7
des in the cohort. Administration of hCG based on 6.1 to 7.0 23 34.8 ± 4.1 8.7 ± 6.0
the size of the leading follicle(s) may result in a 7.1 to 15.0 46 34.5 ± 4.3 19.6 ± 5.9
substantial number of very small follicles with un-
* Values for age of patients represent means ± SD.
known outcome. t Values for male factor represent the means ± SD of follicles
To investigate the effect of follicular size (includ- from couples with subnormal sperm parameters (sperm concen-
ing very small and very large follicles) during con- tration <20 X 106 /mL, motility <50%, or <50% normal morphol-
ogy).
trolled ovarian stimulation for IVF-ET, we exam-
ined the outcome of aspirations of 6,879 follicles
and report herein the effect of follicular size on oo-
cyte recovery, fertilization, cleavage, and embryo cleavage. Processing of oocytes was performed as
quality. previously described (9).
The quality of embryos transferred was assessed
by the number of blastomeres per embryo and em-
bryo morphology according to the following crite-
MATERIALS AND METHODS
ria: grade 1, even-sized blastomeres with no frag-
Data were analyzed from patients undergoing mentation; grade 2, uneven-sized blastomeres with
IVF-ET in the ART program at the University of no or some fragmentation; and grade 3, significant
Pennsylvania between 1987 and 1993. Over 1,100 fragmentation of blastomeres or no division for at
ovarian stimulation cycles from 606 patients were least 24 hours. Intermediate grades of 1.5 and 2.5
studied. The patients ranged in age from 23 to 49 were assigned to embryos with borderline charac-
years. Patients received hMG (Pergonal; Serono, teristics.
Randolph, MA) alone (63%), FSH (Metrodin; Ser- In all cases studied, follicular diameter was mea-
ono) alone (6%), hMG and FSH combined (31%), sured using a General Electric RT 3600 scanner
with 99.4% of all patients receiving a GnRH agonist (General Electric Co., Rancho Cordera, CA)
(GnRH-a, [leuprolide acetate] Lupron; TAP Phar- equipped with a 5-MHz vaginal probe transducer.
maceuticals, Deerfield, IL). Luteal phase GnRH-a To assess the correlation between follicular size
(1 mg/d SC) was begun on cycle day 21 (51%) and and follicular volume, follicular diameter was mea-
flare-up GnRH-a was started on cycle day 3 (49%). sured in two dimensions immediately before ovum
The criteria for hCG (10,000 units, Profasi; Serono) retrieval in a group of 50 follicles. The diameter of
administration have remained constant over the pe- the follicle was then compared with the follicular
riod of the study (leading follicle(s) between 16 and volume aspirated using the formula for the volume
18 mm in diameter). Ultrasound scans were per- of a sphere: v = 4/31rr3 •
formed by either experienced clinical faculty or fel- In the overall analysis of 6,879 follicles, volume
lows under their direct supervision, thereby mini- was divided into eight groups by 1-mL increments.
mizing interexaminer variation. There was no Statistical comparisons among these groups was
statistically significant difference among FF vol- performed using x 2 analysis.
ume groups in mean patient age or in the percent-
age of couples with a male factor (Table 1). As a
RESULTS
general rule, fluid volume at the time of retrieval
was recorded to the nearest 0.5 mL. Only follicles Table 1 shows the number of follicles in each
yielding clear or minimally blood-tinged fluid were group based on the actual volume aspirated. The
included in the analysis. Oocytes used for GIFT range offollicular volume was 0.1 to 15 mL (6 to 31
were excluded. A total of 6,879 follicles aspirated mm). Twenty-seven percent of follicles measured 1
were analyzed for oocyte recovery, fertilization, and to 2 mL (12 to 16 mm), which represented the larg-

1206 Wittmaack et al. Follicular size in IVF cycles Fertility and Sterility
 095~--------------------------+---TI
~W"~-------------------------iD---ri
•
:[
4
lso~~--------~~~~-gt-~t-~1--iiH
,~ • •
"3
..
IC
~

"3
3 • •
~
,;e
u
2 •
0.1-1 1.1-2 2.1-3 3.1-4 4.1-5 5.1-6 6.1-7 7.1-15
~ Follicular fluid volume (ml)
'a
l. Figure 3 Fertilization rates (•) represent the percentage of
fertilized oocytes based on the total number of oocytes retrieved.
2 4 Cleavage rates (D) represent the percentage of cleaved oocytes
Actual follicular fluid vol. (ml) based on the total number of fertilized oocytes. Data are shown
for each FF volume category. Error bars indicate SE ± 5%. **P
Figure 1 Correlation of actual and predicted FF volumes. Ac- < 0.0001 versus FF volume groups 1.1 to 15 mL .
tual volumes were measured at the time of follicle aspiration.
Predicted volumes of the same follicles were calculated from the
US measurement of the diameters using the formula, v = 4/31rr3 •
recovery for each follicular volume group. As
est group. As shown in Table 1, larger follicles were shown, follicles with a volume ::5: 1 mL showed a
obtained in decreasing numbers. The percentage of significantly lower recovery rate than follicles with
very small follicles (::5:1 mL) was much higher than volumes> 1 mL (P < 0.0001). The highest recovery
that of very large follicles (>6 mL), 21% and 1%, rate (83.5%) was observed in follicles with a volume
respectively. of 3 to 4 mL, corresponding to a follicular diameter
Figure 1 shows the relationship between the ac- of 18 to 20 mm. With higher volumes (>7 mL) a
tual and predicted volumes in the group of 50 folli- decrease in recovery rate (P < 0.01) was again ob-
cles measured just before ovum retrieval. A high served. The number of atretic or fractured oocytes
degree of correlation (r = 0. 79) was noted between was similar in each group and ranged from 4% to
the calculated and actual volume observed. This 7%.
suggests that follicular size could be estimated us- All normal-appearing oocytes were inseminated.
ing follicular volume as a surrogate measurement of Figure 3 shows the fertilization and cleavage rates
size. of oocytes based on volume at the time of follicle
Information regarding oocyte recovery, fertiliza- aspiration. Fertilization and cleavage rates were sig-
tion, and cleavage rates was available for each folli- nificantly higher in oocytes obtained from folli-
cle. Figure 2 represents the percentage of oocyte cles > 1 mL compared with follicles ::5: 1 mL (P
< 0.0001). Fertilization rates were relatively con-
90
stant when follicular volume exceeded 2 mL. Cleav-
age rates in follicles > 1 mL continued to increase
85
in successive groups, reaching a maximum of 92%
~80 in the 6 to 7 mL range, corresponding to a diameter
f75 of 23 to 24 mm.
8 70 To evaluate the effect of the leading follicle (i.e.,
e
~65 the largest follicle in a particular patient) on its
~60 cohort, the fertilization and cleavage rates of oo-
55 cytes from the same cohort were determined and
50
correlated with the fluid volume of the leading folli-
0.1-1 1.1·2 2.1-3 3.1-4 4.1-5 5.1-6 6.1-7 7.1-15 cle. Volumes ::5: 2 mL for leading follicles were not
(5.8-12.4) (12~-15.6) (15.7-17.9) (18.0-19.7) (19.8-21.2) (21.3-225) (22.6-23.7) (23.8-30.6)
observed in sufficient quantities to allow statistical
Follicular fluid volume (ml)
analysis. From a range of 2 to 15 mL, lead FF vol-
Figure 2 Percentage of follicles yielding an oocyte at the time ume did not influence significantly fertilization and
of retrieval. Data are shown for each FF volume category. Error cleavage rates of cohort oocytes (Fig. 4).
bars indicate SE ± 5%. *P < 0.01; **P < 0.0001 (versus FF
volume groups 1.1 to 7 mL). Numbers in parentheses represent After cleavage, embryonic development was
corresponding follicular diameter in mm. monitored, and embryos were graded on the basis of

Vol. 62, No.6, December 1994 Wittmaack et al. Follicular size in IVF cycles 1207
 Ben-Rafael et al. (9), who noted a decreased recov-
ery rate when the leading follicle grew beyond 23
mm in hMG and hCG cycles. This contrasts with
clomiphene citrate (CC)-stimulated cycles, in
which increased recovery rates have been reported
in large follicles> 20 mm (17). In addition to higher
recovery rates, our study shows an increased fertil-
ization and cleavage rate of oocytes obtained from
follicles with volumes of> 1 mL. Lopata et al. (18),
Lead follicular fluid volume (ml)
looking at CC- and combination CC- and hMG-
Figure 4 Fertilization rates (•) and cleavage rates (D) are stimulated cycles, found no clear correlation be-
shown for the complete cohort of oocytes based on the fluid tween follicle diameter and the percentage of folli-
volume of the leading follicle at the time of retrieval. Error bars
indicate SE ± 5%. cles yielding embryos for transfer. However, this
study is limited by small numbers. In spontaneous
ovulation cycles Trounson (19) demonstrated are-
lationship between the FF volume and the percent-
morphological appearance at the time of transfer. age of normal embryo development based on all
In addition, the number ofblastomeres of each em- mature oocytes inseminated. The highest percent-
bryo was determined. The average embryo quality age of normal embryos was associated with FF vol-
based on these two parameters appeared to be inde- umes of 5 to 15 mL, suggesting that in spontaneous
pendent of follicular volume (Table 2). The embryo cycles biologically equivalent follicles are larger
quality distribution for each follicular volume than in hMG-stimulated cycles.
group was fairly symmetrical and not bipolar. This Traditionally, the timing of hCG administration
excludes the possibility that similar average values in hMG-stimulated cycles has been based on the
for embryo quality were obtained in the presence of size of the leading follicle(s), ranging from 16 to 18
a very heterogeneous data distribution. mm (11 to 13 mL) (20, 21). It has been suggested
that delaying hCG administration decreases fertil-
ization (20) and PRs (21) in IVF cycles. In contrast,
DISCUSSION
our data show very little effect of lead follicle size on
Ovulation represents the final stage of follicular fertilization and cleavage rates of cohort oocytes.
development and maturation. Morphologically, it is This suggests that the number of adequate size fol-
heralded by rapid distention of the antrum by FF. licles in a cohort is a more valuable parameter than
Ultrasound has been used as a means to monitor lead follicle size. This is especially important in
follicular maturation. Nevertheless, in spontaneous view of the unexpected high frequency at which
cycles, great variations in follicular growth have very small follicles :::; 1 mL (:::;12 mm) are present
been reported (14). Although it is believed generally for aspiration in IVF cycles (21% ). In contrast, very
that larger follicles may yield more mature eggs, it large follicles> 6 mL (>23 mm) represent only 1%
has been shown that very small follicles may yield of all follicles.
oocytes capable of fertilization, cleavage, and preg-
nancy (10).
Table 2 Embryo Quality*
Previous studies have shown a correlation be-
tween actual FF volume at laparoscopy and FF vol- Follicular volume Embryo grade No. of blastomeres
ume estimates based on US follicle diameter mea-
mL
surements before surgery (15, 16). This study
confirms these findings. In fact, the correlation ob- 0.1 to 1.0 2.0 ± 0.6 3.3 ± 1.4
1.1 to 2.0 1.9 ± 0.6 3.5 ± 1.4
tained was higher than that in previous studies, pos- 2.1 to 3.0 1.8 ± 0.6 3.7 ± 1.4
sibly because of the absence of a time delay between 3.1 to 4.0 1.8 ± 0.6 3.5 ± 1.4
US measurements and fluid aspirations. 4.1 to 5.0 1.8 ± 0.6 3.7 ± 1.6
5.1 to 6.0 1.8 ± 0.6 3.7 ± 1.8
We found higher oocyte recovery rates in follicles 6.1 to 7.0 1.8 ± 0.5 3.3 ± 1.1
with volumes of> 1 mL (>12 mm diameter) com- 7.1 to 15.0 1.9 ± 0.5 3.4 ± 1.6
pared with those of :::;1 mL. Above a volume of 7 mL
*Values are means± SD. Embryos were graded and blasto-
(24 mm), however, recovery rates decreased meres were counted at the time of transfer. See Materials and
sharply. This is compatible with the observation by Methods for details.

1208 Wittmaack et al. Follicular size in IVF cycles Fertility and Sterility
 Embryo quality as assessed by embryo grade and
number of blastomeres has been correlated with
PRs by various authors (22, 23). These studies show
a positive effect of good quality embryos on PRs. It
is of interest that we did not find an association
between follicular volume and embryo quality, sug-
gesting that in IVF cycles oocytes from small folli-
cles have the same potential as those from larger
follicles, provided they progress beyond the stages
of fertilization and initial cleavage.
It is believed generally that ART results are de-
pendent on patient age. When data were reanalyzed
using a subgroup of patients under 35 years of age,
results were similar to those found for the whole
study population.
We conclude that follicular size is a useful indica-
tor of subsequent oocyte recovery, fertilization, and
cleavage in IVF cycles. For optimal results, the FF
volume in hMG- and hCG-stimulated cycles should
be > 1 mL, which corresponds to a follicle diameter
of 12 mm, and optimally not larger than 7 mL (24
mm). This broad range of acceptable follicular sizes
will allow greater flexibility in the decision-making
process for the timing of hCG administration in
IVF-ET cycles. Rather than using the concept of
inducing ovulation based on the size of the leading
follicle, we suggest to optimize cycle outcome by
bringing as many follicles as possible into the ac-
ceptable range before hCG administration, regard-
less of lead follicle size. Indeed, one might sacrifice
the leading follicle in an attempt to allow further
growth of very small follicles. Because the number
of follicles > 6 mL (>23 mm) was very low (1 %),
one would not expect a significant number of folli-
cles to be pushed into the >7-mL group, which was
associated with poor oocyte recovery. With the
widespread use of GnRH agonists in stimulation
protocols, spontaneous ovulation can be avoided
under these circumstances. Whether or not these
results can be extrapolated to non-IVF cycles will
require further study.
Acknowledgments. We thank the IVF laboratory staff at the
Hospital of the University of Pennsylvania and Jill Howarth,
B.S., for their help collecting and compiling data for this study.
REFERENCES
1. Medical Research International Society for Assisted Repro-
ductive Technology (SART), The American Fertility Soci-
ety. In vitro fertilization-embryo transfer (IVF-ET) in the
Vol. 62, No.6, December 1994
United States: 1990 results from the IVF-ET Registry. Fer-
til Steril 1992;57:15-24.
2. Navot D, Bergh PA, Williams MA, Garrisi GJ, Guzman I,
Sandler B, et al. Poor oocyte quality rather than implanta-
tion failure as a cause of age-related decline in female fertil-
ity. Lancet 1991;337:1375-7.
3. Botero-Ruiz W, Laufer N, DeCherney AH, Polan ML, Ha-
seltine FP, Behrman HR. The relationship between follicu-
lar fluid steroid concentration and successful fertilization of
human oocytes in vitro. Fertil Steril 1984;41:820-6.
4. Laufer N, Tarlatzis BC, DeCherney AH, Masters JT, Ha-
seltine FP, MacClusky N, et al. Asynchrony between hu-
man cumulus-corona-cell complex and oocyte maturation
after human menopausal gonadotrophin treatment for in
vitro fertilization. Fertil Steril 1984;42:366-72.
5. Toner JP, Philput CB, Jones GS, Muasher SJ. Basal folli-
cle-stimulating hormone level is a better predictor of in vi-
tro fertilization performance than age. Fertil Steril
1991;55:784-91.
6. Stone BA, Marrs RP. Ovarian responses to menopausal go-
nadotropins in groups of patients with differing basal
growth hormone levels. Fertil Steril 1992;58:32-6.
7. Dickey RP, Olar TT, Curole DN, Taylor SN, Rye H. Endo-
metrial pattern and thickness associated with pregnancy
outcome after assisted reproduction technologies. Hum Re-
prod 1992;7:418-21.
8. Simonetti S, Veeck LL, Jones HW Jr. Correlation of follicu-
lar fluid volume with oocyte morphology from follicles stim-
ulated by human menopausal gonadotropin. Fertil Steril
1985;44:177-80.
9. Ben-Rafael Z, Kopf GS, Blasco L, Flickinger GL, Tureck
RW, Strauss JF, et al. Follicular maturation parameters as-
sociated with the failure of oocyte retrieval, fertilization,
and cleavage in vitro. Fertil Steril1986;45:51-7.
10. Jones HW Jr, Acosta AA, Andrews MC, Garcia JE, Jones
GS, Mayer J, et al. Three years of in vitro fertilization at
Norfolk. Fertil Steril 1984;42:826-34.
11. Chiu TTY, Tam PPL. A correlation of the outcome of clini-
cal in vitro fertilization with the inositol content and em-
bryotrophic properties of human serum. J Assist Reprod
Genet 1992;9:524-30.
12. Van Wissen B, Eisenberg C, Debey P, Pennehouat G, Auger
J, Bomsel-Helmreich 0. In vitro DNA fluorescense after in
vitro fertilization (IVF) failure. J Assist Reprod Genet
1992;9:564-71.
13. Loutradis DC, Kiessling AA, Kallianidis K, Siskos K,
Creatsas G, Michalas S, et al. A preliminary trial of human
zygote culture in Ham's F-10 without hypoxanthine. J As-
sist Reprod Genet 1993;10:271-5.
14. Rossavik IK, Gibbons WE. Variability of ovarian follicular
growth in natural menstrual cycles. Fertil Steril
1985;44:195-9.
15. O'Herlihy C, De Crespigny LC, Lopata A, Johnston I, Hoult
I, Robinson H. Preovulatory follicular size: a comparison of
ultrasound and laparoscopic measurements. Fertil Steril
1980;34:24-6.
16. Mantzavinos T, Garcia JE, Jones HW Jr. Ultrasound mea-
surements of ovarian follicles stimulated by human gonado-
tropins for oocyte recovery and in vitro fertilization. Fertil
Steril1983;40:461-5.
17. Quigley MM, Wolf DP, Maklad NF, Dandekar PV, Soko-
Wittmaack et al. Follicular size in IVF cycles 1209
 loski JE. Follicular size and number in human in vitro fertil-
ization. Fertil Steril 1982;38:678-80.
18. Lopata A, Kellow G, Leung P, du Plessis Y., McBain J,
Gronow M, et al. The ovarian follicular response in relation
to fertilization, cleavage and pregnancy success rates. In:
Crosignani PG, Rubin BL, editors. In vitro fertilization. Pro-
ceedings of the Serono Colloquia on Reproduction. Vol. 4.
London: Academic Press, 1983:341-52.
19. Trounson A. In vitro fertilization. In: Martini L, James V,
editors. Current topics in experimental endocrinology. Vol.
5. New York: Academic Press, 1983:43-73.
20. Laufer N, DeCherney AH, Tarlatzis BC, Zuckerman AL,
Polan ML, Dlugi AM, et al. Delaying human chorionic go-
1210 Wittmaack et al. Follicular size in IVF cycles
nadotropin administration in human menopausal gonado-
tropin-induced cycles decreases successful in vitro fertiliza-
tion of human oocytes. Fertil Steril 1984;42:198-203.
21. Clark L, Stanger J, Brinsmead, M. Prolonged follicle stimu-
lation decreases pregnancy rates after in vitro fertilization.
Fertil Steril 1991;55:1192-4.
22. Visser DS, Fourie FleR. The applicability of the cumulative
embryo score system for embryo selection and quality con-
trol in an in-vitro fertilization/embryo transfer programme.
Hum Reprod 1993;8:1719-22.
23. Erenus M, Zouves C, Rajamahendran P, Leung S, Fluker M,
Gomel V. The effect of embryo quality on subsequent preg-
nancy rates after in vitro fertilization. Fertil Steril
1991;56:707-10
Fertility and Sterility