Minimal Stimulation and Natural Cycle In Vitro Fertilization, 1st ed. 2015

3. Natural Cycle IVF: Follicle Physiology and Treatment Costs

Michael von Wolff 

(1)

Division of Gynaecological Endocrinology and Reproductive Medicine, University Women’s Hospital, Effingerstrasse 102, Berne, 3010, Switzerland

Michael von Wolff

Email: michael.vonwolff@insel.ch

Abstract

Several studies have shown that the implantation potential of oocytes derived from natural cycle IVF (NC-IVF) seems to be higher compared to conventional gonadotropin stimulated IVF (cIVF). The following chapter first describes that this difference might be due to differences in the follicular endocrine milieu in NC-IVF and cIVF. Intrafollicular concentration of androgens and the implantation marker anti-Müllerian hormone (AMH) are significantly higher in NC-IVF. These differences are possibly due to suppressed luteinizing hormone (LH) production in cIVF.

The chapter also describes a treatment protocol which is most cost-effective and promises the lowest possible treatment costs per achieved pregnancy. Using such an optimized treatment protocols with only 1.2 consultations before aspiration, with follicle aspiration without the use of analgesics, with flushing of the follicles, and with a pregnancy rates of 13 % per cycle, NC-IVF treatment requires around 1/3 more time to achieve a pregnancy but treatment costs per achieved pregnancy are lower than in cIVF.

Keywords

In vitro fertilizationNatural cycle IVFAnti-Müllerian hormoneTestosteroneCosts

Differences in the Endocrine Milieu in NC-IVF and cIVF

Ever since gonadotropins were introduced into IVF treatment, it has been discussed whether they have an effect on egg cell quality and which preparations or stimulation regimens are more advantageous. The focus of this discussion is the difference between human menopausal gonadotropins (hMG), recombinant follicle-stimulating hormone (r-FSH) with and without recombinant luteinizing hormone (r-LH) supplementation, and the gonadotropin dose.

Furthermore, the discussion is limited to the differences of various gonadotropins. The differences between naturally matured follicles (NC-IVF) and follicles stimulated conventionally with exogenous gonadotropins (cIVF) have hardly been studied. This is quite astonishing as the implantation potential of embryos, derived from NC-IVF follicles, seems to be higher in NC-IVF, suggesting that nature seems to provide us with an ideal model of the physiologically best follicle.

Furthermore, markers which indicate oocyte quality have been sought after for years. These markers include intrafollicular anti-Müllerian hormone (AMH). In several studies, high AMH concentrations correlate with the pregnancy rate (Fanchin et al. 2007; Takahashi et al. 2008; Pabuccu et al. 2009). Which mechanisms stimulate AMH production have only been rudimentally investigated up to now.

Because of this, it is clear that the endocrine milieu of naturally matured follicles should be compared with follicles after high-dose gonadotropin stimulation. The follicular fluid from NC-IVF follicles can be considered to be a model for the ideal follicle as evolution has possibly perfected folliculogenesis and every endocrine manipulation is likely to demonstrate an adverse disruption of the endocrine milieu.

In fact, the analysis of the follicular fluid, collected at the time of follicular aspiration in 40 NC-IVF cycles and in 36 cIVF cycles, using the antagonist protocol, revealed marked differences in the endocrine profile (von Wolff et al. 2014a):

AMH, testosterone (T), androstenedione (A2), estradiol (E2), follicle-stimulating hormone (FSH), and luteinizing hormone (LH) concentrations are significantly different in NC-IVF and in cIVF follicles. These six parameters are shown in detail in box and whisker plot diagrams in Fig. 3.1. AMH, T, and E2 concentrations are around threefold; A2 around 1.5-fold; and LH around 14-fold higher in NC-IVF than in cIVF follicles.

A312222_1_En_3_Fig1_HTML.gif

Fig. 3.1

Differences in the endocrine milieu in follicles of NC-IVF (NC) and cIVF. Concentrations of different hormones in the follicular fluid at the time of follicle aspiration. All concentrations are significantly different in both therapies (n = 76) (Reprinted from von Wolff et al. 2014a)

The marked difference in the concentration of the putative implantation marker AMH in NC-IVF follicles raised the question, if the concentration of AMH correlates with other follicular fluid and serum parameters, suggesting a metabolic link. Such a direct or indirect link was analyzed by a regression analysis of AMH and testosterone. Testosterone concentrations are positively correlated (r = 0.35, p = 0.0002) with AMH concentrations, suggesting that the higher AMH concentrations in NC-IVF than in cIVF follicles are due to higher testosterone concentration in NC-IVF follicles (von Wolff et al. 2014a).

The significantly higher testosterone concentration in NC-IVF follicles might be either due to increased testosterone production due to increased LH activity in NC-IVF or due to an inhibition of the follicular aromatase, inhibiting the conversion of T in E2. A positive correlation of testosterone and LH (r = 0.48, p < 0.0001) (von Wolff et al. 2014a) suggests first a metabolic link between LH and testosterone and second that high testosterone concentration in NC-IVF follicles is probably due to much LH activity in NC-IVF than in cIVF. An inhibition of the aromatase activity with a reduced conversion of T into E2 seems to be unlikely as such an effect would result in an accumulation of testosterone with a non-linear correlation of T and E2, which could not be demonstrated by von Wolff et al. (2014a).

Two essential questions have derived from these results: (1) What significance do increased AMH concentrations have? (2) Which regulatory mechanisms lead to increased AMH concentrations?

The first question, the significance of follicular AMH on oocyte quality, cannot currently be answered easily, as the significance of AMH is unclear. However, it is proven that AMH is produced by granulosa cells (Vigier et al. 1984) and that atretic granulosa cells do not produce AMH (De Vet et al. 2002). The degree of apoptosis on the granulosa cells correlated with the developmental competence of the oocyte (Nakahara et al. 1997). These relationships lead to the hypothesis that a high AMH concentration may have no direct effect on the oocyte but is only a marker for the granulosa cell function and as such, is of relevance for the function of the oocyte. Several studies hypothesize that there is a direct link between the oocyte function and the AMH production of the granulosa cells. The oocyte seems to activate various physiological processes in the surrounding granulosa cells. In a mouse model, it was shown that the oocyte influenced the AMH expression by this mechanism (Salmon et al. 2004).

The second question, which regulatory mechanisms lead to AMH production, can only be answered indirectly at the moment. Andersen and Lossl (2008) indirectly proved that during IVF treatment, induction with human chorionic gonadotropin (hCG before gonadotropin stimulation leads to higher intrafollicular androgen concentrations as well as increased follicular AMH concentrations, i.e., that intrafollicular testosterone possibly stimulated the AMH production. Based on our study results, this means that the increased androgen concentrations in the naturally matured follicles are the reason for the increased AMH concentrations. The precise mechanisms for the stimulation of AMH production are, however, still unknown. Androgens may induce FSH receptor expression in the granulosa cells (Weil et al. 1999). A direct stimulatory effect of LH is also possible. A stimulatory effect of hCG/LH on the AMH production in granulosa cells of polycystic ovaries (PCO) patients but not in the granulosa cells of healthy women was found by Phy et al. (2004). In our study, we detected a strong correlation between the follicular testosterone concentration and the AMH concentration, which supports the supposition of a dependency of the AMH concentration on the testosterone concentration.

This in turn raises the question of which mechanisms lead to an increased follicular testosterone concentration. The increased androgen concentrations in naturally matured follicles can be based upon two different mechanisms. Either AMH inhibits the aromatase, as a result of which androgens accumulate or androgenesis in the theca cells is (mediated by LH) increased.

The second hypothesis implies that androgen synthesis is increased in naturally matured follicles in the theca cells. Thecal androgen synthesis is stimulated by LH. This appears to confirm the hypothesis that in NC-IVF, the LH concentrations in the serum as well as in the follicular fluid are significantly higher, because LH suppression using gonadotropin-releasing hormone (GnRH) analogs or GnRH antagonists, as in cIVF, is not performed.

The unphysiological suppression of LH in cIVF seems to affect a cascade of metabolic changes within the follicle (Fig. 3.2). This possibly leads to lower implantation potential of the oocytes in cIVF and is probably one reason for the higher implantation potential in NC-IVF.

A312222_1_En_3_Fig2_HTML.gif

Fig. 3.2

Model for the effect of LH on the follicular endocrine milieu. The model is based on the data presented in Fig. 3.1 and on the study by von Wolff et al. (2013a) and Vaucher et al. (2013). Suppressed LH, as found in cIVF, due to GnRH agonists or GnRH antagonists result in lower follicular testosterone concentrations and thereby, in lower AMH concentration, leading possibly to a lower implantation potential of the oocyte in cIVF. AMH has been shown to be a marker for the implantation potential of the oocyte (Fanchin et al. 2007) (dark blue: theca cells, light blue: granulosa cells, green: oocyte)

Treatment Protocol in NC-IVF

Theoretically, NC-IVF is performed without any hormones. Even hCG is not used to induce follicle maturation. Practically, this approach is not useful, as the efficacy of the treatment would be far too low to be able to compete with cIVF.

The requirements of NC-IVF are therefore:

1.

2.

3.

4.

To fulfill the first two requirements, the physician can either monitor the follicles every 1–2 days to identify the ideal day for ovulation induction, which requires many expensive and time-consuming consultations. Or, the physician uses a treatment concept that allows reduction of consultations without reducing pregnancy rate.

As gonadotropins and GnRH antagonists are expensive and as Clomiphene citrate at the common dosage of 50 mg/day has negative effects on endometrial function and can induce cyst formations resulting in cycle cancelations in the following cycle, we have introduced a treatment protocol with very low dosages of Clomiphene citrate (von Wolff et al. 2014b).

Patients received 25 mg Clomiphene citrate per day, started on day 6 or 7 until 24 h before ovulation induction with hCG is given (Fig. 3.3). The first consultation takes place on cycle day 10 ± 2. Follicular diameter and endometrial thickness are analyzed by ultrasound and concentrations of estradiol (E2) and LH by serum analysis. The results are used to calculate the expected time of ovulation. In a few cases, a second consultation is required. 5,000 IU of human chorionic gonadotropin is given 36 h before follicle aspiration when the follicle is ≥15 mm and estradiol concentration is ≥700 pmol/L.

A312222_1_En_3_Fig3_HTML.gif

Fig. 3.3

Treatment protocol of NC-IVF as performed in Berne (clomiphene citrate 25 mg/day = CC25)

The results of this treatment protocol are shown in Fig. 3.4. Less than 5 % of patients described mild side effects such as hot flushes or headache. These results clearly indicate the superiority of the treatment protocol in respect to the transfer/cycle ratio. They also demonstrate that Clomiphene citrate did not reduce pregnancy rate.

A312222_1_En_3_Fig4_HTML.gif

Fig. 3.4

Outcome of the 108 NC-IVF and the 103 ccNC-IVF (with clomiphene citrate) therapies. Significant differences of the treatment steps between the treatment groups are marked by an asterisk. Patients underwent both kind of therapies without randomization. One hundred three patients started with an NC-IVF cycle and, if not pregnant after the first cycle, followed by a ccNC-IVF cycle. Fourteen patients started with a ccNV-IVF cycle (Modified according to von Wolff et al. 2014b)

Therefore, to perform treatment cycles with the lowest possible number of consultations, lowest possible treatment costs and highest possible pregnancy chances per cycle, the use of Clomiphene citrate at low dosages should be considered. Accordingly, the calculation of treatment costs and pregnancy rate per treatment time in the chapter below is based on the treatment results achieved by the use of low dosages of Clomiphene citrate.

Follicle Aspiration in NC-IVF

The technique of follicle aspiration plays an important role in NC-IVF as first, it needs to be highly efficient to get the highest possible number of oocytes and second, it should be simple and not very painful in order to allow monthly treatment cycles. To improve the efficacy of follicle aspiration, we have reintroduced follicle flushing in monofollicular NC-IVF.

This change in our treatment protocol is in contrast to the common scientific evidence. A Cochrane analysis even concluded that “there is no evidence that follicular aspiration and flushing is associated with an increase in oocyte yield.” Follicular flushing even seemed to be disadvantageous as “the operative time is significantly longer and more opiate analgesia is required for pain relief during oocyte retrieval” (Wongtra-Ngan et al. 2010). However, this statement was based on different studies in which follicle aspiration was performed in polyfollicular IVF (normal responders) and oligofollicular IVF (low responders) following controlled ovarian hyperstimulation.

We therefore performed a clinical study to analyze if this statement also applies to NC-IVF (von Wolff et al. 2013a): 164 aspirations were performed in monofollicular IVF cycles. Follicles were aspirated without any anesthesia, using 19G single-lumen needles (250 mmHg). Pain intensity during follicle aspiration was low (Fig. 3.5). After initial aspiration, follicles were flushed and aspirated three times each with 2 mL flushing medium with heparin. Total oocyte yield/aspiration was 44.5 % in the aspirate, 20.7 % in the first flush, 10.4 % in the second flush, and 4.3 % in the third flush (Table 3.1). By flushing, the total oocyte yield increased significantly (p < 0.01) by 80.9 % from 44.5 to 80.5 %.

A312222_1_En_3_Fig5_HTML.gif

Fig. 3.5

Pain intensity in NC-IVF aspiration without analgetics/anesthesia. Pain intensity of follicle aspiration using a 19G aspiration needle in relation to venous blood sampling according to a survey performed in Berne

Table 3.1

Effect of follicle flushing on oocyte yield in NC-IVF

 

NC-IVF

Aspirations (n)

164

Mean age (y) ± SD

37.0 ± 3.8 (28–45)

Mean follicular diameter (mm) ± SD

19.0 ± 2.1

Total oocytes/aspirations = n (%)

132/164 (80.5)

 

Aspirate

First flush

Second flush

Third flush

Oocytes in each group/aspirations = n (%)

73/164 (44.5)

34/1645 (20.7)

17/164 (10.4)

7/164 (4.3)

MII oocytes/aspirated oocytes = n (%)

67/73 (91.8)

31/34 (91.2)

16/17 (94.1)

7/7 (100)

Fertilized oocytes (transfer rate)/aspiration = n (%)

33/164 (20.1)

16/164 (9.8)

10/164 (6.1)

4/164 (2.4)

Total fertilized oocytes (transfer rate)/aspiration = n (%)

63/164 (38.4)

Reprinted from von Wolff et al. (2013)

The proportions of MII oocytes/aspirated oocytes were similar in all four groups (91.8 %, 91.2 %, 94.1 %, 100 %). The proportion of fertilized (2PN) oocytes/MII oocytes were also similar in all four groups (49.3 %, 51.6 %, 62.5 %, 57.1 %). Transfer rates in each group/total aspirations were 20.1, 9.8, 6.1, and 2.4 %. By flushing, the total transfer rate increased significantly (p < 0.01) by 91.0 % from 20.1 to 38.4 %.

Three flushings thereby almost doubled not only the number of aspirated oocytes but also the transfer rate in monofollicular IVF. These results indicate that first the oocyte yield can be increased by flushing and second that the oocytes, collected by flushing, are equally mature and fertilizable than those aspirated without flushing.

Luteal Phase in NC-IVF

Theoretically, luteal phase support is not needed in NC-IVF. In contrast, practically most physicians do offer progesterone supplementation in NC-IVF. The main reason is probably due to the habit that progesterone is also substituted in cIVF within the first 2 weeks after follicle aspiration. However, in cIVF, luteal phase support is needed due to the LH drop either induced by the unphysiological estradiol concentration or induced by the lasting effect of GnRH agonists and GnRH antagonists. As both effects do not apply to NC-IVF, luteal phase support is not required in NC-IVF if the follicle is just aspirated.

But what happens if the follicle is not only aspirated but additionally flushed as described above? The process of flushing leads to a reduction of intrafollicular granulosa cells and thereby – at least theoretically – to a reduced pool of progesterone and possibly, estrogen-producing luteal cells. Therefore, if follicles are flushed, progesterone and even possibly, estrogen supplementation should be considered, according to the treatment protocols of thawing cycles using estrogen to prepare the endometrium.

In Berne, we are currently performing a clinical study analyzing the effect of follicle flushing on the progesterone and estrogen concentration in the luteal phase. Results of the study are expected.

Treatment Costs in NC-IVF

In cIVF several new and expensive laboratory techniques have been introduced within the past few years to increase pregnancy rates. Patients are offered techniques such as assisted hatching, spindle and zona analysis, preimplantation screening (PID), etc. However, the effects of these procedures on increasing pregnancy rates have not been clearly proven. Assuming that naturally matured follicles cannot be further optimized, in NC-IVF, most of these techniques would not be required, allowing reduction of treatment costs.

To further reduce treatment costs, the number of consultations before follicle aspiration needs to be minimized. We reduced the number to an average of 1.2 first by using Clomiphene citrate and second by the largely reliable calculation of the time of ovulation by measuring the follicle size, as well as the estradiol and LH levels. To reduce treatment costs even further, we perform the NC-IVF aspiration without anesthesia.

Our pregnancy rate in all treated patients up to the age of 42 years was 13.6 % per cycle (Table 3.1). In a patient group with 2/3 Clomiphene citrate cycles, Aanesen et al. (2010) described a pregnancy rate of 14.7 % per cycle with ca. 2–3 consultations before the aspiration per cycle. Schimberni et al. (2009) identified a pregnancy rate of 9.8 % per cycle in poor responders without the use of Clomiphene citrate with daily ultrasound monitoring from the seventh day of the cycle and with aspiration without local or general anesthesia. By using low doses of Clomiphene citrate, we achieved the pregnancy rates described in other studies but with a significantly lower time and cost expenditure.

We consider that the calculation of the costs of NC-IVF per cycle on the basis of a pregnancy rate of 13 %, a transfer rate per cycle of 55 %, and one consultation before the aspiration, which is performed without anesthesia, to be representative for optimized NC-IVF treatment.

Based on these numbers, we performed a comparison of treatment costs of cIVF and NC-IVF (Tables 3.2 and 3.3). The costs for cIVF were based on the assumption of a pregnancy rate of 30 % per fresh cycle and 20 % per cryo-cycle, in accordance with the ESHRE-IVF register (Ferraretti et al. 2012). Three consultations before the aspiration and one aspiration under anesthesia were calculated. Furthermore, differences between low responding patients with an expected transfer per stimulation and normally responding patients with a fresh transfer and a cryo-transfer allowed typical situations in the IVF practice to be incorporated.

Table 3.2

Estimated costs of cIVF compared with NC-IVF, using low dosages of clomiphene citrate (von Wolff et al. 2014c)

 

cIVF – one fresh cyclea

cIVF – one cryo-cycle following cIVFb

NC-IVF – one fresh cyclec

Total required consultations/cycle (n)

5

3

3

Required labor – physician (min)

105

60

75

Required labor – secretaries and nurses (min)

90

45

60

Required labor – IVF laboratory staff (min)

250

120

195

Required medication (€)

1200,–

70,–

40,–

Required blood tests (E2, LH) (€)

60,–

0

35,–

Required consumables, IVF laboratory (media, ICSI pipettes, aspiration needle, transfer catheter, sperm preparation, etc.) (€)d

184,–

191,–

179,–

Anesthesia and postoperative care (€)

500,–

0

0

Total costs, consumables, anesthesia, blood tests

1744,–

261,–

254,–

Total labor (min)

445

225

330

Total costs (€)e

2188,–

383,–

431,–

acIVF including gonadotropins and GnRH agonists/antagonists

bIncluding cryopreservation by vitrification and cryo-cycles with estrogen/progesterone supplementation

cNC-IVF with clomiphene citrate

d1/3 of cycle fertilization by ICSI and 2/3 of cycle fertilization by insemination. Gas for incubators, laboratory equipment, etc., not included

ePhysician, € 40,–/h; secretaries and nurses, € 30,–/h

Table 3.3

Treatment costs, pregnancy rate, required consultations, and treatment time of different number of cycles in cIVF and NC-IVF, using low dosages of clomiphene citrate (von Wolff et al. 2014c)

 

Total costs/initiated cycle(s)a

Cumulative pregnancy rate/initiated cycle(s)

Cumulative number of required consultations (n)b

Cumulative required treatment time (month)c

cIVF one cycle without cryo-cycles

2188,–

30 %d

5

1

cIVF two cycles without cryo-cycles

4376,–

51 %d

10

3

One cryo-cycle following cIVF

383,–

20 %d

3

1

cIVF plus one cryo-cycle

2571,–

44 %d

8

3

NC-IVF, one cycle

431,–

13 %e

2.6

1

NC-IVF, two cycles

862,–

24 %e

5.2

2

NC-IVF, three cycles

1293,–

34 %e

7.8

3

NC-IVF, four cycles

1724,–

43 %e

10.4

4

NC-IVF, five cycles

2155,–

50 %e

12.0

5

aCosts according to calculation in Table 3.2. In cIVF, total costs are based on the assumption of one transfer per cycle. In NC-IVF, total costs are based on the assumption of one consultation before aspiration, 90 % aspiration rate/cycle, 80 % oocyte collection rate/cycle, 70 % mature oocyte rate/cycle, and 55 % transfer rate/cycle

bCalculations are based on a transfer rate of 100 % in cIVF and cryo-cycles and 55 % in NC-IVF

cIncluding a break of 1 month following a cIVF cycle (fresh transfer) and no break between NC-IVF cycles

dApproximated according to pregnancy rates in the ESHRE register (Ferraretti et al. 2012)

eApproximated according to pregnancy rate of 13 % (13 %): >1 cycle calculated: i.e., 2 cycles, 100–872/100

The calculations show that under the assumptions stated above, NC-IVF treatment is ca. 15 % less expensive with the same pregnancy rate, but the treatment may take ca. 30 % longer for the same pregnancy rate (Fig. 3.6) (von Wolff et al. 2014c).

A312222_1_En_3_Fig6_HTML.gif

Fig. 3.6

Pregnancy rate and treatment time in NC-IVF and cIVF. Expected pregnancy rate (PR) following 1–4 months of treatment and expected pregnancy rate following cumulative treatment costs of 4,000€ (calculation based on data in Table 3.3) in NC-IVF compared with conventional IVF (cIVF). cIVF was calculated for patients achieving only one transfer per stimulation (described as “low responding”) and those achieving one fresh transfer and one cryo-transfer per stimulation (described as “normally responding”) (Modified according to von Wolff et al. 2014c)

A comparison of the costs of NC-IVF and cIVF has only been rudimentarily tried in the literature up to now. Nargund et al. (2001) calculated that the costs of an NC-IVF pregnancy are ca. 75 % lower than cIVF. However, detailed information on cost calculations was not provided in the study. Aanesen et al. (2010) only compared the costs of medication for NC-IVF and cIVF treatments. The cost of medication was given as ca. €1,200 per cycle for cIVF, assuming a daily gonadotropin dose of 150 IU recombinant FSH. In clinical practice, these costs may be significantly higher in a poor responder and may rise to up to €2,000 per cycle.

In summary, our study shows that NC-IVF is cheaper than cIVF treatment in low responders which is associated with high gonadotropins as well as in normal responders. The additional costs as a result of higher twin pregnancy rates and complications because of overstimulation after cIVF were not taken into account. In contrast, the occurrence of pregnancy takes longer on average with NC-IVF treatment.

References

Aanesen A, Nygren KG, Nylund L. Modified natural cycle IVF and mild IVF: a 10 year Swedish experience. Reprod Biomed Online. 2010;20:156–62.CrossRefPubMed

Andersen CY, Lossl K. Increased intrafollicular androgen levels affect human granulosa cell secretion of anti-Müllerian hormone and inhibin-B. Fertil Steril. 2008;89:1760–5.CrossRefPubMed

de Vet A, Laven JS, de Jong FH, Themmen AP, Fauser BC. Antimüllerian hormone serum levels: a putative marker for ovarian aging. Fertil Steril. 2002;77:357–62.CrossRefPubMed

Fanchin R, Mendez Lozano DH, Frydman N, Gougeon A, di Clemente N, Frydman R, et al. Anti-Müllerian hormone concentrations in the follicular fluid of the preovulatory follicle are predictive of the implantation potential of the ensuing embryo obtained by in vitro fertilization. J Clin Endocrinol Metab. 2007;92:1796–802.CrossRefPubMed

Ferraretti AP, Goossens V, de Mouzon J, Bhattacharya S, Castilla JA, Korsak V, Consortium for European Society of Human Reproduction and Embryology (ESHRE), et al. Assisted reproductive technology in Europe, 2008: results generated from European registers by ESHRE. Hum Reprod. 2012;27:2571–84.CrossRefPubMed

Nakahara K, Saito H, Saito T, Ito M, Ohta N, Takahashi T, et al. The incidence of apoptotic bodies in membrana granulosa can predict prognosis of ova from patients participating in in vitro fertilization programs. Fertil Steril. 1997;68:312–7.CrossRefPubMed

Nargund G, Waterstone J, Bland J, Philips Z, Parsons J, Campbell S. Cumulative conception and live birth rates in natural (unstimulated) IVF cycles. Hum Reprod. 2001;16:259–62.CrossRefPubMed

Pabuccu R, Kaya C, Cağlar GS, Oztas E, Satiroglu H. Follicular-fluid anti-Mullerian hormone concentrations are predictive of assisted reproduction outcome in PCOS patients. Reprod Biomed Online. 2009;19:631–7.CrossRefPubMed

Phy JL, Conover CA, Abbott DH, Zschunke MA, Walker DL, Session DR, et al. Insulin and messenger ribonucleic acid expression of insulin receptor isoforms in ovarian follicles from nonhirsute ovulatory women and polycystic ovary syndrome patients. J Clin Endocrinol Metab. 2004;89:3561–6.CrossRefPubMed

Salmon NA, Handyside AH, Joyce IM. Oocyte regulation of anti-Müllerian hormone expression in granulosa cells during ovarian follicle development in mice. Dev Biol. 2004;266:201–8.CrossRefPubMed

Schimberni M, Morgia F, Colabianchi J, Giallonardo A, Piscitelli C, Giannini P, et al. Natural-cycle in vitro fertilization in poor responder patients: a survey of 500 consecutive cycles. Fertil Steril. 2009;92:1297–301.CrossRefPubMed

Takahashi C, Fujito A, Kazuka M, Sugiyama R, Ito H, Isaka K. Anti-Müllerian hormone substance from follicular fluid is positively associated with success in oocyte fertilization during in vitro fertilization. Fertil Steril. 2008;89:586–91.CrossRefPubMed

Vaucher A, Kollmann Z, Bersinger NA, Weiss B, Stute P, Marti U, von Wolff M. Gonadotropin stimulation in in vitro fertilization (IVF) significantly alters the hormone concentrations in follicular fluid – a comparative study between Natural Cycle and conventional IVF. Hum Reprod. 2013;28(S1):i321–323..

Vigier B, Picard JY, Tran D, Legeai L, Josso N. Production of anti-Müllerian hormone: another homology between Sertoli and granulosa cells. Endocrinology. 1984;114:1315–20.CrossRefPubMed

von Wolff M, Kollmann Z, Vaucher B, Weiss B, Bersinger NA. AMH, a putative follicular marker of oocyte quality, is concentrated around 3-fold higher in follicular fluid of Natural Cycle-IVF than in gonadotropin stimulated IVF. Hum Reprod. 2013b;28(S1):i333–4.

von Wolff M, Hua Y-Z, Santi A, Ocon E, Weiss B. Follicle flushing in monofollicular IVF almost doubles the number of transferable embryos. Acta Obstet Gynecol Scand. 2013a;92:346–8.PubMedCentralCrossRef

von Wolff M, Kollmann Z, Flück CE, Stute P, Marti U, Weiss B, Bersinger NA. Gonadotrophin stimulation for in vitro fertilization significantly alters the hormone milieu in follicular fluid: a comparative study between natural cycle IVF and conventional IVF. Hum Reprod. 2014a;29:1049–57.CrossRef

von Wolff M, Nitzschke M, Stute P, Bitterlich N, Rohner S. Low-dosage clomiphene reduces premature ovulation rates and increases transfer rates in natural-cycle IVF. Reprod Biomed Online. 2014b;29:209–15.CrossRef

von Wolff M, Santi A, Rohner S, Ocon E, Stute P, Popovici R, Weiss B. Modified Natural cycle In-vitro Fertilization – an alternative IVF treatment with lower costs per achieved pregnancy but longer treatment time. J Reprod Med. 2014c;59:553–9.

Weil S, Vendola K, Zhou J, Bondy CA. Androgen and follicle-stimulating hormone interactions in primate ovarian follicle development. J Clin Endocrinol Metab. 1999;84:2951–6.CrossRefPubMed

Wongtra-Ngan S, Vutyavanich T, Brown J. Follicular flushing during oocyte retrieval in assisted reproductive techniques. Cochrane Database Syst Rev. 2010;8:CD004634.