Unexplained infertility categorization based on female laparoscopy and total motile sperm count, and its impact on cumulative live‐births after one in‐vitro fertilization cycle. A retrospective cohort study involving 721 cycles

Abstract Purpose To determine how subcategorizing unexplained infertility based on female laparoscopy and total‐motile‐sperm‐count assessment would impact cumulative live‐births after one in‐vitro fertilization (IVF) cycle. Methods Seven hundred twenty one IVF cycles from Jan 2014‐April 2019 performed at a single‐center were retrospectively analyzed. Couples with unexplained infertility having normal uterine and endometrial morphology were subcategorized into three cohorts, UI (1): those with no tuboperitoneal pathology on laparoscopy and total‐motile‐sperm‐count (TMSC) ≧20 million: n = 103; UI (2): tuboperitoneal pathology on laparoscopy or TMSC <20 million, n = 86; and UI(3): tuboperitoneal status not known: n = 114. Controls were severe male factor, bilateral tubal block, and grade 3/4 endometriosis: n = 418. Primary Outcome was cumulative‐live‐birth‐per‐initiated‐IVF cycle (CLBR). Odds ratio for live‐births were adjusted for confounding factors. Results The CLBR in UI1 cohort was significantly lower than controls (29.1% vs 39; OR = 0.62; 95%CI = 0.39‐0.98; P = .04); but similar in UI2 and UI3 vs. controls. (37.2% vs 39.95%; OR = 0.89, 95%CI = 0.55‐1.44; P = .89) and (38.6% vs 39.95%; OR = 0.98, 95%CI = 0.64‐1.55; P = .98). After adjusting for age, infertility duration, past live‐births, and AMH, the adjusted odds for CLBR in UI1 was 0.48 (95%CI = 0.28‐0.82; P = .007). Conclusions Unexplained infertility when defined after a normal laparoscopy and TMSC significantly lowered cumulative‐live‐births‐per‐initiated‐IVF cycle when compared with traditional diagnosis of tubal, endometriosis, or male factor infertility. In UI subcategory with abnormal laparoscopy or TMSC, CLBR remained unaffected. This information could be useful for counseling couples prior to IVF. Large‐scale prospective studies are needed to confirm this observation.


| INTRODUC TI ON
Unexplained infertility (UI) is a diagnosis of exclusion made after standard infertility investigations involving tests of ovulation, tubal patency, and standard semen analysis have failed to reveal an underlying absolute cause as a barrier in causing natural conception. 1 Existing literature is ambiguous about whether the diagnosis of UI has a better prognosis for live-births in IVF when compared with the diagnoses of endometriosis, tubal or male factor infertility. In a retrospective analysis of 121 744 women undergoing their first cycle of autologous IVF between 2000-2007, the authors found that UI had a better prognosis for live births vis-à-vis tubal or male factor infertility. 2 However, in another cohort study involving 9915 women who underwent IVF/ICSI treatment from 2008 to 2012, the predictors found to be significantly associated with reduced chances of IVF/ICSI success among others were tubal factor and unexplained infertility. 3 In yet another retrospective analysis of data collected from a randomized controlled trial involving 738 women, the odds of live-birth following one-cycle IVF were significantly lowered in the presence of tubal and male factor infertility when compared with UI (OR: 0.57; 95%CI: 0.38-0.84 and 0.42, 95% CI 0.22-0.79). 4 This ambiguity in live-birth prognosis could be because UI cannot be considered a monolith etiologically. It is a heterogenous condition encompassing subtle abnormalities in one or more of the following units: the fallopian tube, oocyte, sperm, or the endometrium. 5 Specifically, the tubal defects may include, peritubal adhesions, fimbrial phimosis, or unilateral tubal block 6 ; peritoneal defects may include minimal, mild, or moderate endometriosis; the sperm defects may include but not limited to, sperm concentration, motility or morphology abnormalities, defects in calcium oscillations leading to poor oocyte activation, etc 7 ; oocyte defects may exist at follicular, cellular or molecular level encompassing genuine empty follicle syndrome, and oocyte nuclear or cytoplasmic defects leading to poor embryonic development 8,9 ;and finally, uterine defects, encompassing morphological or molecular endometrial abnormalities, in each case leading to poor implantation. These defects are subtle enough to be missed by a standard infertility work-up.
Laparoscopy has been used to reasonably rule out tubal and peritoneal factors associated with unexplained infertility that cannot be diagnosed with tests such as HSG alone. [10][11][12] A stricter criteria for assessing semen like total motile sperm count (TMSC) can rule out a sperm related factor, better than WHO 2010 lower reference limits, as a barrier in causing natural conception. [13][14][15] It is known that IVF-ICSI has the potential to overcome infertility due to failure of gamete migration to the site of fertilization or due to failure of sperm penetration into the oocyte. It is unlikely, however, to overcome infertility due to defects in oocyte, sperm or endometrium responsible for molecular events occurring post sperm entry into the oocyte such as oocyte activation, 16 sperm decondensation, pronuclear alignment and fusion, cellular cleavage, embryonic genomic activation, blastulation, or implantation. 7,17 The absence of conditions potentially treatable by IVF-ICSI (tubal, endometriosis, oligo-astheno-teratozoospermia) increases the chance of existence of molecular defects either in the gametes or endometrium. It is, therefore, hypothesized that in couples with a subcategory of UI who have normal female laparoscopy and TMSC, the live-birth outcomes after IVF would be poorer when compared with couples undergoing IVF for potentially treatable factors like endometriosis, tubal or male factor infertility.

| ME THODS
The study was approved by the institutional ethics committee vide registration number EC/01/19/1465. Data of all couples undergoing IVF cycles performed between Jan 2014-April 2019 at a tertiary care-multiple-provider set up was collected. The diagnostic categories of male, tubal, endometriosis, and unexplained infertility were included in the study. Exclusions were donor-recipient cycles, preimplantation genetic testing, presence of gross uterine factor (diffuse adenomyosis, grade 2/3 Asherman's syndrome, uncorrected FIGO type 0, 1 or 2 fibroids, mid-cycle endometrial thickness <6 mm) and incomplete cycles; defined when embryos from one oocyte aspiration cycle remained unutilized, in the absence of a live-birth or ongoing pregnancy.
The diagnostic categories of male, tubal, and endometriosis were regrouped into the control cohort. The diagnostic category of UI was subcategorized into three cohorts based on previous laparoscopy and TMSC findings as shown in Figure 1. Basically, previous laparoscopy findings were analyzed to define presence or absence of tuboperitoneal disease. Laparoscopic findings of unilateral tubal block, fimbrial phimosis, peri-adnexal disease, pelvic adhesions, and grade 1/2 endometriosis constituted "tuboperitoneal disease." The absence of above constituted "no tuboperitoneal disease." In the absence of a prior laparoscopy, the status of tuboperitoneal disease was deemed "indeterminate." Semen reports of UI cycles were analyzed next. At least two semen reports (WHO 2010 standard) were available for all couples; one done prior to IVF initiation and one on IVF day. The poorer report was considered to calculate TMSC using the formula: total volume X sperm concentration X percentage of progressively motile sperm. TMSC of ≧20 million was deemed non-sperm factor UI, and TMSC <20 million was labeled as mild-sperm-factor UI. Hamilton et al (2015) have argued and shown that TMSC is a better parameter than WHO 2010 criteria to assess discriminating potential of sperm for natural pregnancy as it takes the absolute value of three semen parameters into consideration simultaneously. 13 In addition, in 518 couples undergoing their first ICSI cycles, TMSC was shown to have a better predictive value over WHO lower reference limits for fertilization, blastulation, and miscarriage rates, and values more than 20 million/mL were suggestive of normal sperm function. 14 Unexplained infertility cycles were finally subcategorized into three cohorts namely UI1: having non-tuboperitoneal and non-sperm factor unexplained infertility; UI2: having mild tuboperitoneal and/ or mild-sperm-factor unexplained infertility; and UI3: having indeterminate factor, labeled so because in the absence of laparoscopy, a tuboperitoneal factor could not be ruled out. The distribution of different sub etiologies within the three cohorts of UI is shown in Table 1.

| Laparoscopy and in-vitro fertilization
At the center of study, laparoscopy is not the first investigative modality to assess fallopian tubal patency. It is usually offered to women with at least three failed IUIs, UI with prior pelvic surgery, long-duration UI usually exceeding two years, and positive pelvic findings on ultrasound or hysterosalpingography. Operative measures are undertaken during laparoscopy to address periadnexal adhesions, cornual block, and endometriosis. Infertility state continuing after a period of one year with active attempts at pregnancy is considered failed laparoscopy, and IVF is offered subsequently.
In women with advanced age, long infertility durations, or with low AMH, an early IVF is offered and sometimes as the first treatment modality.

| In-vitro fertilization protocol
The protocol for IVF and transfer has been defined previously. 18 In brief, all women underwent IVF in agonist or antagonist regimes with gonadotropin doses determined based on the woman's age, Body Mass Index (BMI), AMH, and previous response to ovarian stimulation. Fertilization of oocytes was achieved using conventional insemination or intracytoplasmic sperm injection. Embryos resulting from one cycle were transferred on day 2/3/5 in either fresh or subsequent frozen cycle till a live-birth or ongoing pregnancy was achieved or embryos from one cycle were exhausted.

| Outcomes
Primary outcome variable was live-birth resulting from the transfer of embryos generated per oocyte aspiration cycle. Live-births were defined as all births at or beyond 26 weeks of gestation with survival for at least a day as per the hospital's neonatal intensive care guideline on the age of viability and active neonatal resuscitation policy.
Crude and adjusted odds ratio for live-births were calculated for all UI cohorts against controls.
Other outcomes assessed were total retrieved oocytes, mean total utilizable embryos, total-eggs-to-utilisable-embryo-ratio, totaleggs-to-live-birth-ratio, and mean-transfers-to-one-live-birth. Total utilizable embryos were a sum of embryos/blastocysts transferred and embryos/blastocysts frozen in one oocyte retrieval cycle.

| Statistical analysis
Data were expressed as mean ± 2 SD for gaussian variables and as median, interquartile range (IQR) for non-gaussian variables.
Independent samples t test and one-way ANOVA were used to compare means between two groups and more than two groups, respectively, when the studied variable had a Gaussian distribu-

| Population characteristics
Nine hundred forty seven oocyte retrieval cycles were undertaken in the period of study. Two hundred twenty six were excluded on account of being incomplete (21), employing donor eggs (150) or preimplantation genetic testing (16), oocyte freezing (6)

| Comparison between the four cohorts
The number of cycles with diagnoses of UI1, UI2, UI3, and controls were 103, 86, 114, and 418, respectively. A comparison of pretreatment variables between the four cohorts is given in Table 2. when compared with controls. (Table 3).

| Outcomes
The crude odds ratio for live-births against controls for the UI1 cohort was 0.62 (95%CI 0.39-0.98, P = .043). The adjusted odds ratio for live-births adjusting for confounding variables age, infertility duration, past live-births, past treatments, and AMH was 0.48 (95% CI = 0.28-0.82; P = .004). The crude and adjusted odds ratio for live-births in the UI2 and UI3 subcategory against controls were not significant (Table 4).

| D ISCUSS I ON
Unexplained Infertility (UI) encompasses various conditions arising due to defects in one or the other unit of reproduction and is considered a diagnosis of exclusion incumbent upon normalcy of semen analysis, tubal patency, and ovulation tests. The greater the armamentarium of investigations employed, the narrower the prevalence of UI is likely to be. 19 In a study that employed routine laparoscopy in women with UI, (where prior tubal patency had been established on HSG), 25% additional cases of pelvic pathology could be discovered, encompassing severe tubal disease (4%), peri-adnexal adhesions (8%), and Grade 1/2 endometriosis (13%). 6 Our study showed additional tuboperi- The present study divided UI into a non-tuboperitoneal-non-sperm-factor cohort and a mild-tuboperitoneal/mild-spermfactor cohort, after application of laparoscopy findings and total motile sperm count. In doing so, it attempted to address the hypothesis that IVF-ICSI would be potentially less successful in the non-tuboperitoneal-non-sperm-factor cohort, where embryogenesis or implantation events downstream of sperm penetration into the oocyte were most likely to be affected.
This study showed that couples with UI defined after normal laparoscopy and TMSC of ≧20 million (representing non-tuboperitoneal, non-sperm factor infertility) had fewer live-births per started cycle compared with couples undergoing IVF for a severe tuboperitoneal or severe sperm factor. (OR, 0.62; 95% CI = 0.32-0.98, P = .04).
The difference in live-births remained significant, infact even more enhanced after adjusting for cofounding factors. (OR = 0.48; 95% CI = 0.28-0.82), suggesting that if age, infertility duration, past livebirths, and AMH were equal in the two groups, the odds for a livebirth after one cycle of IVF in couples with UI1 would be reduced to half when compared with couples with a known factor infertility.
Also it was seen that in the UI1 cohort, despite a mean higher number of retrieved eggs, the numbers of utilizable embryos were similar to controls. And despite a mean higher number of transfers in this cohort, the live-births remained lower. Together, the above  a Cumulative live-birth = Live-birth resulting from all transfers from one initiated IVF cycle.

TA B L E 3 IVF outcomes in the four cohorts
b Oocytes to embryo ratio = Total oocytes needed to make one utilizable embryo, obtained by ratio of total retrieved oocytes to total utilizable embryos. c Oocytes to-live-birth-ratio = mean eggs needed to produce one live-birth.
findings appear to suggest, that in unexplained infertility with normal laparoscopy and total motile sperm count, sub-microscopic gamete, or endometrial defects predominate as etiologic factor and hence are potentially less amenable to IVF-ICSI.
Endometrial implantation potential has been found to be lower in some cases of unexplained infertility due to chronic endometritis, 28 alteration of vaginal flora, alteration in endometrial blood flow, 29 or in the expression of proimplantation molecules. 30 But uterine causes are a rare cause of infertility 31 and may not contribute to more than 10%-15% of all cases of true unexplained infertility. 7 Since it was seen that true unexplained infertility relied on supernumerary embryos and multiple transfers to overcome infertility, it is presumed that the defect most likely lay at the embryonic level with some embryos having the potential and most not having the potential for live-birth. Further it may be said that since AMH strongly correlates with retrieved oocyte numbers, which in turn correlates positively with supernumerary embryos, AMH might correlate better with live births in true unexplained infertility.
To check this hypothesis, a correlation of AMH value with cumulative live-births in UI cohorts and controls was performed using ROC curves. (Figure 2) The analysis shows AMH to be a strong predictor of cumulative live-births in UI (1)

| Age, infertility duration and AMH
It is noted that female age is marginally higher by 1. It is also essential to note that median AMH is higher in UI1 cohort compared with controls despite median age being contrarily higher.
The higher AMH in UI1 cohort could reflect a selection bias probably suggesting either an elimination of women with very low AMH or an inadvertent inclusion of more women with PCOS. Could a higher prevalence of PCOS in UI1 have negatively influenced outcomes?
To answer this question, an analysis of AMH distribution among the four cohorts was done ( This is a retrospective analysis and although attempts have been made to address biases, the study needs external validation through prospectively done studies within more strictly selected populations.
In conclusion, unexplained infertility, when defined after a normal laparoscopy and a normal total motile sperm count significantly lowers cumulative live-births-per-started-IVF cycle when compared with traditional diagnosis of tubal factor, endometriosis-associated, or male factor infertility. The difference persists despite adjustments for age, previous live births, AMH, and infertility duration. In unexplained infertility subcategory with abnormal laparoscopy or TMSC, the live-birth rates in IVF remain unaffected when compared with couples with traditional diagnosis of tubal factor, endometriosis-associated, or male factor infertility. This information could be useful for counseling couples prior to IVF. Also, based on the mean embryo to live-birth ratio and the mean transfer to live-birth ratio, the couple with true unexplained infertility, especially those with long-duration infertility, could be counseled that, they may need more total cycles to achieve a livebirth than their other counterparts.