Persistent progesterone elevation in infertility revealing non-classic P450 oxidoreductase deficiency: 2 cases.

A 38-year-old woman with a history of hormone receptor-positive right breast cancer (luminal type, stage I) underwent mastectomy in Year X-5 and received adjuvant tamoxifen therapy until Year X-2. After discontinuing tamoxifen in preparation for pregnancy, she was referred to a reproductive endocrinology clinic for fertility evaluation. Oocyte retrieval was performed in Year X-1, and 9 blastocysts were successfully cryopreserved. During the subsequent cycles, serum progesterone remained elevated, reaching 7.3 nmol/L (2.3 ng/mL) (normal follicular phase: <0.89 nmol/L [<0.28 ng/mL]) during the follicular phase, and gonadotropin suppression with gonadotropin-releasing hormone agonist (buserelin) failed to reduce progesterone to physiological levels, raising suspicion of an extra-ovarian source of progesterone production. Because this sustained progesterone elevation was considered to contribute to endometrial–embryo asynchrony, thereby interfering with embryo transfer, she was referred to our endocrinology department in Year X. Although she also had a history of endometriosis, this condition alone could not readily explain the persistent follicular-phase progesterone elevation, prompting further endocrine evaluation. She had no external genital abnormalities at birth, no apparent skeletal malformations, and no known family history of congenital adrenal disorders. She experienced menarche at age 14 and had a history of endometriosis with bilateral ovarian chocolate cysts diagnosed at age 30, for which she received low-dose oral contraceptives. Prior to the initiation of oral contraceptive therapy, her menstrual cycles had been irregular. She had no clinical signs of hyperandrogenism such as acne or hirsutism and had no previous episodes suggestive of adrenal insufficiency. A 35-year-old woman with a history of anal atresia had been undergoing infertility treatment since Year Y-1. Although intracytoplasmic sperm injection was performed, conception was not achieved. Persistent endometrial thinning and adnexal enlargement were noted during treatment. Repeated progesterone measurements at the referring fertility clinic demonstrated elevated progesterone levels across different menstrual cycle phases, including a level of 6.3 nmol/L (1.99 ng/mL) measured immediately before embryo transfer. At presentation, on day 6 of menses, progesterone was 3.2 nmol/L (1.0 ng/mL) and E2 was 77 pmol/L (20.9 pg/mL) (normal follicular phase: 106-722 pmol/L [28.8-196.8 pg/mL]), indicating a hormonal pattern incompatible with normal follicular physiology. Because these abnormalities were considered to impair endometrial development, she was referred to our endocrinology department in Year Y. She had no external genital abnormalities at birth. The age of menarche was not available from the medical records. Before infertility treatment, she had irregular menstrual cycles of approximately 45 days and had previously received low-dose estrogen–progestin therapy. She had no clinical signs of hyperandrogenism such as acne or hirsutism and no previous episodes suggestive of adrenal insufficiency. Her family history was notable for 2 brothers who had external genital abnormalities suggestive of disorders of sex development according to the patient, although detailed evaluations were not available.

Infertility affects approximately one in 6 adults worldwide [ 1 ], and the use of fertility treatments—including assisted reproductive technologies—has steadily increased over recent decades [ 2 ]. Despite these advances, approximately 15% to 30% of infertile couples have no identifiable cause after standard fertility evaluations and are therefore classified as having unexplained infertility [ 3 ]. Within this population, endocrine disorders represent an important but sometimes overlooked etiology. Congenital adrenal hyperplasia (CAH) is a recognized endocrine cause of female infertility, most commonly due to non-classic 21-hydroxylase deficiency (NC-21OHD) [ 4 ]. Non-classic P450 oxidoreductase deficiency (NC-PORD), a rare and underdiagnosed form of CAH, results from partial impairment of electron transfer from P450 oxidoreductase (POR) to multiple steroidogenic enzymes, including CYP17A1, CYP21A2, and CYP19A1 [ 5 ]. Unlike classic P450 oxidoreductase deficiency (PORD), NC-PORD lacks overt skeletal or genital malformations, and its clinical manifestations may remain subtle until adulthood [ 6 , 7 ]. A key biochemical feature in women with NC-PORD is noncyclic elevation of serum progesterone, which may impair endometrial maturation and interfere with embryo transfer [ 7 ]. However, progesterone elevation alone is not specific to PORD and may also be observed in other forms of CAH [ 4 , 8 ]. Instead, the combination of persistent progesterone elevation, low estradiol (E2) levels, absence of clinical hyperandrogenism, and the presence of ovarian macrocysts is suggestive of impaired multienzyme steroidogenesis characteristic of NC-PORD [ 7 ]. Here, we report 2 women in whom unexplained infertility accompanied by sustained progesterone elevation prompted detailed endocrine evaluation, ultimately leading to the suspicion of NC-PORD as the most likely underlying condition. Reports in which infertility evaluation directly results in the identification of NC-PORD remain rare, and these cases underscore the importance of considering underlying steroidogenic disorders in the workup of unexplained infertility.

After initiation of dexamethasone, serum progesterone decreased to 0.3 nmol/L (0.1 ng/mL). During a subsequent HRT cycle, endometrial thickness improved to 7.2 mm. Dexamethasone was discontinued 1 week before embryo transfer when vaginal progesterone supplementation was initiated. In February of Year X+1, embryo transfer resulted in pregnancy. No glucocorticoid replacement was administered during pregnancy because the patient remained free of symptoms suggestive of adrenal insufficiency. However, the pregnancy ended in a missed miscarriage at 8 weeks and 5 days of gestation. After initiation of dexamethasone, serum progesterone decreased to below the assay detection limit (<0.3 nmol/L [<0.1 ng/mL]). The patient continued assisted reproductive treatment (ART) at the referring fertility clinic. In July of Year Y+1, pregnancy was achieved through ART, and fetal cardiac activity was confirmed in August. However, the pregnancy ended in a missed miscarriage. In March of Year Y+2, pregnancy was achieved again through ART. At that time, dexamethasone was replaced with prednisolone 2 mg/day to reduce the risk of adrenal insufficiency during pregnancy. At the time of writing, the pregnancy is ongoing.

Non-classic PORD can present primarily with infertility, even in the absence of overt adrenal insufficiency or classic features. Persistent progesterone elevation accompanied by relatively low E2 levels and ovarian macrocysts may serve as a key endocrine clue to NC-PORD in patients with unexplained infertility. Early recognition of endocrine abnormalities enables mechanism-based treatment strategies and may improve reproductive outcomes.

Following genetic testing, the patient was considered to have NC-PORD. Because she had no clinical manifestations of adrenal insufficiency, routine glucocorticoid replacement was not initiated, and glucocorticoids were reserved for sick-day use only. Although glucocorticoid therapy was expected to suppress adrenal progesterone production, embryo transfer was initially attempted without steroid replacement in collaboration between endocrinologists and gynecologists. Hormone replacement therapy (HRT) cycles for assisted reproductive treatment were subsequently performed. The first embryo transfer cycle was canceled because of insufficient endometrial development. During a subsequent HRT cycle, serum progesterone was suppressed to 2.2 nmol/L (0.7 ng/mL); however, endometrial thickness remained only 4.7 mm at the time of embryo transfer, and pregnancy was not achieved. To further suppress progesterone production, low-dose dexamethasone (0.25 mg/day) was initiated in Year X. Following genetic testing, the patient was considered to have NC-PORD. Because she had no clinical manifestations of adrenal insufficiency, routine hydrocortisone replacement was not initiated. To suppress adrenal progesterone production, low-dose dexamethasone (0.5 mg/day) was started in Year Y.

Laboratory findings at presentation are shown in Table 1 . Serum progesterone was elevated at 7.3 nmol/L (2.3 ng/mL) during the follicular phase, accompanied by an elevated 17α-hydroxyprogesterone (17-OHP) concentration of 18.2 nmol/L (6.0 ng/mL) (normal: <6.9 nmol/L [<2.3 ng/mL]). Baseline adrenocorticotropic hormone (ACTH), cortisol, dehydroepiandrosterone sulfate (DHEA-S), plasma renin activity, and aldosterone concentrations were within their respective reference ranges ( Table 1 ). Abdominal computed tomography (CT) showed no adrenal morphological abnormalities ( Fig. 1A and 1B ). Pelvic magnetic resonance imaging (MRI) revealed cystic lesions in the left ovary, with the largest measuring approximately 45 mm in diameter ( Fig. 1C ). Abdominal imaging findings in Case 1 and Case 2. (A, B) Abdominal CT of Case 1 showing normal adrenal morphology (arrows). (C) Axial T2-weighted MRI of Case 1 demonstrating multiple left ovarian cysts, the largest measuring approximately 45 mm in diameter (arrow). (D) Coronal abdominal CT of Case 2 showing mild enlargement of the left adrenal limb with a short-axis diameter of approximately 3 mm (arrow). Hematologic, biochemical, and endocrine parameters of Case 1 and Case 2 Hormone levels were measured during the early follicular phase. Reference ranges for LH, FSH, E2, and progesterone correspond to the early follicular phase, whereas those for DHEA-S correspond to women aged 31-40 years. Hormonal assays, including 17-OHP, were performed at an external laboratory (SRL Inc., Tokyo, Japan), and the corresponding reference intervals were derived from their reports. Serum cortisol concentrations were measured using the Elecsys Cortisol II assay (Roche Diagnostics) on a cobas e801 analyzer by ECLIA. Plasma renin activity was measured in Case 1, whereas active renin concentration was measured in Case 2. Abbreviations: 17-OHP, 17α-hydroxyprogesterone; ACTH, adrenocorticotropic hormone; ALT, alanine aminotransferase; AST, aspartate aminotransferase; Cl, chloride; Cr, creatinine; DHEA-S, dehydroepiandrosterone sulfate; E2, estradiol; ECLIA, electrochemiluminescence immunoassay; eGFR, estimated glomerular filtration rate; FSH, follicle-stimulating hormone; FT3, free triiodothyronine; FT4, free thyroxine; Hb, hemoglobin; K, potassium; LH, luteinizing hormone; Na, sodium; ND, not data; Plt, platelet count; TSH, thyroid-stimulating hormone; WBC, white blood cell count. The rapid ACTH stimulation test demonstrated a suboptimal cortisol response, with a peak cortisol level of 351 nmol/L (12.7 μg/dL), which remained below the recently proposed 60-minute cutoff for the Elecsys Cortisol II assay (485 nmol/L [17.6 μg/dL]) [ 9 ]. In contrast, both baseline and stimulated levels of 17-OHP and progesterone were markedly elevated, with peak values of 40.6 nmol/L (13.4 ng/mL) and 59.1 nmol/L (18.6 ng/mL), respectively ( Fig. 2 ). These findings were suggestive of CAH. Results of the rapid ACTH stimulation test in Case 1 and Case 2. (A) Case 1. (B) Case 2. Serum cortisol, 17-OHP, and progesterone levels at baseline and at 30 and 60 minutes after ACTH (250 μg) stimulation. DHEA-S values obtained during the ACTH stimulation test are shown in Case 2. Abbreviations: ACTH, adrenocorticotropic hormone; DHEA-S, dehydroepiandrosterone sulfate; 17-OHP, 17α-hydroxyprogesterone. Genetic testing was performed after obtaining informed consent according to local institutional guidelines. Targeted next-generation sequencing using an adrenal disease gene panel at the Kazusa DNA Research Institute identified heterozygous variants in the POR gene ( c.1370G>A [p.Arg457His] and c.1648C>T [p.Arg550Trp]) ( Table 2 ). Based on American College of Medical Genetics and Genomics/Association for Molecular Pathology (ACMG/AMP) criteria, these variants were classified as pathogenic or likely pathogenic. Although parental genetic testing could not be performed due to advanced parental age and limited accessibility, the combination of clinical, biochemical, and genetic findings was compatible with NC-PORD. Genetic testing results of Case 1 and Case 2 Variant annotation is based on the human reference genome GRCh38 and the POR transcript NM_000941.3 . Population frequencies were derived from gnomAD v4.1.0 (GRCh38). In silico predictions were obtained using Ensembl VEP (SIFT, PolyPhen-2) and MutationTaster (v2025). Functional domain assignment was based on the crystal structure of human P450 oxidoreductase [ 13 ]. ACMG/AMP (2015) criteria were used for variant classification. Abbreviations: ACMG, American College of Medical Genetics and Genomics; AMP, Association for Molecular Pathology; gnomAD, Genome Aggregation Database; InDel, insertion/deletion; MT, MutationTaster; POR, P450 oxidoreductase; SIFT, Sorting Intolerant From Tolerant; VEP, Variant Effect Predictor. Laboratory findings at presentation are shown in Table 1 . Despite being on day 6 of the menstrual cycle, serum progesterone was elevated at 3.2 nmol/L (1.0 ng/mL), with mildly decreased E2 levels. Serum 17-OHP was also elevated at 7.9 nmol/L (2.6 ng/mL). Baseline ACTH, cortisol, DHEA-S, active renin concentration, and aldosterone concentrations were within their respective reference ranges ( Table 1 ). Abdominal CT revealed slight enlargement (3 mm in short diameter) of the left adrenal gland ( Fig. 1D ). Although adnexal enlargement had been noted during infertility treatment, ovarian macrocysts comparable to those observed in Case 1 were not identified on imaging studies. Based on persistently elevated progesterone levels throughout the menstrual cycle, the family history, and imaging findings, a form of CAH was suspected. The rapid ACTH stimulation test demonstrated a suboptimal cortisol response, with a peak cortisol level of 406 nmol/L (14.7 μg/dL), which also remained below the recently proposed 60-minute cutoff for the Elecsys Cortisol II assay. In contrast, both baseline and stimulated levels of 17-OHP and progesterone were markedly elevated, with peak values of 31.5 nmol/L (10.4 ng/mL) and 46.1 nmol/L (14.5 ng/mL), respectively ( Fig. 2 ). These findings indicated impaired adrenal steroidogenesis and were suggestive of NC-PORD. Genetic testing was performed after obtaining informed consent according to local institutional guidelines. Targeted next-generation sequencing using an adrenal disease gene panel at the Kazusa DNA Research Institute identified compound heterozygous variants in the POR gene ( c.1820A>G [p.Tyr607Cys] and c.1905_1925del [p.Arg636_Val642del]) ( Table 2 ). Based on ACMG/AMP criteria, these variants were classified as pathogenic or likely pathogenic. In the context of the clinical and biochemical findings, the results were compatible with NC-PORD.

Unexplained infertility accompanied by persistent progesterone elevation remains an uncommon initial presentation of NC-PORD. While previous reports have primarily focused on fertility management in patients already diagnosed with PORD [ 7 ], reports in which infertility evaluation itself directly led to the diagnosis remain limited. Our 2 cases highlight that sustained progesterone elevation, particularly when accompanied by relatively low E2 levels and ovarian macrocysts, may represent a clinically useful clue prompting targeted endocrine evaluation, including consideration of NC-PORD. Classic PORD typically presents in infancy with genital abnormalities, Antley-Bixler-like skeletal malformations, and adrenal insufficiency [ 6 , 13 ]. In contrast, NC-PORD lacks these features and is often recognized in adolescence or adulthood during evaluation for amenorrhea or infertility [ 7 ]. Both of our patients were consistent with this adult-onset presentation. Distinguishing NC-PORD from NC-21OHD can be challenging because both forms of CAH may present with infertility and elevations in baseline 17-OHP and progesterone [ 4 ]. However, in NC-21OHD, steroidogenesis downstream of CYP17A1 remains preserved, resulting in normal to increased androgen and E2 production [ 14 ]. Although follicular-phase progesterone elevation may be observed, the overall E2–progesterone balance is often relatively maintained, and ovarian macrocysts, which have been reported as a characteristic feature of PORD, are less commonly observed in NC-21OHD [ 15 ]. In contrast, NC-PORD is characterized by impaired electron transfer from POR to multiple microsomal P450 enzymes, resulting in partial deficiencies across steroidogenic pathways [ 5 ]. Consequently, reflecting reduced CYP17A1 activity, progesterone tends to rise disproportionately relative to 17-OHP. Furthermore, impaired CYP19A1 activity leads to decreased E2 production, creating a hormonal environment characterized by disproportionate progesterone elevation. In addition, this relative E2 deficiency, together with a compensatory increase in follicle-stimulating hormone, may contribute to abnormal folliculogenesis, which can manifest as cystic ovarian changes in some cases [ 7 , 16 ]. These hormonal and morphological features may provide important diagnostic clues in clinical practice. Interestingly, Case 2 also had a history of anal atresia. Although rare, anorectal malformations have been reported in patients with PORD and may represent one of its associated congenital manifestations [ 17 ]. While a causal relationship cannot be established in the present case, this finding may represent an additional clinical clue suggesting impaired POR function. The rapid ACTH stimulation test in our patients supported impairment of multiple adrenal enzymes, demonstrating exaggerated increases in progesterone and 17-OHP with a suboptimal cortisol response. Although a similar pattern may be observed in NC-21OHD [ 7 ], the disproportionate elevation of progesterone relative to 17-OHP, together with reduced E2 levels, was more consistent with impaired multienzyme activity characteristic of NC-PORD. The identified POR variants were located within cofactor-binding regions essential for electron transfer, including the flavin adenine dinucleotide (FAD)- and nicotinamide adenine dinucleotide phosphate (NADPH)-binding domains. Variants in these regions have been associated with impaired electron transfer to microsomal P450 enzymes, leading to partial but broad disruption of adrenal and gonadal steroidogenesis [ 12 , 18 ]. This functional impairment is consistent with the observed hormonal profile, characterized by exaggerated ACTH-stimulated increases in progesterone and 17-OHP with suboptimal cortisol responses. Taken together, these findings support a genotype–phenotype correlation consistent with residual but globally impaired POR activity, explaining the adult-onset presentation and the combination of persistent progesterone elevation, relative E2 deficiency, and infertility. Case 1 also had a history of hormone receptor–positive breast cancer. Although POR plays a role in estrogen biosynthesis and metabolism through its interaction with multiple microsomal P450 enzymes [ 18 ], the relationship between POR deficiency and breast cancer risk remains unclear. Given the limited available evidence, any causal association would be speculative and requires further investigation. Our cases highlight that the reproductive pathophysiology of NC-PORD differs from that of NC-21OHD, with important implications for fertility management. In NC-21OHD, glucocorticoid therapy often restores ovulation by suppressing adrenal androgen excess, and natural conception is frequently achieved [ 19 ]. In contrast, spontaneous conception in NC-PORD appears to be uncommon, with only a limited number of successful pregnancies reported following assisted reproductive technologies [ 7 , 16 , 20‐22 ]. In NC-PORD, persistent progesterone elevation and relative E2 deficiency impair both ovulatory function and endometrial receptivity. Elevated progesterone levels at the time of embryo transfer have been associated with reduced implantation and pregnancy rates [ 23 ], underscoring the clinical relevance of this hormonal imbalance. As observed in our patients, this pattern may persist despite gonadotropin-releasing hormone agonist suppression, suggesting that conventional infertility treatments alone may be insufficient. These findings suggest that management strategies effective for NC-21OHD may not fully address the reproductive challenges in NC-PORD. Limited evidence indicates that combining low-dose glucocorticoids with E2 supplementation may improve endometrial conditions in patients undergoing assisted reproduction [ 16 ]. Consistent with these reports, glucocorticoid therapy in our patients effectively suppressed adrenal progesterone production, resulting in physiological progesterone concentrations before embryo transfer. Although the impact of glucocorticoid therapy on pregnancy outcomes cannot be determined from 2 cases, these observations suggest that appropriate endocrine management may contribute to successful fertility treatment in selected patients with NC-PORD. Several limitations should be acknowledged. CYP21A2 was not included in the targeted gene panel, and therefore, NC-21OHD could not be genetically excluded. In addition, parental genetic testing was not performed, and allelic phasing could not be confirmed. These limitations should be considered when interpreting the genetic findings in this study. Non-classic P450 oxidoreductase deficiency should be considered in patients with unexplained infertility accompanied by persistent progesterone elevation, particularly when associated with relatively low E2 levels and ovarian abnormalities. Early recognition of this endocrine pattern may facilitate timely diagnosis and enable mechanism-based management.