Evaluation of coagulation disorders and iron deficiency in women with heavy menstrual bleeding.

LR designed and analyzed the study and wrote the manuscript; YF, CB, GG, MC, and GD saw the patients and collected the data; and LR and YD participated in the writing. All authors approved the manuscript.

This work was supported in part by Groupe LFB.

During the study period, 454 women were enrolled. Ultrasound exploration, performed for all of them, was normal in 270 cases (59%), and non‐informative in 10 (2.2%). The major gynecologic etiologies retrieved were adenomyosis ( n  = 63, 14%), myoma or fibroma ( n  = 49, 10.8%) and polyps ( n  = 26, 5.7%). Only four cases of endometriosis were disclosed. Previous surgery was reported by 344 women (75.8%), associated with bleeding for 19 (4.2%) of them. The most frequent surgeries were oral ( n  = 156, 5%)—wisdom tooth removal ( n  = 148, 32.5%), tonsillectomy/appendicectomy ( n  = 68, 20%), and gynecologic surgery ( n  = 65, 2%). Sixteen patients (3.5%) had history of cesarean section. For the whole cohort, median Higham and ISTH‐BAT scores were, respectively, 321 (IQR 218–555) and 3 (IQR 2–3). In 416 of the 443 informative cases (94%), HMB was the major and sole hemorrhagic symptom (ISTH BAT ≤5). Blood transfusion (Table  1 ) had been necessary at some time for 20 (4.44%) of the 454 women, among whom 6 (1.3%) had received both blood transfusion and intravenous iron infusion, and 40 (8.8%) the latter only. Oral iron supplementation had been prescribed for 341 (75.1%) of the 454 women and for 136 (30%) of the 454 women tranexamic acid therapy was specified. Treatments of iron deficiency. a Abbreviations: IV, intravenous; PRBCU, packed red blood cell units; VWD, von Willebrand disease. Data are presented as number (percentage). All values of laboratory hemostasis tests were within the normal range, except for one woman for whom, upon observation of a 58% prothrombin time, a moderate FVII deficiency was diagnosed. Notably, no FIX, FXI, or isolated FVIII deficiency was observed. Upon review of medical history, 8 (1.76%) women had a known CBD, 6 (1.3%) with a diagnosis of VWD, the woman with FVII deficiency mentioned above, and a hemophilia carrier. These women are singled out in Table  2 , which reports that they shared most of the characteristics of the whole cohort except, as expected, for hemostasis parameters. Indeed, significantly lower values of FVIII:C ( P  = 0.025), VVF:AG ( P  = 0.010), and VWF:GPIbR ( P  = 0.008) were observed for these women with CBD compared with all others. Patient characteristics comparing women with a known congenital bleeding disorder and those without. a Abbreviations: Ag, antigen; BMI, body mass index (calculated as weight in kilograms divided by the square of height in meters); CRP, C‐reactive protein; FIX:C, Factor IX coagulant activity; FVIII:C, Factor VIII coagulant activity; FXI:C, Factor XI coagulant activity; GPIbR, glycoprotein Ib receptor; Hb, hemoglobin; ISTH‐BAT, International Society on Thrombosis and Hemostasis bleeding assessment tool; IU, International Units; ns, not significant; VWF, von Willebrand factor. Data are presented as number and as median (interquartile range). Patients with measured levels. In the 446 women (98.2%) of unknown CBD status (Table  S1 ), VWF exploration retrieved severe deficiency of VWF (VWF:GPIbR ≤30 IU/dL) in 3 (0.7%), a less severe deficiency (VWF:GPIbR >30 and ≤40 IU/dL) in 17 (3.8%), and a moderate deficit (VWF:GPIbR >40 and ≤50 IU/dL) for 24 (5.4%). Overall, 41 (9.2%) of the 446 women were disclosed as having abnormal levels of VWF. These three groups differed significantly from that of women with normal levels of VWF:GPIbR by their significantly lower coagulation factor levels. At the time of testing, among the 446 women who initially were not known to have a CBD, anemia (hemoglobin ≤120 g/L) was observed in 155 (34.7%) and severe (hemoglobin ≤100 g/L) in 38 (8.5%). Moreover, iron deficiency (ferritin <15 μg/L) was present in 191/446 (42.8%), 256/446 (57.3%) being below 20 μg/L and 324/446 (72.5%) being below 30 μg/L (Figure  2 ). Thirteen (2.9%) women had values above 100 μg/L, indicating slight inflammation, associated with anemia in 3 (0.7%). CRP was recorded as “normal” for 371/452 (82%) patients. For the 81 (17.8%) with reported values, the median was 7.1 mg/L (IQR 4.7–11.1 mg/L), with with reported values, the median was 7.1 mg/L (IQR 4.7–11.1 mg/L), with CRP levels greater than 5 mg/L for 59 patients (72.8% of this population, 13% of the whole informative cohort). Ferritin levels (excluding 13 patients >100 μg/L); The dashed line indicates the 30‐μg/L threshold below which hypoferritinemia is considered. The box and whiskers representation indicates the median and 25–75% centiles, plus minimum and maximum values, excluding extreme values.

In this large French series of women with HMB, a low median ISTH‐BAT score, consistent with literature data, 14 confirmed that most of them did not present CBD. Indeed, coagulation tests revealed normal values in the vast majority of cases. Eight women had a known history of CBD. However, a decrease in VWF activity below 50% was identified in 9.2% of the population of women without known CBD, severe in 0.7%. This incidence is similar to that of VWD in the general population. 4 In the cohort study by Zia et al., 15 33% of adolescents with HMB were diagnosed with a bleeding disorder, including 8.5% with VWD. Brown et al. 16 found 7.5% with VWD while this value was 13% in the Italian study by Seravalli et al., 17 where 113 adolescents were retrospectively recruited. In the cohort reported here, all women benefited from coagulation studies, which appeared to be a good way to detect as yet undiagnosed VWD in women with HMB; their use should be generalized. The other factor examined here was the incidence of anemia in women with HMB. Using the classical hemoglobin threshold of 120 g/L, a rate of 35% of overt anemia was disclosed, an incidence much higher than the current estimation of about 15% in the general population. 18 Moreover, taking into account ferritin levels, a larger number of these women were detected as having iron deficiency. Using the WHO threshold of 15 μg/L, 9 hypoferritinemia was detected in almost 43% of the patients, while using the other recommended threshold of 30 μg/L, this figure rose to nearly 73%. This indicates, in spite of frequent iron supplementation, a latent iron deficiency that can be deleterious in the long term. Such iron deficiency without overt anemia has been addressed in a recent expert review by Munro et al., 19 who report an 18% rate of anemia in women donating blood, even greater in adolescents. In conclusion, this study of consecutive patients confirms and updates the relationships between HMB, CBD, anemia, and iron deficiency. It strongly supports the recently updated recommendations urging better detection and exploration of the huge number of women with this condition.

Heavy menstrual bleeding (HMB) is frequent, with a prevalence estimated between 10% and 30%. 1 This condition, which is liable to lead to health‐compromising comorbidities, 2 remains ill‐diagnosed and ill‐managed. A European survey disclosed that 46% of affected women had never sought medical help for this issue, 3 mostly because of the difficulty for women to recognize the volume of their menses. Sociocultural factors also play a substantial role in women's failure to report HMB, in spite of its impact on health and social life. 4 There is, moreover, a broad misunderstanding about necessary explorations in this context from both women and healthcare professionals. 3 Medical causes leading to HMB have been identified and organized in a classification 5 by FIGO (the International Federation of Gynecology and Obstetrics), called the PALM (Polyps, Adenomyosis, Leiomyomas, and Malignancy or atypical endometrial hyperplasia)‐COEIN (Coagulopathies, Ovulatory disorders, primary Endometrial disorders, Iatrogenic and Not otherwise classified) FIGO system 2. HMB could also be related to coagulation disorders, both congenital (congenital bleeding disorders [CBD]) or acquired. 6 A menorrhagia‐specific screening tool has been proposed by the American College of Obstetricians and Gynecologists that should be used more regularly. 2 It comprises basic coagulation tests, Factor VIII (FVIII) activity and von Willebrand Factor (VWF) antigen and activity measurements, as well as a complete blood count and ferritin assay. 7 Indeed, HMB is the major and sometimes sole symptom of adolescents with CBD, leading to a considerably longer time to diagnosis compared with boys with CBD. 6 Moreover, 70% of women with von Willebrand disease (VWD) present with HMB. 8 Recommendations of the National Institute for Health and Care Excellence (NICE) indicate that complete blood count and investigation of coagulation parameters should be carried out first hand, before confirming, or not, iron deficiency by a ferritin assay. 7 For the latter, the World Health Organization (WHO) 15 μg/L cut‐off to define iron deficiency remains the widely accepted definition of hypoferritinemia in the most recent 2020 guidelines, 9 whereas thresholds of 20 or 30 μg/L are also mentioned, and even 70 μg/L in case of inflammation. 9 In terms of therapy, strategies vary with each patient. Women identified with CBD will be taken care of accordingly (desmopressin, tranexamic acid) but may also, as for those without CBD, require hormone therapy or surgical procedures. 8 , 10 , 11 Here, screening for coagulation and anemia was performed in a large cohort of women referred for HMB in a dedicated center for hemostatic disorders. This allowed for a detailed description of clinical and biological characteristics of HMB and evaluation of the incidence of CBD and iron deficiency in this population.

The authors have no conflicts of interest.

A multidisciplinary platform devoted to HMB has been established in the Gynecology‐Obstetrics department of the Lyon Hospices Civils, France. The present study was approved by the local ethics committee as #23‐5192. All women who consulted in this structure for HMB between January 2022 and January 2024 were enrolled in the present study, in order to collect a large series of parameters. Figure  1 displays the flow chart of these investigations. Besides demographic and basic health data, HMB was evaluated with a questionnaire, allowing calculation of a Higham score, using the first pictorial blood loss assessment charts published by Higham et al. 12 This is a visual chart whereby women collect, over a whole cycle, the number of changes of tampon or pad, evaluate the abundance of bleeding, and mention the potential presence of clots and/or flooding. A score higher than 100 denotes menorrhagia (more than 80 mL of blood loss). Bleeding was also evaluated using the International Society on Thrombosis and Hemostasis bleeding assessment tool (ISTH‐BAT). 13 In this tool, menorrhagia is evaluated on a four‐item scale: (1) consultation only, (2) antifibrinolytics or pill use, (3) curettage or iron therapy, and (4) blood transfusion or replacement therapy, or desmopressin, or hysterectomy. A score higher than 5, using all items in the scoring tool, denotes abnormal bleeding. Flow chart of the study. CBC, complete blood count; ISTH‐BAT, International Society on Thrombosis and Hemostasis bleeding assessment tool. All women benefited from transvaginal or pelvic ultrasound explorations and the results were collected, with details of any anomalies detected. Personal antecedents of surgery and hemorrhagic surgery were investigated as well as previous treatments received, such as hormone therapy, transfusions, iron supplementation, and/or tranexamic acid. Among the biologic parameters recorded, laboratory standard coagulation tests included activated partial thromboplastin time (APTT), prothrombin time, and fibrinogen, that were assayed on an ACL‐Top (Werfen®, Barcelona, Spain). VWF was investigated through antigen and activity measurements using VWF:Ag (HemosIL®, Werfen®; normal range 60–200 IU/dL) and VWF activity, i.e. VWF:GPIbR (HemosIL®, Werfen®; normal range: 50–200 IU/dL for non‐O blood group and 40–200 IU/dL for O blood group individuals). FVIII activity (FVIII:C) was measured using an APTT‐based one‐stage clotting assay (FVIII‐deficient plasma, HemosIL®, Werfen®) on ACL‐Top (Werfen®, normal range: 50–140 IU/dL). Hemoglobin level (normal range 120–160 g/L) and platelet count (normal range 150 × 10 9 –400 × 10 9 /L) were obtained with the hematology analyzer Sysmex XN‐10 (Sysmex Corporation, Kobe, Japan). To complete the assessment of anemia, ferritin and C‐reactive protein (CRP) levels were analyzed on an ECLIA Roche cobas®pro (Roche Diagnostics, Mannheim, Germany) instrument (normal ranges 13–150 μg/L and <5 mg/L, respectively). Continuous variables are expressed as medians and interquartile range (IQR), categorical variables are reported as numbers and percentages. Statistics were performed with Medcalc (Ostend, Belgium) software using Mann–Whitney U tests. Values of P less than 0.05 were considered statistically significant.

Table S1.