Abnormal Uterine Bleeding in Young Women with Blood Disorders.

Abnormal uterine bleeding (AUB) is common in adolescents 1 . Immaturity of the hypothalamic pituitary ovarian (HPO) axis is the most common cause of AUB in this age group 2 . Certain aspects of underlying inherited or acquired blood disorders, as discussed in this chapter, exacerbate the “expected” hormonal imbalance at this age, thereby, increasing the morbidity of the underlying problem 3 , 4 . Even though blood disorders may induce AUB, uterine structural and/or endocrine abnormalities tend to be overlooked in the presence of a blood disorder 5 . A multifactorial etiology demands a collaborative approach between hematologists and gynecologists or adolescent medicine physicians 6 , 7 . In this chapter, we will discuss management of AUB in adolescents within 4 clinical contexts: AUB while on anticoagulant therapy, and with inherited bleeding disorders, bleeding management with cytopenias, specifically, thrombocytopenia, and in sickle cell disease. Throughout, areas of controversy and opportunities for further research are highlighted. There has been a long-standing confusion concerning terminologies and definitions surrounding female reproductive tract bleeding 5 . ‘Menorrhagia’, a loosely defined term is used for heavy menstrual bleeding (HMB). ‘Dysfunctional uterine bleeding’ is associated with adolescents and often implies anovulatory bleeding. We will use the International Federation of Gynecology and Obstetrics terminologies and classification system throughout this chapter, and strongly recommend them for use in every day practice 5 . These have been published after years of robust international cooperation and consensus forming in 2011 5 . AUB or HMB are the preferred overarching terms, and the work up of HMB or AUB proceeds within the realm of PALM-COEIN classification, irrespective of age. PALM indicates structural causes : polyp, adenomyosis, leiomyoma, and malignancy, and COEIN indicates non-structural causes : coagulopathy, ovulatory dysfunction, endometrial, iatrogenic, and not yet classified 5 . HMB is defined as periods lasting more than seven days, soaking through a pad or tampon under two hours or soaking through bed clothes, passing clots and ferritin below normal limits or anemia 8 . AUB refers to any departure from normal uterine bleeding, either in volume or duration of flow, regularity or frequency, and in theory, encompasses HMB. The number of adolescents in need of antithrombotic therapy has gradually increased over the past decades 9 . Anticoagulation is associated with AUB in 20–70% of patients 10 . One retrospective study of 90 women aged 15–49 years found that frequency of HMB increased from 17.8% before anticoagulation to 29.5% thereafter 11 . Another consideration is that heavy periods, often the reason for hormone induced VTE in the first place, are associated with stopping hormonal therapy may be intensified by anticoagulants. Estrogens are known to increase the risk for VTE twofold to fourfold in a dose-dependent way 12 , 13 . The risk is also increased with third generation when compared with the first or second-generation progestins 13 . Patients with hormone-induced VTE are instructed to discontinue them, generally at the time of diagnosis. There is controversy regarding whether discontinuation should be immediate upon diagnosis of the VTE or deferred to the time point of discontinuation of anticoagulant therapy. The former recommendation is found in a World Health Organization (WHO) publication from 2010, stating that estrogen-containing oral contraceptives (OCs) should not be used, even on established oral anticoagulation 14 . In contrast, the Scientific and Standardization Committee of the International Society on Thrombosis and Haemostasis (ISTH) in a guidance document, suggested, hormonal therapy can be continued in selected patients but anticoagulants should be continued for the duration of hormonal therapy 15 . This is based on the premise that the anticoagulant effect trumps the thrombogenic potential of estrogens. Key implications in this clinical context, therefore, include: a) the risk of recurrence with hormones while on anticoagulation, and b) medication options for AUB control and pregnancy prevention while on anticoagulants. The risk of continuing hormonal therapy in women younger than 60 years of age (mean age 41.3 years) with VTE, was investigated in those treated with either lowmolecular-weight heparin followed by vitamin K antagonists (VKA) or with a direct oral anticoagulant (DOAC), Rivaroxaban, in one landmark study; 1888 women were analyzed, evenly distributed between the 2 anticoagulant regimens 16 . Of those, 475 women were taking concomitant hormonal treatment, either remaining on it despite the VTE diagnosis (n= 402) or initiated during the study period (n=73). The key finding of this study was the similar incidence rate during the time periods at risk, among those on hormonal therapy (3.7% per year) and those off/without hormonal therapy (4.7% per year), even after adjustment for age, prior hormonal therapy, assigned anticoagulant treatment, and cancer at baseline. In addition, there was no difference in crude incidence densities between women taking estrogen vs. progestin (3.7% per year and 3.8% per year, respectively), suggesting that the patient may continue with the type of hormones of her choice while on anticoagulation. Only 7 VTE events occurred on hormonal therapy (4 on estrogen-containing and 3 on progestin-only therapy) 16 . Despite these findings, there is reluctance among physicians to prescribe estrogen-containing OCs to women on anticoagulation. Moreover, there are no data to guide such management decisions in adolescents with AUB on anticoagulation. A recent systematic review highlighted that anticoagulant dose reduction or temporal cessation for controlling AUB while on anticoagulation was associated with a higher risk of recurrent VTE, though this was based on findings from only one study in this systematic review 10 . Treatment of AUB/HMB in adolescents taking anticoagulants, like in older women, depends on comorbidities, hormonal imbalance, either due to immaturity of the HPO axis or polycystic ovarian syndrome, need for contraception and rarely, the presence of structural uterine changes. Other general principles, specific to adolescents, are discussed in Table 1 . In women of childbearing potential who receive VKAs (e.g. Warfarin), adequate contraception is required because these drugs cross the placenta, potentially leading to bleeding in the fetus and/or severe embryopathy 17 . The DOACs, currently in phase III trials in pediatrics, have as yet unknown adverse effects on fetal development. 18 Even without the teratogenic effects, anticoagulants in therapeutic doses increase the length and severity of menstrual bleeding, which often requires medical management. In adolescents with AUB on anticoagulants, options such as non-steroidal anti-inflammatory drugs (NSAIDs) and estrogen containing OCs, as discussed above, are either contraindicated or not used. A recent systematic review shows that insertion of a progestin-releasing intrauterine device seems to decrease the length and severity of menses in most women on anticoagulation. The investigators of this review also acquired the clinical judgement of two panels of key opinion leaders (10 international AUB and 10 thrombosis experts) and compared expert opinion to current practice patterns by surveying ~300 international physicians (hematology, gynecology, internal medicine cardiology, pulmonology etc.). The expert recommendations were divergent and different in several important points from actual clinical practice, which was quite heterogenous. Only the following aspects of combined expert opinion analyses met the predefined threshold of ≥ 70% agreement: 1) hormonal contraception was not judged to be a risk factor for recurrent VTE during active/on-going anticoagulation; 2) progestin-only pill and progestin releasing IUD were suggested to be best options for patients wishing to initiate contraception on anticoagulants; 3)tranexamic acid was suggested to be useful for treatment of anticoagulant associated AUB, and 4)anticoagulant dose reduction and inferior vena cava filter placement were not recommended management options for AUB, especially not in the first month following a VTE. Invasive approaches such as endometrial ablation are an alternative to hysterectomy in older women with refractory or life-threatening HMB on anticoagulation 19 . However, as opposed to the “general” principles of AUB management in older women that conceivably can be applied to young women with AUB on anticoagulants, invasive management approaches require careful consideration given the lack of recent or any data, respectively. To summarize, key management points include: Women of reproductive age may suffer from HMB when treated with anticoagulants. Adequate information regarding this risk should be given to the patient when initiating anticoagulation. Women taking oral anticoagulants for VTE may use estrogen or progestin hormonal therapy to control menstrual bleeding without increased risk for recurrent VTE. 16 For women with HMB on anticoagulant therapy, tranexamic acid is an alternative to hormonal therapy to reduce the bleeding. The latter does not provide a contraceptive effect. The thrombogenic effect of estrogens does not disappear precipitously upon discontinuation of estrogen containing OCs. The WHO Collaborative Study had noted that the effect disappeared 3 months after stopping OCs 13 . It therefore seems reasonable to recommend discontinuing OCs or oral estrogen substitution at least 1 month before planned discontinuation of anticoagulation rather than at the same time. A collaborative, multidisciplinary approach between thrombosis and AUB experts is paramount for effective health care delivery. During anticoagulation, adequate measures to control AUB and prevent pregnancy, and reproductive toxicity to a fetus are indicated for all women with reproductive potential, including adolescents. Research on this subject is desperately needed to guide clinical practice in young women. Menstruation disrupts blood vessels, restoration of which requires an intact hemostatic system and successful interaction of platelets, clotting factors and fibrinolytic proteins 20 . Although less likely to present with spontaneous bleeding, inherited bleeding disorders (BDs) can become particularly severe after hemostatic challenges such as surgery, trauma, menses or childbirth. Mounting evidence suggests that 10–62% adolescents with HMB may have an underlying BD ( Table 2 ) 7 . The average age of women identified with an underlying BD, however, is 35 years. In other words, the diagnosis of a BD is a relatively late one. Wide ranges of reported prevalence, difficulty in discerning normal menstrual bleeding from HMB and the semi-empiric use of hormonal therapy makes BDs in adolescents challenging to identify 21 . There have been awareness efforts by the American College of Obstetricians and Gynecologists (ACOG) in 2001, and in 2006, by the American Academy of Pediatrics (AAP) in collaboration with ACOG advising that hematologic disorders (particularly, von Willebrand disease (VWD)) be considered in HMB, especially at menarche 22 , 23 . VWD is reviewed in “Von Willebrand disease: Diagnostic Strategies and Treatment Options” by Drs. Ng and DiPaola in this issue. However, despite awareness efforts, BDs remain under-diagnosed in women 7 . Historically, clinical questions that discriminate individuals with BDs from those without have been bleeding following hemostatic challenges such as tonsillectomy or dental extraction and the presence of a BD in the family; however, such exposures are often absent in the pediatrics. Generic bleeding assessment tools (BATs) that quantify bleeding symptoms in adults improve diagnostic accuracy and predict bleeding phenotype for a variety of BDs 24 – 26 . The ISTH BAT was specifically designed to incorporate pediatric specific bleeding symptoms, including frequency and severity of bleeding 27 . Even though the ISTH BAT can discriminate between no BD and a possible BD with acceptable accuracy, it has not been tested in adolescents 28 . The pictorial blood assessment chart (PBAC) is a semi-quantitative method and allows women to track the number of pads or tampons during a menstrual period, and the degree of soiling 29 . A score is generated based on that information, and PBAC scores of ≥ 100 correlate with ≥ 80 mL of menstrual blood loss, the classic definition of HMB. The sensitivity and specificity of the PBAC is 86% and 89% respectively in adult women with HMB 30 , 31 . PBAC was evaluated in adolescents 32 . The mean PBAC score for the entire cohort was 195, but notably different when three groups were analyzed separately; 362, 136 and 44 for subjects in heavy, normal or light periods group respectively. Twenty percent of subjects in the ‘heavy group’ were diagnosed with a BD while there were no patients with a BD in the other two groups. Although analysis of diagnostic utility was not performed, this is the first study in adolescents investigating PBAC 32 . More recently, another HMB specific BAT called the Phillip screening tool was evaluated to screen for BDs in women aged 18–50 years with PBAC score > 100 and a normal pelvic ultrasound 33 . This tool has a sensitivity of 89% for a BD. The sensitivity was lower (62%) for the 25 adolescents included in the study. The overall sensitivity improved to 93% by adding iron deficiency, and to 95%, when the PBAC score was increased to > 185 34 . The utility of this tool is yet to be studied in an exclusive adolescent population. The initial laboratory evaluation of patients with a suspected BD should start with a complete blood count, review of peripheral blood smear for platelet morphology, prothrombin time, partial thromboplastin time, and optionally either fibrinogen or thrombin time 35 . The ACOG and AAP recommend testing for VWD in adolescents with HMB, especially at menarche 23 , 36 . Ideally testing should be repeated at least twice for those with a high suspicion of a BD, and subnormal results should be confirmed 35 . Testing on OCs containing 30–35 μg of estrogen does not affect the laboratory diagnosis of VWD 37 , 38 . However, testing should not be performed during an OC taper or after high dose estrogen pulses because VWF levels are elevated with OCs containing ≥ 50 μg of estrogen 39 . If these tests are normal, further testing should include screening for qualitative platelet disorders with a platelet aggregation and secretion study 40 , 41 . The authors propose a management algorithm, including laboratory testing ( Figure 1 ) developed in the context of multidisciplinary management of young women with HMB 6 . Recommended gynecological testing for HMB includes screening for pregnancy, and chlamydia trachomatis and Neisseria gonorrhea in this age group 42 . A history of irregular bleeding prior to the onset of heavy bleeding should prompt hormone testing to screen for polycystic ovarian syndrome (PCOS) and hypothyroidism 2 , 43 . The role of pelvic examination is debated but an external genital exam is generally undertaken to detect anatomical abnormalities 1 , 44 . Pelvic ultrasonography is generally performed in cases that are refractory to therapy 45 , 46 . A multidisciplinary approach to management, with involvement from a hematologist, an adolescent medicine specialist and/or gynecologist results in optimal treatment outcomes 47 , 48 . OCs, the prototype for other combined hormonal contraceptives, offer both contraceptive and non-contraceptive benefits in adolescents with HMB, and include reduction of menstrual flow 49 , 50 . The 2008 NHLBI VWD treatment guidelines recommend OCs as the first choice for HMB in an adolescent who does not desire pregnancy but may desire future child bearing 51 . Effectiveness of OCs in adolescents with platelet function defects remains to be studied. Extended cycling of OCs to reduce withdrawal bleeding that would otherwise occur every 28 days with conventional OCs is promising 52 . Studies in adult women have demonstrated that menstruating less frequently than monthly is highly acceptable 52 . Use of progesterone-only hormonal therapy, such as progesterone only pills, medroxyprogesterone acetate (MPA) and the progestin implant, also may reduce menstrual flow and are another alternative to OCs in adolescents with HMB but seem less preferable due to weight gain 49 . Increasingly used in adolescents, the levonorgestrel intra-uterine system (LNG-IUS) steadily releases 20 mcg of levonorgesterel per 24 hours for 5 years into the endometrial cavity 53 . IUDs are safe in nulliparous adolescents but require the expertise of an experienced gynecologist 54 . LNG-IUS was shown to be an effective option in adolescents in a number of BDs with minimal complications, high compliance rate, and improvement in HMB and anemia 50 , 55 . Antifibrinolytics decrease the fibrinolytic activity observed in the endometrial tissue of women with HMB 56 . Two antifibrinolytics, epsilon aminocaproic acid (Amicar) and tranexamaic acid (TA) have a well-established safety profile 57 and both reduce menstrual flow and improved QOL among females with HMB and abnormal laboratory hemostasis 58 . Data on the use of NSAIDs for management of HMB in adolescents are non-existent. Expert guidelines suggest they should not be used as first-line therapy for HMB, because they affect platelet function and may further increase menstrual blood loss in patients with undiagnosed BDs 59 . Because menstrual pain is common in adolescents with BDs, the use of NSAIDs in those with severe menstrual cramps with hemostatic therapy deserve further study. In contrast to BDs, such as severe hemophilia reviewed in Evolving Complexity in Hemophilia Management by Dr. Croteau, in this issue, in which prophylaxis is the accepted standard of care, factor replacement such as VWF replacement for VWD, is less frequently used in women with HMB and is another area in need of investigation. The need for surgery in adolescents with HMB is based on the clinical stability of the patient. Examination under general anesthesia, evacuation of blood clots from the uterine cavity and curettage should be considered, for serious or refractory menstrual bleeding. Endometrial ablation is not considered in adolescents due to concerns for future fertility preservation 60 . Endometrial balloon tamponade using a Foley catheter may provide a temporary measure to control bleeding in a hemodynamically unstable patient 61 . Readers are encouraged to refer to other reviews for management protocols for acute HMB on an inpatient and outpatient basis 6 , 59 . There are many conditions in childhood that feature thrombocytopenia, of which those with moderate to severe thrombocytopenia may likely experience AUB. Thrombocytopenia can arise from a production problem, such as that associated with inherited bone marrow failure syndromes (iBMF), idiopathic aplastic anemia (iAA)(which is reviewed in chapter 12 by Savasan), and chemotherapy-induced thrombocytopenia (CIT), or because of platelet destruction such as with immune thrombocytopenia (ITP). The treatment of AUB in patients with thrombocytopenia should always include the treatment of the underlying disorder ( Table 3 ), but there are some unique differences in the approach to those with impaired production vs. increased destruction. The most logical therapy for platelet production disorders may appear to be platelet transfusion, but given the risk of alloimmunization and transfusion-related complications, it may be desirable to pursue adjunctive therapies, especially in disorders that may require future hematopoietic stem cell transplant (HSCT) or in which one might choose to reserve platelet transfusion for life-threatening bleeding. Other therapeutic modalities include hormonal and hemostatic agents. There is little to no evidence in iBMF and iAA for the use of antifibrinolytics such as amicar and TA 62 . Induction of amenorrhea with hormone based therapies or leutenizing-hormone releasing hormone agonist (GnRH-a) or MPA may be useful. In conditions in which it is not expected for the production to improve or when the thrombocytopenia is moderate, one could consider the use of an intrauterine device (IUD) for induction of amenorrhea 63 . Desmopressin Acetate (DDAVP) can also increase the release of VWF making what few platelets are in circulation more effective, but given the limited literature in these disorders, this should be considered when other methods are not effective 64 . Chemotherapy induced thrombocytopenia often leads to a transient window of risk of AUB. Platelet transfusion is recommended prophylactically when platelet count falls below 10,000/mm3 65 .The incidence of moderate to severe HMB is 40% in patients receiving ablative chemotherapy for HSCT 63 . Induction of amenorrhea is commonly pursued in oncology and HSCT patients as the severity and duration of thrombocytopenia is expected to be quite long and the impact of chemotherapy on ovaries may additionally lead to AUB due to HPO-axis disruption. Scarce evidence exists for Amicar and TA in patients undergoing chemotherapy in conjunction with platelet transfusion for the treatment of abnormal bleeding, but this is not specific to menstrual bleeding 62 . GnRH-a and MPA are commonly employed in advance of conditioning chemotherapy +/− radiation 63 , 66 , with the recommendation that GnRH-a is more effective than MPA for menstrual suppression and reduction of HMB 63 , 67 . Due to the risk of VTE in cancer patients, providers may choose to avoid estrogen-based hormonal therapy. Additionally, whether a foreign body such as an IUD serves as a nidus of infection in patients with prolonged immunosuppression is controversial and no studies have been done to determine the exact risk. In ITP, platelet transfusion should be reserved for life threatening bleeding when all the mainstays of ITP therapy have been exhausted. In persistent or chronic ITP, it is important to consider the use of hormonal suppression of menses or slowing of menstrual flow with the use of antifibrinolytic agents 62 . For history of HMB prior to the diagnosis of ITP, it is important to exclude VWD and qualitative platelet dysfunction with a thorough hemostatic evaluation 68 . The gynecological issues in adolescents with SCD include a later onset of menarche, AUB and vaso-occlusive pain associated with menstrual cycles 69 . Menstrual patterns are reported to be normal 70 . An important feature, exclusive to women, is the association of SCD related pain with menstrual cycles 70 . Readers should refer to chapter 2 focused on current therapies for women with SCD. While hormonal fluctuations seem to underlie the painful SCD episodes associated with menses 70 , there is a paucity of data investigating HMB in this group of adolescents. Conceivably, etiological factors underlying HMB in non-SCD patients should also be present in SCD patients. It is possible that mild BDs are mitigated by the SCD related hypercoagulability but a more pronounced bleeding phenotype is present in those who experience HMB and manifests with the monthly challenge of menstruation. Adolescents with SCD and HMB are also at higher risk of iron deficiency anemia that may be masked in the presence of the underlying chronic anemia. This is especially true for the patients on hydroxycarbamide whose higher mean corpuscular volume (MCV) 71 masks the microcytosis characteristic of iron deficiency. Low ferritin levels, the hallmark of iron deficiency anemia 72 , are not present in SCD due to the underlying inflammatory state, making diagnosis difficult. AUB in adolescents with SCD, like other blood disorders, requires a systematic approach that begins with a thorough menstrual history. A careful consideration of other blood disorders, such as underlying VWD, systemic medical illnesses like hypothyroidism, PCOS or structural uterine abnormalities is paramount. Iron deficiency should be screened with ferritin level; though, caution should be exercised with interpretation. A better marker of iron deficiency in SCD may be percent transferrin saturation of less than 25% 73 . There is reluctance to use estrogen containing OCs for hormonal suppression in an already hypercoagulable state, however, WHO guidelines on contraceptives in SCD, also endorsed by ACOG, state that the benefits of combined injectable contraceptives, low-dose OCs, and IUDs outweigh the risks associated with the increased morbidity and mortality associated with pregnancy 74 . Additionally, progesterone-only pills can be used without restriction 14 . It should be noted that data are based on the high unplanned pregnancy risk rather than the safety profile of these pills 75 . Longitudinal, prospective studies on etiology of AUB, uses and complications of contraception, risk of unplanned pregnancy and prevalence of other reproductive issues are urgently needed for adolescents with SCD.

Several issues for adolescents with AUB remain to be studied. A correct appreciation of the importance of AUB and its management are skills that all peditricians should strive to achieve. It is quite clear that failure to do so could result in suffering and risk of serious bleeding. Despite advances in the clinical investigation and management of adult women with AUB and blood disorders, several areas need to be studied in adolescents as covered in the previous sections. These gaps urgently require future research and should be filled in the next years by multicenter collaborative studies.