Menorrhagia is a serious complication in young female oncology patients who suffer from severe thrombocytopenia during myelosuppressive treatment. To the authors' knowledge, little is known regarding the incidence of this complication or the effectiveness of possible therapies for its prevention.
In this retrospective clinical study, after a thorough gynecologic evaluation, young female oncology patients with regular menstrual cycles undergoing myelosuppressive treatments received either depo-medroxyprogesterone acetate (DMPA), or D-tryptophan-6-luteinizing hormone-releasing hormone depot treatment (gonadotropin-releasing hormone agonist [GnRH-a]), or no treatment before the administration of myelosuppresive chemotherapy. Only patients who later developed severe thrombocytopenia (<25,000 platelets per μL) were included in the study. Daily blood counts, menorrhagia, nonvaginal bleeding episodes, and the need for blood products, gynecologic consultations, and other medical interventions were determined.
Of 101 women with cancer who met the inclusion criteria, 42 patients received DMPA, 39 patients received GnRH-a, and 20 patients remained untreated. The mean duration (± standard deviation) of severe thrombocytopenia was 24.76 ± 23.6 days. Four patients were not included because of significant gynecologic pathologies. General bleeding from nongynecologic sites was similar for all groups and was not associated with vaginal bleeding. Severe or moderate menorrhagia was documented in none of the 39 women who received GnRH-a, in 9 patients (21.4%) who received DMPA, and in 9 untreated patients (40%; P = .02). Fewer calls for urgent gynecologic consultations were documented in the GnRH-a group compared with the untreated group (P < .0001).
High-dose chemotherapy, ablative chemotherapy, or total body irradiation followed by stem cell support or bone marrow transplantation (BMT) are common treatment modalities in young patients who have various malignancies. These treatments effectively destroy the patient's bone marrow until the recovery of blood stem cells. During this time, pancytopenia can result in severe morbidity. Administration of bone marrow growth factors facilitates granulocyte and red blood cell line recovery, improves the quality of life, and allows for the continuation of high-dose chemotherapy.1–3 However, currently, there is no clinically available platelet-specific growth factor.4 Therefore, severe thrombocytopenia after high-dose chemotherapy and subsequent stem cell transfer or BMT poses a major clinical problem in the treatment of these hematology and oncology patients and seriously may complicate their medical management.5 Platelet transfusions remain the standard method of preventing or treating thrombocytopenia. However, platelet resources are limited, and their administration is associated with complications.5, 6 In some patients, transfusions of autologous cryopreserved platelets provide a possible solution for support in severe chemotherapy-associated thrombocytopenia.7
In addition to their destructive effects on the bone marrow, chemotherapeutic agents are toxic to the ovaries in female patients of reproductive age.8 Chemotherapy, thus, can lead to hormone disturbances and dysfunctional uterine bleeding.9 Timely, regular menstruation as well as dysfunctional uterine bleeding during episodes of severe thrombocytopenia can cause menorrhagia with marked blood loss and thrombocyte consumption. Such events can lead to significant discomfort, necessitate administration of blood products, delay or interrupt chemotherapeutic treatments, and jeopardize treatment outcome.
In principle, menorrhagia in thrombocytopenic patients can be treated conservatively with various medications.10 Continuous administration of combined oral contraceptive pills or progestational agents can prevent or postpone bleeding episodes during the thrombocytopenic period. However, gastrointestinal injury, nausea, emesis, and diarrhea associated with the treatment can disrupt the steady intestinal absorption of steroid hormones.11 Intramuscular injection of a long-acting progestative agent, such as depo-medroxyprogesterone acetate (DMPA), has been used for many years for the treatment of excessive vaginal bleeding from various causes.12, 13 In the alternative, long-acting preparations of gonadotropin-releasing hormone agonists (GnRH-a) have been used for the prevention of severe menorrhagia during ablative chemotherapy or radiotherapy treatments.14–16 Medical blockage of the hypothalamic-pituitary axis during GnRH-a treatment decreases gonadotropin levels, reducing gonadal hormone levels to the prepubertal range. Because of steroid hormone depletion, the endometrium is almost unexposed to hormone influence, and menstrual activity ceases.
To date, the effectiveness and potential benefit of these treatments for women undergoing myeloablative therapy have not been compared. The objectives of this study were 2-fold: 1) to assess the incidence and severity of significant menorrhagia in thrombocytopenic women undergoing myeloablative therapy and 2) to compare the individual effectiveness of DMPA and GnRH-a for preventing vaginal hemorrhage during the episode of prolonged severe thrombocytopenia.
MATERIALS AND METHODS
Medical information was collected from the computer data base of a university-affiliated tertiary medical center. Files of hematooncology female patients of reproductive age (range, 12–51 years) who were treated by ablative or high-dose chemotherapy during a 7-year period were studied. Files of patients who, during treatment, developed severe thrombocytopenia (defined as a platelet count <25,000 per μL) were reviewed. Only patients with confirmed ovarian function, as shown by the occurrence of menstrual cycles and the results of endocrine studies, were included in the study. Patients who received hormone therapies (oral contraceptives, hormone replacement therapy, etc.) immediately prior to chemotherapy and patients who had evidence of gynecologic pathology, as demonstrated by pelvic sonography, were excluded from the study.
Gynecologic Evaluation and Treatment Groups
During the study period, female hematooncology patients of reproductive age who were scheduled for high-dose chemotherapy were referred routinely for gynecologic evaluation, which was performed by a single gynecologic team. This included a full medical and gynecologic history with detailed evaluation of menstrual cycles and reproductive performance, pelvic examination, ultrasound evaluation of the reproductive system, and hormone profile of gonadotropin levels on Days 3 through 5 of the cycle. When it was not possible to determine gonadotropin levels because of the menstrual cycle stage, plasma estradiol and progesterone levels were measured. If an intrauterine device was present prior to myelosuppressive therapy, then it was removed. Patients received either no treatment at all or received 1 of 2 hormone-preventive treatments prior to the initiation of chemotherapy according to the protocol that was practiced in the department when patients were admitted. During the study period, the preventive hormone protocol has changed, Initially, women did not receive preventive hormone treatment. Thereafter, all subsequent women received intramuscular injections of DMPA 150 mg (Depoprovera&#142;®; Upjohn Company, Kalamazoo, MI) every 3 months. Finally, patients in the 3rd group were given an intramuscular injection of 3.75 mg of D-tryptophan-6-luteinizing hormone-releasing hormone (Decapeptyl&#142;® CR; Ferring, Kiel, Germany) (GnRH-a) once a month. Hormone treatments were started prior to chemotherapy and were maintained as long as the patient's platelet counts were <100,000 platelets per μL. During the entire study period. the chemotherapeutic protocols remained unchanged.
Daily patient records documented by hematology and oncology department physicians evaluated each patient's general condition, nonvaginal bleeding events, and a daily complete blood count. Each day with a platelet count <25,000 platelets per μL was defined as a “day of severe thrombocytopenia,” and the total number of “days of severe thrombocytopenia” was determined for each patient. If, during an episode of “severe thrombocytopenia,” there were 1 or 2 days of platelet counts >25,000 platelets per μL, then those isolated days of “temporary improvement” also were counted as days of severe thrombocytopenia.
During the study period, a single gynecologic team evaluated the severity and duration of vaginal bleeding episodes in all patients. When patients or Hematology Department staff reported excessive vaginal bleeding, an urgent call for gynecologic consultation was initiated, regardless of the time of day. The on-call physician initiated immediate, unplanned treatments as appropriate. Blood products that were administered as well as any additional drugs were documented meticulously.
Generalized nonvaginal bleeding was scored according to clinical symptoms: 1) no bleeding; 2) mild bleeding—mucosal bleeding or skin hemorrhages; 3) moderate bleeding—bleeding from major organ systems (e.g., cystitis or melena); or 4) severe bleeding—life-threatening bleeding episodes. Vaginal bleeding severity also was scored according to 4 categories: 1) no vaginal bleeding; 2) spotting-to-mild vaginal bleeding that did not necessitate intervention; 3) moderate vaginal bleeding—daily blood loss comparable to a patient's regular menstruation but requiring medical intervention because of its duration; or 4) severe vaginal bleeding—significant bleeding with blood loss exceeding the patient's regular menstruation and requiring medical or surgical intervention.
The use of GnRH-a treatment was approved by the institutional ethical committee. The use of DMPA for these indications was the standard of care at the time of treatment.
The distributions of the variables, the number of days of thrombocytopenia, and the units of blood products administered showed a markedly skewed distribution; therefore, nonparametric methods were used in the analysis. First, a Kruskal–Wallis 1-way analysis of variance (ANOVA) was performed as an overall test of homogeneity of the mean levels. Whenever the ANOVA was statistically significant at the .05 level, paired comparisons were made by using the Mann–Whitney test. The distribution symmetry allowed for parametric 1-way ANOVA. Then, multiple comparisons were made by using the Student–Newman–Keuls method. The relations between 3 outcome variables—general bleeding tendency, amount of vaginal bleeding, and number of gynecologist visits—were tested by using the chi-square test.
In total, 101 patients with regular menstrual cycles and with documented normal ovarian function before the administration of high or ablative doses of chemotherapy developed severe thrombocytopenia during treatment (<25,000 platelets per μL). Four patients were excluded because of the presence of gynecologic pathology. One patient who had an endometrial polyp and another patient who had endometrial hyperplasia underwent dilatation and curettage before the administration of chemotherapy. A 3rd patient suffered from irregular, excessive uterine bleeding (menometrorrhgia) before chemotherapy and had a 16-cm, enlarged uterus with large fibroids. She underwent abdominal hysterectomy before BMT. In 1 patient, an early pregnancy was diagnosed during evaluation; it was terminated before chemotherapy was started.
The mean age (± standard deviation) of the 101 patients who met all enrollment criteria was 29.82 ± 8.8 years (range, 12–51 years). Among them, 93 patients (92.1%) received ablative chemotherapy followed by BMT, and 8 patients (7.9%) were treated on high-dose, nonablative chemotherapy regimens. Twenty patients (19.8%) received no hormone treatment, 42 patients (41.6%) received DMPA, and 39 patients (38.7%) received GnRH-a. The ages and chemotherapy treatment protocols were similar for the 3 groups (Table 1). The average number of days of severe thrombocytopenia was 24.76 ± 23.6 days (range, 3–128 days) (Fig. 1).
Table 1. Characteristics of 101 Patients with Thrombocytopenia Treated for Hematologic and Oncologic Malignancies in the 3 Treatment Groups
The 3 groups were comparable statistically with regard to patient age, diagnosis, and chemotherapy regimen. There was no significant difference between the groups with respect to the duration of thrombocytopenia: 27.7 ± 28.2 days for the untreated group, 25.9 ± 24.8 for the DMPA group, and 22.15 ± 20 days for the GnRH-a group (P = .6) (Table 1). There was no difference in the occurrence of moderate-to-severe nonvaginal general bleeding among the 3 groups: 20% of untreated patients (4 patients), 21% of patients in the DMPA group (9 patients), and 21% of patients in the GnRH-a group (8 patients). The average number of platelet units, packed cell units, and units of fresh-frozen plasma used throughout the treatment period also did not differ significantly. In the untreated group, more patients received tranexamic acid for longer periods than in the 2 treatment groups; however, this difference was not statistically significant: The data were incomplete for 1 patient; therefore, the results of general bleeding episodes were evaluated for 100 patients (Table 2).
Table 2. Evaluation of the Amount of Generalized (Nonvaginal) Bleeding and Clinical and Therapeutic Data of 100 Patients
Days of treatment with tranexamic acid ± SD (no. of patients)
10.8 ± 14 (11)
6 ± 9 (20)
6.5 ± 9.9 (15)
7.2 ± 10.6 (46)
Moderate and severe vaginal bleeding was observed in 16.8% of patients in our study. Both treatment groups had a lower rate of menorrhagia than the untreated group, and none of the patients of the GnRH-a group had moderate or severe vaginal bleeding (Table 3). These differences were statistically significant (P = .002).
Table 3. Severity of Vaginal Bleeding During the Period of Thrombocytopenia in the 3 Treatment Groups
No. of patients (%)
DMPA indicates depo-medroxyprogesterone acetate; GnRH-a, gonadotropin-releasing hormone agonist (a D-tryptophan-6 lutenizing hormone-releasing hormone); SD, standard deviation.
Bleeding that did not necessitate intervention.
Bleeding that was comparable to regular menstruation and necessitated medical intervention.
Bleeding that was greater than the regular menstruation required intervention.
Differences between the groups also were observed regarding the need for urgent gynecologic consultation because of vaginal bleeding during the thrombocytopenic episode (P < .0001) (Fig. 2). In the untreated group, 65% of patients required at least 1 gynecologic consultation, and 30.1% of patients in the DMPA group had a gynecologic consultation. However, in the GnRH-a group, only 3 patients had a gynecologic consultation because of mild vaginal bleeding, and all 3 of those patients merely needed reassurance without any medical intervention. Urgent treatment with conjugated estrogens to stop severe menorrhagia was indicated in the untreated and DMPA group, but no treatment was needed in the GnRH-a group (P = .004) (Table 3).
We did not observe any association between generalized nonvaginal bleeding events and episodes of vaginal bleeding (P = .1). Only 1 of 100 patients suffered from moderate-to-severe general bleeding with concurrent moderate-to-severe vaginal bleeding (Table 4).
Table 4. The Correlation between Vaginal Bleeding and Generalized Bleeding*
Generalized bleeding: No. of patients (%)
P = .1.
Data describing 1 patient's tendency to bleed were missing.
Severe, prolonged thrombocytopenia caused by ablative or high-dose chemotherapy treatments in young female cancer patients poses a specific significant clinical problem. Disruption of the physiologic processes of the normal menstrual cycle can result in severe menorrhagia, further increasing morbidity and complicating the already fragile state of these patients. The results of the current study indicate that, when chemotherapy-induced thrombocytopenia occurs in female patients with normal ovarian function, up to 40% of those patients experience episodes of either severe or moderate menorrhagia. Therefore, prevention of significant menorrhagia is of the utmost importance.
The feasibility and effectiveness of medical treatments to prevent abnormal uterine bleeding secondary to hematologic changes after chemotherapy have not been investigated thoroughly. For the current study, examined and compared the efficacy of 2 treatment modalities: DMPA and GnRH-a.
Several earlier studies demonstrated the efficacy and benefits of DMPA as a contraceptive agent.17–19 In addition to its contraceptive effects, DMPA administration usually results in amenorrhea.19 Therefore DMPA (a relatively inexpensive drug) has been used over the last decades to achieve therapeutic amenorrhea in the management of a variety of uterine, endocrine, and hematologic dysfunctions that can lead to abnormal uterine bleeding.12, 13, 20–22 However, some patients who receive DMPA experience irregular uterine bleeding.23 Such irregular bleeding patterns may complicate the tendency for uncontrolled bleeding further in thrombocytopenic patients.
GnRH-a administration significantly decreases pituitary secretion of gonadotropins. This results in the arrest of cyclic ovarian function and a significant reduction in ovarian steroid hormone production. Therefore, the uterine endometrium is almost unexposed to any hormone influence, resulting in amenorrhea. These changes decrease the likelihood and severity of abnormal uterine bleeding. However, GnRH-a may cause an initial flare-up effect (i.e., a rise in pituitary gonadotropins and ovarian hormones), which can lead to the development of ovarian cysts.24 This unwanted effect can be prevented by the use of a GnRH antagonist. It is possible that, in the future, the use of GnRH antagonists may result in effects on ovarian function similar to the effects of GnRH-a without the initial flare-up effect.25–27 Currently, long-acting GnRH antagonists are in the development stage.
All patients included in the current study were women who were menstruating regularly and who were treated for various malignancies. All were free of any gynecologic abnormalities, as assessed by clinical evaluation, hormone measurements, and sonography. All patients suffered from severe thrombocytopenia. Table 1 indicates that the 3 groups were comparable with no significant differences in patient age, the duration of thrombocytopenia, or the severity of nonvaginal bleeding.
In our study, both GnRH-a and DMPA decreased the prevalence of menorrhagia and minimized symptoms in young female cancer patients during episodes of severe thrombocytopenia. However, GnRH-a was far more effective than DMPA in these patients. Differences in response rates among the 3 patient groups—an untreated group, 1 group that received a GnRH-a regimen, and 1 group that received DMPA—were highly significant, as measured by days of excess bleeding, the need for emergency administration of intravenous estrogens, and the need for urgent gynecologic consultations. None of the patients who received GnRH-a developed moderate or severe menorrhagia, whereas 40% of untreated patients and 21.4% of patients who received with DMPA experienced moderate or severe menorrhagia.
Our results demonstrate that there is no correlation between severe menorrhagia and a general bleeding tendency in other sites. Of the 21 patients who had episodes of moderate-to-severe, generalized bleeding, only 1 woman complained of concomitant moderate-to-severe vaginal bleeding (Table 4). Furthermore, of 79 women without significant general bleeding, 16 patients (20.3%) experienced moderate-to-severe menorrhagia (P = .1). Adequate control of endometrial bleeding depends on hormone and local factors that affect the local clot-forming mechanisms. Therefore, treatment to prevent menorrhagia must be added to conventional regimens that are used for thrombocytopenic patients irrespective of the tendency for generalized bleeding. Acute menorrhagia, especially in young, anovulatory women, also can be treated by high-dose estrogens followed later by progestagens.13 It should be borne in mind that women with oncologic disease are at increased risk for thromboembolic phenomena,28, 29 which may be triggered or aggravated by the thromboembolic potential of estrogens.
Blood products are administered to thrombocytopenic patients to compensate for deficient formation of all cell lines and for excessive bleeding. We could not demonstrate significant differences in the number of blood product units administered during follow-up in the current study. The need for platelets, packed cells, and fresh-frozen plasma is based on several factors, including bone marrow production, the patient's general condition, and bleeding episodes—vaginal or nonvaginal. In addition, standard care of these patients includes the prophylactic transfusion of blood products and the aggressive replacement of red blood cells and platelets. All of these factors may have contributed to the similar numbers of blood product units used in each of the groups.
Several small, descriptive, noncomparative studies have demonstrated that the administration of a GnRH-a was safe and effective for preventing vaginal bleeding in women who were undergoing myelosuppressive chemotherapy or radiotherapy.15, 30, 31 Ghalie et al.14 reduced the occurrence of hypermenorrhea in premenopausal women who were undergoing BMT by giving daily injections of leuprolide, a GnRH-a. Those authors stressed the importance of early treatment with GnRH-a, preferably 2 weeks before the administration of chemotherapy. Chiusolo et al. concluded that the GnRH-a leuprolide was highly effective in preventing uterine bleeding in premenopausal women undergoing stem cell transplantation and that it had an excellent toxicity profile and no interaction with hemostatic balance or hepatic function.31 However, the few positive reports on the use of a GnRH-a did not compare it with other therapies. The current results indicate the efficacy of GnRH-a and demonstrate its clinical superiority over DMPA. The apparent clinical advantage is important, because GnRH-a can cause menopausal symptoms and usually is much more expensive than DMPA.
The current study also demonstrated the importance of a routine gynecologic evaluation, including pelvic sonography, for all female cancer patients before the initiation of cytotoxic therapy. Uterine anomalies and unrecognized pregnancies that may have complicated the course of therapy were diagnosed and treated promptly in 4 patients who ultimately were excluded from this study. Our clinical treatment protocol also includes the removal of intrauterine devices, because we believe that their removal reduces the risk of vaginal bleeding and pelvic infection during an episode of pancytopenia.
In this study, we demonstrated the advantage of hormone treatment prior to the administration of high-dose chemotherapy in young female cancer patients who have functioning ovaries and periodic menstruation. Furthermore, the study demonstrated a clinical advantage of GnRH-a over DMPA in preventing and lessening the severity of chemotherapy-associated menorrhagia when it is administered before initiation of the cytotoxic regimen. Because there is no correlation between the occurrence and severity of abnormal uterine bleeding and the general tendency to bleed, it is not feasible to identify patients prospectively who are at higher risk for menorrhagia and administer hormone treatment to a subset of young female oncology patients. Based on these data, we recommend the administration of a GnRH-a to all young cancer patients with regular cycles who undergo high-dose chemotherapy that is likely to cause severe thrombocytopenia.