- Email: [email protected]
Predictive factors for the methotrexate treatment outcome in ectopic pregnancy: A comparative study of 400 cases
Accepted Manuscript Title: Predictive factors for the methotrexate treatment outcome in ectopic pregnancy: a comparative study of 400 cases Author: Lucie Bonin C´ecile Pedreiro St´ephanie Moret Gautier Chene Pascal Gaucherand G´ery Lamblin PII: DOI: Reference:
S0301-2115(16)31027-2 http://dx.doi.org/doi:10.1016/j.ejogrb.2016.11.016 EURO 9687
To appear in:
EURO
Received date: Revised date: Accepted date:
12-7-2016 8-10-2016 15-11-2016
Please cite this article as: Bonin Lucie, Pedreiro C´ecile, Moret St´ephanie, Chene Gautier, Gaucherand Pascal, Lamblin G´ery.Predictive factors for the methotrexate treatment outcome in ectopic pregnancy: a comparative study of 400 cases.European Journal of Obstetrics and Gynecology and Reproductive Biology http://dx.doi.org/10.1016/j.ejogrb.2016.11.016 This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
Predictive factors for the methotrexate treatment outcome in ectopic pregnancy: a comparative study of 400 cases Lucie BONIN (M.D)a, Cécile PEDREIRO (M.D)a, Stéphanie MORET (CRA)b, Gautier CHENE (M.D, Ph.D)a, Pascal GAUCHERAND (M.D, Ph.D)a, Géry LAMBLIN (M.D) a*
a
Department of Obstetrics and Gynecology Surgery, Femme Mère Enfant Hospital, Lyon,
Hospices Civils de Lyon, France b
Clinical Research Associate, Hospices Civils de Lyon, Femme Mère Enfant University Hospital, Department of Obstetrics and Gynecology, Lyon, France
Corresponding author: Géry Lamblin (M.D) Address of corresponding author: Department of Obstetrics and Gynecology, Urogynecology, Femme Mère Enfant University Hospital, 59 Boulevard Pinel, 69677 LyonBron, France. Telephone number: + 33-4-72-35-58-86 / +33-6-61-58-21-93 Fax number: +33-4-27-86-92-67 E-mail address: [email protected]
Contributions to authorship: Each author’s contribution to the manuscript Adresses of co-authors: a,b
Department of Obstetrics and Gynecology, Urogynecology, Femme Mère Enfant
University Hospital, 59 Boulevard Pinel, 69677 Lyon-Bron, France.
Conflict of Interest Statement: none
All authors saw and approved the final version.
Abstract Objective: We sought to evaluate the global success rate of intramuscular methotrexate for the treatment of ectopic pregnancy, identify factors predictive of treatment success or failure, and study methotrexate tolerability in a large patient cohort. Study design: For this single-center retrospective observational study, we retrieved the records of all women who had a clinically or echographically confirmed ectopic pregnancy with a Fernandez score <13 and who were treated according to a 1 mg/kg intramuscular single-dose methotrexate protocol. Medical treatment failure was defined by an obligation to proceed to laparoscopy. Needing a second injection was not considered to be medical treatment failure. Results: Between February 2008 and November 2013 (69 months), 400 women received methotrexate for ectopic pregnancy. The medical treatment protocol was effective for 314 patients, i.e., an overall success rate of 78.5%. A single methotrexate dose was sufficient for 63.5% of the women and a second dose was successful for 73.2% of the remaining women. The medical treatment success rate fell as initial hCG levels climbed. The main factors associated with methotrexate failure included day (D) 0, D4 and D7 hCG levels, pretherapeutic blood progesterone, hematosalpinx at D0 and pain at D7. Early favorable kinetics of hCG levels was predictive of success. Methotrexate treatment was successful in 90% of women who had D0 hCG <1,000 IU/l. Methotrexate tolerability was good, with only 9% of the women reporting non-severe adverse effects. The fertility rate with delivery after medical treatment for ectopic pregnancy was 80.7%. Conclusion: In this study, we showed that an initial hCG value <1,000 IU/l and favorable early HCG kinetics were predictive factors for the successful medical treatment of ectopic pregnancy by methotrexate, and hematosalpinx and pretherapeutic blood progesterone >5
ng/mL at diagnosis were predictive of its failure. We also confirmed good tolerability for single-dose methotrexate protocols.
Keywords: ectopic pregnancy; methotrexate; medical treatment; predictive factor; HCG plasmatic
Predictive factors for the methotrexate treatment outcome in ectopic pregnancy: a comparative study of 400 cases
Introduction
Ectopic, most frequently tubal (>95% of cases [1,2]), pregnancy is a leading cause of maternal mortality, representing 4.9% of maternal deaths in developed countries [3]. To potentially avoid surgery, methotrexate can be used as a first-line treatment for uncomplicated ectopic pregnancies with human chorionic gonadotropin (hCG) levels less than 5,000 IU/l [4,5]. The drawbacks of methotrexate are the possibility of treatment failure and resultantly a prolonged monitoring period. In previous studies, success rates of single-dose protocols of methotrexate in ectopic pregnancy have ranged from 70% for Sagiv et al in 2012 to 94.2% for Stovall and Ling in 1993 [6,7]. The literature presents a number of risk factors for methotrexate failure, the most important of which is pretreatment hCG. Cutoffs in the literature range from 1,000 to 5,000 IU/l [8]. The adverse effects of the treatment are most frequently hematological, digestive, hepatic, pulmonary or buccal in nature [8]. The largest series reported to date are those of Rabischong et al in 2011 [9] and Lipscomb et al in 1999 [10], with respectively 419 and 350 women included. Other series in the literature are smaller, often comprising less than 100 women. Our primary objective with the present study was to evaluate the overall success rate of systemic methotrexate in a large series of women who experienced ectopic pregnancy. Our secondary objectives were to describe patient characteristics, identify predictive factors for methotrexate treatment failure and success, and to assess tolerability.
Materials and methods
We performed a single-center retrospective observational study on women admitted to the obstetrics & gynecology (OBGYN) department of the Women and Children's Hospital in Lyon, France, between February 2008 and November 2013. Using the hospital's pharmacy registry, we extracted all patients for whom the OBGYN department had prescribed an injection of methotrexate during the study period. The data from the patients' computerized medical files were centralized in a spreadsheet (Excel for Mac 2011, Microsoft Corporation, Redmond, WA, USA) for statistical analyses. Patient inclusion criteria were admission for ectopic pregnancy, treatment with 1 mg/kg of intramuscular methotrexate, and documented hCG level. To homogenize the cohort, women with cervical, ovarian, cornual or cesarean scar ectopic pregnancies were excluded. Ectopic pregnancy was diagnosed on direct ultrasound signs (hematosalpinx, or lateral extrauterine gestational sac) associated with an empty uterus and positive hCG, or on clinical signs (pelvic pain, metrorrhagia) associated with an empty uterus and suggestive hCG kinetics over several days (no doubling, little reduction in or stagnation of levels at 48 hours). All ultrasounds were transvaginal. The six operators were senior physicians with extensive experience in the echographic diagnosis of ectopic pregnancy. Management protocols were based on the clinical practice recommendations for ectopic pregnancy published by the French National College of Obstetrics & Gynecology (CNGOF) in 2003 [11]. Contraindications for medical treatment with methotrexate were a Fernandez score >12 [12,13], hemodynamic instability, prior ipsilateral ectopic pregnancy, poor patient comprehension and unacceptable home-to-hospital distance. All patients had a series of pretherapeutic laboratory tests including hCG (reference level), blood progesterone, complete blood count, coagulation (prothrombin time, partial thromboplastin
time,
fibrinogen),
liver
function
(transaminase,
bilirubin,
alkaline
phosphatase), serum creatinine, blood group and irregular antibody screening. Hepatic or renal failure, and significant anemia or thrombocytopenia were contraindications for the administration of methotrexate. We retrospectively calculated the Fernandez score from the data found in the files of the patients (66 patients). All patients received an initial single 1 mg/kg intramuscular injection of methotrexate followed by two hours of surveillance in the OBGYN department. The day of that initial injection was defined as Day (D) 0. Thereafter, the patients were received for clinical followup on Days 2, 4 and 7 with ultrasounds performed if there were any doubts. Biological
surveillance included hCG assay at Days 2, 4, 7 and every seven days thereafter until negative, and a post-methotrexate workup at D7 (identical to the pretreatment workup). A 25 to 30% increase in the hCG level was tolerated at D4 but with a return to pretherapeutic levels expected for D7. Initial injection failure was defined as an increase in hCG level of >30% between D0 and D4, an hCG level at D7 greater than that of D0, or a secondary rebound of the hCG level. When initial injection failure was observed, a second injection could be administered using the same modalities as the first. Medical treatment failure was defined as a necessity for surgical intervention by laparoscopy (we note that needing a second methotrexate injection was not considered to be medical treatment failure). For the analyses of the present study, the patients were divided into two groups according to the success or failure of medical treatment with methotrexate. We contacted the 400 patients by phone and asked them about their fertility. All patients received information on the study and the computerized treatment of their data, but their written informed consent was not necessary as per French regulations concerning non-interventional studies.
Statistical analyses Statistical analyses were performed using SAS version 9.3 (SAS institute Inc., Cary, NC, USA). Quantitative data were described using the mean ± standard deviation and compared using the Student T test. Qualitative data were described by frequency and relative frequency and compared using the Chi² test or the Fisher's exact test when frequency was less than five. The main criterions were assessed by logistic regression. A ROC curve was plotted for the evaluation of the prognostic value of D0 hCG level. Statistical significance was set at p<0.05.
Results
General characteristics In the 69-month study period, 472 women were prescribed methotrexate in the setting of pregnancy at the Lyon Women & Children's Hospital. Of those women, 72 were excluded from the study: 1 who received methotrexate for postpartum rheumatoid arthritis; 22 who had non-tubal ectopic pregnancies, 4 who had failed intrauterine pregnancies diagnosed during follow-up and 45 who had incomplete medical files as concerns the data necessary for the study or who were rapidly lost to follow-up. During the 69-month period, we supported 936 ectopic pregnancies in our hospital, 536 patients were initially treated with surgery and thus 400 treated with methotrexate therapy (42.7%). Figure 1 provides a flow chart illustrating treatments and their outcomes for the series and Table 1 a summary of patient characteristics. The mean age of the patients was 30.6 (±5.2) years and the mean gestational age at the first methotrexate injection was 42.9 (±12.4) days. Of the 400 included patients, 318 (79.5%) needed only a single methotrexate injection and 82 (20.5%) needed two injections. There were 248 cases of echographically visible tubal pregnancies and 152 cases for which the tubal location could not be affirmed by echography but that were nonetheless treated as extrauterine pregnancies. Twelve patients underwent a diagnostic laparoscopy when there was a doubt about the diagnostic. It happened before receiving methotrexate treatment, which was the first-line therapy for these patients afterwards. A history of ectopic pregnancy was noted for 44 (11%) of the women, 17 of whom were treated radically. Ectopic pregnancies occurred in a setting of apparent contraception failure for 34 (8.5%) women. Birth control methods for these women were: intrauterine devices (5 women), oral contraceptives (15), condoms (3) and oral emergency contraception (11). The ectopic pregnancies occurred in the setting of assisted reproduction for 52 (13%) cases, 38 of which were in vitro fertilizations. The mean pretherapeutic hCG level for all patients was 1,104 IU/l (± 1,263). Mean blood progesterone was 7.79 ng/ml (±21.72). Mean gravidity and parity were respectively 2.6 (±1.7) and 0.8 (±1.0). There were 212 nulliparous, 99 primiparous and 89 multiparous women. Motives for seeking consultation were metrorrhagia for 116 (29%) patients, pain for 62 (15.5%), metrorrhagia and pain for 169 (42.2%) and abnormal hCG kinetics for 51 (12.75%). Forty-nine (12.2%) patients were asymptomatic.
Hematosalpinx was detected by ultrasound on D0 in 243 (60.7%) patients: 24 (6%) measured less than 10 mm, 202 (50.5%) between 10 and 30 mm and 17 (4.2%) more than 30 mm. There were no significant differences as concerns hematosalpinx size.
Outcomes A single injection of methotrexate successfully treated 63.5% of the ectopic pregnancies. A second injection was successful for 73.2% of the patients who needed it. Second injections, when needed, were administered after a mean of 8.4 (±10.9) days following the first injections. The mean D0 hCG level in patients who needed a second injection was 1,757 IU/l (±1,942). The overall methotrexate treatment success rate (combining first and second injections) was 78.5% (314 patients). The mean delay to negative hCG was 20.7 (±12.2) days. An inversely proportional relation was observed between the medical treatment success rate and initial hCG levels, with the former falling as the latter climbed (p<0.05) (Figure 2). We also found that the threshold for a 80% success rate was reached for a Fernandez score of 10 or under (calculated for 66 patients for whom the data were available) (Figure 3). The mean Fernandez score was 9,2 for the patients with a successful methotrexate treatment (48 patients), and 10.9 for those with a failure of the medical therapy (18 patients). The difference of Fernandez score between the groups ―success‖ and ―failure‖ was statistically significative (p=0,0002). Surgery after methotrexate failure was necessary for 86 women. Indications (sometimes concurrent) for endoscopy are summarized in Table 2. In the women who needed it, the mean delay to endoscopy was 24.5 (±110.8) days. Among the 86 failures of treatment by methotrexate therapy, 49 patients treated by radical surgery (57%) and 37 patients by conservative surgery (43%).
Analyses Table 3 summarizes the results of univariate analyses in the medical success and failure groups including all patients (one and two methotrexate injections) and Table 4 summarizes these analyses in patients who received two injections of methotrexate. A multivariate analysis was also performed but it identified no statistically significant factors. Figure 4 shows the plotted Receiver Operative Characteristic (ROC) curve and the area under the curve for the evaluation of the prognostic value of D0 hCG level. The cutoff value was set
at 1,000 IU/l, with sensitivity and specificity of 70%. The negative and positive predictive values were respectively 40% and 90%.
Adverse effects In all, 36 (9%) patients experienced adverse effects: digestive disorders (nausea, vomiting, diarrhea) occurred in 18 (4.5%) women, mild hepatic cytolysis in 15 (3.75), thrombocytopenia in 1 (0.25%), headache in 1 (0.25%) and stomatitis in 1 (0.25%). There were no serious adverse events.
Fertility We have contacted 400 patients by phone. We failed to contact 93 patients: i.e. 23% lost to follow-up. For the 307 patients successfully contacted, the 2-year cumulative intrauterine pregnancy rate was 82,1%. We found rates of 80.7% for delivery after 22 weeks of gestation, 20.1% for spontaneous miscarriage, 0.8% for late miscarriage, 4.4% for abortion and 13.2% for ectopic pregnancy.
Comments
A methotrexate single-dose protocol successfully treated 78.5% of the 400 cases of ectopic pregnancy included in our study. Our results align with those of other recent studies where similar single-dose protocols were used in series comprising more than 100 women: Between 2011 and 2014, Rabischong et al [9], Sagiv et al [6], Skubisz et al [14] and Helmy et al [15] reported success rates of respectively 75.4% in a series of 419 patients, 70% in 238 patients, 71% in 206 patients and 81.8% in 198 patients. Patient characteristics in our study and in the above-mentioned studies were also comparable. A history of ectopic pregnancy was noted for 11% of our patients, close to the range of 12.6% to 14% reported in most recent studies [9,16–18]. Our delay to negative hCG was somewhat shorter than those reported in the literature. In our study, 9.75% of the pregnancies were a result of in vitro fertilization, which is comparable to the 7.3% rate reported by Van Mello et al in 2013 [18]. A single methotrexate injection was sufficient for 63.5% of our patients, and a second injection resolved 73.2% of the remaining cases. Factors predictive of failure for the first methotrexate injection were hCG levels at days 0, 2, 4, 7 and 14, blood progesterone level at D0, hematosalpinx at D0 and pain at D7. In contrast, a reduction in hCG between D0 and D2, between D0 and D4 or between D0 and D7 were statistically significant for the prediction of success. Predictive factors for failure after a second injection were the hCG levels at days 0, 2, 4 and 7. As was the case for Cho et al [19] and Helmy et al [15], the presence of clinical signs (pain, metrorrhagia) at diagnosis was predictive of neither methotrexate treatment failure nor its success in our study. The pretreatment hCG level is a known predictor of success for methotrexate treatment, but thresholds reported in the literature vary from 1000 IU/l to as much as 5000 IU/l. In our study, initial hCG values <1000 IU/l were associated with a success rate of 90%. Our threshold value of 1000 IU/l was the same or similar to those retained by several other teams. For example, both Nazac et al [16] and Potter et al [20] retained the 1000 IU/l threshold, and Rabischong et al retained 1300 IU/l for their 2011 study [9]. We also found that the pretherapeutic Fernandez score was a statistically significative predictor of success: there was 80% of success when the score was under 10. In our study, but not in previous studies, hematosalpinx at diagnosis was a predictive factor of methotrexate treatment failure [21]. We also noticed that as the size of echographically visible
lateral extrauterine masses increased so did failure rates, but this association was not statistically significant. This same dynamic has been reported in most studies [6,10,16,20–22]. The predictive value of blood progesterone has received little attention in the literature. Both Nazac et al and Lipscomb et al reported that it had no significant predictive role [10,16]. However, Corsan et al did show, albeit in a small series of 44 patients, that women with blood progesterone >7 ng/ml had a significantly greater risk of methotrexate failure [23]. In our study, the rate of methotrexate failure was significantly higher when blood progesterone was greater than 5 ng/ml. Thus, our study identified pretreatment blood progesterone as a statistically significant predictive factor for methotrexate treatment outcome. Considering its ease and low cost, blood progesterone may be of great interest for daily clinical practice in this setting. Day 2 hCG level has also received little attention in the literature, but our study suggests that it holds at least a hypothetical promise for rapidly identifying women who are at risk of methotrexate treatment failure. For practice however, many authors suggest waiting until D4 to reassay hCG because an initial rise in it is expected in the first few days due to trophoblast lysis [24]. Furthermore, convoking a patient at two days post injection and then again at four days as indicated in most protocols would be impractical. As they are available early in management and easy to interpret, D4 and D7 hCG levels retain their interest as predictive factors for methotrexate efficacy. Equally pertinent and easy to interpret, reductions in hCG levels between D0 and D4 and between D0 and D7 were predictive factors in our study and in several earlier studies [14,17,25–27]. The D0 to D4 hCG reduction is particularly interesting in that it offers early treatment prognosis value and in turn enables a possibility of lightened surveillance for patients likely to have good outcomes or a more speedy second injection for patients who need it (the mean delay to second injection was 8.4 days in our study). In both cases, the cumbersome nature of follow-up—and, one would hope, the number of patients lost to it—could be reduced. In a 2014 study performed by Bottin et al in a series of 91 patients, a 20% decrease in hCG between D0 and D4 was identified as an effective predictor of methotrexate success and thus compatible with a single D30 follow-up visit instead of the normal weekly visit schedule [27]. A D2 control is unnecessary after methotrexate treatment, as we know there is an initial normal rise of the hCG level after the injection and it is not discriminant. Conversely, D4 hCG level is interesting to predict methotrexate outcome, and especially considering the evolution of hCG between D0 and D4. It remains an early predictive factor for outcome of methotrexate treatment.
In our study, pain at D7 was a predictive factor for first-injection methotrexate failure. However, in this setting, pain must be interpreted carefully since it is often present in the normal course of methotrexate treatment, where it may indicate hematoma formation or tubal abortion [24,28]. Because of the complexity of interpreting pain and its subjective nature, we think that this finding has little clinical pertinence. Tolerability of methotrexate was good overall in our study. The rate of adverse effects was 9%, comprising only minor, mostly digestive complaints. The study of fertility after a medical treatment for ectopic pregnancy found a rate of 80.7% for delivery after 22 weeks of gestation and 13.2% for recurrence ectopic pregnancy. These results were quite congruent with those of the Demeter study, which found a rate of intrauterine pregnancy of 67% and a recurrence of ectopic pregnancy of 12 % after medical treatment. Our fertility rate with a delivery outcome was even a little bit higher than expected (80.7%) [29]. The main strength of the study that we report here is the size of our patient series, one of the largest to date, second only to that of Rabischong et al published in 2011 [9]. We also present here recent and applicable data. The main limit of our study is that it is retrospective: although it reflects our clinical practice experience, our study could be affected by certain biases and limits innate to this type of study.
Conclusion
For the present study, we evaluated a large series of women to identify predictive factors for the outcome of methotrexate treatment in ectopic pregnancy. We confirmed that an initial hCG level <1,000 IU/l and a Fernandez score under 11 were predictive of success, and showed that D0 blood progesterone >5 ng/ml was predictive of failure. Considering the ease and low cost of blood progesterone testing, this latter finding merits further confirmatory testing in our opinion. While not strictly decisional, our hCG threshold can be used to inform patients on the likelihood of methotrexate success or failure. The kinetics of hCG between D0 and D4 and between D0 and D7 must be carefully monitored in the days following the initial injection of methotrexate. Tolerability was good in our study and thus should not be an issue when medical treatment with methotrexate is indicated for ectopic pregnancy.
Acknowledgments Supported by a grant from the Bibliothèque scientifique de l’Internat de Lyon and les Hospices Civils de Lyon
Conflict of interest None
Funding None
Table 1. Baseline characteristics of 400 women treated by methotrexate for ectopic pregnancy. Table 2. Indications of surgery in 86 patients with methotrexate (MTX) failure. Table 3. Univariate analysis comparing success and failure of methotrexate treatment in 400 patients with ectopic pregnancy. Table 4. Univariate analysis comparing success and failure of the second injection of methotrexate (MTX) in patients with ectopic pregnancy. Figure 1. Treatment flow chart. Figure 2. Sucess and failure rates for methotrexate (MTX) treatment according to pretherapeutic HCG levels. Figure 3. Sucess and failure rates for methotrexate (MTX) treatment according to Fernandez score. Figure 4. ROC Curve for the evaluation of the prognostic value of D0 hCG level (IU/L) Area under the curve: 75.27% ± 3.07% (95% confidence interval : 69.24% - 81.29%)
References
[1]
Bouyer J, Coste J, Fernandez H, Pouly JL, Job-Spira N. Sites of ectopic pregnancy: a
10 year population-based study of 1800 cases. Hum Reprod 2002;17:3224–30. doi:10.1093/humrep/17.12.3224. [2]
American College of Obstetricians and Gynecologists. ACOG Practice Bulletin No.
94: Medical management of ectopic pregnancy. Obstet Gynecol 2008;111:1479–85. doi:10.1097/AOG.0b013e31817d201e. [3]
Khan KS, Wojdyla D, Say L, Gülmezoglu AM, Van Look PFA. WHO analysis of
causes of maternal death: a systematic review. Lancet Lond Engl 2006;367:1066–74. doi:10.1016/S0140-6736(06)68397-9. [4]
Lesavre M, Curinier S, Capmas P, Rabischong B, Fernandez H. [Treatment of tubal
ectopic pregnancy by methotrexate].
Gyn cologie
bst trique Biol eprod
–9.
doi:10.1016/j.jgyn.2014.12.013. [5]
Marret H, Fauconnier A, Dubernard G, Misme H, Lagarce L, Lesavre M, et al.
[Evidence-based evaluation and expertise of methotrexate off label use in gynaecology and obstetrics wor of the
G F
Gyn cologie
bst trique Biol
eprod
–6.
doi:10.1016/j.jgyn.2014.12.008. [6]
Sagiv R, Debby A, Feit H, Cohen-Sacher B, Keidar R, Golan A. The optimal cutoff
serum level of human chorionic gonadotropin for efficacy of methotrexate treatment in women with extrauterine pregnancy. Int J Gynaecol Obstet Off Organ Int Fed Gynaecol Obstet 2012;116:101–4. doi:10.1016/j.ijgo.2011.09.023. [7]
Stovall TG, Ling FW. Single-dose methotrexate: an expanded clinical trial. Am J
Obstet Gynecol 1993;168:1759–1762; discussion 1762–1765. [8]
Varma R, Gupta J. Tubal ectopic pregnancy. BMJ Clin Evid 2012;2012.
[9]
Rabischong B, Tran X, Sleiman AA, Larraín D, Jaffeux P, Aublet-Cuvelier B, et al.
Predictive factors of failure in management of ectopic pregnancy with single-dose methotrexate: a general population-based analysis from the Auvergne Register, France. Fertil Steril 2011;95:401–404, 404.e1. doi:10.1016/j.fertnstert.2010.08.025.
[10]
Lipscomb GH, McCord ML, Stovall TG, Huff G, Portera SG, Ling FW. Predictors of
success of methotrexate treatment in women with tubal ectopic pregnancies. N Engl J Med 1999;341:1974–8. doi:10.1056/NEJM199912233412604. [11]
CNGOF. Collège National des Gynécologues et Obstétriciens Français. Guidelines
for clinical practice Ectopic pregnancy management
Gyn cologie bst trique Biol eprod
2003;32:3S6–112. [12]
Fernandez H, Lelaidier C, Thouvenez V, Frydman R. The use of a pretherapeutic,
predictive score to determine inclusion criteria for the non-surgical treatment of ectopic pregnancy. Hum Reprod Oxf Engl 1991;6:995–8. [13]
Fernandez H, Bourget P, Ville Y, Lelaidier C, Frydman R. Treatment of unruptured
tubal pregnancy with methotrexate: pharmacokinetic analysis of local versus intramuscular administration. Fertil Steril 1994;62:943–7. [14]
Skubisz M, Dutton P, Duncan WC, Horne AW, Tong S. Using a decline in serum hCG
between days 0-4 to predict ectopic pregnancy treatment success after single-dose methotrexate: a retrospective cohort study. BMC Pregnancy Childbirth 2013;13:30. doi:10.1186/1471-2393-13-30. [15]
Helmy S, Bader Y, Pablik E, Tiringer D, Pils S, Laml T, et al. Cut-off value of initial
serum β-hCG level predicting a successful MTX therapy in tubal ectopic pregnancy: a retrospective cohort study. Eur J Obstet Gynecol Reprod Biol 2014;179:175–80. doi:10.1016/j.ejogrb.2014.05.033. [16]
Nazac A, Gervaise A, Bouyer J, de Tayrac R, Capella-Allouc S, Fernandez H.
Predictors of success in methotrexate treatment of women with unruptured tubal pregnancies. Ultrasound Obstet Gynecol Off J Int Soc Ultrasound Obstet Gynecol 2003;21:181–5. doi:10.1002/uog.9. [17]
Skubisz MM, Li J, Lee J, Wallace EM, Tong S. Decline in βhCG levels between days
0 and 4 after a single dose of methotrexate for ectopic pregnancy predicts treatment success: a retrospective
cohort
study.
BJOG
Int
J
Obstet
Gynaecol
2011;118:1665–8.
doi:10.1111/j.1471-0528.2011.03133.x. [18]
Van Mello NM, Mol F, Verhoeve HR, van Wely M, Adriaanse AH, Boss EA, et al.
Methotrexate or expectant management in women with an ectopic pregnancy or pregnancy of
unknown location and low serum hCG concentrations? A randomized comparison. Hum Reprod Oxf Engl 2013;28:60–7. doi:10.1093/humrep/des373. [19]
Cho GJ, Lee SH, Shin JW, Lee NW, Kim T, Kim HJ, et al. Predictors of success of
repeated injections of single-dose methotrexate regimen for tubal ectopic pregnancy. J Korean Med Sci 2006;21:86–9. doi:10.3346/jkms.2006.21.1.86. [20]
Potter MB, Lepine LA, Jamieson DJ. Predictors of success with methotrexate
treatment of tubal ectopic pregnancy at Grady Memorial Hospital. Am J Obstet Gynecol 2003;188:1192–4. [21]
Krissi H, Peled Y, Eitan R, Bishara A, Goldchmit C, Ben-Haroush A. Single-dose
methotrexate injection for treatment of ectopic pregnancy in women with relatively low levels of human chorionic gonadotropin. Int J Gynaecol Obstet Off Organ Int Fed Gynaecol Obstet 2013;121:141–3. doi:10.1016/j.ijgo.2012.11.020. [22]
Gnisci A, Stefani L, Bottin P, Ohannessian A, Gamerre M, Agostini A. Predictive
value of hemoperitoneum for outcome of methotrexate treatment in ectopic pregnancy: an observational comparative study. Ultrasound Obstet Gynecol Off J Int Soc Ultrasound Obstet Gynecol 2014;43:698–701. doi:10.1002/uog.13255. [23]
Corsan GH, Karacan M, Qasim S, Bohrer MK, Ransom MX, Kemmann E.
Identification of hormonal parameters for successful systemic single-dose methotrexate therapy in ectopic pregnancy. Hum Reprod 1995;10:2719–22. [24]
Luciano AA, Roy G, Solima E. Ectopic Pregnancy. Ann N Y Acad Sci 2001;943:235–
54. doi:10.1111/j.1749-6632.2001.tb03805.x. [25]
Agostini A, Blanc K, Ronda I, Romain F, Capelle M, Blanc B. Prognostic value of
human chorionic gonadotropin changes after methotrexate injection for ectopic pregnancy. Fertil Steril 2007;88:504–6. doi:10.1016/j.fertnstert.2006.11.138. [26]
Nowak-Markwitz E, Michalak M, Olejnik M, Spaczynski M. Cutoff value of human
chorionic gonadotropin in relation to the number of methotrexate cycles in the successful treatment
of
ectopic
pregnancy.
Fertil
Steril
2009;92:1203–7.
doi:10.1016/j.fertnstert.2008.07.1775. [27]
Bottin P, Gnisci A, Crochet P, Butzbach P, Cravello L, Gamerre M, et al. [Prognostic
value of early hCG changes after methotrexate injection for ectopic pregnancy Gyn cologie bst trique Fertil [28]
–7. doi:10.1016/j.gyobfe.2013.08.006.
Lipscomb GH, Puckett KJ, Bran D, Ling FW. Management of separation pain after
single-dose methotrexate therapy for ectopic pregnancy. Obstet Gynecol 1999;93:590–3. doi:10.1016/S0029-7844(98)00523-7. [29]
Fernandez H, Capmas P, Lucot JP, Resch B, Panel P, Bouyer J, et al. Fertility after
ectopic pregnancy: the DEMETER randomized trial. Hum Reprod 2013;28:1247–53. doi:10.1093/humrep/det037.
Table 1. Baseline characteristics of 400 women treated by methotrexate for ectopic pregnancy. Successful MTX treatment (n=314) Age (years) ≤24 28 (9) 25-29 91 (29) 30-34 114 (36) >34 81 (26) Gravidity 2.6 ± 1.6 Parity 0.8 ± 1.0 Previous ectopic pregnancy 33 (10) Gestational age*, days 43.1 ± 12.6 Clinical symptoms* Metrorrhagia 93 (30) Pelvic pain 49 (516) Metrorrhagia and pain 130 (42) Hematosalpinx* 182 (60) <10 mm 20 (6) 10-30 mm 151 (50) >30 mm 11 (4) D0 hCG level 862 ± 1073 <500 179 (57) 500-1499 75 (24) ≥1500 60 (19) D0 blood progesterone (ng/mL) Rate 4.7 ± 5.0 <5 38 (68) Contraception** 29 (9) Spontaneous pregnancy 275 (88) Data are shown as mean ± SD or n (%); significant scores with p < 0.05 are bolded D0 : Day of the first methotrexate injection hCG : human chorionic gonadotropin, IU/L *Data collected on D0 **Before the diagnosis of pregnancy Successful treatment: no surgery needed Failure of treatment: necessity of laparoscopy after the MTX treatment
Failure of MTX treatment (n=86) 16 (19) 27 (31) 23 (27) 20 (23) 2.5 ± 1.7 0.7 ± 0.9 11 (13) 42.3 ± 11.7 23 (27) 13 (15) 39 (45) 61 (73) 4 (5) 51 (61) 6 (7) 1972 ± 1499 15 (18) 23 (27) 46 (55) 8.0 ± 8.7 8 (42) 11 (58) 5 (6) 71 (82)
p
0.05 0.51 0.26 0.55 0.64
0.93 0.03
0.08 <0.0001
<0.0001 0.047 0.046 0.30 0.17
Table 2. Indications of surgery in 86 patients with methotrexate (MTX) failure.
n Clinical signs suggestive of tubal rupture (severe abdominal pain, abdominal guarding, faintness)
60
Occurrence of echographic signs (occurrence or increase of hematosalpinx or hemoperitoneum, appearance of an embryo with cardiac activity)
59
Poor hCG course (increase in hCG level of >30% between Day (D) 0 and D4, hCG level at D7 greater than that of D0, or secondary rebound of the hCG level)
24
Falling hemoglobin levels
2
Refusal of the second MTX injection
1
Table 3. Univariate analysis comparing success and failure of methotrexate treatment in 400 patients with ectopic pregnancy
D2 hCG level D4 hCG level D7 hCG level D14 hCG level D2 hCG
Successful MTX treatment (n=314) 908 ± 1266 802 ± 1252 530 ± 952 145 ± 225 131 (53) 185 (65) 220 (81) 81 (35) 68 (26) 51 (21) 32 (11)
Data are shown as mean ± SD or n (%) ; significant scores with p < 0.05 are bolded D0 : Day of the methotrexate injection hCG : human chorionic gonadotropin, IU/L Successful treatment: no surgery needed Failure of treatment: necessity of laparoscopy after the MTX treatment
Failure of MTX treatment (n=86) 2558 ± 2247 2553 ± 2401 1579 ± 1990 795 ± 869 15 (24) 16 (27) 24 (63) 26 (49) 21 (37) 17 (47) 24.5 ± 110.8 4 (6)
p <0.0001 <0.0001 <0.0001 <0.0001 <0.0001 <0.0001 0.009 0.09 0.12 0.0005 0.17
Table 4. Univariate analysis comparing success and failure of the second injection of methotrexate (MTX) in patients with ectopic pregnancy
D0 hCG D2 hCG D4 hCG D7 hCG D14 hCG hCG drop D2-D0 > 500 hCG drop D4-D0 > 500 hCG drop D7-D0 > 500 Delay between first and second injections, days D0 blood progesterone (ng/mL)
Successful MTX treatment (n=60) 1261 ± 1426 1147 ± 1394 910 ± 1085 581 ± 745 168 ± 219 38 (90) 34 (67) 30 (59) 9.25 ± 12.6
Failure of MTX treatment (n=22) 3045 ± 2536 2530 ± 2350 1901 ± 1857 2191 ± 1499 636 ± 646 13 (72) 7 (50) 0 (0) 6.05 ± 1.79
4.26 ± 4.9
4 ± 4.67
Data are shown as mean ± SD or n (%) ; significant scores with p < 0.05 are bolded D0 : Day of the second methotrexate injection hCG : human chorionic gonadotropin, IU/L Successful treatment: no surgery needed Failure of treatment: necessity of laparoscopy after the MTX treatment
p 0.0001 0.006 0.01 <0.0001 0.002 0.11 0.25 0.008 0.24 0.94
Fig 1
Fig 2
Fig 3
Fig 4
S0301-2115(16)31027-2 http://dx.doi.org/doi:10.1016/j.ejogrb.2016.11.016 EURO 9687
To appear in:
EURO
Received date: Revised date: Accepted date:
12-7-2016 8-10-2016 15-11-2016
Please cite this article as: Bonin Lucie, Pedreiro C´ecile, Moret St´ephanie, Chene Gautier, Gaucherand Pascal, Lamblin G´ery.Predictive factors for the methotrexate treatment outcome in ectopic pregnancy: a comparative study of 400 cases.European Journal of Obstetrics and Gynecology and Reproductive Biology http://dx.doi.org/10.1016/j.ejogrb.2016.11.016 This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
Predictive factors for the methotrexate treatment outcome in ectopic pregnancy: a comparative study of 400 cases Lucie BONIN (M.D)a, Cécile PEDREIRO (M.D)a, Stéphanie MORET (CRA)b, Gautier CHENE (M.D, Ph.D)a, Pascal GAUCHERAND (M.D, Ph.D)a, Géry LAMBLIN (M.D) a*
a
Department of Obstetrics and Gynecology Surgery, Femme Mère Enfant Hospital, Lyon,
Hospices Civils de Lyon, France b
Clinical Research Associate, Hospices Civils de Lyon, Femme Mère Enfant University Hospital, Department of Obstetrics and Gynecology, Lyon, France
Corresponding author: Géry Lamblin (M.D) Address of corresponding author: Department of Obstetrics and Gynecology, Urogynecology, Femme Mère Enfant University Hospital, 59 Boulevard Pinel, 69677 LyonBron, France. Telephone number: + 33-4-72-35-58-86 / +33-6-61-58-21-93 Fax number: +33-4-27-86-92-67 E-mail address: [email protected]
Contributions to authorship: Each author’s contribution to the manuscript Adresses of co-authors: a,b
Department of Obstetrics and Gynecology, Urogynecology, Femme Mère Enfant
University Hospital, 59 Boulevard Pinel, 69677 Lyon-Bron, France.
Conflict of Interest Statement: none
All authors saw and approved the final version.
Abstract Objective: We sought to evaluate the global success rate of intramuscular methotrexate for the treatment of ectopic pregnancy, identify factors predictive of treatment success or failure, and study methotrexate tolerability in a large patient cohort. Study design: For this single-center retrospective observational study, we retrieved the records of all women who had a clinically or echographically confirmed ectopic pregnancy with a Fernandez score <13 and who were treated according to a 1 mg/kg intramuscular single-dose methotrexate protocol. Medical treatment failure was defined by an obligation to proceed to laparoscopy. Needing a second injection was not considered to be medical treatment failure. Results: Between February 2008 and November 2013 (69 months), 400 women received methotrexate for ectopic pregnancy. The medical treatment protocol was effective for 314 patients, i.e., an overall success rate of 78.5%. A single methotrexate dose was sufficient for 63.5% of the women and a second dose was successful for 73.2% of the remaining women. The medical treatment success rate fell as initial hCG levels climbed. The main factors associated with methotrexate failure included day (D) 0, D4 and D7 hCG levels, pretherapeutic blood progesterone, hematosalpinx at D0 and pain at D7. Early favorable kinetics of hCG levels was predictive of success. Methotrexate treatment was successful in 90% of women who had D0 hCG <1,000 IU/l. Methotrexate tolerability was good, with only 9% of the women reporting non-severe adverse effects. The fertility rate with delivery after medical treatment for ectopic pregnancy was 80.7%. Conclusion: In this study, we showed that an initial hCG value <1,000 IU/l and favorable early HCG kinetics were predictive factors for the successful medical treatment of ectopic pregnancy by methotrexate, and hematosalpinx and pretherapeutic blood progesterone >5
ng/mL at diagnosis were predictive of its failure. We also confirmed good tolerability for single-dose methotrexate protocols.
Keywords: ectopic pregnancy; methotrexate; medical treatment; predictive factor; HCG plasmatic
Predictive factors for the methotrexate treatment outcome in ectopic pregnancy: a comparative study of 400 cases
Introduction
Ectopic, most frequently tubal (>95% of cases [1,2]), pregnancy is a leading cause of maternal mortality, representing 4.9% of maternal deaths in developed countries [3]. To potentially avoid surgery, methotrexate can be used as a first-line treatment for uncomplicated ectopic pregnancies with human chorionic gonadotropin (hCG) levels less than 5,000 IU/l [4,5]. The drawbacks of methotrexate are the possibility of treatment failure and resultantly a prolonged monitoring period. In previous studies, success rates of single-dose protocols of methotrexate in ectopic pregnancy have ranged from 70% for Sagiv et al in 2012 to 94.2% for Stovall and Ling in 1993 [6,7]. The literature presents a number of risk factors for methotrexate failure, the most important of which is pretreatment hCG. Cutoffs in the literature range from 1,000 to 5,000 IU/l [8]. The adverse effects of the treatment are most frequently hematological, digestive, hepatic, pulmonary or buccal in nature [8]. The largest series reported to date are those of Rabischong et al in 2011 [9] and Lipscomb et al in 1999 [10], with respectively 419 and 350 women included. Other series in the literature are smaller, often comprising less than 100 women. Our primary objective with the present study was to evaluate the overall success rate of systemic methotrexate in a large series of women who experienced ectopic pregnancy. Our secondary objectives were to describe patient characteristics, identify predictive factors for methotrexate treatment failure and success, and to assess tolerability.
Materials and methods
We performed a single-center retrospective observational study on women admitted to the obstetrics & gynecology (OBGYN) department of the Women and Children's Hospital in Lyon, France, between February 2008 and November 2013. Using the hospital's pharmacy registry, we extracted all patients for whom the OBGYN department had prescribed an injection of methotrexate during the study period. The data from the patients' computerized medical files were centralized in a spreadsheet (Excel for Mac 2011, Microsoft Corporation, Redmond, WA, USA) for statistical analyses. Patient inclusion criteria were admission for ectopic pregnancy, treatment with 1 mg/kg of intramuscular methotrexate, and documented hCG level. To homogenize the cohort, women with cervical, ovarian, cornual or cesarean scar ectopic pregnancies were excluded. Ectopic pregnancy was diagnosed on direct ultrasound signs (hematosalpinx, or lateral extrauterine gestational sac) associated with an empty uterus and positive hCG, or on clinical signs (pelvic pain, metrorrhagia) associated with an empty uterus and suggestive hCG kinetics over several days (no doubling, little reduction in or stagnation of levels at 48 hours). All ultrasounds were transvaginal. The six operators were senior physicians with extensive experience in the echographic diagnosis of ectopic pregnancy. Management protocols were based on the clinical practice recommendations for ectopic pregnancy published by the French National College of Obstetrics & Gynecology (CNGOF) in 2003 [11]. Contraindications for medical treatment with methotrexate were a Fernandez score >12 [12,13], hemodynamic instability, prior ipsilateral ectopic pregnancy, poor patient comprehension and unacceptable home-to-hospital distance. All patients had a series of pretherapeutic laboratory tests including hCG (reference level), blood progesterone, complete blood count, coagulation (prothrombin time, partial thromboplastin
time,
fibrinogen),
liver
function
(transaminase,
bilirubin,
alkaline
phosphatase), serum creatinine, blood group and irregular antibody screening. Hepatic or renal failure, and significant anemia or thrombocytopenia were contraindications for the administration of methotrexate. We retrospectively calculated the Fernandez score from the data found in the files of the patients (66 patients). All patients received an initial single 1 mg/kg intramuscular injection of methotrexate followed by two hours of surveillance in the OBGYN department. The day of that initial injection was defined as Day (D) 0. Thereafter, the patients were received for clinical followup on Days 2, 4 and 7 with ultrasounds performed if there were any doubts. Biological
surveillance included hCG assay at Days 2, 4, 7 and every seven days thereafter until negative, and a post-methotrexate workup at D7 (identical to the pretreatment workup). A 25 to 30% increase in the hCG level was tolerated at D4 but with a return to pretherapeutic levels expected for D7. Initial injection failure was defined as an increase in hCG level of >30% between D0 and D4, an hCG level at D7 greater than that of D0, or a secondary rebound of the hCG level. When initial injection failure was observed, a second injection could be administered using the same modalities as the first. Medical treatment failure was defined as a necessity for surgical intervention by laparoscopy (we note that needing a second methotrexate injection was not considered to be medical treatment failure). For the analyses of the present study, the patients were divided into two groups according to the success or failure of medical treatment with methotrexate. We contacted the 400 patients by phone and asked them about their fertility. All patients received information on the study and the computerized treatment of their data, but their written informed consent was not necessary as per French regulations concerning non-interventional studies.
Statistical analyses Statistical analyses were performed using SAS version 9.3 (SAS institute Inc., Cary, NC, USA). Quantitative data were described using the mean ± standard deviation and compared using the Student T test. Qualitative data were described by frequency and relative frequency and compared using the Chi² test or the Fisher's exact test when frequency was less than five. The main criterions were assessed by logistic regression. A ROC curve was plotted for the evaluation of the prognostic value of D0 hCG level. Statistical significance was set at p<0.05.
Results
General characteristics In the 69-month study period, 472 women were prescribed methotrexate in the setting of pregnancy at the Lyon Women & Children's Hospital. Of those women, 72 were excluded from the study: 1 who received methotrexate for postpartum rheumatoid arthritis; 22 who had non-tubal ectopic pregnancies, 4 who had failed intrauterine pregnancies diagnosed during follow-up and 45 who had incomplete medical files as concerns the data necessary for the study or who were rapidly lost to follow-up. During the 69-month period, we supported 936 ectopic pregnancies in our hospital, 536 patients were initially treated with surgery and thus 400 treated with methotrexate therapy (42.7%). Figure 1 provides a flow chart illustrating treatments and their outcomes for the series and Table 1 a summary of patient characteristics. The mean age of the patients was 30.6 (±5.2) years and the mean gestational age at the first methotrexate injection was 42.9 (±12.4) days. Of the 400 included patients, 318 (79.5%) needed only a single methotrexate injection and 82 (20.5%) needed two injections. There were 248 cases of echographically visible tubal pregnancies and 152 cases for which the tubal location could not be affirmed by echography but that were nonetheless treated as extrauterine pregnancies. Twelve patients underwent a diagnostic laparoscopy when there was a doubt about the diagnostic. It happened before receiving methotrexate treatment, which was the first-line therapy for these patients afterwards. A history of ectopic pregnancy was noted for 44 (11%) of the women, 17 of whom were treated radically. Ectopic pregnancies occurred in a setting of apparent contraception failure for 34 (8.5%) women. Birth control methods for these women were: intrauterine devices (5 women), oral contraceptives (15), condoms (3) and oral emergency contraception (11). The ectopic pregnancies occurred in the setting of assisted reproduction for 52 (13%) cases, 38 of which were in vitro fertilizations. The mean pretherapeutic hCG level for all patients was 1,104 IU/l (± 1,263). Mean blood progesterone was 7.79 ng/ml (±21.72). Mean gravidity and parity were respectively 2.6 (±1.7) and 0.8 (±1.0). There were 212 nulliparous, 99 primiparous and 89 multiparous women. Motives for seeking consultation were metrorrhagia for 116 (29%) patients, pain for 62 (15.5%), metrorrhagia and pain for 169 (42.2%) and abnormal hCG kinetics for 51 (12.75%). Forty-nine (12.2%) patients were asymptomatic.
Hematosalpinx was detected by ultrasound on D0 in 243 (60.7%) patients: 24 (6%) measured less than 10 mm, 202 (50.5%) between 10 and 30 mm and 17 (4.2%) more than 30 mm. There were no significant differences as concerns hematosalpinx size.
Outcomes A single injection of methotrexate successfully treated 63.5% of the ectopic pregnancies. A second injection was successful for 73.2% of the patients who needed it. Second injections, when needed, were administered after a mean of 8.4 (±10.9) days following the first injections. The mean D0 hCG level in patients who needed a second injection was 1,757 IU/l (±1,942). The overall methotrexate treatment success rate (combining first and second injections) was 78.5% (314 patients). The mean delay to negative hCG was 20.7 (±12.2) days. An inversely proportional relation was observed between the medical treatment success rate and initial hCG levels, with the former falling as the latter climbed (p<0.05) (Figure 2). We also found that the threshold for a 80% success rate was reached for a Fernandez score of 10 or under (calculated for 66 patients for whom the data were available) (Figure 3). The mean Fernandez score was 9,2 for the patients with a successful methotrexate treatment (48 patients), and 10.9 for those with a failure of the medical therapy (18 patients). The difference of Fernandez score between the groups ―success‖ and ―failure‖ was statistically significative (p=0,0002). Surgery after methotrexate failure was necessary for 86 women. Indications (sometimes concurrent) for endoscopy are summarized in Table 2. In the women who needed it, the mean delay to endoscopy was 24.5 (±110.8) days. Among the 86 failures of treatment by methotrexate therapy, 49 patients treated by radical surgery (57%) and 37 patients by conservative surgery (43%).
Analyses Table 3 summarizes the results of univariate analyses in the medical success and failure groups including all patients (one and two methotrexate injections) and Table 4 summarizes these analyses in patients who received two injections of methotrexate. A multivariate analysis was also performed but it identified no statistically significant factors. Figure 4 shows the plotted Receiver Operative Characteristic (ROC) curve and the area under the curve for the evaluation of the prognostic value of D0 hCG level. The cutoff value was set
at 1,000 IU/l, with sensitivity and specificity of 70%. The negative and positive predictive values were respectively 40% and 90%.
Adverse effects In all, 36 (9%) patients experienced adverse effects: digestive disorders (nausea, vomiting, diarrhea) occurred in 18 (4.5%) women, mild hepatic cytolysis in 15 (3.75), thrombocytopenia in 1 (0.25%), headache in 1 (0.25%) and stomatitis in 1 (0.25%). There were no serious adverse events.
Fertility We have contacted 400 patients by phone. We failed to contact 93 patients: i.e. 23% lost to follow-up. For the 307 patients successfully contacted, the 2-year cumulative intrauterine pregnancy rate was 82,1%. We found rates of 80.7% for delivery after 22 weeks of gestation, 20.1% for spontaneous miscarriage, 0.8% for late miscarriage, 4.4% for abortion and 13.2% for ectopic pregnancy.
Comments
A methotrexate single-dose protocol successfully treated 78.5% of the 400 cases of ectopic pregnancy included in our study. Our results align with those of other recent studies where similar single-dose protocols were used in series comprising more than 100 women: Between 2011 and 2014, Rabischong et al [9], Sagiv et al [6], Skubisz et al [14] and Helmy et al [15] reported success rates of respectively 75.4% in a series of 419 patients, 70% in 238 patients, 71% in 206 patients and 81.8% in 198 patients. Patient characteristics in our study and in the above-mentioned studies were also comparable. A history of ectopic pregnancy was noted for 11% of our patients, close to the range of 12.6% to 14% reported in most recent studies [9,16–18]. Our delay to negative hCG was somewhat shorter than those reported in the literature. In our study, 9.75% of the pregnancies were a result of in vitro fertilization, which is comparable to the 7.3% rate reported by Van Mello et al in 2013 [18]. A single methotrexate injection was sufficient for 63.5% of our patients, and a second injection resolved 73.2% of the remaining cases. Factors predictive of failure for the first methotrexate injection were hCG levels at days 0, 2, 4, 7 and 14, blood progesterone level at D0, hematosalpinx at D0 and pain at D7. In contrast, a reduction in hCG between D0 and D2, between D0 and D4 or between D0 and D7 were statistically significant for the prediction of success. Predictive factors for failure after a second injection were the hCG levels at days 0, 2, 4 and 7. As was the case for Cho et al [19] and Helmy et al [15], the presence of clinical signs (pain, metrorrhagia) at diagnosis was predictive of neither methotrexate treatment failure nor its success in our study. The pretreatment hCG level is a known predictor of success for methotrexate treatment, but thresholds reported in the literature vary from 1000 IU/l to as much as 5000 IU/l. In our study, initial hCG values <1000 IU/l were associated with a success rate of 90%. Our threshold value of 1000 IU/l was the same or similar to those retained by several other teams. For example, both Nazac et al [16] and Potter et al [20] retained the 1000 IU/l threshold, and Rabischong et al retained 1300 IU/l for their 2011 study [9]. We also found that the pretherapeutic Fernandez score was a statistically significative predictor of success: there was 80% of success when the score was under 10. In our study, but not in previous studies, hematosalpinx at diagnosis was a predictive factor of methotrexate treatment failure [21]. We also noticed that as the size of echographically visible
lateral extrauterine masses increased so did failure rates, but this association was not statistically significant. This same dynamic has been reported in most studies [6,10,16,20–22]. The predictive value of blood progesterone has received little attention in the literature. Both Nazac et al and Lipscomb et al reported that it had no significant predictive role [10,16]. However, Corsan et al did show, albeit in a small series of 44 patients, that women with blood progesterone >7 ng/ml had a significantly greater risk of methotrexate failure [23]. In our study, the rate of methotrexate failure was significantly higher when blood progesterone was greater than 5 ng/ml. Thus, our study identified pretreatment blood progesterone as a statistically significant predictive factor for methotrexate treatment outcome. Considering its ease and low cost, blood progesterone may be of great interest for daily clinical practice in this setting. Day 2 hCG level has also received little attention in the literature, but our study suggests that it holds at least a hypothetical promise for rapidly identifying women who are at risk of methotrexate treatment failure. For practice however, many authors suggest waiting until D4 to reassay hCG because an initial rise in it is expected in the first few days due to trophoblast lysis [24]. Furthermore, convoking a patient at two days post injection and then again at four days as indicated in most protocols would be impractical. As they are available early in management and easy to interpret, D4 and D7 hCG levels retain their interest as predictive factors for methotrexate efficacy. Equally pertinent and easy to interpret, reductions in hCG levels between D0 and D4 and between D0 and D7 were predictive factors in our study and in several earlier studies [14,17,25–27]. The D0 to D4 hCG reduction is particularly interesting in that it offers early treatment prognosis value and in turn enables a possibility of lightened surveillance for patients likely to have good outcomes or a more speedy second injection for patients who need it (the mean delay to second injection was 8.4 days in our study). In both cases, the cumbersome nature of follow-up—and, one would hope, the number of patients lost to it—could be reduced. In a 2014 study performed by Bottin et al in a series of 91 patients, a 20% decrease in hCG between D0 and D4 was identified as an effective predictor of methotrexate success and thus compatible with a single D30 follow-up visit instead of the normal weekly visit schedule [27]. A D2 control is unnecessary after methotrexate treatment, as we know there is an initial normal rise of the hCG level after the injection and it is not discriminant. Conversely, D4 hCG level is interesting to predict methotrexate outcome, and especially considering the evolution of hCG between D0 and D4. It remains an early predictive factor for outcome of methotrexate treatment.
In our study, pain at D7 was a predictive factor for first-injection methotrexate failure. However, in this setting, pain must be interpreted carefully since it is often present in the normal course of methotrexate treatment, where it may indicate hematoma formation or tubal abortion [24,28]. Because of the complexity of interpreting pain and its subjective nature, we think that this finding has little clinical pertinence. Tolerability of methotrexate was good overall in our study. The rate of adverse effects was 9%, comprising only minor, mostly digestive complaints. The study of fertility after a medical treatment for ectopic pregnancy found a rate of 80.7% for delivery after 22 weeks of gestation and 13.2% for recurrence ectopic pregnancy. These results were quite congruent with those of the Demeter study, which found a rate of intrauterine pregnancy of 67% and a recurrence of ectopic pregnancy of 12 % after medical treatment. Our fertility rate with a delivery outcome was even a little bit higher than expected (80.7%) [29]. The main strength of the study that we report here is the size of our patient series, one of the largest to date, second only to that of Rabischong et al published in 2011 [9]. We also present here recent and applicable data. The main limit of our study is that it is retrospective: although it reflects our clinical practice experience, our study could be affected by certain biases and limits innate to this type of study.
Conclusion
For the present study, we evaluated a large series of women to identify predictive factors for the outcome of methotrexate treatment in ectopic pregnancy. We confirmed that an initial hCG level <1,000 IU/l and a Fernandez score under 11 were predictive of success, and showed that D0 blood progesterone >5 ng/ml was predictive of failure. Considering the ease and low cost of blood progesterone testing, this latter finding merits further confirmatory testing in our opinion. While not strictly decisional, our hCG threshold can be used to inform patients on the likelihood of methotrexate success or failure. The kinetics of hCG between D0 and D4 and between D0 and D7 must be carefully monitored in the days following the initial injection of methotrexate. Tolerability was good in our study and thus should not be an issue when medical treatment with methotrexate is indicated for ectopic pregnancy.
Acknowledgments Supported by a grant from the Bibliothèque scientifique de l’Internat de Lyon and les Hospices Civils de Lyon
Conflict of interest None
Funding None
Table 1. Baseline characteristics of 400 women treated by methotrexate for ectopic pregnancy. Table 2. Indications of surgery in 86 patients with methotrexate (MTX) failure. Table 3. Univariate analysis comparing success and failure of methotrexate treatment in 400 patients with ectopic pregnancy. Table 4. Univariate analysis comparing success and failure of the second injection of methotrexate (MTX) in patients with ectopic pregnancy. Figure 1. Treatment flow chart. Figure 2. Sucess and failure rates for methotrexate (MTX) treatment according to pretherapeutic HCG levels. Figure 3. Sucess and failure rates for methotrexate (MTX) treatment according to Fernandez score. Figure 4. ROC Curve for the evaluation of the prognostic value of D0 hCG level (IU/L) Area under the curve: 75.27% ± 3.07% (95% confidence interval : 69.24% - 81.29%)
References
[1]
Bouyer J, Coste J, Fernandez H, Pouly JL, Job-Spira N. Sites of ectopic pregnancy: a
10 year population-based study of 1800 cases. Hum Reprod 2002;17:3224–30. doi:10.1093/humrep/17.12.3224. [2]
American College of Obstetricians and Gynecologists. ACOG Practice Bulletin No.
94: Medical management of ectopic pregnancy. Obstet Gynecol 2008;111:1479–85. doi:10.1097/AOG.0b013e31817d201e. [3]
Khan KS, Wojdyla D, Say L, Gülmezoglu AM, Van Look PFA. WHO analysis of
causes of maternal death: a systematic review. Lancet Lond Engl 2006;367:1066–74. doi:10.1016/S0140-6736(06)68397-9. [4]
Lesavre M, Curinier S, Capmas P, Rabischong B, Fernandez H. [Treatment of tubal
ectopic pregnancy by methotrexate].
Gyn cologie
bst trique Biol eprod
–9.
doi:10.1016/j.jgyn.2014.12.013. [5]
Marret H, Fauconnier A, Dubernard G, Misme H, Lagarce L, Lesavre M, et al.
[Evidence-based evaluation and expertise of methotrexate off label use in gynaecology and obstetrics wor of the
G F
Gyn cologie
bst trique Biol
eprod
–6.
doi:10.1016/j.jgyn.2014.12.008. [6]
Sagiv R, Debby A, Feit H, Cohen-Sacher B, Keidar R, Golan A. The optimal cutoff
serum level of human chorionic gonadotropin for efficacy of methotrexate treatment in women with extrauterine pregnancy. Int J Gynaecol Obstet Off Organ Int Fed Gynaecol Obstet 2012;116:101–4. doi:10.1016/j.ijgo.2011.09.023. [7]
Stovall TG, Ling FW. Single-dose methotrexate: an expanded clinical trial. Am J
Obstet Gynecol 1993;168:1759–1762; discussion 1762–1765. [8]
Varma R, Gupta J. Tubal ectopic pregnancy. BMJ Clin Evid 2012;2012.
[9]
Rabischong B, Tran X, Sleiman AA, Larraín D, Jaffeux P, Aublet-Cuvelier B, et al.
Predictive factors of failure in management of ectopic pregnancy with single-dose methotrexate: a general population-based analysis from the Auvergne Register, France. Fertil Steril 2011;95:401–404, 404.e1. doi:10.1016/j.fertnstert.2010.08.025.
[10]
Lipscomb GH, McCord ML, Stovall TG, Huff G, Portera SG, Ling FW. Predictors of
success of methotrexate treatment in women with tubal ectopic pregnancies. N Engl J Med 1999;341:1974–8. doi:10.1056/NEJM199912233412604. [11]
CNGOF. Collège National des Gynécologues et Obstétriciens Français. Guidelines
for clinical practice Ectopic pregnancy management
Gyn cologie bst trique Biol eprod
2003;32:3S6–112. [12]
Fernandez H, Lelaidier C, Thouvenez V, Frydman R. The use of a pretherapeutic,
predictive score to determine inclusion criteria for the non-surgical treatment of ectopic pregnancy. Hum Reprod Oxf Engl 1991;6:995–8. [13]
Fernandez H, Bourget P, Ville Y, Lelaidier C, Frydman R. Treatment of unruptured
tubal pregnancy with methotrexate: pharmacokinetic analysis of local versus intramuscular administration. Fertil Steril 1994;62:943–7. [14]
Skubisz M, Dutton P, Duncan WC, Horne AW, Tong S. Using a decline in serum hCG
between days 0-4 to predict ectopic pregnancy treatment success after single-dose methotrexate: a retrospective cohort study. BMC Pregnancy Childbirth 2013;13:30. doi:10.1186/1471-2393-13-30. [15]
Helmy S, Bader Y, Pablik E, Tiringer D, Pils S, Laml T, et al. Cut-off value of initial
serum β-hCG level predicting a successful MTX therapy in tubal ectopic pregnancy: a retrospective cohort study. Eur J Obstet Gynecol Reprod Biol 2014;179:175–80. doi:10.1016/j.ejogrb.2014.05.033. [16]
Nazac A, Gervaise A, Bouyer J, de Tayrac R, Capella-Allouc S, Fernandez H.
Predictors of success in methotrexate treatment of women with unruptured tubal pregnancies. Ultrasound Obstet Gynecol Off J Int Soc Ultrasound Obstet Gynecol 2003;21:181–5. doi:10.1002/uog.9. [17]
Skubisz MM, Li J, Lee J, Wallace EM, Tong S. Decline in βhCG levels between days
0 and 4 after a single dose of methotrexate for ectopic pregnancy predicts treatment success: a retrospective
cohort
study.
BJOG
Int
J
Obstet
Gynaecol
2011;118:1665–8.
doi:10.1111/j.1471-0528.2011.03133.x. [18]
Van Mello NM, Mol F, Verhoeve HR, van Wely M, Adriaanse AH, Boss EA, et al.
Methotrexate or expectant management in women with an ectopic pregnancy or pregnancy of
unknown location and low serum hCG concentrations? A randomized comparison. Hum Reprod Oxf Engl 2013;28:60–7. doi:10.1093/humrep/des373. [19]
Cho GJ, Lee SH, Shin JW, Lee NW, Kim T, Kim HJ, et al. Predictors of success of
repeated injections of single-dose methotrexate regimen for tubal ectopic pregnancy. J Korean Med Sci 2006;21:86–9. doi:10.3346/jkms.2006.21.1.86. [20]
Potter MB, Lepine LA, Jamieson DJ. Predictors of success with methotrexate
treatment of tubal ectopic pregnancy at Grady Memorial Hospital. Am J Obstet Gynecol 2003;188:1192–4. [21]
Krissi H, Peled Y, Eitan R, Bishara A, Goldchmit C, Ben-Haroush A. Single-dose
methotrexate injection for treatment of ectopic pregnancy in women with relatively low levels of human chorionic gonadotropin. Int J Gynaecol Obstet Off Organ Int Fed Gynaecol Obstet 2013;121:141–3. doi:10.1016/j.ijgo.2012.11.020. [22]
Gnisci A, Stefani L, Bottin P, Ohannessian A, Gamerre M, Agostini A. Predictive
value of hemoperitoneum for outcome of methotrexate treatment in ectopic pregnancy: an observational comparative study. Ultrasound Obstet Gynecol Off J Int Soc Ultrasound Obstet Gynecol 2014;43:698–701. doi:10.1002/uog.13255. [23]
Corsan GH, Karacan M, Qasim S, Bohrer MK, Ransom MX, Kemmann E.
Identification of hormonal parameters for successful systemic single-dose methotrexate therapy in ectopic pregnancy. Hum Reprod 1995;10:2719–22. [24]
Luciano AA, Roy G, Solima E. Ectopic Pregnancy. Ann N Y Acad Sci 2001;943:235–
54. doi:10.1111/j.1749-6632.2001.tb03805.x. [25]
Agostini A, Blanc K, Ronda I, Romain F, Capelle M, Blanc B. Prognostic value of
human chorionic gonadotropin changes after methotrexate injection for ectopic pregnancy. Fertil Steril 2007;88:504–6. doi:10.1016/j.fertnstert.2006.11.138. [26]
Nowak-Markwitz E, Michalak M, Olejnik M, Spaczynski M. Cutoff value of human
chorionic gonadotropin in relation to the number of methotrexate cycles in the successful treatment
of
ectopic
pregnancy.
Fertil
Steril
2009;92:1203–7.
doi:10.1016/j.fertnstert.2008.07.1775. [27]
Bottin P, Gnisci A, Crochet P, Butzbach P, Cravello L, Gamerre M, et al. [Prognostic
value of early hCG changes after methotrexate injection for ectopic pregnancy Gyn cologie bst trique Fertil [28]
–7. doi:10.1016/j.gyobfe.2013.08.006.
Lipscomb GH, Puckett KJ, Bran D, Ling FW. Management of separation pain after
single-dose methotrexate therapy for ectopic pregnancy. Obstet Gynecol 1999;93:590–3. doi:10.1016/S0029-7844(98)00523-7. [29]
Fernandez H, Capmas P, Lucot JP, Resch B, Panel P, Bouyer J, et al. Fertility after
ectopic pregnancy: the DEMETER randomized trial. Hum Reprod 2013;28:1247–53. doi:10.1093/humrep/det037.
Table 1. Baseline characteristics of 400 women treated by methotrexate for ectopic pregnancy. Successful MTX treatment (n=314) Age (years) ≤24 28 (9) 25-29 91 (29) 30-34 114 (36) >34 81 (26) Gravidity 2.6 ± 1.6 Parity 0.8 ± 1.0 Previous ectopic pregnancy 33 (10) Gestational age*, days 43.1 ± 12.6 Clinical symptoms* Metrorrhagia 93 (30) Pelvic pain 49 (516) Metrorrhagia and pain 130 (42) Hematosalpinx* 182 (60) <10 mm 20 (6) 10-30 mm 151 (50) >30 mm 11 (4) D0 hCG level 862 ± 1073 <500 179 (57) 500-1499 75 (24) ≥1500 60 (19) D0 blood progesterone (ng/mL) Rate 4.7 ± 5.0 <5 38 (68) Contraception** 29 (9) Spontaneous pregnancy 275 (88) Data are shown as mean ± SD or n (%); significant scores with p < 0.05 are bolded D0 : Day of the first methotrexate injection hCG : human chorionic gonadotropin, IU/L *Data collected on D0 **Before the diagnosis of pregnancy Successful treatment: no surgery needed Failure of treatment: necessity of laparoscopy after the MTX treatment
Failure of MTX treatment (n=86) 16 (19) 27 (31) 23 (27) 20 (23) 2.5 ± 1.7 0.7 ± 0.9 11 (13) 42.3 ± 11.7 23 (27) 13 (15) 39 (45) 61 (73) 4 (5) 51 (61) 6 (7) 1972 ± 1499 15 (18) 23 (27) 46 (55) 8.0 ± 8.7 8 (42) 11 (58) 5 (6) 71 (82)
p
0.05 0.51 0.26 0.55 0.64
0.93 0.03
0.08 <0.0001
<0.0001 0.047 0.046 0.30 0.17
Table 2. Indications of surgery in 86 patients with methotrexate (MTX) failure.
n Clinical signs suggestive of tubal rupture (severe abdominal pain, abdominal guarding, faintness)
60
Occurrence of echographic signs (occurrence or increase of hematosalpinx or hemoperitoneum, appearance of an embryo with cardiac activity)
59
Poor hCG course (increase in hCG level of >30% between Day (D) 0 and D4, hCG level at D7 greater than that of D0, or secondary rebound of the hCG level)
24
Falling hemoglobin levels
2
Refusal of the second MTX injection
1
Table 3. Univariate analysis comparing success and failure of methotrexate treatment in 400 patients with ectopic pregnancy
D2 hCG level D4 hCG level D7 hCG level D14 hCG level D2 hCG
Successful MTX treatment (n=314) 908 ± 1266 802 ± 1252 530 ± 952 145 ± 225 131 (53) 185 (65) 220 (81) 81 (35) 68 (26) 51 (21) 32 (11)
Data are shown as mean ± SD or n (%) ; significant scores with p < 0.05 are bolded D0 : Day of the methotrexate injection hCG : human chorionic gonadotropin, IU/L Successful treatment: no surgery needed Failure of treatment: necessity of laparoscopy after the MTX treatment
Failure of MTX treatment (n=86) 2558 ± 2247 2553 ± 2401 1579 ± 1990 795 ± 869 15 (24) 16 (27) 24 (63) 26 (49) 21 (37) 17 (47) 24.5 ± 110.8 4 (6)
p <0.0001 <0.0001 <0.0001 <0.0001 <0.0001 <0.0001 0.009 0.09 0.12 0.0005 0.17
Table 4. Univariate analysis comparing success and failure of the second injection of methotrexate (MTX) in patients with ectopic pregnancy
D0 hCG D2 hCG D4 hCG D7 hCG D14 hCG hCG drop D2-D0 > 500 hCG drop D4-D0 > 500 hCG drop D7-D0 > 500 Delay between first and second injections, days D0 blood progesterone (ng/mL)
Successful MTX treatment (n=60) 1261 ± 1426 1147 ± 1394 910 ± 1085 581 ± 745 168 ± 219 38 (90) 34 (67) 30 (59) 9.25 ± 12.6
Failure of MTX treatment (n=22) 3045 ± 2536 2530 ± 2350 1901 ± 1857 2191 ± 1499 636 ± 646 13 (72) 7 (50) 0 (0) 6.05 ± 1.79
4.26 ± 4.9
4 ± 4.67
Data are shown as mean ± SD or n (%) ; significant scores with p < 0.05 are bolded D0 : Day of the second methotrexate injection hCG : human chorionic gonadotropin, IU/L Successful treatment: no surgery needed Failure of treatment: necessity of laparoscopy after the MTX treatment
p 0.0001 0.006 0.01 <0.0001 0.002 0.11 0.25 0.008 0.24 0.94
Fig 1
Fig 2
Fig 3
Fig 4