Comparison of single and multiple dose methotrexate therapy for unruptured tubal ectopic pregnancy: a prospective randomized study
Abstract
Objective. To compare the success rates of single and multiple dose methotrexate protocols for the treatment of unruptured tubal ectopic pregnancy. Design. Prospective randomized controlled trial. Setting. Maternity and teaching hospital in Turkey. Population. One hundred twenty women treated with methotrexate therapy for unruptured tubal ectopic pregnancy. Methods. Sixty-two women received a single dose and 58 received a multiple dose methotrexate regimen. Main outcome measures. Success rate of methotrexate therapy (women successfully treated with one injection and women who completed four doses). Results. In the single dose group, treatment was considered successful in 50 women (80.6%), whereas in the multiple dose group, 52 women (89.7%) responded to treatment (p = 0.21; OR 0.90, 95%CI 0.77–1.05). The average number of days required for human chorionic gonadotropin (hCG) levels to fall below 5 mU/mL was longer in the single dose (22.3 ± 7.6) compared with the multiple dose group (18.3 ± 10.7) (p = 0.03). In the single dose group fewer or 17 women (24.7%) experienced side-effects compared to 28 (48.3%) of those who had multiple doses (p = 0.02, OR 0.57, 95%CI 0.35–0.92). Conclusion. A multiple dose methotrexate regimen for the treatment of unruptured tubal ectopic pregnancy is not more effective than a single dose one. In addition, multiple doses may cause more side-effects, but the time for hCG levels to fall below 5 mU/mL is shorter.
Introduction
The incidence of ectopic pregnancy (EP) has in general increased in recent decades to the rate of approximately 2% of all pregnancies (1). In the United States the increase was from 4.5/1,000 reported pregnancies in 1970 to 20/1,000 pregnancies in 1992 (1). Conversely, in France, the overall EP rate decreased by 2%, from 96.4/100,000 women aged 15–44 in 1992 to 95.3/100,000 in 2002. However, the rate of ‘reproductive failure’ increased by 17% (2). The rising tendency in the USA is thought to have been related to the increased incidence of pelvic inflammatory disease and the use of assisted reproductive technologies. The prevalence of EP among women admitted to hospital with first trimester bleeding, pain, or both, has ranged from 6 to 16% (3). Measurement of human chorionic gonadotropin (hCG) by sensitive quantitative hCG assays and utilization of high resolution transvaginal ultrasound now enable the early detection of EP (4–6). Accurate diagnosis is essential for operative, medical or expectant management. Early diagnosis increases the chance to apply conservative treatment options that may retain tubal function (7, 8).
Medical treatment of EP with methotrexate is an important step in the management of this serious condition. The success rate has been reported as between 75 and 96% in properly selected patients (9). Methotrexate disturbs the de novo synthesis of purine and pyrimidine by competitively inhibiting the binding of dihydrofolic acid to the dihydrofolate reductase enzyme. Thus reduction to folinic acid, an important cofactor in cell growth pathways, is inhibited (10).
Systemic methotrexate therapy is the most commonly used form of therapeutic administration. Although there is no consensus regarding which method is best, it can be administered as single or multiple dose regimen. Recently a new dose regimen, called the ‘2 dose’ protocol, to minimize the number of injections and surveillance visits has also been suggested (11). Single dose is most frequently used since it requires fewer hospital visits and has fewer side-effects, but a recent meta-analysis revealed no significant differences between single and multiple dose success rates (12). Barnhart et al. showed that a single dose was associated with higher treatment failure rates in their systematic analysis (9). Different success rates may be attributed to clinicians' tendency to choose multiple dose treatment for higher hCG levels (9, 13). Recently, Alleyassin et al. showed similar success rates between multiple and single doses in a randomized study (14). The purpose of this prospective randomized study was to compare the success rates of single and multiple dose methotrexate protocols for the treatment of unruptured tubal ectopic pregnancy.
Material and methods
A total of 211 consecutive women admitted and hospitalized with a suspected ectopic tubal pregnancy to the Etlik Women's Health and Maternity Teaching and Research Hospital for observation as part of hospital policy were enrolled, during the period extending from February 2005 to December 2005. Institutional Review Board approval was obtained for the study.
Initially, all women were evaluated by transvaginal ultrasound and measurement of hCG and progesterone levels, and those diagnosed with EP were recruited. EP was diagnosed in 190 women based on positive visualization of an adnexal mass on transvaginal ultrasound (i.e. if one of the following three gray-scale findings were present: a heterogeneous mass or ‘blob sign’ adjacent to and moving separately from the ovary; a mass with a hyperechoic ring around the gestational sac (bagel sign); a gestational sac with a fetal pole with cardiac activity, i.e. a viable extra-uterine pregnancy) (15), combined with an abnormal serum hCG rise (<50% rise over 2 consecutive days) or plateau (<15% decline within 3 days) and an abnormal serum progesterone level (<20 ng/mL) (5, 16, 17). The mean age, gravida, and parity of the 190 women were 30.9 ± 6.0 (range 16–45), 2.9 ± 1.7 (range 0–9), and 1.3 ± 1.1 (range 0–5), respectively. The gold standard for the diagnosis of EP was the combination of positive transvaginal ultrasound findings for EP and an abnormal serum hCG rise or plateau, and an abnormal serum progesterone level (5, 15–17).
The inclusion criteria for methotrexate treatment were as follows: hemodynamic stability, serum hCG levels reaching a plateau or increased by ≤50% in 48-hour intervals, detection of an adnexal mass 3.5 cm or less in diameter, no history of previous tubal surgery, the patient's desire for a future pregnancy, and willingness to participate in the study. Patients who were clinically unstable (i.e. not hemodynamically stable, n = 49), who had hepatic or renal diseases (n = 1), or who were not willing to participate (n = 20) were excluded (18). The CONSORT statement flow diagram is given in Figure 1.
CONSORT statement flow diagram.
Patients satisfying the selection criteria (n = 120) were prospectively randomized into two groups as Group I (treated with single intramuscular dose of 50 mg/m2 methotrexate n = 62) and Group II (treated with multiple dose methotrexate therapy, n = 58). In the follow-up period, we obtained hCG levels 0, 4, and 7 days after the methotrexate injection. Patients with a decline of 15% in hCG levels between days 4 and 7 were accepted as having a positive response and monitored on an outpatient basis weekly until their hCG levels were below 5 mU/mL. A repeat dose of methotrexate was given if hCG levels did not fall 15% between days 4 and 7 after the initial dose or if an hCG level fall of 15% was not observed during the subsequent weekly hCG level follow-up (13). Group II was treated with methotrexate 1 mg/kg/day intramuscularly on days 1, 3, 5, and 7, and leucovorin (0.1 mg/kg) on days 2, 4, 6, and 8). In the follow-up period, serum hCG levels were obtained at baseline (day 0), and on days 1, 3, 5, and 7 until hCG declined 15% from the previous value. A positive response was defined as an hCG level decrease of 15% in 48 hours or after four doses of methotrexate were given (14). Patients with positive response rates were regarded as successfully treated. In all women, subsequent weekly hCG levels were obtained. In all successfully treated women the time in days until the levels of hCG became less than 5 mU/mL was assessed. Only successfully treated patients (women who needed only one methotrexate injection in group I and women who completed the four doses of methotrexate injections in group II) were considered successfully treated (hCG < 5 mU/mL). In the final analysis, women who needed repeat doses of methotrexate in the single dose group or additional doses in the multiple dose group (10 women) or who needed surgical therapy (either laparoscopy or laparotomy, n = 8) were considered unsuccessfully treated.
Randomization
Computer assisted randomization was used based on the instructions at http://www.randomization.com (the generator randomizes each subject to a single treatment using the method of randomly permuted blocks). The study team was not blinded to the randomized group, at least until the randomization was carried out. Women were allocated randomly to the groups on the first treatment day following the diagnosis of EP and evaluation for inclusion criteria by one of the authors, using sealed envelopes.
Statistical analysis
On the basis of previous study results (9) and the estimated success rates of single dose therapy (the lowest success rate, 75%; proportion positive for group I) and multiple dose therapy (the highest success rate, 96%; proportion positive for group II) as the primary end point, a sample size of 58 patients in each group would have a 90.5% power to detect a difference in success rate of 21.0%, using a Z-test with pooled variance, and α = 0.05 two-sided significance level (using the Power and Precision (tm) program, Biostat, USA).
The results are given as proportions and mean ± standard deviations. Fisher's exact chi-squared test, Student's t-test, and Mann–Whitney U test were used for statistical analysis using SPSS software, version 12.0 for Windows. Statistical significance was set at p < 0.05.
Results
The final analysis included 120 women, with a mean ± SD age of 30.2 ± 6.1 years (range 16–45), gravidity of 2.7 ± 1.5 (range 0–8), parity 1.2 ± 1.0 (range 0–4), and gestational age 52.9 ± 15.4 (range 30–112) days. The main complaint on admission was pelvic pain and abnormal vaginal bleeding with a missed period (74.2% of cases). Pelvic pain with an abnormal vaginal bleeding and pelvic pain with a missed period were observed in 9.2 and 15.8% of women, respectively. Only 0.8% of women experienced only abnormal vaginal bleeding with a missed period.
There were no significant differences between groups I and II with regard to demographic, clinical, and laboratory characteristics (Table 1).
| Characteristics | Group I (n = 62) | Group II (n = 58) |
|---|---|---|
| Mean age (years) | 30.4 ± 6.2 | 30.0 ± 6.1 |
| Mean gravida (number) | 2.8 ± 1.6 | 2.7 ± 1.3 |
| Mean parity (number) | 1.2 ± 1.1 | 1.2 ± 0.9 |
| Mean gestational age (day) | 52.4 ± 15.3 | 53.5 ± 15.7 |
| Mean hCG on admission (IU/L) | 1,973.0 ± 2,494.3 | 2,310.9 ± 3,183.6 |
| Mean progesterone on admission (ng/mL) | 4.6 ± 11.0 | 4.3 ± 4.4 |
| Mean size of mass on sonograph (mm) | 25.5 ± 10.3 | 23.1 ± 11.8 |
| Findings on sonograph (%) | ||
| A mass with a hyperechoic ring around the gestational sac (bagel sign) | 23 (37.1%) | 27 (46.6%) |
| A heterogeneous mass or ‘blob sign’ adjacent to and moving separately from the ovary | 38 (61.3%) | 28 (48.3%) |
| Patients with positive cardiac activity | 1 (1.6%) | 3 (5.2%) |
- Note: Data are expressed as mean ± standard deviation or number of patients and percentages in parentheses. There were no statistically significant differences between group I and II with regard to clinical and laboratory characteristics.
The medical treatment was successful in 102 patients (85%). Of the 62 patients in the single dose group, treatment was considered successful in 50 (80.6%). Of the 58 patients in the multiple dose group, 52 were considered to have responded to the treatment (89.7%). Comparison of the success rates in the two groups revealed non-significant differences (p = 0.21; OR 0.90, 95% CI 0.77–1.05). The comparison of the clinical (age, gravidity, parity, gestational age, and side-effects) or laboratory (hCG, progesterone on admission, and size of mass on ultrasonography) population profile for the ‘failed medical treatment’ single (group I) and multiple dose (group II) methotrexate therapy of unruptured ectopic pregnancy revealed no significant differences.
The subgroup analyses of possible factors affecting the success rates in groups I and II are outlined in Table 2. The success rates in all age categories, gravidity, gestational age, size of mass on ultrasonography, transvaginal ultrasound findings, and hCG values on admission were higher in the single dose group than in the multiple dose group, although significant values were not obtained. Furthermore, the success rate in patients with low parity (≤1) or lower progesterone values on admission (≤2) was higher in the multiple dose than in the single dose group (84.2 vs. 88.6%, p = 0.42 for low parity, 69.0 vs. 90.3%, p = 0.04 for lower progesterone value). The only significant value was obtained for the comparison of lower progesterone values on admission (≤2).
| Characteristics | Group I (n = 62) | Group II (n = 58) | Absolute difference with a 95% CI | p |
|---|---|---|---|---|
| Age (years) | ||||
| ≤28 (n = 42) | 18 (85.7%) | 18 (85.7%) | 0.0% (−22.4 to 22.4%) | 0.34a |
| >28 (n = 78) | 32 (78.0%) | 34 (91.9%) | −13.9% (−29.5 to 2.7%) | 0.08a |
| Gravida (number) | ||||
| ≤2 (n = 59) | 25 (80.6%) | 25 (89.3%) | −8.6% (−26.3 to 10.7%) | 0.29a |
| >2 (n = 61) | 25 (80.6%) | 27 (90.0%) | −9.4% (−27.5 to 9.3%) | 0.25a |
| Parity (number) | ||||
| ≤1 (n = 73) | 32 (84.2%) | 31 (88.6%) | −4.4% (−20.5 to 12.4%) | 0.42a |
| >1 (n = 47) | 18 (75.0%) | 21 (91.3%) | −16.3% (−37.2 to 6.0%) | 0.14a |
| Gestational age (day) | ||||
| ≤48 (n = 58) | 26 (81.3%) | 23 (88.5%) | −7.2% (−25.4 to 12.8%) | 0.35a |
| >48 (n = 62) | 24 (80.0%) | 29 (90.6%) | −10.6% (−29.0 to 7.6%) | 0.21a |
| hCG on admission (IU/L) | ||||
| ≤800 (n = 54) | 23 (85.2%) | 25 (92.6%) | −7.4% (−25.7 to 10.9%) | 0.33a |
| >800 (n = 66) | 27 (77.1%) | 27 (87.1%) | −10.0% (−27.9 to 9.3%) | 0.24a |
| Progesterone on admission (ng/mL) | ||||
| ≤2 (n = 60) | 20 (69.0%) | 28 (90.3%) | −21.4% (−40.6 to 0.08%) | 0.04a |
| >2 (n = 60) | 30 (90.9%) | 24 (88.9%) | 2.0% (−14.2 to 20.0%) | 0.56a |
| Size of mass on sonograph (mm) | ||||
| ≤ 20 (n = 54) | 17 (77.3%) | 29 (90.6%) | −13.3% (−35.0 to 6.1%) | 0.17a |
| > 20 (n = 66) | 33 (82.5%) | 23 (88.5%) | −6.0% (−22.3 to 13.6%) | 0.39a |
| Findings on sonograph (%) | ||||
| A mass with a hyperechoic ring around the gestational sac (bagel sign) (n = 50) | 21 (91.3%) | 26 (96.3%) | −5.0% (−23.4 to 10.9%) | 0.44a |
| A heterogeneous mass or ‘blob sign’ adjacent to and moving separately from the ovary (n = 66) | 29 (76.3%) | 23 (82.1%) | −5.8% (−24.3 to 14.9%) | 0.40a |
| Patients with positive cardiac activity (n = 4) | 0 (0.0%) | 3 (100%) | – | – |
- aFisher exact chi-squared test was used.
- Note: Data are expressed as number of patients and percentages in parentheses.
Of the 12 patients in whom treatment failed in the single dose group, eight received a second dose (50 mg/m2) and the remaining four women underwent surgical treatment (either laparoscopy or laparotomy). Only six of the eight women in whom single dose treatment failed, responded to a second dose, whereas one patient in whom the second dose treatment failed responded to a third dose (50 mg/m2). The remaining woman, in whom the second dosed treatment failed, underwent surgery directly. Therefore, 9.6% of the women required more than one dose of the drug. Of the six women in whom multiple dose treatment failed, four required surgery after the multiple dose methotrexate regimen failed. The remaining two patients responded to one more dose of methotrexate (1 mg/kg methotrexate per day intramuscularly), not the full multiple dose regimen. In the multiple dose group, only 3.5% of the women required more than one dose of the drug.
The mean number of days until the hCG level dropped below 5 mU/mL was significantly greater in the single dose group (22.3 ± 7.6) compared to the multiple dose group (18.3 ± 10.7). The mean difference was 4.0 ± 1.8 (p = 0.03, 95%CI 0.33–7.65, using Student's t-test).
For both groups only the clinically significant side-effects during the initial treatment were included. Side-effects were seen in 45 patients (37.5%) and consisted of abdominal pain, nausea, diarrhea, elevated liver enzymes (serum AST or ALT > 40 IU/L), stomatitis, and dermatitis. In the single dose and multiple dose groups, 17 (27.7%) and 28 (48.3%) patients, respectively, experienced side-effects (p = 0.02, OR 0.57, 95%CI 0.35–0.92, using Fisher's exact chi-squared test).
Discussion
This prospective randomized study focused on a comparison of the success rate of single and multiple dose methotrexate regimens for unruptured tubal ectopic pregnancy, since there is no true consensus regarding which protocol, surveillance, and frequency of administration should be used. Our results suggest that the multiple dose regimen is not superior to the single dose one as a conservative treatment modality for EP, and they indicate a higher rate of side-effects but a shorter number of days for hCG levels to drop below 5 mU/mL in the multiple dose regimen.
The single dose administration of methotrexate is a more practical approach to the multiple dose one (19), although Barnhart et al. found that a single dose was associated with a high failure rate (9). By contrast, in a recent retrospective analysis, Lipscomb et al. stated that there was no significant difference between the two protocols (8). These results were attributed to the clinicians' tendency to choose a single dose for the patients with good prognosis while reserving the multiple dose for patients with a high probability of failure. Moreover, they stated that in meta-analysis the patient population of the multiple dose regimen is from an earlier era of medical treatment and so the patients were chosen strictly; thus the multiple dose treatment could have appeared successful. These selection criteria may have influenced the results of the meta-analysis in the multiple dose regimen.
Since there is disagreement in the literature about the most efficient regimen, randomized blinded clinical trials are needed to reduce potential confounders and biases. A randomized clinical trial designed by Alleyasin et al. revealed no significant differences between the two treatment regimens and concluded that there was a need for larger clinical trials (14). Their study implies that single dose treatment should be the first line treatment in selected group of patients (14). A recent Cochrane review also reported that there were no significant differences between the two treatment regimens in terms of primary treatment success rates (20). The current prospective randomized study may contribute to this ongoing debate since it includes more patients in order to reduce statistical errors. The overall methotrexate success rate was 85% for our study. In the literature, the overall success rate of methotrexate therapy is 75–96% independent of the method of administration. In our analysis, the success rates of the single dose regimen (80.6%) and multiple dose regimen (89.7%) did not differ significantly, which was consistent with the results of two previous studies (8, 14), although a rising tendency was observed in the success rate of the multiple dose regimen compared with the single dose one.
The results of our study suggest that the multiple dose regimen is more effective in disrupting the well organized trophoblastic proliferation. This may explain the shorter detriment time for the multiple dose treated group. The main limitations of the multiple dose regimen is the need for leucovorin rescue treatment, the high frequency of side-effects, and the requirement of intensive monitoring. By contrast, the single dose is practical in application but the term ‘single dose’ is a misnomer since 11–20.6% of patients require a second dose (8, 9, 13, 14, 21). In our study only about 9.1% of patients in the single dose group and 3.5% of those in the multiple dose group required a second dose. The optimal and practical treatment of choice could be other than one dose administration.
In a recent study a new methotrexate regimen was proposed (11). To increase the success rate of the single dose regimen, the two dose regimen was also reported to be a reasonable choice, since it optimized convenience and efficiency for patients, and only 12% of patients were reported to need an additional methotrexate dose. In this study a total of 88 women (88/101; 87.1%) were treated successfully with medical treatment. A total of 73 women with the success rate of 72.3% needed one course (2 doses) of methotrexate and a total of 55 women (53%) reported 160 adverse events during the course of treatment. In our study, a similar success rate of medical treatment of EP and a lower adverse events rate are reported. This recent study also highlights the effectiveness of additional methotrexate administration and the need for new methotrexate regimens similar to the multiple dose one (11).
The aim of the current study was to examine whether it was appropriate to choose multiple dose therapy for all patients when there was a chance to offer only one dose. This randomized clinical trial was designed to help clinicians in deciding whether to choose a single dose or multiple doses; thus the treatment failure rate was assessed. It was demonstrated that the rate of failure of the single dose treatment (19.3%) was higher than that of the multiple dose regimen (10.3%). In spite of the failure rate demonstrated by an OR of 4.75(CI 1.77–12.62) in a meta-analysis (9), in a randomized controlled trial it was noted that single dose therapy has a high failure rate (14). Our results also support these findings. Based on this, it should be emphasized that single dose therapy could be a first line of treatment in well selected patients.
Potentially life-threatening nephrotoxicity, hepatotoxicity, pulmonary damage, and myelosuppression are sometimes seen with the use of methotrexate and can occur with either high-dose therapy as used for malignant diseases or with low-dose therapy as used in other disorders such as rheumatoid arthritis and psoriasis (22). Adverse reactions to methotrexate for EP are usually mild and self-limited. The most common are stomatitis and conjunctivitis. Rare side-effects include gastritis, enteritis, dermatitis, pneumonitis, alopecia, elevated liver enzymes, and bone marrow suppression. Approximately a third of patients in the single dose protocol will have side-effects; this rate is lower than with multiple dose regimens (41.2%) (9). These figures are based on analysis of retrospective study results. Our prospective randomized study also emphasized the high side-effects rate in the multiple dose group compared with the single dose one. This may be attributed to the high cumulative dose of methotrexate used in the multiple dose regimen.
According to the current results, it is not easy to define the target population to benefit from the single dose regimen as the first line of treatment, since the subgroup analysis of factors affecting the success rates in both groups revealed no clinical and laboratory parameters predicting the success rate in group I.
Our findings do not suggest the superiority of the multiple dose methotrexate regimen over the single dose one. One of the limitations of this study was the absence of blinding. The other one is that the power calculation was based on a large anticipated difference. This study highlights the need for further research before deciding on which medical treatment is best for unruptured tubal ectopic pregnancy, as multiple dose methotrexate therapy had a significantly higher rate of side-effects and shorter number of days for hCG levels to drop below 5 mU/mL.
Acknowledgments
Declaration of interest: The authors report no conflicts of interest. The authors alone are responsible for the content and writing of the paper.
