- Email: [email protected]
Predictive factors of relapse in low-risk gestational trophoblastic neoplasia patients successfully treated with methotrexate alone
Accepted Manuscript Predictive factors of relapse in low-risk gestational trophoblastic neoplasia patients successfully treated with methotrexate alone F. Couder, J. Massardier, B. You, F. Abbas, T. Hajri, J.P. Lotz, A.M. Schott, F. Golfier PII:
S0002-9378(16)00233-7
DOI:
10.1016/j.ajog.2016.01.183
Reference:
YMOB 10913
To appear in:
American Journal of Obstetrics and Gynecology
Received Date: 14 September 2015 Revised Date:
13 January 2016
Accepted Date: 22 January 2016
Please cite this article as: Couder F, Massardier J, You B, Abbas F, Hajri T, Lotz J, Schott A, Golfier F, Predictive factors of relapse in low-risk gestational trophoblastic neoplasia patients successfully treated with methotrexate alone, American Journal of Obstetrics and Gynecology (2016), doi: 10.1016/ j.ajog.2016.01.183. 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.
1
ACCEPTED MANUSCRIPT
TITLE
Predictive factors of relapse in low-risk gestational trophoblastic neoplasia patients successfully treated with methotrexate alone
AUTHORS
RI PT
F COUDER a, J MASSARDIER b,c , B YOU b,d , F ABBAS e, T HAJRI b , JP LOTZ bf, AM SCHOTT b,e , F GOLFIER ab
SC
a, Service de Gynécologie-Obstétrique et de Chirurgie Oncologique du Centre Hospitalier Lyon Sud, Hospices Civils de Lyon, Faculté de Médecine Lyon Sud, Université Lyon 1, France
M AN U
b, Centre de Référence des Maladies Trophoblastiques de Lyon, Faculté de Médecine Lyon Sud, Université Lyon 1, France
c, Service d’Obstétrique, Centre Hospitalier Femme-Mère-Enfant, Hospices Civils de Lyon, Université Lyon 1, France
d, Service d’Oncologie Médicale, CITOHL, Service d’Hématologie, Institut de Cancérologie des Hospices Civils de Lyon (IC-HCL), Centre Hospitalier Lyon Sud, Hospices Civils de Lyon, Faculté
TE D
de Médecine Lyon Sud, Université Lyon 1, France
e, Université de Lyon, Hospices Civils de Lyon, HESPER, France f, Service d’Oncologie, Centre Hospitalier Tenon, Assistance Publique-Hôpitaux de Paris,
EP
Université Pierre et Marie Curie, France
Corresponding author:
AC C
François Golfier, Service de Gynécologie-Obstétrique et de Chirurgie Oncologique, du Centre Hospitalier Lyon Sud, Faculté de médecine Lyon Sud, Université Lyon 1, France 165 chemin du Grand Revoyet, 69495 Pierre Bénite, France TL: 04 78 86 41 79
Fax: 04 78 86 15 76
@: franç[email protected]
TABLE 3
ACCEPTED MANUSCRIPT Source of Funding: Institut National du Cancer (INCa), Ligue National Contre le Cancer Conflicts of interest: The authors report no conflict of interest.
Word count:
RI PT
Abstract: 409
AC C
EP
TE D
M AN U
SC
Main text: 2515
2
3
ACCEPTED MANUSCRIPT
ABSTRACT
Background Patients with 2000 FIGO low-risk gestational trophoblastic neoplasia are commonly treated with single-agent chemotherapy. Methotrexate is widely used in this indication
RI PT
in Europe. Analysis of relapse after treatment and identification of factors associated with relapse would help understand their potential impacts on 2000 FIGO score evolution and chemotherapy management of gestational trophoblastic neoplasia
M AN U
SC
patients.
Objective
This retrospective study analyzes the predictive factors of relapse in low-risk gestational
Study Design
TE D
trophoblastic neoplasia patients whose hCG normalized with methotrexate alone.
Between 1999 and 2014, 993 patients with gestational trophoblastic neoplasia were
EP
identified in the French Trophoblastic Disease Reference Center database, of which 465 were low-risk patients whose hCG normalized with methotrexate alone. Using uni- and
AC C
multi-variate analysis we identified significant predictive factors for relapse after methotrexate. The Kaplan-Meier method was used to plot the outcome of patients.
Results
The five-year recurrence rate of low-risk gestational trophoblastic neoplasia patients whose hCG normalized with methotrexate alone was 5.7% (IC [3.86; 8.46]). Univariate analysis identified an antecedent pregnancy resulting in a delivery (HR= 5.96; IC 95% [1.40; 25.4], p= 0.016), a number of methotrexate courses superior to 5 courses (5 to 8
courses versus 1 to 4: HR= ACCEPTED 6.19; IC 95% MANUSCRIPT [1.43; 26.8], p= 0.015; 9 courses and more
4
versus 1 to 4: HR= 6.80; IC 95% [1.32; 35.1], p= 0.022) and hCG normalization delay centred to the mean as predictive factors of recurrence (HR= 1.27; IC 95% [1.09; 1.49], p=0.003). Multivariate analysis confirmed the type of antecedent pregnancy and the number of methotrexate courses as independent predictive factors of recurrence. A low-
RI PT
risk gestational trophoblastic neoplasia arising after a normal delivery had a 8.66-times higher relapse risk than that of a post-mole gestational trophoblastic neoplasia (IC 95% [1.98; 37.9], p= 0.0042). A patient who received 5 to 8 courses had a 6.7-times higher
SC
relapse risk than a patient who received 1 to 4 courses of methotrexate (IC 95% [1.54;
M AN U
29.2], p= 0.011). A patient who received 9 courses or more had a 8.1-times higher relapse risk than a patient who received 1 to 4 courses of methotrexate (IC 95% [1.54; 42.6], p= 0.014).
TE D
Conclusion
Low-risk gestational trophoblastic neoplasia following a delivery and patients who need more than 4 courses of MTX to normalization are at a higher risk of relapse than other
EP
low-risk patients. Allotting a higher score to the “antecedent pregnancy” FIGO item should be considered for post-delivery gestational trophoblastic neoplasia. Further
AC C
analysis of the need for consolidation courses is warranted.
KEY WORDS
Low risk gestational trophoblastic neoplasia, Methotrexate, Relapse, Predictive factors
5
ACCEPTED MANUSCRIPT
INTRODUCTION:
Gestational trophoblastic neoplasia (GTN) includes invasive moles, which arises in any type of pregnancy and typically from a hydatidiform mole, includes the malignant forms
RI PT
of gestational trophoblastic diseases (GTD), namely malignant invasive mole, choriocarcinoma, and, rarely, placental site trophoblastic tumor (PSTT) or epithelioid
SC
trophoblastic tumor (ETT).
After a thorough assessment of the extent of disease, patients are scored/staged
M AN U
according to the FIGO 2000 classification system [1, 2]. Those with a score of 6 or less are classified as having a low risk of resistance to single-agent chemotherapy [2]. Different single-agent protocols are available, and methotrexate (MTX) or Dactinomycin
TE D
are used most often
A debate about changing the FIGO score calculation to consider the heterogeneous risk
EP
of single-agent failure in low-risk GTN patients is currently underway. We previously published predictive factors for resistance to MTX in 154 low-risk GTN patients and
AC C
identified choriocarcinoma histology and hormone chorionic gonadotropin (hCG) clearance as the only two independent predictive factors for resistance while existing studies had also identified post-delivery GTN and FIGO score level as factors associated with resistance [3, 4, 5].Beyond analysis of resistance to single-agent chemotherapy, analysis of relapse after treatment and identification of factors associated with relapse would help to improve our understanding of the disease and bring useful data to the current debate [6,7,8]. Here, we present a cohort of 465 low-risk GTN patients whose hCG normalized with MTX alone in order to define independent predictive factors for
6
ACCEPTED MANUSCRIPT relapse and their potential impact on FIGO 2000 score evolution and chemotherapy
AC C
EP
TE D
M AN U
SC
RI PT
management.
7
MATERIAL AND METHODS ACCEPTED MANUSCRIPT
We identified 465 low-risk GTN patients whose hCG normalized with MTX alone, registered to the French Trophoblastic Disease Reference Center (FTDRC) between November 1999 and May 2014. Among 992 total GTN cases, high-risk GTN, placental
RI PT
site trophoblastic tumor (PSTT), and epithelioid trophoblastic tumor (ETT) patients were excluded (n = 184). Low-risk patients who received treatments other than MTX were excluded, namely those with hysterectomy, MTX-resistant disease requiring
SC
second-line chemotherapy, and spontaneous normalization of hCG levels (n = 322).
M AN U
Twenty-one low-risk patients still under MTX treatment at the time of data extraction were not included. One patient had two GTN events and we decided to keep the first event in the analysis (Figure 1).
TE D
The FTDRC’s functioning has already been described elsewhere [9]. Patients are reported to the FTDRC as molar pregnancy or GTN. They receive the treatment by their physician in their own region according to national guidelines. The FTDRC currently
EP
uses MTX in an 8-day MTX regimen [10] with 1mg/kg intramuscular MTX on days 1, 3, 5, and 7 with 10mg of folinic acid on days 2, 4, 6, and 8 [9, 11, 12]. Day 1 of each course
AC C
repeated every 14 days until hCG levels normalize; two additional consolidation courses of chemotherapy are given after the first hCG normalization. As patients receive the treatment in different centers, hCG concentrations are determined by different laboratories using various immunoassay kits that could impact interpretation of data. GTN diagnosis and FIGO score/stage were established according to 2000 FIGO oncology committee recommendations [2]. Twenty-one patients (4.5%) did not complete MTX treatment either due to poor tolerance or non-observance of the treatment. They were maintained in the study population as hCG normalized without further treatment.
8
MANUSCRIPT Disease relapse was definedACCEPTED as a rise in hCG levels following the completion of MTX treatment in the absence of a new pregnancy.
The following potential predictive factors were investigated using survival log-rank tests: parameters included in the FIGO score except for location of metastasis item
RI PT
because no metastasis other than pulmonary were present; previous chemotherapy item because all patients were chemonaive; age; and other potential predictive covariates except for tumor pathology (which was not available for the vast majority of
SC
post-molar GTN patients), including FIGO score (0 – 4 compared to 5 – 6), premature
M AN U
interruption of MTX defined as patients who stopped the treatment before normalization or consolidation courses because of poor tolerance, delay in hCG normalization centered to the mean 1.9, number of whole MTX courses, and number of
TE D
MTX consolidation courses.
To determine independent predictive factors for relapse, variables found to be statistically significant in univariate analysis were included in multivariate analysis
EP
using the Cox proportional hazards model with backward elimination. A p-value of 0.05 was considered statistically significant. Median follow-up was calculated using a reverse
AC C
Kaplan–Meier estimate. The Kaplan-Meier method estimated a five-year survival rate without relapse. The analyses were carried out by SAS V9.3 software (SAS software, SAS Institute Campus Drive, Cary, NC, USA).
9
ACCEPTED MANUSCRIPT
RESULTS
Four hundred and sixty-five patients with low-risk gestational trophoblastic neoplasia normalized their hCG with MTX alone. The mean follow-up period was 56.2 months
RI PT
(range: 1.7 to 171 months, SD: 40.0). The antecedent pregnancy was a mole in 440 patients (94.6%) and a delivery in 8 patients (1.7%). The mean number of MTX courses was 5.8 (range 1 to 18): 166 patients (35.8%) receive 1 to 4 MTX courses, 236 patients
SC
(50.7%) receive 5 to 8 courses and 63 patients (13.5%) receive more than 9 courses. No
M AN U
patient died from the disease. The characteristics of the patients are summarized in Table 1.
Twenty-four patients have relapsed at a mean interval of 10.9 months (range: 1.7 to 37.7 months). The five-year Kaplan-Meier recurrence rate was 5.7% (IC [3.86; 8.46]).
TE D
Among the 8 patients with low-risk GTN after delivery, 25% have relapsed. Figure 2A displays the Kaplan-Meier curve for relapse-free survival according to antecedent pregnancy type. Relapse rate was significantly higher for low-risk GTN occurring after a
EP
delivery than after any other type of antecedent pregnancy (p=0.0095)
AC C
Relapse rate was also significantly higher for patients who had 5 to 8 courses or more than 9 courses of MTX: 1.2% for patients with 1 to 4 courses, 7.2 % and 7.9 % for 5-8 courses and >= 9 courses respectively (p=0.016). Figure 2B displays the Kaplan-Meir curve for relapse-free survival according to number of MTX courses. (p= 0.0158)
Univariate analysis did not reveal premature interruption of MTX or number of consolidation courses to be risk factors for relapse (Table 2). Further, FIGO score, time interval between last pregnancy and beginning of GTN treatment, hCG level before treatment, presence of metastasis were not found to be risk factors of relapse for low-
10
ACCEPTED MANUSCRIPT risk GTN patients treated with MTX only (Table 2). Three factors were identified as
predictive factors of relapse: Low-risk GTN after delivery p =0.016, the overall number of MTX courses (p= 0.015 for 5-8 courses vs 1-4; p= 0.022 for 9 courses or more vs 1-4) and the hCG normalization delay centred to the mean (1.9 month) (p= 0.003)
RI PT
Multivariate analysis was performed after backward elimination of non-significant predictive factors from univariate analysis (Table 3). Due to a positive correlation and collinearity between number of MTX courses and hCG normalization delay (Pearson’rho
SC
= 0.78; P<0.0001), we also eliminated time to hCG normalization. We identified two
M AN U
independent predictive factors of relapse in low-risk GTN patients whose hCG normalized with MTX alone. (i). A post-delivery low-risk GTN had a 8.66-times higher relapse risk than a post-mole low-risk GTN (IC 95% [1.98; 37.9], p= 0.0042) (ii) A patient who received 5 to 8 courses of MTX had a 6.7-times higher relapse risk
TE D
than a patient who received 1 to 4 courses (IC 95% [1.54; 29.2], p= 0.011). A patient who received 9 courses or more of MTX had a 8.1 times higher relapse risk than a
AC C
.
EP
patient who received 1 to 4 courses (IC 95% [1.54; 42.6], p= 0.014)
11
ACCEPTED MANUSCRIPT
COMMENT
Our results demonstrate that in a very homogeneous population of 2000 FIGO-defined low-risk GTN patients whose hCG normalized with MTX alone, the relapse rate is very low. The five-year Kaplan-Meier relapse rate of 5.7% is in the range of relapse rate (2 to
RI PT
8.1%) from previously published series [6, 9, 13, 14, 15] where the FIGO scoring system was not uniformly used. Although the relevance of comparison with these studies is poor, such reported low rates of relapse associated with the quasi-absence of fatal cases
M AN U
SC
confirm the overall excellent prognosis of MTX-treated low-risk GTN patients.
Despite such low relapse rates, demonstration in our study of antecedent pregnancy type as an independent predictive factor for relapse after MTX alone reveals heterogeneity within the low-risk GTN group. The 8.66-times higher relapse risk associated with an antecedent delivery should question the relevance of treating all low-
TE D
risk GTN with MTX and considering them as real low-risk patients [3]. Indeed, beside this higher risk of relapse after treatment, the risk of MTX resistance while under
EP
treatment has also been shown to differ in low-risk patients with antecedent delivery. Post-delivery choriocarcinoma, almost the exclusive histological form of GTN after such
AC C
a pregnancy (when excluding the rare PSTT and ETT), where most fatal cases occur [16], has also been identified as an independent predictive factor for resistance to MTX in low-risk GTN patients [4, 14]. This raises the question again of considering them as lowrisk. The existing and likely poorly reported tendency to shift towards treatment of postdelivery low-risk GTN as high-risk [3] could lead to an overestimation of cure rate of low-risk GTN patients [13, 14, 17, 18]. Consequently, considering revision of the 2000 FIGO scoring system by increasing the weight allocated to antecedent, so-called “term”
ACCEPTED MANUSCRIPT pregnancy would allow continuing to promote its use by all physicians treating GTN
12
patients [17].
Our results also show that the overall number of MTX courses is an independent
RI PT
predictive factor for relapse after MTX alone. The significantly higher relapse risk associated with patients receiving 5 to 8 courses or more than 9 courses compared to patients receiving 1 to 4 courses of MTX reveals additional heterogeneity among lowrisk GTN patients. A similar limit of 8 courses of chemotherapy has been reported as a
SC
predictive factor for relapse in a more heterogeneous population of low- and high-risk
M AN U
GTN patients [7]. In our study, this predictive value was independent of antecedent pregnancy type and other expected prognostic factors, such as FIGO score and other FIGO score components as asserted by the Cox model, suggesting different biological behaviors among low-risk patients whose hCG normalized with MTX. This supports
TE D
published findings of a proportion of low-risk GTN patients whose hCG returned to normal in a median 5 weeks, with one single course of MTX without any consolidation course together with a crude relapse rate as low as 2% [19]. The proportion of patients
EP
treated with one single course of MTX ranged from 44.8% to 81.5% [19, 20]. This is also consistent with the outcomes of 2 previous studies, in which the strong predictive value
AC C
of early hCG modeled residual production (hCGres) regarding the risk of MTX failure was reported [4, 21]. Indeed modeled hCGres is estimated on the basis of hCG values measured during the first 50 days after the start of treatment, thereby suggesting the strong prognostic value of early hCG decline during the first chemotherapy cycles. Together with our inability to identify the number of consolidation courses as an independent predictive factor for relapse, these results could mean that consolidation courses could be critical only for low-risk GTN patients needing the highest number of chemotherapy courses. Of course, this assumption was not directly raised from the
ACCEPTED present study and must be further assessedMANUSCRIPT on the basis of previously published data
13
[6, 7].
The results reported herein should be interpreted with caution given the small number
RI PT
of patients with relapse or those with low-risk GTN after delivery, although it is the largest study with multivariate analysis of predictive factors for relapse to date. The relatively high number of patients contributed to the good statistical power of analysis. The results we present relate to our cohort but may not be applicable to another group
SC
of patients. The data used in the analysis conformed to “real-life” patient data: treatment
M AN U
in various French hospitals and hCG concentrations determined in different laboratories using various immunoassay kits that may give different results from one patient to another. Despite these limitations, our results are highly statistically significant.
TE D
In conclusion, the relapse risk of FIGO defined low-risk GTN patients was significantly higher when occurring after a delivery than after any type of other pregnancy. While promotion of the FIGO scoring/staging system is essential for international comparison,
EP
allotting a higher score to the “antecedent pregnancy” FIGO item when occurring after a delivery should be considered. The independent association of relapse risk with the
AC C
number of whole MTX courses suggests different biological behaviors among low-risk GTN patients. The number of consolidation courses should be further analyzed in the light of these new data. Patients needing more courses of MTX should be warned that they are at higher risk of relapse.
14
ACCEPTED MANUSCRIPT
Acknowledgement:
Source of Funding: Institut National du Cancer (INCa), Ligue National Contre le Cancer Conflicts of interest: The authors report no conflict of interest.
Author contribution:
RI PT
Conception and design: Florence Couder, Jérôme Massardier, Fatima Abbas, Touria Hajri, François Golfier
Collection and assembly of data: Touria Hajri, Florence Couder, Fatima Abbas, Jérôme
SC
Massardier, François Golfier
M AN U
Data analysis and interpretation: Florence Couder, Jérôme Massardier, Benoit You, Touria Hajri, Fatima Abbas, Anne-Marie Schott, Jean-Pierre Lotz, François Golfier Manuscript writing: All authors
AC C
EP
TE D
Final approval of manuscript: All authors
15
ACCEPTED MANUSCRIPT
REFERENCES
[1] Golfier F, Massardier J., Gustalla J-P. et al. Prise en charge des maladies trophoblastiques gestationnelles. Journal de Gynécologie Obstétrique et Biologie de la Reproduction. 2010: 39(3): 25-32. [2] Ngan HY, Bender H, Benedet JL, Jones H, Montruccoli GC, Pecorelli S. FIGO Committee on Gynecologic Oncology. Gestational trophoblastic neoplasia, FIGO 2000 staging and classification. Int J Gynaecol Obstet. 2003;83 Suppl 1:175-7.
RI PT
[3] Lybol C, Centen DW, Thomas CM et al. Fatal cases of gestational trophoblastic neoplasia over four decades in the Netherlands: a retrospective cohort study. BJOG. 2012 ;119(12):1465-72.
SC
[4] You B, Pollet-Villard M, Fronton L et al. Predictive values of hCG clearance for risk of methotrexate resistance in low-risk gestational trophoblastic neoplasias. Ann Oncol, 2010. 21: 1643-50.
M AN U
[5] Chapman-Davis E, Hoekstra AV, Rademaker AW, Schink JC, Lurain JR. Treatment of nonmetastatic and metastatic low-risk gestational trophoblastic neoplasia: factors associated with resistance to single-agent methotrexate chemotherapy. Gynecol Oncol. 2012 Jun;125(3):572-5.
TE D
[6] C.Lybol, F.C.G.J Sweep, R.Harvey et al. Relapse rates after two versus three consolidation courses of methotrexate in the treatment of low-risk gestational trophoblastic neoplasia. Gynecologic Oncology 2012; 125: 576–9.
EP
[7] Yang J, Xiang Y, Wan X, Yang X. Recurrent gestational trophoblastic tumor: management and risk factors for recurrence. Gynecol Oncol. 2006; 103: 587-90.
AC C
[8] Matsui H, Suzuka K, Yamazawa K et al. Relapse rate of patients with low-risk gestational trophoblastic tumor initially treated with single-agent chemotherapy. Gynecol Oncol. 2005; 96: 616-20. [9] Chalouhi GE, Golfier F, Soignon P et al. Methotrexate for 2000 FIGO low-risk gestational trophoblastic neoplasia patients: efficacy and toxicity. Am J Obstet Gynecol. 2009; Jun; 200: 643-6. [10] Bagshawe KD, Dent J, Newlands ES, et al. The role of low-dose methotrexate and folinic acid in gestational trophoblastic tumors (GTT). Br J Obstet Gynaecol 1989; 96: 795-802. [11] Ngan S, Seckl MJ. Gestational trophoblastic neoplasia management: an update. Curr Opin Oncol 2007;19:486-91. [12] Osborne R, Gerulath A. What is the best regimen for low-risk gestational
ACCEPTED trophoblastic neoplasia? A review. J ReprodMANUSCRIPT Med 2004; 49:602-16.
16
[13] McNeish IA, Strickland S, Holden L et al. Low-risk persistent gestational trophoblastic disease: outcome after initial treatment with low-dose methotrexate and folinic acid from 1992 to 2000. J Clin Oncol. 2002; 20(7): 1838-44.
RI PT
[14] Powles T, Savage PM, Stebbing J et al.. A comparison of patients with relapsed and chemo-refractory gestational trophoblastic neoplasia. Br J Cancer. 2007 Mar 12; 96: 732-7. [15] Kerkmeijer LG, Wielsma S, Massuger LF et al. Recurrent gestational trophoblastic disease after hCG normalization following hydatidiform mole in The Netherlands. Gynecol Oncol. 2007; 106: 142-6.
M AN U
SC
[16] Neubauer NL, Strohl AE, Schink JC, Lurain JR. Fatal gestational trophoblastic neoplasia: An analysis of treatment failures at the Brewer Trophoblastic Disease Center from 1979-2012 compared to 1962-1978. Gynecol Oncol. 2015;138(2):339-42. [17] Seckl M.J., Sebire, N.J and Berkowitz RS. Gestational trophoblastic disease. Lancet, 2010. 376: 717-29.
TE D
[18] Chapman-Davis E, Hoekstra AV, Rademaker AW, Schink JC, Lurain JR. Treatment of nonmetastatic and metastatic low-risk gestational trophoblastic neoplasia: factors associated with resistance to single-agent methotrexate chemotherapy. Gynecol Oncol. 2012;125(3):572-5. [19] Chan KK, Huang Y, Tam KF, Tse KY, Ngan HY. Single-dose methotrexate regimen in the treatment of low-risk gestational trophoblastic neoplasia. Am J Obstet Gynecol. 2006;195(5):1282-6.
EP
[20] Growdon WB, Wolfberg AJ, Goldstein DP et al. Evaluating methotrexate treatment in patients with low-risk post-molar gestational trophoblastic neoplasia. Gynecol Oncol. 2009 ;112(2):353-7.
AC C
[21] You B, Harvey R, Henin E et al. Early prediction of treatment resistance in low-risk gestational trophoblastic neoplasia using population kinetic modelling of hCG measurements. Br J Cancer. 2013 May 14;108(9):1810-6.
17
ACCEPTED MANUSCRIPT All patients N=465
Patient characteristics
N (%)*
384 (82.6) 81 (17.4)
Age mean (std)
32.3 (8.1)
Interval between end of previous pregnancy and start of chemotherapy Missing data < 4 months 4 à 6 months (<7) 7 à 12 months (<13) >=13 months
3 (0.7) 392 (84.3) 63 (13.5) 5 (1.1) 2 (0.4)
Pregnancy history Post Mole Post Term Post Ectopic or termination
440 (94.6) 8 (1.7) 17 (3.7)
FIGO score [0 - 4] [5 - 6]
SC 431 (92.7) 34 (7.3)
TE D
FIGO stage Stade I Stade II Stade III Stade IV
155 (33.3) 150 (32.3) 145 (31.2) 15 (3.2) 0 (0)
M AN U
Serum Hcg before start MTX (IU/L) <= 1000 1000 à 10000 10000 à 100000 100000 à 1000000 > 1000000
RI PT
Age < 40 years >= 40 years
387 (83.2) 17 (3.7) 60 (12.9) 1 (0.2)
464 (99.8) 1 (0.2) 0 (0) 0 (0)
Premature interruption of MTX No Yes
444 (95.5) 21 (4.52)
Number of whole courses of MTX 1 to 4 courses 5 to 8 courses ≥ 9 courses
166 (35.8) 236 (50.7) 63 (13.5)
AC C
EP
Site of metastases None or lung Spleen or kidney Alimentary canal Brain or liver
Number of whole courses of MTX mean (std) Number of consolidation courses 0 courses 1 to 1.5 courses ** 2 courses ≥ 3 courses
5.8 (2.6)
37 (8) 45 (9.7) 343 (73.8) 40 (8.6)
Number of consolidation courses of MTX mean (std)
1.9 (0.7)
HCG normalization delay (month) mean (std)
1.9 (1.4)
Relapse No Yes
441 (94.8) 24 (5.2)
* Number and percent of patients unless stated otherwise. ** 1,5 courses: patients who did not complete the second course of MTX because of bad tolerance
Table 1 : Patient characteristics at baseline
18
ACCEPTED MANUSCRIPT
IC95% HR
P
Age (year) centered at 10
0.96
[0.57 ; 1.59]
0.863
Time interval from antecedent pregnancy 2 - 3 months ≥ 4 months Versus < 2 months
1.43 1.88
[0.54; 3.82] [0.57; 6.17]
0.473 0.296
Last pregnancy outcome Delivery Abortion or unknown Versus mole
5.96 1.18
[1.40; 25.4] [0.16; 8.78]
hCG level before treatment (IU/L) 1000 - 10000 10000 - 100000 100000 - 1000000 Versus < 1000
1.42 1.44 3.43
FIGO score [5 - 6] Versus [0 - 4]
SC
0.016 0.871
[0.49; 4.10] [0.50; 4.16] [0.69; 17.0]
0.513 0.496 0.131
M AN U
Risk Factors
RI PT
Crude Hazard Ratio
1.79
[0.53; 5.99]
0.348
3.87
[0.52; 28.7]
0.185
1.11
[0.15; 8.25]
0.916
1.27
[1.09; 1.49]
0.003
2.01 2.48 2.08
[0.18; 22.2] [0.26; 23.9] [0.28; 15.6]
0.569 0.431 0.476
Number of whole courses of MTX 5 to 8 courses
6.19
[1.43 ; 26.8]
0.015
≥ 9 courses
6.80
[1.32 ; 35.1]
0.022
Number of metastases 0 Versus ≥ 1 Versus complete regimen
TE D
Premature MTX interruption
HCG normalization delay (month) centered at mean (1.9 months)
AC C
EP
Number of consolidation courses 0 courses 1 to 1.5 courses * 2 courses Versus ≥ 3 courses
Versus 1 to 4 courses
Number of whole courses of MTX (continued 1.13 [1.00 ; 1.28] 0.055 value) * 1,5 courses: patients who did not complete the second course of MTX because of bad tolerance
Table 2: Univariate analyses of risk factors in relapsing patients
19
ACCEPTED MANUSCRIPT Adjusted Hazard Ratio
IC95% HR
P
Post term pregnancy
8.66
[1.98 ; 37.9]
0.0042
Post abortion or ectopic
1.07
[0.14 ; 7.98]
0.95
5 to 8 courses
6.7
[1.54 ; 29.2]
0.011
≥ 9 courses
8.1
[1.54 ; 42.6]
0.014
Parameter
Versus post mole Number of whole courses of MTX
SC
Versus 1 to 4 courses
RI PT
Last pregnancy outcome
AC C
EP
TE D
M AN U
Table 3: Multivariate analyses of risk factors in relapsing patients
AC C
EP
TE D
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT
AC C
EP
TE D
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT
AC C
EP
TE D
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT
S0002-9378(16)00233-7
DOI:
10.1016/j.ajog.2016.01.183
Reference:
YMOB 10913
To appear in:
American Journal of Obstetrics and Gynecology
Received Date: 14 September 2015 Revised Date:
13 January 2016
Accepted Date: 22 January 2016
Please cite this article as: Couder F, Massardier J, You B, Abbas F, Hajri T, Lotz J, Schott A, Golfier F, Predictive factors of relapse in low-risk gestational trophoblastic neoplasia patients successfully treated with methotrexate alone, American Journal of Obstetrics and Gynecology (2016), doi: 10.1016/ j.ajog.2016.01.183. 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.
1
ACCEPTED MANUSCRIPT
TITLE
Predictive factors of relapse in low-risk gestational trophoblastic neoplasia patients successfully treated with methotrexate alone
AUTHORS
RI PT
F COUDER a, J MASSARDIER b,c , B YOU b,d , F ABBAS e, T HAJRI b , JP LOTZ bf, AM SCHOTT b,e , F GOLFIER ab
SC
a, Service de Gynécologie-Obstétrique et de Chirurgie Oncologique du Centre Hospitalier Lyon Sud, Hospices Civils de Lyon, Faculté de Médecine Lyon Sud, Université Lyon 1, France
M AN U
b, Centre de Référence des Maladies Trophoblastiques de Lyon, Faculté de Médecine Lyon Sud, Université Lyon 1, France
c, Service d’Obstétrique, Centre Hospitalier Femme-Mère-Enfant, Hospices Civils de Lyon, Université Lyon 1, France
d, Service d’Oncologie Médicale, CITOHL, Service d’Hématologie, Institut de Cancérologie des Hospices Civils de Lyon (IC-HCL), Centre Hospitalier Lyon Sud, Hospices Civils de Lyon, Faculté
TE D
de Médecine Lyon Sud, Université Lyon 1, France
e, Université de Lyon, Hospices Civils de Lyon, HESPER, France f, Service d’Oncologie, Centre Hospitalier Tenon, Assistance Publique-Hôpitaux de Paris,
EP
Université Pierre et Marie Curie, France
Corresponding author:
AC C
François Golfier, Service de Gynécologie-Obstétrique et de Chirurgie Oncologique, du Centre Hospitalier Lyon Sud, Faculté de médecine Lyon Sud, Université Lyon 1, France 165 chemin du Grand Revoyet, 69495 Pierre Bénite, France TL: 04 78 86 41 79
Fax: 04 78 86 15 76
@: franç[email protected]
TABLE 3
ACCEPTED MANUSCRIPT Source of Funding: Institut National du Cancer (INCa), Ligue National Contre le Cancer Conflicts of interest: The authors report no conflict of interest.
Word count:
RI PT
Abstract: 409
AC C
EP
TE D
M AN U
SC
Main text: 2515
2
3
ACCEPTED MANUSCRIPT
ABSTRACT
Background Patients with 2000 FIGO low-risk gestational trophoblastic neoplasia are commonly treated with single-agent chemotherapy. Methotrexate is widely used in this indication
RI PT
in Europe. Analysis of relapse after treatment and identification of factors associated with relapse would help understand their potential impacts on 2000 FIGO score evolution and chemotherapy management of gestational trophoblastic neoplasia
M AN U
SC
patients.
Objective
This retrospective study analyzes the predictive factors of relapse in low-risk gestational
Study Design
TE D
trophoblastic neoplasia patients whose hCG normalized with methotrexate alone.
Between 1999 and 2014, 993 patients with gestational trophoblastic neoplasia were
EP
identified in the French Trophoblastic Disease Reference Center database, of which 465 were low-risk patients whose hCG normalized with methotrexate alone. Using uni- and
AC C
multi-variate analysis we identified significant predictive factors for relapse after methotrexate. The Kaplan-Meier method was used to plot the outcome of patients.
Results
The five-year recurrence rate of low-risk gestational trophoblastic neoplasia patients whose hCG normalized with methotrexate alone was 5.7% (IC [3.86; 8.46]). Univariate analysis identified an antecedent pregnancy resulting in a delivery (HR= 5.96; IC 95% [1.40; 25.4], p= 0.016), a number of methotrexate courses superior to 5 courses (5 to 8
courses versus 1 to 4: HR= ACCEPTED 6.19; IC 95% MANUSCRIPT [1.43; 26.8], p= 0.015; 9 courses and more
4
versus 1 to 4: HR= 6.80; IC 95% [1.32; 35.1], p= 0.022) and hCG normalization delay centred to the mean as predictive factors of recurrence (HR= 1.27; IC 95% [1.09; 1.49], p=0.003). Multivariate analysis confirmed the type of antecedent pregnancy and the number of methotrexate courses as independent predictive factors of recurrence. A low-
RI PT
risk gestational trophoblastic neoplasia arising after a normal delivery had a 8.66-times higher relapse risk than that of a post-mole gestational trophoblastic neoplasia (IC 95% [1.98; 37.9], p= 0.0042). A patient who received 5 to 8 courses had a 6.7-times higher
SC
relapse risk than a patient who received 1 to 4 courses of methotrexate (IC 95% [1.54;
M AN U
29.2], p= 0.011). A patient who received 9 courses or more had a 8.1-times higher relapse risk than a patient who received 1 to 4 courses of methotrexate (IC 95% [1.54; 42.6], p= 0.014).
TE D
Conclusion
Low-risk gestational trophoblastic neoplasia following a delivery and patients who need more than 4 courses of MTX to normalization are at a higher risk of relapse than other
EP
low-risk patients. Allotting a higher score to the “antecedent pregnancy” FIGO item should be considered for post-delivery gestational trophoblastic neoplasia. Further
AC C
analysis of the need for consolidation courses is warranted.
KEY WORDS
Low risk gestational trophoblastic neoplasia, Methotrexate, Relapse, Predictive factors
5
ACCEPTED MANUSCRIPT
INTRODUCTION:
Gestational trophoblastic neoplasia (GTN) includes invasive moles, which arises in any type of pregnancy and typically from a hydatidiform mole, includes the malignant forms
RI PT
of gestational trophoblastic diseases (GTD), namely malignant invasive mole, choriocarcinoma, and, rarely, placental site trophoblastic tumor (PSTT) or epithelioid
SC
trophoblastic tumor (ETT).
After a thorough assessment of the extent of disease, patients are scored/staged
M AN U
according to the FIGO 2000 classification system [1, 2]. Those with a score of 6 or less are classified as having a low risk of resistance to single-agent chemotherapy [2]. Different single-agent protocols are available, and methotrexate (MTX) or Dactinomycin
TE D
are used most often
A debate about changing the FIGO score calculation to consider the heterogeneous risk
EP
of single-agent failure in low-risk GTN patients is currently underway. We previously published predictive factors for resistance to MTX in 154 low-risk GTN patients and
AC C
identified choriocarcinoma histology and hormone chorionic gonadotropin (hCG) clearance as the only two independent predictive factors for resistance while existing studies had also identified post-delivery GTN and FIGO score level as factors associated with resistance [3, 4, 5].Beyond analysis of resistance to single-agent chemotherapy, analysis of relapse after treatment and identification of factors associated with relapse would help to improve our understanding of the disease and bring useful data to the current debate [6,7,8]. Here, we present a cohort of 465 low-risk GTN patients whose hCG normalized with MTX alone in order to define independent predictive factors for
6
ACCEPTED MANUSCRIPT relapse and their potential impact on FIGO 2000 score evolution and chemotherapy
AC C
EP
TE D
M AN U
SC
RI PT
management.
7
MATERIAL AND METHODS ACCEPTED MANUSCRIPT
We identified 465 low-risk GTN patients whose hCG normalized with MTX alone, registered to the French Trophoblastic Disease Reference Center (FTDRC) between November 1999 and May 2014. Among 992 total GTN cases, high-risk GTN, placental
RI PT
site trophoblastic tumor (PSTT), and epithelioid trophoblastic tumor (ETT) patients were excluded (n = 184). Low-risk patients who received treatments other than MTX were excluded, namely those with hysterectomy, MTX-resistant disease requiring
SC
second-line chemotherapy, and spontaneous normalization of hCG levels (n = 322).
M AN U
Twenty-one low-risk patients still under MTX treatment at the time of data extraction were not included. One patient had two GTN events and we decided to keep the first event in the analysis (Figure 1).
TE D
The FTDRC’s functioning has already been described elsewhere [9]. Patients are reported to the FTDRC as molar pregnancy or GTN. They receive the treatment by their physician in their own region according to national guidelines. The FTDRC currently
EP
uses MTX in an 8-day MTX regimen [10] with 1mg/kg intramuscular MTX on days 1, 3, 5, and 7 with 10mg of folinic acid on days 2, 4, 6, and 8 [9, 11, 12]. Day 1 of each course
AC C
repeated every 14 days until hCG levels normalize; two additional consolidation courses of chemotherapy are given after the first hCG normalization. As patients receive the treatment in different centers, hCG concentrations are determined by different laboratories using various immunoassay kits that could impact interpretation of data. GTN diagnosis and FIGO score/stage were established according to 2000 FIGO oncology committee recommendations [2]. Twenty-one patients (4.5%) did not complete MTX treatment either due to poor tolerance or non-observance of the treatment. They were maintained in the study population as hCG normalized without further treatment.
8
MANUSCRIPT Disease relapse was definedACCEPTED as a rise in hCG levels following the completion of MTX treatment in the absence of a new pregnancy.
The following potential predictive factors were investigated using survival log-rank tests: parameters included in the FIGO score except for location of metastasis item
RI PT
because no metastasis other than pulmonary were present; previous chemotherapy item because all patients were chemonaive; age; and other potential predictive covariates except for tumor pathology (which was not available for the vast majority of
SC
post-molar GTN patients), including FIGO score (0 – 4 compared to 5 – 6), premature
M AN U
interruption of MTX defined as patients who stopped the treatment before normalization or consolidation courses because of poor tolerance, delay in hCG normalization centered to the mean 1.9, number of whole MTX courses, and number of
TE D
MTX consolidation courses.
To determine independent predictive factors for relapse, variables found to be statistically significant in univariate analysis were included in multivariate analysis
EP
using the Cox proportional hazards model with backward elimination. A p-value of 0.05 was considered statistically significant. Median follow-up was calculated using a reverse
AC C
Kaplan–Meier estimate. The Kaplan-Meier method estimated a five-year survival rate without relapse. The analyses were carried out by SAS V9.3 software (SAS software, SAS Institute Campus Drive, Cary, NC, USA).
9
ACCEPTED MANUSCRIPT
RESULTS
Four hundred and sixty-five patients with low-risk gestational trophoblastic neoplasia normalized their hCG with MTX alone. The mean follow-up period was 56.2 months
RI PT
(range: 1.7 to 171 months, SD: 40.0). The antecedent pregnancy was a mole in 440 patients (94.6%) and a delivery in 8 patients (1.7%). The mean number of MTX courses was 5.8 (range 1 to 18): 166 patients (35.8%) receive 1 to 4 MTX courses, 236 patients
SC
(50.7%) receive 5 to 8 courses and 63 patients (13.5%) receive more than 9 courses. No
M AN U
patient died from the disease. The characteristics of the patients are summarized in Table 1.
Twenty-four patients have relapsed at a mean interval of 10.9 months (range: 1.7 to 37.7 months). The five-year Kaplan-Meier recurrence rate was 5.7% (IC [3.86; 8.46]).
TE D
Among the 8 patients with low-risk GTN after delivery, 25% have relapsed. Figure 2A displays the Kaplan-Meier curve for relapse-free survival according to antecedent pregnancy type. Relapse rate was significantly higher for low-risk GTN occurring after a
EP
delivery than after any other type of antecedent pregnancy (p=0.0095)
AC C
Relapse rate was also significantly higher for patients who had 5 to 8 courses or more than 9 courses of MTX: 1.2% for patients with 1 to 4 courses, 7.2 % and 7.9 % for 5-8 courses and >= 9 courses respectively (p=0.016). Figure 2B displays the Kaplan-Meir curve for relapse-free survival according to number of MTX courses. (p= 0.0158)
Univariate analysis did not reveal premature interruption of MTX or number of consolidation courses to be risk factors for relapse (Table 2). Further, FIGO score, time interval between last pregnancy and beginning of GTN treatment, hCG level before treatment, presence of metastasis were not found to be risk factors of relapse for low-
10
ACCEPTED MANUSCRIPT risk GTN patients treated with MTX only (Table 2). Three factors were identified as
predictive factors of relapse: Low-risk GTN after delivery p =0.016, the overall number of MTX courses (p= 0.015 for 5-8 courses vs 1-4; p= 0.022 for 9 courses or more vs 1-4) and the hCG normalization delay centred to the mean (1.9 month) (p= 0.003)
RI PT
Multivariate analysis was performed after backward elimination of non-significant predictive factors from univariate analysis (Table 3). Due to a positive correlation and collinearity between number of MTX courses and hCG normalization delay (Pearson’rho
SC
= 0.78; P<0.0001), we also eliminated time to hCG normalization. We identified two
M AN U
independent predictive factors of relapse in low-risk GTN patients whose hCG normalized with MTX alone. (i). A post-delivery low-risk GTN had a 8.66-times higher relapse risk than a post-mole low-risk GTN (IC 95% [1.98; 37.9], p= 0.0042) (ii) A patient who received 5 to 8 courses of MTX had a 6.7-times higher relapse risk
TE D
than a patient who received 1 to 4 courses (IC 95% [1.54; 29.2], p= 0.011). A patient who received 9 courses or more of MTX had a 8.1 times higher relapse risk than a
AC C
.
EP
patient who received 1 to 4 courses (IC 95% [1.54; 42.6], p= 0.014)
11
ACCEPTED MANUSCRIPT
COMMENT
Our results demonstrate that in a very homogeneous population of 2000 FIGO-defined low-risk GTN patients whose hCG normalized with MTX alone, the relapse rate is very low. The five-year Kaplan-Meier relapse rate of 5.7% is in the range of relapse rate (2 to
RI PT
8.1%) from previously published series [6, 9, 13, 14, 15] where the FIGO scoring system was not uniformly used. Although the relevance of comparison with these studies is poor, such reported low rates of relapse associated with the quasi-absence of fatal cases
M AN U
SC
confirm the overall excellent prognosis of MTX-treated low-risk GTN patients.
Despite such low relapse rates, demonstration in our study of antecedent pregnancy type as an independent predictive factor for relapse after MTX alone reveals heterogeneity within the low-risk GTN group. The 8.66-times higher relapse risk associated with an antecedent delivery should question the relevance of treating all low-
TE D
risk GTN with MTX and considering them as real low-risk patients [3]. Indeed, beside this higher risk of relapse after treatment, the risk of MTX resistance while under
EP
treatment has also been shown to differ in low-risk patients with antecedent delivery. Post-delivery choriocarcinoma, almost the exclusive histological form of GTN after such
AC C
a pregnancy (when excluding the rare PSTT and ETT), where most fatal cases occur [16], has also been identified as an independent predictive factor for resistance to MTX in low-risk GTN patients [4, 14]. This raises the question again of considering them as lowrisk. The existing and likely poorly reported tendency to shift towards treatment of postdelivery low-risk GTN as high-risk [3] could lead to an overestimation of cure rate of low-risk GTN patients [13, 14, 17, 18]. Consequently, considering revision of the 2000 FIGO scoring system by increasing the weight allocated to antecedent, so-called “term”
ACCEPTED MANUSCRIPT pregnancy would allow continuing to promote its use by all physicians treating GTN
12
patients [17].
Our results also show that the overall number of MTX courses is an independent
RI PT
predictive factor for relapse after MTX alone. The significantly higher relapse risk associated with patients receiving 5 to 8 courses or more than 9 courses compared to patients receiving 1 to 4 courses of MTX reveals additional heterogeneity among lowrisk GTN patients. A similar limit of 8 courses of chemotherapy has been reported as a
SC
predictive factor for relapse in a more heterogeneous population of low- and high-risk
M AN U
GTN patients [7]. In our study, this predictive value was independent of antecedent pregnancy type and other expected prognostic factors, such as FIGO score and other FIGO score components as asserted by the Cox model, suggesting different biological behaviors among low-risk patients whose hCG normalized with MTX. This supports
TE D
published findings of a proportion of low-risk GTN patients whose hCG returned to normal in a median 5 weeks, with one single course of MTX without any consolidation course together with a crude relapse rate as low as 2% [19]. The proportion of patients
EP
treated with one single course of MTX ranged from 44.8% to 81.5% [19, 20]. This is also consistent with the outcomes of 2 previous studies, in which the strong predictive value
AC C
of early hCG modeled residual production (hCGres) regarding the risk of MTX failure was reported [4, 21]. Indeed modeled hCGres is estimated on the basis of hCG values measured during the first 50 days after the start of treatment, thereby suggesting the strong prognostic value of early hCG decline during the first chemotherapy cycles. Together with our inability to identify the number of consolidation courses as an independent predictive factor for relapse, these results could mean that consolidation courses could be critical only for low-risk GTN patients needing the highest number of chemotherapy courses. Of course, this assumption was not directly raised from the
ACCEPTED present study and must be further assessedMANUSCRIPT on the basis of previously published data
13
[6, 7].
The results reported herein should be interpreted with caution given the small number
RI PT
of patients with relapse or those with low-risk GTN after delivery, although it is the largest study with multivariate analysis of predictive factors for relapse to date. The relatively high number of patients contributed to the good statistical power of analysis. The results we present relate to our cohort but may not be applicable to another group
SC
of patients. The data used in the analysis conformed to “real-life” patient data: treatment
M AN U
in various French hospitals and hCG concentrations determined in different laboratories using various immunoassay kits that may give different results from one patient to another. Despite these limitations, our results are highly statistically significant.
TE D
In conclusion, the relapse risk of FIGO defined low-risk GTN patients was significantly higher when occurring after a delivery than after any type of other pregnancy. While promotion of the FIGO scoring/staging system is essential for international comparison,
EP
allotting a higher score to the “antecedent pregnancy” FIGO item when occurring after a delivery should be considered. The independent association of relapse risk with the
AC C
number of whole MTX courses suggests different biological behaviors among low-risk GTN patients. The number of consolidation courses should be further analyzed in the light of these new data. Patients needing more courses of MTX should be warned that they are at higher risk of relapse.
14
ACCEPTED MANUSCRIPT
Acknowledgement:
Source of Funding: Institut National du Cancer (INCa), Ligue National Contre le Cancer Conflicts of interest: The authors report no conflict of interest.
Author contribution:
RI PT
Conception and design: Florence Couder, Jérôme Massardier, Fatima Abbas, Touria Hajri, François Golfier
Collection and assembly of data: Touria Hajri, Florence Couder, Fatima Abbas, Jérôme
SC
Massardier, François Golfier
M AN U
Data analysis and interpretation: Florence Couder, Jérôme Massardier, Benoit You, Touria Hajri, Fatima Abbas, Anne-Marie Schott, Jean-Pierre Lotz, François Golfier Manuscript writing: All authors
AC C
EP
TE D
Final approval of manuscript: All authors
15
ACCEPTED MANUSCRIPT
REFERENCES
[1] Golfier F, Massardier J., Gustalla J-P. et al. Prise en charge des maladies trophoblastiques gestationnelles. Journal de Gynécologie Obstétrique et Biologie de la Reproduction. 2010: 39(3): 25-32. [2] Ngan HY, Bender H, Benedet JL, Jones H, Montruccoli GC, Pecorelli S. FIGO Committee on Gynecologic Oncology. Gestational trophoblastic neoplasia, FIGO 2000 staging and classification. Int J Gynaecol Obstet. 2003;83 Suppl 1:175-7.
RI PT
[3] Lybol C, Centen DW, Thomas CM et al. Fatal cases of gestational trophoblastic neoplasia over four decades in the Netherlands: a retrospective cohort study. BJOG. 2012 ;119(12):1465-72.
SC
[4] You B, Pollet-Villard M, Fronton L et al. Predictive values of hCG clearance for risk of methotrexate resistance in low-risk gestational trophoblastic neoplasias. Ann Oncol, 2010. 21: 1643-50.
M AN U
[5] Chapman-Davis E, Hoekstra AV, Rademaker AW, Schink JC, Lurain JR. Treatment of nonmetastatic and metastatic low-risk gestational trophoblastic neoplasia: factors associated with resistance to single-agent methotrexate chemotherapy. Gynecol Oncol. 2012 Jun;125(3):572-5.
TE D
[6] C.Lybol, F.C.G.J Sweep, R.Harvey et al. Relapse rates after two versus three consolidation courses of methotrexate in the treatment of low-risk gestational trophoblastic neoplasia. Gynecologic Oncology 2012; 125: 576–9.
EP
[7] Yang J, Xiang Y, Wan X, Yang X. Recurrent gestational trophoblastic tumor: management and risk factors for recurrence. Gynecol Oncol. 2006; 103: 587-90.
AC C
[8] Matsui H, Suzuka K, Yamazawa K et al. Relapse rate of patients with low-risk gestational trophoblastic tumor initially treated with single-agent chemotherapy. Gynecol Oncol. 2005; 96: 616-20. [9] Chalouhi GE, Golfier F, Soignon P et al. Methotrexate for 2000 FIGO low-risk gestational trophoblastic neoplasia patients: efficacy and toxicity. Am J Obstet Gynecol. 2009; Jun; 200: 643-6. [10] Bagshawe KD, Dent J, Newlands ES, et al. The role of low-dose methotrexate and folinic acid in gestational trophoblastic tumors (GTT). Br J Obstet Gynaecol 1989; 96: 795-802. [11] Ngan S, Seckl MJ. Gestational trophoblastic neoplasia management: an update. Curr Opin Oncol 2007;19:486-91. [12] Osborne R, Gerulath A. What is the best regimen for low-risk gestational
ACCEPTED trophoblastic neoplasia? A review. J ReprodMANUSCRIPT Med 2004; 49:602-16.
16
[13] McNeish IA, Strickland S, Holden L et al. Low-risk persistent gestational trophoblastic disease: outcome after initial treatment with low-dose methotrexate and folinic acid from 1992 to 2000. J Clin Oncol. 2002; 20(7): 1838-44.
RI PT
[14] Powles T, Savage PM, Stebbing J et al.. A comparison of patients with relapsed and chemo-refractory gestational trophoblastic neoplasia. Br J Cancer. 2007 Mar 12; 96: 732-7. [15] Kerkmeijer LG, Wielsma S, Massuger LF et al. Recurrent gestational trophoblastic disease after hCG normalization following hydatidiform mole in The Netherlands. Gynecol Oncol. 2007; 106: 142-6.
M AN U
SC
[16] Neubauer NL, Strohl AE, Schink JC, Lurain JR. Fatal gestational trophoblastic neoplasia: An analysis of treatment failures at the Brewer Trophoblastic Disease Center from 1979-2012 compared to 1962-1978. Gynecol Oncol. 2015;138(2):339-42. [17] Seckl M.J., Sebire, N.J and Berkowitz RS. Gestational trophoblastic disease. Lancet, 2010. 376: 717-29.
TE D
[18] Chapman-Davis E, Hoekstra AV, Rademaker AW, Schink JC, Lurain JR. Treatment of nonmetastatic and metastatic low-risk gestational trophoblastic neoplasia: factors associated with resistance to single-agent methotrexate chemotherapy. Gynecol Oncol. 2012;125(3):572-5. [19] Chan KK, Huang Y, Tam KF, Tse KY, Ngan HY. Single-dose methotrexate regimen in the treatment of low-risk gestational trophoblastic neoplasia. Am J Obstet Gynecol. 2006;195(5):1282-6.
EP
[20] Growdon WB, Wolfberg AJ, Goldstein DP et al. Evaluating methotrexate treatment in patients with low-risk post-molar gestational trophoblastic neoplasia. Gynecol Oncol. 2009 ;112(2):353-7.
AC C
[21] You B, Harvey R, Henin E et al. Early prediction of treatment resistance in low-risk gestational trophoblastic neoplasia using population kinetic modelling of hCG measurements. Br J Cancer. 2013 May 14;108(9):1810-6.
17
ACCEPTED MANUSCRIPT All patients N=465
Patient characteristics
N (%)*
384 (82.6) 81 (17.4)
Age mean (std)
32.3 (8.1)
Interval between end of previous pregnancy and start of chemotherapy Missing data < 4 months 4 à 6 months (<7) 7 à 12 months (<13) >=13 months
3 (0.7) 392 (84.3) 63 (13.5) 5 (1.1) 2 (0.4)
Pregnancy history Post Mole Post Term Post Ectopic or termination
440 (94.6) 8 (1.7) 17 (3.7)
FIGO score [0 - 4] [5 - 6]
SC 431 (92.7) 34 (7.3)
TE D
FIGO stage Stade I Stade II Stade III Stade IV
155 (33.3) 150 (32.3) 145 (31.2) 15 (3.2) 0 (0)
M AN U
Serum Hcg before start MTX (IU/L) <= 1000 1000 à 10000 10000 à 100000 100000 à 1000000 > 1000000
RI PT
Age < 40 years >= 40 years
387 (83.2) 17 (3.7) 60 (12.9) 1 (0.2)
464 (99.8) 1 (0.2) 0 (0) 0 (0)
Premature interruption of MTX No Yes
444 (95.5) 21 (4.52)
Number of whole courses of MTX 1 to 4 courses 5 to 8 courses ≥ 9 courses
166 (35.8) 236 (50.7) 63 (13.5)
AC C
EP
Site of metastases None or lung Spleen or kidney Alimentary canal Brain or liver
Number of whole courses of MTX mean (std) Number of consolidation courses 0 courses 1 to 1.5 courses ** 2 courses ≥ 3 courses
5.8 (2.6)
37 (8) 45 (9.7) 343 (73.8) 40 (8.6)
Number of consolidation courses of MTX mean (std)
1.9 (0.7)
HCG normalization delay (month) mean (std)
1.9 (1.4)
Relapse No Yes
441 (94.8) 24 (5.2)
* Number and percent of patients unless stated otherwise. ** 1,5 courses: patients who did not complete the second course of MTX because of bad tolerance
Table 1 : Patient characteristics at baseline
18
ACCEPTED MANUSCRIPT
IC95% HR
P
Age (year) centered at 10
0.96
[0.57 ; 1.59]
0.863
Time interval from antecedent pregnancy 2 - 3 months ≥ 4 months Versus < 2 months
1.43 1.88
[0.54; 3.82] [0.57; 6.17]
0.473 0.296
Last pregnancy outcome Delivery Abortion or unknown Versus mole
5.96 1.18
[1.40; 25.4] [0.16; 8.78]
hCG level before treatment (IU/L) 1000 - 10000 10000 - 100000 100000 - 1000000 Versus < 1000
1.42 1.44 3.43
FIGO score [5 - 6] Versus [0 - 4]
SC
0.016 0.871
[0.49; 4.10] [0.50; 4.16] [0.69; 17.0]
0.513 0.496 0.131
M AN U
Risk Factors
RI PT
Crude Hazard Ratio
1.79
[0.53; 5.99]
0.348
3.87
[0.52; 28.7]
0.185
1.11
[0.15; 8.25]
0.916
1.27
[1.09; 1.49]
0.003
2.01 2.48 2.08
[0.18; 22.2] [0.26; 23.9] [0.28; 15.6]
0.569 0.431 0.476
Number of whole courses of MTX 5 to 8 courses
6.19
[1.43 ; 26.8]
0.015
≥ 9 courses
6.80
[1.32 ; 35.1]
0.022
Number of metastases 0 Versus ≥ 1 Versus complete regimen
TE D
Premature MTX interruption
HCG normalization delay (month) centered at mean (1.9 months)
AC C
EP
Number of consolidation courses 0 courses 1 to 1.5 courses * 2 courses Versus ≥ 3 courses
Versus 1 to 4 courses
Number of whole courses of MTX (continued 1.13 [1.00 ; 1.28] 0.055 value) * 1,5 courses: patients who did not complete the second course of MTX because of bad tolerance
Table 2: Univariate analyses of risk factors in relapsing patients
19
ACCEPTED MANUSCRIPT Adjusted Hazard Ratio
IC95% HR
P
Post term pregnancy
8.66
[1.98 ; 37.9]
0.0042
Post abortion or ectopic
1.07
[0.14 ; 7.98]
0.95
5 to 8 courses
6.7
[1.54 ; 29.2]
0.011
≥ 9 courses
8.1
[1.54 ; 42.6]
0.014
Parameter
Versus post mole Number of whole courses of MTX
SC
Versus 1 to 4 courses
RI PT
Last pregnancy outcome
AC C
EP
TE D
M AN U
Table 3: Multivariate analyses of risk factors in relapsing patients
AC C
EP
TE D
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT
AC C
EP
TE D
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT
AC C
EP
TE D
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT