|
Advertisement |
|
|
||
 |
|
|||||
|
|||||
|
 |
||||||
|
||||||
|
||||||
|
Journal of Clinical Oncology, Vol 24, No 1 (January 1), 2006: pp. 52-58 © 2006 American Society of Clinical Oncology. DOI: 10.1200/JCO.2005.03.3043 Early Identification of Resistance to First-Line Single-Agent Methotrexate in Patients With Persistent Trophoblastic DiseaseFrom the Departments of Chemical Endocrinology and Obstetrics and Gynaecology, Radboud University Nijmegen Medical Centre, Nijmegen; Amphia Medical Centre, Breda; and the Dutch Working Party on Trophoblastic Tumors, Utrecht, the Netherlands Address reprint requests to Chris M. Thomas, PhD, Department of Chemical Endocrinology, 530 ACE, Radboud University Nijmegen Medical Centre, PO Box 9101, 6500 HB Nijmegen, the Netherlands; e-mail: C.Thomas{at}ace.umcn.nl
PURPOSE: A generally accepted definition for resistance to first-line single-agent chemotherapy for persistent trophoblastic disease (PTD) is lacking. In the present study, a normogram for serum human chorionic gonadotropin (hCG) from patients with normalization of serum hCG after first-line single-agent chemotherapy for PTD was constructed to identify patients resistant to this chemotherapy. PATIENTS AND METHODS: Between 1987 and 2004, data from 2,132 patients were registered at the Dutch Central Registry for Hydatidiform Moles. A normal serum hCG regression corridor was constructed for 79 patients with low-risk PTD who were cured by single-agent methotrexate (MTX) chemotherapy (control group). Another group of 29 patients with low-risk PTD needed additional alternative therapies (dactinomycin and multiagent chemotherapy) for failure of serum hCG to normalize with single-agent chemotherapy (study group). RESULTS: Serum hCG measurement preceding the fourth and sixth single-agent chemotherapy course proved to have excellent diagnostic accuracy for identifying resistance to single-agent chemotherapy, with an area under the curve (AUC) for receiver operating characteristic curve analysis of 0.949 and 0.975, respectively. At 97.5% specificity, serum hCG measurements after 7 weeks showed 50% sensitivity. CONCLUSION: In the largest study to date, we describe the regression of serum hCG levels in patients with low-risk PTD successfully treated with MTX. At high specificity, hCG levels in the first few courses of MTX can identify half the number of patients who are extremely likely to need alternative chemotherapy to cure their disease and for whom further treatment with single-agent chemotherapy will be ineffective.
A variety of pathologic types of trophoblast neoplasms are included in gestational trophoblastic disease, comprising villous malformations of trophoblast, hydatidiform mole subdivided in complete and partial hydatidiform mole, and nonvillous malformations of which choriocarcinoma is the most frequent.1 In persistent trophoblastic disease (PTD), trophoblastic activity remains after evacuation of a hydatidiform mole, as shown by subsequent unaltered high or even increasing serum human chorionic gonadotropin (hCG) concentrations. The internationally accepted definition released by the International Federation of Gynecology and Obstetrics (FIGO) defines PTD as a plateau in serum hCG for 3 weeks or an increase for 2 consecutive weeks.2 In the Netherlands, the Dutch Society for Obstetrics and Gynecology added to this definition for PTD the condition that at least one serum hCG measurement should exceed the 95th percentile (P95) of normal hCG regression after hydatidiform mole.3-5 The reported frequency of PTD is 20% in complete hydatidiform mole6 and 0.5% to 9.9% in partial hydatidiform mole.7-10 For patients with PTD, several staging classifications and prognostic scoring systems have been developed.3 To date, the FIGO 2000 scoring system for high- and low-risk PTD is advocated by The International Society for the Study of Trophoblastic Diseases and the International Society for Gynecological Cancer to generate univocal data.2 This scoring includes serum hCG, age, antecedent pregnancy, interval between end of pregnancy and start of treatment, previous failure of chemotherapy, and the localization and number of metastases.2 The Dutch classification for high- and low-risk PTD takes the same items into account, except for the number of metastases and serum hCG concentration.3, 11 Once diagnosed and scored, low-risk PTD is typically treated with single-agent chemotherapy (MTX or dactinomycin).12, 13 If staged as high-risk PTD, multiagent chemotherapy with etoposide, methotrexate, dactinomycin, cyclophosphamide, and vincristine (EMA/CO) is the most widely used therapy.12 Approximately 9% to 33% of patients treated with single-agent chemotherapy for low-risk PTD will require multiagent chemotherapy because resistance to the first-line drug or toxic adverse effects occurred.14-16 Repeated administration of MTX could induce MTX resistance,17 mainly as a result of downregulation of the expression levels of the reduced folate carrier that is responsible for transport of MTX into the cell or amplification of the gene for dihydrofolate reductase. This enzyme reduces dihydrofolate in tetrahydrofolate, which is essential in DNA synthesis.18, 19 To date, an internationally accepted definition for resistance to first-line chemotherapy is lacking. In some clinics, resistance to first-line chemotherapy is defined as a plateau or increase in serum hCG and/or development of new metastases.14-16, 20, 21 After single-agent chemotherapy with MTX, no increase in the number of second tumors has been observed.22 In contrast, patients with PTD who are treated with etoposide-containing multiagent chemotherapy have a 50% increased relative risk of developing secondary malignancies, in particular myeloid leukemia, colon carcinoma, and breast cancer, compared with an age-matched group.23 Furthermore, multiagent chemotherapy hastens the occurrence of menopause by 2 years compared with patients treated with single-agent MTX24 and can be accompanied by the occurrence of alopecia and GI symptoms. To identify patients not responding to single-agent chemotherapy in an early stage and, at the same time, to prevent unnecessary multiagent chemotherapy, we constructed a normal serum hCG regression corridor for patients successfully treated with single-agent chemotherapy.
Patients Between 1987 and 2004, 2,132 patients were registered at the Dutch Central Registry for Hydatidiform Moles. After informed consent, patients were registered by their referring clinician. The present study included, in retrospect, all low-risk patients with proven PTD according to Dutch guidelines (histologic diagnosis of mola hydatidosa or choriocarcinoma with serum hCG plateau or increase for 3 consecutive weekly measurements, with at least one measurement > P95 of normal regression according to an uneventful hCG regression published earlier4) if serum was available in our registry. In the Netherlands, patients with low-risk PTD are treated with MTX 1 mg/kg on days 1, 3, 5, and 7 and leucovorin 15 mg rescue on days 2, 4, 6, and 8, and this schedule is repeated every 14 days.13 Exclusion criteria were hysterectomy when PTD was diagnosed; cure after a second curettage; need for alternative therapies for MTX toxicity (none in this cohort); persistent low serum hCG levels (2 to 10 µg/L) after spontaneous regression or persistent low serum hCG after MTX treatment, for which no multiagent chemotherapy was administered; and recurrence of PTD after single-agent chemotherapy. Of the 108 eligible patients, serum hCG normalized under single-agent MTX treatment in 79 patients (73%; responders). The remaining 29 patients (27%) were treated with multiagent chemotherapy after the clinician decided that the disease failed to respond to single-agent chemotherapy.
Immunoassays
Statistics
Table 1 lists age, prechemotherapy serum hCG, number of courses of single-agent chemotherapy, hCG serum half-life, and, if applicable, number of courses of multiagent chemotherapy in both the control (n = 79) and study (n = 29) groups. No difference was observed between the groups for age (median age, 30 years in both groups). Expectedly, serum hCG concentration before the first course of single-agent chemotherapy was significantly less elevated in the control group compared with the study group (251 v 2,098 µg/L, respectively; P < .001). Significantly fewer courses of single-agent chemotherapy were administered in the control group (median, five courses; range, three to 17 courses) compared with the study group (median, seven courses; range, three to 16 courses; P = .003). As expected, the median serum hCG half-life was significantly shorter in the control group (1.02 weeks; 95% CI, 0.91 to 1.13 weeks) compared with the study group (1.92 weeks; 95% CI, 1.63 to 2.27 weeks; P < .001). The median number of courses of multiagent chemotherapy administered in the study group was four (range, three to seven courses).
For each single-agent treatment course, the P2.5, P50, and P97.5 of serum hCG concentration in the control group was calculated, and the results were plotted into a normogram (Fig 1A). In the control group, 2.5% of patients were cured after 2 weeks (after one MTX course), and 50% were cured after 8 weeks (after four MTX courses). After 17 weeks, the P97.5 of serum hCG in the control group remained at 4.3 µg/L and, thus, not less than the cutoff level of 2 µg/L. This is because, after 17 weeks, there were still some patients who needed additional MTX courses who were, at that time point, not yet normalized.
In Figure 1B, the hCG levels of the two weekly intervals for the 29 patients in the study group were plotted into the normogram derived from the serum hCG measurements of the control group. Before the start of the first MTX course, serum hCG in the study group exceeded the P97.5 of the control group in four (13.8%) of 29 patients. After 7 weeks (before the fourth MTX course), serum hCG in the study group exceeded the P97.5 of the control group in 22 (76%) of 29 patients. After 17 weeks (before the ninth MTX course), serum hCG in the study group exceeded the P97.5 of the control group in 25 (86%) of 29 patients. These results are in line with the finding of a significantly shorter median serum hCG half-life in the control group compared with the study group, as stated previously. Figure 2 shows the normogram for serum hCG+hCGß for the study group. This normogram was plotted in the normogram for serum hCG+hCGß derived from the control group. As shown, 2.5% of patients in the study group had normalization of serum hCG (ie, < 2 µg/L) after 9 weeks (before the fifth MTX course). These patients were administered multiagent chemotherapy for persistent low hCG levels after treatment with MTX. After 17 weeks (before the ninth MTX course), the P50 in the study group was 13.8 µg/L, and the P97.5 was 122 µg/L. The overlapping area between the study and control groups is darkly hatched. The area with hCG more than P97.5 of regression in the control group and with hCG less than P97.5 in the study group is lightly hatched.
Diagnostic accuracy of serum hCG half-life was derived from the calculation of an AUC of a ROC curve (Fig 3A). In addition, the diagnostic accuracies of serum hCG levels before the first (Fig 3B), fourth (Fig 3C), and sixth (Fig 3D) first-line single-agent chemotherapy course were established by the calculation of the AUCs of these ROC curves. The corresponding AUCs for these ROC curves are listed in Table 2. The AUC for serum hCG half-life throughout single-agent treatment was 0.833. The AUC of the serum hCG concentration preceding the first-single chemotherapy course was 0.828, which increased to 0.975 preceding the sixth course (after 11 weeks) of single-agent chemotherapy. This means that, at that time point, an excellent diagnostic accuracy was obtained because, after 11 weeks, 60% of the patients needing alternative therapy could be identified at the 97.5% specificity level.
Using serum hCG before the start of single-agent chemotherapy as a diagnostic tool to identify patients who will need alternative chemotherapeutic treatment for PTD resulted in the identification of 14% of patients who needed alternative therapy at the 97.5% specificity level (hCG serum cutoff, 9,600 µg/L). Serum hCG levels obtained before the start of the fourth course of therapy to predict the need for alternative therapy would identify as much as 50% of patients at the 97.5% specificity level (cutoff, 56 µg/L; Table 2).
The objective of this study was to clarify whether it is possible to identify, at an early stage and from a single hCG measurement, patients with PTD who will not be cured by single-agent chemotherapy. For this reason, we established a serum hCG normogram with data from 79 patients with low-risk PTD who were cured by single-agent MTX chemotherapy. This normogram was applied to hCG measurements from 29 PTD patients who needed alternative therapy after single-agent chemotherapy failed to establish normalization of serum hCG. Using this normogram, we found that it is possible to identify 14% of patients who will need alternative therapy before the start of first-line single-agent chemotherapy with 97.5% specificity. Likewise, on the basis of serum hCG measurements obtained just before the fourth MTX course, we were able to identify 50% of patients who would not respond to single-agent chemotherapy at 97.5% specificity. The international criteria used to diagnose PTD include an abnormal serum hCG regression, with a plateau for 3 weeks or an increase for 2 weeks.2 Normal serum hCG regression corridors for uneventful hydatidiform moles have been constructed, with 50% of patients having normalized serum hCG measurements between 6 and 14 weeks after evacuation.4, 20, 21, 26, 27 The Dutch Society for Obstetrics and Gynecology included the extra condition in the diagnosis for PTD that, in addition to a serum hCG plateau or increase for 3 consecutive weeks, at least one serum hCG measurement should exceed the P95 of normal hCG regression, as derived from an hCG normogram constructed from data of 130 patients with uneventful hydatidiform mole.4 There are two reasons for the inclusion of this additional condition. First, PTD, as defined by exceeding the P95 of normal regression, is diagnosed 2 weeks earlier than if the definition for PTD of an hCG plateau or increase for 3 consecutive measurements is used.21, 27 Second, use of a normal hCG regression corridor prevents overtreatment. One study indicated that 15% of patients with a plateau or increase for 3 weeks but without a serum hCG measurement exceeding the P95 will have spontaneous serum hCG remission afterwards.4 Two retrospective studies showed that, in patients treated for PTD on the basis of a plateau or increase in serum hCG for 3 consecutive weeks, 29% had serum hCG measurements less than the P90 of normal regression20 8% had measurements less than the P95 of normal regression.4 Although firm criteria for the diagnosis of PTD exist and the use of a normogram for normal serum hCG regression identifies patients with PTD earlier and prevents overtreatment, no firm criteria for the diagnosis of resistance to first-line single-agent therapy exist. A generally accepted definition for resistance to first-line chemotherapy is lacking, although some clinics define resistance to chemotherapy by a plateau or increase in serum hCG or by detection of (new) metastases.14, 15, 20, 21 Rotmensch et al20 described the only normogram for normalization of hCG in 21 patients with nonmetastatic trophoblastic disease treated with MTX. They reported that the median time for serum hCG to normalize in successfully treated patients with PTD is 50 days. The authors stated that, in successfully treated patients with low-risk PTD, the median time to normalization is comparable to patients with uncomplicated regression of serum hCG after hydatidiform mole (50 days). Our results are in line with their results because we found that serum hCG normalizes in 50% of patients after five 2-week courses of MTX. Our group previously developed a normogram for serum hCG regression after uneventful hydatidiform mole.4 In that study, median time for normalization was 11 weeks, which is comparable to the five 2-week courses that 50% of patients needed for normalization of hCG after successful single-agent chemotherapy. Serum hCG in two patients in the study by Rotmensch et al20 failed to normalize after primary chemotherapy. Serum hCG of these two patients was greater than the P90 of the normal regression with MTX at the start of the MTX treatment. These investigators did not explore the diagnostic accuracy of serum hCG levels to detect resistance to single-agent chemotherapy in low-risk PTD. Shigematsu et al,21 in their study comprising 24 patients with PTD, indicated that woman who would show resistance to single-agent chemotherapy later on (as concluded from increasing hCG for 2 weeks, a plateau for 3 weeks, or the development of new metastases) exceeded the P95 of normal hCG regression significantly earlier than patients who would respond well to single-agent chemotherapy. With the presented serum hCG regression corridor for patients treated with first-line single-agent MTX, we have an excellent diagnostic tool to identify patients with resistance to first-line single-agent chemotherapy at an early stage with great certainty. Before the fourth course of single-agent chemotherapy (after 7 weeks), serum hCG identifies 50% of patients needing alternative therapy at the 97.5% specificity level. The birthrate in the Netherlands is approximately 200,000 persons per year. With an estimated incidence of hydatidiform mole of 1 per 1,000 to 2,000 pregnancies,28 the 18-year period of our study could yield 1,800 to 3,600 patients with hydatidiform mole, of whom we would expect 180 to 360 to have PTD. Registration with the Dutch Central Registry for Hydatidiform Moles is on a voluntary basis, and we registered 2,132 patients with hydatidiform mole throughout the entire study period, of whom 108 patients with PTD were eligible for inclusion in our study. Because of the low incidence of PTD and, in particular, the need for alternative therapy, it is not feasible to perform a prospective study, and we used our retrospective data for a proposal for a new treatment regimen for patients with PTD. In analogy with the regression curve for serum hCG after evacuation of uneventful hydatidiform mole as earlier published by our group,4 we developed a curve for the early detection of resistance to first-line single-agent MTX chemotherapy. Because overtreatment with multiagent chemotherapy is highly undesirable, we increased the cutoff from P95 to P97.5 for normal regression of patients successfully treated with first-line single-agent chemotherapy. If before the fourth course of single-agent chemotherapy the serum hCG concentration exceeds the P97.5 of this normal regression curve, it can be concluded with 50% sensitivity that a change of treatment is mandatory to gain cure (Fig 2, lightly hatched area). In line with the internationally accepted criteria for treatment of PTD, the course of hCG (plateau or increase) should also be included in the decision to start alternative therapy. The choice of second-line chemotherapy could be multiagent chemotherapy with EMA/CO. In the United Kingdom, McNeish et al15 described a successful regimen in which patients with resistance to MTX-leucovorin were administered either dactinomycin if serum hCG was less than 100 U/L or EMA/CO if serum hCG exceeded 100 U/L. In this study, 58 of 67 patients treated secondarily with dactinomycin were cured by this drug. In conclusion, in the largest study to date, we describe the regression of serum hCG levels in patients with low-risk PTD treated with MTX. From these data, it can be concluded that serum hCG levels before the fourth course of MTX can identify half the number of patients who are highly likely to need alternative therapy to cure their disease and for whom further treatment with single-agent chemotherapy will be ineffective.
The authors indicated no potential conflicts of interest.
We thank Dr Paul N. Span for his expert advice.
Authors' disclosures of potential conflicts of interest and author contributions are found at the end of this article.
1. Bentley RC: Pathology of gestational trophoblastic disease. Clin Obstet Gynecol 46:513-522, 2003[CrossRef][Medline] 2. FIGO Oncology Committee: FIGO staging for gestational trophoblastic neoplasia 2000. Int J Gynaecol Obstet 77:285-287, 2002[CrossRef][Medline] 3. Hancock BW: Staging and classification of gestational trophoblastic disease. Best Pract Res Clin Obstet Gynaecol 17:869-883, 2003[CrossRef][Medline] 4. Yedema KA, Verheijen RH, Kenemans P, et al: Identification of patients with persistent trophoblastic disease by means of a normal human chorionic gonadotropin regression curve. Am J Obstet Gynecol 168:787-792, 1993[Medline] 5. Dutch Society for Obstetrics and Gynaecology: Richtlijn Molazwangerschap 1.1 (Guideline Mole Pregnancy 1.1), 2005. http://www.nvog.nl 6. Berkowitz RS, Goldstein DP: Gestational trophoblastic disease. Cancer 76:2079-2085, 1995[CrossRef][Medline] 7. Bagshawe KD, Lawler SD, Paradinas FJ, et al: Gestational trophoblastic tumours following initial diagnosis of partial hydatidiform mole. Lancet 335:1074-1076, 1990[CrossRef][Medline] 8. Berkowitz RS, Goldstein DP, Bernstein MR: Natural history of partial molar pregnancy. Obstet Gynecol 66:677-681, 1985[Medline] 9. Goto S, Yamada A, Ishizuka T, et al: Development of postmolar trophoblastic disease after partial molar pregnancy. Gynecol Oncol 48:165-170, 1993[CrossRef][Medline] 10. Rice LW, Berkowitz RS, Lage JM, et al: Persistent gestational trophoblastic tumor after partial hydatidiform mole. Gynecol Oncol 36:358-362, 1990[CrossRef][Medline] 11. Dutch Society for Obsterics and Gynaecology: Richtlijn Persisterende Trophoblast Ziekte 1.1 (Guideline Persistent Trophoblastic Disease 1.1), 2005. http://www.nvog.nl 12. Ng TY, Wong LC: Diagnosis and management of gestational trophoblastic neoplasia. Best Pract Res Clin Obstet Gynaecol 17:893-903, 2003[CrossRef][Medline] 13. Osborne R, Gerulath A: What is the best regimen for low-risk gestational trophoblastic neoplasia? A review. J Reprod Med 49:602-616, 2004[Medline] 14. Lurain JR: Treatment of metastatic gestational trophoblastic tumors, in Hancock BW, Newlands ES, Berkowitz R (eds): Gestational Trophoblastic Disease. London, United Kingdom, Chapman & Hall, 1997, pp 199-209 15. 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 20:1838-1844, 2002 16. 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 96:616-620, 2005[CrossRef][Medline] 17. Sobrero A, Aschele C, Rosso R, et al: Rapid development of resistance to antifolates in vitro: Possible clinical implication. J Natl Cancer Inst 83:24-28, 1991 18. Moscow JA, Gong M, He R, et al: Isolation of a gene encoding a human reduced folate carrier (RFC1) and analysis of its expression in transport-deficient, methotrexate-resistant human breast cancer cells. Cancer Res 55:3790-3794, 1995 19. Gorlick R, Goker E, Trippett T, et al: Intrinsic and acquired resistance to methotrexate in acute leukemia. N Engl J Med 335:1041-1048, 1996 20. Rotmensch J, Rosenshein NB, Block BS: Comparison of human chorionic gonadotropin regression in molar pregnancies and post-molar nonmetastatic gestational trophoblastic neoplasia. Gynecol Oncol 29:82-86, 1988[CrossRef][Medline] 21. Shigematsu T, Kamura T, Saito T, et al: Identification of persistent trophoblastic diseases based on a human chorionic gonadotropin regression curve by means of a stepwise piecewise linear regression analysis after the evacuation of uneventful moles. Gynecol Oncol 71:376-380, 1998[CrossRef][Medline] 22. 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 20:1838-1844, 2002 23. Seckl MJ, Rustin GJ: Late toxicity after therapy of gestational trophoblastic tumors, in Hancock BW, Newlands ES, Berkowitz R, et al (eds): Gestational Trophoblastic Disease (online book). International Society for the Study of Trophoblastic Diseases, 2003, pp 367-379. http://www.isstd.org/gtd/20%20word%20for%20print.pdf 24. Rustin GJ, Newlands ES, Lutz JM, et al: Combination but not single-agent methotrexate chemotherapy for gestational trophoblastic tumors increases the incidence of second tumors. J Clin Oncol 14:2769-2773, 1996 25. Bower M, Rustin GJ, Newlands ES, et al: Chemotherapy for gestational trophoblastic tumours hastens menopause by 3 years. Eur J Cancer 34:1204-1207, 1998[CrossRef][Medline] 26. Thomas CM, Segers MF, Houx PC: Comparison of the analytical characteristics and clinical usefulness in tumour monitoring of fifteen hCG(-beta) immunoassay kits. Ann Clin Biochem 22:236-246, 1985[Medline] 27. Schlaerth JB, Morrow CP, Kletzky OA, et al: Prognostic characteristics of serum human chorionic gonadotropin titer regression following molar pregnancy. Obstet Gynecol 58:478-482, 1981[Medline] 28. Behtash N, Ghaemmaghami F, Honar H, et al: Is normal beta-hCG regression curve helpful in the diagnosis of persistent trophoblastic disease? Int J Gynecol Cancer 14:980-983, 2004[CrossRef][Medline] 29. Steigrad SJ: Epidemiology of gestational trophoblastic diseases. Best Pract Res Clin Obstet Gynaecol 17:837-847, 2003[CrossRef][Medline] Submitted June 29, 2005; accepted October 5, 2005.
|
||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
|
|||||||||||
|
|||||||||||
|
|
Copyright © 2006 by the American Society of Clinical Oncology, Online ISSN: 1527-7755. Print ISSN: 0732-183X
|
TOP
ABSTRACT
INTRODUCTION
