Wolfberg, Adam J. MD, MPH‡§¶; Feltmate, Colleen MD*‡¶; Goldstein, Donald P. MD*‡¶; Berkowitz, Ross S. MD*‡¶; Lieberman, Ellice MD, DrPH†‡¶
From the *Division of Gynecologic Oncology, New England Trophoblastic Disease Center, Donald P. Goldstein M.D. Trophoblastic Tumor Registry; †Division of Maternal Fetal Medicine and ‡Department of Obstetrics, Gynecology, and Reproductive Biology, Brigham and Women's Hospital; §Department of Obstetrics and Gynecology, Massachusetts General Hospital; and ¶Harvard Medical School Boston, Massachusetts.
Reprints are not available. Address correspondence to: Adam Wolfberg, MD, MPH, Department of Obstetrics, Gynecology, and Reproductive Biology, Brigham and Women's Hospital, 75 Francis Street, Boston, MA 02115; e-mail: email@example.com.
Received March 7, 2004. Received in revised form May 19, 2004. Accepted May 27, 2004.
OBJECTIVE: Complete hydatidiform molar pregnancies occur in approximately 1 of 1,000 conceptions. After uterine evacuation of the trophoblastic tissue, women are followed up with serial serum human chorionic gonadotropin (hCG) measurements. Patients are considered to have attained remission when their hCG level spontaneously declines to an undetectable level and remains there during a 6-month follow-up period. This standard effectively detects all disease recurrence; however, it is resource intensive, delays child bearing, and is subject to significant noncompliance. Our objective was to determine the risk of disease recurrence after hCG spontaneously declines to undetectable levels.
METHODS: We used a database from the New England Trophoblastic Disease Center to analyze hCG levels in patients with complete molar pregnancies.
RESULTS: Among 1,029 women with complete molar pregnancy and complete data, 15% developed persistent gestational trophoblastic neoplasia. The rate of persistent neoplasm among those whose hCG level fell spontaneously to undetectable levels was 0.2% (2/876, 95% confidence interval 0–0.8%). No women developed persistent gestational trophoblastic neoplasia after their hCG level fell to undetectable levels using an assay with a sensitivity of 5 mIU/mL (n = 82, 95% confidence interval 0–4.5%).
CONCLUSION: Based on our experience with women with complete hydatidiform molar pregnancies whose hCG values spontaneously fell to undetectable levels after molar evacuation, we conclude that the risk of recurrent neoplasm after hCG levels fall to less than 5 mIU/mL approaches zero.
LEVEL OF EVIDENCE: II-2
Complete hydatidiform molar pregnancies arise from an ovum apparently lacking maternal chromosomes that is fertilized by either a sperm that duplicates its own chromosomes, or by 2 sperm.1,2 The resulting conception contains exclusively paternal chromosomes, and does not result in development of a fetus. Molar pregnancies—the most common of the gestational trophoblastic neoplasias(GTNs)—occur approximately 1 in 1,000 pregnancies worldwide, although there are geographic differences in incidence.3,4 Hydatidiform mole is usually diagnosed using ultrasound either incidentally, or after a pregnant patient presents with bleeding early in her pregnancy.
After uterine evacuation of the molar tissue, 18–28% of women with complete hydatidiform molar pregnancies will develop evidence of persistent neoplasia.5–8 To detect persistence, women are followed up with serial serum human chorionic gonadotropin (hCG) measurements. A persistent plateau or rise in the hCG level indicates persistent neoplasm, and the need for chemotherapeutic management. The patient is considered in remission if the hCG values spontaneously decline to undetectable levels and remain there during a 6-month follow-up period.9
Investigators have attempted to identify factors that would predict persistent neoplasm in order to shorten follow-up for those patients who spontaneously attain remission and reduce the interval between the diagnoses of molar pregnancy and the diagnosis of persistent neoplasm. Research into clinical findings,7 histopathologic variables,10 and the ratio of the free β-subunit to the complete hCG molecule11 has not identified highly sensitive or specific predictors of persistent disease.
After patients with complete mole attain undetectable hCG levels, they are followed up with monthly hCG levels to assure sustained remission. Although this extended follow-up process effectively detects all disease recurrences, it is also onerous: it delays child bearing, causes anxiety,12 is resource-intensive, and is complicated by significant noncompliance.13 However, once patients attain undetectable hCG levels, their risk of trophoblastic tumor relapse is unclear.14 The current study was undertaken to review our experience with hCG follow-up in patients with complete moles at the New England Trophoblastic Disease Center and to estimate in particular the risk of tumor recurrence after hCG levels become undetectable.
MATERIALS AND METHODS
We used the Donald P. Goldstein Trophoblastic Tumor Registry of the New England Trophoblastic Disease Center, a clinical unit founded in 1965 to care for patients with GTN, and to serve as a referral center for physicians in the community caring for their own patients with these diseases. We analyzed data from 1973, the year the hCG radioimmunoassay became commercially available in the United States, through 2001. We identified all patients with a diagnosis of complete molar pregnancy on the basis of a pathology report. Patients with partial molar pregnancies or other gestational trophoblastic neoplasms were excluded from our analysis. We also excluded those patients for whom we had incomplete information regarding their pathology, date of molar evacuation, or for whom we had incomplete hCG follow-up data.
We then categorized patients with molar pregnancies into 2 groups on the basis of their subsequent hCG testing: persistent GTN and spontaneous remission. Persistence was diagnosed when a patient's hCG value increased for 2 or more values or if the level failed to decline by more than 10% over 3 weekly hCG values (plateauing).1 Sustained remission was diagnosed after 6 months of follow-up hCG testing after molar evacuation if there was no evidence of rising or plateauing hCG values. However, we evaluated all available follow-up hCG values for subjects who were followed up for more than 6 months after their molar evacuation. Patients who became pregnant during the 6-month follow-up interval (n = 44) were considered to be in remission if they had evidence of a normal pregnancy without evidence of choriocarcinoma or placental site tumor.
We determined the occurrence of persistence among women who reached an undetectable hCG level. Our database does not capture the sensitivity of the assays used, and a number of different assays with sensitivities varying from 5 mIU/mL to 10 mIU/mL were used during our study period. By 1993, assays with a sensitivity of less than 5 mIU/mL were universally used in the United States. To assess the potential impact of changing assays on our results, we performed separate analyses examining the occurrence of persistence among women reaching a value of less than assay before and after 1993. For each period the proportion of women with persistent neoplasm was determined, and the 95% confidence intervals around that rate were calculated. This study was approved by the Human Research Committee at our institution.
We identified 2,120 women followed up at our center for complete hydatidiform mole between 1973 and 2001. Women with less than 6 months of follow-up (n = 877) were excluded from the analysis. Of the 877, 817 (93.2%) were lost after their hCG assay fell to less than assay, and 60 (6.8%) were lost to follow-up without their hCG falling to undetectable. We have no record of a patient who was lost to follow-up subsequently being diagnosed with persistent neoplasm. We also excluded the 173 women with data missing for either their evacuation or their pathology report. Finally, we excluded 41 subjects from the analysis because they did not meet current hCG criteria for persistent neoplasm but received chemotherapy anyway. The 1,029 remaining women comprised our study population, of whom 82 were followed up between 1993 and 2001 and the remainder followed up between 1973 and 1992.
Overall, persistent neoplasm was diagnosed in 153 of the 1,029 women (15%) on the basis of hCG criteria. Women with persistent neoplasm were older than women who spontaneously entered remission (mean 27.7 years compared with 25.9 years, P < .01), and were somewhat more likely to have a hCG level on presentation above 100,000 mIU/mL (20.3% compared with 14.7%, P = .08), although this difference did not reach statistical significance. Women with persistent neoplasm were not more likely to have uterine size greater than dates on presentation, to be multiparous, or to have had a prior mole than women who spontaneously entered remission (Table 1). There were only 2 women who had persistent neoplasm diagnosed after their hCG level fell to less than assay, yielding an overall rate of 0.2% (2/876, 95% confidence interval 0–0.8%). Both of these cases occurred in 1975, when the hCG assay used had a sensitivity of less than 10 mIU/mL. None of the 82 subjects whose hCG level spontaneously fell to undetectable levels after 1993 (when the assay had a sensitivity of less than 5 mIU/mL) were found subsequently to develop persistent disease. Between 1993 and 2001 the rate of recurrent neoplasm after hCG levels fell spontaneously to undetectable levels was 0% (0/82, 95% confidence interval 0–4.5%).
Among the 876 women who went into remission, the median time required for hCG levels to fall to less than assay after molar evacuation was 64 days (Fig. 1). The range was 1–351 days.
We conclude that the risk of recurrent neoplasm after hCG values fall to less than 5 mIU/mL is very small and may even approach zero, based on our experience with 876 women with complete hydatidiform molar pregnancies whose hCG value spontaneously fell to undetectable levels after evacuation. These data suggest that postevacuation monitoring for women with uncomplicated complete molar pregnancies may be significantly shortened without compromising their risk for unrecognized persistent trophoblastic neoplasia, particularly if a radioimmunoassay with a sensitivity of 5 mIU/mL or less is used. An abbreviated follow-up period would result in savings of time and resources, while lessening anxiety for women with this disease.
If follow-up of the nonpersistent molar pregnancy were limited to 2 consecutive undetectable hCG titers measured 1 week apart, follow-up could be reduced from 26 weeks to 11 weeks for 50% of women with nonpersistent disease (more than 40% of all women with molar pregnancies) and to 15 weeks for 75% of women with nonpersistent disease. This would reduce the follow-up period by 60% for half of women with nonpersistent molar pregnancies and by 40% for another 25% of these patients, resulting in significant savings measured in cost, anxiety, and fertility. Overall, the follow-up period for nonpersistent disease would be shortened for 97% of women with this condition.
Although there are no data specific to the burden of this disease, the United States incidence of complete molar pregnancy can be estimated at approximately 6,200 annually, and the worldwide incidence can be estimated to be approximately 200,000 annually by multiplying the incidence of complete molar pregnancy by the incidence of pregnancy.15,16 Complete molar pregnancy is more common in some less-developed countries with overburdened health-care systems. Reduced hCG follow-up would result in a decline in the time and cost of monitoring and would result in a substantial savings of scarce resources.
Our findings are particularly significant for women at high risk for complete molar pregnancy: those who are attempting to become pregnant after age 35, for whom the risk of complete molar pregnancy is double that faced by younger women.3,17,18 These women face a time-limited window of fertility due to their age and consequently would benefit if the duration of hCG follow-up could be safely limited.
Our data are specific to complete hydatidiform molar pregnancies and do not apply to partial molar pregnancies. However, the risk of persistent GTN following a partial mole is appreciably lower at 2–4%.1
After achieving undetectable hCG levels, the risk of persistent trophoblastic neoplasia after complete mole appears to be very low and may approach zero. It would be helpful if other centers with extensive experience in monitoring patients with molar pregnancy would review their experience with the frequency of relapse after attaining hCG values less than 5 mIU/mL. If our data are replicated by other centers, we believe that patient safety will not be compromised if women with complete molar pregnancies are discharged from surveillance after they achieve 2 undetectable hCG levels (hCG less than 5 mIU/mL.) 1 week apart.
1. Berkowitz RS, Goldstein DP. Chorionic tumors. N Engl J Med 1996;335:1740–8.
2. Soper JT, Clarke-Pearson D, Hammond CB. Metastatic gestational trophoblastic disease: prognostic factors in previously untreated patients. Obstet Gynecol 1988;71:338–43.
3. Palmer JR. Advances in the epidemiology of gestational trophoblastic disease. J Reprod Med 1994;39:155–62.
4. Tham BW, Everard JE, Tidy JA, Drew D, Hancock BW. Gestational trophoblastic disease in the Asian population of Northern England and North Wales. BJOG 2003;110:555–9.
5. Lurain JR, Brewer JI, Torok EE, Halpern B. Natural history of hydatidiform mole after primary evacuation. Am J Obstet Gynecol 1983;145:591–5.
6. Kohorn EI. Hydatidiform mole and gestational trophoblastic disease in Southern Connecticut. Obstet Gynecol 1982;59:78–94.
7. Curry SL, Hammond CB, Tyrey L, Creasman WT, Parker RT. Hydatidiform mole: diagnosis, management, and long-term followup of 347 patients. Obstet Gynecol 1975;45:1–8.
8. Yuen BH, Cannon W. Molar pregnancy in British Columbia: estimated incidence and postevacuation regression patterns of the beta subunit of human chorionic gonadotropin. Am J Obstet Gynecol 1981;139:316–9.
9. Berkowitz RS, Goldstein DP, Bernstein MR. Management of nonmetastatic trophoblastic tumors. J Reprod Med 1981;26:208–12.
10. Ayhan A, Tuncer ZS, Halilzade H, Kucukali T. Predictors of persistent disease in women with complete hydatidiform mole. J Reprod Med 1996;41:591–4.
11. Berkowitz R, Ozturk M, Goldstein DP, Bernstein M, Hill L, Wands JR. Human chorionic gonadotropin and free subunits’ serum levels in patients with partial and complete hydatidiform moles. Obstet Gynecol 1989;74:212–5.
12. Wenzel L, et al. Quality of life after gestational trophoblastic disease. J Reprod Med 2002;47:387–94.
13. Feltmate CM, Batorfi J, Fulop V, Goldstein DP, Doszpod J, Berkowitz RS. Human chorionic gonadotropin follow-up in patients with molar pregnancy: a time for reevaluation. Obstet Gynecol 2003;101:732–6.
14. Batorfi J, Vegh G, Szepesi J, Szigetvari I, Doszpod J, Fulop V. How long should patients be followed after molar pregnancy? Analysis of serum hCG follow-up data. Eur J Obstet Gynecol Reprod Biol 2004;112:95–7.
16. Ventura SJ, Mosher WD, Curtin SC, Abma JC, Henshaw S. Trends in pregnancy rates for the United States, 1976-1997: an update. Nat Vital Stat Rep 2001;49:1–9.
17. Sebire NJ, Foskett M, Fisher RA, Rees H, Seckl M, Newlands E. Risk of partial and complete hydatidiform molar pregnancy in relation to maternal age. BJOG 2002;109:99–102.
18. Paradinas FJ, Browne P, Fisher RA, Foskett M, Bagshawe KD, Newlands E. A clinical, histopathological and flow cytometric study of 149 complete moles, 146 partial moles and 107 non-molar hydropic abortions. Histopathology 1996;28:101–10.