Background
Gestational trophoblastic neoplasia is a rare gynecological malignancy often treated at tertiary referral centers. Patients frequently travel long distances to obtain care for gestational trophoblastic neoplasia, which may affect cancer outcomes in these patients.
Objective
We examined the association between distance traveled to obtain care and disease burden at time of presentation as well as recurrence.
Study Design
We performed a retrospective cohort analysis of all patients diagnosed with gestational trophoblastic neoplasia from January 1995 to June 2015 at a high-volume tertiary referral center. Patients were included if they met International Federation of Gynecology and Obstetrics 2000 criteria for postmolar gestational trophoblastic neoplasia or had choriocarcinoma, placental-site trophoblastic tumor, or epithelioid trophoblastic tumor. Sixty patients were identified. Disease burden at presentation was examined using both the World Health Organization prognostic score and International Federation of Gynecology and Obstetrics. Patients who traveled more than 50 miles were considered long-distance travelers based on previous literature on the effect of distance traveled on cancer outcomes. Demographic, clinical, and pathological data were obtained by chart review. Bivariable comparisons were performed using the χ 2 test or Fisher exact test for categorical variables. The t test or Wilcoxon rank-sum test was used to compare continuous variables when normally or not normally distributed.
Results
Most patients presented at stage I (61%) with low-risk gestational trophoblastic neoplasia (70%). Median distance to care was 40 miles (range, 4–384). Eighteen patients (30%) had no insurance and 42 (70%) had either private or public insurance. Patients traveling more than 50 miles for care were more likely to have high-risk gestational trophoblastic neoplasia (46% vs 19%, P = .03), but there was no difference in recurrence (13% vs 11%, P = .89). Patients with high-risk gestational trophoblastic neoplasia lived 63 miles farther (92 vs 28 miles, P < .001) than patients with low-risk gestational trophoblastic neoplasia. Long-distance travelers had a longer period between antecedent pregnancy and gestational trophoblastic neoplasia diagnosis (10 weeks vs 4.5 weeks, P = .009) and were more likely to receive multiagent chemotherapy (86% vs 61%, P = .03).
Conclusion
In this cohort, long distance traveled to obtain care for gestational trophoblastic neoplasia was associated with an increased risk of presenting with high-risk disease and requiring multiagent chemotherapy for treatment. Patients with high-risk gestational trophoblastic neoplasia traveled nearly 100 miles to obtain care. There may be a delay in diagnosis in women traveling more than 50 miles to obtain care; however, we found no difference in recurrence risk for long-distance travelers.
Gestational trophoblastic neoplasia (GTN) is a rare gynecological malignancy that originates from the placenta. Incidence rates vary widely across populations, but rates of gestational trophoblastic disease, including benign and malignant trophoblastic growths, are 110 to 120 per 100,000 pregnancies in the United States, with rates of choriocarcinoma of 2–7 per 100,000 pregnancies.
Treatment of GTN is based on a combination of anatomic International Federation of Gynecology and Obstetrics (FIGO) staging and World Health Organization (WHO) prognostic scoring. Women are divided into 2 strata based on their WHO prognostic score. A score of 0–6 is considered low risk and a score of 7 or greater is considered high risk. Gestational trophoblastic tumors are highly chemosensitive and cure rates are excellent, particularly in women with low-risk disease. Women with high-risk disease require treatment with multiagent chemotherapy.
Like many rare diseases, women often travel considerable distances to receive treatment for GTN. Previously, multiple researchers have shown worse clinical outcomes in a variety of cancer types for patients traveling greater than 50 miles to obtain care. Patients who travel greater than 50 miles to obtain care are more likely to present with later stage disease and have increased mortality.
In addition to distance traveled to care, multiple socioeconomic factors, such as insurance status, have been shown to affect outcomes in other gynecological malignancies. Given the lack of data on the effect of these variables on outcomes in GTN, we aimed to evaluate the effect of distance traveled as well as other socioeconomic factors, such as insurance status, race, and employment, on the disease burden at presentation and outcomes in women being treated for GTN.
Our primary objective was to determine the association between distance traveled to care and disease burden (WHO score and FIGO stage) at presentation. Our secondary objective was to determine the association between distance traveled to care and recurrence.
Materials and Methods
Study design, setting, and participants
The Strengthening the Reporting of Observational Studies in Epidemiology guidelines were followed in the design and reporting of this study. Following institutional review board approval (number 14-2125), we performed a retrospective cohort study of all patients treated for GTN from January 1995 until June 2015 at the University of North Carolina Gynecologic Oncology Clinic in Chapel Hill, NC (UNC-CH). UNC-CH is a high-volume tertiary referral center.
Potential subjects were identified from the records of a weekly multidisciplinary tumor conference. Patients were included in the cohort if they met the FIGO 2000 criteria for postmolar GTN as defined in the following text and/or had a pathological diagnosis of choriocarcinoma, placental-site trophoblastic tumor (PSTT), or epithelioid trophoblastic tumor (ETT). FIGO 2000 criteria for GTN include the following: (1) an human chorionic gonadotropin (hCG) level plateau of ±10% for at least 3 weeks, (2) an elevation in serum hCG level by greater than 10% over 2 weeks, or (3) the presence of metastatic disease before hCG level plateau or rise.
Variables, data sources, and bias
Our primary outcome was disease burden at presentation. This was examined in 2 ways. In the first, the WHO score was used to classify patient as low risk (WHO score < 7) or high risk (WHO score ≥ 7). Additionally, patients were placed into groups based on the anatomic stage of disease as defined per the FIGO 2000 staging as stage I/II vs stage III/IV. Our secondary outcome of disease recurrence was defined as clinical recurrence based on serum hCG levels or biopsy-proven metastatic disease.
The primary exposure of interest was distance traveled to obtain care. This was calculated using the GPS coordinates of the patient’s documented home address and the GPS coordinates of UNC-CH to produce the distance between the 2 locations. Based on multiple previous studies looking at the effect of distance traveled to care on disease outcomes, a distance of 50 miles was chosen as an a priori cutoff for long- and short-distance traveled. Patients who traveled more than 50 miles to obtain care were considered long-distance travelers.
Conversely, patients who traveled less than 50 miles were considered short-distance travelers. Secondary exposures of interest included socioeconomic factors such as race, insurance status, and employment. Insurance status was defined as the insurance held by the patient at the time of presentation. For increased accuracy, referral records from outside clinics were used to confirm the patient’s insurance status at the time of presentation. Patients were divided into insured and uninsured. An individual blinded to the study objective performed chart abstraction to reduce information bias. Demographic, clinical, and pathological data were obtained by chart review.
Sample size and statistical analysis
Prior studies have reported rates of high-risk disease as low as 10% of all GTN patients. However, urban referral centers have reported much higher rates (35%) of high-risk GTN. The sample size for this study was calculated based on an alpha of 0.05 (2 tailed) with a power of 80%. We estimated that women traveling more than 50 miles to obtain care would have a 3-fold increased risk of presenting with high risk GTN (60%) from a baseline of 20%. Thus, we predicted needing 24 women in each exposure group (long distance and short distance traveled for care) to power this study.
Bivariable comparisons were performed using the χ 2 test or Fisher exact test for categorical variables as appropriate. The Student t test or Wilcoxon rank-sum test was used for continuous variables, depending on whether the variables were normally distributed. A value of P < .05 was considered significant for all analyses. All analyses were conducted using Stata statistical software (version 14.0; Stata Corp, College Station, TX).
Results
Overall cohort demographics
Sixty-four potential subjects were identified from January 1995 until June 2015. Of these, 60 patients met criteria for inclusion. The 4 patients excluded did not meet FIGO criteria for postmolar GTN (n = 2), had no available records (n = 1), or pathology review at our institution did not concur with the previously diagnosed PSTT (n = 1).
Of the 60 patients included in this cohort study, there were 41 invasive moles (68%), 13 choriocarcinomas (22%), 4 PSTTs (7%), and 2 ETTs (3.%). FIGO stage at the time of presentation was 61% stage I (n = 34), 0% stage II (n = 0), 25% stage III (n = 14), and 14% stage IV (n = 8). Overall, 70% (n = 42) presented with low-risk GTN and 30% (n = 18) had high-risk GTN at presentation.
The antecedent pregnancies were 52% complete moles (n = 31), 20% partial moles (n = 12), 20% miscarriages at < 20 weeks (n = 12), 7% intrauterine pregnancies > 20 weeks (n = 4), and 10% were unknown or undocumented (n = 6). The median time to GTN diagnosis from antecedent pregnancy, defined as date of beta hCG rise or plateau, was 6 weeks (range, 4–208 weeks). Mean age (SD) was 31.5 years old (±9.8 years). Median distance traveled for care was 40 miles (range, 4–384 miles). Eighteen patients (30%) had no insurance and 42 patients (70%) had either public or private insurance. The median follow-up time was 13.5 months (range, 0–287 months).
Distance traveled to care
Thirty-six patients were classified as short-distance travelers and 24 patients were long-distance travelers. Among short-distance travelers, the mean distance traveled was 28 miles (±11.2). Among long-distance travelers the mean distance was 128 miles (±91.1). There was no difference in age, race, parity, type of GTN, marital status, or insurance status between short-distance and long-distance travelers. Demographic characteristics of each group are summarized in Table 1 .
| Variables | < 50 miles (n = 36) | > 50 miles (n = 24) | Overall (n = 60) | P value |
|---|---|---|---|---|
| Age, y | 31 (±10.8) | 32 (±8.2) | 32 (±9.8) | NS |
| Race | NS | |||
| White | 14 (39%) | 8 (33%) | 22 (37%) | |
| Black | 7 (19%) | 6 (25%) | 13 (22%) | |
| Asian | 3 (8%) | 2 (8%) | 5 (8%) | |
| Hispanic | 12 (33) | 6 (25%) | 18 (30%) | |
| Other | 0 (0%) | 2 (8%) | 2 (3%) | |
| Parity | 1.3 (±1.3) | 1.5 (±1.4) | 1.4 (±1.3) | NS |
| GTN type | NS | |||
| Invasive mole | 29 (81%) | 12 (50%) | 41 (68%) | |
| Choriocarcinoma | 5 (14%) | 8 (33%) | 13 (22%) | |
| PSTT | 1 (3%) | 3 (13%) | 4 (7%) | |
| ETT | 1 (3%) | 1 (4%) | 2 (3%) | |
| Married | 16 (44%) | 9 (38%) | 25 (42%) | NS |
| Insured | 23 (64%) | 19 (79%) | 42 (70%) | NS |
| Employment | NS | |||
| Yes | 11 (31%) | 6 (25%) | 17 (28%) | |
| No | 9 (25%) | 5 (21%) | 14 (23%) | |
| Unknown | 16 (44%) | 13 (54%) | 29 (48%) |
a Continuous variables are reported as means ± SD, and categorical variables are reported as raw numbers with proportions.
Our primary outcome was the effect of distance traveled for care on disease burden at presentation. In regard to the WHO score, long-distance travelers were significantly more likely to present with high-risk disease (relative risk [RR], 2.4, 95% confidence interval [CI], 1.1–5.2). High-risk patients lived 60 miles farther from the hospital (92 vs 28 miles, P < .001).
In regard to FIGO stage, there was no difference in stage at diagnosis between short-distance and long-distance patients ( Table 2 ). There was also no difference in presentation at advanced stage (III-IV), with 48% of long-distance travelers presenting at advanced stage compared with 34% of short-distance travelers (RR, 1.4, 95% CI, 0.7–2.6). However, time from antecedent pregnancy to diagnosis of GTN, defined as the date of the beta hCG rise or plateau, was longer in patients who traveled greater than 50 miles to care (10 weeks vs 4.5 weeks, P = .009).
| Variables | < 50 miles (n = 36) | > 50 miles (n = 24) | Overall (n = 60) | P value |
|---|---|---|---|---|
| Prognostic score | .03 | |||
| 0–6 (low) | 29 (81%) | 13 (54%) | 42 (70%) | |
| ≥ 7 (high) | 7 (19%) | 11 (46%) | 18 (30%) | |
| Stage a | NS | |||
| I | 23 (66%) | 11 (52%) | 34 (61%) | |
| II | 0 (0%) | 0 (0%) | 0 (0%) | |
| III | 9 (26%) | 5 (24%) | 14 (25%) | |
| IV | 3 (9%) | 5 (24%) | 8 (14%) | |
| Recurrence | 4 (11%) | 3 (13%) | NS | |
| Chemotherapy b | .03 | |||
| Single agent | 30 (86%) | 14 (61%) | 44 (73%) | |
| Methotrexate | 28 (93%) | 12 (86%) | 40 (91%) | |
| Actinomycin-D | 2 (7%) | 2 (14%) | 4 (9%) | |
| Multiagent | 5 (14%) | 9 (39%) | 14 (23%) | |
| EMA-CO | 4 (80%) | 6 (67%) | 10 (71%) | |
| EMA-EP | 0 (0%) | 3 (33%) | 3 (21%) | |
| EMA | 1 (20%) | 0 (0%) | 1 (7%) |
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
