Clinical characteristics of the study population
Between July 2011 and April 2018, 330 eligible patients were enrolled in the study, 19 patients were lost to follow-up, and clinical information could be obtained on 311 patients: 107 in the GnRHa group and 204 in the control group. There were 15 patients in the GnRHa group who required initiation of myeloablative chemotherapy with a shorter interval. The shortest follow-up was 3 years and the longest was 5 years. Baseline population characteristics are described in Table 1. In the GnRHa group, 64.49% (69 of 107) received standard chemotherapy prior to the myeloablative regimen, which was significantly fewer than 83.33% (170 of 204 ) in the control group (P<0.001). There were no significant differences in age, TBI exposure, or pre-transplant serum levels of FSH, LH, and estradiol between the two groups.
Primary endpoints: incidence of POI and recovery of menstruation
Overall, 81.67% (254 of 311) of patients had POI; 6.11% (19 of 311) experienced resumption of menstruation; and 12.22% (38 of 311) received HRT but did not have spontaneous resumption of menstruation and yet met the diagnostic criteria for POI. In the GnRHa group, 78.50% (84 of 107) had POI versus 83.33% (170 of 204) in the control group. In contrast, 7.48% (8 of 107) of patients resumed menses spontaneously in the GnRHa group versus 4.39% (11 of 204) in the control group. In the GnRHa group, 14.02% (15/107) of patients whose ovarian function was unknown received HRT within one year of HSCT versus 11.27% (23/204) in the GnRHa group. control. There were no significant differences in the primary follow-up outcome between the two groups (P= 0.568) (Fig. 1).
Considering that there were fewer patients who had received conventional chemotherapy before the myeloablative regimen in the GnRHa group than in the control group, we performed a subgroup analysis to minimize bias. The results showed that 239 patients received chemotherapy prior to the myeloablative regimen for HSCT. In the GnRHa group, 82.61% (57/69) had POI vs. 84.12% (143/170) in the control group, 2.90% (2/69) resumed menstruation vs. 2 .94% (5/170) of the control group, and 14.49% (10/69) had an unknown outcome vs. 12.94% (22/170) of the control group. The difference between the two groups was not statistically significant (P= 0.946) (see Fig. 2a). Among the 72 patients who did not receive chemotherapy before myeloablative chemotherapy and HSCT, 71.05% (27/38) had POI versus 79.41% (27/34) in the control group, 15.79% ( 6/38) returned to menstruation vs. 17.65% (6/34) in the control group, and 13.16% (5/38) had an unknown outcome 2.94% (1/34) in the control group . The difference between the two groups was not statistically significant (P= 0.365) (see Fig. 2b).
Subsequently, binomial log regression was applied to determine the RR of POI and examine possible factors influencing POI (Table 2). Compared with women without GnRHa treatment, the unadjusted RR of POI was 1.29 (95% CI: 0.80–2.07, P= 0.293). There was no statistical significance even after adjusting for factors including conventional chemotherapy before the myeloablative regimen for HSCT and TBI exposure. (adjusted RR: 1.19, 95% CI: 0.73–1.93, P= 0.487).
As expected, mean FSH and E2 levels were significantly lower in the GnRHa group than in the control group after HSCT, but this difference was no longer observed during long-term follow-up (Fig. 3).
Secondary objectives: perimenopausal symptoms
Perimenopausal symptoms associated with hypoestrogenism were reported in 70.42% (219 of 311) of patients. Among patients who received GnRHa co-treatment, 62.62% (67 of 107) complained of perimenopausal symptoms, which was significantly less than 74.51% (152 of 204) in the control group (RR: 1.47 , 95% CI: 1.05–2.06, P= 0.027). The difference remained statistically significant after adjusting for factors including conventional chemotherapy prior to myeloablative regimen for HSCT (RR: 1.488, 95% CI: 1.055–2.098, P= 0.024) and TBI exposure (RR: 1.44, 95% CI: 1.02–2.02, P= 0.036; adjust for two factors simultaneously, RR: 1.46, 95% CI: 1.04–2.06, P= 0.031).
Of the 6 menopausal symptoms, the order of reported symptoms was mood swings (71.64%, 48 of 67), hot flashes (58.21%, 39 of 67), sleep disturbances (29.85%, 20 of 67), sexual dysfunction (26.87%, 18 of 67), joint/muscle pain (2.99%, 2 of 67), and dysuria/urinary frequency (1.49%, 1 of 67) in the group of GnRHa and hot flashes (61.18%, 93 of 152), mood changes (60.53%, 92 of 152), sleep disturbance (39.47%, 60 of 152), sexual dysfunction (35.53% , 54 of 152), joint/muscular pain (12.50%, 19 of 152) and dysuria/urinary frequency (2.63%, 4 of 152) in the control group (Table 3).