We screened 362 female AIS patients who underwent PSF surgery. Among them, 156 were excluded because of Schwab grade III–VI osteotomy (2), TXA or other hemostatic drugs (96), menstrual cycle issues (48; before menarche, irregular menstrual cycle, or menstrual cycle longer than 30 days or shorter than 28 days), and incomplete medical information (10). Study participants included 206 remaining patients. Among them, 41 were included in Group A (progesterone group) and 165 in Group B (no progesterone group) (Fig. 1). No statistically significant differences were found in the demographic and radiographic data (age, height, weight, Risser sign, Lenke classification, Cobb angle, bending Cobb angle, and CS [% used]) between the two groups (Table 1). Preoperative coagulation function variables, including platelet count, fibrinogen level, prothrombin time, thrombin time, and APTT, are presented in Table 2. Coagulation functions were not significantly different between the two groups.
Table 1
Demographic and Radiographic Data
| Group A | Group B | P |
| n = 41 | N = 165 | |
Age (year) | 14.8 ± 2.2 | 14.9 ± 1.8 | 0.64 |
Height (cm) | 161.2 ± 6.4 | 160.4 ± 6.7 | 0.27 |
Weight (kg) | 47.8 ± 7.1 | 48.6 ± 7.4 | 0.43 |
Risser sign | 3.6 ± 1.1 | 3.7 ± 0.9 | 0.79 |
Major Cobb angle (°) | 53.9 ± 15.9 | 52.8 ± 15.2 | 0.16 |
Bending Cobb angle (°) | 29.7 ± 21.2 | 29.2 ± 16.1 | 0.28 |
Lenke classification |
1 | 18 | 80 | 0.68 |
2 | 9 | 37 | 0.53 |
3 | 3 | 11 | 0.66 |
4 | 1 | 4 | 0.43 |
5 | 6 | 19 | 0.34 |
6 | 4 | 14 | 0.26 |
Cell-saver (%) | 90.2% (37/41) | 93.3% (154/165) | 0.72 |
Statistically significant (P < 0.05). |
Table 2
Preoperative coagulation function variables between Group A and Group B
| Group A | Group B | P |
| n = 41 | N = 165 | |
Platelet (X 109/L) | 221.8 ± 57.3 | 243.7 ± 71.9 | 0.40 |
Prothrombin time (s) | 12.6 ± 0.8 | 12.2 ± 0.5 | 0.91 |
Thrombin time (s) | 17.5 ± 0.9 | 17.9 ± 0.8 | 0.56 |
APTT (s) | 31.5 ± 3.2 | 32.7 ± 3.9 | 0.74 |
Fibrinogen (g/L) | 2.4 ± 0.4 | 2.7 ± 0.5 | 0.32 |
APTT, activated partial thromboplastin time. Statistically significant (P < 0.05). |
The surgery–related variables between the two groups showed no significant differences in IBL, TBL, NBL, TBV, operation time, mean intraoperative arterial pressure, intraoperative volume of crystalloid, cell–saver transfusion volume, correction rate, transfusion rate, and postoperative hospital stay (Table 3). The IBL of group A (715 ± 348 ml) was higher than that of group B (649 ± 311 ml); however, the difference was not statistically significant. The Kruskal–Wallis test showed no difference in the number of fusion levels (P = 0.19) and internal fixations (P = 0.26). The mean operative time was 202 ± 60 min in group A and 209 ± 56 min in group B ( P = 0.17). Six patients in group A (14.6%) and 12 in group B (7.3%) required allogeneic blood transfusions during or after surgery; however, the difference was not statistically significant. None of the patients in either group sustained any major perioperative complications. However, five patients experienced nausea after progesterone injection, which could be relieved after rest. No other adverse effects associated with progesterone use, such as headache, or diarrhea, were noted. None of the patients suffered from deep venous thrombosis or liver complications. One patient in group B developed delayed wound infection after surgery, and internal fixation was later removed. Another patient in group B developed superior mesenteric artery syndrome after surgery, which was later treated using diet adjustment. No similar complications were observed in Group A.
Table 3
Surgery–Related Variables between Group A and Group B
| Group A | Group B | P |
| n = 41 | N = 165 | |
Intraoperative blood loss (IBL), mL | 715.6 ± 348.3 | 649.5 ± 311.5 | 0.25 |
Normalized blood loss (NBL) {TBL(ml)/Weight(kg) * No of fusion level} | 1.5 ± 0.9 | 1.4 ± 1.2 | 0.12 |
Total blood loss (TBL), ml | 958 ± 329 | 894 ± 384 | 0.76 |
Total blood volume (TBV), ml | 3107 ± 853 | 3159 ± 1004 | 0.40 |
Operation Time (min) | 202.9 ± 60.2 | 209.6 ± 56.1 | 0.17 |
Mean ABP (mm Hg) | 71.5 ± 5.2 | 72.4 ± 6.6 | 0.12 |
Crystalloid (ml) | 1420 ± 615 | 1490 ± 907 | 0.89 |
Cell-saver transfusion (ml) | 225 ± 116 | 210 ± 152 | 0.27 |
Number of pedicle screw | 15.0 ± 2.8 | 14.8 ± 3.0 | 0.52 |
Number of fusion level | 9.7 ± 2.4 | 9.6 ± 2.6 | 0.19 |
Postoperative Cobb angle (°) | 18.4 ± 7.8 | 17.8 ± 9.8 | 0.26 |
Correction rate (%) | 65.9 ± 12.6 | 66.3 ± 10.9 | 0.89 |
Postoperative hospital stay, day | 6.0 ± 1.2 | 5.8 ± 1.2 | 0.35 |
Preoperative hemoglobin, g/L | 13.5 ± 1.9 | 13.3 ± 1.1 | 0.37 |
Postoperative hemoglobin, g/L | 9.6 ± 2.0 | 10.1 ± 1.3 | 0.68 |
Transfusion rate (%) | 14.6 | 7.3 | 0.46 |
No, number; ABP, arterial blood pressure. Statistically significant (P < 0.05). |
Discussion
Surgery during menstruation is a concern for surgeons because the perioperative blood loss may produce undesirable effects. This may be due to the malfunctioning coagulation system during menstruation. Findikcioglu et al.(14) reported that operative blood loss is higher during the periovulatory phase of rhinoplasty surgery. As reported by Sariguney et al.(13), perioperative blood loss in breast reduction surgery is greater during the perimenstrual phase than during the periovulatory period. However, some studies have shown contradictory results. A study by Lin et al.(21) found that menstruation had no significant effect on operative bleeding in vitreoretinal surgery. It was found by Das et al. that women undergoing open heart surgery during menstruation did not experience an increase in both surgical and menstrual blood loss (22).
There are few studies on menstruation and perioperative blood loss in patients with scoliosis in spinal surgery. A study by Li et al.(23) analyzed 161 cases of AIS by grouping them into four categories: premenstrual (24–30 days), follicular (6–11 days), ovulatory (12–17 days), and luteal (18–23 days). They found that female AIS patients who underwent PSF surgery during the premenstrual phase of the menstrual cycle suffered more IBL. In the study, menstruation was considered a contraindication for PSF surgery; they avoided undergoing PSF surgery during menstruation due to concern about increased perioperative blood loss. However, Chiu et al.(24) compared AIS patients who underwent PSF during the menstrual period with those who underwent PSF during the non–menstrual period and found no significant difference in IBL. In our institution, PSF surgery is usually not performed during menstruation for the following reasons. First, PSF surgery is characterized by a large IBL, during menstruation, however, patients lose a significant amount of blood, which increases the odds of blood transfusion. Second, many AIS patients are not local residents, and when the operation date conflicts with the menstrual date, if they wait till the end of their menstruation to undergo PSF surgery, it usually has financial and emotional implications for patients and their guardians. Third, pain and fatigue associated with menstruation will delay the diagnosis of complications and affect postoperative rehabilitation exercise. As progesterone withdrawal triggers menstruation, we used progesterone to postpone menstruation and staggered the operation and menstruation. Menstruation usually occurs on the third or fourth day after the last progesterone injection, and rehabilitation exercise is usually restored in patients with AIS who underwent PSF surgery.
While progesterone can effectively postpone menstruation, some studies suggested it may increase thrombotic risk. Previous studies have shown that oral contraceptives (including estrogen and progesterone) can increase plasma concentrations of fibrinogen, factor VIIa, factor VIII, free protein S, and APTT–based activated protein C sensitivity ratio, which might increase the risk of thrombotic situations(15, 25). A higher-quality study also suggested that estrogen plus progestogen users have a higher risk of thrombosis than estrogen users alone(26). As a hormone replacement therapy, Canonico et al. found that progesterone alone increased the risk of venous thrombosis in postmenopausal women (27). However, the role of progesterone in venous thromboembolism is controversial. In premenopausal women, Vasilakis et al. found no increase in thrombosis risk when progesterone was used alone for contraception(28). Through network meta–analysis, Stegeman et al. found that the difference in the incidence of venous thrombosis outcomes between women using oral contraceptives and those not using them was only 1.4 per 10000 (16). The discrepancy between progesterone and thrombosis results may be due to the different levels of physiological sex hormones in the study population. In addition, the form of administration, dosage, frequency, and duration of progesterone treatment differed. In our study, all patients who received progesterone treatment had their menstruation successfully postponed. However, we did not find that progesterone injection affected IBL, TBL, and NBL or increased the risk of thrombus–related events. In addition, there was no difference in complications such as infection, prolonged hospital stay, and delayed wound healing between the two groups. This may be due to the short–term (4 days) and relatively small doses (20 mg/day) of progesterone compared to the long–term and large–dose use of progesterone.
Patients who underwent Schwab grade III–VI osteotomy and those who received TXA treatment were excluded from this study. This is because high–grade osteotomy has been proven to be a risk factor for blood loss in PSF surgery(29). TXA has also been proven to reduce blood loss and blood transfusion rates in PSF surgery for AIS patients(30). However, only 96 patients used TXA, and the use of TXA was not consistent in the early stages; therefore, we excluded these patients. Demographic and operative factors were matched between the two groups. All patients underwent surgery performed by the same senior spinal surgeon in our hospital. Perioperative treatment was performed following the same protocol. Confounding factors were minimized. Eventually, the IBL was slightly higher in group A than in group B (715 ± 348 ml vs. 649 ± 311 ml); however, the difference was not statistically significant. Furthermore, there was no significant difference in TBL between groups A and B. To avoid the influence of weight and the number of fusion levels, NBL was calculated; however, there was still no significant difference. These results suggest that progesterone did not affect perioperative blood loss. Allogeneic blood transfusion rates were also higher in group A (14.6%) than in group B (7.3%); however, this difference was not statistically significant. No complications occurred in any patients in group A. Therefore, we postulate that using progesterone to postpone menstruation does not affect perioperative blood loss or complications in AIS patients who underwent PSF surgery.
This study has several limitations. Firstly, it was limited by the inherent limitations of retrospective studies. The use of progesterone during PSF surgery needs to be evaluated in a prospective randomized multicenter study. Second, no hormonal studies were performed on these patients. The menstrual cycle is regulated primarily by four hormones: estrogen, luteinizing hormone, progesterone, and follicle-stimulating hormone. Further studies are needed to explore which hormone is more suitable for postponing menstruation in PSF surgeries and the appropriate dose and mode of administration.