One of the main challenges in management of CSP is massive bleeding. Uterine artery embolization, which has been widely used in controlling hemorrhage in gynecology and obstetrics, is regarded as a good option of treating CSP with minimal invasion, especially when used as an adjuvant therapy with other surgical treatment such as curettage. There have been dozens of studies evaluating the efficacy and safety of UAE used in CSP treatment, either as a single therapy or combined with other methods [5–10, 14–19].
However, there is currently no study pointing out the optimal timing to perform surgical treatment after prophylactic UAE. The previously reported study designs were quite different in the treatment interval between UAE and curettage.
During the UAE procedure, MTX was locally injected through the uterine arteries followed by the main stems of uterine arteries blocked by gelfoam particles [20]. MTX works on killing the embryo and trophoblasitc cells. The reduce of uterine blood supply by UAE was impermanently, since the gelfoam could be resolved within 7–14 days. Depending on this mechanism, most authors theoretically suggested that curettage should be taken within 24–72 hours after UAE to balance the benefit of onset time and the risk of recanalization [5–11]. Currently, it is only a clinical opinion which still needs reliable clinical evidence. In our hospital, gynecologists performed curettage even within 24 hours or longer than 72 hours after UAE in some cases. The treatment intervals were individualized depending on operation schedules, gynecologists’ intentions, or patients’ preferences. Thus, we designed the present study to discuss the most appropriate time to perform curettage after UAE.
In this cohort study, we found that patients in the short interval arm were likely to have better clinical outcomes than those in the long interval arms. This result demonstrated that the delay of curettage following UAE might increase the risk of intra-operative bleeding and other complications. Consistently, the result of multi-variable logistic regression indicated that the treatment interval longer than 72 hours after UAE was one of the risk factors of intra-operative bleeding. This increase of bleeding risk by time delay might be explained by the initiating of collateral circulation or tissue edema due to long-time ischemia. The incidence of bleeding was lower in shorter-interval arms, which appeared that treatment interval within 24 hours might be ideal. However, when we adjusted the cut-off of this variable at 24 or 48 hours in the multi-variable analysis, the statistical results showed no significance. Therefore, we could not get a final conclusion about the best timing of curettage after UAE based on this study, but we strongly suggest that curettage should not be delayed longer than 72 hours under general conditions.
Other risk factors identified as risk factors of bleeding in this combination therapy included thinner myometrial thickness, larger diameter of CSP mass, and type II CSP. Since this finding coincides with our previous study [14] and other reports [21, 22], we did not discuss more about these factors. However, the gestational age and peritrophoblastic blood supply which were identified as potential risk factors in the univariate analysis were not included in the result of our final regression model. This result might be explained by: (1) the synergy of gestational age and diameter of gestational sac (adjusted as a confounding factor); (2) the blood supply shown at the time of diagnosis had been successfully blocked by UAE, no longer influencing the clinical outcome in this combination treatment strategy.
There are limitations of our study. Firstly, this is a retrospective and observational research. The interventions were decided depending on the intentions of clinicians and patients with selective bias. Secondly, the bleeding risk factors in treatment of CSP are multi-factorial and interative. Our study showed that the influence of treatment interval had significantly clinical and statistical meaning, but it might not be an independent factor. Lastly, even though our sample size is large enough as a study of one single medical center, multi-center randomized controlled trials (RCTs) with more reliable evidence are needed to get a better conclusion.