Cox proportional hazards regression analysis
Of the various clinicopathological prognostic factors under study, univariate Cox regression analysis illustrated that FIGO stages IB3, IIA2 and IIIC, LVSI (+), lymph node metastasis, parametrial invasion, surgical margins positive, depth of stromal invasion, tumour size of>4 cm and chemoradiotherapy (P ˂0.001) were associated with a lower 5-year DFS and OS (Table 2).
There was no evidence for a relationship with patient age, histologic type, clinical symptoms or surgical delay time.At a median (IQR) follow-up of 58 (47-75) months, 70 patients (8.75%) had recurrence, and 65 (8.13%) died during the study period.
The association between delayed surgery and survival outcome was confirmed by Cox analysis. Remarkably, a longer waiting period (longer wait time group) was not associated with poorer outcomes in terms of either OS or DFS. In the whole cohort, a longer wait time was not statistically associated with DFS (5-year rates of 91.5% versus 90.9%, HR 0.99, 95% CI 0.62~1.59, P =0.98; Fig. 2A) or OS (5-year rates of 92.9% versus 90.8%, HR 0.68, 95% CI 0.42~1.10, P =0.11; Fig. 2B) when compared to a shorter wait time. In multivariable Cox regression analyses, with adjustment for sex, FIGO stage and histology, the surgical delay time was also not associated with DFS (HR 0.90, 95% CI 0.56~1.45, P =0.662; Table 3, Model I) or OS (HR 0.62, 95% CI 0.38~1.02, P =0.060; Table 3, Model I). This association remained unchanged after adjusting for age, FIGO stage, histology, nodal status, clinicopathological characteristics, postoperative therapy, and clinical symptoms (Table 3, model III).
In addition, when wait time was considered a continuous variable, there was no association with DFS (HR per additional wait week 0.997, 95% CI 0.98~1.01, P = 0.655) or OS (HR per additional wait week 0.984, 95% CI 0.97~1.00, P=0.079) (Table 3).
To test the robustness of our findings, we performed several sensitivity analyses.
In a sensitivity analysis, cases were limited to LVSI (+). There were 30 (14.3%, 30/210) recurrences and 29 (13.8%, 29/210) deaths during the follow-up (median, 60 months). We found similar associations between surgical wait time and 5-year DFS (82.8%, HR 1.62, 95% CI 0.76~3.46, P=0.213) and OS (85.3%, HR 0.87, 95% CI 0.42~1.8, P=0.706) (Fig. 3).
Sensitivity analysis
In a sensitivity analysis, cases were limited to tumours ≤4 cm. There were both 20 (3.86%, 20/518) recurrences and deaths during the follow-up (median, 60 months). We found similar associations between surgical wait time and DFS (96.5%, HR 1.13, 95% CI 0.45~2.82, P=0.798) and OS (95.8%, HR 0.81, 95% CI 0.33~2.01, P=0.652) (Fig. S2).
However, in patients with a tumour size>4 cm, the survival outcome was worse in those with a wait time < 12 weeks, although there was no significant difference in DFS and OS between the two groups (Fig. S3). We considered that this might be due to rapid tumour growth, irregular vaginal bleeding or contact bleeding and other typical clinical symptoms, resulting in earlier medical treatment and shorter waiting times for patients.
4. Discussion
4.1 Main findings and interpretation
Since December 2019, the outbreak of coronavirus disease 2019 (COVID-19) has seriously affected the lives and health of people around the world. Millions of people even die, greatly affecting the timely diagnosis and treatment of other diseases, especially cancer2,3, 13-15.
During the COVID-19 epidemic in China, many patients with early clinical symptoms of cervical cancer were managed and centralized quarantine procedures for a long time, resulting in delayed diagnosis and treatment. Thus, it is likely that there are a number of patients in whom surgery for early cervical cancer is being postponed. Consequently, patients and their families are subjected to tremendous psychological burdens, such as varying degrees of anxiety, worry and panic. For the current COVID-19 pandemic, data examining the outcome of surgical delay time for cervical cancer patients are of utmost importance.
We defined the surgical delay time as the time from the onset of typical clinical symptoms to surgery, which differs from published studies. In China, a large proportion of cervical cancer patients are likely to receive appropriate treatment quickly once they are diagnosed. However, cervical cancer screening is not timely and sufficient in northwestern China. There may be a prolonged period from symptom onset to the first medical visit and surgical treatment. Focus of research should determine the longest acceptable surgical delay time before the patient outcomes are adversely affected rather than examining whether surgical delay affects patient outcomes at all.
To our knowledge, this study is the first to specifically examine whether delayed surgery may be a risk factor for poor postoperative outcomes in patients with early cervical cancer in the Chinese population. The main finding of the current study is that approximately half of the patients waited more than three months for surgery, and a 12-week wait time for radical hysterectomy was not associated with reduced OS and DFS in the long term.
In our study, Kaplan–Meier survival analysis showed that a longer wait time was not associated with 5-year DFS or OS, which was similar to previous findings. A retrospective study from the National Cancer Database examining 2732 cases of only stage IA cervical cancer showed that the 4.5-year OS rate was 95.0% and that a longer hysterectomy wait time was not associated with all-cause mortality risk7. However, another study results indicated that a waiting time longer than 8 weeks was associated with poorer after 5-year survival (HR 3.4, 95% CI 1.3~9.2,P =0.021) in FIGO stages IA2 or IB1 cervical cancer patients, although OS showed no significant difference between short (≤8 weeks) and long (>8 weeks) wait times12.
These contrasting results are probably because all cervical cancer patients in our study and the Kulisara et al. study7-11 had early-stage disease in which prognosis is generally good after radical hysterectomy.
A logical theory to explain our findings is that cervical cancer tumours develop slowly in the early stages, and therefore delaying surgical therapy for a short period of time may not have an influence on oncologic prognosis10, 17. This is based on the finding that disease progression is uncommon among women whose pregnancy has been disrupted by early-stage cervical cancer and that afflicted patients frequently postpone delivery to allow for foetal development throughout pregnancy without jeopardizing mortality5.
4.2 Strengths and limitations
Due to time limitations, the DFS and OS 5 years after surgery are still unclear and require long-term follow-up. Our study also found that LVSI, lymph node metastasis, tumour size and depth of invasion were independent prognostic indicators of 5-year DFS and OS. Our findings have many important implications for future practice. It is suggested that patients with early cervical cancer should have a medical visit as soon as possible after the onset of clinical symptoms and receive radical hysterectomy within 12 weeks.
During the current COVID-19 epidemic, taking “containment” measures is merely a temporary stopgap. A short delay time did not increase the poor outcome of patients with early cervical cancer.
5. Conclusion
In conclusion, our study suggests that a delay time of 12 weeks from the onset of typical clinical symptoms to radical hysterectomy may not increase the risk of long-term disease recurrence and poor survival in patients with early-stage cervical cancer. This information partly supports the recent suggestion to consider postponement of surgery for 6~8 weeks for the management of early cervical cancer during the COVID-19 pandemic. During the COVID-19 pandemic, delayed surgery has not resulted in serious outcomes, and thus, patients should not be overly anxious, irritable or worried.