Salvage Esophagectomy after Definitive Chemoradiotherapy: Safety, Efficacy, and Quality in a Large Population-Based Cohort Study
Jessica L Hudson, Melanie Subramanian, Tara R Semenkovich, G. Alexander Patterson, Benjamin D Kozower, Bryan F Meyers, Varun Puri
Washington University in St Louis, St Louis, MO
OBJECTIVES: Surgical resection after definitive chemoradiotherapy for locally advanced esophageal cancer remains controversial, balancing the risks of surgery with the hopes of improving oncologic or functional outcomes. Using a large retrospective cohort, we explored the oncologic quality as well as safety and survival outcomes of salvage esophagectomy.
METHODS: Patients who underwent esophagectomy after chemoradiotherapy from 2004-2014 were abstracted from the National Cancer Database. Patients who had planned esophagectomy (resection <2 months after completion of radiation, N=9841) were compared to those with salvage esophagectomy (either resection >6 months after radiation completion or total radiation dose >6000 cGy, N=230). There were no statistically significant differences in baseline characteristics or quality, safety, and efficacy outcomes between these two salvage esophagectomy definitions and therefore were combined for statistical analysis.
RESULTS: Salvage esophagectomy was identified in 230 patients, defined by either timing of surgery (>6 months after radiation completion, N=101) or by radiation dosage (treatment volume >6000cGy, N=132). Planned esophagectomy was performed in 9841 patients. As expected, for planned esophagotomy, surgery occurred a mean of 51 ±15.7 days after radiation completion while salvage esophagectomy was performed a mean of 171 ± 127.9 days, p<0.001. Planned esophagectomies received less radiation than salvage esophagectomies (4770±333 cGy) vs 6679±1050 cGy (p<0.001). No statistically significant differences were noted in comorbidities or clinical stage. Compared to salvage, the planned esophagectomy patients had higher socioeconomic factors and sought care in metropolitan centers (P<0.05).
No difference was seen in readmissions (p=0.70) or 90-day mortality (p=0.11) between the planned and salvage esophagectomy groups. Median survival after salvage esophagectomy was 35.6 months (interquartile range 20-79), compared to 37.6 months (interquartile range 17-117) for planned esophagectomy (p=0.292). With salvage esophagectomy, patients were at increased risk of having fewer lymph nodes sampled (p=0.002), having macroscopic tumor (R2) visible at the margins (OR 6.16, p<0.01), and increased 30-day mortality (OR 2.23, p<0.01) when compared to those undergoing a planned resection. Surprisingly, 37.8% of salvage resection patients had complete pathologic response in their esophagectomy specimen compared to 31% of planned esophagectomies (OR 1.38, p= 0.02).
CONCLUSIONS: This is the largest US retrospective cohort to date evaluating the efficacy and safety of salvage esophagectomy. Our findings suggest that when faced with the decision between palliation and unplanned resection, it is reasonable to consider salvage esophagectomy from the perspective of long-term outcomes. However, in this real-word study, quality metrics including lymph node sampling, complete pathologic response, and R2 resection rates for salvage esophagectomy were worrisome. Salvage esophagectomy is likely an underutilized strategy in appropriately selected patients but discerning judgment and diligent perioperative care are imperative.
|Univariate p value||Odds Ratio||Lower limit, 95%CI||Upper limit 95%CI|
|Number of lymph nodes sampled, mean (SD)||14.0 (13.4)||14.1 (13.4)||11.2 (10.7)||0.002|
|Lymph nodes positive||3481 (38.3)||3425 (38.5)||56 (28.4)||0.004||0.63||0.46||0.87|
|Surgical margins positive||503 (5.1)||485 (5.1)||17 (7.7)||0.086||1.55||0.94||2.56|
|Macroscopic residual tumor (R2)||16 (0.2)||14 (0.1)||2 (0.9)||0.006||6.16||1.39||27.25|
|Complete pathologic response||3101 (30.8)||3014 (30.6)||87 (37.8)||0.019||1.38||1.05||1.81|
|30-day unplanned readmission||505 (5.0)||492 (5.2)||13 (5.8)||0.700||1.12||0.63||1.97|
|30-day mortality||280 (3.3)||266 (3.2)||14 (6.8)||0.004||2.23||1.28||3.89|
|90-day mortality||669 (7.8)||647 (7.8)||22 (10.8)||0.113||1.44||0.92||2.25|
|Abbreviations: CI, confidence interval; N, number; SD, standard deviation.|
Back to 2018 Program