A triage algorithm may identify patients with presumed advanced ovarian cancer who represent preoperatively defined candidates for diagnostic laparoscopy.
A triage algorithm may identify patients with presumed advanced ovarian cancer who represent preoperatively defined candidates for diagnostic laparoscopy. The year-long prospective study, conducted at the University of Texas MD Anderson Cancer Center, used laparoscopic assessment to predict no gross postoperative residual disease (R0) and yielded an 86% accuracy rate.
The Anderson algorithm was presented March 29th in Chicago at the 2015 Society of Gynecologic Oncology’s Annual Meeting on Women’s Cancer. “Our hypothesis was that diagnostic laparoscopy is an accurate means of determining the ability to achieve complete gross resection at the time of ovarian cancer debulking,” said Alpa M. Nick, MD, assistant professor of gynecologic oncology and the primary author of the study. “We believe that personalized surgical therapy represents an opportunity for quality improvement.”
The Anderson algorithm calls for:
During the study, whenever a patient presented with presumed advanced ovarian cancer, two surgeons independently used the Fagotti score to describe the disease distribution. Predictive Index Value (PIV) scores greater than or equal to 8 resulted in the patient being referred for neoadjuvant chemotherapy (NACT). Patient outcomes and faculty compliance were tracked prospectively and R0 rates were compared with historical practice. Agreement between the first and second surgeon occurred 97% of the time.
Of 99 patients with suspected advanced ovarian cancer, 11 were not offered laparoscopy due to medical comorbidities. The primary surgeon did not offer laparoscopy to 2 patients (98% compliance). One patient was categorized as “surgery required,” while 20 were surgically unresectable, leaving 65 to receive diagnostic laparoscopy (PIV <8, n = 40; PIV ≥8, n = 25).
The patients’ median age was 66 (range, 39-90 years), the median operative time was 36 minutes (range, 11-102 minutes), and the estimated blood loss was 5 mL (range, 0-50 mL). Nearly all of the tumors were ovarian in origin, and nearly all of the tumors were serous in type.
Median CA-125 values were greater among patients triaged to NACT compared with those who received laparoscopy (893 vs 161). Pretreatment median platelet counts were also greater among the patients who received NACT (379 vs 289.5). Among patients who underwent primary tumor reductive surgery, the postlaparotomy score sheet correlated with the primary surgeon’s laparoscopic score 96% of the time (R = 0.66,P<.01).
The study team concluded that laparoscopy is a reproducible and reliable method of determining ovarian cancer disease distribution. “Adoption of the Anderson algorithm has resulted in an improved rate of R0 cytoreduction at our institution,” Nick said. The same institution showed a historic R0 resection rate of 40% from 2007 through 2012 (P = <.001). “This is promising because improving the rates of R0 cytoreduction should improve patient survival.”
According to Nick, interest in adopting the Anderson algorithm has been high. The MD Anderson team has reportedly been working with academic cancer centers and private providers to expand use of the algorithm.
Nick AM, Coleman RL, Ramirez PT, et al. The Anderson algorithm: personalized surgical therapy for advanced ovarian cancer. Presented at: 2015 Society of Gynecologic Oncology’s Annual Meeting on Women’s Cancer; March 28-31, 2015; Chicago, IL. Abstract No 20.