|Year : 2017 | Volume
| Issue : 3 | Page : 550-555
Role of laparoscopy in predicting surgical outcomes in patients undergoing interval cytoreduction surgery for advanced ovarian carcinoma: A prospective validation study
Than Singh Tomar1, Rema Prabhakaran Nair1, Suchetha Sambasivan1, K M Jagathnath Krishna2, Aleyamma Mathew2, Iqbal Ahmed M1
1 Division of Surgical Oncology, Regional Cancer Centre, Trivandrum, Kerala, India
2 Division of Cancer Epidemiology and Biostatistics, Regional Cancer Centre, Trivandrum, Kerala, India
|Date of Web Publication||24-May-2018|
Dr. Rema Prabhakaran Nair
Division of Surgical Oncology, Regional Cancer Centre, Trivandrum, Kerala
Source of Support: None, Conflict of Interest: None
OBJECTIVE: The objective of this study was to evaluate the safety and efficacy of laparoscopy in detecting inoperable disease in patients undergoing interval cytoreduction (ICR) for advanced ovarian carcinoma (AOC). The primary outcome measured was the performance of laparoscopy-based predictive index value (PIV) score developed by Fagotti et al. The secondary outcomes measured were performance of individual parameters included in PIV score and optimal cytoreduction (OCR) rates in our population. PATIENTS AND METHODS: This is a single-arm, prospective validation trial. Patients undergoing ICR for AOC in our institution were evaluated prospectively with laparoscopy before planned attempt at debulking surgery. Seven laparoscopic parameters included in laparoscopic PIV score were evaluated. Laparoscopic findings were compared with the final outcomes of definitive surgery. OCR was defined as residual disease <1 cm. The efficiency of the individual laparoscopy score was analyzed using receiver operating characteristic (ROC) curves. RESULTS: A total of 73 patients planned for ICR for AEOC were included in the study. Laparoscopic PIV score could successfully predict inoperability in 12 (16.4% of total study population) out of 14 inoperable patients in the total population and thus could avoid 85% of unsuccessful surgeries at a PIV score cutoff of ≥8. Performance of individual parameters included in PIV score was also evaluated. Two parameters out of seven, that is, mesenteric retraction and stomach infiltration had poor performance on ROC curve. Modified PIV score was calculated for each patient after excluding these two parameters. Modified PIV score had similar performance as Fagotti's PIV score at cutoff ≥6 (P = 0.728, for difference in area under the curve). No staging laparoscopy-related serious adverse events were noted in any of the patients. CONCLUSIONS: Laparoscopy is a safe, effective, and accurate method for predicting inoperability in patients undergoing ICR for AEOC.
Keywords: Carcinoma ovary, diagnostic laparoscopy, interval cytoreduction, ovarian neoplasm
|How to cite this article:|
Tomar TS, Nair RP, Sambasivan S, Krishna K M, Mathew A, Iqbal Ahmed M. Role of laparoscopy in predicting surgical outcomes in patients undergoing interval cytoreduction surgery for advanced ovarian carcinoma: A prospective validation study. Indian J Cancer 2017;54:550-5
|How to cite this URL:|
Tomar TS, Nair RP, Sambasivan S, Krishna K M, Mathew A, Iqbal Ahmed M. Role of laparoscopy in predicting surgical outcomes in patients undergoing interval cytoreduction surgery for advanced ovarian carcinoma: A prospective validation study. Indian J Cancer [serial online] 2017 [cited 2019 Dec 10];54:550-5. Available from: http://www.indianjcancer.com/text.asp?2017/54/3/550/233145
| » Introduction|| |
Epithelial ovarian cancer is fifth most common cancer in women and a leading cause of gynecological cancer mortality. It usually presents at an advanced stage because of nonspecificity of symptoms at an early stage and lack of effective screening methods. Successful cytoreductive (debulking) surgery is the cornerstone in the management of advanced ovarian carcinoma (AOC), and residual disease after cytoreduction surgery is the strongest predictor of survival. Rates of optimal cytoreduction (OCR) vary among institutions, surgeons, selection criteria for surgery, and with use of neoadjuvant chemotherapy (NACT). The percentage of patients with OCR in the EORTC-NCIC randomized trial reported by Vergote et al. was 41.6% and 80.6% after primary and interval cytoreduction (ICR) surgery, respectively. Considering importance of OCR on patient survival and varied rates of success achieved in multiple studies, it is always desirable to predict chances of successful surgery preoperatively. Many methods have been used to predict operability in epithelial ovarian cancer including the presence of ascites, CA-125, radiological criteria, and performance status but with partial success. Peritoneal carcinomatosis, diaphragmatic deposits and pelvic deposits are frequent sites of residual disease. These are usually not effectively visualized on contrast-enhanced computed tomography (CECT) scan and lead to higher false negative rate in predicting operability. In addition, post-NACT patients may have residual fibrosis and adhesions which can make accurate prediction of operability on CECT scan further difficult. Most of the above-mentioned sites can be effectively visualized and disease extent assessed directly by laparoscopy, making laparoscopy an attractive minimally invasive modality for predicting operability, and avoiding unnecessary laparotomy.
Many investigators have studied the role of laparoscopy for this purpose. In the pilot study done by Fagotti et al. in 2005, laparoscopy successfully predicted suboptimal debulking (residual tumor ≥1 cm) in 100% of cases and debulking surgery with no macroscopic tumor left in 89% of cases in patients undergoing primary debulking surgery. Use of laparoscopy to predict operability without standardized criteria is associated with subjective variation among different operating surgeons. Fagotti et al. in 2006 developed a prediction model based on retrospective analysis of same group of patients mentioned earlier. As NACT followed by ICR is being used more commonly in patients who present with very advanced disease and poor performance status, there is a need to develop accurate methods to predict operability in this group of patients also. Accurate prediction of inoperability cannot be overemphasized in post-NACT patients, as nontherapeutic laparotomy will add morbidity, increased cost, poor quality of life, and delay in starting second-line chemotherapy.
We undertook this study to evaluate the performance of laparoscopic predictive index value (PIV) score and individual parameters proposed by Fagotti et al. in predicting operability in ICR settings.
| » Patients and Methods|| |
This is a single-arm, prospective validation study conducted from May 2014 to April 2015. We included all consecutive patients undergoing ICR surgery for AEOC at our center. Patients who were considered operable and planned for ICR based on post-NACT CECT scan abdomen and/or responding CA 125 values were included in this study. All patients were under 75 years of age and had biopsy or cytologically proven epithelial ovarian cancer. All of these patients had advanced stage at initial presentation based either on prechemotherapy CECT scan findings or previous surgical details of failed attempt at primary cytoreduction done before referral to our center. Preoperative evaluation was done as per our institute protocol which included complete physical and gynecological examination, postchemotherapy abdominal and pelvic CECT scan, chest X-ray, or CECT scan thorax (if the patient had pleural effusion before NACT) and serum CA 125 level. Patients who had poor Eastern Cooperative Oncology Group performance status (>2) and other conditions such as large mass filling whole abdomen and prior multiple surgeries which would prevent successful laparoscopy were excluded from the study. Informed consent was obtained from all patients.
All the patients underwent diagnostic laparoscopy before standard vertical midline laparotomy under same anesthesia. The same team of dedicated gynecology surgeons operated on all patients. Open Hassons's technique was used for initial port placement, preferably at the umbilicus. Additional ports were inserted depending on intraoperative findings. We evaluated seven laparoscopic parameters, namely, diffuse peritoneal carcinomatosis, confluent diaphragmatic deposits, any liver deposits, omental involvement up to greater curvature, stomach infiltration, bowel involvement requiring resection, and mesenteric retraction. The presence or absence of each parameter was noted during laparoscopy, and the final PIV score was calculated by adding point value of 2 for each parameter present. Time required for laparoscopic evaluation was also noted for each case. Under same anesthesia, patients underwent laparotomy and ICR were attempted. Standard surgery for ICR included extrafascial hysterectomy, bilateral salpingo-oophorectomy, total omentectomy, appendectomy in mucinous histology, peritoneal and diaphragmatic deposit excision if any, and bowel/rectal resection if needed. The final results of debulking surgery were noted for each case at the end of the procedure.
Approval and ethics
This study was approved by the Institutional Review Board and Ethical Committee. This study was registered with clinical trial registry of India before starting accrual. This study was funded by our institute.
Estimated sample size for this study was 68. Five additional patients were also included to compensate for technical failure or inability to do satisfactory laparoscopic evaluation in case of severe adhesions and extensive deposits.
Sensitivity, specificity, positive predictive value (PPV), negative predictive value (NPV), and overall accuracy were calculated for each possible PIV score cutoff. Sensitivity was defined as the total number of inoperable patients correctly identified (true positive) out of total inoperable patients (true positive + false negative). Specificity was calculated as number of operable patients correctly identified (true negative) divided by total operable patients (true negative + false positive). PPV was defined as true positive divided by total positive. NPV was calculated as true negative divided by total negative. In addition “rate of avoidance of unnecessary laparotomy” was calculated for each score by dividing true positive by total study population to know ability of selected cutoff to avoid unnecessary laparotomy in total population. Rate of “inappropriate unexploration” was calculated for each PIV cutoff by dividing false positive by the total population to know rate of “falsely labeling patient inoperable.” False negative rates were calculated by dividing false negative with total inoperable patients to know rate of falsely labeling patient operable out of the total inoperable patients. Receiver operating characteristics (ROCs) curve analysis was done and area under curve (AUC) was calculated by plotting “sensitivity” versus “1-specificity” to check the performance of laparoscopy-based PIV score to predict inoperability. ROC curve was plotted under nonparametric assumption and null hypothesis being true area = 0.5.
Sensitivity, specificity, PPV, NPV, and accuracy were calculated for each individual parameter also, to know its independent performance. Necessary criteria defined for each parameter to remain validated in the final score with point value of 1 was specificity, PPV, and NPV of more than 75%, 50%, and 50%, respectively. In addition, one additional point value was given to parameters which had accuracy of more than 60%. ROC curve analysis was done for each parameter to know its ability to predict inoperability. Ninety-five percentage confidence interval (CI) was calculated for AUC.
| » Results|| |
A total of 73 post-NACT patients underwent laparoscopy before planned definitive surgery. The most commonly used chemotherapy was intravenous carboplatin and paclitaxel. Average numbers of NACT cycles were 4.7 ranging from 2 to 6. Other patient characteristics are included in [Table 1].
Optimal cytoreduction rates and predictive index value score distribution
Out of total 73 patients, 59 (80.83%) achieved OCR and 14 (19.17%) had residual disease of ≥1 cm left behind. All 59 patients who achieved OCR had undergone sequential laparoscopy followed by laparotomy. In 14 patients who did not achieve OCR, 6 had both laparoscopy followed by laparotomy whereas in 8 patients, laparotomy was not attempted in view of laparoscopic finding of extensive miliary deposits on peritoneal surface as subsequent laparotomy was unlikely to change the outcome. [Table 2] and [Figure 1] shows PIV score distribution and final outcomes.
|Table 2: Predictive index value score distribution and final results of surgery|
Click here to view
|Figure 1: Predictive index value score distribution and results of interval cytoreduction|
Click here to view
Performance of Fagotti's predictive index value score
Sensitivity, specificity, PPV, NPV, accuracy, rate of avoidance of unnecessary laparotomy, rate of inappropriate unexploration, and false negative rate for each PIV cutoff are tabulated [Table 3]. None of the included patients had PIV score >10. As evident from [Table 3], score ≥8 appears to be the most optimum cutoff as this score had highest accuracy of 97% with 100% PPV. PIV cutoff of ≥8 had the potential to decrease unnecessary laparotomy rate by 16.40% in total population and can correctly identify 85.71% of total inoperable patients. Rate of falsely labeling inoperable is 0 as this score has 100% PPV. We had only two cases with suboptimal debulking even when they had PIV score <8. The first patient had significant lymphadenopathy infiltrating vessels and another patient had extensive pelvic disease which was infiltrating rectum and pelvic side walls. ROC curve analysis was showing high discriminating index for PIV score with AUC of 0.976, (95% CI 0.932–1.0) [Figure 2].
|Table 3: Performance parameters for predictive index value score at different cutoffs|
Click here to view
|Figure 2: Receiver operating characteristic curve analysis of predictive index value model|
Click here to view
Performance of individual parameters included in predictive index value score
Sensitivity, specificity, PPV, NPV, and accuracy were calculated for each parameter to assess individual performance [Table 4]. Each individual parameter fulfilled predefined criteria of specificity >75%, PPV >50%, and NPV >50% to be included in the final score. In addition, every individual parameter also had accuracy of >60%; hence all parameters fulfilled predefined criteria by Fagotti et al. to be included in the final PIV score with individual point value of 2.
Modified predictive index value score
AUC was calculated for each individual parameter also by ROC analysis to know its individual discriminative power for the prediction of inoperability [Table 4]. Mesenteric retraction and stomach infiltration had low AUC with value of 0.5 included in 95% CI range signifying poor discriminative power. Low performance of these parameters was mainly because of few patients who were positive for these parameters. For the same reason, mesenteric retraction and stomach infiltration also had poor sensitivity of 33.3% and 14.2%, respectively. We calculated modified PIV score ranging from 0 to 10 for each patient using only five parameters after excluding mesenteric retraction and stomach infiltration. [Table 5] and [Table 6] show modified PIV score distribution, OCR rates, and performance parameters at different modified PIV score cutoffs, respectively.
|Table 5: Modified predictive index value score distribution and operability|
Click here to view
|Table 6: Modified predictive index value score and performance parameters|
Click here to view
As evident from [Table 6], modified PIV cutoff ≥6 has better performance with accuracy, PPV, and NPV of 98.63, 100, and 98.33, respectively. This cutoff has potential to decrease unnecessary laparotomy rate from 19.17% to 1.37% of the total population with no false positivity.
Comparison of Fagotti's predictive index value score and modified predictive index value score
Fagotti's PIV score was compared with our modified PIV score by the ROC analysis. Fagotti's score and modified PIV score had high AUC of 0.976 (95% CI 0.932–1.00) and 0.979 (95% CI of 0.937–1.00), respectively. [Figure 3] compares AUC for Fagotti's score (green line) versus modified PIV score (blue line). Both lines were overlapping each other signifying similar discrimination power for both scores. Mann–Whitney test was not showing significant difference in AUC with two-tailed P = 0.728. This signifies that modified PIV score is as accurate as original Fagotti's score with less parameters and hence less complex scoring.
|Figure 3: Receiver operating characteristic curve for Fagotti's predictive index value score (green line) and modified predictive index value score (blue line)|
Click here to view
Adverse events and technical feasibility
Successful laparoscopic scoring was possible in 100% of cases. All parameters were successfully evaluated in all patients except for mesenteric retraction which was not possible to evaluate for three patients because of dense adhesions and post-chemotherapy fibrosis [Table 4]. No laparoscopy-related serious adverse events were noted in any of the patient during the study period.
| » Discussion|| |
Role of laparoscopy for predicting operability in AEOC has been evaluated previously mostly in patients undergoing primary cytoreduction surgery. Fagotti's score was developed after retrospective analysis of 64 patients who underwent laparoscopy followed by an open attempt at cytoreduction surgery. Validation study done by same authors also confirmed same results in the primary settings. After publication of two randomized controlled trials on NACT versus primary debulking surgery showing noninferior results in patients undergoing interval debulking surgery, more and more patients with advanced disease are being treated with NACT followed by ICR. Predicting operability in this group of patients is important as rates of successful surgery after NACT varies and are about 80%. It means 20% of patients who have extensive disease or nonresponding disease are left behind with gross residual disease and morbidity of unnecessary laparotomy. Many methods including clinical response, resolution of ascites, CECT scan, and CA 125 response have been used to predict operability in this group, but false negative rates are high as reflected by rates of sub-OCR reported in various studies. Various studies published on use of CECT scan on predicting operability have selected different criteria for predicting operability and came up with differing results.,, Diffuse peritoneal carcinomatosis, extensive mesenteric involvement, multiple bowel loop involvement, confluent diaphragmatic disease, and porta hepatis deposits are common causes for inoperable disease. Laparoscopy in these settings provides a chance to directly look into the abdomen and decide on operability, but to make an accurate objective decision about operability without standard criteria is difficult. From surgical point of view, the presence of any of the individual parameters as such does not necessarily indicate inoperability and same is evident from performance of individual parameters shown in [Table 4]. Laparoscopic PIV score combines multiple parameters and provides a better global assessment of intra-abdominal disease extent, hence inoperability. PIV score has better discrimination power to predict inoperability than any of the single parameter studied is evident from the high AUC (0.976) on ROC analysis. Two studies have reported on this group of patients. The prospective validation study reported by Fagotti et al. in 2008 had only 22 post-NACT patients. Another study reported by Fagotti et al. in 2010 on the role of laparoscopy in ICR, included 98 patients and reported that only four parameters, namely, bowel infiltration, stomach infiltration, mesenteric retraction, and surface liver deposits qualified to be included in the final scoring system. PIV cutoff of >4 was considered most useful with PPV of 100%, rate of inappropriate unexploration 0%, and drop in rate of unnecessary laparotomy by 10%. As previously discussed in results, our results on PIV score differs from results reported by Fagotti et al. in 2010 in ICR patients. In our study population, PIV score had accuracy ranging from 67% to 97% depending on the cutoff used. PIV score ≥8 appears most appropriate as it has the highest accuracy and PPV of 97% and 100%, respectively. None of the patients, who had PIV score ≥8 achieved OCR in our population, hence, rate of inappropriate unexploration (false positive) is 0 at this cutoff. Score ≥8 identified 12 out of 14 inoperable patients in total population of 73, hence, this cutoff has potential to decrease unnecessary laparotomy rate from 19.17% to 2.77% of total patients (16.40% absolute reduction in total patients). If we select higher cutoff than 8, many inoperable patients will be labeled as operable and benefit of laparoscopy will decrease as sensitivity drops from 85% at cutoff 8 to 35.7% at cutoff 10. PIV score cutoffs <8 [Table 3] has low specificity and high chances of labeling patient as false positive (increased rate of inappropriate unexploration).
Looking into the performance of individual parameters included in PIV score, all parameters satisfied predefined criteria to remain in the final score with point value of 2 [Table 4]. None of the single parameters had accuracy better than combined PIV score of 8. On the ROC curve analysis, omental involvement up to greater curvature, confluent diaphragmatic carcinomatosis, and diffuse peritoneal carcinomatosis had best discrimination power with AUC >0.8. Mesenteric retraction and stomach infiltration had lowest performance with low AUC of 0.641 and 0.563, respectively.
When analysis was done after excluding mesenteric retraction and stomach infiltration [Table 6] best performance was obtained at modified PIV cutoff of ≥6 (i.e., >4). This cutoff is similar to the findings reported by Fagotti et al. in validation study done on ICR patients. All patients included in this study were judged operable on clinical, radiological, and serological assessment, yet we had 19.17% inoperability rate in the total study population. Laparoscopic PIV score and modified PIV score have potential to decrease laparotomy rate by 16.40% and 17.80% of total patients, respectively, which is clinically meaningful for this population.
Successful laparoscopic scoring was possible in all patients. Time required for laparoscopic scoring was noted for each case and ranged from 6 to 32 min with average being 15.5 min. Laparoscopic scoring is not a technically demanding procedure and there was no significant learning curve noted on previous studies. Laparoscopic scoring is a safe procedure as no laparoscopy-related intraoperative injury was noted during the entire study.
As evaluated by other investigators, this technique has a significant role to play in both primary as well as interval settings. The standard treatment approach for AEOC is primary debulking surgery followed by chemotherapy. Rates of optimal debulking vary in primary cytoreduction surgeries. Diagnostic laparoscopy done before primary cytoreduction surgery can identify patients who are not suitable for primary debulking and avoid unnecessary laparotomy. In these settings, diagnostic laparoscopy can also serve as a useful tool to decide on treatment selection, that is, primary debulking surgery and adjuvant chemotherapy versus NACT followed by interval debulking.
| » Conclusions|| |
Laparoscopic PIV score is safe and accurate method for predicting operability in ICR patients. This safe and easily available technique can be utilized to decrease the rate of unnecessary laparotomy and consequent decrease in the duration of hospitalization, improved quality of life, reduction in cost, and complications in inoperable patients. We suggest modified PIV score with five parameters may be used for ICR patients with cutoff of ≥6 (i.e., >4). This simplified score carries same accuracy for predicting inoperability as original PIV score.
Financial support and sponsorship
This study was financially supported by Regional Cancer Centre, Medical College Campus, Trivandrum - 695 011, Kerala, India.
Conflicts of interest
There are no conflicts of interest.
| » References|| |
Siegel R, Naishadham D, Jemal A. Cancer statistics, 2012. CA Cancer J Clin 2012;62:10-29.
Cannistra SA. Cancer of the ovary. N
Engl J Med 2004;351:2519-29.
Bristow RE, Tomacruz RS, Armstrong DK, Trimble EL, Montz FJ. Survival effect of maximal cytoreductive surgery for advanced ovarian carcinoma during the platinum era: A meta-analysis. J Clin Oncol 2002;20:1248-59.
Eisenkop SM, Spirtos NM. What are the current surgical objectives, strategies, and technical capabilities of gynecologic oncologists treating advanced epithelial ovarian cancer? Gynecol Oncol 2001;82:489-97.
Vergote I, Tropé CG, Amant F, Kristensen GB, Ehlen T, Johnson N, et al.
Neoadjuvant chemotherapy or primary surgery in stage IIIC or IV ovarian cancer. N
Engl J Med 2010;363:943-53.
Decloedt J, Vergote I. Laparoscopy in gynaecologic oncology: A review. Crit Rev Oncol Hematol 1999;31:15-26.
Fagotti A, Fanfani F, Ludovisi M, Lo Voi R, Bifulco G, Testa AC, et al.
Role of laparoscopy to assess the chance of optimal cytoreductive surgery in advanced ovarian cancer: A pilot study. Gynecol Oncol 2005;96:729-35.
Fagotti A, Ferrandina G, Fanfani F, Ercoli A, Lorusso D, Rossi M, et al.
Alaparoscopy-based score to predict surgical outcome in patients with advanced ovarian carcinoma: A pilot study. Ann Surg Oncol 2006;13:1156-61.
Fagotti A, Ferrandina G, Fanfani F, Garganese G, Vizzielli G, Carone V, et al.
Prospective validation of a laparoscopic predictive model for optimal cytoreduction in advanced ovarian carcinoma. Am J Obstet Gynecol 2008;199:642.e1-6.
Bristow RE, Duska LR, Lambrou NC, Fishman EK, O'Neill MJ, Trimble EL, et al.
Amodel for predicting surgical outcome in patients with advanced ovarian carcinoma using computed tomography. Cancer 2000;89:1532-40.
Byrom J, Widjaja E, Redman CW, Jones PW, Tebby S. Can pre-operative computed tomography predict resectability of ovarian carcinoma at primary laparotomy? BJOG 2002;109:369-75.
Dowdy SC, Mullany SA, Brandt KR, Huppert BJ, Cliby WA. The utility of computed tomography scans in predicting suboptimal cytoreductive surgery in women with advanced ovarian carcinoma. Cancer 2004;101:346-52.
Fanfani F, Ferrandina G, Corrado G, Fagotti A, Zakut HV, Mancuso S, et al.
Impact of interval debulking surgery on clinical outcome in primary unresectable FIGO stage IIIc ovarian cancer patients. Oncology 2003;65:316-22.
Fagotti A, Fanfani F, Vizzielli G, Gallotta V, Ercoli A, Paglia A, et al.
Should laparoscopy be included in the work-up of advanced ovarian cancer patients attempting interval debulking surgery? Gynecol Oncol 2010;116:72-7.
Fagotti A, Vizzielli G, Costantini B, Lecca A, Gallotta V, Gagliardi ML, et al.
Learning curve and pitfalls of a laparoscopic score to describe peritoneal carcinosis in advanced ovarian cancer. Acta Obstet Gynecol Scand 2011;90:1126-31.
[Figure 1], [Figure 2], [Figure 3]
[Table 1], [Table 2], [Table 3], [Table 4], [Table 5], [Table 6]