Special Feature

Highlights from The 38th Annual Meeting of The Society of Gynecologic Oncologists

By Robert L. Coleman, MD, Professor & Director, Clinical Research, Department of Gynecologic Oncology, University of Texas, M.D. Anderson Cancer Center Department of Gynecologic Oncology, Houston, is Associate Editor for OB/GYN Clinical Alert.

The Annual Meeting of the Society of Gynecologic Oncologists (SGO) represents the one completely focused platform to exchange preclinical and clinical information in the subspeciality of gynecologic oncology. This year more that 1300 health care providers, researchers and industry representatives participated in a conference offering much more than distribution of the latest scientific and clinical advances. The format also provided information and hands-on training for practice management, surgical skills, and education as well as discussion of controversial topics via focused sessions and debates. The scientific program consisted of 319 abstracts; 79 plenary and focused plenary presentation and 240 poster presentations.1 From these I have selected 9 abstracts to briefly review as they represent some of the important advances in the field and will be of interest, I believe, to the readership.

Ovarian Cancer

Significance of pretreatment CA-125 level in advanced ovarian carcinoma: a meta-analysis of seven Gynecologic Oncology Group protocols. Zorn et al [Abstract 3].

Since the landmark paper in 1996 by McGuire and colleagues, paclitaxel and platinum has been the recognized standard of care for primary adjuvant therapy of advanced ovarian cancer. Including this pivotal study, there have been 6 other phase III efforts by the GOG in which this regimen has been evaluated in different cohorts of patients to determine efficacy and gather important data on toxicity and survival. The database provides a unique resource to evaluate important prognostic and predictive factors, such as CA125. The investigators of the current study reviewed CA125 determinations from 1299 patients participating in these trials, most of whom had registration values post-operatively but before administration of chemotherapy. Not surprising, 20% of patients had values in excess of 1000 U/mL; however, CA125 was within normal limits in 8%. Although this might be normally expected in patients with mucinous ovarian cancers, nearly 70% of patients with this histology had an elevated pretreatment CA125 (median 99 U/mL). As a prognostic factor, a one-fold increase in CA125 was associated with a 9% increase in the hazard for disease progression. The association was even more striking among patients debulked to microscopic residual (15% increase in hazard) and in those with endometrioid tumors (17% increase in hazard). The report highlights the variability of CA125 and the important, albeit not perfect, role this biomarker plays in the care of women with ovarian cancer.

Ovarian cancer screening in the Prostate, Lung, Colorectal and Ovarian Cancer Screening Trial: Results from 4 years of annual screening in a randomized trial. Partridge et al. [Abstract 10]

As is well appreciated in other solid tumors, the best treatment strategy is prevention/early detection. Controlling for stage, survival rates of ovarian cancer differ little from those of breast and colon; however, the vast majority—nearly 80%—of new ovarian cases are diagnosed with significant tumor burden distributed throughout the peritoneal cavity or extra-abdominally. Long-term survival is limited in these cases despite advances in surgical technique and chemotherapy. By contrast, survival in stage I disease is excellent (85%-90% at 10 years). This differential has sparked the search for effective screening modalities. Currently, our best technology and algorithms rely on imperfect tools such as transvaginal ultrasound and biomarkers (eg, CA125) and have had limited success. Nonetheless, around the world, large-scale randomized trials are being conducted to determine if general population screening holds merit for this disease.

In this study, investigators participating in the PLCO trial presented results from the screening protocol over the first 4 years. A total of 39,115 women have been randomized to receive transvaginal ultrasound and CA125 on an annual basis. To date, 98 women have been identified with ovarian cancer; of these, 63% were detected by the screening process. The remainder was screen-negative and diagnosed between scheduled screens. The majority of cancer detected by ultrasound were stage I and II; the majority identified by CA125 abnormalities were stage III and IV. Surgical intervention was more common following an abnormal ultrasound as compared to an abnormal CA125; however, the yield of a cancer diagnosis was 10-fold less. Both modalities produced a positive predictive value under 4% (1% for CA125 screening and 3.7% for ultrasound). These numbers reflect the low prevalence of disease in the screening population, as well as, the poor performance of the testing modality. However, the impact of detection even among these few cases awaits reference as the control group for this trial is still blinded. Ultimately, unless mortality is impacted, the strategy will be considered a failure. Currently, strategies such as using serial CA125 determination and probability assignments to sequentially determined values, alone and in the context of an ultrasound screening algorithm, as well as novel biomarker development represent new avenues for this important function.

Gene expression signature predicts for response to chemotherapy in advanced-stage serous papillary ovarian cancer. Ozbun et al. [Abstract 55]

Despite key advances in drug development, selection of individual agents is largely empiric. The quest for identifying precise ways to select agents and/or predict those in whom a particular therapy is not likely to help is an old but persistent effort. In this study, investigators used sophisticated tools to develop and then validate the impact a particular gene signature has on response to treatment. To do this, they microdissected tissue specimens from previously untreated, advanced stage, serous ovarian cancer patients and profiled thousands of genes. The women were classified based on clinical response to primary therapy: "chemosensitive" (progression-free survival (PFS) of at least 6 months), "chemoresistant" (PFS of less than 6 months) and "chemorefractory" (progressed on primary therapy). They identified 31 genes predicting resistance and 105 genes predicting refractory phenotypes. When applied to an independent data set, the classifier lists predicted 90% resistant and 92% refractory gene signatures. Numerous biologically relevant processes were represented in the data mining including collagen, apoptotic, cell survival and DNA repair genes. The ability to distinguish the clinical phenotype of these tumors is an important step in identifying novel targets and pathways for future therapy trials. While the current study does not provide a "menu" of drug choices that may work as therapy, it does demonstrate the potential futility of empiric drug selection. Since these results can be done a priori, future work will help improve the precision of constructing individualized and strategic treatment plans.

Access to gynecologic oncologists and its impact on survival of women with epithelial ovarian cancer. Chan et al. [Abstract 17]

In recent years, a number of articles have begun to appear in the literature detailing and validating, essentially, the mission statement of the Society—that its, to provide specialized and coordinated care to women with gynecologic malignancies. It is disheartening to document that despite these benefits, the majority of women in the US with gynecologic cancers will receive fragmented care largely from those unfamiliar with the full spectrum of specialized services that are offered by gynecologic oncologists (GYO). In this report, data from all patients included in the California Cancer Registry from 1994 to 1996 were analyzed to determine patterns of delivered care. Nearly 1500 patients were included; approximately 34% received care from GYOs. Compared with those who were cared for by non-GYOs, women receiving care from GYOs were more likely to live in urban areas, be more educated, affluent and "white-collar" employees. In addition, they were also more likely to receive primary surgery, present in advanced stage, have higher grade tumors and receive chemotherapy. Women who did not have surgery by a GYO were 4 times less likely to be staged. Despite the stage shift, all categories of women cared for by GYOs had longer survival—nearly double in those with the most advanced disease. Although clearly "preaching to the choir" the investigators highlight the disparity in health care delivery and impact from that access. On January 12, 2007, President Bush signed into law "The Gynecologic Cancer Education and Awareness Act (aka "Johanna's Law"). The result of years of grassroots efforts from concerned family members and advocates as well as members from the gynecologic oncology community, the law provides special recognition of gynecologic malignancy and provides educational opportunities to women and healthcare professionals not familiar with the diseases encompassed under the gynecologic oncology umbrella. Bolstered by data from this study and others identifying similar results, it is our hope the likelihood a woman with gynecologic malignancy will see a gynecologic oncologist will increase substantially.

Endometrial Cancer

Survival of endometrial cancer patients after laparoscopically assisted vaginal hysterectomy or total abdominal hysterectomy: Analysis of risk factors. Leiserowitz et al. [Abstract 7].

Current trends in the patterns of care for women with endometrial cancer include more complete surgical staging, less pelvic radiation, increasing use of adjuvant chemotherapy and more common use of laparoscopy for surgical extirpation. However, a distinct concern limiting universal adoption of the latter was the potential for suboptimal surgical evaluation of "at risk" areas relative to standard ceiliotomy. However, emerging data, such as the current abstract appear to quell at least some of the concern attributed to the potential of reduced survival. In a review of the State of California databases, Leiserowitz collected information on 978 women with endometrial cancer undergoing laparoscopic procedures and 11,765 women undergoing laparotomy. Outcome variables were demographics, comorbidities, nodal evaluation, use and type of adjuvant therapy and survival. The authors found that women undergoing laparoscopy tended to be younger, less sick, and with more favorable tumors compared to those undergoing laparotomy. However, cause-specific survival was significantly better in the laparoscopy group even after considering age, race, stage, comorbid conditions and use of radiation therapy. In the lowest risk groups the difference reached parity; however, in high-risk cohorts the benefit was pronounced. Although high selection bias exists in a study such as this (< 10% of participants underwent laparoscopy) the trends were of interest. Currently the gynecologic oncology community is awaiting maturity of a GOG randomized clinical trial of laparoscopy to laparotomy in women with early stage corpus malignancies. Until these results are known, we need to carefully counsel patients as to the treatment plan for their surgical management, particularly if an endoscopic approach is undertaken.

A cost-effectiveness analysis of prophylactic surgery vs gynecologic surveillance for women from hereditary nonpolyposis colorectal cancer families. Yang et al. [Abstract 68].

Hereditary nonpolyposis colorectal cancer (HNPCC) is a familial syndrome commonly recognized for its association with early age onset colorectal cancer. However, recent investigation suggests carcinoma of the endometrium is just as likely. In addition, women from HNPCC families are at increased risk of ovarian cancer. The current study evaluated whether risk-reducing surgery (as is commonly advocated in women who are BRCA mutation carriers) is cost effective compared with surveillance. In this report, surveillance was described in two protocols: annual exam, endometrial biopsy, transvaginal ultrasound and CA125 or annual exam alone. For the purposes of this study, risk-reducing surgery was to be performed as a hysterectomy, bilateral salpingoophorectomy (H-BSO) at age 30. Both non-discounted and 3% discounted lifetime costs were calculated. In reference to the two surveillance cohorts, the longest expected survival was observed in the prophylactic H-BSO, followed by annual screening and then annual exam. The most costly was annual exam alone—$83,876 lifetime cost for 25.27 discounted life-years. By contrast, prophylactic surgery cost $24,534 per patient for 25.71 discounted life-years and represents the dominant strategy. Although this analysis provides some important risk-reducing information for at risk women, caution must be used in interpretation as quality of life-years gained and indirect costs were not included in the model. Further information from prospectively counseled women would be necessary to clarify important challenges and barriers to an effective screening strategy in women from HNPCC families.

Cervix Cancer

Type III radical hysterectomy for obese women with cervical carcinoma: Robotic vs open. Shafer et al [Abstract 20]

A comparison of total laparoscopic hysterectomy and abdominal radical hysterectomy for cervix cancer Frumovitz et al. [Abstract 21]

While a few authors have suggested the feasibility of total laparoscopic radical hysterectomy over the last 10 years, experience is mounting and these two studies nicely represent a response to the call for comparative data in centers undergoing the "learning curve" of adopting a new procedure into their clinical repertoire. The study by Shafer and colleagues comes from a center known for endoscopic expertise but reports on consecutive patients undergoing robotic surgery to complete a radical hysterectomy and bilateral pelvic lymphadenectomy (RHPLND). In this series, data were reviewed from 31 consecutive patients undergoing robotic RHPLND, 13 of whom were found to meet the criteria for obesity (BMI > 30) and 48 women (11 obese) undergoing open RHPLND in the 6 years preceding institution of robotic surgery. Outcome variables were predominately surgical such as operative time, blood loss, node count, hospital stay and morbidity. In this series, the endoscopic cases were associated with statistically higher node counts, lower operative blood loss and shorter hospitalization. No difference in operative time, BMI or morbidity was observed between the cohorts. The authors concluded robotic surgery is feasible in obese women and offers advantages in women at highest risk for operative complications.

In the report by Frumovitz and colleagues, adoption of laparoscopic RHPLND occurred in a center where several but not all physicians had recently learned the procedure. All patients over an 18-month period undergoing either procedure were analyzed for operative and post-operative characteristics. The sample included 97 patients of whom 26 underwent laparoscopic RHPLND. No differences were observed in BMI, age, tumor parameters, stage or race. Similar to the previous report, lower blood loss and hospitalization were observed in the endoscopically treated patients. Similar lymph node counts, pathologic parameters (parametrial and vaginal cuff margins), and post-operative urinary recovery were observed between the cohorts. In addition, post-operative infectious morbidity was significantly lower for those undergoing the laparoscopic approach. Survival data were immature. To address the issue of operator experience on outcomes, these investigators are currently undertaking a prospective trial evaluating surgical performance of novice surgeons in a mentoring program with attention to the "learning curve."

These two studies represent recent trends in advanced operative laparoscopy in gynecologic cancer. At the Annual Meeting, accompanying these reports were 2 full days of "hands-on" laparoscopic surgical skill courses. The educational format, which combines lectures and mentored cadaveric training is highly effective in providing the necessary tools for clinical adoption of these procedures and the courses, perennially, are among the most popular. Despite the enthusiasm, it is important to recall the series such as those presented are typically comparing procedures done by highly trained attending physicians with an interest in minimally invasive surgery to those done by gynecologic oncologists (often times operative with fellow trainees). Even the most ardent supporter of laparoscopic surgery would admit that there is a learning curve—steep for radical procedures—which must be overcome to reach parity to their results from open laparotomy. However, the interest and outcome data support the trend and a call for an international randomized clinical trial comparing laparoscopic and open radical hysterectomy has been made.

Translational Medicine

Characterization and safety of in vivo small interfering RNA delivery in neutral nanoparticles. Sood, et al [Abstract 6]

One of the most prolific and promising areas in contemporary drug development is the construction of novel target-specific therapeutics affecting key biological processes of cellular function. For example, more than 300 agents have been developed targeting angiogenesis alone in recent years. A key strategy to document the specific functions of a specific gene is through selective silencing using short interfering RNA (siRNA) molecules. Preclinically, this relatively recent discovery has enabled scientists to efficiently investigate the impact of gene silencing and the potential for therapeutic targeting. However, to date, a systemically deliverable siRNA-based compound has not been successfully developed. Sood and colleagues demonstrate in an animal model of intraperitoneal ovarian cancer the ability to "hit" an intracellular target, FAK, selectively in tumor tissues utilizing siRNA. The delivery vehicle was a novel neutrally charged liposomal nanoparticle, which was demonstrated to diffuse deeply into tumors and dysregulate FAK signaling. This was confirmed in multiple tumor models including resistant strains. The efficacy in ovarian cancers was noted as a single agent, which was augmented in combination with a common cytotoxic agent used in humans with ovarian cancer. While distinctly preclinical, it should be noted that a first-in-human construct utilizing this same liposomal nanoparticle and siRNA has been manufactured and soon will be entering human investigation. If successful, the methodology will have broad application in human disease.

Note: Complete Abstracts can be found in:

1. Abstracts, Gynecol Oncol. (2007) 104: S2-S35