Article Text

Download PDFPDF
Pathologic distribution at the time of interval tumor reductive surgery informs personalized surgery for high-grade ovarian cancer
  1. Courtney D Bailey1,
  2. Rebecca Previs2,
  3. Bryan M Fellman3,
  4. Tarrik Zaid4,
  5. Marilyn Huang5,
  6. Alaina Brown6,
  7. Ahmed Enbaya4,
  8. Nyla Balakrishnan7,
  9. Russell R Broaddus8,
  10. Diane C Bodurka4,
  11. Pamela Soliman4,
  12. Nicole D Fleming4,
  13. Alpa Nick9,
  14. Anil K Sood4 and
  15. Shannon Neville Westin4
  1. 1 Obstretrics and Gynecology, Division of Gynecologic Oncology, Augusta University Medical College of Georgia, Augusta, Georgia, USA
  2. 2 Obstretrics and Gynecology, Division of Gynecologic Oncology, Duke Cancer Institute, Durham, North Carolina, USA
  3. 3 Biostatistics, University of Texas MD Anderson Cancer Center, Houston, Texas, USA
  4. 4 Gynecologic Oncology and Reproductive Medicine, University of Texas MD Anderson Cancer Center, Houston, Texas, USA
  5. 5 Obstretrics and Gynecology, Division of Gynecologic Oncology, Sylvester Comprehensive Cancer Center, Miami, Florida, USA
  6. 6 Obstretrics and Gynecology, Division of Gynecologic Oncology, Vanderbilt University Medical Center, Nashville, Tennessee, USA
  7. 7 Public Health, School of Dental Medicine, University of Pittsburgh, Pittsburgh, Pennsylvania, USA
  8. 8 Pathology and Laboratory Medicine, University of North Carolina System, Chapel Hill, North Carolina, USA
  9. 9 Gynecologic Oncology, Tennessee Oncology, Nashville, Tennessee, USA
  1. Correspondence to Dr Shannon Neville Westin, Gynecologic Oncology and Reproductive Medicine, University of Texas MD Anderson Cancer Center, Houston, TX 77030, USA; swestin{at}mdanderson.org

Abstract

Introduction The surgical approach for interval debulking surgery after neoadjuvant chemotherapy has been extrapolated from primary tumor reductive surgery for high-grade ovarian cancer. The study objective was to compare pathologic distribution of malignancy at interval debulking surgery versus primary tumor reductive surgery.

Methods Patients with a diagnosis of high-grade serous or mixed, non-mucinous, epithelial ovarian, fallopian tube or primary peritoneal cancer who underwent neoadjuvant chemotherapy or primary tumor reductive surgery and had at least 6 months of follow-up were identified through tumor registry at a single institution from January 1995 to April 2016. Pathologic involvement of organs was categorized as macroscopic, microscopic, or no tumor. Statistical analyses included Mann-Whitney and Fisher’s exact tests.

Results Of 918 patients identified, 366 (39.9%) patients underwent interval debulking surgery and 552 (60.1%) patients underwent primary tumor reductive surgery. Median age was 62.3 years (range 25.3–92.5). The majority of patients in the interval debulking surgery group were unstaged (261, 71.5%). In the patients who had a primary tumor reductive surgery, 406 (74.6%) had stage III disease. In both groups, the majority of patients had serous histology: 325 (90%) and 435 (78.8%) in the interval debulking and primary tumor reductive surgery groups, respectively. There was a statistically significant difference between disease distribution on the uterus between the groups; 31.4% of the patients undergoing interval debulking surgery had no evidence of uterine disease compared with 22.1% of primary tumor reductive surgery specimens (p<0.001). In the adnexa, there was macroscopic disease present in 253 (69.2%) and 482 (87.4%) of cases in the interval vs primary surgery groups, respectively (p<0.001). Within the omentum, no tumor was present in the omentum in 52 (14.2%) in the interval surgery group versus 91 (16.5%) in the primary surgery group (p<0.001). In the interval surgery group, there was no tumor involving the small and large bowel in 49 (13.4%) and 28 (7.7%) pathologic specimens, respectively. This was statistically significantly different from the small and large bowel in the primary surgery group, of which there was no tumor in 20 (3.6%, p<0.001) and 16 (2.9%, p<0.001) of cases, respectively.

Conclusion In patients undergoing interval debulking surgery, there was less macroscopic involvement of tumor in the uterus, adnexa and bowel compared with patients undergoing primary cytoreductive surgery.

  • ovarian cancer
  • gynecologic surgical procedures

Statistics from Altmetric.com

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.

HIGHLIGHTS

  • There is less macroscopic tumor in uterine, adnexal, and bowel specimens obtained after neoadjuvant chemotherapy.

  • After neoadjuvant chemotherapy, surgical specimens may have microscopic or no residual tumor at interval surgery.

  • 51.6 % of patients had no residual disease in the hysterectomy specimen after neoadjuvant chemotherapy.

Introduction

The approach for primary tumor reductive surgery for ovarian, primary peritoneal, and fallopian tube carcinoma has been used to extrapolate the technique for interval debulking surgery after neoadjuvant chemotherapy. The goal of surgery is optimal cytoreduction to achieve no gross residual disease, as maximum reduction of all visible disease to no gross visible disease has demonstrated improved progressive-free interval and overall survival.1 The National Comprehensive Cancer Network (NCCN) recommends bilateral salpingo-oophorectomy with hysterectomy and gross tumor debulking after neoadjuvant chemotherapy.2 It remains unknown whether interval debulking surgery must include the same procedures as primary tumor reductive surgery. In fact, optimal cytoreduction to no gross residual disease at the time of interval debulking surgery may be possible with fewer procedures, such as bowel resections and upper abdominal procedures, than those required during primary tumor reductive surgery.3–5

Few studies have been performed to assess the need for hysterectomy in primary tumor reductive surgery. Bunting et al 6 found that the majority of gynecological oncologists perform hysterectomy because they believe it is important for staging or because there is a possibility for synchronous tumor in the uterus. However, the necessity of hysterectomy is unclear, given that current evidence suggests that involvement of the uterus may be infrequent both at the time of primary tumor reductive surgery or interval debulking surgery.6 7 Prior studies, however, were limited by small sample size and lacked quantification of the amount of tumor involving the uterus and cervix. Based on these data, cytoreduction to no gross visible disease may be achieved in many cases without hysterectomy, even in the upfront setting. Importantly, retrospective data suggest that patients who underwent a supracervical hysterectomy were not more likely to have a vaginal or cervical recurrence. Patients who had a supracervical hysterectomy, however, were less likely to have tumor erosion into the vagina compared with patients who underwent a total abdominal hysterectomy, although this did not reach statistical significance.8

To our knowledge, no studies have specifically examined the involvement of the uterus and cervix in patients undergoing interval debulking surgery after neoadjuvant chemotherapy. With increasing use of neoadjuvant chemotherapy, the clinical implications of whether residual disease is present in the uterine or cervical specimens could influence whether a total hysterectomy or supracervical hysterectomy should be performed. We hypothesize that forgoing a hysterectomy could lead to a simpler operation with decreased blood loss, shorter operating time, and could potentially decrease long-term complications, such as pelvic organ prolapse and urinary incontinence.9–11 A more personalized surgical approach that may involve less surgery could minimize surgical complications and expose patients to less risk by eliminating procedures that have no proven benefit. However, individual factors such as personal family history or the presence of a germline mutation that predisposes a patient to gynecological malignancies must also be considered during surgical planning. The objective of the study was to characterize the pathologic residual volume of disease in uterine and other surgical specimens at interval debulking surgery after neoadjuvant chemotherapy compared with primary tumor reductive surgery.

Methods

A retrospective cohort study was performed on women diagnosed with ovarian, primary peritoneal, or fallopian tube carcinoma ('ovarian cancer') between January 1995 and April 2016 at the University of Texas MD Anderson Cancer Center. In accordance with the journal’s guidelines, we will provide our data for the reproducibility of this study in other centers if requested. This study was approved by the Institutional Review Board. Patients were identified through the Tumor Registry and Gynecologic Oncology surgical database. Demographic data and surgical pathologic information were collected from the electronic medical record. Patients were included if they had undergone primary or interval tumor reductive surgery for high-grade serous or mixed, non-mucinous, epithelial ovarian cancer, fallopian tube, or primary peritoneal cancer and had at least 6 months of follow-up data. Patients with low-grade tumors or multiple primary cancers were excluded. Patients were excluded if there was no pathologic report available.

Descriptive statistics were used to summarize the demographic, clinical, and surgical characteristics and patients were stratified by the type of surgery (primary tumor reductive surgery vs interval debulking surgery). Wilcoxon rank-sum, χ2, and Fisher’s exact tests were used to compare demographic and clinical characteristics. The proportion of patients with macroscopic, microscopic, and no disease present in each surgical specimen was compared by type of surgery using the χ2 test or Fisher’s exact test. Surgical specimens were categorized as free of residual tumor (no tumor), having microscopic tumor present, or grossly present (macroscopic) tumor on the serosal surface. All analyses were performed with Stata 14.1 (College Station, TX). P-values less than 0.05 were considered statistically significant.

Results

A total of 918 patients were identified who underwent primary tumor reductive surgery or interval debulking surgery after administration of neoadjuvant chemotherapy. The median age was 62.3 years (range 25.3–92.5). In the patients who had a primary tumor reductive surgery, 406 (74.6%) had stage III disease. In both groups, the majority of patients had serous histology: 325 (90%) and 435 (78.8%) in the interval debulking and primary tumor reductive surgery groups, respectively. Of these, 552 (60.1%) of patients had primary cytoreduction, while 336 (39.9%) underwent interval surgery. Nine patients were excluded due to synchronous primary ovarian and endometrial cancers. Five of these had a known diagnosis of endometrial cancer prior to their surgery. Demographic and clinical characteristics of the two cohorts are shown in Table 1. There was no significant difference between BMI, race, or ethnicity between these two groups. Patients in the interval debulking surgery group compared with the primary surgery group were more likely to be unstaged (261, 71.5% vs 13, 2.4%) or have stage IV disease (71, 19.5. % vs 52, 9.6%) (p=0.001) and have primary peritoneal cancer (88, 24.2. % vs 104, 18.6%) (p<0.001).

Table 1

Demographic characteristics: comparison between interval debulking surgery versus primary tumor reductive surgery. Data are numbers (%) unless stated otherwise

Surgical characteristics of the groups are demonstrated in Table 2. The proportion of patients that did not have hysterectomy was similar in both arms (interval debulking surgery 39.7% vs primary tumor reductive surgery 34.8%, p=0.13). Residual disease was significantly different between the two groups, with no gross residual more often in the interval debulking surgery group (52.8% vs 40.9%, p=0.005).

Table 2

Surgical characteristics: comparison between interval debulking surgery versus primary tumor reductive surgery. Data are numbers (%) unless indicated otherwise

The volume of residual disease in pathologic specimens was significantly different in patients undergoing interval debulking surgery compared with primary tumor reductive surgery (Table 3). In the adnexa, there was macroscopic disease present in 253 (69.2%) and 482 (87.4%) of cases in the interval versus primary surgery groups, respectively (p<0.001). Within the omentum, no tumor was present in 52 (14.2%) in the interval surgery group versus 91 (16.5%) in the primary surgery group (p<0.001). Thirty-five per cent of patients (242/918) had a bowel resection during tumor reductive surgery. In the interval surgery group, there was no tumor involving the small and large bowel in 49 (13.4%) and 28 (7.7%) of pathologic specimens, respectively. This was statistically significantly different from the small and large bowel in the primary surgery group, of which there was no tumor in 20 (3.6%, p<0.001) and 16 (2.9%, p<0.001) cases, respectively. Among patients who had hysterectomy, patients who underwent interval debulking surgery were more likely to have no tumor involvement of the uterine serosa compared with patients who underwent primary tumor reductive surgery (Figure 1; 34.9% vs 51.6%, p=<0.001). In addition, macroscopic involvement was found significantly less in uterine specimens at the time of interval debulking surgery compared with primary tumor reductive surgery (31.0% vs 52.0%, p=<0.001). The surface or stroma of the cervix was involved with tumor in only 4.1% and 4.0% of cases in the interval debulking surgery and primary tumor reductive surgery groups, respectively (p=0.89).

Figure 1

Pathologic distribution of disease of uterine specimens.

Table 3

Pathologic residual disease in surgical specimens. Data are numbers (%)

Discussion

Pathologic distribution of disease was significantly different between patients undergoing interval debulking surgery versus those undergoing primary tumor reductive surgery for upfront treatment of advanced stage ovarian cancer. Although traditionally the surgical technique for interval debulking has been based on procedures performed at the time of primary tumor reductive surgery, our data reflect that performing a hysterectomy may not be necessary for patients undergoing interval debulking surgery after neoadjuvant chemotherapy. Surgeons may consider the omission of hysterectomy or rather performing a supracervical hysterectomy given the relative infrequency of cervical involvement. In the absence of visible disease, the patient may not require bowel resection or an upper abdominal procedure. Our study demonstrated a larger number of specimens with macroscopic uterine involvement at the time of primary tumor reductive surgery than previous studies,6 7 with about 50% of specimens involved. This is in comparison to a retrospective case series which showed macroscopic involvement of the uterine serosa in only 30% of specimens.6 This study did not characterize microscopic disease pathologically. Similar to our data, another retrospective study identified that 47.5% of uteri had no involvement of tumor at the time of debulking surgery.7

The risk of a synchronous uterine cancer must also be considered if a hysterectomy is not performed as part of a primary cytoreductive or interval debulking surgery. Although the risk is low, synchronous uterine cancer has previously been found in 5%–10% of ovarian cancer patients, typically of endometrioid histology.6 9 A study from our institution characterized 84 patients diagnosed with a synchronous uterine and ovarian cancer between 1989 and 2002.9 As expected, the majority of these patients had an endometrioid uterine cancer and almost half were symptomatic with abnormal uterine bleeding. Uterine premalignancy is also a consideration and one study of 186 patients found that uterine premalignancies were found in 28% of patients, including endometrial intraepithelial carcinoma in 4% of patients and atypical hyperplasia in 24%.10 This was a relatively small, retrospective study, however, and overall, this has not been well studied in the literature. Risk factors that increased the likelihood of synchronous endometrioid and ovarian cancer included premenopausal status, younger age, and obesity.9 11

To safely consider uterine conservation as an option, concurrent synchronous primary uterine cancer should be ruled out. This can be accomplished via preoperative endometrial sampling and it should be performed if the patient is experiencing any abnormal uterine bleeding. Physicians should counsel patients that if uterine conservation is desired, the possibility of a future hysterectomy could ultimately be required. In our series, there were only nine patients (less than 1%) who were excluded due to synchronous primary of the ovary and endometrium, and five of these were known prior to surgery. Preoperative uterine sampling would likely have found the other four patients prior to surgery if uterine conservation had been desired.

Neoadjuvant chemotherapy also impacted the pathologic distribution of ovarian cancer in the gastrointestinal tract. While 29.9% of large bowel specimens had macroscopic involvement at the time of upfront surgery, 17.2% had macroscopic involvement after neoadjuvant chemotherapy. Patients who had a bowel resection without evidence of macroscopic or microscopic disease underwent resections for suspected disease that was ultimately negative pathologically, had concern for obstruction, or compromised blood supply during dissection. The decision to resect bowel without evidence of disease must be weighed against the risk of complications. In one study, large bowel resections were found to increase postoperative complications in 2%–30.5% of patients, including prolonged ileus, perforated viscous, anastomotic leaks, and short bowel syndrome.3 This study estimated 87% of patients would have required a bowel resection to receive no gross residual disease during primary tumor reductive surgery, while only 49% required a bowel resection at interval debulking surgery.3

The primary strength of our study is the large number of patients in both the primary tumor reductive surgery and interval debulking surgery groups. All pathology specimens were reviewed at the University of Texas MD Anderson Cancer Center, conferring consistent results in the reporting of pathology. Our study has the inherent weaknesses of a retrospective study. This includes inherent bias in the selection of patients for each intervention as approximately two-thirds of patients underwent primary tumor reductive surgery. Additional bias may have been introduced due to exclusion of patients with synchronous primary tumors. However, the incidence of synchronous primary endometrial tumors is known to be low and has been well described at our institution previously.9 Although all pathology was reviewed by a gynecological pathologist, no re-review was performed for the purpose of this study. The study inclusion criteria were broad, therefore the procedures performed were not consistent across all patients. Residual disease was determined only by operative report and there was no confirmation by postoperative imaging.

Overall, this study shows the pathologic distribution of disease varies between specimens obtained at the time of primary tumor reductive surgery and interval debulking surgery after neoadjuvant chemotherapy. The incidence of pathologic involvement of most specimens in our study was decreased after neoadjuvant chemotherapy, indicating a significant pathological response. These specimens included uterus, adnexa, omentum, small and large bowel. These data suggest that fewer procedures may be required after neoadjuvant chemotherapy to achieve no gross residual, including a hysterectomy or bowel resection. This study is hypothesis generating, and additional studies are needed to determine if hysterectomy is truly needed. Further, are there differences in surgical morbidity, survival, and patient-reported outcomes in patients who do not undergo hysterectomy (or who undergo supracervical hysterectomy) at the time of interval debulking surgery? Patients treated with neoadjuvant chemotherapy may have minimal residual disease in pathologic specimens. Surgical staging can provide value when organ involvement is unclear, but patients should be counseled that pathology reports may not show definitive involvement by tumor. Our data suggest that physicians may consider further individualization of surgical approaches for patients who have received neoadjuvant chemotherapy.

References

Footnotes

  • Twitter @PamSolimanMD, @nicoleflemingmd, @Shannon.Westin

  • Contributors CDB, RP, SNW: conception, statistical analysis, critical analysis, drafting/final editing. BMF: statistical analysis, critical analysis, drafting/final editing. TZ, MH, AB, AE, NB, RRB, DCB, PTS, NDF, AMN, AKS: critical analysis, drafting/final editing.

  • Funding NIH K12CA088084 K12 Calabresi Scholar Award, NIH 1P50CA217685-01 SPORE in Ovarian Cancer, NIH P30CA016672 MD Anderson Cancer Center Support Grant, Andrew Sabin Family Fellowship GOG Foundation Scholar Investigator Award.

  • Competing interests RP receives research support from Myriad. AKS is a consultant for Merck, and Kiyatec. AKS receives research funding (M-Trap); and is a shareholder (BioPath). SNW is a consultant for AstraZeneca, Circulogene, Clovis Oncology, Eisai, GSK/Tesaro, Merck, Novartis, Pfizer, and Roche/Genentech. SNW receives research support from ArQule, AstraZeneca, Bayer, Clovis Oncology, Cotinga Pharmaceuticals, Novartis, Roche/Genentech, and GSK/Tesaro. All conflicts of interest are not related to the current work.

  • Patient consent for publication Not required.

  • Provenance and peer review Not commissioned; externally peer reviewed.

  • Data availability statement All data relevant to the study are included in the article.

  • Author note All work was done at MD Anderson Cancer Center.