Article Text

Download PDFPDF

External validation of a 'response score' after neoadjuvant chemotherapy in patients with high-grade serous ovarian carcinoma with complete clinical response
  1. Pier Carlo Zorzato1,
  2. Gian Franco Zannoni2,
  3. Riccardo Tudisco3,
  4. Tina Pasciuto4,
  5. Andrea Di Giorgio5,
  6. Massimo Franchi6,
  7. Giovanni Scambia7 and
  8. Anna Fagotti7
  1. 1 Department of Obstetrics and Gynecology, Ospedale degli Infermi, Ponderano (BI), Biella, Italy
  2. 2 Fondazione Policlinico Universitario A. Gemelli IRCCS, Pathology Division, Department of Woman, Child and Public Health, Rome, Italy
  3. 3 Obstetrics and Gynecology, Catholic University of the Sacred Heart, Rome, Italy
  4. 4 Fondazione Policlinico Universitario A. Gemelli IRCCS, Statistics Technology Archiving Research (STAR) Center, Department of Woman, Child and Public Health, Rome, Italy
  5. 5 Fondazione Policlinico Universitario A. Gemelli IRCCS, Division of Peritoneal and Retroperitoneal Surgery, Rome, Italy
  6. 6 Department of Obstetrics and Gynecology, AOUI Verona, University of Verona, Verona, Italy
  7. 7 Dipartimento Scienze della Salute della Donna e del Bambino, Fondazione Policlinico Universitario A. Gemelli IRCCS, Roma, Italy
  1. Correspondence to Professor Giovanni Scambia, Dipartimento Scienze della Salute della Donna e del Bambino, Fondazione Policlinico Universitario A. Gemelli IRCCS, Roma 8 00168, Italy; giovanni.scambia{at}policlinicogemelli.it

Abstract

Objectives The chemotherapy response score (CRS) has been developed for measuring response to neoadjuvant chemotherapy in tubo-ovarian high-grade serous carcinoma. This study aimed to validate the ability of this three-tier scoring system of pathologic response on omental specimens to determine prognosis in a subgroups of patients who had clinical complete response to neoadjuvant chemotherapy.

Methods This was a retrospective study, conducted in women receiving interval debulking surgery at the Division of Gynecologic Oncology, between December 2007 and April 2017. Inclusion criteria were: high-grade serous ovarian cancer, FIGO stage IIIC/IV, platinum-based neoadjuvant chemotherapy, and clinical complete response after neoadjuvant chemotherapy (normalization in CA125 levels, disappearance of all target and non-target lesions according to RECIST 1.1). CRS was defined by a single pathology review and classified as previously reported: CRS1, no or minimal tumor response with fibroinflammatory changes limited to a few foci ranging from multifocal or diffuse regression-associated fibroinflammatory changes with viable tumor in sheets, or nodules to extensive regression-associated fibroinflammatory changes with multifocal residual tumor; CRS2, appreciable tumor response with viable tumor readily identifiable; and CRS3, complete absence of tumor or nodules with maximum size of 2 mm. CRS was analyzed according to clinical variables and survival.

Results A total of 108 patients were eligible for analysis. The average age was 65 (range 36–85) years. A total of 91 (84.3%) patients had stage IIIC disease and 17 (15.7%) patients had stage IV disease. No statistically significant differences were observed in terms of age, FIGO stage, CA125 serum levels, type of chemotherapy schedules, and number of cycles between the three groups. Patients in the CRS3 group had a longer median progression-free survival (25.8 months) compared with CRS2 or CRS 1 (20.3 vs 17.4 months, respectively; p=0.001). Median overall survival was 68.9 months for CRS3, 35.0 months for CRS2, and 45.9 months for CRS1 (p=0.034).

Conclusion Complete or near-complete pathologic response assessed in the omental specimens of advanced epithelial ovarian carcinoma patients after neoadjuvant chemotherapy (CRS3) is predictive of prolonged progression-free and overall survival. In particular, this is true in women with a clinical complete response.

  • ovarian cancer
  • surgical oncology
  • omentum
  • pathology

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

  • Only 19.4% of patients categorized as having a complete clinical response actually had a complete pathologic response.

  • Complete absence of tumor tissue was found in 40.7% of omental samples and in 28.2% of ovarian samples.

  • Patients in the CRS3 group had a significantly longer progression-free survival compared with those in the CRS2 or CRS1 groups.

Introduction

The standard of care in advanced epithelial ovarian cancer is primary debulking surgery followed by adjuvant chemotherapy. Complete gross tumor resection is the goal of surgery and represents the most powerful predictor of clinical outcome.1 Neoadjuvant chemotherapy and subsequent interval debulking surgery is an alternative treatment in advanced epithelial ovarian cancer which has been demonstrated to offer similar survival results with lower toxicity in selected patients.2 3 In these women, residual tumor at time of interval debulking surgery still has a prognostic value, but a few studies have shown that also pathologic response to neoadjuvant chemotherapy is a strong predictive factor for survival.4 5

Recently, Böhm et al6 have calculated a three-tier scoring system of pathologic response on omental specimens, which can predict the prognostic outcomes in advanced epithelial ovarian cancer patients receiving neoadjuvant chemotherapy. This model has been validated, by other groups, as an effective approach in predicting progression-free survival and platinum sensitivity, but data on overall survival are conflicting.7 Moreover, the efficacy of this score in predicting different outcomes has been tested in women with any clinical response to neoadjuvant chemotherapy, including patients with no pathologic response and thus worse survival.8 9

The aim of this study was to validate the ability of such classification in identifying a subgroup of patients with different prognosis who had clinical complete response to neoadjuvant chemotherapy.

Methods

This was a retrospective study, conducted at the Division of Gynecologic Oncology, Fondazione Policlinico Universitario A. Gemelli IRCCS, Rome, Italy. The institutional review board approved the study and patients gave informed consent for the collection and use of their data for a scientific purpose. Between December 2007 and April 2017, 560 charts of patients with advanced ovarian, tubal, and peritoneal carcinoma receiving interval debulking surgery after neoadjuvant chemotherapy were reviewed. Among these, we selected women who met the following criteria: high-grade serous ovarian cancer, International Federation of Gynecology and Obstetrics (FIGO) stage IIIC/IV, platinum-based neoadjuvant chemotherapy, and complete clinical response after neoadjuvant chemotherapy, according to the criteria of the Gynecological Cancer Intergroup and/or response evaluation criteria in solid tumor criteria (RECIST 1.1) (normalization in CA125 levels, disappearance of all target and non-target lesions).10–12

Clinical and laboratory data including age, FIGO stage, chemotherapy regimen, residual tumor after interval debulking surgery, and CA125 serum levels were retrieved from the patients’ medical records.

All patients were defined as non-elegible for primary debulking surgery, according to clinical, serologic, instrumental data, and/or surgical exploration. Interval debulking surgery was performed either by laparotomy or minimal invasive surgery according to pre-operative evaluation and surgeon preference. Surgical procedures were recorded and graded according to the surgical complexity score by Aletti et al.13 After surgery, patients received post-operative adjuvant chemotherapy.

After completion of primary treatment, all women were triaged to clinical routine follow-up procedures according to National Comprehensive Cancer Network (NCCN) guidelines.14

The resected specimens from interval debulking surger were formalin-fixed and paraffin-embedded according to standard procedures and stained with hematoxylin and eosin (H&E) in the Department of Pathology, Fondazione Policlinico Universitario A. Gemelli IRCCS, Rome, Italy.

A single gynecologic pathologist (GFZ) reviewed all the slides, independently to any data in the original histopathology reports. He assigned a chemotherapy response score (CRS) based on omental assessment, according to Bohm et al.6 In brief, CRS1 corresponds to no or minimal tumor response (no or minimal regression-associated fibroinflammatory changes limited to a few foci), cases in which it is difficult to decide between regression and tumor-associated desmoplasia or inflammatory cell infiltration; CRS2 means appreciable tumor response with viable tumor readily identifiable, ranging from multifocal or diffuse regression-associated fibroinflammatory changes with viable tumor in sheets, streaks, or nodules to extensive regression-associated fibroinflammatory changes with multifocal residual tumor; and CRS3 corresponds to complete absence of tumor cells or individual cells, cell groups, or nodules with maximum size of 2 mm. Moreover, the presence of residual disease was evaluated in the ovaries as well as in all other available sites.

Differences according to CRS status, in terms of clinicopathologic features at diagnosis, and treatment details were analyzed using the Pearson 2 exact test and Kruskal-Wallis test, as appropriate. Progression-free survival was defined as the time elapsed between the date of diagnosis and evidence of recurrence, as assessed by imaging or clinically, or date of last follow-up. Overall survival was defined as the time elapsed between the date of diagnosis and death or date of last censored. Median follow-up was calculated according to the inverted Kaplan-Meier technique.15 Overall survival and progression-free survival curves were estimated by the Kaplan-Meier product limit method16 and compared by log-rank test. For progression-free survival and overall survival, Cox proportional hazards models were used to assess treatment effect at univariate and multivariate analyses. All statistical calculations were performed using the software Stata Version 13.0 (Stata Corp., College Station, TX, USA).

Results

A total of 108 women had all information available, met inclusion criteria, and were included in the final analysis. Fifty-seven patients (52.7%) were evaluated at our institution at the time of primary diagnosis, whereas 51 (47.3%) were referred from other centers after neoadjuvant chemotherapy. Characteristics of the patients enrolled in the study are summarized in Table 1.

Table 1

Patient and pathologic characteristics

At histologic evaluation of the omentum, 24 (22.2%) patients were categorized as CRS1, 18 (16.7%) patients as CRS2, and 66 (61.1%) patients as CRS3.

No statistically significant differences were observed between the three groups in terms of age, FIGO stage, and CA125 serum levels both at the time of diagnosis and after neoadjuvant chemotherapy. Carboplatin/paclitaxel every 3 weeks was the most commonly used neoadjuvant chemotherapy regimen (92.6 %), with or without bevacizumab, which was introduced in combination with standard chemotherapy and as maintenance therapy since its approval in Italy (January 2014). Among the three groups, there was no difference regarding both the type of chemotherapy regimens and number of cycles. The median number of cycles was 4 (range 2–7). Moreover, no difference was found in terms of pathologic response (Table 1), although a higher but not statistically significant rate of CRS3 was found in women having bevacizumab (14/18, 77.7% vs 52/90, 57.7%; p=0.11).

BRCA mutational status, defined as either germline or somatic, was available only for 43 patients, showing no differences in term of pathologic response. All patients received interval debulking surgery with no gross residual disease. Most patients had a low surgical complexity score (96.3%) according to Aletti, and one-third of the patients underwent minimally invasive surgery (Table 2).

Table 2

Surgical procedures performed in the study patients

Even when patients were categorized as having a complete clinical response, only 21 (19.4%) patients had a complete absence of cancer in all resected specimens (defined as complete pathologic response) (Table 3). Complete absence of neoplastic tissue, with no tumor cell or cell groups or nodules of any size, was found in 44 (40.7%) of the omental samples and in 29 (28.2%) of the ovarian samples. No persistence of pathologic disease in additional samples from either/or peritonectomy, lymphadenectomy, bowel resection, and other procedures was observed in 65 patients (60.2%). In particular, among 66 patients with CRS3, 44 (66.7%) had complete cancer regression in the omentum and 21 (31.8%) had complete absence of tumor in all resected sites. Conversely, when the omental tissue was negative for neoplastic cells (44 patients), it was still possible to find vital tumor in other specimens in 23 patients.

Table 3

Histologic findings after neoadjuvant chemotherapy

At the census date of June 2018, 81 (75.0%) patients had recurred and 40 (37.0%) patients had died of disease. Median follow-up was 48.1 months (95% CI 41.8 to 52.1 months) with no significant differences between the three groups (42.4, 55.4, and 48.8 months for CRS1, CRS2, and CRS3, respectively). In the entire series, median progression-free survival was 22.6 (range 6.7–95.8) months and median overall survival was 61.2 (range 10.9–95.8) months. Patients in the CRS3 group had a significantly longer median progression-free survival (25.8 months) compared with CRS2 or CRS1 (20.3 vs 17.4 months, respectively; p=0.001) (Figure 1A). Median overall survival was 68.7 months for CRS3, 35.0 months for CRS2, and 45.9 months for CRS1 (p=0.04) (Figure 1B).

Figure 1

(A) Progression-free survival (PFS) and (B) overall survival (OS) of CRS1, CRS2, and CRS3. CRS, chemotherapy response score.

A subgroup survival analysis was performed considering only patients with CRS3 as a separate group. A significantly longer median progression-free survival was noted for patients with complete pathologic response in all sites (60.7 months) when compared with CRS3 complete and presence of residual disease in other metastatic sites (23.8 months) and CRS3 with minimal pathologic persistence of tumor (22.3 months) (p=0.003) (online supplementary Figure 1A). Patients with complete pathologic response also had better overall survival (median overall survival not reached) versus 65.6 months and 61.2 months, respectively (p=0.35) (online supplementary Figure 1B).

Supplemental material

The prognostic role of pathologic response was confirmed at univariate and multivariate analysis (Table 4). The role of residual tumor at interval debulking surgery was not analyzed, since all our patients underwent complete cytoreduction (RT 0). Moreover, the chemotherapy regimen was not included in the analysis because only 18 patients in the whole series received bevacizumab. Patients categorized as CRS3, compared with patients in the CRS1 and CRS2 groups, had a lower risk of recurrence (HR 0.35, 95% IC 0.20–0.61, p<0.0001) and longer survival (HR 0.38, 95% IC 0.18–0.82, p=0.013).

Table 4

Univariate and multivariate analysis of progression-free and overall survival according to patients' characteristics

Discussion

This study confirms that complete or near-complete pathologic response assessed in the omental specimens of advanced epithelial ovarian cancer patients after neoadjuvant chemotherapy (CRS3) is predictive of prolonged progression-free and overall survival. While this finding has been already described in the literature, the available studies include a wide variety of clinical response (both patients with complete and non-complete (partial or absent) clinical response). Indeed, the inclusion of women with clinical partial or no response could artificially increase the difference in terms of survival between patients with CRS1 and CRS3, thus leading to inconclusive results. We recognize that the retrospective design and the relatively small number of patients are major limitations but, as far as we are aware, this is the first study evaluating the independent prognostic role of the anatomo-pathologic score in a homogeneous population of patients according to clinical criteria. Our study enrolled only women showing clinical complete response, thus overcoming potential limitations of previous studies.

The addition of bevacizumab to standard neoadjuvant chemotherapy increases the rate of CRS3 by approximately 20%, even though this difference is not statistically significant probably because of the small sample size. This information is missing from large prospective randomized trials17 18 and may support the idea that any agent able to improve pathologic response to neoadjuvant chemotherapy is useful for prolonging survival.19 Different rates of CRS found in patients showing clinical complete response underline the inadequacy of radiologic and serologic pre-operative evaluation for distinguishing different types of responses, and suggest that histologic assessment after surgery is a more accurate tool for defining response to chemotherapy and prognosis. Similarly, we confirm the role of interval debulking surgery for achieving complete resection of all microscopic residual disease after neoadjuvant chemotherapy, which is still present in more than 80% of patients considered as having had a clinical complete response. Finally, clinical, radiologic, and serologic criteria have been proven to be effective in selecting a group of patients in whom achieving no residual tumor at the end of surgery is possible with a low surgical complexity score (96.3% of the cases). The striking difference in term of progression-free survival between patients showing complete pathologic response (60.7 months) versus all the others, including CRS3 with complete response but presence of residual disease in other metastatic sites (23.8 months) and CRS3 with minimal pathologic persistence of tumor (22.3 months), confirms that this is the goal to be pursued when administering neoadjuvant chemotherapy. Unfortunately, real pathologic complete response is very rare, accounting for 19.4% of the entire population. Our results also underline that even if omental pathologic assessment after neoadjuvant chemotherapy is useful for prognostic and therapeutic purposes, it is not very precise. Indeed, more than half of the patients completely responding in the omentum still have residual disease at other sites (23/44, 52.7%) and their prognosis is not different with respect to those having minimal residual tumor in the omentum or CRS2 (online supplementary Figure 1). Different organ vascularization or different molecular signatures might explain different response rate in different tumor sites.

Contrary to the work published by Lee and colleagues,20 we noted no statistically significant difference between CRS1 and CRS2 in term of progression-free survival and overall survival, most likely due to the smaller size of CRS2.

In several tumors, including breast and rectal cancer,21–24 the regression after neoadjuvant chemotherapy assessed by imaging techniques and by histologic evaluation of biopsies has allowed clinicians the ability to identify subgroups of patients with different outcomes, representing a turning point into the management of these malignancies, with a reduction of the effective role of surgical intervention for good responders. At least theoretically, the extrapolation of such a principle to ovarian cancer might allow the avoidance of interval debulking surgery in women with extremely favorable prognosis and true complete clinical and pathologic response. Indeed, the role and extension of interval debulking surgery is still debated, and some data suggesting that there is no difference in terms of survival in advanced ovarian cancer patients not receiving surgery for any reason.25 To date, a parameter able to predict the class of tumor regression in ovarian cancer after neoadjuvant chemotherapy does not exist and more efforts are needed in this area. The CRS system proposed by Bohm et al6 is a simple and reproducible method that we have validated in a selected population. It is not ideal, but it has the potential to be a surrogate of a more precise prognostic classification of response to chemotherapy. However, it must be kept in mind that the possibility of avoiding interval debulking surgery is only a speculation and is absolutely not supported by published data. In the future, CRS3 may be the target to consider when designing clinical studies on neoadjuvant chemotherapy. Detection of potential molecular predictive factors for CRS3 will be helpful in identifying patients with different prognoses and assist in treatment personalization.

Supplemental material

References

  1. 1.
  2. 2.
  3. 3.
  4. 4.
  5. 5.
  6. 6.
  7. 7.
  8. 8.
  9. 9.
  10. 10.
  11. 11.
  12. 12.
  13. 13.
  14. 14.
  15. 15.
  16. 16.
  17. 17.
  18. 18.
  19. 19.
  20. 20.
  21. 21.
  22. 22.
  23. 23.
  24. 24.
  25. 25.

Footnotes

  • Twitter @annafagottimd

  • Contributors ZPC: study conception/design; data collection; data analysis/interpretation; manuscript preparation. ZGF: study conception/design; data analysis/interpretation. TR: data collection; data analysis/interpretation. PT: data collection; data analysis/interpretation; statistical analysis. DGA: responsible surgeon or imager; patient recruitment. FM: responsible surgeon or imager; patient recruitment. SG: responsible surgeon or imager; patient recruitment. FA: study conception/design; data analysis/interpretation; responsible surgeon or imager; manuscript preparation; patient recruitment.

  • Funding The authors have not declared a specific grant for this research from any funding agency in the public, commercial or not-for-profit sectors.

  • Competing interests None declared.

  • Patient consent for publication Not required.

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

  • Data availability statement Data are available upon reasonable request.