Article Text

Download PDFPDF

Histopathologic response to neoadjuvant chemotherapy as a prognostic biomarker in tubo-ovarian high-grade serous carcinoma: updated Chemotherapy Response Score (CRS) results
  1. Steffen Böhm1,
  2. Nhu Le2,
  3. Michelle Lockley3,
  4. Elly Brockbank4,
  5. Asma Faruqi5,
  6. Ian Said5,6,
  7. Arjun Jeyarajah4,
  8. Rekha Wuntakal7,
  9. Blake Gilks8 and
  10. Naveena Singh5
  1. 1 Department of Medical Oncology, Barts Health NHS Trust, London, UK
  2. 2 Cancer Epidemiology and Prevention, British Columbia Cancer Agency, Vancouver, British Columbia, Canada
  3. 3 Barts Cancer Institute, Queen Mary University of London, London, UK
  4. 4 Department of Gynaecologial Oncology, Barts Health NHS Trust, London, UK
  5. 5 Department of Pathology, Barts Health NHS Trust, London, UK
  6. 6 Department of Histopathology, Mater Dei Hospital, Msida, Malta
  7. 7 Department of Gynaecological Oncology, Queens Hospital, Romford, UK
  8. 8 Department of Pathology and Laboratory Medicine, University of British Columbia, Vancouver, British Columbia, Canada
  1. Correspondence to Steffen Böhm, Medical Oncology, Barts Health NHS Trust, London EC1A 7BE, UK;{at}


Objective The Chemotherapy Response Scoring (CRS) system was developed to enable reproducible reporting of histologic tumor response in interval debulking specimens following neoadjuvant chemotherapy in advanced stage tubo-ovarian high-grade serous carcinoma. This prognostic biomarker has been included in ovarian cancer pathology reporting guidelines (International Collaboration on Cancer Reporting, College of American Pathologists) and in the upcoming European Society for Medical Oncology-European Society of Gynaecological Oncology (ESMO-ESGO) guidelines for ovarian cancer management. We present follow-up data on the CRS validation initiatives and suggest research with novel therapeutic agents incorporating this biomarker.

Methods The cohort on whom CRS was originally developed was analyzed after an extended follow-up of an additional 36 months. The CRS histopathologic scoring system was applied to omental sections obtained at interval surgery from all 80 patients. Progression-free and overall survival were re-calculated.

Results After a median follow-up of 4.3 years the CRS score predicted progression-free survival with an HR of 0.39 (95% CI 0.21 to 0.70), p = 0.002 adjusted for age, stage, and debulking status (median 1.08 vs 2.27 years for CRS1/2 vs CRS3). CRS was also predictive of overall survival with an HR of 0.17 (95% CI 0.07 to 0.44), p = 0.0002 adjusted for age, stage, and debulking status (median 2.55 vs 5.47 years for CRS1/2 vs CRS3).

Conclusion CRS3 is a reproducible prognostic biomarker for improved progression-free and overall survival in stage 3C or 4 tubo-ovarian high-grade serous carcinoma after neoadjuvant chemotherapy. The score, obtained at interval debulking surgery, can help facilitate research and biomarker driven first-line treatment of patients with advanced ovarian cancer.

  • CRS
  • biomarker
  • chemotherapy response
  • personalised therapy
  • ovarian cancer

Statistics from

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.


  • The Chemotherapy Response Score (CRS) assesses the histopathologic response to neoadjuvant chemotherapy in advanced high-grade serous tubo-ovarian carcinoma.

  • The role of CRS as a prognostic biomarker is acknowledged in the upcoming European Society for Medical Oncology-European Society of Gynaecological Oncology (ESGO-ESMO) Clinical Practice Guidelines.

  • The CRS has relevance for routine management, research and personalised first-line ovarian cancer treatment.


Until now, a measurable marker of tumor biology has not routinely been incorporated into first-line treatment decisions. Complete cytoreduction is an aim of primary and interval debulking surgery alike. However, we should want to know more about the disease in order to make a treatment decision for every patient with ovarian cancer. The so far unchallenged ‘one-size fits all’ chemotherapy concept should evolve in line with emerging scientific evidence. Neoadjuvant chemotherapy and standardized histopathological assessment of tissue after neoadjuvant chemotherapy might provide a valuable platform for rapid and reliable identification of medical treatment effects on an individual patient level.

The recently developed Chemotherapy Response Score (CRS) histopathologically assesses sensitivity to platinum-based therapy immediately after neoadjuvant chemotherapy.1 Assignment of CRS offers the potential of identifying patients with high-grade serous carcinoma for whom a durable long-term response to chemotherapy or cure is possible (CRS3) versus those who will inevitably experience a recurrence and die of disease (CRS1 or 2).1 In order to enter routine clinical practice, however, there must be extensive validation of CRS, in multiple independent centers and cohorts of patients. This will take time as there will be delays associated with accrual of sufficient follow-up data for cohorts identified prospectively.

In this article, we report observed median progression-free survival and overall survival data in the ‘test cohort’ of 60 patients from the original CRS study,1 in whom an additional 3 years of follow-up information was available, plus 20 patients who were not included in the original study due to insufficient follow-up at the time. We also report on ongoing validation projects and summarize how neoadjuvant chemotherapy and CRS could potentially be used as a biomarker platform for research and personalized first-line treatment of tubo-ovarian high-grade serous carcinoma.


All eligible patients were identified within the Barts Gynae Tissue Bank repository; this included the original test cohort cases from the study to develop CRS (N=60).1 This study cohort consists of 80 patients; 60 were included in the previous report,1 but appear in this report with additional follow-up, together with 20 patients not previously reported on. Histologically confirmed International Federation of Gynecology and Obstetrics (FIGO) stage IIIC or IV ovarian, fallopian tube, or primary peritoneal high-grade serous carcinoma and primary platinum-based chemotherapy before interval cytoreductive surgery were required as inclusion criteria. Patients were treated between 2009 and 2015. Fourteen patients received adjuvant bevacizumab (Table 1). Clinical data were collected through chart and tissue repository database review. Institutional review board approval was granted for collection of tissue and clinical information. Chemotherapy, surgery, and patient follow-up were performed as previously described.1 The diagnosis of high-grade serous carcinoma was confirmed on slide review by study pathologists in every case. The chemotherapy response score was histopathologically assigned by study pathologists, using previously published criteria.1 2 Briefly, the section of omentum showing the most residual viable tumor was assigned a score based on the response of the omental tumor to chemotherapy: score 1=abundant tumor with no or minimal perceptible response to chemotherapy; score 2=significant amounts of viable tumor present, showing readily appreciable fibro-inflammatory response secondary to treatment; score 3=complete or near complete response with no tumor or minimal irregularly scattered tumor nests (none >2 mm). For cases of the test cohort, the same score was used as in the initial study. Median (min; max) follow-up time of the entire cohort was calculated from the date of first neoadjuvant chemotherapy to the cut-off date in May 2017. Progression-free survival and overall survival were calculated using the previously published criteria; briefly, time to progression or death was measured from the date of first neoadjuvant chemotherapy.1 Time to death from tubo-ovarian high-grade serous carcinoma was considered for overall survival. Survival analysis was conducted by using Cox proportional hazards regression and the log-rank test to assess the prognostic value of the CRS, adjusting for age, stage, and debulking status. Survival functions were estimated by using the Kaplan-Meier method.

Table 1

Baseline characteristics of the cohort


Baseline characteristics

Baseline characteristics of the extended cohort of 80 patients, including the test cohort, were representative of the target population under routine care conditions and are presented in Table 1. Of the 80 cases, 72.5% (58/80) showed CRS2 and 27.5% (22/80) showed CRS3. No case was recorded as showing no/minimal response (CRS1).

Follow-up and observed event rate

The median follow-up was 4.3 years (min 2.5 years, max 7.7 years). At the data cut-off date, 84% progression-free survival events (67 events/80 patients) and 58% overall survival events (46 events/80 patients) had been observed, providing sufficiently mature data of the entire cohort for these endpoint analyses.

Progression-free and overall survival

Median progression-free survival was 13 months for patients with an intermediate chemotherapy response (CRS2) and 27 months for patients with a complete or near complete histopathological chemotherapy response (CRS3); HR 0.39 (95% CI 0.21 to 0.70), p=0.002 adjusted for age, stage, and debulking status.

Median overall survival was 31 months for patients with CRS2 and 66 months for patients with CRS3; HR 0.17 (95% CI 0.07 to 0.44), p=0.0002 adjusted for age, stage, and debulking status.

Between years 4 and >6 none of the patients with CRS3 had a disease progression event (plateau tail of the progression-free survival curve at approximately 30%) while for CRS2 patients the curve shows an apparent plateau at 4 years accounting for approximately 10% of cases. Overall survival curves begin to separate after 1 year and two of the CRS3 patients are alive and progression-free now for >6 and>7 years. In contrast, none of the patients with CRS2 in this cohort, so far, lived for more than 6 years after first-line treatment for advanced tubo-ovarian high-grade serous carcinoma. Kaplan-Meier curves for progression-free survival (A) and overall survival (B) are provided in Figure 1. Results were consistent with those seen when just considering the original ‘test cohort’ of n=60 (online supplementary data and Figure S1).

Supplemental material

Figure 1

(A) Progression-free survival and (B) overall survival of patients with a chemotherapy response score of CRS 2 (intermediate response) versus patients with a chemotherapy response score of CRS 3 (complete or near complete response) after neoadjuvant chemotherapy and interval debulking surgery. CRS, Chemotherapy Response Score.


We have previously published and validated the CRS, a system that stratifies the histopathological response to neoadjuvant chemotherapy at the time of interval debulking surgery.1 The improved progression-free survival and overall survival associated with CRS3, as reported in 2015, were confirmed in this extended cohort, after a median observation period of 4.3 years. It is important to note at the outset that no case was scored as CRS1 in this cohort. It has been previously found that separating CRS1 and CRS2 has low reproducibility and no significant prognostic value. In current practice the score is informative as a binary system, CRS1/CRS2 vs CRS3, until better methods emerge to separate prognostic categories within the CRS1/CRS2 group, using methods that provide more objective and clinically relevant categorization than by morphology alone. Patients with complete or near-complete response to neoadjuvant chemotherapy (CRS3) demonstrated statistically significant and clinically relevant longer median progression-free survival and overall survival compared with patients with intermediate response to neoadjuvant chemotherapy (CRS2). In addition to the differences in median progression-free survival and overall survival and the hazard ratios, two facts are noteworthy in this cohort: first, the flattened tail of the CRS3 progression-free survival curve after approximately 3.5 years, reflecting the long-term survivors being identified within the CRS3 cohort;and second, the proportion of >20% of patients within this tail—in other words, the relatively small number of patients with tubo-ovarian high-grade serous carcinoma who will be long-term survivors are captured within the CRS3 group.

A high degree of platinum sensitivity, represented by CRS3, seems to be a prerequisite for such long-term relapse-free survival. In this respect, CRS3 is a marker of a subpopulation of tubo-ovarian high-grade serous carcinoma with exceptionally favorable outcomes. At this point we do not know what the distinguishing molecular features of the CRS3 group are, compared with CRS1/2. It may prove to be characterized by enrichment for co-occurrence of mutations in the retinoblastoma gene (RB1) and homologous recombination pathways,3 and we await studies correlating molecular pathology with CRS.

For defining the value of CRS as a new biomarker further validation was requested by our group, the American Society of Clinical Oncology (ASCO) guideline, and the fifth Ovarian Cancer Consensus Conference.1 4 5 First and foremost, reproducibility is key for a robust biomarker. We have shown that the system can be reliably assessed by pathologists with different levels of expertise after online training via the training module, consisting of scanned slides.6 7 Second, independent external validation of the prognostic value is necessary. Indeed, several independent groups have reported their findings and confirmed the prognostic significance of the CRS system for progression-free survival.8–11 To further strengthen this validation work additional initiatives, including an international, multi-center individual patient data meta-analysis, are underway.

This study has limitations that should be acknowledged. The data were retrospectively collected and comprise a relatively small patient cohort from a single center. The cohort is not a completely independent set of patients from what has been previously published, and there is significant overlap with our previous report.1 Strengths include the assignment of the CRS score by the pathologists who devised the system, the relatively long follow-up available, and that it reflects the experience with patients treated in routine clinical practice.

CRS could serve as the first reliable prognostic system of measuring tumor response in metastases at the tissue level. Uptake of CRS in the pathology guidelines of the International Collaboration on Cancer Reporting (ICCR) and College of American Pathologists (CAP) and in the upcoming European Society for Medical Oncology-European Society of Gynecological Oncology (ESMO-ESGO) Clinical Practice Guideline on ovarian cancer will give us more data and insights into the performance of CRS in the coming years.


  1. 1.
  2. 2.
  3. 3.
  4. 4.
  5. 5.
  6. 6.
  7. 7.
  8. 8.
  9. 9.
  10. 10.
  11. 11.


  • BG and NS contributed equally.

  • Funding The Barts Gynae Tissue Bank was supported by Barts and The London Charity.

  • Disclaimer The content of this paper reflects the opinion of the authors.

  • Competing interests None declared.

  • Patient consent Patient consent was obtained.

  • Ethics approval Institutional review board approval was granted for collection of tissue and clinical information.

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