Abstract
Objective:
To determine the clinical characteristics of patients who attained pathologic complete response (pCR) after neoadjuvant chemotherapy (NACT) and to identify specific predictive or prognostic factors associated with pCR.
Methods:
Two distinct populations of patients who underwent NACT followed by interval tumor reductive surgery (TRS) were used in this retrospective study. The first contained 472 patients from a single institution. The second contained only pCR patients (67); those identified from population one, plus 44 obtained through collaborative institutions. Cox analysis and log-rank tests were performed to assess associations between clinical characteristics and pCR outcome, recurrence-free survival (RFS), and overall survival (OS).
Results:
The median RFS and OS in our pCR-only population was 24.2 and 80.8 months, respectively, with a median follow-up time of 32.4 months. In our single institution population, 23 patients attained pCR (4.9%) and had longer RFS compared to non-pCR patients with viable microscopic, optimal, or suboptimal residual disease (24.3 vs. 12.1 vs. 11.6 vs. 9.6 months, p = 0.025, 0.012, 0.008, respectively), and longer OS compared to those with optimal or suboptimal residual disease (54.5 vs. 29.4 vs. 25.7 months, p = 0.027, 0.007, respectively). Patients were more than three-fold likely to attain pCR if their CA125 value was normal at the time of surgery (OR 3.54, 95% CI:1.14 – 11.05, p = 0.029).
Conclusions:
Women with pCR after NACT have significantly longer RFS compared to those with residual viable tumor at the time of interval tumor-reductive surgery, and CA125 is plausible biomarker for identifying these extreme responders preoperatively.
Keywords: CA125, debulking surgery, interval cytoreductive surgery, neoadjuvant chemotherapy, ovarian cancer, pathologic complete response, predictive factors, survival outcomes, tumor-reductive surgery
Introduction
Ovarian cancer is the deadliest gynecologic malignancy, with most cases diagnosed at an advanced stage and with high-grade serous histology.[1, 2] Treatment consists of primary cytoreductive surgery with the intent of obtaining microscopic or ≤1 cm residual disease (optimal), followed by adjuvant platinum-based chemotherapy. Since the early 2000s, however, utilization of neoadjuvant chemotherapy (NACT) in ovarian cancer has steadily increased.[3] One study of 1,538 women treated at six NCI-designated cancer centers reported that use of NACT between 2003 and 2011 increased from 16% to 34% in women with stage IIIC disease and from 41% to 62% in women with stage IV disease.[4] One major reason for this increase is that prior to the late-1990s, upfront surgical intervention was considered the standard of care, regardless of the surgical outcome. More recently, four randomized controlled trials showed no difference in progression-free survival (PFS) or overall survival (OS) when NACT was compared with primary tumor-reductive surgery (TRS).[5-8] Although debate still remains regarding the optimal order of treatment for advanced stage ovarian cancer, these studies cemented NACT as a viable option for patients with unresectable disease, medical comorbidities that preclude aggressive surgery, or poor performance status.
Physicians who treat patients with chemotherapy before surgery have the unique opportunity of measuring response at the histologic level rather than through only biomarkers or imaging. Pathologic complete response (pCR) is defined as the absence of residual invasive cancer on histologic evaluation of all surgical specimens following completion of NACT. In breast cancer, pCR has been extensively studied, with multiple clinical trials showing a significant benefit in PFS and OS in women who attain pCR versus those who have residual cancer in their surgical specimens.[9] Interestingly, it appears that patients with triple-negative breast cancer (TNBC) have the largest survival benefit if pCR is attained.[9] Given the known molecular similarities between TNBC and high-grade serous ovarian cancer,[10] and the increased incidence of pCR in TNBC when treated with platinum agents compared to other breast cancer subtypes,[11, 12] we hypothesized that patients with ovarian cancer who attained pCR would also achieve a survival benefit.
The objectives of our retrospective study were threefold. First, we aimed to define the frequency of pCR in a large single-institution cohort of women who underwent NACT followed by interval TRS for advanced-stage ovarian cancer. Second, we used this cohort to determine whether any preoperative clinical characteristics predicted pCR, as well as whether pCR attainment led to improved survival outcomes. Lastly, we sought to more explicitly describe the clinical characteristics associated with a larger cohort of pCR cases through an international collaboration between five institutions.
Materials and Methods
We developed two distinct analytical populations of patients to achieve our stated objectives. The first population was used to identify the frequency of pCR, assess for any survival advantages, and determine whether any preoperative factors were predictive of attaining pCR. This population, from one institution, comprised all patients with suspected advanced-stage ovarian cancer of any histology who received NACT followed by interval TRS between June 2004 and September 2018. Our second population was used to further identify the clinical characteristics of a large group of pCR-only patients and comprised all pCR cases identified over designated time periods from four collaborating institutions, in addition to the pCR patients from the first population. Institutional Review Board approval was obtained at all institutions.
All patients underwent at least 1 cycle of NACT followed by maximal-effort interval TRS with intent to remove all visible disease. All patients from the single-institution population received adjuvant chemotherapy after interval TRS, regardless of surgical outcome or pathology. Institutional Review Board approval was attained at all participating sites. All patient data were recorded in a password-protected database that included information regarding age at diagnosis, stage of disease, race, BRCA germline mutation status, NACT regimen and number of cycles administered, baseline and individual cycle CA125 and platelet values (when available), dates of surgery and recurrence (if applicable), and disease status at last clinic visit or contact.
Institutional Specific Population Contributions
From MD Anderson Cancer Center, 473 patients received NACT and underwent interval TRS between June 2004 and September 2018, with 23 cases of pCR identified. From Oklahoma Health Sciences Center, 15 pCR cases were identified between May 2009 and January 2018. From University of Southern California, 3 pCR cases were identified between August 2005 and August 2015. From Saitama Medical University International Medical Center (Japan), 11 pCR cases were identified between October 2009 and November 2014. From A.C. Camargo Cancer Center (Brazil), 15 pCR cases were identified between June 2007 and March 2018.
Statistical Analysis
Summary statistics were used to describe the demographic and clinical characteristics of the pCR-only population. Demographic and clinical characteristics of the MD Anderson study population were summarized and compared by pCR and non-pCR groups. Recurrence-free survival (RFS) was estimated using the Kaplan–Meier product-limit estimator and modeled via Cox proportional hazards regression. RFS was measured from the date of surgery to the earliest date of last clinic visit, date of first recurrence, or date of death. Subjects who were recurrence free and alive were censored at date of last clinic visit. Overall survival was measured from the date of surgery to the earliest date of death or date of last contact. Statistical significance was defined at the 0.05 level. All statistical analyses were performed using Stata/MP v16.0 (College Station, Texas).
Results
A total of 67 patients across the five institutions were identified as having attained pCR after NACT (Table 1). The median age of these patients was 62 years, with 63% having stage III and 37% having stage IV disease. Nearly all patients had serous histology (94%). The median number of NACT cycles prior to interval TRS was 4 (range 3-9), and the most common NACT regimen administered at cycle 1 was dose-dense taxane with carboplatin every 3-4 (q3-4) weeks (36%). All patients in this population underwent interval cytoreductive surgery to no gross residual disease (R0), as per the definition of pCR. Most patients identified as white (80%), and 39% of tested patients had either a BRCA1 or BRCA2 germline mutation. The median baseline CA125 value was 741 U/mL, with a subsequent drop to 14 U/mL prior to undergoing interval TRS. The median follow-up time was 32.4 months, and the median RFS and OS were 24.2 and 80.8 months, respectively. At the time of data analysis, 51% (34/67) of patients remained in remission, and 53% were alive with no evidence of disease.
TABLE 1.
Demographic and clinical characteristics of the pCR-only cohort (N=67)
| Characteristic | |
|---|---|
| Institution, N (%) | |
| MD Anderson | 23 (34.3) |
| ACCCC (Brazil) | 15 (22.4) |
| OUHSC | 15 (22.4) |
| SMU (Japan) | 11 (16.4) |
| USC | 3 (4.5) |
| Age at diagnosis (years) | |
| N | 67 |
| Mean (SD) | 61.3 (9.0) |
| Median (Min-Max) | 62.0 (41.0 - 84.0) |
| Race, N (% of known) | |
| White | 52 (80.0) |
| Asian | 12 (18.5) |
| American Indian/Alaskan Native | 1 (1.5) |
| Unknown | 2 |
| BRCA mutation status, N (% of known) | |
| No mutation | 23 (60.5) |
| BRCA1 | 4 (10.5) |
| BRCA2 | 5 (13.2) |
| BRCA1/2 not specified | 6 (15.8) |
| Unknown | 29 |
| Primary disease site, N (%) | |
| Ovary | 59 (88.1) |
| Peritoneum | 7 (10.4) |
| Fallopian tube | 1 (1.5) |
| Stage, N (%) | |
| III | 42 (62.7) |
| IV | 25 (37.3) |
| Histology, N (%) | |
| Serous | 63 (94.0) |
| Clear cell | 1 (1.5) |
| Adenocarcinoma NOS | 1 (1.5) |
| Mixed | 1 (1.5) |
| Neuroendocrine | 1 (1.5) |
| # of NACT cycles | |
| N | 67 |
| Mean (SD) | 4.5 (1.7) |
| Median (Min-Max) | 4 (3 - 9) |
| Treatment at 1st cycle of NACT, N (%) | |
| Weekly taxane / q3-4 week IV carboplatin | 24 (35.8) |
| Weekly taxane / q3-4 week IP carboplatin | 6 (8.9) |
| q3 week taxane / q3-4 week IV carboplatin | 13 (19.4) |
| Taxane / platinum NOS | 18 (26.9) |
| Taxane / platinum / bevacizumab NOS | 4 (6.0) |
| Platinum NOS | 1 (1.5) |
| Other | 1 (1.5) |
| Baseline CA125 value (U/mL) | |
| N | 62 |
| Mean (SD) | 2026.4 (2869.4) |
| Median (Min-Max) | 741.0 (9.7 - 14561.0) |
| CA125 value at interval TRS (U/mL) | |
| N | 37 |
| Mean (SD) | 35.9 (88.1) |
| Median (Min-Max) | 14.3 (5.9 - 540.0) |
| Baseline platelets (103/μL) | |
| N | 50 |
| Mean (SD) | 363.2 (115.1) |
| Median (Min-Max) | 356 (164 - 704) |
| Platelets at interval TRS (103/μL) | |
| N | 23 |
| Mean (SD) | 221.7 (138.4) |
| Median (Min-Max) | 195 (57 - 640) |
ACCCC, A.C. Camargo Cancer Center; IV, intravenous; IP, intraperitoneal; NACT, neoadjuvant chemotherapy; NOS, not otherwise specified; OUHSC, University of Oklahoma Health Sciences Center; q, every; SD, standard deviation; SMU, Saitama Medical University International Medical Center; TRS, tumor-reductive surgery; USC, University of Southern California.
To identify differences between patients with pCR and non-pCR, we next analyzed our single-institution only population containing 472 patients within the defined time period who received NACT followed by interval surgery. We identified 23 patients who attained pCR at interval-TRS, and 449 patients who had viable cancer cells in their pathologic specimens, resulting in a 4.9% rate of pCR (95% CI: 3.1 – 7.2). The median follow-up within this single-institution population was 24.1 months (range 0.4 – 147.8 months), and the median RFS and OS were 12.1 and 41.7 months, respectively. The median RFS was significantly longer in patients with pCR compared to those without (24.3 vs. 11.9 months, p = 0.013, Figure 1a). The median OS was 54.5 months in the pCR group compared to 39.9 months in the non-pCR group (p = 0.150, Figure 1b). Stratifying non-pCR patients by residual disease status revealed that pCR patients had a longer RFS than non-pCR patients who had either microscopic (R0, 24.3 vs. 12.1 months (p = 0.025)), optimal (≤ 1 cm, 11.6 months (p = 0.012)), or suboptimal (> 1 cm, 9.6 months (p = 0.008), Figure 1C) residual disease at interval-TRS. Patients with pCR had a longer OS than those without pCR with optimal or suboptimal residual disease at interval-TRS (54.5 vs. 29.4 months (p = 0.027) and 25.7 months (p = 0.007), respectively; Figure 1d). At the time of their last recorded clinic visits, 68.2% of patients with pCR had no evidence of disease, compared to only 20.9% of patients without pCR.
Figure 1. Survival outcomes based on pCR.
(a, b) Recurrence-free (a) and overall (b) survival patients with pCR versus those without pCR. (c, d) Recurrence-free (c) and overall (d) survival among patients with pCR versus non-pCR with microscopic (R0), optimal (≤ 1 cm), or suboptimal (> 1 cm) residual disease.
There were no significant differences in baseline age, race, primary disease site, histology, or BRCA mutation status between the pCR and non-pCR groups within our single-institution population (Table 2). In both groups, a BRCA germline mutation was present in approximately 25% of patients. Although no significant difference existed in baseline CA125 levels between the two groups, patients who attained pCR had a significantly lower median CA125 level prior to interval-TRS than those without pCR (13.8 vs. 32.0 U/mL, p < 0.001). There was no difference in baseline or interval-TRS platelet values between the groups. Optimal or R0 resection was achieved in 90.9% of the patients in the non-pCR group, and 69.6% had R0 resection. Both groups had a median number of 3 NACT cycles prior to surgery; however, the specific NACT regimen administered at cycle 1 significantly differed, with weekly taxane plus q3-4 week carboplatin being more common in the pCR vs. non-pCR group (47.8% vs. 25.6%, p = 0.006). We found a significant difference in the number of NACT cycles administered between the pCR and non-pCR groups (p = 0.029), with the pCR group having a higher proportion of patients receiving ≥7 cycles compared to the non-pCR group (17.4 vs. 3.8%, Table 2). Furthermore, patients who received ≥7 cycles of NACT were more than 5 times likely to attain pCR than those who received between 1-3 cycles NACT (OR 5.18, 95% CI: 1.51 – 17.77, p = 0.009, Table 3).
TABLE 2.
Demographic and clinical characteristics of the MD Anderson only population (N=472)
| Characteristic | non-pCR (N = 449) |
pCR (N = 23) |
p- vaue |
|---|---|---|---|
| Age at diagnosis, years | 0.754 | ||
| N | 449 | 23 | |
| Mean (SD) | 62.1 (10.6) | 63.1 (10.5) | |
| Median (Min-Max) | 63 (21 - 85) | 63 (44 - 84) | |
| Race, N (% of known) | 0.629 | ||
| White | 378 (86.5) | 20 (95.2) | |
| African American | 31 (7.1) | 0 (0) | |
| Asian | 26 (5.9) | 1 (4.8) | |
| American Indian/Alaskan Native | 2 (0.5) | 0 (0) | |
| Unknown | 12 | 2 | |
| Primary disease site, N (%) | >0.999 | ||
| Ovary / fallopian tube / peritoneum | 441 (98.2) | 23 (100) | |
| Mullerian NOS | 8 (1.8) | 0 (0) | |
| Stage, N (%) | 0.383 | ||
| III | 237 (52.8) | 10 (43.5) | |
| IV | 212 (47.2) | 13 (56.5) | |
| Histology, N (%) | 0.710 | ||
| Serous | 411 (91.5) | 22 (95.7) | |
| Non-serous | 38 (8.5) | 1 (4.3) | |
| BRCA mutation status, N (% of known) | 0.816 | ||
| No mutation | 204 (76.4) | 12 (75) | |
| BRCA1 | 35 (13.1) | 2 (12.5) | |
| BRCA2 | 21 (7.9) | 2 (12.5) | |
| Variance of unknown significance | 7 (2.6) | 0 (0) | |
| Unknown | 182 | 7 | |
| Baseline CA125 value (U/mL) | 0.101 | ||
| N | 434 | 23 | |
| Mean (SD) | 2543.3 (6824.2) | 2100.3 (3672.9) | |
| Median (Min-Max) | 946.1 (16.0 - 111578.0) | 620.4 (9.7 - 14561.0) | |
| CA125 value at interval TRS (U/mL) | <0.001 | ||
| N | 143 | 21 | |
| Mean (SD) | 97.8 (259.7) | 23.0 (23.1) | |
| Median (Min-Max) | 32.0 (6.0 - 2326.0) | 13.8(5.9 – 94.0) | |
| Baseline platelets (103/μL) | 0.273 | ||
| N | 396 | 23 | |
| Mean (SD) | 402.9 (139.8) | 359.8 (111.4) | |
| Median (Min-Max) | 380 (7 - 928) | 367 (164 - 618) | |
| Platelets at interval TRS (103/μL) | 0.121 | ||
| N | 143 | 22 | |
| Mean (SD) | 228.2 (76.4) | 219.5 (141.2) | |
| Median (Min-Max) | 225 (11 - 498) | 193 (57 - 640) | |
| # of NACT cycles | 0.283 | ||
| N | 449 | 23 | |
| Mean (SD) | 4.0 (1.5) | 4.4 (1.9) | |
| Median (Min-Max) | 3 (1 - 12) | 3 (3 - 9) | |
| Specific of # NACT cycles, N (%) | 0.029 | ||
| 1, 2, or 3 cycles | 264 (58.8) | 12 (52.2) | |
| 4, 5, or 6 cycles | 168 (37.4) | 7 (30.4) | |
| ≥ 7 cycles | 17 (3.8) | 4 (17.4) | |
| Treatment at 1st cycle of NACT, N (%) | 0.006 | ||
| Weekly taxane / q3-4 week IV carboplatin | 115 (25.6) | 11 (47.8) | |
| q3 week taxane / q3-4 week IV carboplatin | 147 (32.7) | 4 (17.3) | |
| Taxane / platinum NOS | 175 (39.0) | 5 (21.7) | |
| Taxane / platinum / bevacizumab NOS | 7 (1.6) | 1 (4.4) | |
| Taxane NOS / bevacizumab NOS | 1 (0.2) | 0 (0) | |
| Platinum NOS | 2 (0.5) | 1 (4.4) | |
| Taxane NOS | 1 (0.2) | 0 (0) | |
| Other | 1 (0.2) | 1 (4.4) | |
| Residual disease at interval TRS, N (% of known) | 0.003 | ||
| Microscopic (R0) | 305 (69.7) | 23 (100) | |
| Optimal (≤ 1 cm) | 93 (21.2) | 0 (0) | |
| Suboptimal (> 1 cm) | 40 (9.1) | 0 (0) | |
| Unknown | 11 | 0 |
IV, intravenous; SD, standard deviation; TRS, tumor-reductive surgery.
TABLE 3.
Demographic and clinical prognostic factors for attaining pCR (MD Anderson only population)
| OR | 95% CI | p-value | |
|---|---|---|---|
| Race | |||
| Other | 1.00 | 1.00 - 1.00 | |
| White | 3.12 | 0.41 – 23.70 | 0.271 |
| Stage | |||
| III | 1.00 | 1.00 - 1.00 | |
| IV | 1.45 | 0.62 – 3.38 | 0.386 |
| Histology | |||
| Serous | 1.00 | 1.00 - 1.00 | |
| Non-serous | 0.49 | 0.06 - 3.75 | 0.493 |
| BRCA mutation status | |||
| No mutation | 1.00 | 1.00 - 1.00 | |
| BRCA1 | 0.97 | 0.21 - 4.53 | 0.971 |
| BRCA2 | 1.62 | 0.34 - 7.73 | 0.546 |
| Variance of unknown significance | 1.00 | 1.00 - 1.00 | |
| Total # of NACT cycles | 1.18 | 0.92 - 1.49 | 0.187 |
| Specific # of NACT cycles | |||
| 1, 2, or 3 cycles | 1.00 | 1.00 – 1.00 | |
| 4, 5, or 6 cycles | 0.92 | 0.35 – 2.38 | 0.858 |
| ≥ 7 cycles | 5.18 | 1.51 – 17.77 | 0.009 |
| Cycle 1 chemotherapy regimen | |||
| q3 week taxane / q3-4 week IV carboplatin | 1.00 | 1.00 - 1.00 | |
| Weekly taxane / q3-4 week IV carboplatin | 3.52 | 1.09 – 11.33 | 0.035 |
| Platinum NOS / taxane NOS | 1.05 | 0.28 - 3.98 | 0.943 |
| Baseline CA125 value (U/mL) | |||
| > 35 | 1.00 | 1.00 - 1.00 | |
| ≤ 35 | 33.33 | 6.25 – 100 | < 0.001 |
| CA125 value at interval TRS (cutoff of 35 U/mL) | |||
| > 35 | 1.00 | 1.00 - 1.00 | |
| ≤ 35 | 3.57 | 1.14 – 11.11 | 0.029 |
| CA125 value at interval TRS (cutoff of 14 U/mL) | |||
| > 14 | 1.00 | 1.00 - 1.00 | |
| ≤ 14 | 3.70 | 1.43 – 9.09 | 0.007 |
| Baseline platelets (103/μL) | |||
| > 400 | 1.00 | 1.00 - 1.00 | |
| ≤ 400 | 1.20 | 0.51 – 2.86 | 0.669 |
| Platelets at interval TRS (103/μL) | |||
| > 400 | 1.00 | 1.00 - 1.00 | |
| ≤ 400 | 0.36 | 0.07 – 2.00 | 0.243 |
CI, confidence interval; OR, odds ratio; TRS, tumor-reductive surgery.
We next determined whether any clinical or demographic characteristics were associated with attaining pCR or with survival, using our single-institution population. Race, stage, histology, BRCA mutation status, baseline and interval-TRS platelet levels, and total number of NACT cycles received were not associated with attaining pCR (Table 3). Interestingly, compared to patients who received paclitaxel every 3 weeks, those who received weekly paclitaxel (“dose-dense”) had more than three times the odds of attaining pCR (odds ratio [OR] 3.52, 95% confidence interval [CI]: 1.09 – 11.33, p = 0.035). Although rare in our dataset and also in clinical practice, patients who had a baseline CA125 of ≤ 35 U/mL (n= 7/457, 1.5%) had a significantly higher chance of attaining pCR (OR 30.25, 95% CI: 6.32 – 144.78, p < 0.001). Of these 7 patients, 6 had tumors of serous histology and 1 had carcinosarcoma. In addition, we found that patients whose CA125 had normalized at the time of interval TRS (n= 95/164, 57.9%) (≤ 35 U/mL), or fallen to ≤ 14 U/mL (n= 44/164, 26.8%) (the median interval TRS CA125 value in our pCR-only cohort), had a significantly higher chance of attaining pCR than those who did not (OR 3.54, 95% CI: 1.14 – 11.05, p = 0.029; OR 3.67, 95% CI: 1.43 – 9.39, p = 0.007; respectively). As expected, both BRCA mutation status and baseline platelet level were associated with improved survival (Tables 4 and 5). Specifically, patients with either a BRCA1 or BRCA2 mutation had a significantly improved median RFS compared to wild-type (17.9 and 12.7 months vs. 10.9 months, respectively; log rank = 0.042); however no difference in OS was seen. Patients with a baseline platelet value of ≤ 400 x103/μL had a significantly longer OS than those with baseline platelets > 400 x103/μL (46.4 vs 37.2 months, p = 0.003) and a marginal increase in RFS (12.9 vs. 11.1 months, p = 0.054). No association between baseline or interval-TRS CA125 values and RFS or OS were identified.
TABLE 4.
Demographic and clinical prognostic factors for recurrence-free survival (MD Anderson only population)
| N | Median (months) |
Log- rank |
HR | 95% CI | p- value |
|
|---|---|---|---|---|---|---|
| Race | 0.326 | |||||
| Other | 59 | 10.3 | 1.00 | 1.00 - 1.00 | ||
| White | 384 | 12.2 | 0.86 | 0.64 - 1.16 | 0.327 | |
| Stage | 0.221 | |||||
| III | 237 | 12.5 | 1.00 | 1.00 - 1.00 | ||
| IV | 220 | 11.5 | 1.14 | 0.93 - 1.39 | 0.222 | |
| Histology | 0.241 | |||||
| Serous | 420 | 12.4 | 1.00 | 1.00 - 1.00 | ||
| Non-serous | 37 | 9.7 | 1.24 | 0.87 - 1.77 | 0.243 | |
| BRCA mutation status | 0.042 | |||||
| No mutation | 210 | 10.9 | 1.00 | 1.00 - 1.00 | ||
| BRCA1 | 37 | 17.9 | 0.64 | 0.42 - 0.97 | 0.035 | |
| BRCA2 | 23 | 12.7 | 0.59 | 0.36 - 0.97 | 0.039 | |
| Variance of unknown significance | 7 | 13.1 | 0.78 | 0.34 - 1.76 | 0.544 | |
| Remaining tumor size | 0.045 | |||||
| pCR | 23 | 24.3 | 1.00 | 1.00 - 1.00 | ||
| Non-pCR microscopic (R0) | 297 | 12.1 | 1.95 | 1.09 – 3.50 | 0.025 | |
| Non-pCR optimal (≤ 1 cm) | 90 | 11.6 | 2.18 | 1.19 – 4.01 | 0.012 | |
| Non-pCR sub-optimal (> 1 cm) | 38 | 9.6 | 2.45 | 1.26 – 4.79 | 0.008 | |
| Baseline CA125 value (U/mL) | 0.064 | |||||
| > 35 | 437 | 12.1 | 1.00 | 1.00 - 1.00 | ||
| ≤ 35 | 7 | 27.9 | 0.40 | 0.15 - 1.09 | 0.073 | |
| CA125 value at interval TRS (cutoff of 35 U/mL) | 0.786 | |||||
| > 35 | 67 | 14.2 | 1.00 | 1.00 - 1.00 | ||
| ≤ 35 | 90 | 14.1 | 1.05 | 0.75 - 1.46 | 0.787 | |
| Baseline platelets (103/μL) | 0.054 | |||||
| > 400 | 181 | 11.1 | 1.00 | 1.00 - 1.00 | ||
| ≤ 400 | 227 | 12.9 | 0.81 | 0.65 - 1.00 | 0.055 | |
| Platelets at interval TRS (103/μL) | 0.095 | |||||
| > 400 | 6 | 8.1 | 1.00 | 1.00 - 1.00 | ||
| ≤ 400 | 152 | 14.1 | 0.47 | 0.19 - 1.16 | 0.103 |
CI, confidence interval; HR, hazard ratio; TRS, tumor-reductive surgery.
TABLE 5.
Demographic and clinical prognostic factors for overall survival (MD Anderson only population)
| N | Median (months) |
Log- rank |
HR | 95% CI | p- value |
|
|---|---|---|---|---|---|---|
| Race | 0.978 | |||||
| Other | 60 | 41.1 | 1 | 1.00 - 1.00 | ||
| White | 396 | 41.7 | 0.99 | 0.67 - 1.48 | 0.978 | |
| Stage | 0.567 | |||||
| III | 245 | 41.1 | 1.00 | 1.00 - 1.00 | ||
| IV | 225 | 41.7 | 0.92 | 0.70 - 1.21 | 0.567 | |
| Histology | 0.064 | |||||
| Serous | 432 | 41.7 | 1.00 | 1.00 - 1.00 | ||
| Non-serous | 38 | 41.7 | 1.53 | 0.97 - 2.40 | 0.066 | |
| BRCA mutation status | 0.205 | |||||
| No mutation | 214 | 45.7 | 1.00 | 1.00 - 1.00 | ||
| BRCA1 | 37 | 57.8 | 0.62 | 0.33 - 1.17 | 0.139 | |
| BRCA2 | 23 | 60.9 | 0.57 | 0.26 - 1.24 | 0.158 | |
| Variance of unknown significance | 7 | NE | 0.41 | 0.06 - 2.96 | 0.377 | |
| Remaining tumor size | < 0.001 | |||||
| pCR | 23 | 54.5 | 1.00 | 1.00 - 1.00 | ||
| Non-pCR microscopic (R0) | 303 | 44.2 | 1.48 | 0.65 – 3.35 | 0.352 | |
| Non-pCR optimal (≤ 1 cm) | 93 | 29.4 | 2.58 | 1.12 – 5.95 | 0.027 | |
| Non-pCR sub-optimal (> 1 cm) | 40 | 25.7 | 3.41 | 1.39 – 8.36 | 0.007 | |
| Baseline CA125 value | 0.341 | |||||
| > 35 | 448 | 41.7 | 1.00 | 1.00 - 1.00 | ||
| ≤ 35 | 7 | 100.6 | 0.58 | 0.18 - 1.81 | 0.348 | |
| CA125 value at interval TRS | 0.878 | |||||
| > 35 | 69 | 39.9 | 1.00 | 1.00 - 1.00 | ||
| ≤ 35 | 95 | 42.8 | 0.97 | 0.66 - 1.42 | 0.878 | |
| Baseline platelets (103/μL) | 0.003 | |||||
| > 400 | 182 | 37.2 | 1.00 | 1.00 - 1.00 | ||
| ≤ 400 | 235 | 46.4 | 0.64 | 0.48 - 0.86 | 0.003 | |
| Platelets at interval TRS (103/μL) | 0.706 | |||||
| > 400 | 7 | 42.8 | 1.00 | 1.00 - 1.00 | ||
| ≤ 400 | 158 | 39.7 | 1.19 | 0.48 - 2.93 | 0.706 |
CI, confidence interval; HR, hazard ratio; NE, not evaluable; TRS, tumor-reductive surgery.
To further explore the trends in CA125 values during NACT, we next identified how many patients had a “normalized” CA125 value after 3 or 4 cycles of chemotherapy. Using our single-institution cohort, we identified all patients who had both a baseline CA125 of > 35 U/mL and any recorded CA125 value after 3 or 4 cycles of chemotherapy. We found that of 135 patients who had received 3 cycles, 50 (37.0%) had normal CA125 values (≤ 35 U/mL) while 85 (63%) remained elevated (>35 U/mL). Of the 50 patients with normal CA125 values after 3 cycles, 10% (5/50) attained pCR, compared to only 2.4% (2/85) of those whose CA125 remained elevated. Similarly examining patients who received 4 cycles of NACT, we again found that the proportion of patients with pCR was higher in those with normalized CA125 values compared to those without (14.3% (4/28) vs. 4.7% (2/43)). Next, we sought to determine if there was an enrichment in CA125 “normalizers” in our multi-institution, pCR-only population. We found that after 3 cycles of NACT, 75% (15/20) of pCR patients had normalized CA125 values, and further, after 4 cycles of NACT, 75% (9/12) of pCR patients had normalized values.
Discussion
In our multi-institutional pCR-only population, the median OS was 81 months, which is substantially longer than typically seen for patients with advanced-stage ovarian cancer, even those who undergo upfront optimal cytoreduction (median OS 32 – 66 months).[13-15] Notably, we found that 39% of patients with pCR tested in this population had a germline BRCA mutation, which is higher than the established prevalence of germline BRCA mutations in all ovarian cancer patients (10-20%).[16-18] No data was available regarding other possible germline mutations or somatic BRCA mutations which may have also contributed to the increased survival. We also found that the median CA125 value prior to interval-TRS was 14 U/mL, which prompted us to examine CA125 trends in our single-institution matched population.
In our single-institution population, we observed that 4.9% (23/472) of patients attained pCR and experienced a significant RFS benefit compared to those who did not. This is in line with previous reports which have also identified an RFS benefit in pCR patients, and cite pCR rates between 6.5 and 19.4%, although notably with a lower inclusion of patients (91 to 322).[19-22] We similarly found that pCR patients have an improved OS compared to non-pCR patients with optimal (≤1 cm) or suboptimal (>1 cm) residual disease at interval-TRS.[19, 20]
Patients with pCR also more commonly received dose-dense/weekly taxane compared to non-pCR patients (47.8 vs. 25.6%), with univariate analysis confirming a greater likelihood of attaining pCR when weekly taxane vs. every-3-week dosing was used. Although previous trials have shown PFS and OS advantage with dose-dense taxane,[23] the more recent ICON8 clinical trial, which included patients receiving NACT, showed no PFS difference between the dose-dense and every-3-week regimens.[24]
Examining CA125 trends in our single-institution population revealed that patients with pCR had a significantly lower CA125 value at the time of interval-TRS, and patients with values ≤ 35 U/mL at the time of surgery had more than a threefold increase in the odds of attaining pCR (OR 3.57, p = 0.029). CA125 level may serve as a particularly useful biomarker in predicting pCR, as not only were patients with a normal (≤35 U/mL) pre-surgery CA125 value more likely to attain pCR, but as the CA125 value decreased, the likelihood of pCR increased. When we applied the median CA125 value at interval TRS for our pCR-only group (14 U/mL) to our single-institution population, we found an even stronger association with attaining pCR (OR 3.70, p = 0.007).
The number of NACT cycles administered in the single-institution population also revealed interesting findings. There was a higher proportion of patients in the pCR group who received ≥7 cycles of NACT compared to the non-pCR group, and a more than 5-fold increase in the chance of attaining pCR in those who underwent ≥7 cycles of NACT compared to 1-3 cycles. While the main reasons for the use of >6 cycles of NACT included persistently elevated CA125, lack of response on interval imaging, or continued medical comorbidities precluding aggressive surgical debulking, these findings are certainly provocative and raise the question of the ideal amount of NACT cycles to give patients. As discussed above, patients whose CA125 normalized to ≤35 had a more than 3-fold increase in the chance of attaining pCR. Although this data is observational and should be interpreted with caution, barring chemotherapy-related toxicities, consideration could be given to administering NACT until CA125 normalization.
Additionally, as patients who have lower pre-surgical CA125 values have a higher likelihood of attaining pCR, they may be more suitable candidates for a laparoscopic approach to interval TRS. The Laparoscopic cytoreduction After Neoadjuvant ChEmotherapy (LANCE) trial will examine whether minimally invasive surgery is non-inferior to laparotomy in terms of RFS in patients with advanced-stage ovarian cancer who had partial or complete response to NACT and a normalization of their CA125 value. Our study demonstrates that normalization of CA125 values prior to interval TRS, especially when occurring after 3 or 4 cycles of NACT, leads to the highest likelihood of attaining pCR.
While our data was consistent with previous reports demonstrating improved OS in ovarian cancer patients with a normal baseline platelet level (defined as platelets > 400 x103/μL) compared to those with baseline thrombocytosis [25, 26], we did not identify any associations with attaining pCR based on baseline or pre-surgery platelet values. Although there was an enrichment of patients with BRCA gene mutations in our collaborative, pCR-only population; in our single-institution comparison population, we found no difference in the proportion of patients with BRCA mutations between those with pCR or without. Furthermore, in our logistic regression analysis, we did not find any association between BRCA positivity and the attainment of pCR, however, this may be limited due to the relatively small number of patients with pCR in this population (n=23).
While pCR is a relatively rare occurrence, it could potentially be used to identify ideal candidates for maintenance drug trials. Even in patients who undergo interval-TRS to no gross residual, suspicion is high that viable cancer cells remain in the patient, particularly when their pathologic specimens have viable tumor present. Concordantly, we found that patients with pCR had a significantly improved RFS when compared to non-pCR patients who underwent an R0 interval-TRS, suggesting that patients with pCR seem more likely to be truly cancer free as opposed to just cytoreduced. These patients would allow researchers to specifically measure the effectiveness of agents in preventing relapse, as opposed to controlling the growth of the residual viable cancer cells present.
A major strength of this study was the large number of patients with pCR analyzed from our multi-institutional international collaboration, allowing us to describe the characteristics and treatment regimens of 67 patients who attained pCR. One limitation, however, is that in our single-institutional population only 23 of 472 patients attained pCR, limiting our ability to compare the two groups. Furthermore, the low usage of bevacizumab and PARP inhibitors in this study may limit generalizability given the rapidly changing landscape of ovarian cancer treatment. As this study is retrospective, inherent limitations such as missing medical record data and misclassification bias exist. Compounding this was the lack of corresponding data from the collaborating institutions regarding the number of women who did not attain pCR during the timeframes. While our single-institution population contained both pCR and non-pCR groups, allowing for an appropriate assessment of pCR rate and survival differences, the multi-institution population did not, and must be interpreted accordingly. Finally, there was no centralized pathology review for the international collaboration portion of the study.
In women who have undergone maximal-effort interval TRS following NACT, those who attain pCR demonstrate a significant two-fold RFS benefit compared to those without pCR, even those with R0 surgical outcomes. Overall survival was significantly improved in patients with pCR compared to those without pCR with optimal or suboptimal residual disease. Preoperative CA125 value may be a useful marker to identify these extreme responders.
Highlights:
Of women undergoing interval tumor-reductive surgery after neoadjuvant chemotherapy, 4.9% had pCR.
Women who attained pCR had a significantly longer recurrence-free survival compared to those who did not.
CA125 levels at the time of interval tumor-reductive surgery may be a useful biomarker to identify women who will attain pCR.
Acknowledgements
We thank Sunita Patterson, Senior Scientific Editor in MD Anderson’s Research Medical Library, for editing this article.
Sources of Funding:
This research was in part supported by the MD Anderson Ovarian Cancer Moon Shot, the National Institutes of Health through CA016672 (MD Anderson’s Cancer Center Support Grant; used the Clinical Trials Office and Biostatistics Resource Group), CA217685, CA209904, the American Cancer Society, the Ovarian Cancer Research Alliance, and the Frank McGraw Memorial Chair in Cancer Research. KFH is supported by a training fellowship from the Gulf Coast Consortia, through the Computational Cancer Biology Training Program (CPRIT Grant No. RP170593). SNW is supported by a GOG Scholar Investigator Award. JAR is supported by grant from The National Institutes of Health National Cancer Institute (K08 CA234333).
Footnotes
Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
Conflicts of Interest: During the conduct of the study, BF reports grants from NIH; KFH reports grants from CPRIT; AKS reports grants from NIH; and SNW reports grants from NIH and GOG Foundation.
Outside the submitted work, RLC reports grants and personal fees from AstraZeneca, grants from Merck, personal fees from GSK, grants and personal fees from Clovis, grants and personal fees from Genmab, grants and personal fees from Roche/Genentech, grants and personal fees from Janssen, personal fees from Agenus, personal fees from Regeneron, and personal fees from OncoQuest; NDF reports personal fees from Tesaro, personal fees from BMS/Pfizer, and personal fees from GSK; KH reports grants and personal fees from MSD, grants from Daiichi-Sankyo, personal fees from Chugai, grants and personal fees from Takeda, personal fees from Mochida, personal fees from Eisai, grants from Abbott, and grants from Ono; CJL reports personal fees from Agenus; KMatsuo reports other from Chugai; KMoore reports personal fees and other from Astra Zeneca, grants, personal fees and other from Genentech/Roche, grants, personal fees and other from Immunogen, grants, personal fees and other from Clovis, grants, personal fees and other from GSK/Tesaro, other from Pfizer, personal fees from Aravive, personal fees from VBL Therapeutics, personal fees from Onco Med, grants and other from Lilly, personal fees from Eisai, personal fees from Vavotar, personal fees from Abbvie, personal fees from Tarveda, personal fees from Myriad, personal fees from Rubius, personal fees from Elevar, personal fees from Merck, personal fees from Mersana, personal fees from Sorrento, and personal fees from OncXerna; AKS reports grants from M-Trap, other from Biopath, personal fees from Merck, personal fees from Kiyatec, and personal fees from Astra Zeneca; SNW reports grants and personal fees from AstraZeneca, grants and personal fees from Clovis Oncology, grants and personal fees from GSK/Tesaro, grants and personal fees from Roche/Genentech, grants and personal fees from Novartis, personal fees from Merck, personal fees from Pfizer, personal fees from Eisai, grants from Cotinga Pharmaceuticals, grants from Bayer, grants from ArQule, personal fees from CIrculogene, personal fees from Agenus, grants from Mereo, and grants from Bio-Path.
The remaining authors report no conflict of interest.
Abstract presentation: A portion of the findings were presented at the 50th Annual Meeting of the Society of Gynecologic Oncology; Honolulu, HI; March 16-19, 2019.
REFERENCES
- 1.Siegel RL, Miller KD, and Jemal A, Cancer statistics, 2019. CA Cancer J Clin, 2019. 69(1): p. 7–34. [DOI] [PubMed] [Google Scholar]
- 2.Torre LA, et al. Ovarian cancer statistics, 2018. CA Cancer J Clin, 2018. 68(4): p. 284–296. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Matsuo K, et al. Possible candidate population for neoadjuvant chemotherapy in women with advanced ovarian cancer. Gynecol Oncol, 2021. 160(1): p. 32–39. [DOI] [PubMed] [Google Scholar]
- 4.Meyer LA, et al. Use and Effectiveness of Neoadjuvant Chemotherapy for Treatment of Ovarian Cancer. J Clin Oncol, 2016. 34(32): p. 3854–3863. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Kehoe S, et al. Primary chemotherapy versus primary surgery for newly diagnosed advanced ovarian cancer (CHORUS): an open-label, randomised, controlled, non-inferiority trial. Lancet, 2015. 386(9990): p. 249–57. [DOI] [PubMed] [Google Scholar]
- 6.Vergote I, et al. Neoadjuvant chemotherapy or primary surgery in stage IIIC or IV ovarian cancer. N Engl J Med, 2010. 363(10): p. 943–53. [DOI] [PubMed] [Google Scholar]
- 7.Onda T, et al. Comparison of survival between primary debulking surgery and neoadjuvant chemotherapy for stage III/IV ovarian, tubal and peritoneal cancers in phase III randomised trial. Eur J Cancer, 2020. 130: p. 114–125. [DOI] [PubMed] [Google Scholar]
- 8.Fagotti A, et al. Randomized trial of primary debulking surgery versus neoadjuvant chemotherapy for advanced epithelial ovarian cancer (SCORPION-NCT01461850). Int J Gynecol Cancer, 2020. 30(11): p. 1657–1664. [DOI] [PubMed] [Google Scholar]
- 9.Cortazar P, et al. Pathological complete response and long-term clinical benefit in breast cancer: the CTNeoBC pooled analysis. Lancet, 2014. 384(9938): p. 164–72. [DOI] [PubMed] [Google Scholar]
- 10.Koboldt DC, et al. Comprehensive molecular portraits of human breast tumours. Nature, 2012. 490(7418): p. 61–70. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Petrelli F, et al. The value of platinum agents as neoadjuvant chemotherapy in triple-negative breast cancers: a systematic review and meta-analysis. Breast Cancer Res Treat, 2014. 144(2): p. 223–32. [DOI] [PubMed] [Google Scholar]
- 12.Zhang P, et al. Better pathologic complete response and relapse-free survival after carboplatin plus paclitaxel compared with epirubicin plus paclitaxel as neoadjuvant chemotherapy for locally advanced triple-negative breast cancer: a randomized phase 2 trial. Oncotarget, 2016. 7(37): p. 60647–60656. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Eisenkop SM, et al. Relative influences of tumor volume before surgery and the cytoreductive outcome on survival for patients with advanced ovarian cancer: a prospective study. Gynecol Oncol, 2003. 90(2): p. 390–6. [DOI] [PubMed] [Google Scholar]
- 14.Chi DS, et al. What is the optimal goal of primary cytoreductive surgery for bulky stage IIIC epithelial ovarian carcinoma (EOC)? Gynecol Oncol, 2006. 103(2): p. 559–64. [DOI] [PubMed] [Google Scholar]
- 15.Aletti GD, et al. Aggressive surgical effort and improved survival in advanced-stage ovarian cancer. Obstet Gynecol, 2006. 107(1): p. 77–85. [DOI] [PubMed] [Google Scholar]
- 16.Schorge JO, et al. SGO White Paper on ovarian cancer: etiology, screening and surveillance. Gynecol Oncol, 2010. 119(1): p. 7–17. [DOI] [PubMed] [Google Scholar]
- 17.Weissman SM, Weiss SM, and Newlin AC, Genetic testing by cancer site: ovary. Cancer J, 2012. 18(4): p. 320–7. [DOI] [PubMed] [Google Scholar]
- 18.Alsop K, et al. BRCA mutation frequency and patterns of treatment response in BRCA mutation-positive women with ovarian cancer: a report from the Australian Ovarian Cancer Study Group. J Clin Oncol, 2012. 30(21): p. 2654–63. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19.Petrillo M, et al. Prognostic role and predictors of complete pathologic response to neoadjuvant chemotherapy in primary unresectable ovarian cancer. Am J Obstet Gynecol, 2014. 211(6): p. 632 e1–8. [DOI] [PubMed] [Google Scholar]
- 20.Zorzato PC, et al. External validation of a 'response score' after neoadjuvant chemotherapy in patients with high-grade serous ovarian carcinoma with complete clinical response. Int J Gynecol Cancer, 2020. 30(1): p. 67–73. [DOI] [PubMed] [Google Scholar]
- 21.Muraji M, et al. Histopathology predicts clinical outcome in advanced epithelial ovarian cancer patients treated with neoadjuvant chemotherapy and debulking surgery. Gynecol Oncol, 2013. 131(3): p. 531–4. [DOI] [PubMed] [Google Scholar]
- 22.Liang MI, et al. Prognostic role of pathologic response and cytoreductive status at interval debulking surgery after neoadjuvant chemotherapy for advanced epithelial ovarian cancer. J Surg Oncol, 2019. 120(4): p. 779–785. [DOI] [PubMed] [Google Scholar]
- 23.Katsumata N, et al. Long-term results of dose-dense paclitaxel and carboplatin versus conventional paclitaxel and carboplatin for treatment of advanced epithelial ovarian, fallopian tube, or primary peritoneal cancer (JGOG 3016): a randomised, controlled, open-label trial. Lancet Oncol, 2013. 14(10): p. 1020–6. [DOI] [PubMed] [Google Scholar]
- 24.Clamp AR, et al. Weekly dose-dense chemotherapy in first-line epithelial ovarian, fallopian tube, or primary peritoneal carcinoma treatment (ICON8): primary progression free survival analysis results from a GCIG phase 3 randomised controlled trial. Lancet, 2019. 394(10214): p. 2084–2095. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25.Stone RL, et al. Paraneoplastic thrombocytosis in ovarian cancer. N Engl J Med, 2012. 366(7): p. 610–8. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26.Allensworth SK, et al. Evaluating the prognostic significance of preoperative thrombocytosis in epithelial ovarian cancer. Gynecol Oncol, 2013. 130(3): p. 499–504. [DOI] [PMC free article] [PubMed] [Google Scholar]