Abstract
Background
The study objective is to examine the impact of obesity on frontline carboplatin dosing in the neoadjuvant and adjuvant settings and to evaluate the association of dosing with survival among epithelial ovarian cancer (EOC) patients.
Methods
We selected 1527 women diagnosed with EOC from January 1, 2011 to October 20, 2021 from a nationwide electronic health record-derived de-identified database. The dose reduction of frontline carboplatin was defined as a relative dose intensity (RDI) < 0.85. Cox proportional hazards regression was used to estimate hazard ratios (HR) and 95% confidence intervals (CI) for the association of RDI with survival overall and by histology.
Results
Women with a BMI ≥ 30 kg/m2 versus <30 kg/m2 were more likely to be underdosed (RDI < 0.85) with frontline carboplatin. Underdosing of carboplatin in the neoadjuvant setting was associated with worse survival among women with serous tumours (HR = 1.98, 95% CI = 1.15, 3.42). Underdosing of carboplatin in the adjuvant setting was not associated with survival.
Discussion
In the real-world setting, underdosing of carboplatin in the neoadjuvant setting was associated with inferior survival among women with serous tumours. With the increasing utilisation of neoadjuvant chemotherapy in EOC, actual weight-based dosing of carboplatin may be important to improve outcomes in this patient population.
Subject terms: Ovarian cancer, Cancer epidemiology
Introduction
While the dismal five-year survival rate of 48% [1] among women with ovarian cancer is partially attributed to diagnosis with advanced-stage disease (5-year survival 29%) [1], tumour biology and treatment differences likely contribute. Frontline treatment for ovarian cancer includes cytoreductive surgery and a platinum- and taxane-based chemotherapy [2]. Calculation of chemotherapy dosing using body weight or body surface area (BSA) [3] may be adjusted in obese patients due to concern for treatment-related toxicities [3, 4], despite The American Society of Clinical Oncology (ASCO) recommendations for full weight-based dosing in this patient population [2]. Obese women are more likely to be underdosed in the adjuvant setting [5, 6] which may be associated with increased mortality [6]. However, little is known about the effect of underdosing ovarian cancer patients in the neoadjuvant setting despite the increase in its utilisation as a frontline treatment strategy over the last two decades [7, 8]. Thus, the objective of this study is to determine how obesity impacts frontline chemotherapy dosing in both the neoadjuvant and adjuvant settings and how dosing impacts outcomes among epithelial ovarian cancer (EOC) patients in a real-world electronic health record (EHR)-derived database.
Methods
Study population
This study used the nationwide longitudinal EHR-derived Flatiron Health database, comprising de-identified patient-level structured and unstructured data, curated via technology-enabled abstraction [9, 10]. During the study period, the de-identified data originated from approximately 280 U.S. cancer clinics (~800 sites of care). This study included 1527 patients diagnosed with invasive ovarian, fallopian tube, or peritoneal cancer who had at least two documented clinical visits, on different days, occurring from January 1, 2011 to October 20, 2021. The majority of patients in the database originate from community oncology settings; relative community/academic proportions may vary depending on the study cohort. The data are de-identified and subject to obligations to prevent re-identification and protect patient confidentiality.
Patient and clinical characteristics
Patient data on age at diagnosis (continuous, years), race and ethnicity (non-Hispanic White, non-Hispanic Black, Hispanic, and Other), histology (serous, endometrioid, mucinous, clear cell, and epithelial not otherwise specified [NOS]), stage (early [Stage 1, 2], late [Stage 3, 4]), BMI at diagnosis (kg/m2), first-line therapy (neoadjuvant chemotherapy, upfront surgical debulking), and debulking status (optimal or no or <1 cm residual disease, suboptimal ≥1 cm residual disease) were included in downstream analyses.
Relative dose intensity
For first-line therapy, detailed by-cycle chemotherapy dosing information for carboplatin in the neoadjuvant and adjuvant settings were used to calculate relative dose intensity (RDI; Supplementary Table 1). For each cycle of carboplatin, the RDI was calculated by dividing the actual dose by the expected dose, according to dosing recommendations in the NCCN guidelines [2]. First, creatinine clearance (CrCl) was calculated using the Cockcroft–Gault equation: CrCl = ((140 – age in years) × weight at the time of the cycle administration in kg)/((72 × serum creatinine in mg/dL) × 0.85 (for females)). To avoid overestimation of CrCl, a measured serum creatinine of less than 0.7 mg/dL was set to 0.7 mg/dL, and CrCl was capped at 125. Next, the expected dose was calculated using the Calvert formula (target AUC × (CrCl + 25)). As recommended by NCCN guidelines and to reflect current clinical practice, the expected dose was capped at 900 and 750 for targets of AUC = 6 and AUC = 5, respectively [2]. If the measurement of weight, serum creatinine, or target AUC was missing for any cycle, we used measured weight, serum creatinine, or target AUC within 31 days of the cycle start date and before surgery for neoadjuvant chemotherapy, and within 31 days of the cycle start date and surgery as a proxy. There was excellent consistency of the carboplatin RDI across all cycles in the neoadjuvant and adjuvant settings within frontline therapy regimens (intraclass correlation coefficient [ICC] = 0.94 and 0.97, respectively among women with upfront chemotherapy and ICC = 0.95 for the adjuvant setting for women with upfront surgery). Therefore, within each frontline therapy regimen, the RDI for each cycle was summed and divided by the number of cycles completed in both the neoadjuvant and adjuvant settings. An average RDI < 0.85 was used to define dose reduction [5, 6, 11].
Multiple imputation
Data were missing for key prognostic variables, stage (n = 83, 5%) and debulking status (n = 342, 22%). Multiple imputation was used to impute the missing data for these variables via the ‘mice’ R package. Within each frontline treatment group (neoadjuvant chemotherapy vs. upfront surgical debulking), multiple imputation was performed using a Cox proportional hazard regression model, including carboplatin RDI, race and ethnicity, age at diagnosis, stage, debulking status, and histology. A dataset containing these covariates, the predictors, and the outcome as well as the Nelson–Aalen estimate of the cumulative hazard is fed to the algorithm. The fraction of missing information is used to set the number of imputations.
Statistical analyses
The distribution of patient and clinical characteristics, and RDI for carboplatin in the neoadjuvant and adjuvant settings were compared by pre-treatment BMI (< 25 kg/m2, ≥25 to <30 kg/m2, ≥30 kg/m2). To examine overall survival, dates of death were based on a composite mortality variable that utilises data from the EHR, the Social Security Death Index, and commercial death data and is highly concordant with the National Death Index [12]. Follow-up time was measured as the number of days from the date of first treatment to date of death or last contact. Kaplan-Meier survival curves and log-rank tests were used to compare overall survival (OS) by RDI for carboplatin in neoadjuvant and adjuvant settings stratified by frontline therapy regimen (neoadjuvant chemotherapy or primary surgical debulking). Within each frontline therapy regimen, multivariable Cox proportional hazard regression was performed to estimate hazard ratios (HR) and 95% confidence intervals (CI) for the association of RDI with risk of all-cause mortality while adjusting for relevant prognostic factors (age at diagnosis, race and ethnicity, stage, debulking status, histology, and year of diagnosis). For the analyses examining women treated with neoadjuvant chemotherapy, both RDI for carboplatin in the neoadjuvant and adjuvant settings were included in the same model. The proportional hazards assumption was tested using time x covariate interactions (e.g., follow-up time × stage) for each covariate individually and collectively. No violation of proportional hazards was observed among women receiving upfront chemotherapy. However, a violation of proportional hazards was observed for stage and histology in the models of women receiving upfront surgery (P < 0.05) because of time-varying survival patterns for these covariates. To address nonproportionality, stage and histology were included as strata variables in the Cox models of women with upfront surgery to allow for different baseline hazard functions by these covariates. We additionally assessed the association of carboplatin RDI and survival by histology (serous vs. non-serous) and by pre-treatment BMI ( < 25 kg/m2 vs. ≥ 25 kg/m2). We also examined whether our findings were consistent using multiple imputation versus a complete case analytic approach, and additionally examined whether patient characteristics differed among women with versus without complete data. A two-sided P < 0.05 was used to define statistical significance. Analyses were performed using R (Version 4.1.2.).
Results
A total of 8519 EOC patients diagnosed from January 1, 2011 to October 20, 2021 were available in Flatiron Health. Of those, 5389 had by-cycle chemotherapy information and met Flatiron Health guidelines for inclusion. We further restricted the dataset to exclude women with borderline tumours or unknown histology as well as women who did not receive both surgery and chemotherapy as frontline therapy or women who received other platinum-based therapies as frontline therapy besides carboplatin (e.g., cisplatin). Of the 2971 women remaining, we excluded women with missing data for variables used to calculate the RDI (n = 797) and women who received non-guideline-adherent frontline care (n = 647; e.g., two lines of chemotherapy prior to surgery, more than nine total cycles of frontline chemotherapy). Thus, 1527 invasive EOC patients met study eligibility criteria and had available data on frontline chemotherapy to be included in this analysis (Fig. 1).
Fig. 1. Consort diagram.
Flow diagram of the inclusion of women with ovarian cancer in the study.
The median age of diagnosis was 64 years (interquartile range [IQR] = 56, 71), and most women were non-Hispanic White (67%; Table 1). Women were more commonly diagnosed in 2011–2015 (38%), followed by 2016–2018 (35%) and 2019–2021 (27%). The majority of patients had serous tumours (67%) and were diagnosed with late-stage disease (66%). A little over a fourth of patients received neoadjuvant chemotherapy (26%) as their frontline therapy regimen, and most were optimally debulked with their cytoreductive surgery (84%). Compared to women with a BMI < 25 kg/m2, women with a BMI ≥ 30 kg/m2 were more likely to be diagnosed at a younger age and to be non-Hispanic Black.
Table 1.
Distribution of patient characteristics overall and by pre-treatment BMI.
| Pre-treatment BMI | ||||
|---|---|---|---|---|
| Patient characteristics | Total (N = 1527) | <25 kg/m2 (n = 547) | ≥25 to <30 kg/m2 (n = 438) | ≥30 kg/m2 (n = 542) |
| Median (IQR) or n (%) | Median (IQR) or n (%) | Median (IQR) or n (%) | Median (IQR) or n (%) | |
| Age at diagnosis, years | 64 (56, 71) | 65 (57, 73) | 66 (58, 72) | 61 (54, 68) |
| Year of diagnosis | ||||
| 2011–2015 | 581 (38) | 210 (38) | 164 (37) | 207 (38) |
| 2016–2018 | 541 (35) | 199 (36) | 159 (36) | 183 (34) |
| 2019–2021 | 405 (27) | 138 (25) | 115 (26) | 152 (28) |
| Race and ethnicity | ||||
| Non-Hispanic White | 1020 (67) | 359 (66) | 299 (68) | 362 (67) |
| Non-Hispanic Black | 78 (5) | 15 (3) | 25 (6) | 38 (7) |
| Hispanic | 104 (7) | 40 (7) | 31 (7) | 33 (6) |
| Other | 325 (21) | 133 (24) | 83 (19) | 109 (20) |
| Histology | ||||
| Serous | 1026 (67) | 374 (68) | 300 (68) | 352 (65) |
| Endometrioid | 144 (9) | 35 (6) | 44 (10) | 65 (12) |
| Clear cell | 136 (9) | 57 (10) | 33 (8) | 46 (8) |
| Mucinous | 46 (3) | 18 (3) | 14 (3) | 14 (3) |
| Epithelial, NOS | 175 (11) | 63 (12) | 47 (11) | 65 (12) |
| Stagea | ||||
| Stage 1 and 2 | 513 (34) | 187 (34) | 147 (34) | 179 (33) |
| Stage 3 and 4 | 1014 (66) | 360 (66) | 291 (66) | 363 (67) |
| Upfront first-line therapy | ||||
| Neoadjuvant chemotherapy | 391 (26) | 131 (24) | 108 (25) | 152 (28) |
| Surgical debulking | 1136 (74) | 416 (76) | 330 (75) | 390 (72) |
| Debulking statusa | ||||
| Optimal | 1279 (84) | 456 (83) | 363 (83) | 460 (85) |
| Suboptimal | 248 (16) | 91 (17) | 75 (17) | 82 (15) |
| Survival time, months | 30 (16, 51) | 31 (17, 51) | 30 (16, 53) | 29 (16, 50) |
BMI body mass index, IQR interquartile range, NOS not otherwise specified.
aImputed variable. Stage was missing for 83 participants, and debulking status was missing for 342 participants.
Among women who received upfront chemotherapy (n = 391), 17% (n = 67) experienced underdosing of carboplatin in the neoadjuvant setting and 24% (n = 93) in the adjuvant setting (Supplemental Table 2), and of the women underdosed in the neoadjuvant setting (n = 67), the majority were also underdosed in the adjuvant setting (73%). Similarly, 21% of women who had surgical debulking as frontline therapy were underdosed with carboplatin in the adjuvant setting (n = 1136). Irrespective of type of frontline therapy, women with a BMI ≥ 30 kg/m2 were more likely to experience carboplatin dose reduction in the neoadjuvant and adjuvant settings (Fig. 2). For example, among women who received upfront chemotherapy, 8% of women with a BMI < 25 kg/m2 were underdosed with carboplatin in the neoadjuvant setting compared to 11% for women with a BMI ≥ 25 to <30 kg/m2 and 29% for women with a BMI ≥ 30 kg/m2 (Supplemental Table 2).
Fig. 2. Ridgeline plot of the average carboplatin RDI stratified by type of frontline therapy and pre-treatment BMI.
Carboplatin RDI in the neoadjuvant setting among women receiving upfront chemotherapy (a), in the adjuvant setting among women receiving upfront chemotherapy (b), and in the adjuvant setting among women receiving upfront surgery (c). The blue dotted line represents the RDI value of 0.85 indicating dose reduction. RDI relative dose intensity, BMI body mass index.
After adjustment for relevant prognostic factors, women experiencing carboplatin underdosing in the neoadjuvant setting had worse survival than women who were adequately dosed (HR = 1.51, 95% CI = 0.95, 2.40), although this association was not statistically significant (Table 2). Stratifying by histology revealed that underdosing with carboplatin in the neoadjuvant setting was significantly associated with worse survival among women with serous tumours receiving upfront chemotherapy as frontline therapy (HR = 1.98, 95% CI = 1.15, 3.42) but not among women with non-serous tumours (HR = 0.77, 95% CI = 0.25, 2.36). No associations between RDI in the adjuvant setting and survival were observed for women receiving either upfront chemotherapy or upfront surgical debulking overall or by histology.
Table 2.
HRs and 95% CIs of the association of RDI and survival stratified by first-line therapy overall and by histology.
| Total | Serous | Non-serous | ||||
|---|---|---|---|---|---|---|
| Patient characteristic | Cases (deaths) | HR (95% CI) | Cases (deaths) | HR (95% CI) | Cases (deaths) | HR (95% CI) |
| Upfront chemotherapy (n = 391)a | ||||||
| Neoadjuvant RDI | ||||||
| ≥85% | 324 (147) | 1.00 (referent) | 249 (113) | 1.00 (referent) | 75 (34) | 1.00 (referent) |
| <85% | 67 (32) | 1.51 (0.95, 2.40) | 51 (28) | 1.98 (1.15, 3.42) | 16 (4) | 0.77 (0.25, 2.36) |
| Adjuvant RDI | ||||||
| ≥85% | 298 (133) | 1.00 (referent) | 228 (103) | 1.00 (referent) | 70 (30) | 1.00 (referent) |
| <85% | 93 (46) | 0.95 (0.63, 1.41) | 72 (38) | 0.86 (0.53, 1.38) | 21 (8) | 1.32 (0.56, 3.09) |
| Upfront surgical debulking (n = 1136)b | ||||||
| Adjuvant RDI | ||||||
| ≥85% | 896 (285) | 1.00 (referent) | 565 (211) | 1.00 (referent) | 331 (74) | 1.00 (referent) |
| <85% | 240 (71) | 0.92 (0.71, 1.19) | 161 (55) | 0.94 (0.70, 1.27) | 79 (16) | 0.86 (0.49, 1.50) |
HR hazard ratio, CI confidence interval, RDI relative dose intensity.
aAdjusted for age at diagnosis, race and ethnicity, stage, debulking status, histology, and year of diagnosis. Models among women with serous tumours were not adjusted for histology.
bAdjusted for age at diagnosis, race and ethnicity, debulking status, and year of diagnosis. Stage and histology were included as strata terms due to violations of the proportional hazard assumption. Models among women with serous tumours were not adjusted for histology.
After restricting to patients with complete data on each variable (n = 1125), we repeated our analyses and observed similar findings (Supplemental Table 3). The association of carboplatin underdosing in the neoadjuvant setting with worse survival was present in both the overall population (HR = 2.10, 95% CI = 1.16, 3.79) and among serous tumours (HR = 2.27, 95% CI = 1.18, 4.36). When comparing the women included vs. excluded from the complete case analysis (Supplemental Table 4), included women were more likely to be diagnosed in 2011–2015 (P = 0.008) and to have serous tumours and late-stage disease compared to excluded women (P < 0.001). Thus, the stronger RDI-outcome association in the neoadjuvant setting for the complete case approach is expected as many of the non-serous tumours were excluded from this analysis, where no association with survival was observed.
Stratifying by pre-treatment BMI revealed no differences in the association of adjuvant RDI with survival among women treated with either neoadjuvant chemotherapy or upfront surgical debulking (Supplemental Table 5). However, underdosing of carboplatin in the neoadjuvant setting was associated with worse survival among women with a BMI < 25 kg/m2 (HR = 2.51, 95% CI = 0.68, 9.32) but not statistically significant.
Discussion
In this retrospective analysis of EOC patients in the Flatiron Health database, we demonstrated that obese patients are more likely to be underdosed with carboplatin in both the neoadjuvant and adjuvant settings compared to their normal-weight counterparts. Among women receiving neoadjuvant chemotherapy as frontline therapy, underdosing in the neoadjuvant setting was associated with an increased likelihood of death, particularly among women with serous tumours. Underdosing of carboplatin in the adjuvant setting was not associated with survival irrespective of frontline therapy regimen and histology.
Several studies have evaluated the impact of obesity on chemotherapy dosing and survival in EOC patients with conflicting findings [5, 6, 13–17]. Although studies consistently show that obese women are more likely to be underdosed with platinum-based chemotherapy [5, 6, 13, 17], dose reduction has been associated with worse survival in some [6, 16, 17] but not all studies after adjusting for other prognostic factors [5, 13–15]. These conflicting findings may be due, in part, to small sample sizes and heterogeneous study populations as many studies focus on selected patient populations, such as women with advanced-stage disease [14–17] or those with serous tumours [5]. In one of the largest studies conducted to date that included all EOC patients irrespective of clinical or pathologic characteristics, Bandera, et al. [6] found that a lower RDI for carboplatin in the upfront adjuvant setting was an independent risk factor for ovarian cancer mortality, advocating for full weight-based dosing among ovarian cancer patients. Our study did not corroborate these findings as carboplatin dosing in the adjuvant setting was not associated with survival. However, it is important to note that unlike many of the studies discussed above, our study leveraged data from the Flatiron Health database which includes patients treated at predominantly community-based practices/hospitals and not a large medical group or academic centre where patients likely have access to gynaecologic oncologists which is a known contributor to improved clinical outcomes [18–20].
Despite our null findings in the adjuvant setting, we observed inferior survival for women underdosed with carboplatin in the neoadjuvant setting. To our knowledge, only one prior study included women receiving neoadjuvant chemotherapy [13] and found that progression-free survival was similar irrespective of dose reduction in both the neoadjuvant and adjuvant settings; however, a multivariable analysis was not conducted to investigate this association while adjusting for confounding prognostic factors. As neoadjuvant chemotherapy is becoming more common as a treatment strategy for women with more aggressive, advanced-stage disease that are less likely to achieve optimal debulking with upfront cytoreductive surgery [21], it is critical that additional contemporary studies investigate how chemotherapy dosing impacts outcomes among EOC patients receiving neoadjuvant chemotherapy for frontline therapy.
We found that the inferior survival for underdosing of carboplatin in the neoadjuvant setting was specifically observed among women with serous tumours. Serous tumours are the most common histologic subtype of EOC, and high-grade serous carcinoma, which represents >95% of serous tumours, is recognised as an aggressive tumour that is often diagnosed at a late stage where the disease has disseminated throughout the peritoneal cavity [22, 23]. High-grade serous tumours typically respond well to frontline platinum-based chemotherapy, although most patients will relapse within a few years after initial therapy [21, 24]. Thus, underdosing of frontline carboplatin in this patient population may have a more profound impact on outcomes than among women with non-serous tumours, as observed in the present study. As previously mentioned, due to the aggressive nature of high-grade serous tumours, women with high-grade serous ovarian carcinoma are more likely to receive neoadjuvant chemotherapy as part of their initial management. In the present study, we observed that women receiving neoadjuvant chemotherapy were more likely to have serous tumours than women receiving upfront surgical debulking (78% vs. 63%). Flatiron Health did not collect grade information for ovarian tumours and thus, we were unable to distinguish between high- and low-grade serous tumours. However, as most serous tumours are high-grade [22], our results for serous tumours are likely reflective of high-grade serous carcinoma with minimal bias due to low-grade tumours.
While an association was noted for underdosing of carboplatin in the neoadjuvant setting and survival among women with serous tumours, no associations with survival were observed among women with non-serous tumours for either frontline therapy regimen. This finding is not surprising as some of the histologic subtypes represented in the ‘non-serous’ group are less responsive to chemotherapy (e.g., clear cell, mucinous) [24–26], and thus, it would be unexpected for underdosing of chemotherapy to substantially impact the survival of this patient population. The small sample size of non-serous histologic subtypes, particularly among women receiving neoadjuvant chemotherapy as frontline therapy, impacted our power to investigate associations within each of those subtypes separately. Larger studies with deeply characterised clinical information are needed to further investigate whether the association between chemotherapy dosing and outcomes varies according to histologic subtype of ovarian cancer.
We explored whether our findings differed according to pre-treatment BMI and noted that the association of underdosing in the neoadjuvant setting with worse survival was particularly marked among women with a BMI < 25 kg/m2, although the confidence intervals were wide due to small sample sizes for these sub-group analyses. Similarly, Bandera et al. [6] observed that the inferior survival observed for dose reduction of women with ovarian cancer treated with upfront surgical debulking was most pronounced among normal-weight women compared to overweight and obese women. It is not clear why the association of carboplatin underdosing with poor survival was more pronounced among women with a BMI < 25 kg/m2; however, this may be due to differences in physiologic factors, such as age or creatinine levels, or toxicities across BMI that may impact dosing decisions. We did not have information on the reasons for dose reduction in the present study to disentangle the complex relationships between BMI, toxicity and chemotherapy dosing.
Our study is strengthened by the utilisation of a large, national database including patients treated in predominantly community clinical settings. This work is also one of few contemporary studies to examine the impact of chemotherapy dosing on outcomes in the era of increased utilisation of neoadjuvant chemotherapy as frontline therapy. The most important limitation of this study is its retrospective nature, and the associated inherent risks of selection and reporting biases. Data ascertainment is dependent on proper coding and accurate documentation in the medical record. We also do not have information on treatment that was received outside of the Flatiron Health network. We were unable to examine associations with ovarian cancer-specific or progression-free survival as cause of death and disease progression were not available on this cohort of patients. A few key prognostic factors, stage and debulking status, were missing for a subset of patients. Thus, we performed multiple imputation to impute missing values for these variables. Our imputed results were fairly consistent with findings from the complete case analysis approach, and our primary conclusion of inferior survival for women who were underdosed with carboplatin in the neoadjuvant setting remained irrespective of analytic approach. This study only included patients who received guideline-adherent care and carboplatin and taxane chemotherapy regimens, and thus, our results may not be generalisable to all patients.
Conclusions
In conclusion, obesity is associated with underdosing of carboplatin in the frontline among EOC patients in both neoadjuvant and adjuvant settings, and underdosing of carboplatin in the neoadjuvant setting is associated with worse survival, particularly among women with serous tumours. Full, weight-based dosing of carboplatin based on ASCO guidelines may be important to improve outcomes in this cohort of patients.
Supplementary information
Author contributions
LCP and JC designed the research study. CCL and LCP analysed the data. CCL and LCP created the tables and figures. ALM and LCP wrote the manuscript. All authors reviewed and approved the final manuscript.
Funding
This publication is supported by a 2019 Moffitt Team Science Award and the Cancer Centre Support Grant P30 CA 076292. This work was also supported by the Collaborative Data Services Core at Moffitt Cancer Centre (P30 CA 076292).
Data availability
The data that support the findings of this study have been originated by Flatiron Health, Inc. These de-identified data may be made available upon request and are subject to a license agreement with Flatiron Health; interested researchers should contact “DataAccess@flatiron.com” to determine licensing terms.
Competing interests
The authors declare no competing interests.
Ethics approval and consent to participate
Institutional Review Board approval of the study protocol was obtained prior to the study conduct and included a waiver of informed consent.
Consent for publication
Not applicable.
Footnotes
Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
These authors contributed equally: Jing-Yi Chern, Lauren C. Peres.
Supplementary information
The online version contains supplementary material available at 10.1038/s41416-023-02259-1.
References
- 1.Siegel R, Jemal A. Cancer statistics, 2020. CA Cancer J Clin. 2020;70:7–30. doi: 10.3322/caac.21590. [DOI] [PubMed] [Google Scholar]
- 2.Armstrong D, Alvarez R, Bakkum-Gamez J, Barroilhet L, Behbakht K, Berchuck A, et al. NCCN guidelines insights: ovarian cancer, version 1.2019. J Natl Compr Cancer Netw. 2019;17:896–909. doi: 10.6004/jnccn.2019.0039. [DOI] [PubMed] [Google Scholar]
- 3.Body surface area for adjustment of drug dose. Drug Ther Bull. 2010;48:33–6. https://pubmed.ncbi.nlm.nih.gov/20200147/. [DOI] [PubMed]
- 4.Gurney H. Obesity in dose calculation: a mouse or an elephant? J Clin Oncol. 2007;25:4703–4. doi: 10.1200/JCO.2007.13.1078. [DOI] [PubMed] [Google Scholar]
- 5.Au-Yeung G, Webb P, DeFazio A, Fereday S, Bressel M, Mileshkin L. Impact of obesity on chemotherapy dosing for women with advanced stage serous ovarian cancer in the Australian Ovarian Cancer Study (AOCS) Gynecol Oncol. 2014;133:16–22. doi: 10.1016/j.ygyno.2014.01.030. [DOI] [PubMed] [Google Scholar]
- 6.Bandera E, Lee V, Rodriguez-Rodriguez L, Powell B, Kushi L. Impact of chemotherapy dosing on ovarian cancer survival according to body mass index. JAMA Oncol. 2015;1:737–45. doi: 10.1001/jamaoncol.2015.1796. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Melamed A, Hinchliff E, Clemmer J, Bregar A, Uppal U, Bostock I, et al. Trends in the use of neoadjuvant chemotehrapy for advanced ovarian cancer in the United States. Gynecol Oncol. 2016;143:236–40. doi: 10.1016/j.ygyno.2016.09.002. [DOI] [PubMed] [Google Scholar]
- 8.Lheureux S, Braunstein M, Oza A. Epithelial ovarian cancer: evolution of management in the era of precision medicine. CA Cancer J Clin. 2019;69:280–304. doi: 10.3322/caac.21559. [DOI] [PubMed] [Google Scholar]
- 9.Ma X, Long L, Moon S, Adamson B, Baxi S. Comparison of population characteristics in real-world clinical oncology databases in the US: Flatiron Health, SEER, and NPCR. medRxiv [Preprint] 2020. Available from: 10.1101/2020.03.16.20037143.
- 10.Birnbaum B, Nussbaum N, Seidl-Rathkopf K, Agrawal M, Estevez M, Estola E, et al. Model-assisted cohort selection with bias analysis for generating large-scale cohorts from the EHR for oncology research. arXiv:2001.09765 [cs.CY].[Preprint]. Available from: 10.48550/arXiv.2001.09765.
- 11.Griggs J, Mangu P, Temin S, Lyman G. Appropriate chemotherapy dosing for obese adult patients with cancer: American Society of Clinical Oncology Clinical Practice Guideline. J Oncol Pr. 2012;8:e59–e61. doi: 10.1200/JOP.2012.000623. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Curtis M, Griffith S, Tucker M, Taylor M, Capra W, Carrigan G. Development and validation of a high-quality composite real-world mortality endpoint. Health Serv Res. 2018;53:4460–76. doi: 10.1111/1475-6773.12872. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Sivakumaran T, Mileshkin L, Grant P, Na L, DeFazio A, Friedlander M, et al. Evaluating the impact of dose reductions and delays on progression-free survival in women with ovarian cancer treated with either three-weekly or dose-dense carboplatin and paclitaxel regimens in the national prospective OPAL cohort study. Gynecol Oncol. 2020;15:47–53. doi: 10.1016/j.ygyno.2020.04.706. [DOI] [PubMed] [Google Scholar]
- 14.Liutkauskiene S, Janciauskiene R, Jureniene K, Grizas S, Malonyte R, Juozaityte E. Retrospective analysis of the impact of platinum dose reduction and chemotherapy delays on the outcomes of stage III ovarian cancer patients. BMC Cancer. 2015;15:105. doi: 10.1186/s12885-015-1104-5. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Nagel C, Backes F, Hade E. Effect of chemotherapy delays and dose reduction on progression free and overall survival in the treatment of epithelial ovarian cancer. Gynecol Oncol. 2012;124:221–4. doi: 10.1016/j.ygyno.2011.10.003. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Anuradha S, Donovan P, Webb P, Brand A, Goh J, Friedlander M, et al. Variations in adjuvant chemotherapy and survival in women with epithelial ovarian cancer—a population-based study. Acta Oncol. 2016;55:226–33. doi: 10.3109/0284186X.2015.1054950. [DOI] [PubMed] [Google Scholar]
- 17.Hanna R, Poniewierski M, Laskey R, Lopez M, Shafer A, Van Le L, et al. Predictors of reduced relative dose intensity and its relationship to mortality in women receiving multi-agent chemotherapy for epithelial ovarian cancer. Gynecol Oncol. 2013;129:74–80. doi: 10.1016/j.ygyno.2012.12.017. [DOI] [PubMed] [Google Scholar]
- 18.Earle C, Schrag D, Neville B, Yabroff K, Topor M, Fahey A, et al. Effect of surgeon speciality on processes of care and outcomes for ovarian cancer patients. J Natl Cancer Inst. 2006;98:172–80. doi: 10.1093/jnci/djj019. [DOI] [PubMed] [Google Scholar]
- 19.Engelen M, Kos H, Willemse P, Aalder J, de Vries E, Schaapveld M, et al. Surgery by consultant gynecologic oncologists improves survival in patients with ovarian carcinoma. Cancer. 2006;106:589–98. doi: 10.1002/cncr.21616. [DOI] [PubMed] [Google Scholar]
- 20.Bouchard-Fortier G, Gien L, Sutradhar R, Chan W, Krzyzanowska M, Liu S, et al. Impact of care by gynecologic oncologists on primary ovarian cancer survival: a population-based study. Gynecol Oncol. 2022;164:522–8. doi: 10.1016/j.ygyno.2022.01.003. [DOI] [PubMed] [Google Scholar]
- 21.Matulonis U, Sood A, Fallowfield L, Howitt B, Sehouli J, Karlan B. Ovarian cancer. Nat Rev Dis Prim. 2016;2:16061. doi: 10.1038/nrdp.2016.61. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22.Peres L, Cushing-Haugen K, Köbel M, Harris H, Berchuck A, Rossing MA, et al. Invasive epithelial ovarian cancer survival by histotype and disease stage. J Natl Cancer Inst. 2019;111:60–68. doi: 10.1093/jnci/djy071. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23.Lisio M, Fu L, Goyeneche A, Gao A, Telleria C. High-grade serous ovarian cancer: basic sciences and therapeutic standpoints. Int J Mol Sci. 2019;20:952. doi: 10.3390/ijms20040952. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24.Jelovac D, Armstrong D. Recent progress in the diagnosis and treatment of ovarian cancer. CA Cancer J Clin. 2011;61:183–203. doi: 10.3322/caac.20113. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25.Pectasides D, Pectasides E, Psyrri A, Economopoulos T. Treatment issues in clear cell carcinoma of the ovary: a different entity? Oncologist. 2006;11:1089–94. doi: 10.1634/theoncologist.11-10-1089. [DOI] [PubMed] [Google Scholar]
- 26.Hess V, A’Hern R, Nasiri N, King D, Blake P, Barton D, et al. Mucinous epithelial ovarian cancer: a separate entity requiring specific treatment. J Clin Oncol. 2004;22:1040–4. doi: 10.1200/JCO.2004.08.078. [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Data Availability Statement
The data that support the findings of this study have been originated by Flatiron Health, Inc. These de-identified data may be made available upon request and are subject to a license agreement with Flatiron Health; interested researchers should contact “DataAccess@flatiron.com” to determine licensing terms.