Abstract
Introduction
Complete gross resection (CGR) at interval debulking surgery (IDS) is a critical surgical endpoint in ovarian cancer, but its likelihood remains difficult to estimate before surgery in patients treated with neoadjuvant chemotherapy (NACT). The hemoglobin, albumin, lymphocyte, and platelet (HALP) score integrates nutritional, inflammatory, immune, and anemia-related status, but its association with CGR after NACT remains unclear. This study evaluated the association of peri-NACT HALP with CGR at IDS.
Methods
We retrospectively analyzed 239 patients with advanced high-grade serous ovarian cancer who received platinum-based NACT followed by IDS. HALP was calculated before NACT and before IDS, and HALP difference was defined as the latter minus the former. The ability of HALP-related indices to distinguish CGR was examined using receiver operating characteristic analyses and logistic regression models. Exploratory analyses further examined KELIM score, chemotherapy response score, and survival outcomes in relation to these indices.
Results
Compared with the non-CGR group, the CGR group showed higher HALP at both sampling points and a larger peri-NACT difference. The area under the curve was 0.761 for HALP before NACT, 0.816 for HALP before IDS, and 0.593 for HALP difference. In multivariate analysis, HALP before NACT >19.3 and HALP before IDS >33.4 remained independently associated with lower odds of non-CGR, whereas HALP difference did not. Combined pre/post HALP status further stratified CGR rates. Exploratory analyses also suggested associations between HALP-related indices and NACT response indicators and survival outcomes.
Conclusion
HALP before NACT and before IDS were associated with CGR after NACT, whereas HALP difference showed limited discriminatory performance. These exploratory associations require further external validation.
Keywords: ovarian cancer, hemoglobin, albumin, lymphocyte, platelet score, neoadjuvant chemotherapy, complete gross resection, predictive biomarker
Introduction
For advanced ovarian cancer, upfront cytoreductive surgery with subsequent platinum-based chemotherapy is an established treatment approach. When extensive disease, operative risk, or a low prospect of complete primary cytoreduction argues against upfront surgery, patients may instead receive neoadjuvant chemotherapy (NACT) before interval debulking surgery (IDS) to reduce disease burden and facilitate resection.1 Within this treatment pathway, the principal operative goal at IDS is complete gross resection (CGR), because any visible residual disease may negate part of the prognostic advantage associated with complete tumor removal.2,3 Patients who are unlikely to achieve CGR represent a clinically important subgroup, as they may undergo morbid surgery without clear survival benefit and miss the opportunity for timely treatment modification. Therefore, preoperative identification of patients unlikely to achieve CGR remains a critical issue in advanced ovarian cancer treated with NACT.
Existing strategies for anticipating cytoreduction outcome draw on clinical features, tumor markers, or imaging, yet their suitability for repeated use during NACT is limited. Radiomic signatures and image-based tumor-burden measures can aid prediction, but they depend on specialized acquisition protocols, computational resources, and expert interpretation.4 Differences in imaging workflows between institutions further restrict their routine use, particularly where resources are limited. Diagnostic laparoscopy allows direct inspection of disease distribution, but its invasiveness and procedural risk make serial assessment impractical.5 CA125 kinetics provide another approach: the CA125 elimination rate constant K (KELIM) is an accessible marker with reported value for NACT response and the surgical result at IDS.6,7 Its calculation, however, relies on repeated CA125 measurements across several cycles, so a dependable estimate may not be available early in treatment. KELIM is also less informative when baseline CA125 is low or the tumor does not produce CA125.
Host systemic inflammation and immune-nutritional status have been increasingly recognized as clinically relevant factors in ovarian cancer.8 Accordingly, blood-based biomarkers derived from routine laboratory parameters have gained increasing attention because they are readily accessible, cost-effective, and suitable for repeated assessment. Blood-based inflammatory indices such as neutrophil-to-lymphocyte ratio (NLR), platelet-to-lymphocyte ratio (PLR), monocyte-to-lymphocyte ratio (MLR), systemic immune-inflammation index (SII), and pan-immune-inflammation value (PIV) have been investigated in relation to prognosis, chemotherapy response, platinum resistance, and cytoreduction outcomes.9 Our previous study further demonstrated the value of peri-treatment assessment of systemic inflammatory indices by showing that PIV measured before NACT, after the first NACT cycle, and before IDS was associated with CGR at IDS.9 However, PIV and related indices are derived entirely from circulating blood cell counts and therefore predominantly reflect systemic inflammatory balance, without directly incorporating host nutritional or anemia-related status. The hemoglobin, albumin, lymphocyte, and platelet (HALP) score may address this limitation by integrating hemoglobin, albumin, lymphocyte count, and platelet count into a single index reflecting anemia-related physiological status, nutritional reserve, immune competence, and platelet-associated inflammatory activity.10 This integrated profile may be particularly relevant in advanced ovarian cancer, where nutritional impairment and systemic inflammation frequently coexist.11 Although HALP has been studied across multiple malignancies,10 evidence in ovarian cancer remains limited and has mainly focused on baseline or preoperative HALP.12,13 The peri-NACT profile of HALP and its association with CGR at IDS remain unclear.
We therefore evaluated three HALP measures, obtained before NACT, before IDS, and as their difference, in relation to CGR at IDS in advanced ovarian cancer. The associations between HALP-related indices and established NACT response indicators, including KELIM and chemotherapy response score, as well as survival outcomes, were also assessed.
Methods
Patients
The cohort comprised patients treated with first-line platinum-based NACT and subsequent IDS at Peking Union Medical College Hospital from January 2018 through December 2023. Eligibility required high-grade serous carcinoma of ovarian, fallopian tube, or primary peritoneal origin; International Federation of Gynecology and Obstetrics (FIGO) stage IIIC-IV disease; NACT delivered at our center; and IDS after 2–4 NACT cycles. Complete clinical records and follow-up information were also required. We excluded patients with other histologic subtypes, one or more than four NACT cycles, another concurrent malignancy, acute or chronic infection, hematologic or autoimmune disease, or missing laboratory measurements at a required timepoint.
The final analytic cohort partially overlapped with that of our previously published PIV study.9 However, the present study applied different eligibility and laboratory-data requirements, addressed a distinct biomarker-related research question, and followed a separate analysis plan. None of the statistical results or conclusions reported in the previous PIV study are duplicated here. The Institutional Review Board of Peking Union Medical College Hospital approved this retrospective study (approval number: K6963), which was performed in accordance with the Declaration of Helsinki. Given the retrospective design, the requirement for informed consent was waived. All patient data were de-identified before analysis and handled in a secure manner to protect confidentiality.
Data Collection and Definitions
Data Source and Collection
Two investigators independently abstracted the required information from electronic medical records. The source materials included inpatient notes, laboratory and pathology reports, operative documentation, treatment records, and follow-up files, and discrepancies were checked against the source record.
Demographic and Clinical Variables
Demographic and clinical variables included age at diagnosis, body mass index (BMI), medical comorbidities, American Society of Anesthesiologists (ASA) physical status, FIGO stage, ascites volume, NACT regimen, number of NACT cycles, adjuvant chemotherapy (ACT) regimen, total number of chemotherapy cycles, bevacizumab use, and poly (adenosine diphosphate-ribose) polymerase (PARP) inhibitor use. Medical comorbidities included hypertension, cardiovascular or cerebrovascular disease, diabetes, chronic liver disease, chronic lung disease, chronic kidney disease, and other chronic conditions recorded at baseline.
Laboratory Parameters and HALP Calculation
KELIM was derived from serial CA125 measurements collected during NACT through the web-based calculator available at https://www.biomarker-kinetics.org/CA-1256 Values of 1 or above were categorized as favorable, corresponding to more rapid CA125 clearance and greater chemosensitivity. Laboratory parameters were extracted at two predefined treatment timepoints: before the initiation of NACT and before IDS. Baseline measurements were taken no more than 2 weeks before the first NACT cycle, and preoperative values were obtained within 1 week before IDS whenever available. The extracted parameters included hemoglobin, serum albumin, lymphocyte count, and platelet count. The HALP score was calculated using the formula: HALP = hemoglobin (g/L) × albumin (g/L) × lymphocyte count (× 109/L)/ platelet count (× 109/L). HALP before NACT and HALP before IDS were calculated from laboratory values at the corresponding timepoints. HALP difference was calculated as HALP before IDS minus HALP before NACT, representing the absolute change in HALP during NACT.
Surgical and Pathological Characteristics
Operative details were abstracted from the surgical reports. Cytoreduction status was determined according to the residual tumor documented at the completion of IDS. CGR was defined as no visible residual disease after surgery, whereas non-CGR was defined as any macroscopic residual tumor remaining after surgery. Procedure burden was graded with the Aletti surgical complexity score.14 Complications occurring within 30 days were classified by Clavien-Dindo grade.15
Pathological response to NACT was evaluated using the chemotherapy response score (CRS) based on histological findings in the omentum at IDS.16 CRS 1 represented absent or minimal regression, CRS 2 indicated an appreciable response with viable tumor remaining, and CRS 3 denoted complete or near-complete regression. For analytical purposes, CRS 1 and CRS 2 were grouped together and compared with CRS 3.
Follow-Up and Study Outcomes
Follow-up visits were scheduled at 3-month intervals for 2 years, at 6-month intervals in years 3–5, and yearly thereafter. The follow-up database was closed in June 2025. Disease progression or recurrence was assessed according to Response Evaluation Criteria in Solid Tumors or CA125 progression criteria defined by the Gynecological Cancer InterGroup.17
The primary objective was to evaluate the associations and discriminatory performance of HALP before NACT, HALP before IDS, and HALP difference in relation to CGR at IDS. Secondary analyses explored how HALP-related measures corresponded to KELIM, CRS, and survival. Progression-free survival (PFS) extended from the first NACT treatment to progression, recurrence, or death from any cause. Overall survival (OS) extended from the first NACT treatment to death from any cause. For participants without the relevant event, observations were censored on the date of the latest clinic visit or telephone follow-up.
Statistical Analysis
As this was an exploratory study and no prior effect-size estimates were available in this specific clinical context, a prespecified sample-size calculation was not performed. A post hoc power analysis was conducted using independent-samples t tests (α = 0.05, two-tailed) to evaluate the adequacy of our sample size for detecting differences in HALP parameters between the CGR and non-CGR groups. The analysis showed > 99.9% power for HALP before NACT and HALP before IDS, with large effect sizes (Cohen’s d = 0.89 and 1.10, respectively), whereas HALP difference yielded 73.0% power with a smaller effect size (Cohen’s d = 0.35) (Supplementary Table S1). Although power was lower for HALP difference, the corresponding findings were interpreted cautiously, while the HALP before NACT and HALP before IDS comparisons were well powered.
Continuous variables were summarized as medians with interquartile ranges (IQRs) or means with standard deviations (SDs); categorical variables were summarized as counts and percentages. Group differences were tested with Student’s t-test, the Mann–Whitney U-test, Fisher’s exact test, or the χ2-test, as appropriate.
Receiver operating characteristic (ROC) curves were generated for the three HALP measures to determine how well each distinguished CGR from non-CGR at IDS. Discrimination was summarized by the area under the curve (AUC) and its 95% confidence interval (CI). For each measure, the Youden index identified the cutoff that jointly optimized sensitivity and specificity. The AUCs and cutoffs were internally validated using 1000 stratified bootstrap resamples, with mean AUCs and median cutoffs reported alongside their 95% CIs. The original cohort-derived cutoffs were retained to categorize patients into higher and lower HALP groups for the association analyses. Dynamic changes in HALP status from before NACT to before IDS were displayed using a Sankey diagram. To further evaluate the combined effect of baseline and preoperative HALP status, patients were categorized into four groups according to HALP before NACT and HALP before IDS. CGR rates were compared across these combined HALP strata.
Candidate preoperative factors for non-CGR were screened in separate univariate logistic models, and factors with P ≤ 0.1 were carried forward to the multivariable analysis. NACT cycle number was additionally included in all multivariable specifications, irrespective of its univariate P value, to address potential treatment heterogeneity. The primary model simultaneously included HALP before NACT, HALP before IDS, and HALP difference. To assess the robustness of the results to the potential mathematical coupling arising from the simultaneous inclusion of HALP before NACT, HALP before IDS, and their difference, four sensitivity models were fitted using the same non-HALP covariates. Model 2 included HALP before NACT and HALP before IDS, whereas model 3, 4, and 5 included HALP before NACT alone, HALP before IDS alone, and HALP difference alone, respectively. Associations are expressed as odds ratios (ORs) with 95% CIs. Given the potential correlation among HALP-related indices, multicollinearity among variables included in the multivariable logistic regression model was assessed using variance inflation factors. The number of non-CGR events per estimated parameter ranged from 11.1 to 14.8 across model 1–5, indicating an acceptable balance between the event count and model complexity. Exploratory subgroup analyses were performed across preoperative factors included in the logistic regression analysis, with unadjusted ORs and 95% CIs calculated within each subgroup. Time-to-event distributions were estimated by the Kaplan-Meier method and compared with Log rank tests. Cox proportional hazards models were fitted separately for PFS and OS. Variables with P ≤ 0.1 in univariate Cox analysis were included in multivariate models. Hazard ratios (HRs) with 95% CIs were calculated.
Analyses and figures were produced in R (version 4.0.0) together with GraphPad Prism (version 10.0). All tests were two-tailed, and P < 0.05 was considered statistically significant.
Results
Patient Characteristics
A total of 239 patients were included in the final analysis after applying the eligibility and exclusion criteria. Among them, 150 patients achieved complete gross resection (CGR) after IDS, whereas 89 patients had macroscopic residual disease and were classified as non-CGR (Figure 1).
Figure 1.
Patient selection and enrollment flowchart.
Abbreviations: IDS, interval debulking surgery; NACT, neoadjuvant chemotherapy; FIGO, International Federation of Gynecology and Obstetrics; CGR, complete gross resection.
Cohort characteristics are presented in Table 1. The two surgical-outcome groups were broadly similar with respect to demographic profile, baseline disease features, and NACT treatment characteristics.
Table 1.
Patient Characteristics According to Cytoreduction Status
| Characteristic | Total (n = 239) | CGR (n = 150) | Non-CGR (n = 89) | P |
|---|---|---|---|---|
| Age (years) | 0.147 | |||
| <65 | 169 (70.7) | 111 (74.0) | 58 (65.2) | |
| ≥65 | 70 (29.3) | 39 (26.0) | 31 (34.8) | |
| BMI (kg/m2) | 0.092 | |||
| <25 | 181 (75.7) | 119 (79.3) | 62 (69.7) | |
| ≥25 | 58 (24.3) | 31 (20.7) | 27 (30.3) | |
| Medical comorbidities | 0.429 | |||
| No | 150 (62.8) | 97 (64.7) | 53 (59.5) | |
| Yes | 89 (37.2) | 53 (35.3) | 36 (40.5) | |
| ASA grade | 0.092 | |||
| 1 | 6 (2.5) | 5 (3.3) | 1 (1.1) | |
| 2 | 185 (77.4) | 121 (80.7) | 64 (71.9) | |
| 3 | 48 (20.1) | 24 (16.0) | 24 (27.0) | |
| FIGO stage | 0.512 | |||
| IIIC | 154 (64.4) | 99 (66.0) | 55 (61.8) | |
| IV | 85 (35.6) | 51 (34.0) | 34 (38.2) | |
| Ascites ≥ 1000 mL | 0.905 | |||
| No | 79 (33.0) | 50 (33.3) | 29 (32.6) | |
| Yes | 160 (67.0) | 100 (66.7) | 60 (67.4) | |
| NACT regimen | 1.000 | |||
| Doxorubicin/platinum | 3 (1.3) | 2 (1.3) | 1 (1.1) | |
| Taxane/platinum | 236 (98.7) | 148 (98.7) | 88 (98.9) | |
| NACT cycles | 0.582 | |||
| 2 | 83 (34.7) | 52 (34.7) | 31 (34.8) | |
| 3 | 128 (53.6) | 78 (52.0) | 50 (56.2) | |
| 4 | 28 (11.7) | 20 (13.3) | 8 (9.0) | |
| CA125 before NACT (median and IQR, U/mL) | 1147.0 (521.9–2143.0) | 1154.0 (564.2–2201.8) | 1104.0 (481.0–1993.0) | 0.636 |
| CA125 before IDS (median and IQR, U/mL) | 63.9 (26.1–139.8) | 48.5 (20.0–102.5) | 103.0 (47.7–214.2) | < 0.001 |
| KELIM score | < 0.001 | |||
| <1 | 92 (38.5) | 43 (28.7) | 49 (55.1) | |
| ≥1 | 147 (61.5) | 107 (71.3) | 40 (44.9) | |
| CRS | < 0.001 | |||
| 1/2 | 187 (78.2) | 103 (68.7) | 84 (94.4) | |
| 3 | 52 (21.8) | 47 (31.3) | 5 (5.6) | |
| Surgical complexity score | 0.353 | |||
| High | 25 (10.5) | 16 (10.7) | 9 (10.1) | |
| Intermediate | 159 (66.5) | 104 (69.3) | 55 (61.8) | |
| Low | 55 (23.0) | 30 (20.0) | 25 (28.1) | |
| 30-day postoperative complication | 0.722 | |||
| Clavien-Dindo 1-2 | 40 (16.7) | 24 (16.0) | 16 (18.0) | |
| Clavien-Dindo 3 | 6 (2.5) | 3 (2.0) | 3 (3.4) | |
| None | 193 (80.8) | 123 (82.0) | 70 (78.6) | |
| ACT regimen | 0.279 | |||
| Doxorubicin/platinum | 6 (2.5) | 2 (1.3) | 4 (4.5) | |
| Taxane/platinum | 233 (97.5) | 148 (98.7) | 85 (95.5) | |
| Total chemotherapy cycles | 0.456 | |||
| <6 | 4 (1.7) | 2 (1.3) | 2 (2.2) | |
| 6 | 62 (25.9) | 43 (28.7) | 19 (21.4) | |
| >6 | 173 (72.4) | 105 (70.0) | 68 (76.4) | |
| Bevacizumab use | 0.013 | |||
| No | 165 (69.0) | 95 (63.3) | 70 (78.7) | |
| Yes | 74 (31.0) | 55 (36.7) | 19 (21.3) | |
| PARP inhibitor use | < 0.001 | |||
| No | 122 (51.0) | 62 (41.3) | 60 (67.4) | |
| Yes | 117 (49.0) | 88 (58.7) | 29 (32.6) | |
| HALP before NACT (median and IQR) | 15.7 (9.9–25.9) | 20.8 (11.9–30.6) | 11.4 (7.9–15.7) | < 0.001 |
| HALP before IDS (median and IQR) | 33.3 (24.0–44.5) | 39.4 (30.5–50.6) | 23.3 (18.3–32.2) | < 0.001 |
| HALP difference (median and IQR) | 14.6 (5.8–25.9) | 17.6 (7.0–30.2) | 12.0 (3.5–22.0) | 0.017 |
Notes: Unless specified otherwise, data are n (%).
Abbreviations: CGR, complete gross resection; BMI, body mass index; ASA, American Society of Anesthesiologists; FIGO, International Federation of Gynecology and Obstetrics; NACT, neoadjuvant chemotherapy; IQR, interquartile range; IDS, interval debulking surgery; KELIM, CA125 elimination rate constant K; CRS, Chemotherapy Response Score; ACT, adjuvant chemotherapy; PARP, poly (adenosine diphosphate-ribose) polymerase; HALP, hemoglobin, albumin, lymphocyte, and platelet score.
Response measures collected after NACT showed clear differences between the cytoreduction groups. CA125 before IDS was significantly lower in the CGR group than in the non-CGR group (median 48.5 vs 103.0 U/mL, P < 0.001). Favorable KELIM scores (≥ 1) occurred in 71.3% of the CGR group and 44.9% of the non-CGR group (P < 0.001). CRS 3 occurred in 31.3% of patients with CGR versus 5.6% of those with non-CGR, confirming a group difference in pathological response (P < 0.001). Surgical complexity scores and 30-day postoperative complication rates were similar between groups. Postoperative bevacizumab use and PARP inhibitor use were more frequent in the CGR group than in the non-CGR group (36.7% vs 21.3%, P = 0.013; 58.7% vs 32.6%, P < 0.001, respectively). The cohort was followed for a median of 35.9 months (range 6.9–91.8); the corresponding medians were 35.8 months (range 6.9–91.8) for CGR and 36.4 months (range 6.9–91.8) for non-CGR.
HALP-Related Indices Were Associated with Cytoreduction Status
HALP-related indices differed significantly between patients with CGR and non-CGR (Table 1). In the overall cohort, median HALP increased from 15.7 before NACT to 33.3 before IDS, indicating an overall improvement in HALP status during treatment. This increase was more pronounced in patients who achieved CGR, whose median HALP rose from 20.8 before NACT to 39.4 before IDS, compared with an increase from 11.4 to 23.3 in the non-CGR group. Accordingly, median values were higher in the CGR group for HALP before NACT and before IDS (both P < 0.001) and for HALP difference (17.6 vs 12.0, P = 0.017).
ROC analysis showed moderate discrimination for HALP before NACT (AUC 0.761, 95% CI 0.701–0.822, cutoff 19.3) and greater discrimination for HALP before IDS (AUC 0.816, 95% CI 0.762–0.871, cutoff 33.4), whereas HALP difference showed limited discrimination (AUC 0.593, 95% CI 0.521–0.665, cutoff 30.5) (Figure 2A–C). Bootstrap internal validation yielded mean AUCs of 0.761 (95% CI 0.705–0.817), 0.817 (95% CI 0.759–0.867), and 0.591 (95% CI 0.528–0.660), respectively, closely approximating the corresponding apparent AUCs. The median bootstrap cutoffs were 19.3 (95% CI 10.1–20.5), 33.4 (95% CI 23.4–39.0), and 22.9 (95% CI 9.2–36.4), respectively, and all original cohort-derived cutoffs fell within their corresponding 95% bootstrap CIs (Supplementary Table S2). Therefore, the original cohort-derived cutoffs were retained for subsequent analyses. We further analyzed changes in categorical HALP status before and after NACT (Figure 2D and E). The proportion of patients with high HALP increased from 39.3% before NACT to 49.8% before IDS, suggesting that some patients shifted to a more favorable HALP status during treatment. When pre-NACT and pre-IDS HALP status were considered together, CGR rates showed a clear gradient across the four strata. Patients with persistently low HALP had the lowest CGR rate (29.2%), whereas those with persistently high HALP had the highest CGR rate (98.4%). Importantly, patients with high HALP at either timepoint also showed relatively high CGR rates, including 71.0% in the high-before-NACT/low-before-IDS group and 71.4% in the low-before-NACT/high-before-IDS group (P for trend < 0.001).
Figure 2.
Discriminatory performance of HALP-related indices for complete gross resection after IDS. (A–C) ROC curves for HALP before NACT, HALP before IDS, and HALP difference in distinguishing cytoreduction status. Red circles and arrows indicate the optimal cutoff points determined using the Youden index; (D) Sankey diagram showing changes in HALP status from before NACT to before IDS and subsequent cytoreduction status; (E) CGR rates across combined HALP strata.
Abbreviations: HALP, hemoglobin, albumin, lymphocyte, and platelet score; IDS, interval debulking surgery; ROC, receiver operating characteristic; NACT, neoadjuvant chemotherapy; CGR, complete gross resection; AUC, area under the curve; CI, confidence interval.
Univariate and multivariate logistic regression analyses were performed to identify preoperative factors associated with non-CGR (Table 2, model 1 including HALP before NACT, HALP before IDS, and HALP difference). In univariate analysis, higher CA125 before IDS was associated with an increased likelihood of non-CGR, whereas favorable KELIM status (≥ 1) and higher values of all three HALP-related indices were associated with lower odds of non-CGR. In the multivariable model, CA125 before IDS ≥ 100 U/mL was associated with higher odds of non-CGR (OR 2.20, 95% CI 1.03–4.71, P = 0.042), whereas KELIM ≥ 1 was associated with lower odds (OR 0.39, 95% CI 0.18–0.85, P = 0.018). Compared with 2 NACT cycles, 3 cycles were not associated with non-CGR (OR 0.95, 95% CI 0.44–2.08, P = 0.904), whereas 4 cycles were associated with lower odds of non-CGR (OR 0.24, 95% CI 0.07–0.82, P = 0.023). Importantly, both HALP before NACT > 19.3 and HALP before IDS > 33.4 retained independent associations with lower odds of non-CGR (OR 0.11, 95% CI 0.05–0.26, P < 0.001; OR 0.15, 95% CI 0.06–0.35, P < 0.001, respectively). In contrast, HALP difference > 30.5 was significant in univariate analysis but lost significance in the multivariate model (OR 0.60, 95% CI 0.18–1.99, P = 0.401). No substantial multicollinearity was observed among variables included in the primary multivariable model, with all variance inflation factors ≤ 1.48. In the sensitivity analyses, HALP before NACT and HALP before IDS remained independently associated with non-CGR when modeled jointly (model 2) or separately (model 3 and 4) (Supplementary Table S3–S5). HALP difference was also independently associated with lower odds of non-CGR when included as the sole HALP-related variable in the adjusted model (model 5: OR 0.21, 95% CI 0.08–0.54, P = 0.001; Supplementary Table S6). Subgroup analyses showed that high HALP-related indices, including HALP before NACT > 19.3, HALP before IDS > 33.4, and HALP difference > 30.5, were consistently associated with lower odds of non-CGR across preoperative factors, although some subgroup estimates did not reach statistical significance (Supplementary Figure S1).
Table 2.
Univariate and Multivariate Logistic Regression Analysis of Preoperative Factors Associated with Non-CGR (Model 1: HALP Before NACT, HALP Before IDS, and HALP Difference Included Jointly)
| Characteristic | Univariate Logistic Regression Analysis | Multivariate Logistic Regression Analysis | ||
|---|---|---|---|---|
| OR (95% CI) | P | OR (95% CI) | P | |
| Age (years) | ||||
| <65 | Reference | |||
| ≥65 | 1.52 (0.86–2.69) | 0.148 | ||
| BMI (kg/m2) | ||||
| <25 | Reference | Reference | ||
| ≥25 | 1.67 (0.92–3.05) | 0.093 | 1.55 (0.70–3.40) | 0.279 |
| ASA grade | ||||
| 1 | Reference | |||
| 2 | 2.64 (0.30–23.12) | 0.379 | ||
| 3 | 5.00 (0.54 ~ 46.05) | 0.155 | ||
| FIGO stage | ||||
| IIIC | Reference | |||
| IV | 1.20 (0.70–2.07) | 0.512 | ||
| Ascites ≥ 1000 mL | ||||
| No | Reference | |||
| Yes | 1.03 (0.59–1.81) | 0.905 | ||
| NACT cycles | ||||
| 2 | Reference | Reference | ||
| 3 | 1.08 (0.61–1.90) | 0.803 | 0.95 (0.44–2.08) | 0.904 |
| 4 | 0.67 (0.26–1.71) | 0.402 | 0.24 (0.07–0.82) | 0.023 |
| CA125 before NACT (U/mL) | ||||
| <1000 | Reference | |||
| ≥1000 | 0.90 (0.53–1.52) | 0.681 | ||
| CA125 before IDS (U/mL) | ||||
| <100 | Reference | Reference | ||
| ≥100 | 2.91 (1.67–5.06) | < 0.001 | 2.20 (1.03–4.71) | 0.042 |
| KELIM score | ||||
| <1 | Reference | Reference | ||
| ≥1 | 0.33 (0.19–0.57) | < 0.001 | 0.39 (0.18–0.85) | 0.018 |
| Surgical complexity score | ||||
| High | Reference | |||
| Intermediate | 0.94 (0.39–2.27) | 0.891 | ||
| Low | 1.48 (0.56–3.92) | 0.429 | ||
| HALP before NACT | ||||
| ≤19.3 | Reference | Reference | ||
| >19.3 | 0.10 (0.05–0.21) | < 0.001 | 0.11 (0.05–0.26) | < 0.001 |
| HALP before IDS | ||||
| ≤33.4 | Reference | Reference | ||
| >33.4 | 0.11 (0.06–0.21) | < 0.001 | 0.15 (0.06–0.35) | < 0.001 |
| HALP difference | ||||
| ≤30.5 | Reference | Reference | ||
| >30.5 | 0.21 (0.09–0.53) | < 0.001 | 0.60 (0.18–1.99) | 0.401 |
Abbreviations: CGR, complete gross resection; OR, odds ratio; CI, confidence interval; BMI, body mass index; ASA, American Society of Anesthesiologists; FIGO, International Federation of Gynecology and Obstetrics; NACT, neoadjuvant chemotherapy; IDS, interval debulking surgery; KELIM, CA125 elimination rate constant K; HALP, hemoglobin, albumin, lymphocyte, and platelet score.
Association Between HALP-Related Indices and NACT Response Indicators
Given the observed association between HALP-related indices and CGR, we further explored whether HALP-related indices were associated with established NACT response indicators, including KELIM score and CRS (Figure 3). Overall, patients in the high HALP groups tended to show more favorable NACT response profiles, as reflected by higher proportions of KELIM ≥ 1 and CRS 3 across all three HALP-related indices. The association between HALP before NACT and KELIM score reached statistical significance, with favorable KELIM scores observed in 72.3% of patients with HALP before NACT > 19.3 compared with 54.5% of those with HALP before NACT ≤ 19.3 (P = 0.006). HALP before IDS was significantly associated with pathological response, with CRS 3 observed in 28.6% of patients with HALP before IDS > 33.4 compared with 15.0% of those with HALP before IDS ≤ 33.4 (P = 0.011).
Figure 3.
Association between HALP-related indices and NACT response indicators. (A) KELIM scores; (B) CRS grades.
Abbreviations: HALP, hemoglobin, albumin, lymphocyte, and platelet score; NACT, neoadjuvant chemotherapy; KELIM, CA125 elimination rate constant K; CRS, Chemotherapy Response Score; IDS, interval debulking surgery.
Prognostic Relevance of HALP-Related Indices
Survival curves were compared across cytoreduction outcomes and HALP categories (Figure 4). CGR was associated with longer PFS (P < 0.001) and OS (P = 0.003) relative to non-CGR (Figure 4A and B). HALP-related indices also stratified patients by survival outcomes. For each HALP measure, the higher-value group had more favorable PFS and OS curves than the lower-value group (Figure 4C–H). The respective log-rank P values for PFS and OS were < 0.001 and 0.002 for HALP before NACT, < 0.001 and < 0.001 for HALP before IDS, and 0.026 and 0.011 for HALP difference. Cox regression analyses further evaluated the associations between HALP-related indices and survival (Supplementary Table S7). In multivariable analysis, HALP before NACT > 19.3 was independently associated with longer PFS (HR 0.65, 95% CI 0.44–0.95, P = 0.028), whereas none of the HALP-related indices was independently associated with OS. To address potential overadjustment, sensitivity analyses were performed after excluding CGR status, total chemotherapy cycles, PARP inhibitor use, and bevacizumab use, which were considered potential post-treatment variables. After excluding these variables, the effect estimates for all three HALP-related indices shifted further below 1 for both PFS and OS. The association between HALP before NACT and PFS was strengthened (HR 0.59, 95% CI 0.41–0.85, P = 0.005), whereas HALP before IDS and HALP difference remained nonsignificant. For OS, HALP before NACT (HR 0.63, 95% CI 0.36–1.08, P = 0.095) and HALP before IDS (HR 0.56, 95% CI 0.31–1.02, P = 0.058) showed stronger associations, although neither reached statistical significance (Supplementary Table S8).
Figure 4.
Survival outcomes stratified by cytoreduction status and HALP-related indices. Kaplan-Meier curves for PFS and OS according to (A, B) cytoreduction status; (C, D) HALP before NACT; (E, F) HALP before IDS; and (G, H) HALP difference.
Abbreviations: HALP, hemoglobin, albumin, lymphocyte, and platelet score; PFS, progression-free survival; OS, overall survival; NACT, neoadjuvant chemotherapy; IDS, interval debulking surgery; CGR, complete gross resection; No, number.
Discussion
This retrospective analysis links peri-treatment host nutritional, inflammatory, and immune status with the surgical result after NACT for advanced ovarian cancer. Using HALP as an integrated blood-based marker, we found that patients with more favorable HALP profiles before treatment and before IDS were more likely to achieve CGR. The combined pre/post HALP status showed differences in CGR rates across the four groups. In addition, the observed associations with KELIM, CRS, and survival outcomes suggest that HALP-related indices may reflect clinically relevant aspects of treatment response and patient prognosis. Taken together, HALP may offer a readily obtained peri-treatment measure to support preoperative estimation of CGR after NACT.
Within the NACT treatment pathway, the value of IDS is closely tied to leaving no visible residual disease. Identifying patients who are likely to achieve CGR before IDS is therefore important for treatment planning, surgical timing, patient counseling, and risk stratification. Current approaches for estimating cytoreducibility include KELIM based on CA125 kinetics, radiologic assessment, and composite clinical prediction models. Although these methods provide valuable information, each has inherent limitations. KELIM requires serial CA125 measurements and is inapplicable in CA125-negative cases.6,18 Imaging-based assessment is essential for mapping disease distribution, but small-volume residual lesions and treatment-related tissue changes may still be difficult to evaluate accurately.4,19 In addition, some prediction models rely on detailed radiologic, laparoscopic, or intraoperative variables, which may reduce their feasibility in routine practice or across different centers.5 These limitations highlight the need for new and readily accessible biomarkers for preoperative prediction of CGR.
Low cost and routine availability have made blood-derived inflammatory indices attractive for studying prognosis and treatment response in oncology. Ovarian cancer studies have evaluated NLR, PLR, MLR, SII, and PIV against survival, chemotherapy response, platinum resistance, and cytoreduction status.9,20–22 However, these indices are derived from two to four circulating blood cell populations and primarily reflect systemic inflammatory or immune cell balance. This has prompted interest in composite indices that incorporate additional dimensions of host status beyond inflammation alone. HALP extends this concept by placing hemoglobin and albumin together with lymphocyte and platelet counts in one measure. By incorporating hemoglobin and albumin, HALP captures anemia-related physiological status and nutritional reserve in addition to immune and platelet-related inflammatory activity.10 Its nutritional component may be especially relevant in advanced ovarian cancer, where ascites, peritoneal spread, gastrointestinal dysfunction, and systemic inflammation can promote pretreatment malnutrition and hypoalbuminemia.11 Changes in nutrition-related parameters during NACT may parallel tumor control, attenuation of tumor-related inflammation, and restoration of host reserve.23 This is clinically important because impaired nutritional status has been associated with poorer treatment tolerance, postoperative morbidity, and unfavorable oncologic outcomes across multiple malignancies.24 In ovarian cancer, albumin-based and lymphocyte-based indices such as prognostic nutritional index (PNI) have been linked to survival, and recent evidence suggests that nutritional indices may also be associated with response to NACT.23 However, HALP is not a tumor-specific biomarker. HALP and its individual components can be influenced by nutritional status, anemia, systemic inflammation, chemotherapy-related myelosuppression, transfusion, albumin administration, and other supportive care. Accordingly, the observed associations of HALP with CGR, KELIM, and CRS in the present study may reflect overall host fitness and treatment-related changes in addition to tumor response, rather than a direct measure of tumor burden, resectability, or chemosensitivity.
HALP has been increasingly studied across multiple solid tumors. In general, lower HALP levels have been associated with unfavorable survival outcomes and adverse clinicopathological features, supporting its potential value as a marker of host nutritional-inflammatory status.10 Consistent with these observations, HALP-related indices stratified PFS and OS in Kaplan-Meier analyses in our cohort. However, these associations were attenuated after multivariable adjustment. This may be because HALP is not a direct tumor-specific determinant of survival but instead reflects nutritional reserve, systemic inflammation, treatment tolerance, and overall host condition. These factors are also related to chemotherapy response, the likelihood of achieving CGR, and the receipt of subsequent treatment, through which HALP may be indirectly associated with survival. Accordingly, adjustment for CGR, total chemotherapy cycles, and subsequent therapies may have attenuated part of this indirect association. This interpretation was supported by the stronger effect estimates observed after these variables were excluded. HALP before NACT nevertheless remained independently associated with PFS, suggesting that pretreatment host condition may retain prognostic relevance beyond these treatment-related pathways. In contrast, HALP before IDS and HALP difference may be more strongly influenced by treatment-related changes and therefore showed limited independent associations with survival. Evidence on HALP in ovarian cancer remains limited. One ovarian cancer study found lower HALP at more advanced stages,12 while another linked a lower preoperative value to suboptimal cytoreduction and poorer survival in advanced epithelial disease.13 Our study extends these observations by focusing on CGR, a clinically meaningful surgical endpoint after IDS, and by evaluating HALP across multiple treatment-related measures, including HALP before NACT, HALP before IDS, HALP difference, and combined pre/post HALP status. HALP before NACT and HALP before IDS were both independently associated with CGR, whereas HALP difference showed limited discriminatory performance. When HALP status at the two timepoints was considered jointly, CGR rates differed across the four groups. Patients with low HALP at both timepoints had the lowest CGR rate, whereas those with high HALP at both timepoints had the highest rate. The two groups with discordant HALP status showed similar CGR rates, regardless of whether HALP changed from low to high or from high to low. These findings suggest that considering HALP status at both timepoints may identify clinically distinct groups, but they do not indicate that the direction or magnitude of change during NACT is independently informative. However, because this combined analysis was based on cohort-derived cutoffs, it should be considered exploratory and interpreted cautiously.
The associations between HALP-related indices and NACT response markers further support the clinical relevance of HALP in this setting. In our cohort, patients with higher HALP generally showed a more favorable response profile, although not all comparisons reached statistical significance. HALP before NACT was significantly associated with favorable KELIM, suggesting that baseline host nutritional-inflammatory-immune status may be related to early chemotherapy sensitivity. By contrast, HALP before IDS was significantly associated with CRS 3, indicating that HALP measured closer to surgery may better reflect the pathological response achieved after NACT. Prior ovarian cancer evidence relates NLR, PLR, MLR, SII, and PIV to response, platinum resistance, and survival.9,20–22 The PNI and controlling nutritional status (CONUT) score have likewise shown associations with stage, chemotherapy response, cytoreduction, platinum sensitivity, and survival.23,25 Higher HALP values may therefore identify patients with a more favorable host profile for chemotherapy response and subsequent complete cytoreduction. Given its exploratory, association-focused design, the present study was not intended to establish causality or to determine whether HALP provides additional predictive value beyond established response measures (such as KELIM and CRS) or other clinical and biochemical variables. These questions require evaluation in further studies.
As a routinely available blood-based index that can be measured repeatedly during NACT, HALP may identify patients whose nutritional and inflammatory status remains unfavorable despite treatment. When persistently low HALP is accompanied by poor CA125 kinetics or radiologic concern for residual disease, more comprehensive reassessment before surgery may be warranted. Low HALP may also highlight potentially modifiable host-related factors, including malnutrition, systemic inflammatory burden, anemia-related physiological reserve, and reduced functional capacity. These considerations may support nutritional assessment, anemia and inflammation management, and prehabilitation-oriented care.26,27 In addition, persistently low HALP together with poor treatment response may prompt discussion of treatment modification within established guideline-based frameworks, such as additional NACT cycles or alternative systemic regimens.28–30 However, the present findings demonstrate association rather than clinical utility and do not establish that HALP-guided management improves surgical or oncologic outcomes. The role of HALP in guiding preoperative reassessment, supportive optimization, or treatment modification therefore remains theoretical and requires prospective evaluation.
Despite these findings, several limitations should be acknowledged. First, the clinical implementation of HALP is subject to methodological and validation challenges. Although HALP is derived from routine laboratory tests, between-center reproducibility may be affected by differences in collection procedures, timing, fasting conditions, analytic platforms, reagents, processing, and storage.31 The cutoff values were derived and evaluated in the same cohort. Although bootstrap internal validation yielded AUC estimates similar to the apparent estimates, internal resampling cannot establish whether the discriminatory performance and cutoff values are generalizable to other populations. Moreover, although the event-to-parameter ratios were acceptable and alternative logistic specifications yielded consistent findings, the modest sample size and exploratory, data-driven analyses leave a residual risk of overfitting and instability in the adjusted association estimates. Second, the retrospective single-center design is vulnerable to selection bias, information bias, and unmeasured confounding. Center-specific patient characteristics, treatment practices, and surgical decision-making may limit the generalizability of the observed associations and cohort-derived cutoffs. Exclusion of patients with missing laboratory measurements or insufficient follow-up may have further reduced cohort representativeness, while changes in clinical practice during the six-year enrollment period may have introduced temporal bias. Although nearly all patients received taxane/platinum-based NACT and cycle number was included in the multivariable logistic analyses, dose intensity, dose reductions, treatment delays, and supportive care were not systematically captured. Differences among surgeons in assessing and documenting residual disease may also have introduced outcome misclassification. In addition, HALP may be influenced by subclinical infection, undiagnosed inflammatory or autoimmune conditions, nutritional support, anemia management, transfusion, albumin supplementation, granulocyte colony-stimulating factor use, chemotherapy-related myelosuppression, and other supportive treatments. Residual confounding by overall host condition and treatment-related factors therefore cannot be excluded. Third, this exploratory study was designed to examine associations rather than to develop or validate a clinical prediction model. Although HALP was evaluated alongside relevant clinical and biochemical variables in the adjusted analyses, no prespecified reference prediction model was constructed; therefore, formal analyses of incremental discrimination or calibration were not performed. The observed relationships of HALP with KELIM and CRS were intended to provide clinical context rather than head-to-head comparisons of predictive performance. Whether HALP provides additional predictive value beyond clinical or biochemical variables remains uncertain. HALP was also not directly compared with PIV or other inflammatory indices in a common analytic cohort, and its relative value requires further evaluation. Finally, the survival analyses were exploratory and require validation in independent cohorts. The OS analysis was additionally limited by the relatively small number of deaths.
Conclusions
Within this advanced ovarian cancer cohort, HALP measured before NACT and before IDS was associated with the likelihood of CGR at IDS, whereas HALP difference showed limited discriminatory performance. Combined pre/post HALP status further distinguished groups with different CGR rates. HALP is derived from inexpensive routine blood tests and may supplement established preoperative assessments, but it requires external validation before clinical use.
Funding Statement
Financial support was provided by the Fundamental Research Funds for Central Universities, Chinese Academy of Medical Sciences & Peking Union Medical College (3332026074), the Beijing Xisike Clinical Oncology Research Foundation (Y-zai2022/zd-0088 and Y-Gilead2024-ZD-0082), and the National Key Clinical Specialty Construction Project (U114000). Neither funding body was involved in designing the study, acquiring or analyzing the data, deciding whether to publish the findings, or drafting the manuscript.
Data Sharing Statement
The datasets used and/or analysed during the current study are available from the corresponding author (Fengzhi Feng) on reasonable request.
Ethical Approval
The Institutional Review Board of Peking Union Medical College Hospital authorized this retrospective investigation (K6963), which adhered to the principles of the Declaration of Helsinki. The Board granted a waiver of informed consent because the study used retrospectively collected data.
Author Contributions
All authors made a significant contribution to the work reported, whether that is in the conception, study design, execution, acquisition of data, analysis and interpretation, or in all these areas; took part in drafting, revising or critically reviewing the article; gave final approval of the version to be published; have agreed on the journal to which the article has been submitted; and agree to be accountable for all aspects of the work.
Disclosure
The authors declare no potential conflicts of interest.
References
- 1.Colombo N, Sessa C, du Bois A, et al. ESMO-ESGO consensus conference recommendations on ovarian cancer: pathology and molecular biology, early and advanced stages, borderline tumours and recurrent disease†. Ann Oncol. 2019;30(5):672–15. doi: 10.1093/annonc/mdz062 [DOI] [PubMed] [Google Scholar]
- 2.Vermeulen CKM, Tadesse W, Timmermans M, Kruitwagen RFPM, Walsh T. Only complete tumour resection after neoadjuvant chemotherapy offers benefit over suboptimal debulking in advanced ovarian cancer. Eur J Obstet Gynecol Reprod Biol. 2017;219:100–105. doi: 10.1016/j.ejogrb.2017.10.019 [DOI] [PubMed] [Google Scholar]
- 3.Yang SP, Chen JX, Xu JY, Lei J, Wu SG, Zhou J. The prognostic effect of residual tumor for advanced epithelial ovarian cancer treated with neoadjuvant chemotherapy or primary debulking surgery. Cancer Med. 2022;11(14):2836–2845. doi: 10.1002/cam4.4642 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Panico C, Avesani G, Zormpas-Petridis K, Rundo L, Nero C, Sala E. Radiomics and radiogenomics of ovarian cancer: implications for treatment monitoring and clinical management. Radiol Clin North Am. 2023;61(4):749–760. doi: 10.1016/j.rcl.2023.02.006 [DOI] [PubMed] [Google Scholar]
- 5.T GD, Cuccu I, DE Angelis E, et al. Laparoscopic prediction of primary cytoreducibility of epithelial ovarian cancer. Minerva Obstet Gynecol. 2024;76(6):548–557. doi: 10.23736/S2724-606X.24.05452-6 [DOI] [PubMed] [Google Scholar]
- 6.You B, Robelin P, Tod M, et al. CA-125 ELIMination Rate Constant K (KELIM) is a marker of chemosensitivity in patients with ovarian cancer: results from the Phase II CHIVA trial. Clin Cancer Res. 2020;26(17):4625–4632. doi: 10.1158/1078-0432.CCR-20-0054 [DOI] [PubMed] [Google Scholar]
- 7.Ambrosino FA, Pisano C, Napoli MD, et al. KELIM score: a marker of chemosensitivity in ovarian cancer. Cancer Treat Rev. 2026;143:103071. doi: 10.1016/j.ctrv.2025.103071 [DOI] [PubMed] [Google Scholar]
- 8.Mu J, Wu Y, Jiang C, Cai L, Li D, Cao J. Progress in applicability of scoring systems based on nutritional and inflammatory parameters for ovarian cancer. Front Nutr. 2022;9:809091. doi: 10.3389/fnut.2022.809091 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Su H, Tian M, Wang Y, et al. Exploratory analysis of serial pan-immune-inflammation value and its early dynamics during neoadjuvant chemotherapy as a predictor of complete cytoreduction at interval debulking surgery in advanced ovarian cancer: a retrospective cohort study. World J Surg Oncol. 2026;24(1):236. doi: 10.1186/s12957-026-04359-7 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Farag CM, Antar R, Akosman S, Ng M, Whalen MJ. What is hemoglobin, albumin, lymphocyte, platelet (HALP) score? A comprehensive literature review of HALP’s prognostic ability in different cancer types. Oncotarget. 2023;14:153–172. doi: 10.18632/oncotarget.28367 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Shao Z, Fu L, Lu Y, Zhou H, Zhu Y, Zhu T. Ovarian cancer and malnutrition: a literature review. Transl Cancer Res. 2025;14(5):3239–3254. doi: 10.21037/tcr-2025-758 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Genc SO, Gulturk EA, Kurt B. Predictive value of PLR, PNI, and HALP scores in ovarian cancer staging. Cancer Manag Res. 2025;17:293–299. doi: 10.2147/CMAR.S508442 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Atjimakul T, Saeaib N, Tunthanathip T, Thongsuksai P. Significance of pretreatment hemoglobin-albumin-lymphocyte-platelet index for the prediction of suboptimal surgery in epithelial ovarian cancer. World J Oncol. 2024;15(2):268–278. doi: 10.14740/wjon1778 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Carter M, Chen AR, Pitt JB, et al. Classification systems of surgical complexity: a scoping review of the literature. J Surg Res. 2025;306:570–579. doi: 10.1016/j.jss.2024.12.049 [DOI] [PubMed] [Google Scholar]
- 15.Golder H, Casanova D, Papalois V. Evaluation of the usefulness of the Clavien-Dindo classification of surgical complications. Cir Esp. 2023;101(9):637–642. doi: 10.1016/j.ciresp.2023.01.012 [DOI] [PubMed] [Google Scholar]
- 16.Böhm S, Faruqi A, Said I, et al. Chemotherapy response score: development and validation of a system to quantify histopathologic response to neoadjuvant chemotherapy in tubo-ovarian high-grade serous carcinoma. J Clin Oncol. 2015;33(22):2457–2463. doi: 10.1200/JCO.2014.60.5212 [DOI] [PubMed] [Google Scholar]
- 17.Rustin GJ, Vergote I, Eisenhauer E, et al. Definitions for response and progression in ovarian cancer clinical trials incorporating RECIST 1.1 and CA 125 agreed by the Gynecological Cancer Intergroup (GCIG). Int J Gynecol Cancer. 2011;21(2):419–423. doi: 10.1097/IGC.0b013e3182070f17 [DOI] [PubMed] [Google Scholar]
- 18.Li C, Cui Q, Wang X, Yao S, Tu H, Chen M. CA-125 elimination rate constant K (KELIM) as a promising predictor of complete cytoreduction after neoadjuvant chemotherapy in advanced ovarian cancer patients: a retrospective study from two Chinese hospitals. BMC Cancer. 2024;24(1):609. doi: 10.1186/s12885-024-12252-3 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19.Ye S, Wang Y, Li R, Shao Z, Zhu T. Predictive value of 18F-FDG-PET/CT for pathologic complete response (pCR) of lesions after neoadjuvant chemotherapy in advanced ovarian cancer. World J Surg Oncol. 2025;24(1):25. doi: 10.1186/s12957-025-04138-w [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Zhang CL, Jiang XC, Li Y, et al. Independent predictive value of blood inflammatory composite markers in ovarian cancer: recent clinical evidence and perspective focusing on NLR and PLR. J Ovarian Res. 2023;16(1):36. doi: 10.1186/s13048-023-01116-2 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Bektaş O, Bektaş K. Prognostic value of systemic inflammatory markers in predicting neoadjuvant platinum response and platinum resistance in advanced epithelial ovarian cancer: a peritoneal cancer index-based analysis. Gynecol Oncol. 2025;199:147–151. doi: 10.1016/j.ygyno.2025.07.001 [DOI] [PubMed] [Google Scholar]
- 22.Wang J, Yin S, Chen K. Predictive value of the systemic immune-inflammation index for the efficacy of neoadjuvant chemotherapy and prognosis in patients with stage III ovarian cancer-a retrospective cohort study. Gland Surg. 2022;11(10):1639–1646. doi: 10.21037/gs-22-459 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23.Rubio-Cordero J, Moreno-Flores G, Ramirez-Morales R, et al. Associations between the prognostic nutritional index and response to neoadjuvant chemotherapy in patients with epithelial ovarian cancer. Int J Gynecol Cancer. 2025;35(11):102109. doi: 10.1016/j.ijgc.2025.102109 [DOI] [PubMed] [Google Scholar]
- 24.Nishikawa H, Goto M, Fukunishi S, Asai A, Nishiguchi S, Higuchi K. Cancer Cachexia: its Mechanism and Clinical Significance. Int J Mol Sci. 2021;22(16):8491. doi: 10.3390/ijms22168491 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25.Karakaş S, Demirayak G, Önder AB, et al. The association between the preoperative prognostic nutritional index and the controlling nutritional status score on tumor stage, chemotherapeutic response and overall survival in ovarian cancer. Nutr Cancer. 2022;74(5):1770–1779. doi: 10.1080/01635581.2021.2022170 [DOI] [PubMed] [Google Scholar]
- 26.Benna-Doyle S, Baguley BJ, Laing E, Kiss N. Nutritional interventions during treatment for ovarian cancer: a narrative review and recommendations for future research. Maturitas. 2024;183:107938. doi: 10.1016/j.maturitas.2024.107938 [DOI] [PubMed] [Google Scholar]
- 27.Maurer T, Belau MH, von Grundherr J, et al. Randomised controlled trial testing the feasibility of an exercise and nutrition intervention for patients with ovarian cancer during and after first-line chemotherapy (BENITA-study). BMJ Open. 2022;12(2):e054091. doi: 10.1136/bmjopen-2021-054091 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28.Shi Y, Xiong W, Xia B. Neoadjuvant therapy in ovarian cancer: emerging paradigms and therapeutic frontiers. Crit Rev Oncol Hematol. 2026;219:105065. doi: 10.1016/j.critrevonc.2025.105065 [DOI] [PubMed] [Google Scholar]
- 29.Fu YP, Lin H, Ou YC, Wu CH, Fu HC. Bevacizumab as a mitigating factor for the impact of high systemic immune-inflammation index on chemorefractory in advanced epithelial ovarian cancer. BMC Cancer. 2024;24(1):1377. doi: 10.1186/s12885-024-13087-8 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 30.Ferrero A, Vassallo D, Geuna M, et al. Immunonutrition in ovarian cancer: clinical and immunological impact? J Gynecol Oncol. 2022;33(6):e77. doi: 10.3802/jgo.2022.33.e77 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 31.Raju V, Reddy R, Javan AC, et al. Tracking inflammation status for improving patient prognosis: a review of current methods, unmet clinical needs and opportunities. Biotechnol Adv. 2025;82:108592. doi: 10.1016/j.biotechadv.2025.108592 [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Data Availability Statement
The datasets used and/or analysed during the current study are available from the corresponding author (Fengzhi Feng) on reasonable request.