Pretreatment serum lactate dehydrogenase is an independent prognostic factor for patients receiving neoadjuvant chemotherapy for locally advanced cervical cancer

Abstract For locally advanced cervical cancer (LACC), hypoxia is a characteristic property. This study aimed to investigate whether baseline lactic dehydrogenase (LDH) level, which is a marker of hypoxia, had clinical value in determining neoadjuvant chemotherapy (NACT) response and prognosis for LACC patients. The study cohort included 418 patients with a median follow‐up of 37.5 months. Cox proportional hazards models were used to assess the prognostic value of baseline LDH levels. Multivariate logistic regression analysis was performed to identify independent predictors of complete response after NACT. Backward stepwise selection with the Akaike information criterion was used to identify factors that could be entered into the multivariate regression model. Compared with patients with LDH levels <252.0 μ/L, patients with LDH levels ≥252.0 μ/L were more likely to have an elevated level of squamous cell carcinoma antigen, lymphatic vascular space involvement, lymph node metastasis, and positive parametrium and achieved lower complete remission rates. Baseline LDH levels ≥252.0 μ/L was an independent prognosticator for recurrence‐free survival (adjusted hazard ratio [HR], 3.56; 95% confidence interval [CI] 2.22–5.69; P < 0.0001) and cancer‐specific survival (adjusted HR, 3.08; 95% CI, 1.89–5.01; P < 0.0001). The predictive value of baseline LDH value remained significant in the subgroup analysis. LDH level ≥252.0 μ/L was identified as an independent predictor of complete remission after NACT (adjusted odds ratio [OR], 0.29; 95% CI, 0.15–0.58; P < 0.0001). Baseline LDH ≥252.0 μ/L is an independent prognostic predictor for patients receiving neoadjuvant chemotherapy for LACC. It helps distinguish patients with different prognosis and select patients who are more likely to benefit from NACT.

the other hand, there are several objections because for patients with locally advanced cervical cancer (LACC) who do not response to NACT, the delay in curative treatment, the development of radio-resistant cellular clones and cross-resistance with radiotherapy may exert negative impact on patient survival [7]. Given these controversies, NACT is not recommended as a routine treatment for LACC patients in current National Comprehensive Cancer Network (NCCN) Clinical Practice guideline [1]. The discrepancy raises the possibility that NACT may improve antitumor outcomes in a subset of patients. Therefore, appropriate biomarkers are needed to select patients who are most likely to benefit from such treatment.
For locally advanced solid tumors, hypoxia is a characteristic property due to rapid cancer cell proliferation, high metabolic demands, and functional angiogenesis [8].
There is clear evidence that hypoxia can promote cancer development and it is involved in the resistance to treatment via the formation of new blood vessels [9]. Lactate dehydrogenase (LDH) is known to be a marker of hypoxia, which plays an important role in the proliferation and metastasis of tumor cells [10]. Moreover, pretreatment serum LDH levels have been found to correlate with the prognosis of patients with malignant diseases [11]. However, to the best of our knowledge, the clinical significance of LDH has been never investigated in patients with LACC. Therefore, we conducted a large cohort study to investigate the prognostic and predictive value of LDH levels for patients treated with NACT for LACC.

Materials and Methods
Patients After Institutional Review Board (IRB) approval was obtained at the Sun Yat-Sen Memorial Hospital, the institutional database was utilized to identify cases. The medical records of all women who received NACT and subsequent class III radical hysterectomy for cervical cancer between January 2005 and June 2010 were reviewed. Patients were reclassified based on the FIGO (Federation International of Gynecology and Obstetrics) 2014 staging system [4]. Inclusion criteria were as follows: histologically confirmed squamous cell carcinoma and adenocarcinoma, FIGO stage IB2 and IIA2 disease, blood collection for LDH measurements prior to NACT, and signed informed consent provided. Exclusion criteria were as follows: patients not completing the planned cycles of NACT, patients not receiving radical surgery after NACT, and patients receiving any previous treatment for cervical or uterine malignancies. Data collected included demographic information; operative, chemotherapy, and radiotherapy notes; histopathologic reports; and follow-up notes. Pretreatment evaluation included physical and gynecologic examination, chest radiography, pelvic ultrasonography, and laboratory tests. Further investigation was performed when indicated. Gynecologic examination was carried out by at least two senior gynecologists. Maximum tumor diameter was determined by clinical measurement. Two authorized pathologists from our institution who were blinded to study outcomes reviewed all cervical pathology.
The NACT regimens were employed as follows: TP, paclitaxel+cisplatin; FP, 5-fluouracil+cisplatin; TC, paclitaxel+carboplatin; and BVP, bleomycin+vincristine+cis platin. All patients underwent type III radical hysterectomy according to the Piver-Rutledge classification with pelvic lymphadenectomy within 4 weeks after the last cycle of NACT. Pathological responses were retrospectively evaluated, and a complete response (CR) was defined as no evidence of viable tumor cells on the tumorous area [12]. CCRT was performed if patients had the following risk factors: positive parametrium, positive lymph nodes, involved surgical margins, greater than one-third stromal invasion, and lymphatic vascular space involvement (LVSI) [4].
Blood samples were collected for laboratory tests within 1 week before initiation of NACT. Serological LDH levels were measured using a Hitachi Automatic Analyzer 7600-020 (Hitachi High-Technologies, Tokyo, Japan). Normal serum LDH enzyme activities were defined to 108.0-252.0 μ/L. Based on pretreatment serum LDH levels, patients were classified into high LDH group (HL group, LDH ≥252.0 μ/L) and normal LDH group (NL group, LDH <252.0 μ/L). Serum squamous cell carcinoma antigen (SCCA) was assessed with an immunoradiometric assay kit (Imx; Abbott Diagnostics, Abbott Park, IL). A cut-off value of 3.5 ng/mL was used to stratify patients into normal and abnormal group [13]. The intraassay variation was <5% for all variables measured. Laboratory personnel performing these assays were blinded to study outcomes.
Follow-up visit including complete history and physical examination and Papanicolau smear of the vaginal vault was recommended every 3 months for 2 years, every 6 months for the next 3 years, and once per year thereafter. Follow-up information was obtained from office visits or telephone interviews. Tumor recurrence was diagnosed on the basis of clinical symptoms, physical examinations, biopsy, and imaging methods including positron emission tomographycomputed tomography (PET-CT), magnetic resonance imaging (MRI), and computed tomography (CT). The primary endpoint of this study was to investigate whether the baseline serum LDH was a prognostic factor for recurrence-free survival (RFS) and cervical cancer-specific survival (CSS). RFS was measured from the date of NACT until the date of recurrence or last follow-up. CSS was calculated as the time interval between the date of NACT and the date of death from cervical cancer or the date of last follow-up.

Statistical analyses
Data were analyzed using STATA/SE version 12.0 statistical software (Stata Corp. LP, College Station, TX) and SPSS version 14.0 (SPPS Inc., Chicago, IL). Continuous variables were presented as the median and range, while Categorical variables were presented as the number and percentages. The Kolmogorov-Smirnov test was used to determine the distribution of continuous variables. Student's t-test was used to compare normally distributed continuous variables, whereas Mann-Whitney U test was used for data with nonnormal distribution. Chi-square test (χ 2 ) or Fisher exact test were used to analyze the frequency distribution between categorical variables where appropriate. RFS and CSS were estimated using the Kaplan-Meier method and compared with the log-rank test. Multivariate analysis (enter method) was performed to identify independent predictors for survival outcomes with Cox proportional hazards model, and hazard ratios (HRs) and 95% confidence intervals (CIs) were presented. The assumption of proportional hazards was tested based on Schoenfeld residuals [14]. A binary logistic regression model for multivariate analysis was also used to determine independent predictor for CR after NACT, expressed with odds ratio (OR), and 95% CI. Akaike information criteria with backward selection were used to select variables that were entered into the multivariate model. All statistical tests were two-tailed, and a P value of <0.05 was considered to be statistically significant.

Characteristics of the study population
The final study cohort included 418 patients with a median follow-up of 37.5 months (range: 4-65 months). The median age at the diagnosis of cervical cancer was 52.0 years (range: 24-80 years). The median serum LDH level for the entire cohort was 194.0 μ/L with a range 63 to 634 μ/L. An overview of clinic-pathologic characteristics of all patients is given in Table 1. Of the included patients, 322 (77.03%) had serum LDH levels <252.0 μ/L, whereas 96 (22.97%) had serum LDH levels ≥252.0 μ/L. Compared with patients in the NL group, patients in the HL group were more likely to have lymphatic vascular space involvement (LVSI) (65.6% vs. 52.2%, P = 0.020, r = 0.114), lymph node metastasis (59.4% vs. 29.8%, P < 0.0001, r = 0.258), and positive parametrium (9.4% vs. 2.2%, P < 0.003, r = 0.158). The proportion of patients who achieved CR after NACT was significantly lower in the HL group in comparison with those in the NL group (11.5% vs. 32.3%, P < 0.0001, r = −0.196). More patients in the HL group complicated with anemia, although it did not reach statistical significance (60.4% vs. 49.1%, P = 0.051, r = 0.096).

Subgroup analysis
We further evaluated the prognostic effects of LDH on RFS and CSS according to patient baseline characteristics, using a Cox proportional hazards regression model. Table 4 summarizes the results of subgroup analysis. The Factors associated with CR after NACT CR after NACT has been confirmed as a reliable surrogate endpoint of survival for patients with LACC. Therefore, an additional logistic regression analysis was conducted to determine factors predicting CR after NACT, and the results are summarized in Table 5. At univariate analysis, pretreatment LDH levels ≥252.0 μ/L (OR, 0.27; 95% CI, 0.14-0.53; P < 0.0001) and pretreatment SCCA levels ≥3.5 ng/mL (OR, 0.35; 95% CI, 0.23-0.55; P < 0.0001) were significantly associated with decreased likelihood of CR after NACT. Four variables were included in the multivariate analysis according to stepwise selection based on AIC. Body mass index (BMI) ≥25 kg/m 2 (OR, 0.35; 95% CI, 0.23-0.55; P < 0.0001), LDH levels ≥252.0 μ/L (OR, 0.35; 95% CI, 0.23-0.55; P < 0.0001), and SCCA levels ≥3.5 ng/mL (OR, 0.35; 95% CI, 0.23-0.55; P < 0.0001) were independently associated with decreased incidence of CR after NACT.

Discussion
An inverse relationship between LDH levels and length of survival has also been identified in many tumor types, including melanoma [15], breast cancer [16], myeloma [17], hepatocellular carcinoma [18], seminoma [19], nasopharyngeal cancer [20], lung cancer [21], colorectal cancer [10], renal cancer [22], oral cancer [23], and pancreatic cancer [24]. For gynecologic malignancies, a strong association between the elevated expression of LDH and an aggressive phenotype has been noted in patients with ovarian and uterine carcinoma [25,26]. In consistent with these findings, our study demonstrated that LACC patients with elevated LDH levels were more likely to have positive SCCA, LVSI, lymph node metastasis, and parametrium invasion. Furthermore, compared with patients with baseline LDH less than 252.0 μ/L, patients with LDH ≥252.0 μ/L had a statistically significant 3.6-fold risk of cancer recurrence and 3.1-fold risk of cancer-specific death, and this association was independent of other potential prognostic factor. The prognostic influence of elevated LDH levels was consistent across all the LACC patient subgroups. Additionally, we found pretreatment LDH levels <252.0 μ/L was an independent predictor of CR after NACT.
Previous research has explored the biological mechanisms that are responsible for the association between elevated LDH levels and ominous prognosis in cancer patients. Possible explanations are as follows. First, tumor cells utilize glycolysis instead of mitochondrial oxidative phosphorylation to generate ATP that is required for the increased energy demand of the rapidly proliferating tumor cells [27]. As a key enzyme in the process of glycolysis, LDH converts pyruvate and NADH to lactate and NAD+, determining the maintenance of glycolytic flow and consequently the production of ATP [28]. Because LDH can be transcriptionally upregulated by hypoxia inducible factor 1α (HIF-1α) and hypoxia in the tumor microenviroment is sufficient to stimulate the activation of HIF, there is a positive feedback loop between HIF and LDH under hypoxic conditions [29,30]. Therefore, elevated levels of LDH indicate an aggressive phenotype. Second, an increased serum LDH has been reported to reflect a heavy tumor burden [31][32][33]. Owing to the heterogeneity of tumor cells, tumors with heavier load contain tumor cells with greater diversities [34]. Thus, LDH-positive patients are more susceptible to treatment resistance. Additionally, the vascular density is significantly higher in patients with elevated LDH levels which suggest an aggressive angiogenesis [35]. As angiogenesis is essential for tumor proliferation and metastasis, patients with increased LDH levels are more likely to have a poor prognosis.
What is noteworthy is that cutoffs for LDH were heterogeneous in previous studies. In this study, we used 252.0 μ/L as the LDH cutoff. The impact of variations in LDH cutoffs has been evaluated in a published metaanalysis [11]. In the study, Zhang et al. pooled data from 68 studies and included 31,857 cancer patients. They concluded that high LDH is associated with an adverse prognosis in solid tumors and the variations in LDH cutoffs have no impact on its prognostic effect.
For patients with LACC, achieving optimal pathological response on surgical specimen is a strong predictor of good clinical outcome [36,37]. The study by Alessandro et al. [38] was the largest one to date that has assessed the benefit of NACT. Based on the long-term follow-up data (median follow-up time: 12.7 years), the authors proposed response to NACT as a surrogate endpoint of survival for LACC patients. Given these findings and our own observations of the prognostic effect of CR for patients with LACC, we conducted an additional multivariate analysis and found that pretreatment LDH levels ≥252.0 μ/L was independently associated with decreased likelihood of CR after NACT (OR, 0.36; 95% CI, 0.23-0.57; P < 0.0001). The result is in line with previous reports that showed LDH is a marker of response to NACT for breast cancer patients and oral cancer patients [23,39]. In vitro studies observed LDH is involved in resistance to chemotherapy, which may be an interpretation for the difference in CR rates by LDH levels [40,41].
This study have several strengths including: (1) it was not only the first one to specifically explore the prognostic value of LDH in LACC but also the largest one to test the prognostic value of LDH in patients with gynecologic cancer; (2) all patients were newly diagnosed, so possible influence from disproportionate pretreatment that patients might receive can be ruled out; (3) all patients were from a single institution, so uniform treatment protocol can be ensured.
This study had several limitations. First, its observational design prevents us from discounting completely any residual factors of confusion that may influence the levels of LDH such as bone disease. Second, data about serial dynamic serum LDH levels are lacking. Finally, the findings of this study may be specific to Asian populations.

Conclusion
In summary, our study suggests that baseline LDH ≥252.0 μ/L is an independent prognostic predictor for LACC patients treated with NACT. Furthermore, LACC patients with LDH levels ≥252.0 μ/L are less likely to achieve CR after NACT. Further study with adequate statistical power is needed to confirm and validate our findings. If validated, baseline LDH, an inexpensive and readily available laboratory parameter, could be utilized as a biomarker that can help physicians further categorize LACC patients with different prognosis and define the appropriate patient subgroup for NACT.