Prognostic implications of dynamic serum lactate dehydrogenase assessments in nasopharyngeal carcinoma patients treated with intensity-modulated radiotherapy

The prognostic value of dynamic serum lactate dehydrogenase (LDH) levels in patients with nasopharyngeal carcinoma (NPC) treated with intensity-modulated radiotherapy (IMRT) hasn’t been explored. We retrospectively analyzed 1,428 cases of NPC treated with IMRT with or without chemotherapy. Elevated pre- and/or post-treatment LDH levels were found to be associated with unfavorable overall survival (OS), disease-free survival (DFS) and distant metastasis-free survival (DMFS), but not with local relapse-free survival (LRFS). The dynamic variations in LDH levels were prognostic factors for OS, DFS and DMFS, but not for LRFS. Multivariate analysis revealed that the N category, T category, post-treatment serum LDH level and age were independent prognostic factors for OS. Our results demonstrated that dynamic variations in LDH levels were associated with risk of distant failure and death, which may shed light on the dynamics of the disease and the response to therapy. We consider that LDH measurements will be of great clinical importance in the management of NPC, especially, when considering “decision points” in treatment algorithms. Therefore, we strongly recommend that LDH levels should be determined before and after treatment in NPC patients and the results integrated into decisions regarding treatment strategies.

NPC management 6 . The baseline LDH level correlates well with tumor stage, outcome prediction and early detection of liver metastasis in patients with NPC, and has been developed as a tool for the non-invasive assessment of the tumor burden in NPC patients [6][7][8] .
Most scoring systems presented so far use LDH as a "static" prognostic variable determined at the time of diagnosis 8,9 , and the clinical significance of dynamic LDH levels measured at different time points has not been fully explored. The dynamics of the disease and responses to therapy may be of great clinical importance, especially when considering "decision points" in treatment algorithms such as residual gross tumor at the end of radiotherapy. Therefore, in this study, we retrospectively reviewed the medical records of patients with newly diagnosed, non-metastatic NPC treated with radical intensity-modulated radiation therapy (IMRT). The predictive and prognostic roles of pre-treatment and post-treatment serum LDH levels were evaluated, particularly with respect to the establishment of prognostic subsets of NPC and appropriate treatment strategies.

Materials and Methods
Patient selection. We retrospectively reviewed the records of all NPC patients who were treated with IMRT at the Cancer Center of Sun Yat-sen University (Guangzhou, People's Republic of China) between November 2009 and February 2012. This study was approved by the institutional review board. We excluded 383 patients because their records contained insufficient information. Thus, we reviewed the cases of a total of 1,428 patients with newly diagnosed, histologically proven, non-metastatic NPC.
Clinical staging. All patients underwent pretreatment evaluations that included a complete history, physical examinations, hematology and biochemistry profiles, MRI of the nasopharynx and neck, chest radiography, abdominal sonography, and whole-body bone scan using single photon-emission computed tomography. All MRI and clinical data were reviewed to minimize heterogeneity in restaging. Two radiologists specializing in head and neck cancers separately evaluated all scans, and any disagreements were resolved by consensus. The American Joint Committee on Cancer staging system (7th edition) was used for stage classification 10 . Treatment. All patients were treated with radical IMRT over the entire course of treatment. Details regarding the IMRT techniques used have previously been reported 11 . In general, the treatment plans were determined according to tumor stage and general health of the patient. During the study period, institutional guidelines recommended radiotherapy alone for patients in stage I, concurrent chemoradiotherapy for those in stage II, and concurrent chemoradiotherapy with or without neoadjuvant/adjuvant chemotherapy for those in stage III to IVb defined by the 7th edition of the UICC/AJCC staging system for NPC. Of the 1,088 patients with stage III or stage IVA-B disease (classified as T3-T4 or N2-N3), 1,034 (95.0%) received chemotherapy, including various regimens of concurrent chemotherapy in combination with either induction chemotherapy or adjuvant chemotherapy in conjunction with a platinum-based therapeutic clinical trial. Concurrent chemotherapy consisted of 80-100 mg/m 2 every 3 weeks or 40 mg/m 2 weekly cisplatin. Induction or adjuvant chemotherapy consisted of 2-4 cycles of cisplatin with 5-luorouracil (80 mg/m 2 cisplatin on day 1 and 800 mg/m 2 per day 5-fluorouracil on days 1-5), or cisplatin with docetaxel every 3 weeks (70 mg/m 2 cisplatin and 70 mg/m 2 docetaxel on day 1), or cisplatin with 5-luorouracil and docetaxel (60 mg/m 2 cisplatin on day 1, 60 mg/m 2 docetaxel on day 1 and 600 mg/m 2 per day 5-fluorouracil on days 1-5). When possible, salvage treatments, including afterloading, surgery and chemotherapy, were provided in the event of a documented relapse or persistent disease. Patient follow-up. Serum LDH levels were measured within 14 days of any therapeutic intervention (with the exception of diagnostic biopsy of primary tumors or cervical lymph nodes) and within 4 weeks after the completion of treatment. The specific procedures of LDH analysis have been described in our previous study 6 . Normal serum LDH enzyme activities ranged from 109-245 IU/l, and the coefficient of variance for the LDH measurements was < 5.0%.
Patients returned for follow-up appointments at least every 3 months during the first 2 years and every 6 months thereafter, until death. The follow-up duration was calculated from the first day of therapy to the day of death or to the day of the last examination. The following end points (time to the first defining event) were assessed: overall survival (OS), disease-free survival (DFS), distant metastasis-free survival (DMFS) and local relapse-free survival (LRFS). Statistical analysis. The Statistical Package for the Social Sciences, version 16.0 (SPSS, Chicago, IL) was used for all statistical analyses. The independent-samples t-test was used to calculate differences in continuous variables between the various patient groups. The pre-treatment and post-treatment serum LDH levels were compared using paired-samples t-tests. Actuarial rates were calculated using the Kaplan-Meier method, and differences were compared using the log-rank test. Multivariate analysis using a Cox proportional hazards model was used to test for independent significance by backward elimination of insignificant explanatory variables. Host factors (including age, sex, pathology and clinical stage) were included as covariates in all tests. The criterion for statistical significance was set at α = 0.05 and P-values were based on two-sided tests.

Results
Patient characteristics. The characteristics of the included patients have been listed in Table 1. Among the 1,428 study patients, 1,066 were male and 362 were female (male/female ratio, 2.94:1). The median patient age was 44.0 years (range, 14-78 years). On histological examination, 99.5% of the patients were diagnosed with type II or III disease, according to the guidelines of the World Health Organization (WHO) 12  LDH level at diagnosis and its prognostic significance. At the time of diagnosis, 93.6% (1,337/1,428) of the NPC patients had serum LDH levels within the normal limits (i.e., 109-245 IU/l). The remaining 6.4% (91/1,428) patients exhibited elevated serum LDH levels. The median and mean pre-treatment serum LDH levels for the entire cohort were 170.00 IU/l and 178.06 ± 2.28 IU/l, respectively (range, 84-453 IU/l). The pre-treatment LDH levels progressively increased with worsening clinical stage, and the mean LDH values at diagnosis in patients with stage I-IV disease were 162.16 ± 27.78 IU/l, 168.94 ± 35.31 IU/l, 176.34 ± 53.96 IU/l and 186.09 ± 54.38 IU/l, respectively (P < 0.001). Of the 91 patients with elevated serum LDH levels at the time of diagnosis, 12 (13.2%) had stage II disease, 37 (40.7%) had stage III disease and 42 (46.2%) had stage IV disease.
The 4-year OS rate among the patients with normal pretreatment serum LDH levels (90.9%) was significantly higher than the rate in patients with elevated LDH levels (80.0%; P = 0.001). Significant differences between patients with normal and elevated baseline LDH levels were also detected in the case of the 4-year DFS (83.4% and 62.3%; P < 0.001) and DMFS (90.1% and 70.5%; P < 0.001). However, the 4-year LRFS was the same regardless of the LDH value at diagnosis (91.7% and 90.8%; P = 0.61).
The paired-samples t-test revealed that the LDH level significantly declined after treatment (P < 0.001). The median and mean difference in the pre-and post-treatment LDH levels was 16 and 18.67 ± 1.60 IU/l. To evaluate the changes in LDH levels after treatment, we divided the patients into two groups according to the median difference: patients in whom the median decrease in LDH level after treatment was > 16 IU/l and patients in whom this decrease was < 16 IU/l or in whom there was an increase the LDH level after treatment. Patients in the former group had a 4-year OS of 92.8%, which was significantly better than survival rate in the latter group (88.9%, P = 0.007). Similarly, the DMFS rates in the two groups were significantly different (91.5% vs. 86.3% respectively, P = 0.005). The difference in DFS was only marginally significant (84.3% vs. 79.9%, respectively, P = 0.058), while no difference was detected in LRFS between the two groups (91.6% vs. 91.7%, P = 0.471).

Multivariate analysis.
Multivariate analysis was performed to adjust for various prognostic factors. The following parameters were included in a Cox proportional hazards model: age (< 44 years vs. ≥ 44 years), sex, T classification, N classification, chemotherapy (present vs. absent), pre-treatment LDH levels (≤ 245 IU/l vs. > 245 IU/l) and post-treatment LDH levels (≤ 245 IU/l vs. > 245 IU/l). The N category, T category, post-treatment serum LDH levels and age were found to be independent prognostic factors for OS (P < 0.001, P < 0.001, P = 0.002 and P = 0.040 respectively). The multivariate analysis also revealed that the N category, T category and pre-and post-treatment serum LDH levels were significant predictors of DFS and DMFS. However, only the WHO histological type was an independent indicator of LRFS (Table 3).

Discussion
Several prognostic variables, including LDH, have been identified to predict survival in NPC 8,13 . While the prognostic value of LDH levels at the time of diagnosis is well established, little is known about the value of post-treatment LDH levels as a marker of disease response. In the current study, we determined the pre-and post-treatment LDH levels in a cohort of NPC patients in order to determine the prognostic value of LDH as a dynamic parameter.    The results of our study showed that an elevated baseline LDH level was associated with distant failure and disease-specific survival. Many studies have shown that the LDH level at diagnosis is of prognostic significance in NPC, which was confirmed in the present study 4,7,8,14 . Liaw et al. retrospectively investigated 465 NPC patients and demonstrated that serum LDH levels were higher in patients who had distant metastasis or disease relapse than in patients who were in remission and in normal controls 4 . A phase III randomized study from Hong Kong compared the outcomes of concurrent chemoradiation versus radiation therapy alone in 348 NPC patients with advanced nodal disease, and found that serum LDH was a significant predictor of distant failure 15 . On the basis of these reports, we concluded that the baseline serum LDH level is a useful prognostic factor for NPC.
Although the prognostic significance of LDH in the case of OS, DFS and DMFS has been confirmed, the value of pretreatment LDH in predicting local relapse varies between studies 4,7 . In the current study, LDH level was not a significant factor for locoregional control on univariate and multivariate analyses. This result was consistent with that of our previous study, which included 465 NPC patients treated with IMRT 6 . However, Wan et al. have reported that among a cohort of NPC patients from a randomized controlled trial, pretreatment serum samples were a prognostic predictor of LRFS 7 . Moreover, in a prognostic scoring system for locoregional control proposed by Cheng et al., pretreatment LDH level was identified as an important parameter for LRFS prediction in NPC patients 9 . This difference among our study and the above studies may be attributable to differences in the treatments administered; Wan et al. and Cheng et al. used conventional two-dimensional radiotherapy, while our study involved IMRT. Compared with conventional radiotherapy, IMRT generates more conformal dose coverage in the target volume and therefore results in ideal local control 16 . In the current study, the 4-year LRFS was as high as 91.7%. Thus, it is possible that the improved local control rates obtained with IMRT eliminated the prognostic impact of LDH levels on LRFS.
The prognosis of a cancer depends on the biological aggressiveness of the tumor, the characteristics of the host and the therapeutic interventions. The prognostic factors of NPC have been an important research focus, and numerous investigations in this area have been published. Among the suggested disease-related prognosticators, the following have been proved to be significant prognostic factors in NPC: stage at diagnosis, plasma/serum Epstein-Barr virus DNA, serum LDH, tumor volume, cranial nerve palsy and the presence of tumor-associated genes in the peripheral blood cells [17][18][19][20] . However, the application of these pretreatment markers in clinical practice with consequent adjustments of the treatment regimen has never been shown to be feasible. Therefore, the search for new markers that can be used to predict and monitor early tumor-treatment responses will greatly impact patient management.
Our study suggests that post-treatment serum LDH levels were an independent prognosticator. Although it wasn't the unique prognosticator, it indicated that besides the familiar risk factors such as TNM categories and age at diagnosis, the dynamic LDH levels during the treatment period should be considered, especially to define patients with a high risk of metastasis who warrant more aggressive systemic therapy to improve long-term outcomes when considering the addition of adjuvant or maintenance therapy at the end of radiotherapy. This is the first time that the prognostic value of post-treatment serum LDH levels has been explored in patients with newly diagnosed NPC treated with radical IMRT. Our conclusion was consistent with that of another study conducted in a cohort of patients with metastatic NPC. In 2012, Jin et al. reported that patients with elevated post-treatment LDH levels had worse survival than those with normal LDH levels. In their study, patients with normal pre-and post-treatment LDH values showed the highest response rate and the most favorable prognosis 21 . Notably, on the basis of our data, we recommend that LDH levels should be evaluated in NPC patients after the finish of radiotherapy. An increase in LDH at this time should lead to a thorough reevaluation of the disease status.
Serum LDH is potentially an ideal prognosticator among NPC patients treated with radical IMRT. During the search for prognostic indicators, a strong emphasis must be paid to ensure that the indicator is useful for both clinicians and researchers, while also being applicable in highly endemic regions all over the world. The  convenience, repeatability and low cost of LDH assessments make this factor one of the most applicable in clinical practice. Furthermore, the prognostic indicator is important for not only predicting outcomes but also determining the proper management strategy for NPC patients. Since distant failure continues to be a critical point affecting patient prognosis, prognostic factors should focus on the risk of distant failure. Elevated pre-and/or post-treatment serum LDH levels predict an increased risk of both distant failure and suboptimal survival. More intensive systemic treatment strategies that can be used in this group of patients must be investigated. This study identified that pre-and post-treatment LDH levels are negative prognostic indicators in NPC. High LDH is of great clinical significance in many other solid tumors as well as hematological malignancies 22,23 . However, the exact mechanism underlying this phenomenon is not understood. High serum LDH levels usually indicate the presence of a hypoxic environment associated with a large tumor burden 24 , and the oxygenation status of a tumor has been shown to be an important determinant of clinical outcomes among patients who receive radiotherapy and chemotherapy 25 . Alternatively, it was reported recently that increased LDH may reflect aberrant oncogene activity, since it is known that the myc and PI3K/Akt/mTOR pathways regulate cellular LDH expression levels 26 . The activation of these pathways was demonstrated to be an unfavorable prognostic factor in NPC 27 . Finally, LDH may be related to host factors, such as cellular turnover, cachexia and inflammation due to progressive tumor growth 22 .
To the best of our knowledge, this analysis is the first to demonstrate the prognostic role of dynamic serum LDH changes in patients with newly diagnosed, non-metastatic NPC treated with radical IMRT. Elevated preand post-treatment serum LDH levels were confirmed to be useful predictors of distant metastasis and poor overall survival. Thus, we strongly recommend that LDH determinations be carried out in NPC patients before and after radiotherapy, and that the test results be integrated into decisions regarding treatment strategy.