Usefulness of ¹⁸F-fluorodeoxyglucose positron-emission tomography/computed tomography combined with the platelet-lymphocyte ratio in predicting the prognosis of nasopharyngeal carcinoma

Abstract

Objectives:

To investigate the value of ¹⁸F-fluorodeoxyglucose (FDG) positron-emission tomography (PET)/computed tomography (CT) combined with the platelet-lymphocyte ratio (PLR) in predicting the prognosis of nasopharyngeal carcinoma (NPC).

Methods:

This was a retrospective analysis of the data of 73 patients with NPC who underwent ¹⁸F-FDG PET/CT before treatment from January 2010 to December 2014. The maximum standard uptake value (SUVmax) of NPC and the PLR within 1 week before treatment were both measured. The Mann-Whitney U-test was used to compare the differences between the SUVmax and PLR among the different clinical characteristics of patients with NPC and the 5-year progression-free survival (PFS) rate; according to the receiver operating characteristic (ROC) curve, the best cutoff values of the SUVmax and PLR were obtained and used to group patients. The Kaplan-Meier method and Log-rank test were used to conduct univariate analysis of 5-year PFS in patients with NPC, and Cox regression was used to conduct multivariate analysis; differences in the 5-year PFS of patients with different SUVmax values combined with the PLR were compared.

Results:

The SUVmax and PLR of patients with disease progression within 5 years were higher than those of patients without disease progression (p = 0.006 and p = 0.026). SUVmax = 9.7 and PLR = 132.98 had the best prognostic diagnostic efficiency for patients. Cox multivariate analysis showed that the SUVmax and PLR are independent factors affecting the prognosis of NPC. The 5-year PFS of patients with SUVmax <9.7 was significantly higher than that of patients with SUVmax ≥9.7 in the high PLR group (PLR ≥132.98) and in the low PLR group (PLR <132.98) (59.3% vs  29.4%, p = 0.033 and 90.9% vs 42.9%, p = 0.006, respectively). For patients with SUVmax <9.7, the 5-year PFS of the high PLR group was significantly lower than the low PLR group (59.3% vs  90.9%, p = 0.016); for patients with SUVmax ≥9.7, there was no significant difference in 5-year PFS between the high PLR group and the low PLR group (29.4% vs  42.9%, p = 0.406).

Conclusions:

Both the SUVmax of the primary tumor and the PLR before treatment have an important influence on the prognosis of NPC. Combining the SUVmax and the PLR can more accurately predict the prognosis of patients with NPC.

Advances in knowledge

In this study, we evaluated the prognostic value of combining pretreatment tumor ¹⁸F-FDG uptake on PET/CT imaging and PLR in NPC patients. We found that both SUVmax and PLR are independent factors for the PFS of NPC patients, and a low SUVmax (SUVmax <9.7) combined with a low PLR (PLR <132.98) revealed significant PFS benefit.

Introduction

Nasopharyngeal carcinoma (NPC) is one of the most common malignant tumors of the head and neck in China, and its distribution has obvious regional differences, with a high incidence in southeastern China, such as Guangdong and Guangxi¹. Due to sensitivity to radiotherapy, radiotherapy alone or combined with chemotherapy is currently the main treatment for NPC, but there is still a high recurrence rate after radical treatment.² TNM clinical staging is currently the most commonly used diagnostic criterion to judge prognosis, but it only relies on the anatomical information of the tumor. There are still significant differences in the prognosis of patients in the same stage^3,4; therefore, more accurate indicators are needed.

In recent years, the prognostic value of inflammation indicators for tumors has received increasing attention. Numerous inflammatory indicators, such as the neutrophil-lymphocyte ratio (NLR), platelet-lymphocyte ratio (PLR), and C-reactive protein (CRP), have been found to be associated with the prognosis of malignant tumors, such as lung cancer, breast cancer, and head and neck squamous cell carcinoma.^5–7 Some studies have shown that the PLR is related to the prognosis of NPC.⁸ However, inflammatory indicators reflect the overall state of the body, and they ignore the impact of tumor heterogeneity on the prognosis when these indicators are used for prognosis evaluation. ¹⁸F-FDG PET/CT uses glucose as a metabolic substrate, which can reflect the glucose metabolism of tumors. Multiple studies have shown that the maximum standardized uptake value (SUVmax) of tumor lesions is related to the proliferation activity, invasiveness and degree of differentiation.^9–11 This study aimed to study the predictive value of the SUVmax combined with the PLR in the prognosis of NPC.

Methods and materials

This study was approved by our institutional ethical committee and all patients’ clinical features were obtained from the medical records. The requirement for informed consent was waived owing to the retrospective nature of the analysis.

Study population

This was a retrospective analysis of the data of NPC patients who were treated at the First People’s Hospital of Foshan City from January 1, 2010, to December 31, 2014. All patients underwent ¹⁸F-FDG PET/CT examination before treatment，and all patients were confirmed by pathology. The inclusion criteria were as follows: (1) positive lesions of NPC detected in ¹⁸F-FDG PET/CT examination, (2) no distant metastasis, (3) no history of immune disease or blood system disease, (4) no complicated infectious diseases or active inflammation, (5) no complicated severe liver or kidney disease, and (6) no history of malignant tumor disease.

The complete blood count (CBC) of all patients was tested within 1 week before treatment, and the neutrophil, lymphocyte, and platelet counts were recorded. We calculated PLR as the counts of platelet divided by the lymphocyte counts, and NLR as counts of neutrophil divided by the lymphocyte counts. Patients with the following were excluded: (1) incomplete clinical data, (2) loss to follow-up after treatment, and (3) the appearance of a second primary tumor during follow-up after treatment. Finally, a total of 73 patients were enrolled in this study. The age range was 19 to 82 years old, with an average age of 50 (41.5–61.5) years old. All patients underwent T staging and clinical staging according to the American Joint Committee on Cancer (AJCC) Seventh Edition TNM classification and staging system.¹²

Treatment

All patients in this study received three-dimensional conformal radiation therapy (3D-CRT) or intensity-modulated radiation therapy (IMRT) radical radiotherapy, stage I patients received radiotherapy alone, stage II patients received cisplatin-based concurrent radiotherapy, and stages III and IVa patients received docetaxel, cisplatin and 5-fluorouracil (TPF)/cisplatin, and 5-fluorouracil (PF)-based induction chemotherapy and cisplatin based on concurrent radiotherapy and chemotherapy. The radiotherapy dose was as follows: primary nasopharyngeal lesions and neck metastatic lymph nodes (68 ~ 70) Gy, clinical target area (CTV) (54 ~ 60) Gy, single dose 1.8–2.2 Gy. According to the patient’s clinical stage and adverse reactions, two to three cycles of induction chemotherapy were given. The TPF chemotherapy regimen was docetaxel 60 mg m⁻², i.v. drip on the first day; cisplatin 60 mg/m2, i.v. drip on the first day; and 5-fluorouracil 600 mg m⁻² per day continuous infusion on days 1 to 5, repeated for one cycle every 21 days. The PF chemotherapy regimen was cisplatin 60 mg m⁻², i.v. infusion on the first day; 5-fluorouracil 600 mg m⁻² per day, continuous infusion on the first to fifth days, repeating a cycle every 21 days. Concurrent cisplatin chemotherapy regimen: cisplatin 100 mg m⁻², i.v. infusion on the 1st, 22nd, and 43rd days.

PET/CT examination

The GEMINI PET/CT scanner from PHILIPS Company was used. ¹⁸F-FDG was provided by Guangzhou Isotope Center of the China Academy of Atomic Energy, with a radiochemical purity >95%. All patients were fasted for more than 6 h before the examination, and their fasting blood glucose was ≤150 mg dl⁻¹. ¹⁸F-FDG 4.44 ~ 5.55 MBq kg⁻¹ was injected via the cubital vein, and PET/CT imaging was performed after 60 min of peaceful rest. The scanning range was from the skull to the upper femur. PET adopts 3D acquisition mode, 3 min/bed, and the data are acquired by the Ramla3D method to acquire attenuation correction images. The CT acquisition parameters were as follows: voltage 120 kV, current 200 mAS, pitch 1.0, layer thickness 6 mm, and reconstruction thickness 5 mm.

Image analysis

All PET/CT images were analyzed by two nuclear medicine physicians who have vast experience in NPC research. According to the PET/CT fusion image, the region of interest (ROI) of the NPC lesion was delineated, and the SUVmax of the lesion was automatically measured by the workstation.

Follow-up and prognostic evaluation

When radiotherapy was over, 73 patients were followed up by a combination of outpatient review and telephone. The follow-up time node was December 31, 2019, and if local recurrence, regional metastasis, distant metastasis or death occurred, follow-up ended immediately. The follow-up time ranged from 8 to 102 months, and the average time was 67 (33.5, 81.0) months. The 5-year progression-free survival (PFS) was used as a survival evaluation index. PFS was defined as the time from the first day of treatment to the occurrence of local recurrence, metastasis, death, or termination of follow-up.

Statistical methods

SPSS v.19.0 software was used for statistical analysis. SUVmax, age, NLR, and PLR were expressed as the median M (P25, P75). The comparison between the two groups of data was performed by the Mann-Whitney U-test; the SUVmax and PLR were analyzed by ROC curve analysis to find the best critical value, and the patients were grouped according to the critical value. The Kaplan-Meier method and Log-rank test were used for survival analysis and univariate analysis. Cox regression was performed for multivariate analysis. p ＜ 0.05 indicates a significant difference.

Results

Follow-up

A total of 29 patients had recurrence or metastasis, and the 5-year PFS was 60.27% (44/73), of which 15 patients had local recurrence in the nasopharynx, eight patients had regional lymph node metastasis, six patients had distant metastasis, and no patient died as the follow-up endpoint.

The relationship between pretreatment SUVmax, NLR, PLR, and clinical characteristics and prognosis

The SUVmax of patients with high T stage and clinical stage NPC was significantly higher than that of patients with low stage (Z = −4.281, p = 0.000; Z = −3.553, p = 0.000). Patients with disease progression within 5 years had a higher SUVmax and PLR than those without disease progression (Z = −2.769, p = 0.006; Z = −2.223, p = 0.026), while the NLR was not associated with patient prognosis (Z = −0.751, p = 0.453) (Table 1). Through correlation analysis, there was no correlation between SUVmax and NLR (p = 0.154), and between SUVmax and PLR (p = 0.637), respectively.

Table 1.

Comparison of pretreatment SUVmax, NLR and PLR among different clinical features and prognoses of patients with NPC

Characteristics	N	SUVmax	p value	NLR	p value	PLR	p value
Sex
Male	56	7.50 (5.35,11.88）	0.734	2.67 (1.76,4.13）	0.154	142.86 (107.40,178.40）	0.932
Female	17	7.80 (5.20,9.05）		1.91 (1.75,3.04）		138.89 (105.67,189.16）
Age (year)
˂50	36	7.80 (5.55,10.60）	0.817	2.28 (1.61,4.50）	0.294	132.92 (104.52,172.13）	0.206
≥50	37	6.60 (5.10,12.75）		3.00 (1.80,3.86）		150.00 (113.48,187.67）
T staging
T1-2	27	5.70 (3.80,6.60）	0.000	2.32 (1.75,3.86）	0.433	138.99 (104.98,172.13）	0.253
T3-4	46	9.10 (6.40,12.33）		2.79 (1.75,4.00）		150.00 (113.39,189.16）
Total staging
I + II	19	5.90 (3.80,6.60）	0.000	2.51 (1.75,3.86）	0.567	142.86 (98.62,172.13）	0.615
III + IV	54	8.60 (5.70,12.10）		2.56 (1.75,4.00）		140.88 (109.98,186.93）
PFS (month)
˂60	29	10.30 (6.55,12.80)	0.006	2.97 (1.75,4.08)	0.453	157.79 (134.87,209.09)	0.026
≥60	44	6.55 (4.80,8.90)		2.51 (1.75,3.69)		129.02 (104.98,170.00)

NLR, Neutrophil-lymphocyte ratio; NPC, Nasopharyngeal carcinoma; PFS, Progression-free survival rate; PLR, Platelet-lymphocyte ratio;SUVmax, Maximum standard uptake value.

Prognostic factors

Using ROC curve analysis, the critical value of SUVmax that predicted 5-year PFS is 9.70, its area under the curve (AUC) is 0.692 (0.564 ~ 0.821) (p = 0.006), the Youden index is 0.370, and the corresponding sensitivity and specificity are 55.20 and 81.80%, respectively. PLR predicts 5-year PFS with a cutoff value of 132.98, the AUC is 0.655 (0.525 ~ 0.786) (p = 0.026), the Youden index is 0.316, and the corresponding sensitivity and specificity are 79.30 and 52.30%, respectively. NLR was not associated with patient prognosis, and its AUC is 0.552 (0.414 ~ 0.691) (p = 0.453). Among the clinical factors of patients, tumor T stage, clinical stage, SUVmax, and PLR before treatment were the factors that affected the prognosis of NPC (Table 2). According to the results of multivariate Cox regression analysis, the SUVmax and PLR of the lesion before treatment were independent factors that determined the prognosis of patients with NPC, while T stage and clinical stage were not independent factors (Table 3).

Table 2.

Univariate analysis of pretreatment SUVmax, PLR, and clinical features affecting the prognosis of NPC patients

Characteristics	N	Percentage of 5-year PFS	p value
Sex
Male	56	58.92	0.747
Female	17	64.71
Age (years)
˂50	36	63.89	0.751
≥50	37	56.76
T staging
T1-2	27	74.07	0.046
T3-4	46	52.17
Total staging
I + II	19	78.95	0.042
III + IV	54	53.70
SUVmax
˂9.7	49	73.47	0.001
≥9.7	24	33.33
PLR
˂132.98	29	79.31	0.008
≥132.98	44	47.73

NPC, Nasopharyngeal carcinoma; PFS, Progression-free survival rate; PLR, Platelet-lymphocyte ratio;SUVmax, Maximum standard uptake value.

Table 3.

Multivariate analysis of prognostic factors for the 5-year PFS of NPC patients

Factors	HR	95% CI for HR	p value
T staging (T3-4 versus T1-2)	1.880	0.472 ~ 7.486	0.371
Total staging (III + IV versus I + II)	3.797	0.753 ~ 19.161	0.106
SUVmax (≥9.7 versus ˂9.7)	2.538	1.137 ~ 5.665	0.023
PLR (≥132.98 versus ˂132.98)	3.525	1.276 ~ 9.735	0.015

CI, Confidence interval; HR, Hazard ratio;NPC, Nasopharyngeal carcinoma; PFS, Progression-free survival rate; PLR, Platelet-lymphocyte ratio; SUVmax, Maximum standard uptake value.

The value of the combination of SUVmax and PLR in prognosis prediction

When SUVmax < 9.7, the 5-year PFS of patients with a low PLR (PLR <132.98) was significantly higher than that of patients with a high PLR (PLR ≥132.98) (90.9% vs  59.3%, p = 0.016). When SUVmax ≥ 9.7, the difference of 5-year PFS between patients with low PLR and patients with high PLR was not statistically significant (42.9% vs  29.4%, p = 0.406). No matter PLR ≥ 132.98 or PLR < 132.98, the 5-year PFS of patients with low SUVmax (SUVmax <9.7) was significantly higher than that of patients with high SUVmax (SUVmax ≥9.7) (59.3% vs  29.4%, p = 0.033 and 90.9% vs 42.9%, p = 0.006, respectively) (Figure 1).

Figure 1.

The effect of combining SUVmax and PLR on the 5-year PFS in patients with nasopharyngeal carcinoma (a) the 5-year PFS of SUVmax < 9.7 & PLR < 132.98 vs. SUVmax < 9.7 & PLR ≥ 132.98; (b) the 5year PFS of SUVmax ≥ 9.7 & PLR < 132.98 vs. SUVmax ≥ 9.7 & PLR ≥ 132.98; (c) the 5year PFS of PLR < 132.98 & SUVmax < 9.7 vs. PLR < 132.98 & SUVmax ≥ 9.7; (d) the 5-year PFS of PLR ≥ 132.98 & SUVmax < 9.7 vs. PLR ≥ 132.98 & SUVmax ≥ 9.7. SUVmax, maximum standard uptake value; PLR, platelet-lymphocyte ratio; PFS, progression-free survival rate.

Discussion

Recurrence and metastasis are important factors affecting the survival of NPC patients. In recent years, with the improvement in treatment technology, especially the application of IMRT, the curative effect of radical radiotherapy has been significantly improved, but 5% ~ 15% of patients will still have local recurrence, and approximately 15% ~ 30% of patients will have distant metastases.¹³ Individualized treatment to improve the therapeutic effect and long-term survival rate of NPC is of great significance, and this depends on the accurate assessment of prognostic diagnostic indicators. TNM staging can no longer meet clinical needs.

This study shows that the SUVmax of the primary tumor is an independent prognostic factor for patients with NPC, and the 5year PFS of patients with a high SUVmax is significantly reduced. Tumoral heterogeneity has an enormous effort on the treatment and outcome of NPC patients, such as cell proliferation and aggressiveness. Multiple studies have shown that the SUVmax of tumors is related to the proliferation activity of tumor cells.^9,14,15 The increased proliferation activity of tumor cells and the acceleration of nuclear antigen and protein metabolism will require more energy support; therefore, glucose is the main energy source, and the degree of glucose metabolism of tumors will increase. A meta-analysis¹⁶ showed that the expression of Ki67, which revealing the proliferative activity of tumor, has an obvious relation with overall survival (OS) and PFS in patients with NPC. The uptake of FDG can also be used to assess its aggressiveness,¹⁰ and the increase in SUVmax of the lesion may indicate that NPC cells have stronger metastatic potential. The SUVmax of NPC lesions is also related to the expression of Glucose transporter (GLUT)−1 in the lesion. Deron¹⁷ found that the SUVmax of head and neck tumors was closely related to the expression of GLUT-1 in tumor cells, and the high expression of GLUT-1 was associated with the expression of GLUT-1 in NPC lymph node metastasis. Epstein-Barr virus (EBV) infection was closely related.¹⁸ These are important factors that affect the survival rate of patients with NPC. In addition, this study also found that patients with NPC with larger tumors and higher stages had a higher SUVmax, which may be related to the proliferation of tumor cells. As the tumor grows, the number of tumor cells inside the tumor will increase. At the same time, the hypoxia of the cells inside the tumor is aggravated, and the energy consumed to maintain the same proliferation rate will also increase. Hypoxia can also lead to an increase in FDG uptake.¹⁹ High glucose metabolism is not unique phenomena for tumor cells. Inflammation can also cause increased ¹⁸F-FDG uptake.²⁰ Some tumor studies found that inflammatory marks are correlated with the SUVmax of the lesion because tumors can cause an immune response and cause local inflammation.²¹ However, this study did not find correlations between SUVmax and PLR and between SUVmax and NLR in NPC. The local inflammation caused by the tumor may not be the main factor influencing the increase in ¹⁸F-FDG uptake in NPC lesions, but it is still not certain for other inflammatory markers not being included in the study. Epstein-Barr virus (EBV) infection may be another factor affecting the ¹⁸F-FDG uptake of the lesion. Chang²² concluded that the level of circulating EBV DNA was significantly associated with SUV of the primary tumor in the study of NPC, But the study of Alessi²³ came to a different conclusion. As the data of EBV DNA were not obtained in this study, it needs to be confirmed by further research.

The PLR is another important factor affecting the prognosis of patients with NPC. Platelets promote the development of cancer. Cancer cells can induce platelet activation. Platelets contain a variety of cell growth factors, including platelet-derived growth factor (PDGF) and vascular endothelial growth factor (VEGF), which control the growth of blood vessels. Factors promote tumor blood vessel growth, accelerate tumor growth,²⁴ and secrete some inflammatory mediators, such as platelet factor-4, interleukin-8, CXCL5, and CXCL7, to guide tumor metastasis²⁵ while helping tumor cells escape the body’s immune monitoring. Animal studies have shown that when the platelet count and function decrease, the metastatic activity of tumors is significantly reduced, while the growth and metastasis of tumor cells in mice with platelet defects are inhibited.²⁶ Lymphocytes play an important role in immune monitoring and antitumor activity. NK cells and CD4 +and CD8+T lymphocytes can eliminate tumor cells and inhibit tumor metastasis. This has been found in research on head and neck squamous cell carcinoma. The numbers of CD4 +and CD8+T lymphocytes in tumor patients are significantly reduced.²⁷ The results of Kawaguchi²⁸ also found that CD8 +lymphocytes can predict the sensitivity of NPC to chemotherapy and radiotherapy, and high CD8 +lymphocyte levels indicate that patients with NPC have high sensitivity to radiotherapy and chemotherapy.

The NLR is another indicator of inflammation that has been studied in recent years. Pre-treatment NLR was considered to be significantly related to the prognosis of patients with NPC.²⁹ The mechanism may be that the increase in the NLR indicates an increase in inflammation, which is beneficial for promoting the progression of aggressive tumors. But this study had not found NLR have any impact on the prognosis of patients with NPC. The reason may be that radiotherapy will cause obvious inflammation at the tumor site, leading to changes in the local inflammatory environment of the original tumor. Changes in the NLR before and after radiotherapy are more suitable for predicting the prognosis of NPC.³⁰ In addition, the level of CRP³¹ and EBV-DNA²³ may also be related to the prognosis of NPC. However, whether they are independent factors affecting prognosis of NPC requires further confirmation.

NPC patients with higher T staging and higher total staging have poor prognosis. But the present study indicated that the SUVmax of the primary tumor and the PLR before treatment are better than TNM staging in predicting the prognosis of NPC. Combined with SUVmax and PLR analysis, it can combine the tumor characteristics of NPC lesions and the immune status of the body to more objectively evaluate the patient’s circumstance to make an accurate prediction of the prognosis. In this study, the 5-year PFS was used as the research time point. It was found that the incidence of recurrence and metastasis in patients with SUVmax <9.7 was significantly lower than that of patients with SUVmax ≥9.7. Unlike the SUVmax, the PLR has a different impact on the prognosis of patients with different SUVmax levels. The incidence of recurrence and metastasis in patients with a low SUVmax but high PLR was significantly higher than that of patients with a low SUVmax and low PLR, while there was no significant difference in 5-year PFS between patients with different PLR values in the high SUVmax group. At present, there is no research report that combines the SUVmax and PLR before treatment to predict the prognosis of NPC. Our findings suggested the combination of SUVmax and PLR may improve the prognostic stratification. More active treatment should be taken for patients with high SUVmax. On the contrary, patients with low SUVmax and low PLR are suitable for a mild treatment and avoid complications caused by unnecessary radical therapy. The results may provide a basis for the formulation of individualized treatment plan for NPC patients.

This study had several limitations. The major disadvantage is that this is a single-center retrospective study and there may be bias in patient selection. Another limitation is the small number of patients included in the study. Whether the diagnostic cutoff values of the SUVmax and PLR in the study are appropriate remains to be confirmed. In the future, multicentre prospective studies are needed to improve the results of this study.

Conclusion

Our study showed that both the SUVmax of the primary tumor and the PLR before treatment have an important influence on the prognosis of NPC. There is no obvious correlation between the two. Combining the SUVmax and the PLR can more accurately stratify the prognosis risk for patients with NPC. It is beneficial to formulate more appropriate individualized treatment plans to improve the prognosis and quality of life.