Inverse Association between Platelet–Lymphocyte Ratio and Prognosis in Terminally Ill Cancer Patients: A Preliminary Study

Abstract

Background:

An elevated platelet–lymphocyte ratio (PLR) is an indicator for worse outcomes in cancer, but its significance at the end of life remains unclear.

Objective:

This study aimed to investigate the value of PLR as an independent prognostic factor in terminally ill cancer patients.

Methods:

This retrospective cohort study included 312 terminal cancer patients and was conducted in a palliative care unit of a tertiary cancer center. Patient demographic data, clinical information, and laboratory values, including complete blood cell count, were obtained. Survival was analyzed using the Kaplan–Meier method and log-rank test. The Cox proportional hazards model was used to identify independent prognostic factors for survival.

Results:

Median survival was 16 days in patients with PLR ≥200 and 9 days in patients with PLR <200 (p = 0.008). Results of multivariate analysis showed that the following factors predicted worse survival: poor performance status (adjusted hazard ratio [aHR], 2.16; 95% confidence interval [CI], 1.50–3.09), azotemia (aHR, 1.43; 95% CI, 1.01–2.02), hypoalbuminemia (aHR, 1.55; 95% CI, 1.07–2.26), hyperbilirubinemia (aHR, 1.67; 95% CI, 1.23–2.29), elevated lactate dehydrogenase (aHR, 1.58; 95% CI, 1.11–2.26), high neutrophil–lymphocyte ratio (aHR, 1.49; 95% CI, 1.09–2.04), and low PLR (aHR, 1.46; 95% CI, 1.08–1.97).

Conclusion:

Although elevated PLR indicates worse outcomes in patients with solid tumors, decreased PLR was an independent prognostic factor for poor survival in cancer patients at the end of life.

Introduction

Although predicting the remaining life expectancy of patients with advanced cancer is a challenge, communicating this information is important in a palliative care setting.¹ A subjective approach uses the clinician's survival estimate and patient's clinical symptoms.² However, an objective biomarker from the peripheral blood would be useful for this purpose.

A solid body of evidence supports the pivotal role of systemic inflammation in cancer occurrence and progression.^3,4 Accordingly, an inflammation-related biomarker has been shown to be an independent prognostic indicator in cancer patients⁵ and may predict survival at the end of life.^6,7 Recently, the platelet–lymphocyte ratio (PLR), which is easily calculated from the complete blood cell count (CBC), has been reported to have an independent prognostic value in solid tumors.⁸ A meta-analysis of studies comprising 12,754 cancer patients demonstrated that a high PLR was associated with poor survival in patients with various solid tumors, and the association tended to be stronger in metastatic disease than in early-stage disease.⁸ However, it is not clear whether this association remains at the very end of life.

Lymphopenia is consistently associated with poor prognosis in cancer.^9,10 However, platelet count linking with prognosis in cancer might be different across the trajectories. While an elevated platelet count is generally related to poor prognosis in several types of cancer,¹¹ in terminally ill cancer patients low platelet count is associated with shorter survival time.^12,13 Thus, we hypothesized that, in the terminal phase, the association between PLR and survival would be null or rather reversed. This preliminary study aimed to evaluate the relationship between PLR and survival in terminally ill patients with advanced cancer.

Methods

Study design and participants

This retrospective cohort study was conducted in a palliative care unit within the Incheon regional cancer center. All participants were terminal cancer patients who were likely to die within months according to the oncologist's estimates, and they were not currently receiving any curative treatment (i.e., chemo- or radiotherapy). They and/or their families had been informed of being in a terminal state by their physician, and they had accepted the referral to palliative care unit. In case of our unit, approximately two-thirds of patients were referred from other departments in our center. From an electronic database, we selected 358 patients who had died from December 2012 to December 2014. Of these patients, the analysis included 312 patients who had available PLR data. This study was approved by the Institutional Review Board (IRB) of Gachon University Gil Medical Center. The requirement for obtaining informed consent was waived by the IRB because the study entailed chart review only with no identifying patient information, and the data were obtained from routine clinical care.

Data collection

Patient information collected on admission to the palliative care unit included age, gender, cancer site, evidence of infection, and functional status (assessed using the Eastern Cooperative Oncology Group Performance Status [ECOG PS], observer-rated scale of physical ability ranging from 0 to 4).¹⁴ Blood samples were taken within days of referral to our unit, and they were used to measure CBC, liver function (albumin, total bilirubin, and prothrombin time international normalized ratio), renal function (creatinine), lactate dehydrogenase (LDH), and C-reactive protein (CRP). The neutrophil–lymphocyte ratio (NLR) and PLR were calculated. Survival time was defined as the period from the day of evaluation to the day of death. Laboratory values were categorized based on their reference range or distribution (median or quartile).

Statistical analysis

The Kaplan–Meier method and the log-rank test were used to compare median survival time of groups divided according to PLR (<200 vs. ≥200). The Cox proportional hazards method estimated the hazard ratio (HR) and 95% confidence interval to identify independent prognostic factors associated with survival. Variables significant in the univariate analysis (p < 0.05) were entered into the final model to calculate adjusted HR (aHR). We constructed a best-fit multivariate model using a stepwise variable selection procedure. In all analyses, p < 0.05 generated in two-sided tests was considered significant. All statistical tests were performed using Stata statistical software version SE9 (StataCorp, College Station, TX).

Results

Table 1 shows median survival according to patient characteristics. Of the 312 patients, 166 (53.2%) were male and 164 (52.6%) were older than 65 years. The most common site of cancer was the lung (n = 70, 22.4%), followed by pancreatico-biliary tract (n = 57, 18.3%). Most patients (n = 233, 74.7%) had poor performance status (ECOG PS ≥3) and 124 (39.7%) had evidence of current infection. Overall median survival was 12 days (interquartile range, 5–28).

Table 1.

Survival Time According to Patient Characteristics (n = 312)

		Median survival days (95% CI)	p^a
Age, years
≤65	148	12 (5–23)	0.213
>65	164	11 (5–30)
Sex
Male	166	10 (5–21)	0.019
Female	146	13 (5–30)
Cancer site
Esophagus/stomach	48	12 (5–31)	0.770
Colorectal	35	14 (6–20)
Liver	27	9 (4–15)
Pancreatico-biliary tract	57	13 (6–20)
Breast	20	6 (3–22)
Lung	70	11 (5–31)
Urogenital	28	12 (7–28)
Others	27	9 (4–21)
ECOG performance status
1	13	49 (38–57)	<0.001
2	62	17 (9–35)
3	147	11 (5–25)
4	86	8 (4–17)
Current infection
No	188	16 (6–30)	<0.001
Yes	124	7 (4–16)
Leukocytosis (≥10 × 10³/mm³)
No	138	14 (6–29)	0.029
Yes	174	9 (5–21)
Neutrophilia (>75%)
No	76	21 (10–33)	0.004
Yes	236	9 (5–21)
Lymphopenia (<20%)
No	34	21 (9–30)	0.228
Yes	278	11 (5–25)
Anemia (<11.5 g/dL for women; <13.0 g/dL for men)
No	45	18 (5–35)	0.162
Yes	267	11 (5–24)
Thrombocytopenia (<150 × 10³/mm³)
No	204	16 (6–30)	<0.001
Yes	108	7 (4–16)
Azotemia (>1.2 mg/dL)
No	231	13 (6–29)	<0.001
Yes	77	7 (4–17)
Hypoalbuminemia (<3.5 g/dL)
No	58	23 (10–42)	<0.001
Yes	254	10 (4–21)
Hyperbilirubinemia (>1.2 mg/dL)
No	170	17 (6–31)	<0.001
Yes	141	9 (4–18)
Prolonged PT INR (>1.2)
No	135	17 (6–36)	<0.001
Yes	131	10 (4–20)
Elevated LDH level (>485 U/L)
No	51	26 (13–36)	0.001
Yes	163	10 (4–22)
C-reactive protein level (mg/dL)^b
Low (<8.2)	155	15 (6–31)	0.002
High (≥8.2)	151	9 (4–20)
Neutrophil–lymphocyte ratio
Low (<10)	163	19 (6–33)	<0.001
High (≥10)	149	8 (4–17)
PLR
≤200	149	9 (4–21)	0.008
>200	152	16 (6–30)

Log-rank test.

Median value.

CI, confidence interval; LDH, lactate dehydrogenase; PLR, platelet–lymphocyte ratio; PT INR, prothrombin time international normalized ratio.

Factors associated with significantly shorter survival times in univariate analysis (Table 1) were male sex, poor functional status (ECOG PS), current infection, leukocytosis, neutrophilia, thrombocytopenia, azotemia, hypoalbuminemia, hyperbilirubinemia, prolonged PT-INR, elevated LDH level, high CRP level, high NLR, and low PLR. Median PLR was 207 (interquartile range, 120–317). Low PLR was associated with shorter survival (median survival 16 days with PLR ≥200 vs. 9 days with PLR <200; p = 0.008) (Fig. 1).

FIG. 1.

Kaplan–Meier survival curve of patients categorized by platelet–lymphocyte ratio (PLR). Patients with PLR <200 (dotted line) had significantly shorter survival time than those with PLR ≥200 (solid line; p = 0.008 by log-rank test).

Variables that were significant (p < 0.05) in univariate analysis were included in multivariate analysis using the Cox proportional hazards model (Table 2). The results confirmed the following factors as predictors of worse survival: poor ECOG PS (aHR, 2.16), azotemia (aHR, 1.43), hypoalbuminemia (aHR, 1.55), hyperbilirubinemia (aHR, 1.67), elevated LDH (aHR, 1.58), high NLR (aHR, 1.49), and low PLR (aHR, 1.46).

Table 2.

Independent Prognostic Factors for Survival, According to the Cox Proportional Hazards Model

	Univariate analysis		Multivariate analysis ^a
	HR (95% CI)	p	HR (95% CI)	p
Elderly (>65 years)	0.87 (0.69–1.09)	0.226
Male	1.31 (1.04–1.65)	0.023
Poor performance status (ECOG, 3–4)	1.94 (1.47–2.56)	<0.001	2.16 (1.50–3.09)	<0.001
Current infection	1.47 (1.17–1.86)	0.001
Leukocytosis	1.28 (1.02–1.61)	0.034
Neutrophilia	1.45 (1.11–1.90)	0.006
Lymphopenia	1.25 (0.86–1.81)	0.242
Anemia	1.25 (0.91–1.73)	0.174
Thrombocytopenia	1.52 (1.20–1.93)	0.001
Azotemia	1.55 (1.19–2.02)	0.001	1.43 (1.01–2.02)	0.042
Hypoalbuminemia	2.07 (1.51–2.82)	<0.001	1.55 (1.07–2.26)	0.021
Hyperbilirubinemia	1.72 (1.36–2.18)	<0.001	1.67 (1.23–2.29)	0.001
Prolonged PT INR	1.70 (1.32–2.19)	<0.001
High CRP level (≥8.2 ng/dL)	1.42 (1.13–1.79)	0.003
Elevated LDH level	1.72 (1.23–2.39)	0.001	1.58 (1.11–2.26)	0.011
High NLR (≥10)	1.70 (1.34–2.14)	<0.001	1.49 (1.09–2.04)	0.013
Low PLR (≤200)	1.34 (1.07–1.69)	0.011	1.46 (1.08–1.97)	0.013

Including selected variables that were significant (p < 0.05) in the univariate analysis.

CRP, C-reactive protein; ECOG, Eastern Cooperative Oncology Group; HR, hazard ratio; NLR, neutrophil–lymphocyte ratio.

Discussion

To the best of our knowledge, this is the first study to evaluate PLR as an independent prognostic indicator in terminally ill cancer patients. We found that low PLR predicted worse survival. Although elevated PLR predicts poor outcomes in patients with early or advanced solid tumors,^8,15–18 our study showed an inverse relationship between PLR and life expectancy in cancer patients near the end of life.

Lymphocytes, which play an important role in cancer immune surveillance,^19,20 are associated with treatment response and prognosis in cancer.^10,21 Thus, lymphopenia is associated with shorter survival in terminal cancer patients^9,22 and is part of a prognostic score.⁵ However, in our study, neutrophilia was more predictive of survival than lymphopenia, which may be due to a higher vulnerability to infection in this patient population.

A complex body of evidence highlights the contribution of platelets to cancer growth, angiogenesis, and metastasis.^23,24 Several studies have also showed that elevated blood platelet count is associated with clinical outcome in cancer patients.^25,26 However, in terminally ill cancer patients, a decreased platelet count can occur through mechanical replacement of bone marrow by the tumor, liver failure, or disseminated intravascular coagulation.²⁷ The inverse relationship between PLR and survival observed in our study could be explained by a decrease in platelet count that surpassed the progression of lymphopenia.

Both NLR and PLR values can be derived from the CBC and may therefore provide simple and inexpensive biomarkers for cancer prognostication. These ratios, but not the values used to derive the ratios (neutrophil, lymphocyte, or platelet count), were independent predictors of survival in patients with terminal cancer. Given that NLR and PLR are independently associated with survival despite the fact that they are both derived from the CBC, we surmised that they reflect different aspects influencing survival beyond systemic inflammation. The median value of PLR in this study was ∼200, close to the cutoff (range, 150–300) used by previous studies evaluating the relationship between PLR and survival in patients with earlier stage.⁸ In contrast, the median NLR value in our study was ∼10, which was much higher than the cutoff (range, 2–5) used in previous studies.²⁸ This suggests that NLR increases continuously as the cancer progresses, whereas PLR initially increases but then decreases at some point before the end of life. To better understand these findings, future studies are needed to evaluate the levels of these biomarkers during the final weeks of life and explore their relationships with other prognostic factors.

A limitation of this study is that it was based on retrospective chart review in a single center. Several potential confounders (i.e., scoring systems such as the Palliative Prognostic Score, clinician estimates, or other serologic markers) could be missed or measured imperfectly.²⁹ Larger multicenter and preferably prospective studies are warranted to validate our findings.

In conclusion, for end-of-life care during which minimal tests should be performed, a simple blood biomarker to predict the remaining life expectancy is preferable. Our data indicate that PLR, a cost-effective and readily available biomarker, can provide additional information on the prognosis of terminally ill cancer patients.

Footnotes

Acknowledgments

This work was supported by the Senior-friendly Product R&D program funded by the Ministry of Health and Welfare thorough the Korean Industry Development Institute (HI14C1435). The funder had no role in the design of the study; the collection, analysis, or interpretation of the data; the writing of the manuscript; nor the decision to submit the manuscript for publication.

Author Disclosure Statement

No competing financial interests exist.

References

Maltoni

, Caraceni

, Brunelli

,; Steering Committee of the European Association for Palliative C: Prognostic factors in advanced cancer patients: Evidence-based clinical recommendations—A study by the Steering Committee of the European Association for Palliative Care. J Clin Oncol, 2005; 23:6240–6248.

Gripp

, Moeller

, Bolke

, et al.: Survival prediction in terminally ill cancer patients by clinical estimates, laboratory tests, and self-rated anxiety and depression. J Clin Oncol, 2007; 25:3313–3320.

Mantovani

, Allavena

, Sica

, Balkwill

: Cancer-related inflammation. Nature, 2008; 454:436–444.

Elinav

, Nowarski

, Thaiss

, et al.: Inflammation-induced cancer: Crosstalk between tumours, immune cells and microorganisms. Nat Rev Cancer, 2013; 13:759–771.

McMillan

: The systemic inflammation-based Glasgow Prognostic Score: A decade of experience in patients with cancer. Cancer Treat Rev, 2013; 39:534–540.

Glare

, Sinclair

, Downing

, et al.: Predicting survival in patients with advanced disease. Eur J Cancer, 2008; 44:1146–1156.

Ahn

, Hwang

, Lee

, et al.: Neutrophil-lymphocyte ratio predicts survival in terminal cancer patients. J Palliat Med, 2016; 19:437–441.

Templeton

, Ace

, McNamara

, et al.: Prognostic role of platelet to lymphocyte ratio in solid tumors: A systematic review and meta-analysis. Cancer Epidemiol Biomarkers Prev, 2014; 23:1204–1212.

Glare

, Virik

: Independent prospective validation of the PaP score in terminally ill patients referred to a hospital-based palliative medicine consultation service. J Pain Symptom Manage, 2001; 22:891–898.

10.

Gooden

, de Bock

, Leffers

, et al.: The prognostic influence of tumour-infiltrating lymphocytes in cancer: A systematic review with meta-analysis. Br J Cancer, 2011; 105:93–103.

11.

Buergy

, Wenz

, Groden

, Brockmann

: Tumor-platelet interaction in solid tumors. Int J Cancer, 2012; 130:2747–2760.

12.

Suh

, Ahn

: Lactate dehydrogenase as a prognostic factor for survival time of terminally ill cancer patients: A preliminary study. Eur J Cancer, 2007; 43:1051–1059.

13.

Kim

, Ha

, Hwang

, et al.: Random serum cortisol as a predictor for survival of terminally ill patients with cancer: A preliminary study. Am J Hosp Palliat Care, 2016; 33:281–285.

14.

Oken

, Creech

, Tormey

, et al.: Toxicity and response criteria of the Eastern Cooperative Oncology Group. Am J Clin Oncol, 1982; 5:649–655.

15.

Koh

, Bhoo-Pathy

, Ng

, et al.: Utility of pre-treatment neutrophil-lymphocyte ratio and platelet-lymphocyte ratio as prognostic factors in breast cancer. Br J Cancer, 2015; 113:150–158.

16.

Chen

, Dai

, Yin

, et al.: Negative impact of preoperative platelet-lymphocyte ratio on outcome after hepatic resection for intrahepatic cholangiocarcinoma. Medicine (Baltimore), 2015; 94:e574.

17.

Shirai

, Shiba

, Sakamoto

, et al.: Preoperative platelet to lymphocyte ratio predicts outcome of patients with pancreatic ductal adenocarcinoma after pancreatic resection. Surgery, 2015; 158:360–365.

18.

Jiang

, Zou

, Hu

, et al.: The elevated pretreatment platelet-to-lymphocyte ratio predicts poor outcome in nasopharyngeal carcinoma patients. Tumour Biol, 2015; 36:7775–7787.

19.

Shankaran

, Ikeda

, Bruce

, et al.: IFNgamma and lymphocytes prevent primary tumour development and shape tumour immunogenicity. Nature, 2001; 410:1107–1111.

20.

Dunn

, Old

, Schreiber

: The immunobiology of cancer immunosurveillance and immunoediting. Immunity, 2004; 21:137–148.

21.

Loi

, Sirtaine

, Piette

, et al.: Prognostic and predictive value of tumor-infiltrating lymphocytes in a phase III randomized adjuvant breast cancer trial in node-positive breast cancer comparing the addition of docetaxel to doxorubicin with doxorubicin-based chemotherapy: BIG 02–98. J Clin Oncol, 2013; 31:860–867.

22.

Vigano

, Bruera

, Jhangri

, et al.: Clinical survival predictors in patients with advanced cancer. Arch Intern Med, 2000; 160:861–868.

23.

Gay

, Felding-Habermann

: Contribution of platelets to tumour metastasis. Nat Rev Cancer, 2011; 11:123–134.

24.

Bambace

, Holmes

: The platelet contribution to cancer progression. J Thromb Haemost, 2011; 9:237–249.

25.

Gao

, Zhang

, Xia

: Increased platelet count is an indicator of metastasis in patients with nasopharyngeal carcinoma. Tumour Biol, 2013; 34:39–45.

26.

Hernandez

, Donohue

, Anderson

, et al.: The significance of thrombocytosis in patients with locally advanced cervical carcinoma: A Gynecologic Oncology Group study. Gynecol Oncol, 2000; 78:137–142.

27.

Nash

, Turner

, Scully

, Kakkar

: Platelets and cancer. Lancet Oncol, 2002; 3:425–430.

28.

Templeton

, McNamara

, Seruga

, et al.: Prognostic role of neutrophil-to-lymphocyte ratio in solid tumors: A systematic review and meta-analysis. J Natl Cancer Inst, 2014; 106:dju124.

29.

Stone

, Lund

: Predicting prognosis in patients with advanced cancer. Ann Oncol, 2007; 18:971–976.