Ann Surg Oncol (2014) 21:1998–2004 DOI 10.1245/s10434-014-3533-9
ORIGINAL ARTICLE – GASTROINTESTINAL ONCOLOGY
Value of Preoperative Inflammation-Based Prognostic Scores in Predicting Overall Survival and Disease-Free Survival in Patients with Gastric Cancer Paolo Aurello, MD, Simone Maria Tierno, MD, Giammauro Berardi, MD, Federico Tomassini, MD, Paolo Magistri, MD, Francesco D’Angelo, MD, and Giovanni Ramacciato, MD Third Division of General Surgery, Sant’Andrea Hospital, University of Rome ‘‘La Sapienza’’, Rome, Italy
ABSTRACT Background. This study was designed to identify which are the best preoperative inflammation-based prognostic scores in terms of overall survival (OS) and disease-free survival (DFS) in patients with gastric cancer. Methods. Between January 2004 and January 2013, 102 consecutive patients underwent resection for gastric cancer at S. Andrea Hospital, ‘‘La Sapienza’’, University of Rome. Their records were retrospectively reviewed. Results. After a median follow up of 40.8 months (8– 107 months), patients’ 1-, 3-, and 5-year OS rates were 88, 72, and 59 %, respectively. After R0 resection, the 1-, 3-, and 5-year DFS rates were 93, 74, and 56 %, respectively. A multivariate analysis of the significant variables showed that only the modified Glasgow prognostic scores (p \ 0.001) and PI (p \ 0.001) were independently associated with OS. Regarding DFS, multivariate analysis of the significant variables showed that the modified Glasgow prognostic score (p = 0.002) and prognostic index (p \ 0.001) were independently associated with DFS. Conclusions. The results of this study show that modified Glasgow prognostic score and prognostic index are independent predictors of OS and DFS in patients with gastric cancer.
Gastric cancer is the fourth most common malignancy and the second most common cause of cancer-related death worldwide.1 Tumor stage, tumor size, grade of differentiation,
Ó Society of Surgical Oncology 2014 First Received: 27 September 2013; Published Online: 11 February 2014 S. M. Tierno, MD e-mail: [email protected]
and vascular invasion are well-established prognostic factors.2 The relationship between hypoalbuminemia and poor prognosis has also been established.3–5 In addition, recently there has been growing interest in the host’s inflammatory response to tumor, as evidenced by increased concentrations of C-reactive protein (CRP). Previous studies have demonstrated the relationship between gastric cancer and chronic inflammation, although the exact mechanism involved is not clear.6–11 The presence of an inflammatory response is crucial in the development of cancer-associated malnutrition, resulting in poor performance status and increased mortality in patients with cancer.12 On the basis of the relationship between inflammation and cancer progression, several inflammation-based scores have been proposed as prognostic. The main scores that have been associated with cancer-specific survival are the modified Glasgow Prognostic Scores (mGPS), neutrophil lymphocyte ratio (NLR), platelet lymphocyte ratio (PLR), prognostic index (PI), and prognostic nutritional index (PNI).13–17 To our knowledge, there are no studies elucidating which of these prognostic scores is more suitable in predicting outcomes in patients with gastric cancer. The aim of this study was to compare the prognostic value of these preoperative inflammation-based prognostic scores (the mGPS, NLR, PLR, PI, and PNI) in terms of overall survival (OS) and disease-free survival (DFS) in patients with gastric cancer in various stages of disease. METHODS The records of a series of 126 consecutive patients undergoing resection for gastric cancer between January 2004 and January 2012 at S. Andrea Hospital, ‘‘La Sapienza’’, University of Rome, were retrospectively reviewed. All patients had undergone gastrectomy and modified D2 lymphadenectomy with curative intent according to tumor location and extent of disease. Combined resection of other organs, such as
Value of Preoperative Inflammation-Based Prognostic Scores
1999
TABLE 1 Inflammation-based prognostic score construction Systemic inflammation-based prognostic score
TABLE 2 Patients’ clinicopathologic characteristics (n = 102) Score
The modified Glasgow prognostic score (mGPS)
Variable
Data
Age (years)
69 ± 10.6
C-reactive protein B1 mg/dl and albumin C3.5 g/dl
0
Sex (M/F)
62/40
C-reactive protein B1 mg/dl and albumin \3.5 g/dl
0
Tumor stage (I/II/III/IV)
34/15/35/18
C-reactive protein [1 mg/dl
1
pT1/T2/T3/T4
21/34/31/16
C-reactive protein [1 mg/dl and albumin \3.5 g/dl
2
pN0/N1/N2/N3
37/20/17/28
Neutrophil lymphocyte ratio (NLR)
Tumor location
Neutrophil count: lymphocyte count \5:1
0
Upper and diffuse
28
Neutrophil count: lymphocyte count C5:1
1
Middle and lower
74
Platelet lymphocyte ratio (PLR)
Surgical procedure
Platelet count: lymphocyte count \150:1
0
Platelet count: lymphocyte count 150–300:1 Platelet count: lymphocyte count [300:1
1 2
Prognostic index (PI)
Total gastrectomy Subtotal gastrectomy Number of lymph nodes retrieved
50 52 27.9 ± 13
Histologic grade of differentiation (G1/G2/G3/G4) 16/22/54/10
C-reactive protein B1 mg/dl and white cell count B11 9 109/l
0
C-reactive protein B1 mg/dl and white cell count [11 9 109/l
1
C-reactive protein [1 mg/dl and white cell count B11 9 109/l
1
C-reactive protein [1 mg/dl and white cell count [11 9 109/l
2
LNR
0.10 (0.00–0.90)
Albumin (g/dl)
3.3 ± 0.5
CRP (mg/dl)
0.98 (0.01–6.19)
WBC (103/ll)
6.7 ± 2.4
PLT count (103/ll)
265.6 ± 100.7
CEA (ng/ml)
2.39 (0.21–948)
CA 19.9 (Ul/ml)
14.5 (0.08– 2,990)
Albumin (g/dl) ? 5 9 total lymphocyte count 9 109/l C45 0
mGPS (0/1/2)
49/25/28
Albumin (g/dl) ? 5 9 total lymphocyte count 9 109/l \45 1
PI (0/1/2) PNI (0/1)
53/27/22 40/62
PLR (0/1/2)
45/44/13
NLR (0/1)
74/28
Prognostic nutrition index (PNI)
the spleen, pancreas, and colon, was performed in case of direct invasion. None of the patients included in the study had liver or distant metastases. T-stage and nodal status were determined on the basis of histopathologic evaluation by using the American Joint Committee on Cancer Staging Manual, 7th edition. The lymph node ratio (LNR) was calculated according to the relationship between positive nodes and total nodes of the specimen. No patient included in the study underwent neoadjuvant chemotherapy, and all patients with stage II or higher disease were treated with adjuvant chemotherapy after surgery, as proposed in the Associazone Italiana di Oncologica Medica guidelines (http://www.aiom.it). Demographic details, operation type, histopathologic characteristics of the tumor, and survival data were collected. Blood samples were obtained before initial treatment for measurement of CRP, albumin, white blood cell count, neutrophils, lymphocytes, platelet count, carcinoembryonic antigen, and carbohydrate antigen 19-9. The mGPS, NLR, PRL, PI, and PNI were constructed as described in Table 1. Patients who had clinical infections or other inflammatory conditions; whose clinical, laboratory, and histopathologic data were missing; and who died within 30 days after the operation were excluded from the study. Ultimately, 102 patients were eligible. All patients were evaluated 1 month after surgery. The last follow-up was
WBC white blood cell count, PLT platelet, CEA carcinoembryonic antigen, CA carbohydrate antigen
in May 2013. Clinicopathologic characteristics of the 102 patients are shown in Table 2. Statistical Analysis OS was defined as the interval between the date of operation and the date of death or last patient visit. DFS was defined as the time from operation until tumor relapse, either local or distant. Comparison of categorical variables was performed by using the v2 test. Kaplan–Meier curves were generated for OS and DFS to compare patients with each inflammation-based score, and differences in the survival rate between the groups were compared by the log–rank test. A backward stepwise Cox regression model was used to identify variables influencing OS and DFS. Multivariate analysis was performed by using variables that have a significant independent relationship with OS and DFS. Significance was defined as a p value of \0.05. All statistical analysis were performed with SPSS for Mac version 17.0 (SPSS, Inc., Chicago, IL).
2000
P. Aurello et al.
0.4
0.2
0.0 20
40
60
80
100
0.6
0.4
0.8
0.4
0.2
0.0
0.0 0
20
40
Months
60
80
100
120
Pls 0 1 2 0-interrupted 1-interrupted 2-interrupted
0.6
0.2
120
0
20
40
60
80
100
120
Months
Months
D
1.0
PNI 0 1 0-interrupted 1-interrupted
0.8
0.6
0
C 1.0
Overall survival
0.8
Overall survival
B
NLRs 0 1 0-interrupted 1-interrupted
1.0
Overall survival
A
E 1.0
1.0 0.8
0.6 mGPS 0 1 2 0-interrupted 1-interrupted 2-interrupted
0.4
0.2
Overall survival
Overall survival
0.8
0.6 PLRs 0 1 2 0-interrupted 1-interrupted 2-interrupted
0.4
0.2
0.0
0.0 0
20
40
60
80
100
120
0
20
40
Months
60
80
100
120
Months
FIG. 1 Relationship between inflammation-based prognostic scores and overall survival in patients with gastric cancer: a NLR, b PNI, c PI, d mGPS, e PLR
RESULTS A total of 53 (51.9 %) patients had an increased CRP at presentation ([1 mg/dl), 55 (53.9 %) patients had hypoalbuminemia (\3.5 g/dl), and 28 (27.4 %) patients had both increased CRP and hypoalbuminemia. A total of 29 (28.4 %) patients had an increased white blood cell count at presentation ([11 9 109/l). Forty-nine patients (48.1 %) were allocated to mGPS 0, 25 (24.5 %) patients to mGPS 1, and 28 (27.4 %) to mGPS 2. Forty-five patients (44.2 %) were allocated to PLR 0, 44 (43.1 %) patients to PLR 1, and 13 (12.7 %) to PLR 2. Seventy-four patients (72.5 %) were allocated to NLR 0 and 28 (27.5 %) to NLR 1. Forty patients (39.3 %) were allocated to PNI 0 and 62 (60.7 %) patients to PNI 1. Fifty-three patients (51.9 %) were allocated to PI 0, 27 (26.5 %) patients to PI 1, and 22 (21.6 %) to PI 2. Symptoms at presentation were bleeding in 32 (31.3 %) patients, gastric outlet obstruction in 12 (11.7 %), and pain and discomfort in 58 (56.8 %). Radiologic lymph node invasion at the time of diagnosis was suspected in 62
(60.7 %) patients. Seven patients required a combined splenectomy, 2 underwent a right hemicolectomy, and 2 patients underwent a pancreaticoduodenectomy because of direct tumor invasion. After pathologic examination, the mean tumor diameter was 2.7 ± 1.2 mm, and all specimens’ transection lines were free from tumor cells (R0). A total of 68 patients underwent postoperative chemoradiotherapy because of tumor stage II or higher according to the Associazone Italiana di Oncologica Medica guidelines (http://www.aiom.it). The median duration of follow-up was 40.8 months (range 8–107 months). Of 102 patients, 62 patients (60.8 %) had died and 40 (39.2 %) patients were alive at the end of follow-up. All patients in the study died from cancer-related causes. The 1-, 3-, and 5-year OS rates were 88, 72, and 59 %, respectively. After R0 resection, the 1-, 3-, and 5-year DFS rates were 93, 74, and 56 %, respectively. The relationship between the inflammation-based prognostic scores and OS is shown in Fig. 1. The relationship between these scores and the DFS is shown in Fig. 2.
Value of Preoperative Inflammation-Based Prognostic Scores
NLRs 0 1 0-interrupted 1-interrupted
1.0
0.6
0.4
0.8
0.2
1.0
0.8
0.6
0.4
20
40
60
80
100
120
0
20
40
Months
60
80
100
120
Months
D
0.4
Pls 0 1 2 0-interrupted 1-interrupted 2-interrupted
0.0
0.0 0
0.6
0.2
0.2
0.0
0
20
40
60
80
100
120
Months
E 1.0
1.0
0.8
0.8
0.6
0.4
mGPS 0 1 2 0-troncata 1-interrupted 2-interrupted
0.2
0.0
Disease free survival
Disease free survival
C
PNI 0 1 0-interrupted 1-interrupted
1.0
Disease free survival
Disease free survival
0.8
B
Disease free survival
A
2001
0.6
0.4
PLRs 0 1 2 0-interrupted 1-interrupted 2-interrupted
0.2
0.0 0
20
40
60
80
100
120
0
20
Months
40
60
80
100
120
Months
FIG. 2 Relationship between inflammation-based prognostic scores and disease-free survival in patients with gastric cancer: a NLR, b PNI, c PI, d mGPS, e PLR
The univariate analysis showed that LNR (p = 0.03), TNM (p = 0.01), NLR (p = 0.003), PI (p \ 0.001), mGPS (p \ 0.001), and PLR (p \ 0.001) were associated with OS. A multivariate analysis of these significant variables showed that only the mGPS (p \ 0.001) and PI (p \ 0.001) were independently associated with OS (\T3[ Table 3). Regarding the DFS, the univariate analysis showed that LNR (p = 0.002), number of lymph nodes retrieved (p = 0.05), NLR (p \ 0.001), PNI (p = 0.01), PI (p \ 0.001), mGPS (p \ 0.001), and PLR (p \ 0.001) were significant. A multivariate analysis of these significant variables showed that the mGPS (p = 0.002) and PI (p \ 0.001) were independently associated with DFS (Table 3). At a subgroup analysis, important clinicopathologic prognostic characteristics were significantly different between patients with mGPS 0, 1, or 2 (LNR, p \ 0.001; tumor stage, p \ 0.001; grading, p = 0.03). More in detail, patients with mGPS 0 had the following clinicopathologic characteristics: mean LNR, 0.12 ± 0.1; 16 of 49 (32.6 %) patients had more advanced cancer (stage III–IV); and 26 of 49 (53.1 %) patients had a higher tumor grading (grade III–IV). Patients allocated in mGPS 1
had a mean LNR of 0.16 ± 0.2, 13 of 25 (52 %) patients had more advanced cancer, and 15 of 25 (60 %) patients had a higher tumor grade. Concerning patients with an mGPS score of 2, the mean LNR was 0.49 ± 0.2, 26 of 28 (92.8 %) had a more advanced cancer, and 23 of 28 (82.1 %) patients had a higher tumor grade. DISCUSSION Mantovani et al.9 demonstrated the role of the host inflammatory response in the development and progression of cancer. The major role of the acute-phase response is to remove the insult from the body and to reinstate the initial status; however, if this response persists, such as in case of cancer, it contributes to the development of the disease. According to the concept that ongoing inflammation mediates a progressive loss of vital protein, leading to the death of patients with advanced cancer, Forrest and colleagues first developed a scoring system based on the combination of CRP and albumin in patients with inoperable non-small-cell lung cancer.13,18 On further investigation, the GPS was modified and termed mGPS, on the observation that
2002
P. Aurello et al.
TABLE 3 Prognostic factors for overall survival (OS) and disease-free survival (DFS) in patients with gastric cancer Variables
Univariate analysis
Multivariate analysis
Univariate analysis
Multivariate analysis
OS
OS
DFS
DFS
Hazard ratio (95 % CI)
p value
Hazard ratio (95 % CI)
p value
Hazard ratio (95 % CI)
p value
Age
1.04 (1.00–1.07)
0.10
1.00 (0.97–1.04)
0.62
Sex
0.80 (0.42–1.53)
0.51
0.95 (0.48–1.89)
0.90
Hazard ratio (95 % CI)
p value
Tumor stage I
1
0.01
1
0.36
1
0.20 (0.08– 0.52)
0.01
0.73 (0.16– 3.37)
0.69
III
0.27 (0.07–0.97)
0.04
0.46 (0.08–2.51)
0.37
0.16 (0.03–0.70)
IV
0.60 (0.28–1.26)
0.17
0.39 (0.14–1.10)
0.07
0.70 (0.33–1.50)
0.37
0.06 (0.07–0.59)
0.11
0.32 (0.06–1.33)
0.21
Tumor location
0.1 (0.03– 0.31)
0.98
II
0.1 0.21
Surgical procedure
1.62 (0.86–3.05)
0.12
1.75 (0.89–3.43)
0.10
Lymph nodes retrieved
1.03 (0.97–1.02)
0.84
1.02 (1.0–1.05)
0.05
1.02 (0.99–1.03)
0.12
0.80 (0.14–4.40)
0.80
Grading G1
1
0.73
1
0.98
G2
2.85 (0.33–24.72)
0.34
1.08 (0.09–12.02)
0.95
G3 G4
2.60 (0.32–20.5) 3.52 (0.4–26.13)
0.36 0.21
1.61 (0.1–13.48) 3.51 (0.48–24.21)
0.65 0.21
LNR
3.14 (1.07–9.18)
0.03
5.84 (1.92–17.63)
0.002
Albumin (g/dl)
0.65 (0.39–1.07)
0.09
0.71 (0.41–1.23)
0.22
CRP (mg/dl)
1.22 (0.93–1.59)
0.14
1.08 (0.77–1.51)
0.63
WBC (103/ll)
1.11 (0.99–1.25)
0.53
1.13 (0.99–1.28)
0.57
PLT count (103/ll)
0.99 (0.99–1.00)
0.63
1 (0.99–1.03)
0.66
CEA (ng/ml)
0.99 (0.99–1.00)
0.41
0.99 (0.99–1.04)
0.44
1 (0.99–1.00)
0.99
1(0.99–1.01)
0.89
CA 19.9 (Ul/ml)
0.59 (0.10–3.51)
0.56
mGPS 0
1
\0.001
1
\0.001
1
\0.001
1
0.002
1
0.13 (0.06–0.31)
\0.001
1.70 (1.20–3.42)
0.005
0.03 (0.01–0.11)
\0.001
1.78 (0.82–2.85)
0.002
2
0.67 (0.33–1.37)
0.02
1.91 (1.38–3.18)
0.008
0.33 (0.15–0.73)
0.006
1.73 (1.16–2.39)
0.001
PI 0 1 2 PNI (0/1)
1
\0.001
0.04 (0.01–0.12) 0.48 (0.24–0.94)
\0.001 0.03
0.52 (0.26–1.04)
0.06
1 0.04 (0.01–0.20) 0.37 (0.16–0.82)
\0.001 \0.001 0.01
1
\0.001
0.37 (0.01–0.11) 0.38 (0.18–0.79)
\0.001 0.009
0.33 (0.14–0.76)
0.01
1
\0.001
0.06 (0.01–0.28) \0.001 0.21 (0.08–0.55) 0.001 0.11 (0.03–0.40)
0.1
PLR 0
1
\0.001
1
0.26
1
\0.001
1
0.19
1
0.09 (0.03–0.25)
\0.001
0.43 (0.10–1.73)
0.23
0.06 (0.02–0.27)
\0.01
0.38 (0.08–1.63)
2
0.66 (0.30–1.43)
0.3
1.13 (0.45–2.79)
0.79
0.1 (0.22–1.12)
0.01
2.05 (0.71–5.89)
0.18
NLR (0/1)
0.39 (0.21–0.72)
0.003
1.51 (0.69–3.28)
0.29
0.28 (0.14–0.54)
\0.001
0.95 (0.35–2.56)
0.91
0.19
WBC white blood cell count, PLT platelet, CEA carcinoembryonic antigen, CA carbohydrate antigen
hypoalbuminemia without an increased CRP concentration was rare and that hypoalbuminemia on its own was not associated with poor survival.19–23 Recently, Proctor et al.24 demonstrated that the mGPS is one of the most useful scoring systems in determining long-term outcomes in patients with different malignancies.
The systemic inflammatory response also contributes to the changes in the relative levels of circulating white blood cells as indicated by the increase of neutrophils and the decrease of lymphocytes in the progression of cancer. Previous studies have shown that the relationship between neutrophilia and lymphopenia, also called the NLR, is a
Value of Preoperative Inflammation-Based Prognostic Scores
prognostic independent factor in gastric cancer.25,26 Recently, Mano et al.27 have demonstrated that NLR is an independent predictor of survival after hepatectomy in patients with hepatocellular cancer. Jung et al.25 have shown the same findings in patients with late-stage gastric cancer undergoing resection. In their article, the statistical significance of the NLR ratio was evident only in patients with stage III gastric cancer on both OS and DFS . Ishizuka et al.28 and Pinato et al.29 have shown that a combination of albumin and lymphocyte count, called PNI, is associated with the survival of patients with hepatocellular cancer. Smith et al.15 have demonstrated that the PLR is a significant prognostic marker in patients with pancreatic cancer, and Kasymjanova et al.16 have shown that the PI, as evidenced by a combination of serum CRP and white cell count, is a significant prognostic factor in patients with lung cancer. Our study shows that the scores based on CRP, such as mGPS and PI, are more accurate in determining good and poor prognosis in patients with gastric cancer compared with NLR and PLR, which are based on components of the circulating white cell count or in combination with albumin (PNI). This could be explained according to the fact that the prototypical acute phase protein is CRP, which is a nonspecific but sensitive marker of systemic inflammatory response. This process is induced in hepatocytes mainly by interleukin (IL)-6, a proinflammatory cytokine, stimulated by IL-1b in an autocrine manner. In gastric cancer, high serum IL-6 levels have shown to be correlated with progression of disease; this could be due to the secretion of IL-6 by gastric cancer cells and/or stromal cells.30 Because advanced cancer is often associated with a systemic inflammatory response, high CRP levels are common in patients with advanced disease. This important issue is furthermore clearly visible in this study thanks to the results obtained with the subgroup analysis. In fact, patients with a higher mGPS score (constructed with albumin and CRP) are patients that have been diagnosed with a more advanced cancer stage, with a higher LNR, and with a higher tumor grade at pathologic examination (LNR, p \ 0.001; tumor stage, p \ 0.001; tumor grade, p = 0.03). Crumley and colleagues demonstrated that preoperative evaluation of systemic inflammatory response, as evidenced by an increased CRP level, is important for the prognosis of patients with gastric cancer.5 It is of interest that the combination of CRP and albumin (mGPS), whose relationship with survival has been validated by previous studies, had similar prognostic value to that of CRP and white cell count (PI). These results would suggest that the white blood cell count, rather than NLR or PLR, may be a useful addition to mGPS. Most gastric cancer patients, even after curative resection, experienced recurrence with a rate of 60 % and a median length of survival after recurrence of 7.4 months in case of distant metastases and 10.4 months in case of local
2003
recurrence.31 Identification of sensitive markers for patients who have undergone curative gastrectomy and who are at high risk of recurrence would provide useful information for planning follow-up after surgery or intensive adjuvant chemotherapy. In this study the Kaplan–Meier analysis showed that a higher inflammation-based prognostic score was associated with poorer DFS. In particular, the mGPS and PI were independent predictors of outcome. It is well known that the TNM classification is a good indicator for patients with gastric cancer because it includes pathologic findings that are revealed after surgery. In contrast, the inflammation-based prognostic scores are able to classify patients with gastric cancer before surgery, although their mechanism has not been clearly elucidated. Probably, the host local inflammatory response to the tumor might be associated with systemic inflammatory response, and, in particular, inflammatory cells may interact directly and indirectly with the primary tumor to promote angiogenesis, extracellular matrix remodeling, and preparation of the metastatic niche.32 Otherwise, in accordance with previous studies, the inflammation-based prognostic scores NLR, PNI, and PLR were statistically significant at the univariate analysis, thus demonstrating their prognostic value in predicting cancerspecific outcome in patients with gastric cancer.15,16,25–27 Further studies are needed to clarify the role of systemic inflammatory response in patients with gastric cancer and to clearly identify the possibility that inflammation could become a therapeutic target as the tumor itself. This study has some limitations, mostly consisting of its relatively small sample size, retrospective design, and single-center study. Therefore, a large-scale prospective validation study is needed to confirm these results. In summary, this study’s results show that mGPS and PI are independent markers of poor prognosis in patients with gastric cancer and are superior to the other inflammationbased prognostic scores in terms of prognostic evaluation ability. A measurement of these inflammation-based prognostic scores should be included in the routine preoperative assessment of all patients with gastric cancer. Moreover, the preoperatively accurate stratification of patients is important for planning the best treatment and follow-up for patients with gastric cancer. DISCLOSURE All of the authors declare that they have no potential commercial conflicts of interest relevant to this article.
REFERENCES 1. Jemal A, Siegel R, Xu J, Ward E. Cancer statistics, 2010. CA Cancer J Clin. 2010;60:277–300. 2. Kelley JR, Duggan JM. Gastric cancer epidemiology and risk factors. J Clin Epidemiol. 2003;56:1–9.
2004 3. McMillan DC, Watson WS, O’Gorman P, Preston T, Scott HR, McArdle CS. Albumin concentrations are primarily determined by the body cell mass and the systemic inflammatory response in cancer patients with weight loss. Nutr Cancer. 2001;39:210–3. 4. Onate-Ocana LF, Aiello-Crocifoglio V, Gallardo-Rincon D, et al. Serum albumin as a significant prognostic factor for patients with gastric carcinoma. Ann Surg Oncol. 2007;14:381–9. 5. Crumley AB, Stuart RC, McKernan M, McMillan DC. Is hypoalbuminemia an independent prognostic factor in patients with gastric cancer? World J Surg. 2010;34:2393–8. 6. Roxburgh CS, McMillan DC. Role of systemic inflammatory response in predicting survival in patients with primary operable cancer. Future Oncol. 2010;6:149–63. 7. Lee DY, Hong SW, Chang YG, Lee WY, Lee B. Clinical significance of preoperative inflammatory parameters in gastric cancer patients. J Gastric Cancer. 2013;13:111–6. 8. Coussens LM, Werb Z. Inflammation and cancer. Nature. 2002;420:860–7. 9. Mantovani A, Allavena P, Sica A, Balkwill F. Cancer-related inflammation. Nature. 2008;454:436–44. 10. Colotta F, Allavena P, Sica A, Garlanda C, Mantovani A. Cancerrelated inflammation, the seventh hallmark of cancer: links to genetic instability. Carcinogenesis. 2009;30:1073–81. 11. Chiba T, Marusawa H, Ushijima T. Inflammation-associated cancer development in digestive organs: mechanisms and roles for genetic and epigenetic modulation. Gastroenterology. 2012; 143:550–63. 12. Yamashita H, Katai H. Systemic inflammatory response in gastric cancer. World J Surg. 2010;34:2399–400. 13. Forrest LM, McMillan DC, McArdle CS, Angerson WJ, Dagg K, Scott HR. A prospective longitudinal study of performance status, an inflammation-based score (GPS) and survival in patients with inoperable non-small-cell lung cancer. Br J Cancer. 2005;92:1834–6. 14. Walsh SR, Cook EJ, Goulder F, Justin TA, Keeling NJ. Neutrophil-lymphocyte ratio as a prognostic factor in colorectal cancer. J Surg Oncol. 2005;91:181–4. 15. Smith RA, Bosonnet L, Raraty M, et al. Preoperative plateletlymphocyte ratio is an independent significant prognostic marker in resected pancreatic ductal adenocarcinoma. Am J Surg. 2009;197:466–72. 16. Kasymjanova G, MacDonald N, Agulnik JS, et al. The predictive value of pre-treatment inflammatory markers in advanced nonsmall-cell lung cancer. Curr Oncol. 2010;17:52–8. 17. Nozoe T, Ninomiya M, Maeda T, Matsukuma A, Nakashima H, Ezaki T. Prognostic nutritional index: a tool to predict the biological aggressiveness of gastric carcinoma. Surg Today. 2010;40:440–3. 18. Forrest LM, McMillan DC, McArdle CS, Angerson WJ, Dunlop DJ. Evaluation of cumulative prognostic scores based on the systemic inflammatory response in patients with inoperable nonsmall-cell lung cancer. Br J Cancer. 2003;89:1028–30.
P. Aurello et al. 19. Crumley AB, McMillan DC, McKernan M, McDonald AC, Stuart RC. Evaluation of an inflammation-based prognostic score in patients with inoperable gastro-oesophageal cancer. Br J Cancer. 2006;94:637–41. 20. Elahi MM, McMillan DC, McArdle CS, Angerson WJ, Sattar N. Score based on hypoalbuminemia and elevated C-reactive protein predicts survival in patients with advanced gastrointestinal cancer. Nutr Cancer. 2004;48:171–3. 21. Mohri Y, Tanaka K, Ohi M, et al. Inflammation-based prognostic score as a predictor of postoperative gastric cancer recurrence. Anticancer Res. 2012;32:4581–4. 22. McMillan DC. The systemic inflammation-based Glasgow Prognostic Score: a decade of experience in patients with cancer. Cancer Treat Rev. 2013;39:534–40. 23. Kinoshita A, Onoda H, Imai N, et al. Comparison of the prognostic value of inflammation-based prognostic scores in patients with hepatocellular carcinoma. Br J Cancer. 2012;107:988–93. 24. Proctor MJ, Morrison DS, Talwar D, et al. A comparison of inflammation-based prognostic scores in patients with cancer. A Glasgow Inflammation Outcome Study. Eur J Cancer. 2011;47: 2633–41. 25. Jung MR, Park YK, Jeong O, et al. Elevated preoperative neutrophil to lymphocyte ratio predicts poor survival following resection in late stage gastric cancer. J Surg Oncol. 2011;104: 504–10. 26. Aizawa M, Gotohda N, Takahashi S, Konishi M, Kinoshita T. Predictive value of baseline neutrophil/lymphocyte ratio for T4 disease in wall-penetrating gastric cancer. World J Surg. 2011;35: 2717–22. 27. Mano Y, Shirabe K, Yamashita Y, et al. Preoperative neutrophilto-lymphocyte ratio is a predictor of survival after hepatectomy for hepatocellular carcinoma: a retrospective analysis. Ann Surg. 2013;258:301–5. 28. Ishizuka M, Kubota K, Kita J, Shimoda M, Kato M, Sawada T. Impact of an inflammation-based prognostic system on patients undergoing surgery for hepatocellular carcinoma: a retrospective study of 398 Japanese patients. Am J Surg. 2012;203:101–6. 29. Pinato DJ, North BV, Sharma R. A novel, externally validated inflammation-based prognostic algorithm in hepatocellular carcinoma: the prognostic nutritional index (PNI). Br J Cancer. 2012;106:1439–45. 30. Kim DK, Oh SY, Kwon HC, et al. Clinical significances of preoperative serum interleukin-6 and C-reactive protein level in operable gastric cancer. BMC Cancer. 2009;9:155. 31. Rohatgi PR, Yao JC, Hess K, et al. Outcome of gastric cancer patients after successful gastrectomy: influence of the type of recurrence and histology on survival. Cancer. 2006;107:2576–80. 32. DeNardo DG, Johansson M, Coussens LM. Immune cells as mediators of solid tumor metastasis. Cancer Metastasis Rev. 2008;27:11–8.
ORIGINAL ARTICLE – GASTROINTESTINAL ONCOLOGY
Value of Preoperative Inflammation-Based Prognostic Scores in Predicting Overall Survival and Disease-Free Survival in Patients with Gastric Cancer Paolo Aurello, MD, Simone Maria Tierno, MD, Giammauro Berardi, MD, Federico Tomassini, MD, Paolo Magistri, MD, Francesco D’Angelo, MD, and Giovanni Ramacciato, MD Third Division of General Surgery, Sant’Andrea Hospital, University of Rome ‘‘La Sapienza’’, Rome, Italy
ABSTRACT Background. This study was designed to identify which are the best preoperative inflammation-based prognostic scores in terms of overall survival (OS) and disease-free survival (DFS) in patients with gastric cancer. Methods. Between January 2004 and January 2013, 102 consecutive patients underwent resection for gastric cancer at S. Andrea Hospital, ‘‘La Sapienza’’, University of Rome. Their records were retrospectively reviewed. Results. After a median follow up of 40.8 months (8– 107 months), patients’ 1-, 3-, and 5-year OS rates were 88, 72, and 59 %, respectively. After R0 resection, the 1-, 3-, and 5-year DFS rates were 93, 74, and 56 %, respectively. A multivariate analysis of the significant variables showed that only the modified Glasgow prognostic scores (p \ 0.001) and PI (p \ 0.001) were independently associated with OS. Regarding DFS, multivariate analysis of the significant variables showed that the modified Glasgow prognostic score (p = 0.002) and prognostic index (p \ 0.001) were independently associated with DFS. Conclusions. The results of this study show that modified Glasgow prognostic score and prognostic index are independent predictors of OS and DFS in patients with gastric cancer.
Gastric cancer is the fourth most common malignancy and the second most common cause of cancer-related death worldwide.1 Tumor stage, tumor size, grade of differentiation,
Ó Society of Surgical Oncology 2014 First Received: 27 September 2013; Published Online: 11 February 2014 S. M. Tierno, MD e-mail: [email protected]
and vascular invasion are well-established prognostic factors.2 The relationship between hypoalbuminemia and poor prognosis has also been established.3–5 In addition, recently there has been growing interest in the host’s inflammatory response to tumor, as evidenced by increased concentrations of C-reactive protein (CRP). Previous studies have demonstrated the relationship between gastric cancer and chronic inflammation, although the exact mechanism involved is not clear.6–11 The presence of an inflammatory response is crucial in the development of cancer-associated malnutrition, resulting in poor performance status and increased mortality in patients with cancer.12 On the basis of the relationship between inflammation and cancer progression, several inflammation-based scores have been proposed as prognostic. The main scores that have been associated with cancer-specific survival are the modified Glasgow Prognostic Scores (mGPS), neutrophil lymphocyte ratio (NLR), platelet lymphocyte ratio (PLR), prognostic index (PI), and prognostic nutritional index (PNI).13–17 To our knowledge, there are no studies elucidating which of these prognostic scores is more suitable in predicting outcomes in patients with gastric cancer. The aim of this study was to compare the prognostic value of these preoperative inflammation-based prognostic scores (the mGPS, NLR, PLR, PI, and PNI) in terms of overall survival (OS) and disease-free survival (DFS) in patients with gastric cancer in various stages of disease. METHODS The records of a series of 126 consecutive patients undergoing resection for gastric cancer between January 2004 and January 2012 at S. Andrea Hospital, ‘‘La Sapienza’’, University of Rome, were retrospectively reviewed. All patients had undergone gastrectomy and modified D2 lymphadenectomy with curative intent according to tumor location and extent of disease. Combined resection of other organs, such as
Value of Preoperative Inflammation-Based Prognostic Scores
1999
TABLE 1 Inflammation-based prognostic score construction Systemic inflammation-based prognostic score
TABLE 2 Patients’ clinicopathologic characteristics (n = 102) Score
The modified Glasgow prognostic score (mGPS)
Variable
Data
Age (years)
69 ± 10.6
C-reactive protein B1 mg/dl and albumin C3.5 g/dl
0
Sex (M/F)
62/40
C-reactive protein B1 mg/dl and albumin \3.5 g/dl
0
Tumor stage (I/II/III/IV)
34/15/35/18
C-reactive protein [1 mg/dl
1
pT1/T2/T3/T4
21/34/31/16
C-reactive protein [1 mg/dl and albumin \3.5 g/dl
2
pN0/N1/N2/N3
37/20/17/28
Neutrophil lymphocyte ratio (NLR)
Tumor location
Neutrophil count: lymphocyte count \5:1
0
Upper and diffuse
28
Neutrophil count: lymphocyte count C5:1
1
Middle and lower
74
Platelet lymphocyte ratio (PLR)
Surgical procedure
Platelet count: lymphocyte count \150:1
0
Platelet count: lymphocyte count 150–300:1 Platelet count: lymphocyte count [300:1
1 2
Prognostic index (PI)
Total gastrectomy Subtotal gastrectomy Number of lymph nodes retrieved
50 52 27.9 ± 13
Histologic grade of differentiation (G1/G2/G3/G4) 16/22/54/10
C-reactive protein B1 mg/dl and white cell count B11 9 109/l
0
C-reactive protein B1 mg/dl and white cell count [11 9 109/l
1
C-reactive protein [1 mg/dl and white cell count B11 9 109/l
1
C-reactive protein [1 mg/dl and white cell count [11 9 109/l
2
LNR
0.10 (0.00–0.90)
Albumin (g/dl)
3.3 ± 0.5
CRP (mg/dl)
0.98 (0.01–6.19)
WBC (103/ll)
6.7 ± 2.4
PLT count (103/ll)
265.6 ± 100.7
CEA (ng/ml)
2.39 (0.21–948)
CA 19.9 (Ul/ml)
14.5 (0.08– 2,990)
Albumin (g/dl) ? 5 9 total lymphocyte count 9 109/l C45 0
mGPS (0/1/2)
49/25/28
Albumin (g/dl) ? 5 9 total lymphocyte count 9 109/l \45 1
PI (0/1/2) PNI (0/1)
53/27/22 40/62
PLR (0/1/2)
45/44/13
NLR (0/1)
74/28
Prognostic nutrition index (PNI)
the spleen, pancreas, and colon, was performed in case of direct invasion. None of the patients included in the study had liver or distant metastases. T-stage and nodal status were determined on the basis of histopathologic evaluation by using the American Joint Committee on Cancer Staging Manual, 7th edition. The lymph node ratio (LNR) was calculated according to the relationship between positive nodes and total nodes of the specimen. No patient included in the study underwent neoadjuvant chemotherapy, and all patients with stage II or higher disease were treated with adjuvant chemotherapy after surgery, as proposed in the Associazone Italiana di Oncologica Medica guidelines (http://www.aiom.it). Demographic details, operation type, histopathologic characteristics of the tumor, and survival data were collected. Blood samples were obtained before initial treatment for measurement of CRP, albumin, white blood cell count, neutrophils, lymphocytes, platelet count, carcinoembryonic antigen, and carbohydrate antigen 19-9. The mGPS, NLR, PRL, PI, and PNI were constructed as described in Table 1. Patients who had clinical infections or other inflammatory conditions; whose clinical, laboratory, and histopathologic data were missing; and who died within 30 days after the operation were excluded from the study. Ultimately, 102 patients were eligible. All patients were evaluated 1 month after surgery. The last follow-up was
WBC white blood cell count, PLT platelet, CEA carcinoembryonic antigen, CA carbohydrate antigen
in May 2013. Clinicopathologic characteristics of the 102 patients are shown in Table 2. Statistical Analysis OS was defined as the interval between the date of operation and the date of death or last patient visit. DFS was defined as the time from operation until tumor relapse, either local or distant. Comparison of categorical variables was performed by using the v2 test. Kaplan–Meier curves were generated for OS and DFS to compare patients with each inflammation-based score, and differences in the survival rate between the groups were compared by the log–rank test. A backward stepwise Cox regression model was used to identify variables influencing OS and DFS. Multivariate analysis was performed by using variables that have a significant independent relationship with OS and DFS. Significance was defined as a p value of \0.05. All statistical analysis were performed with SPSS for Mac version 17.0 (SPSS, Inc., Chicago, IL).
2000
P. Aurello et al.
0.4
0.2
0.0 20
40
60
80
100
0.6
0.4
0.8
0.4
0.2
0.0
0.0 0
20
40
Months
60
80
100
120
Pls 0 1 2 0-interrupted 1-interrupted 2-interrupted
0.6
0.2
120
0
20
40
60
80
100
120
Months
Months
D
1.0
PNI 0 1 0-interrupted 1-interrupted
0.8
0.6
0
C 1.0
Overall survival
0.8
Overall survival
B
NLRs 0 1 0-interrupted 1-interrupted
1.0
Overall survival
A
E 1.0
1.0 0.8
0.6 mGPS 0 1 2 0-interrupted 1-interrupted 2-interrupted
0.4
0.2
Overall survival
Overall survival
0.8
0.6 PLRs 0 1 2 0-interrupted 1-interrupted 2-interrupted
0.4
0.2
0.0
0.0 0
20
40
60
80
100
120
0
20
40
Months
60
80
100
120
Months
FIG. 1 Relationship between inflammation-based prognostic scores and overall survival in patients with gastric cancer: a NLR, b PNI, c PI, d mGPS, e PLR
RESULTS A total of 53 (51.9 %) patients had an increased CRP at presentation ([1 mg/dl), 55 (53.9 %) patients had hypoalbuminemia (\3.5 g/dl), and 28 (27.4 %) patients had both increased CRP and hypoalbuminemia. A total of 29 (28.4 %) patients had an increased white blood cell count at presentation ([11 9 109/l). Forty-nine patients (48.1 %) were allocated to mGPS 0, 25 (24.5 %) patients to mGPS 1, and 28 (27.4 %) to mGPS 2. Forty-five patients (44.2 %) were allocated to PLR 0, 44 (43.1 %) patients to PLR 1, and 13 (12.7 %) to PLR 2. Seventy-four patients (72.5 %) were allocated to NLR 0 and 28 (27.5 %) to NLR 1. Forty patients (39.3 %) were allocated to PNI 0 and 62 (60.7 %) patients to PNI 1. Fifty-three patients (51.9 %) were allocated to PI 0, 27 (26.5 %) patients to PI 1, and 22 (21.6 %) to PI 2. Symptoms at presentation were bleeding in 32 (31.3 %) patients, gastric outlet obstruction in 12 (11.7 %), and pain and discomfort in 58 (56.8 %). Radiologic lymph node invasion at the time of diagnosis was suspected in 62
(60.7 %) patients. Seven patients required a combined splenectomy, 2 underwent a right hemicolectomy, and 2 patients underwent a pancreaticoduodenectomy because of direct tumor invasion. After pathologic examination, the mean tumor diameter was 2.7 ± 1.2 mm, and all specimens’ transection lines were free from tumor cells (R0). A total of 68 patients underwent postoperative chemoradiotherapy because of tumor stage II or higher according to the Associazone Italiana di Oncologica Medica guidelines (http://www.aiom.it). The median duration of follow-up was 40.8 months (range 8–107 months). Of 102 patients, 62 patients (60.8 %) had died and 40 (39.2 %) patients were alive at the end of follow-up. All patients in the study died from cancer-related causes. The 1-, 3-, and 5-year OS rates were 88, 72, and 59 %, respectively. After R0 resection, the 1-, 3-, and 5-year DFS rates were 93, 74, and 56 %, respectively. The relationship between the inflammation-based prognostic scores and OS is shown in Fig. 1. The relationship between these scores and the DFS is shown in Fig. 2.
Value of Preoperative Inflammation-Based Prognostic Scores
NLRs 0 1 0-interrupted 1-interrupted
1.0
0.6
0.4
0.8
0.2
1.0
0.8
0.6
0.4
20
40
60
80
100
120
0
20
40
Months
60
80
100
120
Months
D
0.4
Pls 0 1 2 0-interrupted 1-interrupted 2-interrupted
0.0
0.0 0
0.6
0.2
0.2
0.0
0
20
40
60
80
100
120
Months
E 1.0
1.0
0.8
0.8
0.6
0.4
mGPS 0 1 2 0-troncata 1-interrupted 2-interrupted
0.2
0.0
Disease free survival
Disease free survival
C
PNI 0 1 0-interrupted 1-interrupted
1.0
Disease free survival
Disease free survival
0.8
B
Disease free survival
A
2001
0.6
0.4
PLRs 0 1 2 0-interrupted 1-interrupted 2-interrupted
0.2
0.0 0
20
40
60
80
100
120
0
20
Months
40
60
80
100
120
Months
FIG. 2 Relationship between inflammation-based prognostic scores and disease-free survival in patients with gastric cancer: a NLR, b PNI, c PI, d mGPS, e PLR
The univariate analysis showed that LNR (p = 0.03), TNM (p = 0.01), NLR (p = 0.003), PI (p \ 0.001), mGPS (p \ 0.001), and PLR (p \ 0.001) were associated with OS. A multivariate analysis of these significant variables showed that only the mGPS (p \ 0.001) and PI (p \ 0.001) were independently associated with OS (\T3[ Table 3). Regarding the DFS, the univariate analysis showed that LNR (p = 0.002), number of lymph nodes retrieved (p = 0.05), NLR (p \ 0.001), PNI (p = 0.01), PI (p \ 0.001), mGPS (p \ 0.001), and PLR (p \ 0.001) were significant. A multivariate analysis of these significant variables showed that the mGPS (p = 0.002) and PI (p \ 0.001) were independently associated with DFS (Table 3). At a subgroup analysis, important clinicopathologic prognostic characteristics were significantly different between patients with mGPS 0, 1, or 2 (LNR, p \ 0.001; tumor stage, p \ 0.001; grading, p = 0.03). More in detail, patients with mGPS 0 had the following clinicopathologic characteristics: mean LNR, 0.12 ± 0.1; 16 of 49 (32.6 %) patients had more advanced cancer (stage III–IV); and 26 of 49 (53.1 %) patients had a higher tumor grading (grade III–IV). Patients allocated in mGPS 1
had a mean LNR of 0.16 ± 0.2, 13 of 25 (52 %) patients had more advanced cancer, and 15 of 25 (60 %) patients had a higher tumor grade. Concerning patients with an mGPS score of 2, the mean LNR was 0.49 ± 0.2, 26 of 28 (92.8 %) had a more advanced cancer, and 23 of 28 (82.1 %) patients had a higher tumor grade. DISCUSSION Mantovani et al.9 demonstrated the role of the host inflammatory response in the development and progression of cancer. The major role of the acute-phase response is to remove the insult from the body and to reinstate the initial status; however, if this response persists, such as in case of cancer, it contributes to the development of the disease. According to the concept that ongoing inflammation mediates a progressive loss of vital protein, leading to the death of patients with advanced cancer, Forrest and colleagues first developed a scoring system based on the combination of CRP and albumin in patients with inoperable non-small-cell lung cancer.13,18 On further investigation, the GPS was modified and termed mGPS, on the observation that
2002
P. Aurello et al.
TABLE 3 Prognostic factors for overall survival (OS) and disease-free survival (DFS) in patients with gastric cancer Variables
Univariate analysis
Multivariate analysis
Univariate analysis
Multivariate analysis
OS
OS
DFS
DFS
Hazard ratio (95 % CI)
p value
Hazard ratio (95 % CI)
p value
Hazard ratio (95 % CI)
p value
Age
1.04 (1.00–1.07)
0.10
1.00 (0.97–1.04)
0.62
Sex
0.80 (0.42–1.53)
0.51
0.95 (0.48–1.89)
0.90
Hazard ratio (95 % CI)
p value
Tumor stage I
1
0.01
1
0.36
1
0.20 (0.08– 0.52)
0.01
0.73 (0.16– 3.37)
0.69
III
0.27 (0.07–0.97)
0.04
0.46 (0.08–2.51)
0.37
0.16 (0.03–0.70)
IV
0.60 (0.28–1.26)
0.17
0.39 (0.14–1.10)
0.07
0.70 (0.33–1.50)
0.37
0.06 (0.07–0.59)
0.11
0.32 (0.06–1.33)
0.21
Tumor location
0.1 (0.03– 0.31)
0.98
II
0.1 0.21
Surgical procedure
1.62 (0.86–3.05)
0.12
1.75 (0.89–3.43)
0.10
Lymph nodes retrieved
1.03 (0.97–1.02)
0.84
1.02 (1.0–1.05)
0.05
1.02 (0.99–1.03)
0.12
0.80 (0.14–4.40)
0.80
Grading G1
1
0.73
1
0.98
G2
2.85 (0.33–24.72)
0.34
1.08 (0.09–12.02)
0.95
G3 G4
2.60 (0.32–20.5) 3.52 (0.4–26.13)
0.36 0.21
1.61 (0.1–13.48) 3.51 (0.48–24.21)
0.65 0.21
LNR
3.14 (1.07–9.18)
0.03
5.84 (1.92–17.63)
0.002
Albumin (g/dl)
0.65 (0.39–1.07)
0.09
0.71 (0.41–1.23)
0.22
CRP (mg/dl)
1.22 (0.93–1.59)
0.14
1.08 (0.77–1.51)
0.63
WBC (103/ll)
1.11 (0.99–1.25)
0.53
1.13 (0.99–1.28)
0.57
PLT count (103/ll)
0.99 (0.99–1.00)
0.63
1 (0.99–1.03)
0.66
CEA (ng/ml)
0.99 (0.99–1.00)
0.41
0.99 (0.99–1.04)
0.44
1 (0.99–1.00)
0.99
1(0.99–1.01)
0.89
CA 19.9 (Ul/ml)
0.59 (0.10–3.51)
0.56
mGPS 0
1
\0.001
1
\0.001
1
\0.001
1
0.002
1
0.13 (0.06–0.31)
\0.001
1.70 (1.20–3.42)
0.005
0.03 (0.01–0.11)
\0.001
1.78 (0.82–2.85)
0.002
2
0.67 (0.33–1.37)
0.02
1.91 (1.38–3.18)
0.008
0.33 (0.15–0.73)
0.006
1.73 (1.16–2.39)
0.001
PI 0 1 2 PNI (0/1)
1
\0.001
0.04 (0.01–0.12) 0.48 (0.24–0.94)
\0.001 0.03
0.52 (0.26–1.04)
0.06
1 0.04 (0.01–0.20) 0.37 (0.16–0.82)
\0.001 \0.001 0.01
1
\0.001
0.37 (0.01–0.11) 0.38 (0.18–0.79)
\0.001 0.009
0.33 (0.14–0.76)
0.01
1
\0.001
0.06 (0.01–0.28) \0.001 0.21 (0.08–0.55) 0.001 0.11 (0.03–0.40)
0.1
PLR 0
1
\0.001
1
0.26
1
\0.001
1
0.19
1
0.09 (0.03–0.25)
\0.001
0.43 (0.10–1.73)
0.23
0.06 (0.02–0.27)
\0.01
0.38 (0.08–1.63)
2
0.66 (0.30–1.43)
0.3
1.13 (0.45–2.79)
0.79
0.1 (0.22–1.12)
0.01
2.05 (0.71–5.89)
0.18
NLR (0/1)
0.39 (0.21–0.72)
0.003
1.51 (0.69–3.28)
0.29
0.28 (0.14–0.54)
\0.001
0.95 (0.35–2.56)
0.91
0.19
WBC white blood cell count, PLT platelet, CEA carcinoembryonic antigen, CA carbohydrate antigen
hypoalbuminemia without an increased CRP concentration was rare and that hypoalbuminemia on its own was not associated with poor survival.19–23 Recently, Proctor et al.24 demonstrated that the mGPS is one of the most useful scoring systems in determining long-term outcomes in patients with different malignancies.
The systemic inflammatory response also contributes to the changes in the relative levels of circulating white blood cells as indicated by the increase of neutrophils and the decrease of lymphocytes in the progression of cancer. Previous studies have shown that the relationship between neutrophilia and lymphopenia, also called the NLR, is a
Value of Preoperative Inflammation-Based Prognostic Scores
prognostic independent factor in gastric cancer.25,26 Recently, Mano et al.27 have demonstrated that NLR is an independent predictor of survival after hepatectomy in patients with hepatocellular cancer. Jung et al.25 have shown the same findings in patients with late-stage gastric cancer undergoing resection. In their article, the statistical significance of the NLR ratio was evident only in patients with stage III gastric cancer on both OS and DFS . Ishizuka et al.28 and Pinato et al.29 have shown that a combination of albumin and lymphocyte count, called PNI, is associated with the survival of patients with hepatocellular cancer. Smith et al.15 have demonstrated that the PLR is a significant prognostic marker in patients with pancreatic cancer, and Kasymjanova et al.16 have shown that the PI, as evidenced by a combination of serum CRP and white cell count, is a significant prognostic factor in patients with lung cancer. Our study shows that the scores based on CRP, such as mGPS and PI, are more accurate in determining good and poor prognosis in patients with gastric cancer compared with NLR and PLR, which are based on components of the circulating white cell count or in combination with albumin (PNI). This could be explained according to the fact that the prototypical acute phase protein is CRP, which is a nonspecific but sensitive marker of systemic inflammatory response. This process is induced in hepatocytes mainly by interleukin (IL)-6, a proinflammatory cytokine, stimulated by IL-1b in an autocrine manner. In gastric cancer, high serum IL-6 levels have shown to be correlated with progression of disease; this could be due to the secretion of IL-6 by gastric cancer cells and/or stromal cells.30 Because advanced cancer is often associated with a systemic inflammatory response, high CRP levels are common in patients with advanced disease. This important issue is furthermore clearly visible in this study thanks to the results obtained with the subgroup analysis. In fact, patients with a higher mGPS score (constructed with albumin and CRP) are patients that have been diagnosed with a more advanced cancer stage, with a higher LNR, and with a higher tumor grade at pathologic examination (LNR, p \ 0.001; tumor stage, p \ 0.001; tumor grade, p = 0.03). Crumley and colleagues demonstrated that preoperative evaluation of systemic inflammatory response, as evidenced by an increased CRP level, is important for the prognosis of patients with gastric cancer.5 It is of interest that the combination of CRP and albumin (mGPS), whose relationship with survival has been validated by previous studies, had similar prognostic value to that of CRP and white cell count (PI). These results would suggest that the white blood cell count, rather than NLR or PLR, may be a useful addition to mGPS. Most gastric cancer patients, even after curative resection, experienced recurrence with a rate of 60 % and a median length of survival after recurrence of 7.4 months in case of distant metastases and 10.4 months in case of local
2003
recurrence.31 Identification of sensitive markers for patients who have undergone curative gastrectomy and who are at high risk of recurrence would provide useful information for planning follow-up after surgery or intensive adjuvant chemotherapy. In this study the Kaplan–Meier analysis showed that a higher inflammation-based prognostic score was associated with poorer DFS. In particular, the mGPS and PI were independent predictors of outcome. It is well known that the TNM classification is a good indicator for patients with gastric cancer because it includes pathologic findings that are revealed after surgery. In contrast, the inflammation-based prognostic scores are able to classify patients with gastric cancer before surgery, although their mechanism has not been clearly elucidated. Probably, the host local inflammatory response to the tumor might be associated with systemic inflammatory response, and, in particular, inflammatory cells may interact directly and indirectly with the primary tumor to promote angiogenesis, extracellular matrix remodeling, and preparation of the metastatic niche.32 Otherwise, in accordance with previous studies, the inflammation-based prognostic scores NLR, PNI, and PLR were statistically significant at the univariate analysis, thus demonstrating their prognostic value in predicting cancerspecific outcome in patients with gastric cancer.15,16,25–27 Further studies are needed to clarify the role of systemic inflammatory response in patients with gastric cancer and to clearly identify the possibility that inflammation could become a therapeutic target as the tumor itself. This study has some limitations, mostly consisting of its relatively small sample size, retrospective design, and single-center study. Therefore, a large-scale prospective validation study is needed to confirm these results. In summary, this study’s results show that mGPS and PI are independent markers of poor prognosis in patients with gastric cancer and are superior to the other inflammationbased prognostic scores in terms of prognostic evaluation ability. A measurement of these inflammation-based prognostic scores should be included in the routine preoperative assessment of all patients with gastric cancer. Moreover, the preoperatively accurate stratification of patients is important for planning the best treatment and follow-up for patients with gastric cancer. DISCLOSURE All of the authors declare that they have no potential commercial conflicts of interest relevant to this article.
REFERENCES 1. Jemal A, Siegel R, Xu J, Ward E. Cancer statistics, 2010. CA Cancer J Clin. 2010;60:277–300. 2. Kelley JR, Duggan JM. Gastric cancer epidemiology and risk factors. J Clin Epidemiol. 2003;56:1–9.
2004 3. McMillan DC, Watson WS, O’Gorman P, Preston T, Scott HR, McArdle CS. Albumin concentrations are primarily determined by the body cell mass and the systemic inflammatory response in cancer patients with weight loss. Nutr Cancer. 2001;39:210–3. 4. Onate-Ocana LF, Aiello-Crocifoglio V, Gallardo-Rincon D, et al. Serum albumin as a significant prognostic factor for patients with gastric carcinoma. Ann Surg Oncol. 2007;14:381–9. 5. Crumley AB, Stuart RC, McKernan M, McMillan DC. Is hypoalbuminemia an independent prognostic factor in patients with gastric cancer? World J Surg. 2010;34:2393–8. 6. Roxburgh CS, McMillan DC. Role of systemic inflammatory response in predicting survival in patients with primary operable cancer. Future Oncol. 2010;6:149–63. 7. Lee DY, Hong SW, Chang YG, Lee WY, Lee B. Clinical significance of preoperative inflammatory parameters in gastric cancer patients. J Gastric Cancer. 2013;13:111–6. 8. Coussens LM, Werb Z. Inflammation and cancer. Nature. 2002;420:860–7. 9. Mantovani A, Allavena P, Sica A, Balkwill F. Cancer-related inflammation. Nature. 2008;454:436–44. 10. Colotta F, Allavena P, Sica A, Garlanda C, Mantovani A. Cancerrelated inflammation, the seventh hallmark of cancer: links to genetic instability. Carcinogenesis. 2009;30:1073–81. 11. Chiba T, Marusawa H, Ushijima T. Inflammation-associated cancer development in digestive organs: mechanisms and roles for genetic and epigenetic modulation. Gastroenterology. 2012; 143:550–63. 12. Yamashita H, Katai H. Systemic inflammatory response in gastric cancer. World J Surg. 2010;34:2399–400. 13. Forrest LM, McMillan DC, McArdle CS, Angerson WJ, Dagg K, Scott HR. A prospective longitudinal study of performance status, an inflammation-based score (GPS) and survival in patients with inoperable non-small-cell lung cancer. Br J Cancer. 2005;92:1834–6. 14. Walsh SR, Cook EJ, Goulder F, Justin TA, Keeling NJ. Neutrophil-lymphocyte ratio as a prognostic factor in colorectal cancer. J Surg Oncol. 2005;91:181–4. 15. Smith RA, Bosonnet L, Raraty M, et al. Preoperative plateletlymphocyte ratio is an independent significant prognostic marker in resected pancreatic ductal adenocarcinoma. Am J Surg. 2009;197:466–72. 16. Kasymjanova G, MacDonald N, Agulnik JS, et al. The predictive value of pre-treatment inflammatory markers in advanced nonsmall-cell lung cancer. Curr Oncol. 2010;17:52–8. 17. Nozoe T, Ninomiya M, Maeda T, Matsukuma A, Nakashima H, Ezaki T. Prognostic nutritional index: a tool to predict the biological aggressiveness of gastric carcinoma. Surg Today. 2010;40:440–3. 18. Forrest LM, McMillan DC, McArdle CS, Angerson WJ, Dunlop DJ. Evaluation of cumulative prognostic scores based on the systemic inflammatory response in patients with inoperable nonsmall-cell lung cancer. Br J Cancer. 2003;89:1028–30.
P. Aurello et al. 19. Crumley AB, McMillan DC, McKernan M, McDonald AC, Stuart RC. Evaluation of an inflammation-based prognostic score in patients with inoperable gastro-oesophageal cancer. Br J Cancer. 2006;94:637–41. 20. Elahi MM, McMillan DC, McArdle CS, Angerson WJ, Sattar N. Score based on hypoalbuminemia and elevated C-reactive protein predicts survival in patients with advanced gastrointestinal cancer. Nutr Cancer. 2004;48:171–3. 21. Mohri Y, Tanaka K, Ohi M, et al. Inflammation-based prognostic score as a predictor of postoperative gastric cancer recurrence. Anticancer Res. 2012;32:4581–4. 22. McMillan DC. The systemic inflammation-based Glasgow Prognostic Score: a decade of experience in patients with cancer. Cancer Treat Rev. 2013;39:534–40. 23. Kinoshita A, Onoda H, Imai N, et al. Comparison of the prognostic value of inflammation-based prognostic scores in patients with hepatocellular carcinoma. Br J Cancer. 2012;107:988–93. 24. Proctor MJ, Morrison DS, Talwar D, et al. A comparison of inflammation-based prognostic scores in patients with cancer. A Glasgow Inflammation Outcome Study. Eur J Cancer. 2011;47: 2633–41. 25. Jung MR, Park YK, Jeong O, et al. Elevated preoperative neutrophil to lymphocyte ratio predicts poor survival following resection in late stage gastric cancer. J Surg Oncol. 2011;104: 504–10. 26. Aizawa M, Gotohda N, Takahashi S, Konishi M, Kinoshita T. Predictive value of baseline neutrophil/lymphocyte ratio for T4 disease in wall-penetrating gastric cancer. World J Surg. 2011;35: 2717–22. 27. Mano Y, Shirabe K, Yamashita Y, et al. Preoperative neutrophilto-lymphocyte ratio is a predictor of survival after hepatectomy for hepatocellular carcinoma: a retrospective analysis. Ann Surg. 2013;258:301–5. 28. Ishizuka M, Kubota K, Kita J, Shimoda M, Kato M, Sawada T. Impact of an inflammation-based prognostic system on patients undergoing surgery for hepatocellular carcinoma: a retrospective study of 398 Japanese patients. Am J Surg. 2012;203:101–6. 29. Pinato DJ, North BV, Sharma R. A novel, externally validated inflammation-based prognostic algorithm in hepatocellular carcinoma: the prognostic nutritional index (PNI). Br J Cancer. 2012;106:1439–45. 30. Kim DK, Oh SY, Kwon HC, et al. Clinical significances of preoperative serum interleukin-6 and C-reactive protein level in operable gastric cancer. BMC Cancer. 2009;9:155. 31. Rohatgi PR, Yao JC, Hess K, et al. Outcome of gastric cancer patients after successful gastrectomy: influence of the type of recurrence and histology on survival. Cancer. 2006;107:2576–80. 32. DeNardo DG, Johansson M, Coussens LM. Immune cells as mediators of solid tumor metastasis. Cancer Metastasis Rev. 2008;27:11–8.
