**7. Application of SIR in complex scoring systems for gastric cancer**

Elevated concentration of fibrinogen in the serum of gastric carcinoma patients has negative prognostic value regarding several aspects of survival—overall and cancer-free survival. The independent prognostic value of increased fibrinogen level has been demonstrated in 351 surgically treated gastric cancer patients. The hazard ratio was 2.61 (95% CI = 1.18–5.76) as reported by Suzuki et al. [111]. The independent prognostic role was confirmed in another large surgically treated cohort of 1196 gastric cancer patients [112]. Applying ROC-identified cut-off (3.9 g/L), high fibrinogen level was significantly associated with overall survival in multivariate analysis [113]. In patients who underwent curative gastrectomy, hyperfibrinogenaemia (reaching or exceeding 350 mg/dL) was associated not only with overall but also cancer-free survival. By multivariate analysis, fibrinogen level again was an independent

Classic studies have explored the diagnostic meaning of hyperfibrinogenaemia resulting in conclusion that fibrinogen level is significantly elevated in gastric cancer patients but not in individuals having gastric or duodenal peptic ulcer. Such reports stem back as far as to 1975 [114]. Later, it was repeatedly confirmed that fibrinogen levels in gastric cancer are higher than in controls, even if the tumour was non-metastatic. The mean levels in cancer patients vs. control individuals were 505 vs. 336 mg/dL [115]. Nowadays, the ongoing research has identified fibrinogen fragments that could potentially serve as serum markers of gastric cancer. Fibrinogen fragments, e.g., carboxyl terminal fraction of fibrinogen alpha, have been tested as a serum marker of gastric cancer in comparison with healthy controls and individuals affected by chronic gastritis [116, 117]. A 15-amino acid peptide of the fibrinogen alpha chain, fibrinostatin, has anti-angiogenic

Regarding the local events within the tumour, fibrinogen has been identified in tumour stroma as early as 1984 [119, 120] while fibrin and D-dimers are found in the invasive front [120].

Fibrinogen level parallels the tumour burden, correlates with advanced TNM stage [112] and is associated with adjacent organ involvement [121]. In a recent considerable cohort of 1090 gastric cancer patients treated by gastrectomy, high fibrinogen level (exceeding the ROCidentified cut-off at 3.9 g/L) was significantly associated with tumour size, T, N and TNM stage [113]. Fibrinogen shows statistically significant associations with the invasion depth of gastric cancer confirmed by several other studies focusing on T [122–124]. Several studies have identified meaningful association with presence of metastasis in lymph nodes [122–124]. The association with tumour spread has also been confirmed, regarding the presence of dis-

The logical next step is incorporation of fibrinogen measurements into combined scores that could be used to assess the prognosis or tumour spread. A complex score comprising evaluation of hyperfibrinogenemia (exceeding 400 mg/dL) and elevated NLR (exceeding 3.0) was associated with shorter survival. The combined score showed significantly different results in patients developing progressive disease despite chemotherapy or chemoradiotherapy [103]. Similar score comprising evaluation of hyperfibrinogenemia (reaching or exceeding 305 mg/dL) and elevated NLR (reaching or exceeding 2.34) was significantly associated with invasion depth, lymph node metastasis, lymphovascular invasion and stage [110]. Coagulation score based on

properties; thus therapeutic applications have been hypothesised [118].

prognostic factor along with pTN [33].

178 Gastric Cancer

tant metastases [122].

SIR parameters have been incorporated in diverse complex scores that allow reaching higher diagnostic value (see also **Tables 6**–**7**).

A complex score, based on fibrinogen (cut-off 400 mg/dL) and NLR (cut-off 3.0) levels, was applied to predict the effect of chemotherapy or chemoradiotherapy in advanced gastric cancer. The created score indeed was significantly higher in patients having cancer progression during treatment; it also was an independent prognostic factor by multivariate analysis [103]. The same authors have elaborated similar combined score, based on the same parameters which by different cut-off levels are adjusted for another research target. The fibrinogen-NLR score at cut-off 305 mg/dL and 2.34, respectively, was significantly associated with depth of tumour invasion, lymph node metastasis, lymphatic and venous invasion and tumour stage. The 5-year survival rates by score categories 0 vs. 1 vs. 2 were 92.9, 84.1 and 66.5%; the differences being statistically significant [110].

The coagulation score, recently proposed by Kanda et al., distinguished high-risk patients having low overall and cancer-free survival. High coagulation score was also an independent prognostic factor for recurrence and was associated with liver metastasis as the initial recurrence [125]. It is in accordance with the observation that D-dimer is associated with metastatic tumour spread both in murine gastric carcinoma models and in patients having visceral metastasis [130].

The score developed by Ishizuka et al. was based on platelet count and NLR to predict postoperative survival. The score classified patients into 3 groups: 0 vs. 1 vs. 2 had post-operative survival of 1676 vs. 1310 vs. 1050 days. The differences were statistically significant. The cancer-specific survival also was significantly different by the score levels. The sensitivity and accuracy of the presented score in regard to survival was higher than the informativity of clinical and pathological parameters—carcinoembryonic antigen CEA, CA19-9, venous and lymphatic invasion and lymph node metastasis [29].

NLR-PLR score can be used to assess overall survival in gastric cancer patients diagnosed at stage I–II. This score was an independent prognostic factor while mGPS, the prognostic nutritional index and combination of platelet count and NLR were not. The score had the highest area under ROC curve in comparison with the listed other scores. The hazard ratio associated with NLR-PLR score was 1.51 (95% CI = 1.02–2.24). Interestingly, there was a trend to shorter mean OS in stage I patients having NLR-PLR score of 2 than in stage II patients scored 0: 89 months vs. 127 months. The score retained prognostic value in stage I and II [31].


Abbreviations: PLR, platelet to lymphocyte ratio; ROC, receiver operating characteristic curve; NLR, neutrophil to lymphocyte ratio; T, local spread of primary gastric cancer by tumour-nodes-metastasis (TNM) classification; CSS, cancer-specific survival; CFS, cancer-free survival; mGPS, modified Glasgow prognostic score. 1 PNI = albumin (g/L) + 5 × total lymphocyte count (×10<sup>9</sup> /L).

**Table 6.** Application of SIR in complex scoring systems for gastric cancer.

The score based on albumin and NLR was elaborated to improve the evaluation of overall survival. The resulting score was independently associated with overall survival. It had higher diagnostic value than NLR, PLR and GPS, as shown by higher area under ROC curve. The overall survival by score values 0 vs. 1 vs. 2 was 44.9% vs. 29.8% vs. 20.3%, respectively [30].


The score based on albumin and NLR was elaborated to improve the evaluation of overall survival. The resulting score was independently associated with overall survival. It had higher diagnostic value than NLR, PLR and GPS, as shown by higher area under ROC curve. The overall survival by score values 0 vs. 1 vs. 2 was 44.9% vs. 29.8% vs. 20.3%, respectively [30].

**Target Score description References**

and tumour size

Lymph node metastases Independent predictive factors (for

Overall survival Albumin (cut-off 35 g/L), NLR

Prognosis and cancer characteristics NLR (cut-off 2.34) and fibrinogen

1.83)

Overall and cancer-specific survival Platelet count and NLR [29]

Overall survival NLR, mGPS and patient-generated

**Table 6.** Application of SIR in complex scoring systems for gastric cancer.

Overall survival Canton score: 1

PNI = albumin (g/L) + 5 × total lymphocyte count (×10<sup>9</sup>

PLR (cut-off 106, based on ROC analysis), age, grade, depth of invasion

lymph node metastasis) that can be determined preoperatively: NLR (cut-off 1.59), PLR (cut-off 155.67), T/ depth of invasion, macroscopic type

Nomogram including independent

(1) for CSS: NLR, age, tumour stage, presence of lymph node metastases, presence of distant metastases; (2) for CFS: NLR, tumour stage, presence of distant metastases, family history of gastric cancer; CA 19-9 level.

NLR (cut-off 3.0) and fibrinogen

index (PNI; cut-off 48), platelet count (cut-off 3 × 1011/L) and NLR (cut-off

Coagulation score: increased level of

subjective global assessment score

/L).

Abbreviations: PLR, platelet to lymphocyte ratio; ROC, receiver operating characteristic curve; NLR, neutrophil to lymphocyte ratio; T, local spread of primary gastric cancer by tumour-nodes-metastasis (TNM) classification; CSS,

fibrinogen and D-dimers

cancer-specific survival; CFS, cancer-free survival; mGPS, modified Glasgow prognostic score.

prognostic nutritional

(Bormann), tumour size

(cut-off 2.3)

predicting factors:

(cut-off 400 mg/dL)

(cut-off 305 mg/dL)

NLR (cut-off 2.1), PLR (cut-off 120) [31]

[72]

[55]

[30]

[47]

[103]

[110]

[45]

[125]

[26]

Lymph node metastases in early

Survival of early gastric cancer

Cancer-specific survival (CSS) and cancer-free survival (CFS)

Overall survival and chemotherapy

Overall and cancer-free survival

Metachronous liver metastases

gastric cancer

patients

180 Gastric Cancer

response

Recurrence

1

0–11 Low risk


Abbreviations: CRP, C-reactive protein; NLR, neutrophil to lymphocyte ratio; PLR, platelet to lymphocyte ratio; mGPS, modified Glasgow prognostic score; T, local spread of primary gastric cancer by tumour-nodes-metastasis (TNM classification), G, grade; PNI, prognostic nutritional index; PLT, platelet count. 1 PNI = albumin (g/L) + 5 × total lymphocyte count (×10<sup>9</sup> /L).

**Table 7.** The definitions of complex scores.

The inflammation and nutrition-based score was elaborated to predict overall survival in patients diagnosed with metastatic gastric cancer. According to this score, patients were classified into favourable, intermediate and high-risk groups exhibiting the median overall survival of 27.6 vs. 13.2 vs. 8.2 months. The respective two-year survival rates were 52% vs. 16% vs. 3%. The ROC curve analysis confirmed that the novel score has higher informativity than any of its components [26].

Deng et al. elaborated complex nomograms to predict cancer-specific and cancer-free survival in surgically treated gastric cancer patients [47].

Pang et al. developed complex system to predict lymph node metastases based on those tumour and systemic parameters that were independently associated with N+ and could be detected preoperatively. The point system was based on hazard ratios detected by logistic regression analysis. Youden Index was applied to detect the cut-off of the novel combined system. Finally, the developed score had specifity of 72.4%, sensitivity 82.7%, positive predictive value 88.7% and negative predictive value 61.5%. Besides the informative value of the score itself, the mathematical model of score design is flawless [55].

Lou et al. developed score to predict lymph node metastases in early gastric cancer. The scoring system reached reasonable accuracy of 0.817 when evaluating prospective cases [72].

The Canton score was created to predict overall survival after gastrectomy. The novel score possessed higher AUC than the classic parameters. The HR for Canton score values 1 vs. 2 vs. 3 (in comparison to 0) were 1.08 (95% CI = 0.80–1.45) vs. 1.55 (95% CI = 1.15–2.10) vs. 1.64 (95% CI = 1.14–2.36) as reported by Sun et al. [45].
