3.2.1. Performance Status

Performance status is a crucial factor in treatment decision-making for patients with NSCLC. ECOG 0–1 is required in clinical trials, but the patient population is much more diverse in clinical practice. According to the real-world data, the prevalence of patients with ECOG ≥ 2 who are treated with nivolumab ranges from 3% to 46% [6–15]. The prognostic value of performance status has been well documented.

Multivariate survival analysis in the Italian EAP cohort of patients with non-squamous NSCLC showed that ECOG 2 performance status is an independent prognostic factor for early death (*p* < 0.0001) [6]. Poor performance status (ECOG 2), compared with ECOG 0, was also identified as an independent prognostic factor for death in the Italian EAP squamous NSCLC cohort (HR 2.76, 95% CI 1.65–4.62; *p* < 0.0001) [7]. In this cohort, the risk of death was also higher in patients with ECOG 1 than in patients with ECOG 0 (HR 1.57, 95% CI 1.17–2.11; *p* = 0.003) [7]. Similar results were obtained in the analysis of a group of 901 Japanese patients, in which 17.4% of the patients had an ECOG score of 2, 3, or 4 (HR 0.39, 95% CI 0.51–0.8; *p* < 0.0001) [8]. Poor performance status was also a risk factor for short PFS (HR 0.64, 95% CI 0.51–0.8; *p* < 0.0001) [8]. Multivariate analysis of the Portuguese EAP data identified performance status as the only independent prognostic factor (*p* < 0.0006) [13]. In a univariate analysis, the risk of death was much lower in patients with ECOG 0–1 than in patients with ECOG 2 (HR 3.8, 95% CI 2.3–6.07; *p* < 0.0001) [13]. Another univariate analysis reported an OS of 3.4 months (95% CI 2.3–4.4) in patients with ECOG 2 versus 11.79 months (95% CI 8.5–15.07) in patients with ECOG 1. The median OS for patients with ECOG 0 was not reached [14]. Some data related to the prognostic value of performance status are summarized in Table 3.


**Table 3.** Prognostic value of performance status in patients treated with nivolumab—real-world data.

OS—overall survival, ECOG—Eastern Cooperative Oncology Group, HR—hazard ratio. \* Multivariate analysis.

#### 3.2.2. Liver Metastases

A pooled analysis of the CheckMate 017 and CheckMate 057 trials with updated results from more than 3 years of follow-up included a subgroup analysis of patients with liver metastases [16]. Liver metastases were found in 23% of 854 patients at baseline. Although nivolumab had a confirmed OS benefit in patients with liver metastases (HR 0.68, 95% CI 0.50–0.91), the median OS for patients with liver metastases was 6.8 months in the nivolumab group and 5.9 months in the docetaxel group (HR 0.68, 95% CI 0.50–0.91), while for the entire patient population the median OS was 11.1 months for the nivolumab group and 8.1 months for the docetaxel group (HR 0.70, 95% CI 0.61–0.81) [16]. Patients with and without liver metastases were not directly compared.

Liver metastases were determined to be an independent negative prognostic factor for OS in multivariate analyses of some real-world data [6–8]. For Italian patients with squamous NSCLC, the HR was 1.44 (95% CI 1.04–1.98; *p* = 0.03) [7], and for non-squamous NSCLC the odds ratio (OR) for early death was 0.47 (95% CI 0.35–0.61; *p* < 0.0001) [6]. However, in another publication, the negative prognostic value of liver metastases was not confirmed [9]. In a retrospective analysis of 215 patients with NSCLC who received nivolumab, atezolizumab, or pembrolizumab (19.1% of patients had liver metastases), there was a higher risk of death in patients with liver metastasis than in those without (HR 2.04, 95% CI 1.33–3.13) [17]. Additional negative prognostic factors for patients with NSCLC and liver metastases were low albumin level, poor performance status, driver mutation, and having five or more liver metastases.
