Statistical Results Interpretation Summary Table

Category

Statistical Method / Measure

Results Statement

Clinical Interpretation of the Effect Estimate

Descriptive statistics

Mean ± SD

The mean age of patients with NSCLC was 63.4 ± 9.8 years.

The average patient age was 63.4 years, with most individuals falling within approximately 10 years around the mean.

Descriptive statistics

Median (IQR)

The median follow-up time was 26 months (IQR: 16–41).

A typical patient was followed for just over 26 months, with most patients having follow-up durations between approximately 1.5 and 3.5 years.

Descriptive statistics

n (%)

Male patients accounted for 61.2% (n = 184) of the cohort.

Approximately 61% patients in the study population were male.

Baseline comparison

t test

There was no significant difference in age between the pembrolizumab and chemotherapy groups (63.1 ± 9.6 vs 63.8 ± 10.1 years, P = 0.47).

The mean age differed by approximately 0.7 years between groups, indicating similar age distributions/ comparable age.

Baseline comparison

χ² test

PD-L1 expression levels were comparable between the two groups (χ² = 1.82, P = 0.18).

The distribution of PD-L1 expression categories was similar between treatment groups.

Continuous outcome comparison

Mean difference

Pembrolizumab significantly prolonged/increased PFS compared with chemotherapy (8.4 vs 5.6 months; mean difference = 2.8 months, P = 0.01).

Compared with chemotherapy, pembrolizumab was associated with an average prolongation of PFS by 2.8 months.

Categorical outcome comparison

Proportion

The objective response rate was higher in the pembrolizumab group (38.5% vs 21.4%, P = 0.005).

Compared with chemotherapy, pembrolizumab was associated with a higher likelihood of achieving an objective response, with an absolute difference of 17.1%.

Correlation analysis

Pearson correlation

PD-L1 expression was moderately correlated with progression-free survival (r = 0.42, P < 0.001).

A correlation coefficient of 0.42 indicates a moderate positive association, with higher PD-L1 expression generally corresponding to longer PFS.

Linear regression

Linear regression (β)

Higher PD-L1 expression was associated with longer PFS (β = 0.18 months per 10% increase, 95% CI: 0.09–0.27, P < 0.001).

A β of 0.18 indicates that for every 10% increase in PD-L1 expression, mean PFS increases by approximately 0.18 months.

Logistic regression

Odds ratio (OR)

Pembrolizumab treatment was associated with a higher likelihood of objective response, (OR = 2.32, 95% CI: 1.45–3.71, P < 0.001).

An OR of 2.32 indicates that the odds of achieving an objective response were approximately 2.3 times higher with pembrolizumab than with chemotherapy.

Survival analysis

Kaplan–Meier

Kaplan–Meier curves demonstrated improved PFS in the pembrolizumab group (median PFS: 8.4 vs 5.6 months).

The median time to disease progression was delayed by approximately 2.8 months in the pembrolizumab group.

Survival comparison

Log-rank test

PFS differed significantly between the two groups (log-rank P = 0.01).

The timing of disease progression over follow-up differed between treatment groups.

Cox regression

Hazard ratio (HR)

Pembrolizumab significantly reduced the risk of disease progression (HR = 0.65, 95% CI: 0.48–0.88, P = 0.004).

An HR of 0.65 indicates that, at any given time point, the risk of disease progression was approximately 35% lower with pembrolizumab.

Multivariable analysis

Adjusted Cox model

Pembrolizumab remained an independent protective factor after adjustment (adjusted HR = 0.68, 95% CI: 0.51–0.90, P = 0.01).

After adjustment for covariates, the risk of disease progression remained approximately 32% lower in the pembrolizumab group.

Subgroup analysis

Consistent PFS benefit was observed across PD-L1 subgroups (P for interaction = 0.41).

Similar magnitudes of risk reduction were observed across PD-L1 subgroups.

Interaction analysis

Interaction test

No significant interaction was observed between treatment and PD-L1 expression (P for interaction = 0.41).

The effect size of pembrolizumab did not meaningfully differ across PD-L1 levels.

Sensitivity analysis

Results remained robust after excluding patients with follow-up < 3 months (HR range: 0.62–0.70).

Across sensitivity analyses, the estimated reduction in progression risk ranged from approximately 30% to 38%.

Non-significant result

OS was longer in the pembrolizumab group but not statistically significant (HR = 0.82, 95% CI: 0.64–1.05, P = 0.09).

Pembrolizumab was associated with a lower estimated risk of death (HR = 0.82), but the confidence interval included unity and the difference did not reach statistical significance.

Trend

A trend toward improved overall survival was observed in the pembrolizumab group compared with the control group (median OS: 20.1 vs 17.3 months).

The median overall survival was prolonged by approximately 2.8 months.

Effect summary

Effect estimate

Pembrolizumab demonstrated a clinically meaningful treatment effect (HR < 1 with a narrow 95% CI).

An HR below 1 with a relatively narrow confidence interval suggests a stable estimate of reduced progression risk.

Adjusted summary

Multivariable model

Pembrolizumab was independently associated with improved outcomes after adjustment (P < 0.05).

The estimated treatment effect remained statistically different from the null after covariate adjustment.

Risk comparison

Relative Risk (RR)

Pembrolizumab reduced the risk of disease progression compared with chemotherapy (RR = 0.72, 95% CI: 0.60–0.86).

An RR of 0.72 indicates that the probability of disease progression was approximately 72% of that observed with chemotherapy.

Risk difference

Absolute risk difference

The absolute risk of progression was 18% lower in the pembrolizumab group.

Out of every 100 patients treated, approximately 18 fewer experienced disease progression.

Clinical relevance

Number Needed to Treat (NNT)

The NNT= 6 to prevent one progression event was 6.

On average, treating 6 patients prevents one additional progression event.

Survival analysis

RMST difference

RMST over 24 months was longer in the pembrolizumab group (difference = 1.9 months, P = 0.02).

Within the first 24 months of follow-up, patients treated with pembrolizumab experienced, on average, approximately 1.9 additional months without progression compared with those receiving chemotherapy.

Event frequency

cumulative Incidence

Disease progression occurred in 42.3% of the pembrolizumab group versus 58.1% of controls.

The cumulative proportion of patients experiencing progression was reduced by approximately 15.8 percentage points.

Event rate

Incidence rate

The incidence rate of disease progression was 0.32 events per person-year in the pembrolizumab group compared with 0.48 events per person-year in the control group.

Regression model

R²

The linear regression model explained 32% of the variance in PFS (R² = 0.32).

Approximately 32% of the variability in PFS was accounted for by the variables included in the model.

Model discrimination

C-statistic / AUC

The model demonstrated good discrimination (C-statistic = 0.74).

A C-statistic of 0.74 indicates a 74% probability of correctly distinguishing between patients with and without progression.

Uncertainty

Standard Error (SE)

The estimated mean difference had an SE of 0.6 months.

The sampling variability of the estimated mean difference is approximately ±0.6 months.

Sensitivity analysis

E-value

The E-value for the observed HR was 2.1.

An unmeasured confounder would need to be associated with both treatment and outcome by a risk ratio of approximately 2.1 to fully explain the observed effect.

Multiple comparison

Adjusted P value

The association remained significant after multiplicity adjustment (adjusted P = 0.03).

The effect estimate remained statistically distinguishable from the null after correction for multiple testing.

Model assumption

PH assumption test

The proportional hazards assumption was not violated (P = 0.62).

The hazard ratio can be reasonably interpreted as approximately constant over time.

Results Section · Commonly Used Phrase

I. Study Population and Baseline Characteristics

Scenario	Common Results Statements
Sample size	A total of 302 patients with non–small cell lung cancer were included, of whom 156 received pembrolizumab.
Baseline comparability	No statistically significant differences were observed between the two groups with respect to age, sex, disease stage, or smoking status.
Baseline imbalance	Compared with the control group, a higher proportion of patients with high PD-L1 expression was observed in the pembrolizumab group (P = 0.03).
Missing data	A total of 12 patients (4.0%) had missing values for at least one clinical variable.

II. Efficacy Outcomes

1️⃣ Survival Outcomes (PFS / OS)

Scenario	Common Results Statements
PFS description	The median progression-free survival in the pembrolizumab group was 8.4 months.
OS description	The median overall survival in the overall cohort was 18.6 months.
Between-group comparison	Progression-free survival was significantly longer in the pembrolizumab group than in the control group (P = 0.01).
Kaplan–Meier analysis	Kaplan–Meier survival curves demonstrated a significant difference in progression-free survival over follow-up between the two groups.
Risk reduction	Pembrolizumab treatment was associated with a significantly reduced risk of disease progression.

2️⃣ Tumor Response (ORR / DCR)

Scenario	Common Results Statements
ORR	The objective response rate in the pembrolizumab group was 38.5%.
DCR	The disease control rate in the pembrolizumab group reached 64.2%.
Between-group difference	Compared with the chemotherapy group, the objective response rate was significantly higher in the pembrolizumab group.
Subgroup response	Patients with high PD-L1 expression demonstrated a significantly higher objective response rate than those with low PD-L1 expression.

III. Safety Outcomes

Scenario	Common Results Statements
Overall adverse events	At least one adverse event occurred in 72.1% of patients receiving pembrolizumab.
Severe adverse events	Grade ≥3 adverse events were observed in 18.4% of patients.
Between-group comparison	No statistically significant difference in the incidence of severe adverse events was observed between the two groups.
Immune-related events	The most common immune-related adverse events included rash and thyroid dysfunction.

IV. Correlation Analysis

Scenario	Common Results Statements
Positive correlation	PD-L1 expression level was significantly positively correlated with progression-free survival.
Negative correlation	Baseline tumor burden was negatively correlated with overall survival.
Strength of correlation	The correlation was of moderate strength (r = 0.41).
Adjustment	The association remained after adjustment for age and disease stage.

V. Regression Analysis

1️⃣ Univariate Analysis

Scenario	Common Results Statements
Univariate Cox regression	In univariate Cox regression analysis, pembrolizumab treatment was significantly associated with a reduced risk of disease progression.
Univariate logistic regression	In univariate analysis, high PD-L1 expression was significantly associated with objective response.

2️⃣ Multivariable Analysis (Key Results)

Scenario	Common Results Statements
Multivariable Cox regression	Multivariable Cox regression analysis demonstrated that pembrolizumab treatment was an independent protective factor against disease progression.
Adjusted results	The association remained statistically significant after adjustment for age, sex, disease stage, and PD-L1 expression.
Control of confounding	The results remained stable after controlling for potential confounding factors.
Model stability	No significant multicollinearity was detected among the variables included in the model.

VI. Subgroup Analysis

Scenario	Common Results Statements
Consistency across subgroups	Consistent treatment benefits of pembrolizumab were observed across all predefined subgroups.
Interaction	No statistically significant interaction between treatment effect and subgroup variables was observed.
Specific population	The treatment benefit appeared to be more pronounced among patients with PD-L1 expression ≥50%.

VII. Sensitivity Analysis

Scenario	Common Results Statements
Purpose	Multiple sensitivity analyses were performed to assess the robustness of the findings.
Exclusion analysis	After excluding patients with follow-up shorter than 3 months, the main results did not materially change.
Alternative models	The primary conclusions remained consistent when alternative statistical models were applied.

VIII. Non-significant and Trend Results (Commonly Used)

Scenario	Common Results Statements
Non-significant result	No statistically significant difference in overall survival was observed between the two groups.
Trend	A trend toward longer overall survival was observed in the pembrolizumab group.
Explanation	This finding may be partly attributable to the limited sample size.

IX. Summary Statements for the End of the Results Section

Scenario	Common Results Statements
Overall summary	In summary, pembrolizumab demonstrated favorable efficacy and an acceptable safety profile in patients with non–small cell lung cancer.
Emphasis	These findings support pembrolizumab as an effective treatment option for this patient population.
Transition	These results provide a basis for further investigation of the long-term benefits of pembrolizumab.