Trial Outcomes & Findings for Radiation Therapy With Durvalumab or Cetuximab in Treating Patients With Locoregionally Advanced Head and Neck Cancer Who Cannot Take Cisplatin (NCT NCT03258554)

Participants were first registered for screening. Central pathology review confirmation of p16 determination was an enrollment requirement for participants with oropharyngeal or unknown primary tumors. Overall,196 of 200 screened participants were enrolled.

Radiation Therapy With Durvalumab or Cetuximab in Treating Patients With Locoregionally Advanced Head and Neck Cancer Who Cannot Take Cisplatin

DLTs were collected to verify the safety of durvalumab with RT in this population. Safety was determined if ≤ 2 of 8 participants in the cohort had any DLT, in which case the study would proceed to phase II with that dose schedule (DS). The probability for the DS to be deemed too toxic, given a true toxicity rate ≥ 45%, is at least 78%. With a true toxicity rate ≤ 20%, the probability for the DS do be deemed safe is 80%. The full DLT definition does not fit here, but includes all grade 5 AEs, grade 3 or 4 AEs definitely or probably related to durvalumab (DPRD) except for specified AEs and situations, and incomplete or \> 2-week delay completing RT due to immune toxicity DPRD. AEs were graded according to National Cancer Institute (NCI) Common Terminology Criteria for Adverse Events (CTCAE) version 4.0, which grade severity from 1=mild to 5=death. Two alternate DSs with a delayed 2nd dose (to reduce/avoid doses concurrent with RT) would be tried if the initial DS was too toxic.

Progression (failure) is defined as local, regional, or distant disease progression, or death from any cause. Progression was assessed by imaging, clinical assessment, or biopsy. Failure time is defined as time from randomization to failure or last follow-up (censored). Failure rates are estimated using the Kaplan-Meier method. The protocol specifies that the distributions of failure times be compared between the arms, which is reported in the statistical analysis results. Two-year estimates are provided. Analysis was planned to occur after 69 failure events had been reported.

Overall survival (OS) time is defined as time from randomization to the date of death from any cause or last known follow-up (censored). Overall survival rates are estimated by the Kaplan-Meier method. The protocol specifies that the distributions of failure times be compared between the arms, which is reported in the statistical analysis results. Two-year estimates are provided.

Locoregional progression is defined as local or regional progression or recurrence, death due to study cancer or unknown causes without documented progression. Progression was assessed by imaging, clinical assessment, or biopsy. Failure time is defined as time from randomization to failure, distant metastasis (competing risk), deaths from other causes (competing risk), or last follow-up (censored). Failure rates are estimated using the cumulative incidence method. The protocol specifies that the distributions of failure times be compared between the arms, which is reported in the statistical analysis results. Two-year estimates are provided.

Failure is defined as the occurrence of distant metastasis. Failure time is defined as time from randomization to first occurrence of distant metastasis, local or regional progression or recurrence (competing risk), death (competing risk), or last follow-up (censored). Failure rates are estimated using the cumulative incidence method. The protocol specifies that the distributions of failure times be compared between the arms, which is reported in the statistical analysis results. Two-year estimates are provided.

Failure is defined as Death from second primary, protocol treatment, or "other cause". Failure time is defined as time from randomization to first occurrence of failure, death due to study cancer or unknown cause (competing risk), or last follow-up (censored). Failure rates are estimated using the cumulative incidence method. The protocol specifies that the distributions of failure times be compared between the arms, which is reported in the statistical analysis results. Two-year estimates are provided.

Fludeoxyglucose F-18 positron emission tomography/computed tomography (FDG-PET/CT) scans at baseline and four months after the end of radiation therapy are compared. Per RECIST 1.1: * Complete response: * Disappearance of all lesions and pathologic lymph nodes * Partial response: * ≥ 30% decrease sum of the longest diameters * No new lesions * No progression of non-target lesions

Common Terminology Criteria for Adverse Events (version 5.0) grades adverse event severity from 1=mild to 5=death. Summary data is provided in this outcome measure; see Adverse Events Module for specific adverse event data.

Assessed using European Organization for Research and Treatment of Cancer Core Questionnaire (EORTC QLQ)/Head-and-Neck module (H\&N35), EuroQol- 5 Dimension (EQ5D), M. D. Anderson Dysphagia Inventory (MDADI), patient reported outcomes (PRO)-CTCAE, geriatric screening (G8). The mean summary score of the EORTC QLQ/H\&N35, EQ5D, MDADI, PRO-CTCAE, G8, CCI, and the subscales will be determined. The mean change from baseline at each time point will be summarized using mean and standard deviations for each arm. Mean change from baseline will be compared between the arms using a two sample t test. If data normality assumptions are not met, the Wilcoxon rank sum test will be used to test the hypothesis. Mean change from baseline will be tested using an omnibus F test followed by individual comparisons of change scores at different time points within each treatment group. The same analysis will be conducted for between group comparisons at each time point.

The mean change from baseline at each time point will be summarized using mean and standard deviations for each arm. Mean change from baseline will be compared between the arms using a two sample t test. If data normality assumptions are not met, the Wilcoxon rank sum test will be used to test the hypothesis.

Failure for progression-free survival is defined as local, regional, or distant disease progression, or death from any cause. Progression was assessed by imaging, clinical assessment, or biopsy. Failure time is defined as time from randomization to failure or last follow-up (censored). PD-L1 expression is defined by a combined positive score (CPS) ≥ 1, assessed by masked central analysis of baseline tissue specimens. CPS = \[(number of tumor cells positive for PD-L1) / (number of tumor cells positive for PD-L1 + number of tumor cells negative for PD-L1)\] multiplied by 100, yielding a possible score of 0 to 100. Treatment effect hazard ratios within PD-L1 subgroup were estimated by Cox proportional hazards model and progression-free survival estimates were estimated by Kaplan-Meier method. Analysis was planned to occur after 69 failure events had been reported and tissue specimen analysis was complete.

Failure for progression-free survival is defined as local, regional, or distant disease progression, or death from any cause. Progression was assessed by imaging, clinical assessment, or biopsy. Failure time is defined as time from randomization to failure or last follow-up (censored). Positive p16 status was defined as more than 70% strong nuclear or nuclear and cytoplasmic staining of tumor cells, confirmed by central pathology review. Treatment effect hazard ratios within p16 subgroup were estimated by Cox proportional hazards model and progression-free survival estimates were estimated by Kaplan-Meier method. Analysis was planned to occur after 69 failure events had been reported and tissue specimen analysis was complete.

Will perform multiparametric flow cytometry on peripheral mononuclear cells (PBMC) from patient-derived blood samples to quantify changes in immune cell frequency and activation status before, during and after radiation combined with cetuximab or radiation combined with durvalumab. To assess the statistical significance of pre- to post-treatment changes in levels of immunoglobulin (Ig)Gs as well as other markers, normalized signal intensities (log2) will be tested using a paired t test within an arm and two sample t test between the two arms. If the distribution assumption is violated for the t tests, nonparametric tests, such as the Wilcoxon signed-rank test will be considered. Lastly, the association between PFS and OS and post-treatment changes in frequency and activation state of T-cells, clonality and diversity of T cell receptor (TCR), and IgG levels will be evaluated using a two-sided Wald test at 0.05 level on the basis of Cox models.

PRO-CTCAE is not intended for expedited reporting, real time review or safety reporting. PRO-CTCAE data are exploratory and not currently intended for use in data safety monitoring or adverse event stopping rules.

The mean change from baseline at each time point will be summarized using mean and standard deviations for each arm. Mean change from baseline will be compared between the arms using a two sample t test. If data normality assumptions are not met, the Wilcoxon rank sum test will be used to test the hypothesis. Mean change from baseline will be tested using an omnibus F test followed by individual comparisons of change scores at different time points within each treatment group. The same analysis will be conducted for between group comparisons at each time point.