**4. Discussion**

In this study, we analyzed *TP53* mutations in relation to clinical outcome in a large cohort of *EGFR*-mutated NSCLC patients receiving first or second generation TKIs as a first line therapy. Our results confirm that exon 8 *TP53* mutations are associated with a shorter PFS, in all settings of treatment.

Moreover, such a negative effect was also observed in the subgroup of patients treated with third generation after the development of T790M mutation.

Numerous studies demonstrated the role of *TP53* mutations in predicting poor prognosis of advanced NSCLC patients [9,11–15], and this was confirmed also in the subgroup of NSCLC patients carrying *EGFR* mutations [8,9,16]. In particular, different recent studies showed that the concurrent presence of *TP53* mutations negatively affects response to TKIs in *EGFR*-mutated NSCLC patients, suggesting a role for these gene mutations in determining primary resistance to these drugs [6,7,17–20]. *TP53* is the most frequently mutated gene in lung adenocarcinoma, with mutation rates reported up to 55% [13,21–23], with a predominantly clonal expression [24]. In our case series, we found 30% of patients carrying a *TP53* mutation in the exons 5–8, the same percentage we previously reported in an independent case series. It is well known that different *TP53* mutations lead to changes in the P53 protein that may have diverse biological significance [9,10,25], and mutations in the DNA-binding domain (exons 5–8), are frequently associated with gain-of-function properties, resulting in pro-oncogenic features of the P53 protein [26]. In our previous work, we found that exon 8 *TP53* mutations were able to predict worse response to *EGFR* TKIs, especially in the subgroup of patients with *EGFR* exon 19 deletion. In the present study we confirmed the negative prognostic value of *TP53* exon 8 mutation in an independent cohort of *EGFR*-mutated NSCLC treated with both first and second generation TKIs in the first line setting. In the present study, the prognostic value of exon 8 *TP53* mutation was evident independently from the type of *EGFR* mutation. In a combined analysis, we showed that the effect was evident on overall survival in *EGFR*-mutated patients who developed T790M at progression after first line TKIs and osimertinib.

These results, in agreement with those reported by Kim et al. [7], suggest a negative predictive role of *TP53* mutation in all lines of therapy and in relation to all TKIs. Furthermore, our results are consistent with a recent study that found that *TP53* mutations in exon 8 are associated with shorter OS of patients receiving a TKI as a first line treatment [27].

In another study, missense mutations in *TP53* gene resulted in shorter PFS in *EGFR* mutated patients treated with TKIs but showed no associations with PFS and OS in patients undergoing surgical resection [28]. According to Xu et al., who reported *TP53* mutations in 88% of NSCLC *EGFR*-mutated patients that responded for <6 months to an *EGFR* TKIs, with respect to 13% of responders for >24 months [29], our results show a higher rate of *TP53* mutations in non-responders group, with no *TP53* mutated patients in the long responder group.

To investigate the role of *TP53* mutations in predicting clinical outcome of patients treated with third generation TKIs, we considered 42 patients' developed T790M mutation to first line treatment with first or second generation TKI and received a third generation drug in the second or further line of therapy. In this subgroup, we found a diminished PFS in patients carrying *TP53* mutations in exon 8, even though without statistical significance, probably due to the small number of analyzed patients; exon 8 *TP53* mutated patients had a significantly shorter OS, with respect to wt *TP53* patients and patients with mutations in other exons of *TP53*. This observation is consistent with previous observations, that identified *TP53* mutations (not only in exon 8) as a negative prognostic predictor [7,16]. This result was not confirmed by a study from Labbé et al., that found no differences in ORR of patients treated with third generation TKIs, based on *TP53* mutation status; this could be for the small size of the analyzed case series [28]. In the light of the paradigm shift brought by FLAURA trial [5], there is a need to identify which biomarkers could predict primary resistance to osimertinib as a first line therapy; if confirmed in a larger case series treated with third generation TKI in the first line, these results could help to better stratify patients, suggesting an *EGFR*-independent mechanism of resistance, as others have already highlighted [30].
