Response to the Letter to the Editor: Clinical and Pathological Implications of Tumor Genomics of Predominant Histologic Subtypes in Lung Adenocarcinoma

In Response:

We appreciate the comments from Dr. Xu and colleagues¹ regarding our recently published manuscript “The underlying tumor genomics of predominant histologic subtypes in lung adenocarcinoma.”² As noted, we found that lepidic tumors have a higher incidence of actionable EGFR mutations, compared with other histologic subtypes. No patient in our study received neoadjuvant tyrosine kinase inhibitor (TKI) therapy, and only 3 patients (1 acinar, 1 micropapillary [MIP], 1 solid [SOL] histologic subtype), all for an EGFR exon 19del mutation, received adjuvant TKI therapy. Thus, only 3.6% of patients (3/83) with an actionable EGFR mutation received adjuvant TKI therapy. We agree that spread through air spaces (STAS) is an important predictor of locoregional and distant recurrence. We found that STAS was more common in the MIP/SOL group, and we also observed that alterations in the PI3K oncogenic pathway were associated with an increased risk of recurrence in this same group. We agree with our colleagues that further experiments and analyses are indicated to ascertain what, if any, relationships exist between STAS and PI3K pathway alterations. There were no differences in rates of locoregional recurrence between the acinar/papillary and MIP/SOL groups, but there was a significant increase in distant recurrences in the MIP/SOL group. The majority of recurrences in our series were distant. Unfortunately, we do not have the exact organ site of recurrence in our data set and therefore did not perform pathway analyses with site of distant recurrence.

Finally, mutational signatures are extracted using somatic mutations in a sample and assigned to a set of predefined canonical signatures from COSMIC. As is common practice, both synonymous and nonsynonymous mutations are included in mutational signatures analyses. In contrast, only nonsynonymous mutations are used to compute tumor mutational burden. Tumor mutational burden and mutational signatures and their proposed etiology across histologic subtypes are shown in our paper’s Figure 3A. To address the question of which mutational signatures are associated with histologic subtype, please see our paper’s Figure 3C, where selected mutational signatures for each subtype grouping are displayed.

We appreciate the interest in our paper and look forward to future studies that will generate new knowledge regarding the mechanisms underlying the aggressive biologies associated with selective lung adenocarcinoma histologic subtypes.