Unsuspected collision of synchronous lung adenocarcinomas: a potential cause of aberrant driver mutation profiles.
N. Rekhtman,L. Borsu,B. Reva,M. Arcila,Gregory Riely,M. Ladanyi,A. Drilon
DOI: https://doi.org/10.1097/JTO.0b013e3182a471c3
IF: 20.121
Journal of Thoracic Oncology
Abstract:A 75 year-old woman, 50 pack-year ex-smoker, presented with a lung nodule on a routine follow-up chest CT scan for prior colon adenocarcinoma (Figure 1A). A wedge resection revealed an apparently single 1.5 cm nodule on gross examination. Microscopic examination revealed an adenocarcinoma of lung origin, as supported by morphology (Figure 1B) and diffusely positive immunohistochemistry for TTF-1 and Napsin A. The tumor consisted of two geographically-contiguous areas - a 6 mm area with solid and acinar patterns, and a peripheral 9 mm area with a predominantly lepidic pattern (Figure 1C,D). Although the two areas could represent heterogeneous histology in a single adenocarcinoma, the dumbbell shape of the lesion raised the possibility of a collision of two immediately-adjacent separate tumors. The two areas were dissected out and submitted for EGFR and KRAS molecular testing separately. Sequencing revealed a KRAS G12A mutation in the 6 mm area, and a novel KRAS L19F/T20S double-mutation in the 9 mm are (Figure 2A). The latter was confirmed to represent same-allele double-mutation - a complex compound substitution (L19_T20>FS) - by next-generation sequencing (Figure 2B), resulting in potential functional cooperativity based on amino acid modeling (Figure 3). Overall, these molecular data established that the 6 mm and 9 mm areas had distinct KRAS mutations, thus representing two colliding synchronous primaries rather than histologically-heterogeneous areas of a single adenocarcinoma.
Figure 1
(A) Chest CT scan: apparently single nodule (arrow) with solid and ground-glass components. (B) Low-power and (C, D) higher power microscopic images of H&E-stained section, revealing two geographically-contiguous areas with distinct growth patterns ...
Figure 2
Sanger sequencing electropherograms, revealing a KRAS G12A(c.35G>C) mutation in the 6 mm nodule (A) and KRAS L19F(c.57G>T)/T20S(c.58A>T) double mutation in the 9 mm nodule (B). (C) Split-screen view of aligned sequencing reads ...
Figure 3
Amino acid modeling of KRAS L19F/T20S (L19_T20>FS) double mutation. L19F and T20S are highly conserved and belong to the core region of KRAS. L19 is predicted to interact with GTP/GDP-binding residues (N116, A146, and K16 that in turn interact ...
This case illustrates a rare phenomenon of collision tumors, in which two distinct tumors co-occur at the same or immediately-adjacent anatomic locations.1 The exact incidence of collision of synchronous lung carcinomas in unknown, but given that 4–20% of patients with lung cancer have multiple primary tumors,2 the odds of two tumors occurring immediately-adjacent to each other are not negligible. As illustrated here, collisions are inapparent on radiologic and gross examination. While microscopic examination is generally effective at identifying collisions of different tumor types, recognition of this phenomenon for synchronous lung adenocarcinomas is particularly challenging because these tumors are at-baseline highly heterogeneous (i.e. composed of a mixture of lepidic, acinar, solid etc patterns), making it difficult to distinguish histologic heterogeneity in a single adenocarcinoma from two colliding, contiguous tumors. Peripheral lepidic pattern is commonly seen in lung adenocarcinomas. In this case, a dumbbell-shaped outline of the lesion raised the possibility of a collision, but a definitive distinction between a single morphologically-heterogeneous adenocarcinoma versus a collision could not be made based on microscopic findings. This case demonstrates of the utility of driver mutation analysis in elucidating tumor clonal relationships in the setting of a possible collision, similar to what has been recently proposed for the utility of molecular profiling for the distinction of synchronous/metachronous primaries from intrapulmonary metastases.3 We also illustrate the use of next-generation sequencing to determine that the two KRAS mutations in the 9 mm nodule resided on same allele, thus excluding the possibility they were a result of yet another unsuspected collision of tumors with distinct KRAS mutations.
In the era of predictive molecular testing, collision may be an under-recognized, albeit rare, cause for aberrant driver mutation profiles in lung adenocarcinomas. Specifically, while there is strong evidence that major driver mutations in EGFR and KRAS are distributed homogeneously throughout individual tumors, there is a number of reports to the contrary. One explanation for such findings has been geographic variability in mutant to wild-type allele ratio resulting from variable amplification of the mutant allele and/or variable admixture of non-neoplastic cells, leading to inconsistent detection of mutations (i.e. mutation ‘pseudo-heterogeneity’).4,5 This case illustrates another potential mechanism for mutation ‘pseudo-heterogeneity’. In our case, random microdissection with a presumption that all dissected samples were derived from a single tumor would have resulted in the impression of intra-tumoral heterogeneity of KRAS mutations. Alternatively, molecular testing of the conglomerate nodule in the absence of microdissection would have resulted in ‘pseudo-multiplicity’ of KRAS G12A and L19F/T20S mutations. In contradistinction, the L19F/T20S double-mutation illustrates a rare example of true functionally-significant mutation multiplicity.
In summary, this case illustrates unsuspected collision as a previously unreported potential cause of aberrant molecular findings in lung adenocarcinomas, and highlights the importance of careful histologic examination of tissue submitted for molecular testing.