EGFR-Driven Lung Adenocarcinoma Acquires and Loses Multiple Resistance Drivers Over Time
A single lung adenocarcinoma can behave like several diseases at once when treatment pressure reshapes the dominant clone. In this 68-year-old never-smoking woman with EGFR exon 19 deletion lung adenocarcinoma, serial biopsies captured a striking sequence: initial response to EGFR-directed therapy, later appearance of concurrent EGFR T790M, EML4-ALK fusion, and MET exon 14 skipping, and then a third biopsy in which those alterations disappeared while KRAS and BRAF bypass mutations emerged.
The case is less about one mutation than about motion. Tumor biology shifted under sequential selective pressure, and each tissue sample documented a different chapter of that evolution.
First Response, Then Acquired Resistance Under EGFR Suppression
At diagnosis, chest CT showed a right upper lobe mass with mediastinal and right hilar adenopathy. Biopsy confirmed invasive adenocarcinoma with a micropapillary component, and molecular testing identified an EGFR exon 19 deletion. Because multistation N2 disease was present, surgery was not pursued. She received one cycle of pemetrexed plus cisplatin and then began gefitinib, with a partial response as the best early outcome.
Toxicity complicated the course quickly. Abnormal liver tests after about 2 months led to a diagnosis of gefitinib-related liver injury, and treatment shifted to icotinib. Even with that interruption, the first-line strategy held disease in check for 10.7 months before a left scapular metastasis appeared.
Plasma NGS from December 2019 already hinted at resistance biology. The assay showed the original EGFR exon 19 deletion, newly detectable EGFR T790M, and low-level PIK3CA E545K. Microsatellite status was stable and blood tumor mutational burden was low. Those biomarkers fit a tumor that remained largely oncogene-driven but was beginning to branch into resistant subclones.
Osimertinib then produced prolonged control. The second progression-free interval lasted 61 months, an unusually durable stretch that makes the later molecular evolution even more informative.
Triple-Driver Resistance Appears on Re-Biopsy
When progression returned in January 2025, repeat biopsy preserved the adenocarcinoma phenotype and excluded small-cell transformation. PD-L1 expression was below 1%. Tissue NGS then revealed an unusual coexistence of EGFR T790M, EML4-ALK fusion, and a MET exon 14 splice-region variant predicted to produce exon 14 skipping, alongside the persistent EGFR exon 19 deletion.
Two details stand out clinically. First, this was not a clean one-driver escape route; it was a multiclone pattern with overlapping bypass pathways. Second, the fusion was identified by DNA-based sequencing only, without orthogonal confirmation by ALK immunohistochemistry, FISH, or RNA testing. The MET alteration also lacked RNA-level confirmation. In practical terms, the tissue supported actionable resistance biology, but not every alteration carried the same level of confirmation.
The treatment choice matched the molecular profile. Aumolertinib plus ensartinib yielded stable disease, with the primary lesion essentially unchanged on serial CT and no new bone lesions. Progression-free survival was approximately 13 months. Serum carcinoembryonic antigen and cytokeratin 19 fragment rose gradually despite radiographic stability, a reminder that biochemical drift may precede obvious anatomic progression in a heterogeneous tumor.
One sentence captures the turning point: the dominant resistant clones were no longer dominant.
KRAS and BRAF Bypass Mutations Replace the Earlier Resistance Pattern
By February 2026, imaging showed enlargement of the primary lesion and a possible right renal metastasis, although the renal lesion lacked pathological confirmation. A third biopsy in March 2026 again showed adenocarcinoma without evidence of neuroendocrine transformation. The molecular profile had changed sharply.
The EML4-ALK fusion, MET exon 14 alteration, and EGFR T790M were no longer detected. In their place, the original EGFR exon 19 deletion had expanded, and several downstream bypass events had appeared: KRAS G12R, KRAS L19F, BRAF D594N, and BRAF E501K. Additional findings included PIK3CA E545K, CTNNB1 S33C, MYC amplification, and CDKN2A loss.
Biologically, this pattern is plausible. Once the tumor was pressured by sequential EGFR and ALK/MET inhibition, resistant cells that could route signaling through the MAPK pathway had a growth advantage. KRAS and BRAF alterations can restore proliferative signaling downstream of receptor blockade, making them classic bypass routes in oncogene-addicted tumors. The case therefore illustrates not only resistance to one drug, but a tumor ecosystem that repeatedly reconfigures its signaling network.
The authors interpret these changes as clonal evolution rather than direct proof of complete eradication of earlier clones. Different specimens, different assay platforms, and spatial heterogeneity all limit how firmly disappearance can be equated with true elimination.
Ivonescimab-Based Therapy Produced Stable Disease, but Attribution Is Limited
After discussion and consent, she started pemetrexed, carboplatin, and ivonescimab, a PD-1 and VEGF bispecific antibody. Grade 2 myelosuppression and rash occurred and were managed symptomatically. Carboplatin was later stopped because of recurrent myelosuppression, and treatment continued with pemetrexed plus ivonescimab.
Follow-up imaging showed no further growth of the primary lesion, no new lesions, stable cervical nodes, unchanged bone scintigraphy, and no change in the right renal lesion. The best response remained stable disease, and she was still receiving therapy at the last follow-up in October 2026.
Because chemotherapy was given concurrently, the activity of ivonescimab cannot be separated from the effect of pemetrexed and carboplatin. Even so, the regimen achieved meaningful disease control in a tumor with low PD-L1 expression, low blood tumor mutational burden, and extensive prior targeted therapy exposure.
Why This Case Matters for Resistance Biology
This report documents a rare longitudinal view of EGFR-mutant lung adenocarcinoma under treatment pressure across three tissue biopsies and three molecular assays. The sequence of findings is clinically instructive: first a classic EGFR-driven tumor, then a multi-driver resistant state with EGFR T790M, ALK, and MET signaling, and finally a bypass-dominant tumor with KRAS and BRAF alterations after dual targeted inhibition.
For clinicians, the case supports repeat biopsy when progression occurs after long targeted control. Resistance patterns can shift over time, and the actionable target at one relapse may vanish by the next. For pathologists and molecular tumor boards, it also shows the limits of single-sample certainty in a spatially heterogeneous cancer.
The unanswered clinical issue is not whether resistance emerges — it does — but which clone is driving the tumor at the exact moment a new treatment decision is made. This case shows how difficult that answer can be, even with serial tissue sampling and broad genomic profiling.