Presentation and investigations
A 62-year-old homemaker, who never smoked, with no known family history of cancer and good past health, presented in January 2024 with cough for 7 months post–COVID-19, progressive weight loss over 2 months, and recent-onset left chest pain. Chest X-ray (CXR) showed left pleural effusion, which was drained at a public hospital, with effusion cytology confirming lung adenocarcinoma with EGFR exon 20 insertion (ex20ins) by Sanger sequencing.
Upon discharge with 2 L/min oxygen, she had a PET-CT scan at a private clinic, which showed a 3.4 x 2.9 x 4.2 cm left hilar mass, bilateral lung nodules of up to 2 cm, and extensive nodal and bone metastases. (Figure 1A) MRI of the brain showed multiple small brain metastases. Her carcinoembryonic antigen (CEA) level was 240 ng/mL.
The patient experienced progressive severe bone pain at her sacral region and lower back, and progressive shortness of breath requiring 4 L/min oxygen. She was readmitted for further care in February 2024. CXR showed recurrence of moderate left pleural effusion with left hydro-pneumothorax, small right pleural effusion, and extensive bilateral reticulonodular lung shadows with lymphangitis carcinomatosis. Owing to bone pain and dyspnoea, her Eastern Cooperative Oncology Group performance status (ECOG PS) was 3. Neurologically, she had difficulty finding words, was forgetful and unable to sit up due to dizziness.
Treatment with amivantamab
On 7 February 2024, standard doses of amivantamab (1,400 mg), pemetrexed and carboplatin as per the PAPILLON regimen were started, with premedication given as per drug insert (antihistamine, antipyretic, and glucocorticoid), along with granulocytic colony-stimulating factor (GCSF) cover and denosumab.1,2
Dosing in week 2 of cycle 1 was omitted for the patient to receive palliative irradiation to the pelvis for pain, and to the thoracic spine for foraminal compression with radiculopathy. After week 3 of cycle 1, she had complete resolution of bilateral pleural effusion and the left hydro-pneumothorax on CXR, while oxygen could be weaned to 2 L/min, at which point the patient was discharged.
At week 1 of cycle 2, the amivantamab dose was reduced to 1,050 mg, along with a 30 percent dose reduction for pemetrexed and carboplatin, as the patient developed grade 1 thrombocytopenia and neutropenia and grade 2 mucositis during cycle 1. In week 3 of cycle 2, amivantamab dosing still needed omitting as the patient suffered from grade 2 mucositis and grade 3 acneiform rash despite pre-emptive oral care with doxycycline, topical clindamycin, povidone-iodine mouthwash, and analgesics for mucositis pain.
Chemotherapy was resumed to full dose at cycle 3, whilst continuing GCSF cover, and amivantamab dose was kept at 1,050 mg, as the patient still suffered from significant mucositis and rash. At this point, she had adequate oxygen saturation and was completely weaned off long-term oxygen therapy, and her ECOG PS had improved to 2, with no difficulty sitting upright nor in finding words. Her CEA level decreased to 119 ng/mL.
At cycles 5 and 6, the patient could walk with a stick and had an ECOG PS of 1, but her CEA began to increase from 119 ng/mL to peak value of 193 ng/mL. A PET-CT body scan and an MRI of the brain were performed after cycle 6, with both scans confirming partial response according to Response Evaluation Criteria in Solid Tumours. Her mucositis and rash had improved to grade 1 and 2, respectively, with the same supportive medications. Upon discussion with her family, the full 1,400 mg dose of amivantamab was resumed; carboplatin was stopped after cycle 6.
At full dose of amivantamab, the patient’s mucositis had completely resolved by cycle 8, and her rash had subsided to grade 1 by cycle 10. After cycle 10, a repeat PET-CT body scan and MRI of the brain in September 2024 showed further tumour size reduction and diminished FDG uptake. (Figure 1B) The patient no longer required a walking aid, while her ECOG PS remained at 1 due to fatigability.
Her blood CEA levels continued to fall slowly and consistently over the ensuing 8 months, from 190 to 38 ng/mL on 14 February 2025. Latest CXR showed residual atelectasis in the left lung base and slightly enlarged left hilar, which remained stable for the past 8 months.
The patient received cycle 17 of amivantamab on 14 March 2025 and is to continue with it until disease progression. Since starting the PAPILLON regimen, her PFS has reached 12 months and is ongoing at the time of writing, and she had not required any radiotherapy to the central nervous system.
Discussion
EGFR ex20ins is the third most common EGFR alteration, with a reported real-world incidence of 1–4 percent among Asian patients with non-small-cell lung cancer (NSCLC).3 Among Hong Kong’s NSCLC patients, the incidence of EGFR ex20ins is 3.5 percent – one of the highest rates in the region, which highlights the importance of biomarker testing and sequencing strategies. Timely identification of less common mutations remains one of the main challenges in NSCLC: while all patients should receive RNA–next-generation sequencing (NGS) testing as per National Comprehensive Cancer Network (NCCN) and European Society for Medical Oncology (ESMO) guidelines, some patients still undergo only polymerase chain reaction (PCR) testing, the results of which may be falsely negative for less common actionable biomarkers.4,5
Rationale for amivantamab use
While NSCLC patients with common EGFR mutations have a median overall survival (mOS) of up to 38.6 months, analyses of real-world data obtained from patients with EGFR ex20ins showed an mOS of 16.2–24.3 months.1 According to a systematic literature review, in patients with EGFR ex20ins treated with chemotherapy alone, the objective response rate (ORR) is 23–29 percent and median progression-free survival (mPFS) is 3.4–6.9 months.6
On the other hand, the phase III PAPILLON study in 308 patients with previously untreated, advanced NSCLC with EGFR ex20ins showed an mPFS of 11.4 vs 6.7 months (hazard ratio [HR], 0.40; 95 percent confidence interval [CI], 0.30–0.53; p<0.001) and a complete or partial response rate at data cut-off of 73 vs 47 percent (rate ratio, 1.50; 95 percent CI, 1.32–1.68; p<0.001) in patients on amivantamab plus carboplatin and pemetrexed (chemotherapy) vs chemotherapy alone.1 (Figure 2)
Amivantamab is a first-in-class bispecific antibody that targets both EGFR and MET, which not only causes EGFR ligand blocking, but also receptor degradation, and possesses immune cell–directing activity, such as antibody-dependent cellular cytotoxicity and trogocytosis.7 Collectively, these mechanisms can bypass the ligand-site resistance against tyrosine kinase inhibitors in patients with EGFR ex20ins, address MET as a resistance mechanism, and recruit effector cells to exert an anticancer effect.1 Amivantamab’s multipronged mechanism of action is likely responsible for the 18-month PFS rate of 31 vs 3 percent reported in the PAPILLON trial for patients treated with amivantamab plus chemotherapy vs chemotherapy alone.1
Our patient presented with life-threatening respiratory failure, but otherwise had good organ function. Amivantamab with chemotherapy was the unequivocal option for a rapid-acting and highly effective treatment. Amivantamab was also the only targeted drug available at the time for EGFR ex20ins NSCLC, and remains the only treatment with compelling data across multiple phase III trials, and has the longest duration of follow-up in EGFR ex20ins NSCLC to date.8-10
OS data and treatment sequencing considerations
The interim overall survival (OS) analysis of the PAPILLON trial showed a HR of 0.67 (95 percent CI, 0.42–1.09; p=0.11) in favour of amivantamab plus chemotherapy vs chemotherapy alone despite a high frequency of crossover to second-line amivantamab monotherapy following disease progression in the chemotherapy group.1 (Figure 3) Additional research suggests that as second-line amivantamab may be unavailable in some clinical settings, the intent-to-treat–based analysis may underestimate the benefit of first-line amivantamab plus chemotherapy. After adjusting for the effect of second-line amivantamab, the OS HRs (95 percent CI) from inverse probability of censoring weighting, two-stage estimation, and rank-preserving structural failure time modelling methods were 0.52 (0.28– 0.94), 0.55 (0.31–0.92), and 0.60 (0.32–1.12), respectively.11
It is important to note that although 42 percent of patients with disease progression on chemotherapy alone were allowed to cross over to second-line amivantamab, they remained at increased risk of death compared with the control arm, which means they may have missed the window of opportunity to benefit most from amivantamab.11
Practical tips on minimizing and managing toxicity
The most common adverse events (AEs) reported in the PAPILLON trial were neutropenia (59 percent), paronychia (56 percent), and rash (54 percent) in the amivantamab plus chemotherapy group, and anaemia (55 percent), neutropenia (45 percent), and nausea (42 percent) in the chemotherapy alone group.1
To minimize the risk of infections resulting from the PAPILLON regimen’s chemotherapy-related haematologic toxicity, it is recommended that patients’ vaccinations are up-to-date before commencing treatment.
During the first two cycles of the PAPILLON regimen, which require weekly dosing, patients need to use prophylactic mouthwash and moisturize their face and upper body at least once daily. Based on experience, early initiation of oral doxycycline or minocycline at the onset of grade 2 rash is crucial to help patients push through this period. Timolol eye drops can be used with occlusion dressing to improve paronychia. Nasal saline douching can relieve nasal mucosal friability and discharge from nasal mucosal ulceration. Steroid-based shampoo and oral antihistamine are often necessary to help patients with scalp rash sleep through the night.
This patient’s experience of declining skin and mucosal toxicity 5–6 months into treatment, with no further new rash developing despite re-escalation of amivantamab to the standard dose, is very common in my practice. Of note, while dose interruptions of any trial agent were reported in 69 percent of patients, only 7 percent of PAPILLON patients discontinued amivantamab due to AEs.1
In the PAPILLON trial, the incidence of infusion-related reactions (IRRs) was 42 percent in the amivantamab-chemotherapy group and 1 percent in the chemotherapy group.1 The phase II SKIPPirr study evaluated several prophylactic strategies to reduce the incidence and severity of IRRs, of which high-dose dexamethasone (8 mg BID) proved the most effective, demonstrating an approximately threefold decrease in the incidence of IRRs vs historical data (22.5 vs 67.4 percent).12 It is therefore essential to utilize the SKIPPirr premedication regimen and have cardiopulmonary resuscitation medication and equipment readily available to manage any potential IRRs during the first infusion of amivantamab, especially in patients with pre-existing cardiorespiratory failure. I also screen nonurgent patients for cardiac risk as per the 2022 European Society of Cardiology Guidelines on cardio-oncology, to ensure that any pre-existing cardiac conditions are well controlled before initiating amivantamab.13
Conclusion
This patient’s response to the PAPILLON regimen has been both efficient and deep, with her ECOG PS improving from 3 at baseline to 2 during cycle 3 and to 1 from cycle 5 onwards, while her PFS is >12 months and still ongoing at the time of writing. Amivantamab’s considerable skin and mucosal toxicity tends to be temporary and begins to subside 5–6 months into treatment in the vast majority of patients, even in those rechallenged at full dose. Given the results of the PAPILLON trial, all patients with EGFR ex20ins NSCLC should be considered for first-line treatment with amivantamab and chemotherapy.
