Mild Baseline Lung Disease and a HER2-Targeted Drug's Own Lung Toxicity
A HER2-mutant lung cancer with a genuinely effective targeted option is complicated by a second, quieter finding on her own staging scan: mild lung scarring, in a disease whose best drug carries a known risk of causing exactly that kind of injury itself.
Marjorie H., 61, sewed for a living for over thirty years — first as a tailor’s assistant, later running her own small alterations business — until rheumatoid arthritis, diagnosed in her forties, made the fine hand work too painful to continue professionally. She still quilts most weekends, working shorter sessions and switching between smaller sections when her hands need a break. Investigation of a persistent cough found a right upper lobe mass; biopsy confirmed adenocarcinoma with a HER2 exon 20 insertion on tissue testing, a driver mutation for which trastuzumab deruxtecan is the best-established targeted option, based on the durable responses DESTINY-Lung02 reported in exactly this population. Her rheumatoid arthritis itself has been quiet for the past two years on the same methotrexate-and-biologic regimen, with no swollen joints on exam today and inflammatory markers only mildly elevated, in a range her rheumatologist has called her personal baseline for years.
Her staging CT, read carefully because of her rheumatoid arthritis history, showed mild bibasilar reticular changes — early, likely RA-associated interstitial lung disease, asymptomatic, with pulmonary function testing that is not quite normal and not yet abnormal enough to act on: FVC 91 percent predicted, DLCO 78 percent predicted — the FVC clearly preserved, the DLCO a few points under the conventional 80 percent floor, which is exactly the pattern early interstitial change produces before it produces symptoms. Both are essentially unchanged from spirometry done two years ago for an unrelated preoperative workup, so this is a stable baseline rather than an active decline. That stability is what matters, and it matters specifically because trastuzumab deruxtecan carries its own labeled risk of drug-induced interstitial lung disease, sometimes severe, occasionally fatal — a risk DESTINY-Lung02 managed by excluding patients with clinically significant ILD at enrollment. Marjorie’s ILD is real but currently silent, which places her in the gap between the trial’s actual exclusion criterion and the more cautious reading some clinicians apply to any radiographic finding at all — a gap the trial was never designed to resolve for her specifically.
A radiographic finding the trial never had to rule on
Start trastuzumab deruxtecan. DESTINY-Lung02 did not exclude every patient with any lung abnormality — it excluded clinically significant, symptomatic ILD, which she does not have. She is asymptomatic, her FVC is preserved, and her DLCO is a handful of points off the floor and has not moved in two years. For an exon 20 insertion there is no comparably effective alternative; withholding the one drug shown to produce durable responses in her specific mutation because of a radiographic finding outside the trial’s own exclusion criterion is over-applying caution the trial itself did not ask for.
You are reading the DESTINY-Lung02 exclusion criterion correctly — she would not have been screened out of that trial.
But the trial’s own safety data still showed drug-related ILD in a meaningful share of enrolled patients who, by definition, started without significant baseline lung disease. Pre-existing fibrotic change is a recognized risk factor for more severe drug-induced ILD across the broader trastuzumab deruxtecan literature, not a theoretical concern specific to this one trial’s inclusion criteria. Treating “not formally excluded” as equivalent to “no added risk” understates what her CT is actually showing us.
Both of those are correctly stated and do not actually contradict each other — she is not excluded by the letter of the trial, and her CT is still a real, drug-relevant risk factor. The honest answer is not picking one framing over the other, it is building a monitoring plan proportionate to a real but unquantified increase in risk.
Baseline HRCT and pulmonary function testing are already done. Set a low threshold for holding the drug at the first sign of new cough, dyspnea, or hypoxia rather than waiting for radiographic confirmation, and have an explicit conversation with Marjorie about what early reporting actually needs to look like — waiting to be sure is exactly the mistake this particular toxicity punishes hardest.
Agreed: start trastuzumab deruxtecan at standard dose given the exon 20 insertion and the lack of a comparably effective alternative, with a structured monitoring plan — scheduled HRCT surveillance, explicit patient education on early cough, dyspnea, or hypoxia, and a pre-specified low threshold for holding the drug rather than waiting for radiographic confirmation of ILD.
Hold the drug immediately, obtain HRCT and start corticosteroids per the established management algorithm, and do not rechallenge without multidisciplinary review.
Continue standard surveillance intervals and reduce HRCT frequency accordingly.
Not agreed: the Pulmonologist remained on record preferring a more conservative starting approach given the added risk, accepting the monitoring plan as reasonable mitigation without agreeing that standard-dose trastuzumab deruxtecan was clearly the better starting point over holding it in reserve behind chemotherapy.