Case Presentation: Treating Patients With HER2+ Breast Cancer Case Presentation

LE is a 63-year-old woman with bone-predominant breast cancer who recently transferred her care to us. She presented 14 years ago with an estrogen receptor (ER)-positive, progesterone receptor (PR)-negative, and HER2-negative tumor, for which she underwent lumpectomy, with pathology revealing a 1.8 cm tumor and 14/34 nodes positive. A PET/CT demonstrated multiple (>10) spine lesions at the outset of her therapy, and extensive blastic lesions were again observed on computerized tomography (CT) imaging months later.

Despite strong suspicion for metastatic disease, she received ddAC/T followed by breast/chest wall radiation. She then went on to take letrozole for 9 years before her disease progressed in the spine. This was managed with tamoxifen for another 2 years, at which point progression in the spine was once again noted, and she switched to Palbociclib and anastrozole. Within 3 months, more disease was observed in the bone and she transitioned to capecitabine (flat dose, 1500 mg BID, days 1-14 of 21-day cycles), which controlled the disease for 9 months. She then took everolimus/exemestane for 4 months, at which point FDG PET identified progression in the bone, as well as a liver lesion—her first visceral metastasis. At this point LE had entered our care, and we obtained a biopsy of the liver mass.

Interestingly, this revealed a change in phenotype to ER-negative, HER2 equivocal by IHC but positive by FISH (6.5:2.9 signals/cell, ratio of 2.2). Given that this HER2+ diagnosis arose in a patient with many lines of prior therapy, she did not qualify for any trials. We treated with paclitaxel and trastuzumab, resulting in resolution of the liver lesion by imaging, with a “complete” metabolic response, including bone sites, observed at 6 months. As the patient was developing neuropathy, we withdrew paclitaxel and continued maintenance trastuzumab, with ongoing response so far.

NCCN guidelines advise metastatic biopsy at first recurrence of distal disease. We know less about how often we should biopsy as the tumor develops resistance and the patient progresses through various therapies. This case illustrates the value of biopsy when tumor behavior shifts. We know that tumors evolve; in one study where 382 patients with primary and metastatic biopsy tissue was available, the disease sites were discordant in 33% of cases; 9% demonstrated gain of HER2 overexpression, as in our case. For this patient, the efficacy of treatment with trastuzumab/paclitaxel at visceral and bony sites suggests that HER2 targeting was productive.

Our case also highlights several unknowns, however. When we do detect a shift in phenotype, what is the best therapy? In our case, we administered first-line HER2-directed treatment with success, but would other targeted agents have been more appropriate, as the patient was actually on her fifth line of treatment? And if we continue to treat with HER2-directed agents, will she continue to respond, or could her phenotype shift again? Is it possible that some tumors are simply more mutable than others?

For our patient, we biopsied when the tumor became more aggressive, which raised suspicion that the tumor’s underlying biology had altered. We try to spare patients painful and invasive biopsy procedures when possible, so do not biopsy at each point of progression. Liquid biopsy is attractive as a relatively cheap and non-invasive option, but yield depends heavily on tumor burden, the molecular data that such biopsies provide lacks the histologic information provided by a tissue sample – and often stays stable over multiple assays. Functional assays using RNA-seq may provide a more dynamic picture of tumor activity, but these await solid validation in the clinical setting. Perhaps a combination of such a liquid biopsy with molecular imaging (eg, novel PET assays such as FES- PET) could help to characterize tumors and guide therapy—or at least indicate when tissue biopsy might be most fruitful.