Ivan de Kouchkovsky, MD, University of California, San Francisco (UCSF), continues his discussion on the link between squamous histology and outcomes in metastatic urothelial carcinoma. He highlights the median survival and response rates after immune checkpoint inhibitors (ICIs) and enfortumab vedotin (EV) for patients with squamous histology, as well as the genetic alterations that were often enriched in squamous samples.
The median survival and response rates were substantially shorter after ICIs and EV for patients with squamous histology. Can you share the data points you reported and their implications?
Dr. de Kouchkovsky: In our study, we identified 160 patients, with 120 having pure urothelial carcinoma and 20% exhibiting urothelial carcinoma with some degree of squamous differentiation or histology. Among the 40 patients with urothelial carcinoma and squamous histology, 10% had pure squamous histology, 20% had predominant squamous histology, and 70% had only a focal squamous differentiation component in their biopsy.
When comparing baseline patient demographics and clinical characteristics at the time of metastatic diagnosis, we observed that patients with urothelial carcinoma and squamous histology tended to be slightly older. Specifically, at the initiation of ICIs, these patients were not only older but also appeared to have more advanced disease, indicated by lower hemoglobin and baseline albumin levels, along with a higher neutrophil-to-lymphocyte ratio. However, there were no significant differences observed when analyzing patients at the time of initiating EV, though this subgroup was smaller.
Regarding the results, it’s noteworthy that patients with urothelial carcinoma featuring squamous differentiation had worse outcomes, with inferior overall survival from the time of initial metastatic diagnosis. This trend held true for patients with predominant or pure squamous histology, as well as those with only a focal squamous histology component.
When examining patients specifically at the time of ICI initiation, we noticed a trend toward lower objective response rates in those with squamous histology, along with shorter median progression-free survival and overall survival compared to their pure urothelial carcinoma counterparts.
Similarly, patients starting EV exhibited a significantly lower objective response rate of only 17% for those with urothelial carcinoma featuring squamous histology, in contrast to 70% in patients with pure urothelial carcinoma. Additionally, these patients had shorter progression-free survival on EV and, while not statistically significant, tended to have inferior overall survival as well.
You wrote that “Squamous samples were enriched for CDKN2A, CDKN2B, and PIK3CA, while pure samples were enriched for ERBB2 alterations.” What does this molecular profiling indicate about urothelial carcinoma genetic makeup?
Dr. de Kouchkovsky: In our analysis, we initially suspected that clinical outcomes would differ between the two histologic groups. We aimed to investigate potential molecular differences, specifically genomic disparities, that might shed light on the observed outcome variations. We conducted next-generation sequencing, which was available for approximately 70% of patients. A smaller subset had tumor mutation burden and PD-L1 expression assessments, reflecting evolving clinical practices regarding these biomarkers in ICI treatment for bladder cancer. Our primary focus was on the genomic profile, particularly on alterations present in 10% or more of cases.
We examined around 17 genes, many of which are typically implicated in bladder cancer, including TERT promoter gene, TP53, CDKN2A, and CDKN2B, and compared their prevalence across the 2 histologic subtypes. We observed that tumors with squamous differentiation had a higher prevalence of alterations in CDKN2A, CDKN2B, PIK3CA, and MTAP genes but fewer alterations in the ERBB2 gene, which governs HER2 expression.
Existing literature on this topic presents mixed findings. Some analyses, including a large study from Memorial Sloan Kettering, found no significant differences in the genomic profile of urothelial carcinoma with squamous differentiation when comparing pure and variant histologies. However, the metastatic status and the percentage of squamous differentiation in these cases were not reported. Smaller studies have reported distinct genomic profiles in urothelial carcinoma but with different genes than those we observed. Our analysis should be considered hypothesis-generating due to its smaller sample size and needs validation in larger cohorts. Nonetheless, it hints at potential mechanisms that could explain the observed outcome differences.
As we progress in developing targeted therapies for bladder cancer, exploring genomic differences among various histologies becomes increasingly important. Notably, there is a growing interest in HER2-targeted therapies, including antibody-drug conjugates like trastuzumab deruxtecan and disitamab vedotin. At UCSF, we are conducting a clinical trial investigating a HER2-targeted antibody-drug conjugate.
HER2 expression is influenced by numerous complex factors, not solely genomic alterations. Going forward, it will be intriguing to consider HER2 expression in patients with squamous histology. Our findings suggest that these tumors may have lower HER2 expression, potentially leading to inferior outcomes with HER2-targeted therapies.