NEW YORK – Tumor-infiltrating lymphocytes (TILS), while known to play an anti-tumor role in many cancer types, were surprisingly associated with worse overall survival among HER2-positive breast cancer patients after neoadjuvant therapy, according to a recent study out of three Italian cancer centers.
Federica Miglietta, an oncologist at the Istituto Oncologico Veneto (IOV) and research fellow at the University of Padua in Italy, presented the findings during the European Organization for Research and Treatment of Cancer (EORTC)'s European Breast Cancer Conference on Friday. Her research showed, contrary to what's been seen in other breast cancer subtypes, that high levels of TILs in residual disease were associated with poorer survival outcomes among HER2-positive breast cancer patients treated with chemotherapy plus HER2-targeted therapy in the neoadjuvant setting.
Measuring TILs in patients' residual disease has yielded conflicting results in the past, Miglietta said during her presentation. For instance, higher levels of TILs in residual disease have been documented as a positive prognostic biomarker for triple-negative breast cancer patients, but studies in HER2-positive breast cancer have been conflicting to date, suggesting at times that the biomarker can play a positive prognostic role, while pointing toward negative prognostic implications in other studies.
To determine whether an increased quantity of TILs is a sign of improved or worsened survival outcomes, Miglietta and her colleagues retrospectively studied residual disease samples in a cohort of HER2-positive breast cancer patients. They analyzed samples taken from stage I-III HER2-positive breast cancer patients treated between 2001 and 2021 across three Italian cancer centers: IOV-IRCCS in Padova; the Azienda Unita Sanitaria Locale di Reggio Emilia; and the IRCCS Humanitas Research Hospital at the Humanitas Cancer Center in Milan.
"The main aims of our study were to deepen the prognostic association between tumor-infiltrating lymphocytes in residual disease and overall survival in HER2-positive breast cancer," she said, adding that she and her colleagues also wanted to assess a prognostic biomarker that combined levels of residual cancer burden and the number of TILs in patients' residual disease samples.
Of 295 patients who'd all received chemotherapy and anti-HER2 agents — consisting of at least Genentech's Herceptin (trastuzumab) — as neoadjuvant therapy, 195 patients had residual disease at the time of surgery, and the researchers were able to determine both residual cancer burden (RCB) and TIL levels in the residual disease (RD-TILs) in 158 patient samples total. Of note, roughly 70 percent of patients with residual disease after neoadjuvant treatment also had hormone receptor-positive tumors.
To stratify patients into biomarker groups, Miglietta and colleagues first sorted the samples into three RCB categories based on the amount of cancer that remained following neoadjuvant treatment, which pathologists scored using hematoxylin and eosin staining. These categories were RCB Class I, which meant minimal residual disease; RCB Class II, meaning intermediate residual disease, and RCB Class III, meaning extensive residual disease. Of the samples measured, the majority — 62.2 percent — were considered RCB Class II.
To measure TIL levels in the residual disease samples, Miglietta and colleagues then used a hematoxylin and eosin staining-based framework that had been proposed by the International Working Group on Immune Biomarkers and published as a recommendation in the journal Seminars in Cancer Biology in 2018. They categorized samples into two groups, "low RD-TILs" or "high RD-TILs," by adopting 15 percent surface area coverage on the residual tumor as the optimal cutoff for their particular cohort.
Among the patient samples analyzed, 82.4 percent fell into the low RD-TIL category.
Survival associations for RCB, RD-TIL
Before combining RCB and RD-TIL into a composite biomarker, Miglietta and colleagues looked into the two separate biomarkers and their individual associations with patients' overall survival outcomes.
As expected, residual cancer burden had a strong association with overall survival: When patients' RCB fell into RCB Class I, the overall survival rate was improved versus Class II, which was improved versus Class III. Indeed, after six years, the overall survival rates were 93 percent, 86.3 percent, and 62 percent for patients who had Class I, II, and III RCB, respectively.
For the RD-TIL biomarker, meanwhile, the researchers found that high levels of TILs in the residual disease samples were associated with poorer overall survival outcomes. Among patients with low RD-TILs according to the 15 percent cutoff, the six-year overall survival rate was 83.7 percent, versus 67.9 percent among patients with high RD-TILs.
After performing a multivariate analysis to rule out contributing survival factors, Miglietta noted that both residual cancer burden and RD-TILs preserved their significant prognostic association with overall survival.
Composite biomarkers
To create a composite biomarker incorporating both RCB and RD-TIL — which they ultimately dubbed RCB-TIL — Miglietta and colleagues used a model that generated a score. Pairing the score with patients' survival data, they found that the composite biomarker had a significant association with overall survival outcomes, both as a continuous score and when broken into three categories: RCB-TIL I, RCB-TIL II, and RCB-TIL III.
The six-year overall survival rates, when based on the composite RCB-TIL score, were 89.8 percent for patients whose tumors were considered RCB-TIL I; 80.2 percent for those considered RCB-TIL II; and 64.4 percent for those considered RCB-TIL III.
"The prognostic performance of this composite marker was numerically higher than that of residual cancer burden and significantly higher than that of tumor-infiltrating lymphocytes in residual disease," Miglietta said.
In an exploratory analysis, she and her colleagues also looked into the change in TIL levels from baseline during neoadjuvant treatment. Here, they found that, although overall TIL levels didn't seem much different between baseline and residual disease samples, they increased to a greater extent when they were low to begin with. Miglietta called this a "peculiar dynamic."
Specifically, almost all of the patients who had high TIL levels in their samples following neoadjuvant therapy showed an increase in TILs from baseline — meaning they converted from low TIL levels at baseline to high levels at the residual disease measuring point.
"This observation may suggest that, while the baseline TILs may reflect the intrinsic immunogenicity of HER2-positive breast cancer, the immune microenvironment of residual disease may … reflect an acquired mechanism [from neoadjuvant treatment]," she said.
Ultimately, in Miglietta's view, the fact that higher TIL levels in residual disease had negative prognostic implications following neoadjuvant therapy might suggest a unique immune cancer cell interaction that's specific to HER2-positive breast cancer.
"Indeed, in this context, the simultaneous presence of residual disease and high immune infiltration may suggest the presence of a vicious loop taking place between cancer cells and immune cells polarized toward immunosuppressive and protumorigenic features," she said.
Implications, next steps
Looking ahead, Miglietta suggested that her findings may ultimately help better refine endpoints used in clinical trials of neoadjuvant HER2-positive breast cancer treatment. The additional information could help shift the way researchers evaluate outcomes in this patient population and treatment setting from a "purely quantitative perspective towards a more qualitative and dynamic appraisal of residual disease," she said.
The composite marker incorporating residual cancer burden and TIL levels in residual disease may help identify patients who face a greater risk of cancer relapse, and for whom additional therapy might be appropriate.
In Miglietta's view, these biomarkers could be incorporated as a tool fairly easily, since RCB and RD-TIL are both highly feasible, validated, and standardized. "The integration of these two biomarkers into the composite score we developed in our study may improve our ability to identify HER2-positive patients at [risk of] poor outcomes, adding information beyond RCB and TILs separately," she told Precision Oncology News. "It could also guide in the selection of patients who may benefit from escalated strategies in the post-neoadjuvant setting."
Of course, given the retrospective nature of the research, Miglietta acknowledged that additional validation studies would be necessary before these potential clinical implications could become reality. The most crucial next step will be prospectively validating these findings in larger patient cohorts.
Second to further validation, Miglietta suggested it would be important to study the composition of the immune infiltrate to determine "whether there actually is an unbalance towards immunosuppressive features" in the cancer. Finally, since the majority of the patients in the cohort she and her colleagues studied had hormone-receptor positive disease, Miglietta said she'd want to look at a group of patients enriched with additional breast cancer subtypes. "It's possible that [hormone receptor status] could affect TIL levels following neoadjuvant treatment, so future studies would need to include multiple subtypes," she said.