ctDNA, Liquid Biopsy and Molecular Subtypes

(UroToday.com) San Raffaele University Hospital, Milan Dr. Gibb was the first speaker of this session. He began explaining the aims of the Decipher Bladder Genomic Classifier (Decipher from now on) a molecular subtyping assay designed to predict the risk of upstaging, benefit from neoadjuvant CT, and identify high-risk, neuroendocrine-like disease. Decipher divides by now bladder cancer into 5 subtypes luminal, infiltrated luminal, basal claudin-low, and neuroendocrine-like characterized by a molecular profiling and clinical behavior “much more consistent with the small cell neuroendocrine lung cancer”. Gibbs talked then about the upstaging issue: up to 40% of patients staged as cT1/T2 with TURBT or CT-scan result have pT3+ and/or N+ disease. According to the data of 206 patients of the MOL study the upstaging rate of Non-luminal tumors was higher compared to the upstaging rate of luminal tumors (51 % vs 34% for T1/T2 disease).1

“There are 22k MIBC patients who are potential candidates for Neo Adjuvant Chemotherapy (NAC), but should all these patients receive NAC considering that we have a 40% response rate and 5% benefit?” said Gibb. He then showed some emerging data found in literature about the luminal molecular subtype, which appears to have a less aggressive clinical presentation in many cases, inferring that maybe some of these patients could have good outcomes just with cystectomy alone. Based on the different cohorts (601 patients) of the NACmeta study, “one of the largest NAC vs no-NAC datasets assembled to date”, he showed that in non-selected patients there was a 7% OS at 3 years benefit for NAC vs no-NAC (p=0.03), in line with the evidence found in the literature.

However, while non-luminal patients seem to confirm this benefit, with a 10% net OS benefit at 3 years (p=0.01), luminal patients do not seem to show a significant difference (negative difference of 2% OS at 3 years for luminal patients treated with NAC vs no-NAC p= 0,7). Gibb continued displaying the potential clinical utility of molecular subtyping to identify high-risk neuroendocrine-like (NE-like) diseases. NE-like represents the “worst bad apple” in bladder cancer, associated with worse prognosis and outcomes to standard therapies compared to other subtypes it represents up to 6% of all tumor types and is not identified through the routine pathological screening. The last topic of his talk was focused on how molecular subtype could help predict the benefit of pembrolizumab, and underlie the emerging encouraging results about ICI benefit for claudin-low tumors.

The next speaker of this session was El Naggar, presenting the topic of monitoring response to treatment through ctDNA. Thanks to whole exome sequencing of the patient’s tissue SignateraTM test, identify the top 16 clonal mutations found in tumor tissue not susceptible to treatment’s pressure selection and using them as a patient-tailored assay to test for the presence of circulating tumor DNA (ctDNA) into patient’s blood. El Naggar showed how ctDNA could be useful in different settings: as patient-tailored assessment of bladder cancer’s monitoring like to monitor neoadjuvant response, to assess postsurgical MRD, the effectiveness of adjuvant treatment, to monitor recurrence, or also in metastatic treatment using ctDNA kinetics to evaluate response to immunotherapy.
El Naggar showed the results published by Christensen et al2 on the usage of ctDNA as a method of detection methods for metastatic relapse and to monitor therapeutic efficacy. Of the 68 enrolled patients, those ctDNA positive showed a worse RFS and OS compared to ctDNA negative patients at all time-point (before and during neoadjuvant chemotherapy and after cystectomy). El Naggar continued talking about the use of ctDNA as a pre-specified exploratory biomarker in the IMvigor-010 study, presented by T. Powles. In the observational arm, as presented by T. Powles at ESMO IO 2022 DFS and OS of ctDNA-negative patients have “a far superior result” compared to ctDNA+ patients (DFS HR 6.30 95%, CI 4.45, 8,92; p= <0.0001. OS HR, 8.00 95% CI: 4.92, 12.99; p=<0.0001). even though IMvigor-010 was a negative study, not meeting its primary endpoint, looking to an unstratified post-hoc analysis ctDNA+ patient who received atezolizumab had “a substantial survival advantage” compared to observation-only patients (Median OS 29.8 vs 14.1, HR 0.59). Subsequently, he showed that the clearance of ctDNA at 3rd cycle was associated with improved outcomes both in OS and DFS in the atezolizumab arm.

In the phase II INSPIRE trial, Bratman et al showed that ctDNA dynamics correlated with PFS, ORR, and OS. The ctDNA+ who had a clearance either complete or partial performed better compared to ctDNA+ patients who experienced an increase in ctDNA levels. El Naggar concluded his talk by mentioning the results of the GALAXY study. ctDNA+ patients experienced worse outcomes compared to ctDNA-. Moreover, ctDNA+ patients who received adjuvant treatment had better DFS compared to patients who did not receive any adjuvant therapy. On the contrary, there were no significant treatment benefits for MRD- negative patients.

After El Naggar, J. Ross went up the stage presenting the last talk of this session. He started by introducing liquid biopsy’s proof of concept and the evolution of precision medicine through the last years. The primary role of liquid biopsy is to identify genomic alterations that can indicate a potential benefit to target therapies, but the ease to detect different mutations is not equal. Short variant mutations are easy to detect, on the opposite amplifications are heavily dependent on how much DNA has been extracted. The amount of shedded or detectable ctDNA is variable depending on factors such as tumor stage, histology, vascularity, and treatment.
San Raf BRCA-0.jpg
For this reason, we have to be very careful when selecting patients to perform a liquid biopsy, because “nothing can harm the patient more, than the lost time of trying a liquid biopsy that will fail”, said Dr. Ross.

The liquid biopsy is not a formal germline test, however, the liquid biopsy is highly informative of the germline status as both the ctDNA and the wbcDNA will harbor the germline mutation, “the variant allele frequency should be at or near 50% and for that, the Liquid Biopsy report will alert the ordering physician that formal germline testing on another blood sample is recommended”, explained Dr. Ross.

In Prostate Cancer, PSA is usually associated with tumor burden (this correlation is weaker in high-grade tumors, precise Dr. Ross), according to the data published by Antonarakis,3 when PSA is < 5 ng/ml the chances of a non-informative liquid biopsy are “so high that it should not be performed”, especially in patients with Gleason score ≤ 7.

“Not all BRCA mutations are the same”, continued Dr. Ross, talking about the differences between the different types of mutations among different cancer types. In prostate cancer, which is dominated by BRCA2 mutation loss vs BRCA 2 point mutation in mCRPC. He then showed a case report published by Vanderweele DJ et al, of a patient with a heavily pretreat high burden mCRPC with lung and liver metastases, and a homozygous BRCA 2 loss, who achieved a “durable dramatic sustained complete response” with the PARP Inhibitor Veliparib.

In patients with a base substitution de-activating mutation of BRCA2, one mechanism of resistance to PARPi emerges due to BRCA2 reversion mutations, which can restore the missing protein, with an activity good enough to overcome the effect of the PARPi. Dr. Ross suggested that in patients with BRCA2 homozygous loss, a reversion mutation can’t happen due to gene absence. If we were able to identify these patients early, remarked Dr. Ross maybe they could receive PARPi instead of certainly CT and maybe NHA.

He then continued showing a case study written by Prof. Necchi and sent to the ASCO GU 2023 meeting where the liquid biopsy of a patient with Plasmacytoid Urothelial Carcinoma showed A w532* CDH1 mutation, which appeared to be associated with resistance to immunotherapy and sensitivity to chemotherapy.

Dr. Ross talked thereafter about the issue of clonal Hematopoiesis and how it could add a layer of complexity when interpreting liquid biopsy results. Variant mutations identified with a liquid biopsy could come from clonal hematopoietic cells instead of ctDNA, hence it is crucial not to misinterpret these findings. “Liquid biopsy has a great potential to improve precision and the ability to personalize the treatment for a cancer patient” with this Dr. Ross ended his talk and the session.

Presented by:

  • Dr. Ewan Gibb Principal Scientist, Ph.D., Bladder Cancer Lead Veracyte Vancouver
  • Dr. Adam El Naggar, MD, Medical Director of Oncology, Natera Memphis
  • Prof. Jeffrey S. Ross, M.D., D.SC. Jones-Rohner Endowed Professor of Pathology, Oncology and Urology SUNY Upstate Medical University, Syracuse, NY, Medical Director Foundation Medicine, Inc., Cambridge, MA

Written by: Emanuele Crupi, MD, Resident in Medical Oncology at San Raffaele University Hospital, Milan, @emanuele_crupi on Twitter during the 1st San Raffaele Urologic Oncology Retreat, Friday, Nov 25, 2022


  1. Lotan Y et al., Molecular Subtyping of Clinically Localized Urothelial Carcinoma Reveals Lower Rates of Pathological Upstaging at Radical Cystectomy Among Luminal Tumors. Eur Urol. 2019 Aug;76(2):200-206.
  2. Christensen et al., Early Detection of Metastatic Relapse and Monitoring of Therapeutic Efficacy by Ultra-Deep Sequencing of Plasma Cell-Free DNA in Patients With Urothelial Bladder Carcinoma. J Clin Oncol. 2019 Jun 20;37(18):1547-1557.
  3. Antonarakis et al. Clinical and pathological features associated with circulating tumor DNA content in real-world patients with metastatic prostate cancer. Prostate. 2022 May;82(7):867-875.