BONSAI-2 study: Nivolumab as therapeutic option after cabozantinib failure in metastatic collecting duct carcinoma patients

Abstract

Background:

The BONSAI phase II trial recently demonstrated the activity of cabozantinib in metastatic collecting duct patients. The outcomes of patients in this setting treated with immunotherapy as second-line is unknown. The aim of the present report was to describe outcomes of patients enrolled in the BONSAI trial that received nivolumab as second-line treatment.

Material and methods:

We describe the oncological outcomes in terms of overall response rate, progression-free survival, overall survival and safety. We excluded patients that did not receive any second-line treatment or were treated with agents other than nivolumab.

Results:

We identified five patients of whom one was excluded due to lack of data. Three patients obtained clinical benefit (one partial response, two stable disease); the second-line progression-free survival (nivolumab) ranged from 2.8 to 19.9 months to and second-line overall survival ranged from 5.1 to 26.5 months. No new safety signals were observed.

Conclusions:

Nivolumab may be considered as second-line therapy option after cabozantinib failure in selected metastatic collecting duct carcinoma patients.

Introduction

Collecting duct carcinoma (CDC or Bellini tumor) is a rare type of non–clear cell renal cell carcinoma (ncRCC). ¹ The diagnosis often occurs at metastatic stage when prognosis is very poor. ¹ While the treatment paradigm for the clear-cell counterpart has been evolved dramatically over the last 20 years, unfortunately metastatic CDC (mCDC) still remains an orphan disease. Patients with mCDC are in fact under-represented in prospective trials, therefore treatment choices are mostly based on small series and case reports. ²

Recently, BONSAI phase II trial showed encouraging results with the use of the multitargeted oral tyrosine kinase inhibitor cabozantinib for the first-line treatment of patients with mCDC. ³ The study met the primary end point, reaching a noticeable objective response rate (ORR) of 35%, and greater to those previously reported for platinum-gemcitabine chemotherapy. ³ Of twenty-three patients treated in BONSAI trial, sixteen patients discontinued the study drug for radiologic progressive disease and were considered for subsequent therapy. Unfortunately, the best subsequent treatment has not been defined yet. Immune-checkpoint inhibitors (ICI) alone or in combination are now recommended as standard-of-care options for clear-cell renal cell carcinoma.^4,5 Cases of excellent response to ICI are reported in literature also for previously treated mCDC patients.^6-8 Among others, nivolumab, a monoclonal antibody directed against the programmed cell death-1 (PD-1), has been used in patients affected by mCDC pretreated with cytotoxic chemotherapy, obtaining interesting results both in terms of ORR and improving patients performance status. ⁷ A phase IIIb study of atezolizumab enrolling 1004 patients with locally advanced or metastatic pre-treated urothelial or non-urothelial carcinoma of the urinary tract included eight CDC patients. ⁹ Unfortunately, data regarding the outcome of CDC patients are not available.

Here we present an exploratory post-hoc analysis from the prospective BONSAI trial in the subgroup of patients who have progressed to the first-line therapy with cabozantinib to assess the oncological and safety outcomes of patients treated with nivolumab as second-line.

Materials and methods

BONSAI-2 study is a pre-specified sub-analysis of BONSAI trial with the purpose of describing the outcomes of patients with mCDC treated with nivolumab after failure of the first-line therapy with cabozantinib. The study design details for BONSAI trial have been published previously. ³ In brief, patients progressed to cabozantinib have been treated with nivolumab intravenously at a flat dose of 480 mg every 28 days until progression, unacceptable toxicity or withdrawal of patient consent. Treatment beyond progression was permitted at investigator judgement in case of clinical benefit.

In this post-hoc analysis overall response rate (ORR), progression-free survival (PFS), overall survival (OS), safety and tolerability were evaluated. Overall response rate was measured according to Response Evaluation Criteria in Solid Tumors (RECIST), version 1.1 by investigator assessment; PFS was defined as the time from first dose of cabozantinib or nivolumab to first documentation of disease progression or death, whichever occurred first; OS was defined as time from cabozantinib or nivolumab starting to death for any cause. For safety assessments, all adverse events and serious adverse events were monitored and recorded with the use of the Common Terminology Criteria for Adverse Events, version 5.0. ¹⁰ Lastly, prognostic risk parameters were analyzed, including International Metastatic RCC Database Consortium (IMDC) risk score and Neutrophil/Lymphocyte Ratio (NLR). IMDC risk score was evaluated before treatment with cabozantinib and nivolumab taking into account clinical (Karnofsky performance status; time from diagnosis to the start of treatment) and laboratory (hemoglobin, platelet count, neutrophil count, corrected calcium) criteria. ³ NLR was obtained by dividing serum absolute neutrophil count by the lymphocyte count in peripheral blood.

The trial was performed in accordance with the International Council for Harmonisation Good Clinical Practice Guidelines and Declaration of Helsinki. Institutional review boards and independent ethics committees approved the study. All patients provided written informed consent before study procedures.

Results

Patients

About 23 patients were treated in the BONSAI trial. Among 23 patients 22 experienced progressive disease. Among 16 patients with radiological progression of disease during cabozantinib therapy, five were treated with nivolumab as subsequent line. One patient was excluded from the present report due to lack of clinical data. Among the remaining 11 patients only one received another active treatment (i.e. chemotherapy), whereas the other ten patients received only supportive care. Clinical and laboratory features are summarized in Table 1.

Table 1.

Clinical and laboratory features.

CDC, n = 4	Pt 1	Pt 2	Pt 3	Pt 4
Clinical and laboratory features
Age at diagnosis, yo	73	61	78	53
Sex	male	female	female	male
Radical nefrectomy	yes	no	yes	yes
Metastases at diagnosis	no	yes	yes	no
Metastatic sites pre-cabozantinib
lung		X
brain
lymph nodes				X
liver
bone		X	X
renal loggia	X	X
Best response to cabozantinib	stable disease	progression of disease	partial response	stable disease
Metastatic sites pre-nivolumab
lung	X	X
brain
lymph nodes	X			X
liver	X
bone	X	X	X
renal loggia	X	X
Best response to nivolumab	progression of disease	stable disease	partial response	stable disease
Clinical benefit with nivolumab	yes	yes	yes	yes
IMDC score
pre-cabozantinib	intermediate	intermediate	intermediate	intermediate
pre-nivolumab	intermediate	intermediate	intermediate	intermediate
NLR
pre-nivolumab	3.8	2.2	2.9	3.0

CDC, collecting duct carcinoma; IMDC, International Metastatic RCC Database Consortium risk score; NLR, neutrophil-lymphocyte ratio; Pt, patient; yo, years old.

The age of participants ranged from 53 to 78 years. Half of the patients were male. Three patients underwent radical nephrectomy. Two patients were already metastatic at the diagnosis.

Before treatment with cabozantinib only one patient had more than one metastatic site, while before treatment with nivolumab two patients developed multiple metastatic sites.

With regard to the analysis of data with a prognostic role, IMDC risk score was intermediate in all patients and NLR level ranged from 2.2 to 3.8.

Three patients underwent radical nephrectomy before cabozantinib treatment. Two patients received radiation therapy about 30 days before starting treatment with nivolumab.

Efficacy and safety

At data cutoff of 1 June 2022, the best response with cabozantinib was a partial response (PR) obtained in one patient (patient n.3). The same patient showed the best response with nivolumab and it was also a partial response (PR). Two patients obtained stable disease and one patient disease progression as best response. Disease progression occurred in all patients (Table 1).

The first-line PFS (cabozantinib) ranged from 2.8 to 18.0 months. The second-line PFS (nivolumab) ranged from 2.8 to 19.9 months. Patient n.3 showed longer PFS with nivolumab while patient n.1 had poor benefit with nivolumab with the fastest time of disease progression (Figure 1).

Figure 1.

Progression-free survival from first-line (cabozantinib) and from second-line (nivolumab). mo, months; PFS, Progression-free survival; Pt, patient.

The OS measured from first-line (start of cabozantinib) ranged from 18.9 to 34.3 months. The OS measured from second-line ranged from 5.1 to 26.5 months. Interestingly, patient n.2 who did not experience response with cabozantinib obtained a clinical benefit with nivolumab, both in terms of PFS and OS (Figure 2).

Figure 2.

Overall survival. mo, months; OS, overall survival; Pt, patient.

All patients experienced clinical benefit in terms of control of cancer-related symptoms. Treatment beyond progression with nivolumab was continued in three patients, one of whom is still alive (patient n.4). Only patient n.2 was treated with additional post-nivolumab therapies: she received chemotherapy with cisplatin and gemcitabine in third-line and everolimus in fourth-line.

Three patients reported at least one adverse event (AE). One patient presented a grade (G) 1 diarrhea, one patient a G1 asthenia, and one patient presented pruritus G2 and pneumonia G3. One patient permanently discontinued nivolumab owing to AE (pneumonia G3).

Discussion

CDC is a very rare subtype of ncRCC with an incidence less than 1% of kidney tumors. It typically occurs in the fifth decade of life, more frequently in men. Often, the diagnosis occurs when the disease is already metastatic and due to its aggressiveness and poor responsiveness to currently available therapeutic weapons (mainly chemotherapy with cisplatin and gemcitabine) the median OS is less than one year. ¹¹

Recently, the BONSAI trial showed encouraging results with use of cabozantinib in untreated mCDC patients, with an objective response rate of 35% and a median PFS of 6.0 months.^3,11 Unfortunately, when disease progression occurs it has not known which is the best subsequent treatment in the second-line setting, due to poor available data, due to the rarity of the disease. ³

The exploration of Tumor Inflammation Signature (TIS) related to the clinical response to PD-1 inhibitors, encouraged the use of immunotherapy in these patients.^2,12

Table 2 shows a series of studies and case reports that show partial and complete responses obtained in patients with mCDC undergoing immune-checkpoint inhibitors. The rationale for using immunotherapy in some studies was based on high PDL-1 expression as a potential predictor of response. Most studies in which mCDC patients were treated with immunotherapy evaluated the efficacy of nivolumab in monotherapy or in combination with ipilimumab; while to our knowledge efficacy data on pembrolizumab are still lacking. Patients may have previously been treated with tyrosine kinase inhibitors (TKIs) such as in data reported by Koshkin et al ¹⁵ and by Yasuoka et al ⁷ who reported data results from a previously heavily pretreated patient who received axitinib in second-line and nivolumab in fourth-line.

Table 2.

Overview of immunotherapy in metastatic collecting duct carcinoma.

Reference	n mCDC patients	Treatment	Treatment-line	Responders
Watanabe et al. (2019) 13	1	nivolumab + ipilimumab	first-line	1 (CR)
Fuu et al. (2022) 14	1	nivolumab + ipilimumab	first-line	1 (CR)
Koshkin et al. (2018) 15	4	nivolumab	first-line or subsequent lines*	1 (PR)
Yasuoka et al. (2018) 7	1	nivolumab	fourth-line**	1 (PR)
Danno et al. (2021) 16	2	nivolumab + ipilimumab	first-line	1 (PR)
Vogelzang et al. (2020) 17	1	nivolumab	second-line	1 (PR)
Tykodi et al. (2022) 18	2	nivolumab + ipilimumab	first-line	0
Mizutani et al. (2017) 6	1	nivolumab	fourth-line***	1 (CR lung, SD lymph node)
Present case series	4	nivolumab	second-line (after cabozantinib)	1 (PR)

CR, complete response; mCDC, metastatic collecting duct carcinoma; PR, partial response.

^*previous treatment lines could include chemotherapy, TKIs or mTOR inhibitor.

^**patient pre-treated with cisplatin plus gemcitabine (first-line treatment), axitinib (second-line treatment), carboplatin plus paclitaxel (third-line treatment).

^***patient pre-treated with cisplatin plus gemcitabine (first-line treatment), paclitaxel plus gemcitabine (second-line treatment), temsirolimus (third-line treatment).

Thus, we reported prospective data about four patients treated with the sequence of nivolumab after progression to cabozantinib as sub-analysis of BONSAI trial.

Interestingly, all patients experienced clinical benefit in terms of control of cancer-related symptoms during treatment with nivolumab regardless of the best radiological response obtained. Patient n.1 had the worst outcome; he started cabozantinib with recurrence only in the renal loggia, while at disease progression he developed multiple metastatic sites, starting immunotherapy with a high disease burden. In addition, he had the highest NLR value among the laboratory data. Patient n.2 who did not respond to cabozantinib, obtained a significant benefit with nivolumab. Patient n.3 achieved a benefit of almost twice with nivolumab compared to cabozantinib in terms of PFS. Patient n.4 is still alive; he was the youngest patient and had only lymph node metastasis. There does not appear to be a clear correlation between the response obtained in the first line with cabozantinib and that obtained in the second line with nivolumab (Table 1).

As far as safety profile, only one patient temporarily discontinued immunotherapy due to G3 pneumonia and no new safety signals emerged.

These results confirmed the heterogeneous behavior of these tumors emphasizing the importance of researching molecular and immunological features exploitable to define a genomic and immune profile that can help identify the most targeted treatment.

The analysis of gene expression and tumor microenvironment including PD-L1 expression and tumor infiltrating lymphocytes could highlight characteristics related to non-responders or responders patients, ascribing a prognostic or a predictive role to these parameters. The BONSAI trial provided molecular analysis for all enrolled patients, which are currently ongoing.

Considered the observed outcomes of patients with CDC in the BONSAI trial and in the present report, evaluation of nivolumab plus cabozantinib as an upfront combination may be reasonable. Even though clear cell RCC is a different tumor entity, the synergistic effect of cabozantinib and nivolumab was demonstrated even in first-line setting. ¹⁹

Limits of the study are closely linked to the study design that is a pre-specified but not pre-planned analysis (despite data having been collected prospectively); data regarding outcome of patients that received different treatments from nivolumab or best supportive care after cabozantinib failure are not available. Moreover, in the BONSAI trial only six patients received a second-line therapy after cabozantinib, suggesting that the results here reported involved a very selected population (as inferred, for example, from the fact that patients 3 and 4 had indolent disease, considering the long PFS achieved with both cabozantinib and nivolumab) and should therefore be interpreted with caution.

Conclusions

Nivolumab may be considered a therapeutic option in selected mCDC patients developing tumor progression to first-line therapy with cabozantinib.