Relapsed/Refractory Chronic Lymphocytic Leukemia Patients Treated with Fixed Duration Venetoclax-Rituximab: Assessment of Response with Ultrasound, and Relationship with Minimal Residual Disease

A fixed duration of venetoclax-rituximab (VenR) resulted in a significant benefit of both PFS and in the attainment of an undetectable minimal residual disease (uMRD) compared with bendamustine-rituximab in relapsed/refractory (R/R) chronic lymphocytic leukemia (CLL) patients. The 2018 International Workshop on CLL guidelines, outside the context of clinical trials, suggested ultrasonography (US) as a possible imaging technique to evaluate visceral involvement, and palpation to evaluate superficial lymph nodes (SupLNs). In this real-life study we prospectively enrolled N = 22 patients. Patients were assessed by US, to determine nodal and splenic response in R/R CLL patients treated with a fixed duration VenR. We found an overall response rate, complete remission, partial remission, and stable disease, of 95.4%, 68%, 27.3%, and 4.5%, respectively. Responses were also correlated with risk categories. The time to response, and the time to clearance of the disease in the spleen, in abdominal LN (AbdLNs), and in SupLNs were discussed. Responses were independent from LN size. The correlation between response rate with MRD were also investigated. US allowed to detect a substantial CR rate correlated with uMRD.


Introduction
Chronic lymphocytic leukemia (CLL) is the most common adult leukemia in Western countries and remains largely incurable despite recent treatment options [1][2][3]. Allogenic stem cell transplantation is unsuitable for most patients [4]. Historically the standard of treatment has been chemoimmunotherapy (CIT) that achieves disease control and prolongs survival [5]. Recently novel targeted therapies in first line or in relapsed/refractory (R/R) CLL patients have improved outcome over CIT [5]. Targeted therapies comprise Bruton tyrosine kinase [BTK] inhibitors, ibrutinib and acalabrutinib, and the phosphatidylinositol 3-kinase [PI3K] inhibitors, idelalisib and duvelisib [5]. The anti-apoptotic protein BCL2 is overexpressed in CLL cells [6,7], and it represents a therapeutic target. Venetoclax is a highly selective potent BCL2 inhibitor that acts regardless of the presence of TP53 mutation. The combination of venetoclax with rituximab (anti-CD20 antibody) has shown to overcome microenvironment-induced resistance to venetoclax [8]. Rituximab can obtain clearance 2 of 11 of minimal residual disease (MRD). In patients treated with CIT, undetectable minimal residual disease (uMRD, <1 CLL cell per 10 4 leucocytes) is linked to more favorable overall survival (OS) and progression free survival (PFS) [9].
In the MURANO study a fixed duration venetoclax plus rituximab (VenR) has resulted in improved PFS and OS compared to bendamustine plus rituximab (BR) treatment in patients with R/R CLL [3]. Investigators have found 93.3% overall response rate (ORR) with 26.8% complete remissions (CR) or CR with incomplete hematological recovery (CRi). After 9 months of treatment the clearance of MRD in peripheral blood (PB) was achieved in 62.4% of patients [6].
The 2018 International Workshop on Chronic Lymphocytic Leukemia (IWCLL) guidelines [9] suggested ultrasonography (US) as a possible imaging technique to evaluate visceral involvement, and palpation to evaluate superficial lymph nodes (SupLNs) in general practice [9]. Ultimate ultrasonography machines are equipped with high resolution linear probes, and US features of normal and diseased SupLNs have been published [10][11][12][13]. Recently, US features of nodal and splenic involvement by CLL has been described [14], and US has been used to assess nodal and splenic response to venetoclax therapy [15].
This study is aimed to assess with US, nodal and splenic response in R/R CLL patients treated with a fixed duration VenR, and to compare the US results with MRD response.
The endpoints of this real-life study are the evaluation of (i) the response rates ORR, CR, partial remission (PR), stable disease (SD), and progressive disease (PD); (ii) the time to response; (iii) the response rate of the LNs with the longest diameter < 5 cm vs. ≥5 cm [16,17]; (iv) the correlation of nodal and splenic response with PB-MRD.

Materials and Methods
This study was conducted in the Hematology Unit of the University of Pisa, Italy, and all patients signed a written consent form (Ethical Committee approval Nr. ID22968, 22 September 2022). We prospectively enrolled N = 22 R/R CLL patients.
The diagnosis of CLL and the definition of R/R patients was made according to the 2018 International Workshop on Chronic Lymphocytic Leukemia guidelines [9]. After 5 weeks ramp-up, all patients received a fixed duration VenR treatment in accordance with what previously described [8]. In detail, venetoclax was administered with a 5-week schedule of gradual increase in the dose (ramp-up) from 20 mg per day to 400 mg per day. After completion of the ramp-up period for venetoclax, administration of rituximab was initiated at 375 mg/m 2 for the first dose (day1 of cycle 1), and 500 mg/m 2 thereafter (day 1 of cycle 2 through cycle 6) in 28-day treatment cycles, while administration of venetoclax was continued. Patients received a total of six 28-day cycles of VenR, followed by singleagent venetoclax (400 mg) once daily for a total of 2 years. The 2-year venetoclax treatment period was calculated from day 1 of cycle 1 after venetoclax dose ramp-up [6].
The US assessment was performed for each patient using a General Electric (GE) Logiq e10s ultrasound sonographer. A 2-5 MHz bandwidth GE convex probe was used for abdominal scanning, and a linear matrix array GE probe bandwidth 4-16 MHz was used for SupLN assessment. The physician who performed the US examinations, was a hematologist member and teacher at the Italian School for basic and emergency ultrasound (SIUMB) of the University of Pisa with more than 20 years' experience in lymphoproliferative disorders and US [14,15].
In this study the spleen has been defined in complete response only when it fulfilled both the dimensional criteria (longitudinal diameter and cross-sectional area) [14,15,20,23], and the attainment of a normal splenic echostructure as previously described [14]. In AbdLNs the echostructure cannot be defined as for SupLNs [12], thus the definition of the response is based only on their dimensions (LNs < 1.5 cm) [9]. When patients were enrolled in the study a total of N = 561 SupLNs and N = 86 AbdLN were analyzed with US before treatment was initiated (T0, baseline). For each patient a median of N = 28 (range 12-32) SupLNs have been scanned with US. Both the number and the echostructure of each LN were recorded and were checked at every following US assessment. If a further US assessment was necessary (for example if treatment had to be temporarily discontinued for any causes) all the SupLNs and AbdLNs found were analyzed before the treatment was resumed. US assessment of each patient was performed at the time points described in Table 1. At each US assessment time point, patients were also assessed for PB-MRD status with flow cytometry (Table 1) [2,9]. Undetectable MRD was defined as below the threshold of one tumor cell per 10 4 white cells [6]. We analyzed PB-MRD because it was shown to be a surrogate of bone marrow MRD as previously described [6]. If a patient discontinued or temporarily discontinued the treatment, the interval of discontinuation was recorded, and US was repeated the day the treatment was re-initiated, and US and MRD follow up assessment continued at each of the time points established at baseline, without any change in the schedule of follow up.

Statistical Analysis
Raw data of the population characteristics are shown in Table 2 and the categorical ones were described by absolute and relative (%) frequency. To analyze categorical data, a chi square test and z-test for two proportions were performed. Significance was fixed at 0.05 and all analyses were carried out with the SPSS v.28 technology.

Patients Characteristics
From December 2021 to July 2022 N = 22 R/R CLL patients were prospectively enrolled in this study. Patients' characteristics, and risk categories are listed in Table 2.

Response
With a median follow up of 12 months (range 6-24 months), we found a nodal (SupLNs and AbdLNS) and splenic response in N = 21/22 patients with 95.4% ORR. We found N = 15/22 (68%) CR, N = 6/22 PR (27.2%), and SD N = 1/22 (4.5%) (Figure 1). We did not detect PD during treatment.  In patients whose response at time of the US assessment was PR (N = 6/22), we found that if the spleen and AbdLNs were pre-treatment involved, they obtained CR. Nevertheless, patients were considered in PR because of the persistence of superficial CLL diseased LNs (Figure 2). Figure 3 shows the time to CR. We found that in patients who obtained nodal and splenic CR (N = 15), N = 13/15 (86.7%) patients reached a CR by 9 months from the beginning of the treatment.
We analyzed if the risk categories of our study population were associated with the quality of nodal and splenic response. We found that IGHV mutational status had no impact on the attainment of nodal and splenic CR (p = 0.717). Nevertheless, we found that patients without Del (17) or TP53 mutated had a higher percentage of CR, with a p-value showing a trend toward a statically significant difference (p = 0.147, and p = 0.101, respectively).
In patients whose response at time of the US assessment was PR (N = 6/22), we found that if the spleen and AbdLNs were pre-treatment involved, they obtained CR. Nevertheless, patients were considered in PR because of the persistence of superficial CLL diseased LNs (Figure 2). In panel G there is still a partial involvement by CLL: the LN has oval shape (L/S > 2), the cortex is thickened and inhomogeneous and reticulated (white arrow), and the posterior part more thickened than the anterior part). The hilum is visible but is compressed and dislocated by the thickened cortex. Figure 3 shows the time to CR. We found that in patients who obtained nodal and splenic CR (N = 15), N = 13/15 (86.7%) patients reached a CR by 9 months from the beginning of the treatment.

Correlation of Nodal and Splenic Response with PB-MRD Response
Fifteen out of 15 patients who attained nodal and splenic CR and two patients in PR, reached uMRD at a median time of 3 months (range 3-15 months). Overall N = 17/21 (80%) patients reached uMRD.

Time to Response
We evaluated the median time to CR of the spleen, of AbdLNs, and SupLNs. We found that it was 6 months (range 3-12 months), 3 months (range 3-9 months), and 6 months (range 3-15 months), respectively (Figure 4).
We analyzed if the risk categories of our study population were associated with the quality of nodal and splenic response. We found that IGHV mutational status had no impact on the attainment of nodal and splenic CR (p = 0.717). Nevertheless, we found that patients without Del (17) or TP53 mutated had a higher percentage of CR, with a p-value showing a trend toward a statically significant difference (p = 0.147, and p = 0.101, respectively).

Correlation of Nodal and Splenic Response with PB-MRD Response
Fifteen out of 15 patients who attained nodal and splenic CR and two patients in PR, reached uMRD at a median time of 3 months (range 3-15 months). Overall N = 17/21 (80%) patients reached uMRD.

Time to Response
We evaluated the median time to CR of the spleen, of AbdLNs, and SupLNs. We found that it was 6 months (range 3-12 months), 3 months (range 3-9 months), and 6 months (range 3-15 months), respectively (Figure 4).

Lymphnodes Dimensions and Response
We found that dimensions of LNs < 5 cm or ≥ 5 cm did not impact on the probability of achieving CR (Figure 2A-C). We found that 94.1% of patients with LNs < 5 cm and 81.8 % with LNs ≥ 5 cm reached CR at T ≤ 6 months, respectively (p-value = 0.688). For T > 6 months, 5.9 % of patients with LNs < 5 cm and 18.2 % with LNs ≥ 5 cm reached CR, respectively (p-value = 0.688) ( Figure 5A,B).

Lymphnodes Dimensions and Response
We found that dimensions of LNs < 5 cm or ≥ 5 cm did not impact on the probability of achieving CR (Figure 2A-C). We found that 94.1% of patients with LNs < 5 cm and 81.8% with LNs ≥ 5 cm reached CR at T ≤ 6 months, respectively (p-value = 0.688). For T > 6 months, 5.9 % of patients with LNs < 5 cm and 18.2 % with LNs ≥ 5 cm reached CR, respectively (p-value = 0.688) ( Figure 5A,B).

Treatment Discontinuation
Four patients had a temporary discontinuation of VenR during the combination treatment phase, for 3.5, 1.5 and 3 months because of SARS-CoV-2 infection and for 16 days because of febrile neutropenia. An additional US assessment of these patients was performed right before the treatment was reintroduced, and it was found that that the quality of response to treatment (N = 3 CR and N = 1 PR) was unchanged despite discontinuation, without any US detectable PD.

Treatment Discontinuation
Four patients had a temporary discontinuation of VenR during the combination treatment phase, for 3.5, 1.5 and 3 months because of SARS-CoV-2 infection and for 16 days because of febrile neutropenia. An additional US assessment of these patients was performed right before the treatment was reintroduced, and it was found that that the quality of response to treatment (N = 3 CR and N = 1 PR) was unchanged despite discontinuation, without any US detectable PD.

3.7.Additional Observations
Using a high resolution US linear probe, we found that in several SupLNs the echostructure showed features of involvement by CLL present only in a small portion of the LN, which was never described before. This echostructure feature was detected in N = 8 patients (36.3% of our casuistic), and in 4/8 it was a transient US status, because eventually achieved a status of US nodal CR ( Figure 1A-C), and accordingly reached a PB-uMRD. Two patients out of 4, with a partial US residual involvement by CLL ( Figures 1D-F and 2D-G) attained uMRD.

Discussion
CLL remains a malignancy largely incurable despite advances in new treatment options [5]. In R/R patients, BTKi ibrutinib is an effective therapeutic option [24], but it requires extended administration until unacceptable toxicity or disease progression. Fixed

Additional Observations
Using a high resolution US linear probe, we found that in several SupLNs the echostructure showed features of involvement by CLL present only in a small portion of the LN, which was never described before. This echostructure feature was detected in N = 8 patients (36.3% of our casuistic), and in 4/8 it was a transient US status, because eventually achieved a status of US nodal CR ( Figure 1A-C), and accordingly reached a PB-uMRD. Two patients out of 4, with a partial US residual involvement by CLL ( Figures 1D-F and 2D-G) attained uMRD.

Discussion
CLL remains a malignancy largely incurable despite advances in new treatment options [5]. In R/R patients, BTKi ibrutinib is an effective therapeutic option [24], but it requires extended administration until unacceptable toxicity or disease progression. Fixed duration therapies may achieve deep and durable clinical responses and may reduce the development of resistant sub clones [25].
The achievement of CR with respect to PR has been shown to correlate with the duration of response and to lead to a longer OS [27]. Seymour et al. reported ORR and SD in 93.3% and 2.1% of patients, respectively [6]. In this study we found ORR in 95.4% and SD in 4.5% of patients without statistically significant difference between the two studies (p= 0.938 for ORR and p = 0.969 for SD). In our real-life study we found that 68% of patients reached nodal and splenic CR.
In IWCLL 2018 guidelines nodal CR is defined in clinical practice as the absence of significant lymphoadenopathy by physical examination [9]. The definition of nodal CR with US of SupLNs takes into account multiple echostructure parameters [10][11][12][13][14]21,24]. Recently, in a prospective study it was shown that US could detect a number of CLL diseased SupLNs higher than palpation, and US could size SupLNs more accurately than palpation [14]. Moreover, it was shown that CLL-SupLNs could become lipoplastic LNs reaching CR during venetoclax therapy, although their dimensions still excided 15 mm [15].
The levels of MRD were shown to be highly predictive of duration of remission after veneteoclax cessation [28]. Seymour et al. in their study reported 62.4% uMRD, and, accordingly, we found uMRD in 80% of patients examined without a statistically significant differences (p = 0.162) [6]. Kater et al. showed that MRD had no effect on PFS in patients in CR/CRi, all having either uMRD or low MRD positivity [28]. Thus, the attainment of CR is of valuable importance.
In our study, we found a high correlation between nodal and splenic CR and uMRD. All patients who achieved nodal and splenic CR also achieved uMRD at median time of 9 months (Figures 3 and 4). We found a correlation between the depth of response with Del(17) and TP53 mutation, but not with IGHV mutational status, according to what previously described [6,28]. However, in patients with Del (17) and TP53 mutation we found only a trend toward a statistical significance probably due to the sample size [6,28].
Wierda et al. [29] and O'Brien et al. [17] showed that patients with LN dimension ≥5 cm have a lower PFS and reached less CR than patients with LN dimension <5 cm. In our study we did not find a correlation between LNs size (LNs < or ≥ 5 cm) and the probability of patients either to reach nodal CR nor a correlation with the time to CR ( Figure 5). An extended follow up of our study population will show if patients with LNs ≥ 5 cm will experience a higher relapse risk.
Moreover, we report on SupLNs with echostructure consistent with a partial involvement by CLL detected with high resolution linear probe. This partial involvement of SupLNs has never been described before in patients affected by CLL. These echostructure features appear to either stabilize into a persistent PR ( Figures 1D-F and 2D-G) or appear a transient status turning into a complete nodal CR ( Figure 1A-C). This transient change in the echostructure of SupLNs might be explained by a total of ten US examinations (1 at baseline T0, three during combination treatment and six from EoCT until the end of second year).
Mato et al. [30] showed that treatment interruption, which mostly occurred during the combination treatment phase, had not statistically significant impact on clinical outcome. In our study, according to Mato et al., four patients temporarily discontinued the treatment during the combination phase without changes of the quality of response achieved.
In conclusion, to the best of our knowledge, in this real-life study we have assessed for the first time in general practice the responses to a fixed dose VenR in R/R CLL patients by US. Ultrasound detected echostructural changes consistent with nodal and splenic CR in a substantial proportion of patients, along with the achievement of uMRD. US has allowed us to follow up patients with frequent time points during the 2-year study period due to the absence of radiation hazard. In a real-life scenario of response assessment of CLL patients, US could be a useful tool and an added value to support palpation.