Introduction
Antibody-radionuclide-conjugates (ARCs) based on CD20 antibodies have been used routinely for the treatment of non-Hodgkin lymphoma (NHL), and two ARCs are currently FDA-approved;
131iodine-tositumomab (Bexxar) and
90yttrium-ibritumomab tiuxetan (Zevalin) [
1].
177Lu-lilotomab satetraxetan or Betalutin® (Nordic Nanovector ASA, Oslo, Norway) is a novel ARC that targets the internalizing CD37 antigen, which is expressed on normal and malignant B-cells [
2,
3]. During B-cell development, the CD37 antigen is found on mature B-cells, but it is absent on plasma cells and normal stem cells [
4,
5]. The ARC therapy is currently under investigation in the phase 1/2a LYMRIT-37-01 trial for patients with relapsed CD37+ B-cell NHL. Four different combinations of pre-dosing and pre-treatment have been investigated in the phase 1 study. Two arms with “cold” lilotomab antibody pre-dosing of 40 mg fixed dosage (arm 1) and 100 mg/m
2 body surface area (BSA) dosage (arm 4), and two without (arms 2 and 3). In addition, all patients were pre-treated with different regimens of rituximab, which targets the CD20 antigen, before the
177Lu-lilotomab satetraxetan injection. Investigating arms 1 and 2, we have previously shown that red bone marrow (RM) is the primary dose-limiting organ for the treatment, and that hematological toxicity was more severe for patients receiving higher RM doses [
6]. RM absorbed doses were lower in arm 1 vs. arm 2. Tumor absorbed doses have been previously reported for
177Lu-lilotomab satetraxetan patients, without revealing any significant differences between the first two arms [
7]. Theoretically, the absorbed dose for a given tissue can be increased or decreased by adjusting the amount of radioactivity prescribed to a patient; however, the absorbed doses for all other tissues will be shifted by the same factor. The ratio of tumor to organs-at-risk absorbed doses is therefore a parameter of vital interest when determining the pre-dosage and pre-treatment regimen that optimizes the biodistribution.
The aim of this work was to use the SPECT/CT data to determine tumor and normal tissue absorbed doses for 177Lu-lilotomab satetraxetan patients in all four arms of the phase 1 trial. Furthermore, potential variations in biodistribution and ratios of tumor to RM absorbed doses were to be determined.
Discussion
In this study, we have investigated normal tissue and tumor absorbed doses for 177Lu-lilotomab satetraxetan therapy following four different pre-treatment and pre-dosing regimens. For all four arms, RM was found the primary dose-limiting organ, and both pre-dosing amounts with lilotomab investigated had a mitigating effect on RM absorbed dose. Increasing the amount of lilotomab did reduce the RM and spleen absorbed doses, however, the decrease was not significant. The ratio of tumor to RM absorbed dose was found to significantly increase for both the patient group given 40 mg of lilotomab and the group given 100 mg/m2 of lilotomab compared to patients not pre-dosed with lilotomab.
In our previous work, we have shown that the RM absorbed dose decreased when 40 mg of lilotomab was given as pre-dosing before
177Lu-lilotomab satetraxetan therapy [
6]. In the current study, the approximately fivefold increase in the amount of lilotomab that was given to patients in arm 4 was found to cause a further minor decrease in RM absorbed dose. Only four patients were included from each of the two arms and a larger number of patients is required for statistical verification of this trend. Still, a lower RM absorbed dose is also supported by that a higher radioactivity level could be given to patients in arm 4 (Fig.
1). There is a concern that pre-dosing with cold antibody could block the CD37 antigen on tumor tissues as well, but on a lesion level there was no significant difference in the tumor absorbed dose for arm 1, arm 4, and not pre-dosed patients. However, the overall variation in tumor absorbed doses was highest in arm 4, so we cannot exclude that uptake was influenced by the lilotomab pre-dosing for some lesions (Fig.
3a).
The amount of unlabeled antibody that produced the highest ratio of tumor to whole-body absorbed dose was investigated for individual patients in the first
131I-tositumomab trials [
10]. In the current work, we calculated the ratio of tumor to RM absorbed dose to compare all
177Lu-lilotomab satetraxetan therapy regimens. The ratio doubled for patients receiving 40 mg lilotomab and doubled again for the patients given 100 mg/m
2 BSA of lilotomab, indicating that the higher pre-dosing level can optimize the therapeutic effect. It should, however, be noted that the ratio parameter considers the mean tumor absorbed dose across all lesions per patient, and large intra-patient variations in tumor absorbed dose can possibly be obscured. The variation is visualized in Fig.
3d. The limited number of patients in each group with more than two tumors eligible for dosimetry does not allow for statistical comparisons, but for arm 4 the intra-patient tumor absorbed doses variation appears somewhat larger, and the largest intra-patient range (710 cGy) was found in this arm (Suppl. Table
1). Here, we aimed to perform an overall assessment of the different groups rather then effect prediction for individual patients. For such prediction studies, a more suiting parameter could perhaps be the ratio of the patient
minimum tumor absorbed dose to RM absorbed dose. However, the same trend is shown using this parameter; increasing values for non-pre-dosed patients, arm 1, and arm 4 (data not shown). A larger uncertainty will possibly be introduced by such a parameter, since not all lesions are eligible for dosimetry.
There were some differences in absorbed dose for other normal tissues, but the only significant difference was for the spleen. This absorbed dose was significantly lower for arm 4 patients than for patients not pre-dosed with lilotomab (Table
2). All organs received absorbed doses within commonly assumed tolerance levels, e.g., the highest spleen absorbed dose across all patients was 6.5 Gy (patient 005), which is lower than the absorbed doses observed to have an effect for other lutetium-177-based treatments [
11,
12]. Accordingly, no signs of non-hematological toxicities were observed for the included patients. For the calculation of ratios between tumors and organs-at-risk, we therefore focused on RM as the most important normal tissue.
Pre-dosing with unlabelled antibody as a means of improving biodistribution has been demonstrated effective for ARCs targeting CD20. Treatment with
131I-tositumomab was described to be preceded by 450 mg cold tositumomab [
13] and
90Y-ibritumomab tiuxetan treatment uses 250 mg/m
2 BSA cold anti-CD20 rituximab as pre-dosing [
14]. The theory is that unlabeled antibody will bind the circulating non-malignant B cells expressing target antigens. Administration of pre-dosing therefore prevents rapid ARC sequestration in the spleen and will, as a result, prolong the ARCs’ plasma half-life [
15]. Reduction of ARC uptake in the spleen is clearly visible, for example
131I-tositumomab [
13], and while the spleen uptake even without lilotomab pre-dosing was lower for
177Lu-lilotomab satetraxetan, a corresponding reduction can also be observed for this treatment (Fig.
2). We have found that the cumulative activity in blood was higher, and the clearance was lower, for arm 1 patients than for arm 2 patients [
6]. The higher amounts of lilotomab given patients in arm 4 increased the cumulative radioactivity in blood even further [
16]. This may also explain the somewhat higher mean tumor absorbed dose in arm 4; if the concentration in the blood increases, this may lead to overcoming the binding site barrier and increase diffusion into the tumor [
17]. The observed changes in biodistribution may indicate that the same mechanics are applicable for pre-dosing with lilotomab before
177Lu-lilotomab satetraxetan treatment as for previous ARC pre-dosing regimens. For the patients enrolled so far, the maximum amount of lilotomab pre-dosing has been 224 mg (Table
1). This is lower than the pre-dosing amounts given for
131I-tositumomab or
90Y-ibritumomab tiuxetan, but can be considered in relative agreement with that the antigen expression of CD37 has been measured approximately half of the CD20 expression in vitro [
5,
18]. It is uncertain whether lilotomab pre-dosing levels above 100 mg/m
2 BSA could prove beneficial. While the increased tumor to RM ratios encourage such investigations, the larger absorbed dose variation for lesions may advise against a continued escalation of pre-dosing amounts. In studies of an iodine-131 labeled anti-CD37 antibody, MB-1, three different protein amounts were investigated, 0.5, 2.5, and 10 mg/kg, which corresponds to 35, 175, and 700 mg for a 70 kg patient [
19,
20]. The highest amount yielded the most favorable biodistribution in the majority of patients. Interestingly, the cold antibody was given simultaneously as the ARC (not as pre-dosing), and one should then believe that both cold and radiolabeled antibodies would bind non-malignant B cells and tumors with the same relative effect. This difference in administration does make direct comparisons with the
177Lu-lilotomab satetraxetan pre-dosing regimens challenging. However, this calls for a closer investigation of the amount of radiolabeled antibody given. For arm 1, the amount of
177Lu-lilotomab satetraxetan was relatively invariable, and for arm 4 the amount was less than 5% of the amount given as cold lilotomab pre-dosing. A clear deviation was found for one of the patients that did not receive lilotomab pre-dosing, as this patient was given approximately twice the amount of radiolabeled antibody compared to the rest (patient 18, 16.4 mg, supplementary Table
1). While we cannot rule out that variable amounts of radiolabeled antibody will impact the biodistribution, the RM and tumor absorbed doses for patient 18 were within the range of the other arm 3 patients. The absolute amounts of radiolabeled CD37 antibody given are probably too low for any measurable effects of possible differences.
Two arms excluding lilotomab as pre-dosing have been investigated. In the present work, the absorbed doses have been reported separately for these two arms, since the rituximab timing varied between the arms. However, no large differences were observed (Table
2), and the data were combined for some of the analyses and discussion. Pre-dosing with the anti-CD20 targeting rituximab on the same day as
177Lu-lilotomab satetraxetan was investigated in arm 3 because rituximab will also bind Fcγ receptors [
21], and although lilotomab is a mouse antibody, it also binds to subtypes of human Fcγ receptors. Our results show that the rituximab pre-dosing will not introduce the same protective effect for RM as pre-dosing with the same anti-CD37 antibody as is part of the ARC, indicating that the blocking mechanics discussed above are antibody-specific. This is in accordance with that rituximab has been found to block radiolabeled anti-CD20 antibodies, but not radiolabeled anti-CD45 antibodies [
22].
In general, biodistribution and dosimetry studies allow for the determination of uptake in organs-at-risk and tumors, and hence the selection of an optimal pre-treatment and pre-dosing regimen. It is an open question as to whether this process should be conducted before activity level escalation (often called dose-escalation) is performed. This could minimize the number of patients in arms that are later judged less effective. While the correlation of RM absorbed dose and hematological toxicity has been demonstrated for the current phase 1/2a trial [
6], the observed variation in tumor absorbed dose prevents fully reliable clinical translation, and further studies are needed to investigate tumor absorbed dose vs. patient outcome. If this issue is resolved, a tumor to RM absorbed dose ratio parameter could prove valuable for predictive purposes.