Antibody-drug conjugates for the treatment of lymphoma: clinical advances and latest progress

Theoretically, an ideal antigen is broadly and homogeneously expressed on malignant cells and hardly present on normal cells, thereby maximizing efficacy and minimizing systemic toxicity. In most cases, ADCs are taken up by selective internalization via receptor-mediated endocytosis and yield intracellular cytotoxicity. The expression of homogeneous target antigens is not absolutely required for ADC efficiency, as heterogeneous tumors may benefit from bystander killing, which can affect proximally located antigen-negative tumor cells [11]. The effectiveness of the ADC-mediated bystander effect mostly relies on factors such as the extent of ADC internalization, the type of linkers, and the physicochemical properties of the attached payloads [5].

Antibodies used for ADCs for lymphoma are made to be chimeric, humanized, or fully human to generate the best immunogenicity profile. Owing to its favorable stability, immunoglobulin G (IgG) is the predominant class used in ADCs, and IgG1 represents a major subtype in the clinic [12]. IgG can induce antibody-dependent cell-mediated cytotoxicity (ADCC), activation of antibody-dependent cell-mediated phagocytosis (ADCP), and the complement pathway by triggering FcγR-expressing cells [13]. Nevertheless, the strong induction of these effects may maximize tumor uptake at the expense of unacceptable systemic toxicity. The IgG4 isotype and Fc-mutated variants of the IgG1 isotype are designed to attenuate toxicity [14]. Non-IgG binding proteins may be a substitute that can be tailored for individual applications with varying degrees of stability and optimized efficacy [15, 16]. The IgM type has also been considered in development because of its greater number of engineered functional sites and higher drug-antibody ratio (DAR) [17]. In addition, triple variable domain Fab, with a small molecular weight, has been proposed as a new format that contributes to deeper tumor tissue distribution, affording stable generation of ADCs [18]. Nevertheless, the design of antibody formats needs to strike a favorable balance between reduced molecular weight and decreased plasma half-life, which can facilitate and vitiate tumor uptake, respectively.

The cytotoxic payloads, as the final effector component, should meet core requirements, such as having strong potency in the minute range within which they are released. The payloads that have been validated in the clinic are categorized into two major groups, namely, DNA-interacting agents (e.g., duocarmycin, calicheamicin, doxorubicin, camptothecin analogs, and pyrrolobenzodiazepine [PBD] dimers) or tubulin inhibitors (e.g., maytansines and auristatins). Inotuzumab ozogamicin (InO) employs calicheamicin as an efficient payload and has been investigated in multiple lymphoid malignancies, including diffuse large B cell lymphoma (DLBCL), mantle cell lymphoma (MCL), and follicular lymphoma (FL) [19‐21]. PBD dimers target DNA minor grooves and therefore cause DNA alkylation. ADCT-301 and ADCT-402, which deliver PBD dimers, are under development [22, 23].

Tubulin inactivators, the auristatins monomethyl auristatin E (MMAE), and monomethyl auristatin F (MMAF) impede microtubule polymerization, leading to cell apoptosis. MMAE is employed in Pola and BV [24, 25], and MMAF is utilized in denintuzumab mafodotin [26]. In addition, MMAE, unlike MMAF, yields bystander effect in vivo [27]. Maytansine derivatives are mainly divided into two categories: DM1 and DM4. DM1 agents are highly potent maytansinoids (emtansine and mertansine) that have broad killing effects on non-Hodgkin lymphoma (NHL) xenografts in vivo [28]. DM4 agents comprise soravtansine and ravtansine (e.g., coltuximab ravtansine), which can augment the bystander effect of adjacent cells in vivo, thereby eradicating tumors [29, 30].

Compared with anti-microtubule payloads, DNA-damaging agents can provide higher potency and enable ADCs to target less abundant tumor antigens [31]. A trend of assessing DNA-interacting agents as ADC payloads has been emerging [32, 33]. Indolinobenzodiazepine dimers (termed IGNs), a novel chemical class of DNA-interfering payloads, feature enhanced bystander effect and tolerability as well as expanded therapeutic utility [34]. Seco-CBI dimers, a class of duocarmycin analogs, are presently being explored in NHL and are conjugated to anti-CD22 mAb, affording a more durable response and higher activity in indolent NHL [35].

Linkers provide a functional handle for efficient conjugation to antibodies. At present, cleavable and non-cleavable linkers are two major classes widely applied in the clinical setting. Due to their ability to liberate diffusible payloads, cleavable linkers are implemented in a broad range of tumor types. Three main types of cleavable linkers respond to physiological stimuli, such as low pH: acid-labile hydrazone linkers, disulfide linkers, and protease-cleavable linkers. N-succinimidyl-4-(2-pyridyldithio)-pentanoate (SPP) and N-succinimidyl-4-(2-pyridylthio)-butanoate (SPDB) represent two classes of disulfide linkers that are unstable at high glutathione concentrations (e.g., used in the maytansinoid-based ADC naratuximab emtansine). Protease-cleavable linkers, the most stable type, comprise dipeptide sequences such as valine-citrulline (Val-Cit) and valine-alanine (Val-Ala) that are generally cleaved by cathepsin B intracellularly (e.g., that used in Pola). Nonetheless, sufficient extracellular proteolytic cleavage of the Val-Cit linker is also observed, indicating that endocytosis of ADC is not required under certain conditions and that developing non-internalizing antibodies may be a potential option for increasing antitumor activity [36].

Concerning the pursuit of increased plasma stability, non-cleavable linkers appear to be a better option for ADCs. Non-cleavable linkers mainly depend on sufficient processing of mAbs within the lysosome after ADC internalization to release the toxin due to the lack of cell permeability. These linkers are accompanied by limited bystander effect and have lower membrane permeability.

The attachment between mAb and linker is a crucial parameter of ADCs as it determines the drug-to-antibody ratio (DAR) and consequently the stability, efficacy, and homogeneity of the ADCs. Site-specific methods are proposed to optimize the DAR and maintain favorable pharmacokinetics while maximizing payloads in the development of ADCs. Several site-specific conjugation approaches that utilize integrated non-natural amino acids, engineered cysteine residues [37] or enzymatic modifications, including transglutaminases [38], sortase A [39], glycosyltransferase [40] and formylglycine-generating enzyme [41], have been reported. The chemoenzymatic approach can harvest a high yield of ADCs with suitable DARs and attractive conjugation efficiency by achieving site-specific conjugation and generating precisely controlled modification sites [42].

The site-specific method has enabled the generation of multivalent conjugated drugs. For example, a heterotrifunctional linker was designed to prepare a dual-cytotoxic (MMAE and DM1) drug conjugate in a site-specific manner, highlighting the potential of ADCs with distinct mechanisms of action for targeting neoplasms and attaining obvious synergy compared with single agents [43]. Additionally, dual-drug combinations also offer possibilities for circumventing ADC-related drug resistance in practice [44]. Therefore, selection and design strategies are complementary and particularly powerful when used in combination.

CD30, a transmembrane glycoprotein belonging to the tumor necrosis factor (TNF) receptor superfamily, is expressed on a small subset of activated B and T lymphocytes and restricted to normal tissues. CD30 is of high interest as a therapeutic target for antibody-based treatments owing to its excessive and selective expression on cancer cells [45]. The prognostic and therapeutic impacts of CD30 expression have been investigated in Hodgkin lymphoma (HL), anaplastic large cell lymphoma (ALCL), cutaneous T cell lymphoma (CTCL), and primary mediastinal B cell lymphoma (PMBCL) [7, 46].

BV comprises a CD30 targeting chimeric IgG1 mAb, an MMAE payload moiety, and a Val-Cit linker. A series of studies have explored the clinical benefits of BV in diverse settings, such as second-line therapy for relapsed or refractory (R/R) disease, consolidation therapy, salvage therapy, and frontline therapy in HL (Table 1). The encouraging results led to the approval of BV by the U.S. FDA for patients with relapsed classical Hodgkin lymphoma (cHL) and relapsed ALCL [47]. In a pivotal phase II study of BV monotherapy, the overall response rate (ORR) was 75%, with 34% complete response (CR) in patients with R/R HL who failed autologous stem cell transplantation (ASCT) [48]. For patients achieving a CR, the median duration of response (DOR) was 20.5 months, with 5-year progression-free survival (PFS) and overall survival (OS) rates of 52% and 64%, respectively. The most common side effects related to the drug were peripheral neuropathy (PN), nausea, fatigue, neutropenia, and diarrhea. In the phase III KEYNOTE-204 study (NCT02684292) involving patients with R/R cHL who had received more than two prior therapies and were treated with BV or pembrolizumab (a PD-1 inhibitor) monotherapy, the ORR was 54% for the BV arm and 65% for the pembrolizumab arm [49]. The reported median PFS was 8.2 months for BV and 12.6 months for pembrolizumab. For patients experiencing severe adverse effects (AEs), the proportion was comparable in the BV and pembrolizumab arms (24% vs. 23%). This trial suggested the potential benefit of BV and pembrolizumab combination therapy in patients who are heavily pretreated. Nonetheless, some patients will ultimately develop BV resistance and may partially attribute to the upregulation of NF-kappaB [50] or multidrug resistance pump (MDR1) [51]. The combination therapy of BV and cyclosporine A (an MDR1 inhibitor) was recently examined in a phase I trial (NCT03013933) to combat BV resistance in cHL, which reported an ORR of 75%, a CR rate of 42%, and a modest toxicity profile [51].

The phase III AETHERA trial (NCT01100502) measured the efficacy and safety profiles of BV in patients with cHL in the post-transplant consolidation setting [52]. The 5-year PFS rates were 59% in patients with BV and 43% in patients with placebo. The most common AEs associated with the therapeutic agent included neurotoxicity (67%), infection (60%), and neutropenia (35%). The majority of PN and neutropenia cases were reversible and managed with dose delay or reduction. Recently, a phase II study evaluated the results of 59 participants with R/R aggressive HL receiving BV and nivolumab (Nivo), a PD-1 inhibitor, as post-ASCT consolidation therapy [53]. The estimated 18-month OS and PFS rates were 98% and 95%, respectively. Treatment was well tolerated, with common AEs of PN (51%) and neutropenia (42%). The encouraging results suggested that BV plus Nivo  may provide prolonged remission in patients at an advanced stage.

The incorporation of BV in salvage therapy for patients with R/R diseases has been evaluated by a few studies. One phase II study (NCT02280993) including 55 R/R cHL patients treated with BV plus dexamethasone, cisplatin, and cytarabine (DHAP) with a PFS rate of 74% and OS rate of 95% at the 2-year follow-up was recently reported [54]. For patients who proceeded to ASCT, 42 patients achieved a metabolic CR and 5 achieved a metabolic partial response (PR). A BV and bendamustine regimen also provided a robust efficacy benefit in HL patients in a single-arm multicenter phase II study [55]. The ORR was 84%, with 30 patients (79%) achieving a CR and 2 patients (5%) achieving a PR. In addition, 33 patients underwent ASCT. The estimated 3-year OS and PFS rates were also promising, at 88% and 67%, respectively. Skin reactions were rather frequent (65%) and should be monitored during treatment. Recently, BV plus Nivo therapy, the first chemo-free combination, was reported to have an ORR of 85% and a CR rate of 67% for all treated patients with cHL in a phase I/II trial (NCT02572167) [56]. The estimated 2-year PFS rate in all enrolled patients was 78%. For patients who proceeded to ASCT after receiving BV + Nivo, the proportion was 91%, suggesting that further exploration of this combination treatment as a bridge to ASCT in a larger cohort is warranted.

The approval for frontline use in combination with chemotherapy was based on the impressive activity and manageable safety profiles observed in the phase III ECHELON-1 trial in advanced-stage cHL [57]. Updated data from the ECHELON-1 cohort (NCT01712490) demonstrated an overall risk–benefit ratio favoring the BV plus doxorubicin, vinblastine, and dacarbazine (A + AVD) arm over the doxorubicin, bleomycin, vinblastine, and dacarbazine (ABVD) arm in patients with previously untreated cHL irrespective of prognostic risk score [58]. At a 5-year follow-up, a durable efficacy benefit was found in participants receiving A + AVD compared with ABVD, with an improved PFS rate (82% vs. 75%), albeit with a comparable ORR (86% vs. 83%) and CR rate (73% vs. 70%). A higher incidence of PN was seen in the A + AVD (67% vs. 43%) arm with an increased rate of febrile neutropenia (19% vs. 8%) and a decreased rate of pulmonary-related toxicity (2% vs 7%). ABVD therapy was considered unsuitable for further development based on a relatively high incidence of pulmonary-related toxicity. In a single-arm, phase II trial (NCT02758717), elderly patients who were considered unsuitable for standard chemotherapies received BV + Nivo with an ORR of 61% and a CR rate of 48% [59]. Neurotoxicity was noted in 14 (30%) of 46 patients and may have accounted for dose adjustments. Currently, a phase III trial (NCT03907488) of the A + AVD arm versus the Nivo + AVD arm is undergoing in patients with newly diagnosed advanced-stage HL.

BV has been evaluated in various types of R/R NHL, including mycosis fungoides (MF) and primary cutaneous ALCL. Efficacy analysis was conducted in a pooled population of 128 patients with CD30-positive MF or primary cutaneous ALCL who failed prior systemic therapies and were enrolled in the multicenter randomized phase III trial (ALCANZA) [60]. ORR lasting at least 4 months (ORR4) intensely favored the BV arm compared to the control (methotrexate or bexarotene) arm (56.3% vs. 12.5%). Notably, this favorable clinical efficacy of the BV group was observed for other endpoints, such as CR rate (16% vs. 2%) and median PFS (17.2 vs. 3.5 months), and manifested as a reduction in tumor burden. Grade 3 to 4 adverse events were reported less frequently in the BV group than in the control group (41% vs. 47%), while PN was seen frequently in the BV group (67% vs. 6%), most of which were grade 1 to 2. Studies combining BV with other immunomodulatory agents or chemotherapies have revealed improvements in outcomes, compared with BV monotherapy. The initial outcomes of a phase II study of BV and lenalidomide (Len) in heavily pretreated subjects with R/R CTCL and R/R PTCL reported a CR rate of 17%, a PR rate of 25%, and an ORR of 38% [61]. Recruitment of both CTCL and PTCL patients for this trial is ongoing (NCT03409432). BV has also been shown to be effective in patients with newly diagnosed PTCL. The ECHELON-2 trial (NCT01777152) observed a consistent response and an OS benefit in aggressive PTCL (predominantly ALCL) with the BV, cyclophosphamide, doxorubicin, and prednisone (A + CHP) regimen compared with a proven standard therapy (CHOP) [62]. The A + CHP group had a superior 5-year PFS rate (51% vs. 43%), an improved CR rate (68% vs. 56%) and ORR (83% vs. 72%), and slightly increased toxicity [63]. The addition of BV in frontline therapy for ALCL has also been evaluated in several studies in the hope of improving cure rates. A recently reported phase II trial enrolled 68 pediatric patients with newly diagnosed ALCL and treated them with BV in combination with standard chemotherapy [64]. The 2-year event-free survival (EFS) rate was 79%, and the 2-year OS rate was 97%, with no cases of severe neuropathy occurred, suggesting that the addition of BV to standard chemotherapies may be tolerable and used to prevent relapses in children.

Additionally, the BV plus Len regimen has also been explored in patients with R/R DLBCL who relapsed after hematopoietic stem cell transplantation (HSCT) and were in need of novel therapies. Data from a phase I trial showed an ORR of 57% in 37 subjects and an ORR of 73% in the CD30-positive subgroup (n = 15) [65]. The encouraging results of BV plus Len regimen led to assessments of efficacy in combination with Len and rituximab versus placebo in combination with Len and rituximab in a phase III study (NCT04404283) for patients with R/R DLBCL. Moreover, a combination with immune checkpoint inhibitors could be a feasible option for R/R B-NHL [66]. The BV plus Nivo regimen has shown activity in patients with R/R PMBCL who have received at least two prior therapies, and exhibited advantages over the use of PD-1 inhibitor or BV alone [7]. Efficacy analysis of 30 patients in the phase I/II study (NCT02581631) reported an ORR of 73%, with 37% of patients achieving a CR. Notable treatment-emergent adverse events (TEAEs) included PN (10%) and thrombocytopenia (10%). Three patients discontinued treatment owing to severe PN. The updated data of the trial showed an ORR of 70% (50% CR) and a tolerable safety profile in 10 patients with R/R mediastinal gray zone lymphoma (GZL) receiving BV + Nivo [67]. The time to CR was 1.2 months and 5 patients who achieved a CR were bridged to HSCT. Based on the high CR rate, this regimen may provide an alternate option for bridging to hematopoietic cell transplantation. Furthermore, the BV combined with rituximab, cyclophosphamide, doxorubicin, and prednisone (R-CHP) regimen was also assessed as frontline therapy in a phase I/II multicenter trial (NCT01994850) using 6 cycles of treatment to assess toxicity and efficacy for 29 patients with CD30-positive B cell lymphomas, including 22 PMBCL, 5 DLBCL, and 2 GZL [68]. After systemic treatment was administered, the ORR was 100%, with 86% of patients achieving a CR by the time therapy was completed. The 2-year PFS and OS rates were 85% and 100%, respectively. Subtype analysis revealed that the 2-year PFS rate in patients with PMBCL was 86%. The treatment may provide durable responses for patients with CD30-positive B cell lymphomas and warrants further exploration. Thus, BV-based therapies may be efficacious and safe treatment options for patients at a high risk of relapse.

CD79b is a component of the B-cell receptor (BCR) complex required for proper cellular localization, trafficking, and signal transduction [69]. CD79b is restricted to B cells and highly prevalent in B cell leukemia and lymphoma with minimal expression in normal tissue, appearing ideally suited to serve as the targeting moiety [70].

The Pola structure contains an anti-CD79b mAb site-specifically coupled with MMAE through a Val-Cit linker similar to the structure of the anti-CD22 ADC pinatuzumab vedotin (Pina). The multicenter, phase II ROMULUS study (NCT01691898) illustrated that rituximab plus Pola (R-pola) showed significant benefit over rituximab plus Pina (R-pina) in patients with R/R DLBCL and R/R FL, and both arms compared favorably with other rituximab-based immunochemotherapy arms [71]. Regardless of CD79b or CD22 expression level, R-pola was related to a more favorable outcome in the FL subgroup, including an ORR of 70% (45% CR) and a median PFS of 15.3 months. In DLBCL patients, the R-pola arm showed a comparable ORR (54% vs. 60%) and CR rate (21% vs. 26%) with the control arm. In addition, R-pola was associated with fewer grade 3 or higher side effects in both the DLBCL and FL cohorts.

Currently, there is no standard treatment strategy for some patients with R/R DLBCL who are transplant-ineligible and for those who relapse after HSCT. In a multicenter phase II trial (NCT02257567), patients were randomly assigned to the Pola plus bendamustine and rituximab (pola-BR) arm or bendamustine plus rituximab (BR) arm. The results showed that compared with the BR group, the ORR (45% vs. 18%), CR rate (40% vs. 17.5%), median OS (12.4 vs. 4.7 months), and PFS (9.5 vs. 3.7 months) in the pola-BR arm were significantly improved, and the AEs were acceptable [25]. Another phase Ib/II trial (NCT02611323) investigated the efficacy and tolerability of a novel triplet combination (Pola, venetoclax, and rituximab) in 57 participants with R/R DLBCL [72]. The reported ORR and CR rate were 65% and 31%, respectively, with a median DOR of 5.8 months. Notable AEs causing dose reduction or interruption of any drug occurred in 18% and 61% of patients, respectively. Further evaluation of this combination is warranted due to the limited number of patients recruited. Based on the aforementioned data, Pola is also under exploration in patients with R/R MCL in an ongoing phase II trial (NCT04659044) that is recruiting patients receiving the same combination therapy of Pola, venetoclax, and rituximab.

The benefit of adding Pola in combination with other immunotherapies has been explored in a few studies. A phase Ib/II study (NCT02600897) sought to determine the efficacy and safety profiles of a triplet combination of Pola, obinutuzumab (G), and lenalidomide (Pola-G-Len) in patients with R/R FL [73]. The ORR was 76%, with a CR rate of 65% in the primary efficacy population (n = 46). Of those patients who were refractory to their last treatment, 71% achieved a CR. Grade 3 or worse adverse events included neutropenia (50%), thrombocytopenia (23%), infections (16%), and anemia (14%). TEAEs causing a dose reduction or cycle cessation arose in 19 (34%) and 41 (73%) patients, respectively, and the majority were attributed to intolerability of Len. A more extended period of follow-up, through and beyond maintenance treatment, is ongoing. Despite the desirable antitumor activity of triplet therapy, optimization of doses or prophylactic use of granulocyte colony stimulating factor (G-CSF) is needed to reduce the incidence of AEs.

An open-label, non-randomized phase Ib/II study (NCT01992653) focused on incorporating Pola into the R-CHP or G-CHP regimen as one of the frontline treatments in adult patients with DLBCL, the desirable outcomes of which contributed to FDA approval [74]. Among patients receiving Pola in combination with chemotherapy, an encouraging response with an ORR of 89% (77% CR) was achieved. Instances of severe myelosuppression included neutropenia (30%), febrile neutropenia (18%), and thrombocytopenia (9%). In light of the encouraging data mentioned above, a phase III, double-blind POLARIX trial (NCT03274492) is undergoing to investigate whether Pola combination therapy (Pola-R-CHP) can be clinically advantageous in certain DLBCL subtypes. In addition, Pola-R-CHP has shown a more favorable efficacy than the proven therapy (R-CHOP) as the frontline regimen for patients with previously untreated DLBCL. Other uncompleted phase I/II studies have investigated various types of lymphomas, such as newly diagnosed double- or triple-hit lymphoma (NCT04479267) and untreated aggressive B cell lymphoma (NCT04231877).

CD19 is a 95 kDa glycoprotein that is critically involved in the processes of B cell proliferation, differentiation, activation, and antibody production, and it can also promote BCR signal transduction. As a biomarker, CD19 is prevalently expressed in B cell malignancies and is thought to be the most reliable surface biomarker for B cells [75].

Loncastuximab tesirine (ADCT-402), which was recently approved for use in patients with R/R large B cell lymphoma, comprises a humanized mAb specifically directed to CD19, a PBD dimer, and a cleavable disulfide-bond linker. The results of 183 patients with R/R NHL who received loncastuximab tesirine in the phase I dose-expansion study (NCT02669017) reported an ORR of 46% and a CR rate of 27% with a median DOR of 5.4 months [76]. Among patients in the DLBCL, MCL, and FL subgroups, the ORRs were 42%, 47%, and 79%, respectively. Notably, the ORR was 56% in the subgroup of elderly patients (≥ 75 years old) with DLBCL, indicating encouraging efficacy. The patients tolerated the therapy well because the treatment-related toxicities were largely hematologic and generally manageable with dose delays, followed by fatigue, nausea, edema, and hepatotoxicity, while dose-limiting toxicities (DLTs) were reported in only 4 patients. The rapid onset of response also supported the further development of loncastuximab tesirine when compared with the new combination therapy of tafasitamab plus Len in patients with R/R DLBCL (1.5 vs. 2.0 months) [77]. Given the small number of patients in the FL subgroup (n = 14), larger randomized studies are needed for validation. The updated analysis of data from a phase II trial (NCT03589469) reported an ORR of 48%, a CR rate of 24% and a median DOR of 10.3 months in 145 evaluable patients with R/R DLBCL with a 26% frequency of severe neutropenia and a 17% frequency of grade 3 to 4 glutamyltransferase (GGT) increase [78]. The data also revealed an ORR of 46% among patients who failed prior CD19-directed chimeric antigen receptor T cell (CAR-T) therapy, indicating that these immunotargeted agents may be used consequentially in high-risk patients. A survival advantage was also noted in patients with double- or triple-hit DLBCL, with an ORR of 33% (notably all CRs) and a median DOR of 13.4 months. Other phase I trials have explored loncastuximab tesirine as part of immunochemotherapy, such as with rituximab (NCT04384484) or ibrutinib.

CD22, whose expression is restricted to mature B lineage cells, imposes modulatory effects on diverse signaling pathways that allow appropriate B cell homeostasis, survival, and activation [83]. CD22 is a recycling endocytic receptor that shuttles toxins between the cell surface and endosomes, leading to rapid internalization [84]. The anti-CD22-based treatment already provides a path forward to overcome the poor prognosis of R/R B lineage lymphomas.

Several second-generation mAbs targeting CD20, a broadly explored target in the DLBCL population, have been clinically tested, but no significant benefits have been shown for mAb therapies over the standard R-CHOP regimen [99]. Hence, a strong rationale exists for examining the roles of CD20-ADCs rather than CD20-mAbs in clinical applications.

CD25 is the α subunit of the IL-2 receptor (IL-2Rα) and can increase the affinity of ligand binding and active oncogenic signaling pathways [102]. Indeed, upregulated CD25 expression is related to a more aggressive course and an inferior outcome in DLBCL, FL, HL, and CLL [103‐106]. It has also been shown to promote lymphomagenesis and drug resistance in T cell lymphomas [107].

Anti-CD30-LDM, an innovative ADC with an intact anti-CD30 antibody conjugated to LDM via a non-cleavable linker, presents attractive tumor-targeting capability and antitumor efficacy both in vitro and in vivo [110]. Treatment with anti-CD30-LDM remarkably inhibited tumor growth in mice in a dose-dependent manner without apparent AEs. Anti-CD30-LDM exerts its cytotoxic effects by inducing DNA damage instead of by blocking the polymerization of tubulin, highlighting its potential for overcoming resistance induced by BV. Additionally, anti-CD30-LDM augmented programmed cell death-1 ligand 1 (PD-L1) presentation in cell lines, indicating possible antitumor synergy between anti-CD30-LDM and immunotherapy agents. Consequently, it could be chosen for further testing as a single agent or combination therapy.

CD37 is a member of the tetra-spanning superfamily, which directly mediates survival and apoptotic signaling, and is expressed on normal and malignant mature B cells, similar to CD20. CD37 has been introduced as an appealing target for treating DLBCL because the expression level  of CD37 on neoplastic cells in patients with DLBCL correlates with PFS and OS [111]. To date, the development of anti-CD37 ADCs has shown limited progression due to limited clinical efficacy.

CD70 is a member of the TNF receptor superfamily and acts as a functional receptor binding to soluble CD27. CD70 is expressed on some types of lymphomas and solid tumors, where it correlates with poor prognosis [119, 120]. Three anti-CD70 ADCs had reached phase I clinical development in patients with CD70-positive R/R B-NHL and metastatic renal cell carcinoma (RCC). Unfortunately, none of these ADCs was retained for further evaluation due to a lack of clinical benefit and excess toxicity.

SGN-CD70A, an ADC, contains an antibody against CD70 linked to the PBD dimer via a protease-cleavable linker. A first-in-human phase I study described early-onset severe side effects and undesirable efficacy in patients with R/R CD70-positive NHL, including DLBCL, MCL, and grade 3b FL [121]. The application of SGN-CD70A has been limited due to the occurrence of grade 4 thrombocytopenia, albeit with long-lasting remission. MDX-1203 (BMS-936561) comprises a fully human IgG1 antibody specific for CD70 linked through a protease-cleavable linker attached to a duocarmycin derivative. The dose-escalation phase I trial was discontinued on account of limited antitumor activity for patients who received more than three prior systemic administrations [122]. Vorsetuzumab mafodotin (SGN-75) is a humanized anti-CD70 ADC with an IgG1 mAb, MMAF payload, and a non-cleavable linker. Given its unacceptable toxicity (idiopathic thrombocytopenic purpura), a phase I study in patients with R/R CD70-positive NHL or metastatic RCC was terminated [123]. Considering that alleviating thrombocytopenia is currently not possible, the applicability of CD70-targeting ADCs remains limited.

CD74 is a type II transmembrane glycoprotein that is involved in the presentation of endogenous antigens and the process of immune regulation. The expression of CD74 protein is found in a wide variety of normal tissues and lymphoid malignancies, supporting the application of anti-CD74 antibodies for targeted immunomodulatory therapies [124].

Only 2 CD74-targeting ADCs (STRO-001 and hLL1-DOX) have reached clinical development. STRO-001 is a humanized and glycosylated anti-CD74 ADC that is generated by novel cell-free protein synthesis technology and integrates a non-cleavable linker and maytansinoid payload. It exhibits a high degree of homogeneity and stability because of its site-specific conjugation. Preliminary data from an open-label phase I study (NCT03424603) of STRO-001 in heavily pretreated patients with advanced B cell malignancies provided early signs of efficacy and a good safety profile [125]. The ORR was 25% in 16 evaluable patients with no ocular or neuropathy disorders observed. The study recently began to dose patients at planned levels of 2.5 mg/kg and 3.5 mg/kg. Milatuzumab doxorubicin (hLL1-DOX) was also explored in a phase I/II trial (NCT01585688) of patients with R/R B-NHL or CLL, but unsatisfactory results led to trial discontinuation [126]. SP7676, another anti-CD74 ADC, has a high degree of stability in circulation via site-specific conjugation. It elicits a strong response, as evidenced by 100% of animals achieving complete regression of tumors in lymphoma models, and SP7676 achieved the intended pharmacodynamic effect and produced tumor regression in the "double-hit" lymphoma models [127]. Nevertheless, the development of these drugs is still in the early stage, and further exploration is needed to ascertain their value in clinical applications.

ROR1 is an oncofetal glycoprotein expressed on CLL, MCL, a variety of malignant tumors, and early-stage B cells but is rarely expressed in normal adult tissues. Moreover, a high level of ROR1 expression in cancer cells appears to be involved in the inhibition of apoptosis and predicts an inferior outcome in primary refractory DLBCL, low-grade FL, and Richter´s syndrome [129].

ADCs under preclinical exploration and demonstrating activity in animal models of lymphomas include those targeting several antigens that have not been described in the aforementioned parts (CD38, CD123, CD185, CD205, and HLA-DR) (Table 2).