Targeting cancer with antibody-drug conjugates: Promises and challenges

ABSTRACT

Antibody-drug conjugates (ADCs) are a rapidly expanding class of biotherapeutics that utilize antibodies to selectively deliver cytotoxic drugs to the tumor site. As of May 2021, the U.S. Food and Drug Administration (FDA) has approved ten ADCs, namely Adcetris®, Kadcyla®, Besponsa®, Mylotarg®, Polivy®, Padcev®, Enhertu®, Trodelvy®, Blenrep®, and Zynlonta™ as monotherapy or combinational therapy for breast cancer, urothelial cancer, myeloma, acute leukemia, and lymphoma. In addition, over 80 investigational ADCs are currently being evaluated in approximately 150 active clinical trials. Despite the growing interest in ADCs, challenges remain to expand their therapeutic index (with greater efficacy and less toxicity). Recent advances in the manufacturing technology for the antibody, payload, and linker combined with new bioconjugation platforms and state-of-the-art analytical techniques are helping to shape the future development of ADCs. This review highlights the current status of marketed ADCs and those under clinical investigation with a focus on translational strategies to improve product quality, safety, and efficacy.

Introduction

Antibody-drug conjugates (ADCs) are a rapidly expanding class of anticancer therapeutics, consisting of an antibody attached, via a chemical linker, to a potent cytotoxic agent also named as “payload.” The antibody is designed to target a specific antigen (receptor) that is highly expressed in tumor cells. ADCs deliver a drug with high selectivity to tumors, thereby minimizing their systemic exposure, potentially leading to an improved therapeutic index (greater efficacy and less side effects). The majority of ADCs follow a similar mode of action that involves antibody mediated receptor binding, ADC internalization, and subsequent payload release and execution of cytotoxicity (Figure 1). The success of ADCs relies on several critical factors: 1) target antigens (e.g., CD30, HER2, CD22, CD33 CD79b, Nectin 4, trophoblast-cell surface antigen 2 (Trop2), B-cell maturation antigen (BCMA), CD19), 2) type of antibody (e.g., IgG1, IgG2, IgG4, nanobody, bispecific antibody), 3) type of payload (e.g., monomethyl auristatin E (MMAE), DM4, calicheamicin, DM1, monomethyl auristatin F (MMAF)), 4) type of linker (e.g., valine-citrulline, Sulfo-SPDB, hydrazone linker), 5) conjugation platform (e.g., lysine-, cysteine-, and site-specific conjugation), 6) target indications (e.g., breast cancer, lymphoma, leukemia, urothelial cancer, lung cancer, ovarian cancer). The complexity of ADCs requires state-of-the-art analytical techniques to adequately characterize and control product quality and manufacturing consistency. This review highlights the recent advances in the clinical development of ADCs and the translational strategies associated with ADC manufacture and quality assessment. Strategies to reduce toxicities of ADCs, including dosing regimens and payload-linker optimization, have been extensively discussed in previous reports,^{1, 2} and are not within the scope of this review.

Figure 1.

Cellular Processing of ADCs. Most ADCs undergo similar mechanisms to release the cytotoxic payload. In general, ADCs are designed for internalization and are processed via the endocytic pathway resulting in release of the payload and cytotoxic effect

The clinical pipeline for ADCs

To date, ten ADCs have been approved by the FDA, namely Adcetris®, Kadcyla®, Besponsa®, Mylotarg®, Polivy®, Padcev®, Enhertu®, Trodelvy®, Blenrep®, and Zynlonta™, with exclusively oncology indications (Table 1, Figure 2a). In addition, more than 80 ADCs are currently under active clinical development as monotherapy or combinational therapy for the treatment of various tumor types.

Table 1.

ADCs approved for clinical use

ADC	Target	Antibody	Linker	Payload	Indication	Manufacturer	Approval Year
Adcetris®	CD30	Chimeric IgG1	Valine-citrulline	MMAE	Previously untreated stage III or stage IV classical Hodgkin’s Lymphoma (cHL); relapsed or refractory cHL; cHL after failure of auto-HSCT or failure of at least two prior multi-agent chemotherapy regimens; systemic anaplastic large cell lymphoma, primary cutaneous anaplastic large cell lymphoma other CD30-expressing peripheral T-cell lymphomas	Seattle Genetics (Seagen)	2011
Kadcyla®	HER2	Humanized IgG1	SMCC	DM1	HER2-positive, metastatic breast cancer	Genentech	2013
Besponsa®	CD22	Humanized IgG4	ActBut	Calicheacmicin	Monotherapy in adults with relapsed or refractory B-cell precursor acute lymphoblastic leukemia (ALL)	Pfizer	2017
Mylotarg®	CD33	Humanized IgG4	ActBut	Calicheacmicin	Single-agent and combinational therapy in newly-diagnosed CD33-positive acute myeloid leukemia (AML) in adults and relapsed or refractory CD33-positive AML in adults and pediatric patients (≥2 years).	Pfizer	2000; 2017
Polivy®	CD79b	Humanized IgG1	Valine-citrulline	MMAE	Combinational use with bendamustine and a rituximab product in adult patients with relapsed or refractory diffuse B-cell lymphoma (DBCL)	Genentech	2019
Padcev®	Nectin-4	Human IgG1	Valine-citrulline	MMAE	Adult patients with locally advanced or metastatic urothelial cancer	Astellas Pharma, inc.	2019
Enhertu®	Her2	Humanized IgG1	Tetrapeptide	exatecan-derivative topoisomerase I inhibitor (DXd)	Adult patients with unresectable or metastatic HER2-positive breast cancer	Daiichi Sankyo	2019
Trodelvy®	Trop-2	Humanized IgG1	Hydrolysable CL2A	SN-38 Topo I inhibitor	Adult patients with metastatic triple-negative breast cancer who have received at least two prior therapies for metastatic disease.	Immunomedics	2020
Blenrep®	BCMA	Humanized IgG1	maleimidocaproyl	MMAF	Adult patients with relapsed or refractory multiple myeloma who have received at least 4 prior selected therapies	GSK	2020
Zynlonta™	CD19	Humanized IgG1	Valine-alanine	SG3249 PBD dimer	adult patients with relapsed or refractory large B-cell lymphoma after two or more lines of systemic therapy, including DLBCL not otherwise specified, DLBCL arising from low grade lymphoma, and high-grade B-cell lymphoma	ADC Therapeutics	2021

Each ADC listed has been approved for treatment of oncological indications in the clinical setting.

4-(4′-acetylphenoxy) butanoic acid (AcBut); 7-ethyl-10-hydroxycamptothecin (SN-38); Antibody–drug conjugate (ADC); B-cell maturation antigen (BCMA); classical Hodgkin’s lymphoma (cHL); GlaxoSmithKline (GSK); Mertansine (DM1); Monomethyl auristatin E (MMAE); Monomethyl auristatin F (MMAF); Pyrrolobenzodiazepine (PBD); Succinimidyl 4-(N-maleimidomethyl) cyclohexane-1-carboxylate (SMCC); Trophoblast-cell surface antigen 2 (Trop2)

Figure 2.

Structure of ADCs Approved for Clinical Use. Design of each approved ADC, highlighting antibody isotype, linker chemistry, payload class and DAR are provided. (A) ADCs approved for hematological malignancies include Adcetris®, Polivy®, Mylotarg®, Besponsa®, Blenrep®, and Zynlonta™. (B)ADCs approved for solid tumors include Kadcyla®, Padcev®, Enhertu® and Trodelvy®

Figure 2.

Continued

FDA Approved ADCs

Of the ten ADCs approved for clinical use (Table 1), six are indicated for treatment of hematological malignancies. Brentuximab vedotin (Adcetris®) is an ADC produced by Seattle Genetics (now known as Seagen). The anti-CD30 (cAC10) ADC consists of ~4 MMAE molecules conjugated through cysteines of reduced interchain disulfide bonds via a protease-cleavable linker (Figure 2a). ³ Brentuximab vedotin was granted accelerated approval in 2011 and full approval in 2015 for the treatment of classical Hodgkin’s lymphoma, systemic anaplastic large cell lymphoma, and peripheral T-cell lymphoma.

In 2017, inotuzumab ozogamicin (Besponsa®), a Pfizer product, was approved for treatment in adults with relapsed or refractory (R/R) B-cell precursor acute lymphoblastic leukemia. The ADC targets the CD22 surface marker using an IgG4 to deliver approximately 6 calicheamicin molecules. The payload is conjugated via surface-exposed lysines using an acid-labile linker.⁴ This design nearly mirrors that of gemtuzumab ozogamicin (Mylotarg®), an anti-CD33 ADC with an average of 2–3 calicheamicin payloads conjugated to the antibody. Gemtuzumab ozogamicin was the first ADC to receive accelerated approval in 2000, contingent on fulfilling the post-marketing requirement of a randomized trial to confirm clinical benefit (S0106; NCT00085709) (Figure 2a).^5,6 However, the trial did not confirm the clinical benefit but instead raised safety concerns due to an increase in treatment-related fatalities compared to the control group receiving standard chemotherapy. The leading causes of fatality in the treatment arm were associated with infection and hemorrhage. Ultimately, the results of the trial led to the voluntary withdrawal of the application by Pfizer in 2010.^7,8 Following modifications to the dosing schedule, which was associated with decreased incidence of hepatotoxicity and early mortality, and to address the critical unmet need for acute myeloid leukemia patients, gemtuzumab ozogamicin was granted approval again in 2017.⁹

Between 2019 and 2021, the FDA granted accelerated approval of polatuzumab vedotin (Polivy®) for R/R diffuse large B-cell lymphoma (DLBCL), belantamab mafodotin (Blenrep®) for R/R multiple myeloma, and loncastuximab tesirine (Zynlonta™) for R/R B-cell lymphoma, respectively. Polatuzumab vedotin is an ADC produced by Genentech with an average of 3.5 MMAE molecules conjugated to cysteines of reduced interchain disulfide bonds on an anti-CD79b antibody (Figure 2a).¹⁰ Belantamab mafodotin is a first-in-class anti-BCMA ADC produced by Astellas Pharma, Inc. The ADC carries approximately 4 MMAF molecules conjugated via a non-cleavable linker to the cysteines of the afucosylated anti-BCMA antibody (Figure 2b). Belantamab mafodotin exhibits versatile mechanisms of action (MOA), including inducing cell death by delivering the MMAF molecules to the target cell and inducing both antibody-dependent cellular cytotoxicity (ADCC) and antibody-dependent cellular phagocytosis (ADCP).¹¹ Loncastuximab tesirine, or Lonca-T, is an anti-CD19 ADC manufactured by ADC Therapeutics for R/R B-cell lymphomas, including DLBCL.¹² The ADC is the latest to be approved as of May 2021 and is the first to carry the pyrrolobenzodiazepine (PBD) dimer toxin (indicated as SG3249). Approximately 2.3 SG3249 molecules are attached to the antibody via a cathepsin-cleavable valine-alanine linker, facilitating DNA minor groove interstrand crosslinking of target cells following payload release (Figure 2a).

The remaining four ADCs are approved for treatment of solid tumors, including three for breast cancers and one for urothelial cancer. Ado-trastuzumab emtansine (Kadcyla®), also notated as T-DM1, is a conjugation of ~3.5 maytansinoid DM1 molecules to the anti-HER2 antibody trastuzumab via surface-exposed lysines (Figure 2b).¹³ T-DM1, produced by Genentech, received FDA approval in 2013 for the treatment of HER-2 positive metastatic breast cancer (mBC), with additional approved uses including monotherapy and combinational administration as well as an adjuvant treatment for early breast cancer. Similar to belantamab mafodotin, T-DM1 induces cell death by release of the payload and ADCC activity retained in the parent mAb that inhibits HER2-signaling.¹⁴ Despite such anti-tumor activity, resistance to T-DM1 remains a challenge and will be discussed further in later sections.¹⁵

To address this barrier, trastuzumab deruxtecan (Enhertu®), an anti-HER2 ADC with several unique properties, was approved in 2019. Produced by Daiichi Sankyo for the treatment of HER2-positive metastatic breast cancer following a prior trastuzumab-based regimen, trastuzumab deruxtecan (T-Dxd) uses the same parent IgG1 antibody as T-DM1, but it is conjugated to approximately 8 molecules of an exatecan-derivative topoisomerase I inhibitor, Dxd, via a protease-cleavable tetrapeptide linker (Figure 2b).¹⁶ Release of the payload and influx into neighboring tumor cells exerts anti-tumor activity in heterogenous cell populations with varying levels of HER2 expression (high or low).¹⁷

Another ADC recently approved for breast cancer is sacituzumab govitecan (Trodelvy®), an anti-Trop2 ADC produced by Immunomedics.¹⁸ Sacituzumab govitecan is a first-in-class ADC for the treatment of metastatic triple-negative breast cancer (TNBC) in patients who have received two prior treatments for metastatic disease including chemotherapy, targeted, or immunotherapy. Sacituzumab govitecan is another example of an ADC product with a high drug-to-antibody ratio (DAR), consisting of ~7.6 SN-38 molecules, a moderately toxic topoisomerase I inhibitor, using a novel, hydrolysable linker called CL2A through cysteines (Figure 2b).¹⁸ SN-38 is the active drug form of the clinically used anticancer agent, CPT-11 or irinotecan. Interestingly, SN-38 was found to be more potent than CPT-11, but less potent than cytotoxic agents conventionally used in ADCs, including calicheamicin and MMAE derivatives.¹⁹ The use of moderately toxic payloads is being investigated as a method to increase payload concentration and overcome the challenges of stability and efficacy with higher DAR ADCs.

Of the four ADCs approved for solid tumors, enfortumab vedotin (Padcev®) is the only product approved for a solid tumor aside from breast cancer. Produced and marketed through a collaboration between Astellas Pharma Inc. and Seagen, enfortumab vedotin is a first-in-class therapeutic indicated for the treatment of Nectin-4 positive urothelial cancer.²⁰ The ADC consists of a human IgG1 against nectin-4, a member of the nectin family of immunoglobulin-like adhesion molecules known to mediate Ca⁺-independent cell–cell adhesion through the recruitment of cadherins and modulation of cytoskeletal arrangements (Figure 2b).²¹ Approximately 3.8 MMAE molecules are conjugated through cysteines via the same cleavable linker technology previously used in other ADCs produced by Seagen.²⁰

Novel ADCs in clinical trials

More than 80 ADCs are currently in active clinical trials, with a majority in phase I and I/II (Table 2, Figure 3a). Over 80% of the clinical trials are investigating ADC safety and efficacy in various solid tumors, while the remaining trials involve hematological malignancies (Figure 3b). This may suggest a shift in recent years toward investigational ADCs for solid tumors following the earlier success of T-DM1 and recent approvals of T-Dxd, sacituzumab govitecan, and enfortumab vedotin. Of this list, there are approximately 40 different targets with several ADCs against the same target (Table 2, Figure 3c).

Table 2.

ADCs currently under clinical investigation

Cytotoxic Payload	ADC	Target	Conjugation	Phase	Conditions	Clinical Trial	Reference
Tubulin disruptor/ anti-mitotic	ASN004	5T4	Cysteine	I	Advanced solid tumors	NCT04410224	22
	IMGC936	ADAM9	Site-specific	I	Advanced solid tumors	NCT04622774	23
	ABGN-107	AG7	Undisclosed	I	Gastric, Colorectal, Pancreatic or Biliary Cancer	NCT02908451	24
	AGS-16C3F	ENPP3	Cysteine	II	Metastatic Renal Cell Carcinoma	NCT02639182	25
	HUMAX-AXL-ADC	AXL	Cysteine	I/II	Ovarian Cancer, Cervical Cancer, Endometrial Cancer, Non-Small Cell Lung Cancer (NSCLC), Thyroid Cancer. and Melanoma Sarcoma	NCT02988817	26
	BA3011	AXL	Undisclosed	II I/II	NSCLC Solid tumors	NCT04681131 NCT03425279	27
	CX-2009	CD166	Undisclosed	II	Advanced Breast Cancer	NCT03149549 NCT04596150
	OBT076	CD205	Cysteine	I	Breast Cancer	NCT04064359	28
	TRPH-222	CD22	Site-specific	I	R/R B-Cell Lymphoma	NCT03682796	29
	SGN-CD228A	CD228	Cysteine	I	Advanced solid tumors	NCT04042480	30
	F0002-ADC	CD30	Lysine	I	R/R hematologic malignancies	NCT03894150	31
	Debio 1562	CD37	Lysine	II	R/R Diffused large B-cell lymphoma (DLBCL) and other forms of non-Hodgkin lymphoma	NCT02564744	32
	STI-6129	CD38	Site-specific	I	R/R Systemic AL Amyloidosis	NCT04316442
	FOR46	CD46	Cysteine	I I	R/R Multiple myeloma (MM) Metastatic castration-resistant prostate cancer	NCT03650491 NCT03575819	33
	IMGN-901	CD56	Lysine	II	R/R Wilms tumor, rhabdomyosarcoma, neuroblastoma, pleuropulmonary blastoma, malignant peripheral nerve sheath tumor, or synovial sarcoma	NCT02452554	34
	CX-2029	CD71	Undisclosed	I/II	Solid tumors or DLBCL	NCT03543813	35
	STRO-001	CD74	Site-specific	I	Advanced B-cell malignancies	NCT03424603	36
	SAR408701	CEACAM5	Lysine	II II II I III I/II	Advanced solid tumors Non-squamous NSCLC	NCT04659603 NCT04524689 NCT04394624 NCT03324113 NCT04154956 NCT02187848	37
	ABBV-399	c-Met	Cysteine	II I II	Advanced solid tumors Non-squamous NSCLC	NCT03574753 NCT02099058 NCT03539536	38
	RC108	c-Met	Undisclosed	I	Advanced malignant solid tumors	NCT04617314
	ABT-414	EGFR	Cysteine	II/III	Glioblastoma	NCT02573324	39
	MRG003	EGFR	Undisclosed	II II II	Recurrent or metastatic squamous cell carcinoma of head and neck, unresectable, locally advanced or metastatic biliary tract cancer, and advanced NSCLC	NCT04838548 NCT04838964 NCT04868162	40
	STRO-002	FolRα	Site-specific	I	Ovarian and endometrial cancers	NCT03748186	41
	MORAB-202	FolRα	Cysteine	I/II I	Solid tumors	NCT04300556 NCT03386942	42
	IMGN853	FolRα	Lysine	II II III II I III I II I/II	Endometrial, epithelial ovarian, fallopian tube, primary peritoneal, and triple negative breast cancers	NCT03832361 NCT03835819 NCT04296890 NCT04274426 NCT03552471 NCT04209855 NCT02996825 NCT04606914 NCT02606305	43
	CDX-011	gpNMB	Cysteine	II	Recurrent or refractory osteosarcoma	NCT02487979	44
	OBI-999	Globo H	Site-specific	I/II	Advanced solid tumor	NCT04084366	45
	PF-06804103	HER2	Site-specific	I	Solid tumors	NCT03284723	46
	ZW49	HER2	Undisclosed	I	HER2-expressing tumors	NCT03821233	47
	RC48	HER2	Cysteine	II I I/II II II I/II II III II I/II III	Metastatic breast, gastric, biliary tract, and urothelial cancers	NCT04329429 NCT04280341 NCT04311034 NCT03809013 NCT04073602 NCT04264936 NCT03556345 NCT04400695 NCT03500380 NCT03052634 NCT04714190	48,49
	ALT-P7	HER2	Site-specific	I	Breast cancer	NCT03281824	50
	ARX788	HER2	Site-specific	I II	Breast and stomach neoplasms	NCT03255070 NCT04829604	51
	FS-1502	HER2	Undisclosed	I	Advanced solid tumors and metastatic breast cancer	NCT03944499
	A166	HER2	Site-specific	I/II	R/R HER2-expressing cancers	NCT03602079	50
	MRG002	HER2	Undisclosed	I/II II	Advanced solid tumors, metastatic gastric/gastroesophageal junction cancer, and advanced metastatic breast cancer	NCT04492488 NCT04742153	52
	BAT8001	HER2	Undisclosed	I/II I III	Advanced breast cancer	NCT04151329 NCT04189211 NCT04185649	53
	W0101	IGF-1 R	Cysteine	I/II	Advanced or metastatic solid tumors	NCT03316638	54
	SGN-B6A	integrin-beta6	Undisclosed	I	Advanced solid tumors	NCT04389632	55
	SGN-LIV1A	LIV-1	Cysteine	I/II I II I I	Advanced or metastatic triple negative breast cancer	NCT03310957 NCT01969643 NCT04032704 NCT03424005 NCT01042379	56
	BAY 94–9343	Mesothelin	Lysine	II I I/II I/II I/II	R/R ovarian, fallopian tube, or primary peritoneal cancers, advanced pancreatic cancer, and pleural mesothelioma	NCT03926143 NCT03102320 NCT03126630 NCT03587311 NCT03816358	57
	BMS-986148	Mesothelin	Undisclosed	I/II	Advanced solid tumors	NCT02341625	58
	RC88	Mesothelin	Undisclosed	I	Advanced solid tumors	NCT04175847
	XMT-1536	NaPi2b	Cysteine	I I/II	Ovarian cancer and NSCLC	NCT03319628 NCT04396340	59
	XMT-1592	NaPi2b	Site-specific	I/II	Ovarian cancer and NSCLC	NCT04396340	60
	ARX517	PSMA	Site-specific	I	Advanced solid tumors	NCT04662580
	VLS-101	ROR1	Lysine	II I	Solid tumors and hematological malignancies	NCT04504916 NCT03833180	61
	SGN-STV	STn	Undisclosed	I	Advanced solid tumors	NCT04665921
	HUMAX®-TF-ADC	TF	Cysteine	II I/II	Cervical cancer	NCT03438396 NCT03786081	62
	JS108	Trop2	Undisclosed	I	Advanced solid tumors	NCT04601285
DNA Damaging	SYD1875	5T4	Site-specific	I	Solid tumors	NCT04202705
	MEDI2228	BCMA	Site-specific	I	R/R MM	NCT03489525	63
	IMGN632	CD123	Site-specific	I/II I/II	Acute lymphocytic leukemia, blastic plasmacytoid dendritic cell neoplasm, myeloproliferative neoplasm, and acute myeloid leukemia	NCT03386513 NCT04086264	64
	ADCT-602	CD22	Site-specific	I/II	R/R B-cell acute lymphoblastic lymphoma	NCT03698552	65
	ADCT-301	CD25	Cysteine	II I II II	Acute myeloid lymphoma, myelodysplastic syndrome, myeloproliferative neoplasm, R/R Hodgkin lymphoma, and R/R DLBCL	NCT04639024 NCT03621982 NCT04052997 NCT03589469	66
	MGC018	CD276	Cysteine	I/II	Advanced solid tumors	NCT03729596	67
	TR1801	c-Met	Site-specific	I	Solid tumors	NCT03859752	68
	ABBV-321	EGFR	Site-specific	I	Advanced solid tumors	NCT03234712	69
	SYD985	HER2	Cysteine	I II I III I	Metastatic breast cancer and endometrial carcinoma	NCT04602117 NCT04205630 NCT04235101 NCT03262935 NCT01042379	70
	NBE-002	ROR1	Site-specific	I/II	Advanced solid tumors	NCT04441099	71
Topo I	DS-7300a	B7-H3	Cysteine	I/II	Advanced solid tumors	NCT04145622
	DS-6157a	GPR20	Cysteine	I	Gastrointestinal stromal tumors	NCT04276415	72
	U3-1402	HER3	Cysteine	II II I I/II II I	Metastatic breast, colorectal, and non-small cell lung cancers	NCT04699630 NCT04479436 NCT03260491 NCT02980341 NCT04619004 NCT04676477	73
	DS-1062	TROP2	Cysteine	II I I III I I/II	NSCLC (advanced or metastatic) and triple negative breast cancer	NCT04484142 NCT04612751 NCT04526691 NCT04656652 NCT03401385 NCT03742102	74
	DS-6000	CDH6	Undisclosed	I/II	Renal cell carcinoma and ovarian cancers	NCT04707248
	SKB264	TROP2	Site-specific	I/II	Advanced or metastatic solid tumors	NCT04152499	75
RNA pol II	HDP-101	BCMA	Site-specific	I/II	R/R MM	NCT04879043	76
TLR agonists	BDC-1001	HER2	Undisclosed	I/II	HER-2 expressing advanced malignancies	NCT04278144
	SBT6050	HER2	Undisclosed	I	Solid tumors	NCT04460456	77
BCL2 family protein inhibitor	ABBV-155	CD276	Cysteine	I	R/R solid tumors	NCT03595059	78
Undisclosed	CC-99712	BCMA	Undisclosed	I	R/R MM	NCT04036461
	JBH492	CCR7	Undisclosed	I	Chronic lymphocytic leukemia and non-Hodgkin lymphoma	NCT04240704
	M1231	EGFR/MUC1	Undisclosed	I	Solid tumors, metastatic NSCLC, and esophageal squamous cell carcinoma	NCT04695847
	B003	HER2	Undisclosed	I	Metastatic breast cancers	NCT03953833
	BB-1701	HER2	Undisclosed	I	Locally advanced/metastatic solid tumors	NCT04257110
	DP303c	HER2	Undisclosed	II II I	Advanced ovarian and gastric cancers and solid tumors	NCT04828616 NCT04826107 NCT04146610
	GQ1001	HER2	Site-specific	I	Advanced solid tumors	NCT04450732
	SHR-A1811	HER2	Undisclosed	I I/II I	Advanced gastric or gastroesophageal junction adenocarcinoma, advanced NSCLC and colorectal cancer	NCT04513223 NCT04818333 NCT04446260
	ARX517	PSMA	Site-specific	I	Advanced solid tumors	NCT04662580
	BA3021	ROR2	Undisclosed	I/II	Solid tumors	NCT03504488
	MRG004A	Tissue factor	Undisclosed	I/II	Advanced or metastatic solid tumors	NCT04843709
	ABBV-011	Undisclosed	Undisclosed	I	R/R Small cell lung cancer	NCT03639194
	SHR-A1904	Undisclosed	Undisclosed	I	Advanced solid tumors	NCT04877717

Each ADC listed is currently under investigation in one or more active clinical trials as of 15 May 2021. The ADCs listed are all registered with clinicaltrials.gov with phase 1-3 trials of “Not yet recruiting”, “Recruiting”, “enrolling by invitation”, and “Active, not recruiting” status investigating use in cancer indications. ADCs marketed for clinical use and developmental ADCs with trials that have been terminated, withdrawn, completed, or are of unknown status were excluded from the table. Disclosed information regarding target, payload action, and conjugation technique are provided or otherwise noted as “Undisclosed”. Data shown was derived from the U.S. National Library of Medicine ClinicalTrials.gov (access date 15 May 2021, search terms of “antibody drug conjugate” and “cancer”)).

Diffuse large B-cell lymphoma (DLBCL); Multiple myeloma (MM); Non-small cell lung cancer (NSCLC); Relapsed and/or refractory (R/R)

Figure 3.

Novel ADCs in Clinical Trials for Oncology. There are currently 82 novel ADCs in 150 active clinical trials registered with clinicaltrials.gov for cancer patients. (a) Most of the ADCs are currently under investigation in phase 1 trials, while a small percentage has advanced to phase 3. (b) Of the 150 ongoing trials, more than 80% are evaluating ADC safety and efficacy in solid tumors whereas less than 20% are trials for hematological malignancies. (c) There are 43 disclosed targets organized here by the number of ADCs designed to recognize them. Most of these targets are under evaluation by a single ADC, while some are being investigated by several different ADCs. (d) Of the 82 novel ADCs, most employ tubulin disrupting payloads, followed by DNA-damaging molecules, topoisomerase I inhibitors, and finally unique payloads such as TLR agonists, a BCL2-xL inhibitor, and an RNA polymerase II inhibitor. Many payloads remain undisclosed. (e) Most ADCs under clinical investigation either utilize the conventional cysteine conjugation strategy or site-specific conjugation platforms while few conjugate to surface lysines. Many techniques remain undisclosed. Data shown was derived from the U.S. National Library of Medicine ClinicalTrials.gov (access date 15 May 2021, search terms of “antibody drug conjugate” and “cancer”)

HER2 is currently one of the most attractive targets for ADC development, with three anti-HER2 ADCs currently in phase III trials. One such anti-HER2 ADC is RC48, produced by RemeGen, joining an IgG1 anti-HER2 antibody, hertuzumab, to approximately four MMAE molecules via a protease-cleavable valine-citrulline linker through cysteine conjugation.⁴⁸ In preclinical studies, RC48 demonstrated antitumor activity at lower doses in trastuzumab and lapatinib sensitive and resistant xenograft models. Superior inhibition was also observed when compared to T-DM1.⁴⁸ Early clinical studies have shown a manageable safety profile in multiple phase I trials for HER2-positive malignancies. Notably, RC48 advanced as a potential therapeutic for the treatment of metastatic or unresectable urothelial carcinoma, demonstrating promising results in a phase II pivotal trial (NCT03507166), including an overall response rate (ORR) of 51.2% in pretreated HER-2 positive locally advanced or metastatic urothelial carcinoma.⁴⁹

Other similarities among ADCs under clinical evaluation include the class of payload. Most use a payload to induce tubulin disruption and mitotic arrest, while a small number cause DNA damage (Figure 3d). Topoisomerase inhibitors, as seen in approved ADCs such as trastuzumab deruxtecan and sacituzumab govitecan, have also begun to appear more frequently in clinical trials, though they still represent a very small percentage. A number of novel payloads are emerging that target specific proteins or receptors. For example, several ADCs, such as the anti-HER2 immune-stimulating antibody conjugate BDC-1001, use a toll-like receptor 7/8 agonist as a payload to elicit immune-mediated tumor efficacy.⁷⁹ BDC-1001 may activate human myeloid antigen-presenting cells within the tumor environment in addition to inducing ADCC and ADCP functions. Currently, BDC-1001 is being investigated in a phase I/II clinical trial for HER2-expressing solid tumors (NCT04278144). Another novel payload class targets the BCL-xL anti-apoptotic protein. ABBV-155 (mirzotamab clezutoclax) is the sole ADC under investigation that uses this class of payload and is designed to target tumors expressing CD276.⁷⁸ ABBV-155 is being evaluated in a phase I trial for R/R solid tumors alone or in combination with taxane therapy (NCT03595059). RNA polymerase II inhibitors, such as amanitin derivatives, can halt cellular transcription processes and protein synthesis, resulting in apoptosis and cell death. HDP-101 is a BCMA-targeting ADC utilizing this derivative under clinical evaluation in a phase I/II trial in R/R multiple myeloma patients (NCT04879043).⁷⁶

Conjugation methodology can directly affect the quality of the ADC, and subsequently the safety and efficacy profiles of the product. There are three main methods of conjugation, including through cysteines of reduced interchain disulfide bonds, surface-exposed lysines, and site-specific techniques. Of the investigational ADCs in active clinical trials, most are manufactured via conventional cysteine conjugation or proprietary site-specific technology licensed by the manufacturers. Only a small portion of ADCs in development use the conventional lysine conjugation methodology, likely due to the vast heterogeneity that can result, as will be discussed in later sections (Figure 3e).

As site-specific conjugation technology can vary among developers, it is worth mentioning several ADCs produced via unique platforms. TRPH-222 is an anti-CD22 ADC conjugated to a maytansinoid payload using the SMARTag™ (Specific Modifiable Aldehyde Recombinant Tag) technology. This platform uses the chemoenzymatic method to engineer a reactive aldehyde (formylglycine) into the mAb for aldehyde-specific conjugation, herein resulting in a controlled maximum DAR of 2.²⁹ TRPH-222 is currently in a phase I trial for R/R B-cell lymphoma, though early results have demonstrated this ADC to be well tolerated (NCT03682796). XMT-1592 is an anti-NaPi2b ADC that is currently under investigation in a phase I/II study for NaPi2b-expressing tumors (NCT04396340). The ADC is produced by Mersana Therapeutics using the Dolasynthen platform that targets the glycan-remodeled Asn297 for site-specific conjugation. The payload auristatin F-hydroxypropylamide (AF-HPA) is membrane-permeable and capable of bystander killing, but it is further metabolized to the membrane-impermeable auristatin F (AF), locking the payload molecules within the cell to achieve “controlled bystander effect” (termed DolaLock).⁸⁰ Preclinical data showed time-dependent accumulation of both AF-HPA and AF in cultured cancer cell lines and in xenograft tumors.^81,82

Several novel antibody platforms are being applied to ADC development strategies. Variations in antibody size, such as the scFv-Fc format used in the ASN004 ADC, could demonstrate an advantage in permeability of solid tumors.²² Two PROBODY drug conjugates (PDCs) are under investigation for tumors expressing CD71 and CD166. Both surface markers are highly expressed in tumor tissues, while also ubiquitously expressed in normal tissues as well. PDCs are masked conjugates that restrict normal tissue recognition and are unmasked by tumor proteases, thereby restricting on-target toxicity outside the tumor site.^83,84 CX-2029, an anti-CD71 PDC, is currently under evaluation in a phase I/II trial for solid tumors or DLBCL (NCT03543813). CX-2009 is an anti-CD166 PDC in a phase I/II trial for unresectable solid tumors (NCT03149549) and a phase II trial to assess activity as a monotherapy or combinational therapy in TNBC (NCT04596150).

Challenges in the development of ADCs

Despite the growing number of ADC approvals, challenges remain in the development of ADCs that demonstrate both superior safety and efficacy in the clinic. One unexpected challenge many developers face during clinical evaluation is the inability to demonstrate benefits over the control arm, such as occurred with MM-302. MM-302 was an anti-HER2 mAb conjugated to liposomal doxorubicin.⁸⁵ The phase II HERMIONE trial (NCT02213744) was designed to determine the benefit of MM-302 treatment with trastuzumab compared to standard care chemotherapy as either gemcitabine, capecitabine, or vinorelbine in HER2-positive locally advanced mBC.⁸⁶ However, the study was terminated due to lack of benefit over comparator treatments. Another ADC to report similar circumstances was AbbVie’s rovalpituzumab tesirine (Rova-T), which targeted cancer-stem cell-associated delta-like protein 3 (DLL3).⁸⁷ Rova-T consisted of an IgG1 anti-DLL3 mAb conjugated to two PBD dimers via a valine-citrulline dipeptide linker. The ADC was intended to treat small cell lung cancer, which is known to overexpress DLL3 in 80% of small cell lung cancer patients with no expression on normal tissues.⁸⁷ Encouraging results in the phase I trial reported 18% ORR in assessable patients and a 38% ORR in patients with high DLL3 expression (NCT01901653), but safety and efficacy concerns were raised due to the results of the phase II trial TRINITY (NCT02674568) in which the primary endpoint was not achieved and high toxicity rates were reported. The most frequent event among patients was pleural effusion, which is considered to be a toxicity associated with PBD dimers.^88,89 Ultimately, the results of the phase III trials, TAHOE (NCT03061812) and MERU (NCT03033511), in which a lack of survival benefit over the control arm was observed, led to the complete discontinuation of the development of Rova-T by AbbVie.⁹⁰

Further challenges in the development of ADCs as therapeutic agents involve toxicities that can be attributed to constituents of the ADC product. Such events have been investigated, mainly focusing on different adverse effects that can be attributed to specific payloads. For example, use of the calicheamicin payload has been associated with increased incidences of liver injury and hepatotoxicity.^7,91 Specifically, increased incidences of veno-occlusive disease, also referred to as sinusoidal obstruction syndrome, and drug-induced liver injury were observed during clinical trials and post-approval use of gemtuzumab ozogamicin, despite dose reduction efforts that led to its re-approval in 2017. Similar occurrences have also been observed with the use of inotuzumab ozogamicin.^92,93 A comprehensive review published in 2016 summarized key clinical toxicities of other approved and developmental ADCs.¹ In general, their findings showed peripheral neuropathy and neutropenia induced by MMAE, which is consistent with adverse events listed for those approved ADCs carrying an MMAE payload. MMAF was associated with ocular toxicities, which is listed as a precaution for the administration of belantamab mafodotin. Differences in clinical toxicities observed between payloads of the same class, e.g., MMAE and MMAF, may indicate linker-associated contributions to these events. Incidences of neutropenia and gastrointestinal system effects have been observed with some ADCs carrying DM1, including T-DM1 and IMGN-901, with increased levels of liver enzymes occurring in some patients administered T-DM1.^1,34 Neutropenia may be a common event among ADCs carrying a topoisomerase I inhibitor, as is observed with trastuzumab deruxtecan, sacituzumab govitecan, and even some developmental ADCs such as U3-1402.^1,94 Myelosuppression, effusion, and inflammation were observed with ADCs carrying PBD dimers, such as Lonca-T and the previous Rova-T and may require concomitant meditation to reduce the incidence of side effects.^69,95 As new payloads continue to emerge in clinical development, clinical data are awaited to understand safety profiles specific to those agents. Understanding events that may be associated with specific payloads can not only aid developers in ADC design but also spur the development of more novel payloads that are less likely to induce harmful events in patients.

Over time, tumors can develop mechanisms to overcome drug efficacy, thereby limiting the success of the treatment. As ADCs are multifunctional therapeutics, some pathways of resistance can develop against individual components of the ADC (Figure 4). One mechanism of resistance could emerge from modulations in antigen recognition by the antibody. This could result from downregulation of the target from the cell surface, rendering the ADC relatively unable to exert their cytotoxic effect.⁹⁶ Several preclinical studies have generated models of acquired resistance, in which cells consistently treated with the ADC over time eventually showed decreases in target antigen protein expression along with other effects.^97,98 In this regard, novel formats of mAbs that can be incorporated into ADCs, such as bispecific or biparatopic mAbs that target two different antigens or nonoverlapping epitopes on the same target antigen, respectively, could aid in overcoming antigen-specific mechanisms of resistance. Li et al. synthesized a biparatopic anti-HER2 ADC conjugated to tubulysin. Preclinical data indicated its ability to restrict tumor growth in four T-DM1-resistant cell lines, though it is not clear whether this is entirely due to the antibody format, or if the novel linker and payloads included in the ADC design contributed as well.⁹⁹ ZW49, a new biparatopic anti-HER2 ADC, is currently undergoing evaluation in a phase I clinical study (NCT03821233). M1231, a bispecific anti-EGFR/MUC1 ADC is also in a phase I trial (NCT04695847).

Figure 4.

Mechanisms of ADC Resistance. Like other therapeutics, tumor cells may develop resistance against ADCs. (1) One mechanism, common to ADCs and monoclonal antibody therapeutics is a reduction in antigen binding, most notably by decreased antigen expression. (2) Most ADCs are internalized following antigen binding, however if internalization of the antigen-ADC complex is impaired, efficacy of the ADC can be reduced. (3) Following internalization, the antibody of the ADC is degraded leaving only the payload (cleavable linkers) or a linker-payload complex (non-cleavable linkers). Defects in the lysosomal degradation process can prevent release of the payload. (4) A common mechanism of resistance to ADCs is the elimination of the payload via drug transporters prior to payload-induced cytotoxic effect. Many traditional payloads of ADCs are substrates of these transporters. (5) Alterations to payload-specific cytotoxicity or cell death pathways can prevent eradication of the tumor cell

Another common mechanism of drug resistance is the removal of the payload via ATP-binding cassette transporters.¹⁰⁰ Many of the cytotoxic warheads used in ADCs may be substrates for these pumps, which can cause drug efflux out of the target cell and a reduction in drug efficacy.¹⁰¹ Clinical data have demonstrated that efflux pumps contribute to the reduced efficacy of gemtuzumab ozogamicin.^102,103 For instance, calicheamicin has been shown to be a substrate of multi-drug resistance mutation 1 (MDR1), and MDR1 expression and activity has been associated with response to gemtuzumab ozogamicin with similar preclinical results observed for inotuzumab ozogamicin.^102–104 Increased drug transporter protein expression has also been observed in T-DM1 resistant cells in addition to decreases in surface antigen expression.¹⁰⁵ Thus, the ability of the small molecule to bypass efflux pump-mediated drug resistance should be considered during the selection of the cytotoxic payload. Other mechanisms of drug resistance may be influenced by any of the several steps involved in the ADC MOA: 1) defects in internalization, trafficking, and recycling, 2) lysosomal degradation leading to impairment of drug release, or 3) alterations in cell death pathways (Figure 4).¹⁰⁰ Preclinical evaluation of these and other potential mechanisms is critical to optimizing ADC development and improving clinical benefit. More attention should be paid to the technical considerations involving ADC design that ultimately influence cellular uptake and processing of the ADC.

Key considerations for ADC design

ADCs use three components to achieve greater clinical benefit, i.e., the mAb, the cytotoxic payload, and a chemical linker. By combining a targeting molecule with a cytotoxic payload, conceivably, the therapeutic window of ADCs is wider than treatment with small-molecule drugs alone. Several reviews have highlighted the complexity of ADCs and the challenges in developing products with improved therapeutic index.^106–109 Optimization of ADC design is critical and requires a more mechanistic understanding of the ADC and its components to heighten the clinical benefit of ADCs.

1. Target antigen

Improvement of ADC safety and efficacy profiles relies significantly on selection of the target antigen and its interaction with the mAb of the ADC. Two critical parameters involved in the selection of the target antigen are tumor specificity and expression level.^110,111 Ideally, the chosen target will exhibit a high level of tumor-specific or disease-specific expression and be minimal to absent in normal tissues. Specificity of the target is critical to reducing toxicity of ADCs, and thus plays a substantial role in their overall success. For oncological indications, the antigen can be expressed as a surface receptor on tumor cells, tumor stem cells, or within the tumor vasculature and microenvironment.^112,113 In best cases, the antigen will also be expressed homogenously across tumor cells at similar levels.¹¹¹ ADCs with sufficient control of bystander effect may overcome the challenge of heterogenous cell populations within a tumor.

2. Monoclonal antibody

After selecting a target, mAbs are produced and screened based on selectivity, tumor penetrating ability, and isotype.¹¹¹ ADCs, both in development and approved, belong to the IgG1, IgG2, or IgG4 subclasses. These subclasses differ in cross-linking capabilities and biological activity, including ADCC and complement-dependent cytotoxicity (CDC) effector functions.¹¹⁴ IgG1 is commonly used due to its enhanced delivery capabilities and additional effector functions compared to IgG2 and IgG4.¹¹⁵ However, when considering the target characteristics and the proposed MOA, effector function may not, in some cases, be desirable, and IgG2 and IgG4 antibodies may be preferred. Isotype selection can also play a role in drug-linker conjugation, particularly when conjugating via cysteine residues.

2.1. Size of the monoclonal antibody

After selection of target antigen and antibody isotype, it is pertinent to consider the size constraints of the antibody for targeting tumors. Classical ADCs include a full-length antibody molecule, which may present challenges for the uptake and permeability of some solid tumors.¹¹⁶ To generate ADCs with improved uptake and penetration, several strategies in novel ADC design have pivoted toward the use of smaller formats of the antibody, including Fab-drug conjugates, scFv-drug conjugates, and diabody-drug conjugates.^117–119 However, these smaller formats may be associated with faster clearance compared to the full-length IgG.¹¹⁶ Although the theoretical potential for these smaller formats exists, much work is still needed to prove a clear benefit.

2.2. Antibody modifications

Another important factor to address when considering ADC design is the posttranslational modifications (PTMs) of the antibody. Like most proteins, antibodies are subject to modifications both during antibody production and storage. Modifications can affect the stability, structure, and biological activity of the antibody, and consequently the ADC.¹²⁰ PTMs, such as deamidation, sialylation, and c-terminal lysine cleavage, can affect the net charge of the mAbs.¹²¹ These changes can lead to the production of charge variants and heterogeneity of the ADC with wider consequences for antibody structure and biologic activity. With regard to the ADC, these changes can interfere with target-ADC binding and ADC entry into the tumor cells, resulting in lower efficacy of the ADC molecule.^122,123 ADCC or CDC functions may be inhibited through PTMs, further hindering the ADC efficacy. To ensure batch-to-batch consistency, it is critical that the ADC charge profile and other modifications that may have been introduced are thoroughly characterized during development.

2.3. ADC internalization

Most ADCs are designed against a target antigen that displays efficient internalization via receptor-mediated endocytosis to facilitate ADC entry upon recognition.¹²⁴ Receptor internalization has long stood as a requirement for effective ADC design to enable release of the cytotoxic payload with limited effects on healthy cells.¹²⁵ To design a successful internalizing ADC, target accessibility, density, internalization rate, and intracellular trafficking of the ADC must be assessed. In general, ADCs against targets expressed on solid tumors have more physical barriers to overcome to reach the antigen following administration compared to hematological malignancies in which the targets are readily exposed to circulating ADCs.¹²⁶ Further, targets can sometimes “shed” from the surface and be released into the blood, posing challenges against loss of ADC in circulation, clearance by the liver, and overall lower efficacy.¹²⁷ Determination of the receptor expression (receptor copies/cell), internalization rate, and rate of recycling can all directly affect ADC entry into target cells and can be difficult to address.

While it is a known fact that the targeting mAb should exhibit high affinity toward the antigen, establishing a minimum threshold for target binding can be variable.¹⁰⁹ As stated earlier, target density and internalization rates are key to ADC entry, metabolism, and payload accumulation within the tumor cell, but these may also be challenging to optimize. Efforts have been made to explore the potential of non-internalizing ADCs that target structural components of the environment surrounding the tumor cell.^128,129 Such an approach may help overcome the penetration barriers of solid tumors by targeting an antigen highly expressed within the tumor stroma.¹¹³ In such cases, proteases shed from nearby apoptotic cells allow for the release of the payload which, due to its smaller size, can cross the membrane of tumor cells.¹³⁰ A recent study showed anti-tumor activity in vivo of a non-internalizing ADC toward Gal-3BP protein that is secreted by cancer cells and localized to the cell surface. Due to accumulation at the surface of cancer cells, toxicity to normal tissues was limited, suggesting that non-internalizing ADCs can exhibit both potency and safety.¹³¹ Similarly, ABBV-085, produced by AbbVie, is an anti-LRRC15 ADC that was recently evaluated in clinical trials.¹³² Leucine-rich repeat containing 15 (LRRC15) is a member of the LRR superfamily with expression primarily on the surface of cancer-associated fibroblasts and stromal cells.¹³³ In preclinical studies, ABBV-085 demonstrated anti-tumor activity in several LRRC15-positive cancer models, as well as LRRC15 stromal fibroblast-positive/cancer-negative models.¹³³ The cell permeable properties of the two MMAE molecules conjugated to the antibody allowed for bystander activity, while an increase in immune infiltrate was also observed in the tumor microenvironment, both contributing to the efficacy of the ADC. Despite such promising preclinical data, only 14.8% of sarcoma patients treated with ABBV-085 at the recommended phase I b dose demonstrated a confirmed partial response, while 29.6% maintained stable disease, and progressive disease was observed in 40.7%.¹³² As of May 2021, no clinical trials that include ABBV-085 are ongoing. More studies are needed to look into the effectiveness of non-internalizing and tumor microenvironment-directed ADCs compared to those that are classically internalized by tumor cells in the clinical setting.

3. Cytotoxic payload

While the mAb is arguably the most important component in ensuring ADC efficiency, the cytotoxic payload is responsible for the execution of tumor cell killing.¹¹² The cytotoxic payload (sometimes referred to as the “warhead”) is typically a small-molecule drug with the purpose of eliciting cell killing of the targeted tumor cells/tissues. The first generations of ADCs used drugs approved for clinical use, including doxorubicin, and resulted in low clinical activity.¹³⁴ The next wave of ADCs adopted the use of more potent small-molecule drugs that were too toxic as a stand-alone treatment, but showed promise in efficacy when selectively delivered to target cells with IC50s in the 0.01–0.1 nM range.¹³⁴ Even so, due to biodistribution, uptake, and loss of conjugation in circulation, it is estimated that only 1–2% of ADC payload will reach the intracellular target.¹⁰⁸ Thus, the potency of the payload must be high (ideally in the subnanomolar range) so that even at a lower accumulated concentration, the ADC can still eradicate the target cells. To achieve this goal, current ADCs mostly incorporate potent molecules that either disrupt tubulin polymerization or induce DNA-damage (Figure 3d).²⁴ Understanding the MOA of the ADC payloads and its applicability to the target is critical. While many ADCs in development currently use anti-mitotic tubulin disruptors for their selective eradication of rapidly proliferating cells, these payloads may not be effective toward targets that are not highly proliferative. It is worth noting the emergence of various toxic molecules conjugated to antibodies that are currently being investigated in clinical trials. Payloads such as topoisomerase inhibitors are gaining interest as cytotoxic agents that may be less toxic, allowing for higher DAR ADCs. This effect can be observed in the recently approved HER2-targeting ADC, trastuzumab deruxtecan, in which both a high DAR and reduced toxicity of the payload were used to produce a molecule with increased stability, efficient cytotoxic effect, and improved safety profile compared to T-DM1. Other payloads such as PBD dimers that exhibit high potency are also emerging for ADC design. These agents, as seen in the recently approved Lonca-T, can exert cytotoxicity at low concentrations with other advantages including efficient bystander cell killing and the potential for low systemic toxicity due to such short half-lives, dependent on several factors, such as conjugation strategies. This often results in low DAR species (e.g., DAR2) and lower dosing compared to ADCs carrying less potent payloads to balance the anti-tumor activity and safety profile of PBD-ADCs. Novel payloads such as immunostimulatory agents, RNA polymerase II inhibitors, and pro-apoptotic BCL-xL inhibitors are also emerging. Further, the choice of payload should also consider potential drug resistance mechanisms. Although payloads such as MMAE and calicheamicin have been shown to be good substrates of P-glycoprotein, others such as PBD dimers and some topoisomerase I inhibitors have been shown to exhibit anti-tumor activity in multi-drug resistant cancer cells.^135,136

While aggregation due to unfolding and exposure of certain hydrophobic residues are concerns that exist for the parent mAb, this challenge is heightened with regard to ADCs due to conjugation methods and linker-payload additions.¹³⁷ Research groups have demonstrated the effects of small-molecule drugs on the hydrophobicity of the ADCs, making the drug more prone to aggregation, particularly under thermal stress.^138,139 As with unconjugated mAbs, aggregation decreases the activity of the ADC and can render the molecule less effective. Apart from aggregation due to linkage, hydrophobic drugs that are conjugated to the mAb can, if exhibiting efficient hydrophobicity, enter neighboring cells upon release from specific chemical linkers and induce killing of non-target cells. For ADCs carrying PBD or MMAE such as the vedotin ADCs, the payload’s cell permeability allows for a bystander killing effect within a heterogenous population. T-Dxd has also been reported to cause bystander killing via drug efflux into neighboring antigen-negative tumor cells.¹⁷

Several conjugation methods address the issue of hydrophobicity by using hydrophilic spacers, linkers, or payloads.^140,141 In a recent study, Satomaa et al. demonstrated the enhanced stability of a novel hydrophilic payload that allowed for higher DAR achievement and low toxicity as a free drug while maintaining high cytotoxicity in target cells.¹⁴² This auristatin glycoside, β-D-glucuronyl-monomethylauristatin E also showed efficient internalization, metabolic processing, and bystander killing effect following conversion to MMAE through cellular metabolism.¹⁴² Further study into a hydrophobicity balance is needed to promote the efficacy of ADCs, hinging on both linker-payload choice and conjugation characteristics.

4. Linker chemistry and conjugation methods

The chemical linker is a critical component of the ADC that joins the mAb and the cytotoxic payload. The linker facilitates ADC stability in circulation until the ADC reaches the target cell and the payload is released.¹⁴³ There are two classes of linkers: cleavable and non-cleavable.¹⁴⁴ Cleavable linkers can be cleaved in response to certain environment factors to release the free drug into the cytosol.¹⁴⁵ This includes hydrazine linkers that are cleaved in response to the acidic environment of the endosome and lysosome, exhibited in gemtuzumab ozogamicin. Cleavable linkers can also be cleaved in the presence of proteases or reducing agents, such as cathepsin B or high levels of glutathione.¹⁴⁵ For non-internalizing ADCs, drug release relies on extracellular cleavage by glutathione and proteases that have been shed as a result of tumor cell death.¹⁴⁶ Non-cleavable linkers are resistant to proteolytic degradation and rely on the full degradation of the antibody to release the attached linker-payload complex. This requires the payload to remain active, while linker bound.¹⁴⁴ Because of this, non-cleavable linkers have been proposed as a strategy to overcome drug resistance as the linker-payload complex is no longer a substrate for MDR1.¹⁴⁷ Therefore, the proposed MOA of the ADC can be a determinant for linker choice.

For some ADCs, the chemical linker may also serve to balance the hydrophobicity between the mAb and payload, therefore reducing potential aggregation. In this regard, analyzing the bioanalytical significance of all components of an ADC is important for evaluating the safety and efficacy of the drug. Hydrophilic linkers and spacers, including cyclodextrins, polyethylene glycol, and other polymers, may play a role in improving the stability of circulation, potency toward the target cells, and overall pharmacokinetics of the conjugate.^148–150

In addition to selectively choosing a chemical linker, the method by which the payloads are conjugated to the antibody is essential in modulating the homogeneity and potency of the ADC.¹⁵¹ Until recently, conventional methods relied upon lysine and interchain cysteines to conjugate cytotoxic molecules to the antibody. In the case of lysine conjugation, heterogeneity was unavoidable due to the large number of lysines available for conjugation compared to cysteine conjugation.¹⁴⁴ Due to the lack of control of conjugation site and quantity, lysines proximal to Fc binding can be affected by drug conjugation, resulting in lower efficiency of binding and cytotoxicity of the ADC.¹⁵²

Currently, most ADCs in use or under development rely on interchain disulfide cysteines for conjugation, in which the 4 (IgG1 and IgG4) or 6 (IgG2) interchain disulfide bonds are reduced by an excess reducing agent, namely tris(2-carboxyethyl)phosphine or dithiothreitol.¹⁵³ This spares disruption of intrachain disulfide bonds while freeing sulfhydryl groups from cysteine residues participating in interchain disulfide bonds (Figure 3e). The resulting product is a mixture of ADCs containing 0–8 drugs per parent IgG1 or IgG4 and 0–12 per IgG2, with predominantly even numbered DAR (0, 2, 4, 6, 8, 10, 12) species within the ADC mixture. Homogeneity of ADCs has improved because substantially fewer cysteines are available for conjugation following reduction compared to lysines. However, even with more homogenous methods, control over DAR and drug-load distribution (DLD) can still be enhanced. Optimizing the DAR and DLD is critical for the pharmacokinetics of the ADC and eradication of target cells.¹⁵⁴ Ensuring homogeneity across all ADCs produced is a key aspect of quality control for developers and manufacturers to advocate for the safety of the product. Early studies initially indicated that DAR of 2–4 drug molecules per antibody is ideal for ensuring stability in circulation and efficacy.¹⁵⁵ ADCs with too few conjugated payloads may exhibit low potency, while increased off-target toxicity and rapid clearance were previously observed in ADCs with higher DAR.¹⁵⁵ However, the recent approvals of trastuzumab deruxtecan and sacituzumab govitecan have challenged this previously defined limit of 4, as both carry nearly eight payloads per antibody. Further, there is a broad range of average DAR in ADCs under clinical evaluation, with as low as 1 payload per antibody such as BDC-1001 and as many as 15 such as ASN004. The DAR may also influence dosing, antibody concentration to be administered, and subsequent tumor uptake of the ADC. ADCs of low DAR may be administered at higher doses depending on payload potency, which delivers a higher antibody concentration to facilitate ADC penetration into solid tumors. ADCs of high DAR may be administered at lower doses, which may lead to a lower antibody concentration and poorer tumor uptake. This notion was supported by in vitro studies involving co-administration of DAR0 or the naked antibody.^156,157

Novel site-specific conjugation methods using unique linker chemistries that yield homogenous ADCs of desired DARs have emerged.¹⁰⁷ One technique involves installing natural or unnatural amino acids into the antibody sequence for strict control over DAR and DLD. The most notable approach to engineering natural amino acids is THIOMAB™, which inserts cysteines at specific sites to allow for thiol conjugation.¹⁵⁸ The resulting ADCs, referred to as THIOMAB™ -drug conjugates or TDCs, have shown improved homogeneity compared to conventionally conjugated ADCs.

Engineering of unnatural amino acids has included examples such as p-acetylphenylalanine and p-azidomethyl-L-phenylalanine, yielding ADCs in which DAR and DLD could be regulated.^159,160 Another strategy is the SMARTag™ technology, which uses chemoenzymatic reactions to install an aldehyde tag for site-specific conjugation,¹⁴⁴ as mentioned above. Here, the conjugation site is a formylglycine (aldehyde) residue produced through enzymatic oxidation of a cysteine in a specific pentapeptide consensus sequence in the mAb.^29,161 A similar engineering method installs natural or synthetic carbohydrate moieties onto the glycan as points of target for drug conjugation.¹⁶² Not only does this technique address homogeneity concerns, but it also provides consistency in loading despite the heterogeneity of N-glycan forms of immunoglobulins. Thompson et al. conjugated PBD with DARs of 4 to azide-modified GalNAc to demonstrate the utility of glycoengineering in ADC design.¹⁶³ The glycoengineered ADCs exhibited potent killing both in vitro and in vivo. With this method, developers enzymatically alter the glycan profile by introducing particular carbohydrate moieties for drug conjugation, yielding evidence of increased homogeneity and potency over conventional conjugation methods. Using enzymes that recognize specific engineered amino acid sequences to cleave and covalently attach drug molecules can also improve control over ADC homogeneity.

Another site-specific conjugation method that has gained attention both among researchers and biotechnology companies is disulfide rebridging.¹⁶⁴ This method is attractive due to its ability to control DAR and DLD without the need of re-engineering the mAb. The technique takes advantage of the conventional cysteine coupling method to conjugate a bifunctional payload.¹⁶⁵ As a result, one drug molecule is coupled per interchain disulfide bond. Using this method, developers achieve consistent DARs of 4 and 6, depending on the immunoglobulin isotype, with expected DLDs. This technique has shown promise to address homogeneity concerns as demonstrated by Bryant and colleagues.¹⁶⁶ One drawback to this method is the use of additional chemicals, requiring additional purification methods and analytical characterization of the final product. Further, as with other described methodologies, the success of each technique in producing a homogenous product is dependent on other factors, such as the nature of the antibody and payload.

Quality assessment

ADCs are complex molecules with unique critical quality attributes (CQAs), including DAR, DLD, the amount of unconjugated payload or unconjugated antibody, antigen binding, and cellular activity, in addition to the quality requirements for naked mAbs. To ensure product quality and manufacturing consistency, each CQA must be adequately evaluated. Several analytical platforms are adopted from mAb analysis, with state-of-the-art analytical techniques being developed to assess ADC-specific CQAs, such as high-resolution native mass spectrometry (MS), native ion-mobility (IM) MS, and two-dimensional high performance liquid chromatography (2D-HPLC) (Table 3).^167,168

Table 3.

Analytical characterization of ADC CQAs

Quality Attributes	Analytical Methods
	Cysteine Conjugates	Lysine Conjugates	Site-Specific Conjugates
DAR, DLD, and unconjugated species (DAR-0)	HIC Native IM-MS, native SEC-MS, or native sheathless CE-MS SEC-HIC two-dimensional HPLC (2D-LC) HIC-RPLC-MS or HIC-SEC-MS (2D-LC-MS) HIC-SEC-IM-MS (2D-LC-IM-MS)	UV/Vis spectroscopy (only for average DAR) CIEF	HIC RPLC RPLC-MS SEC-MS on deglycosylated ADC
Conjugation sites	RPLC-MS/MS or sheathless CE-MS/MS	RPLC-MS/MS (combining tryptic peptide mapping with Asp-N and/or Glu-C peptide mapping)	RPLC-MS/MS
Posttranslational modifications (PTMs)	HILIC-MS or sheathless CE-MS/MS on Fc-fragments HIC-RPLC-MS on reduced ADC	mCE-MS CEX-RPLC-MS	RPLC-MS/MS
Free drug species	RPLC on ADC-free sample SEC, SEC-RPLC, or solid phase extraction LC (SPE) coupled with RPLC-MS (SPE-RPLC-MS) analysis of untreated ADC sample	RPLC on ADC-free sample SEC or SEC-RPLC analysis of untreated ADC sample	RPLC on ADC-free sample
Size variants	SEC SEC-HIC or SEC-RPLC	SEC SEC-PRLC CE-SDS	SEC Analytical ultra-centrifugation (AUC) CE-SDS
Charge variants	CIEF	CIEF	IEX

The listed techniques have been used to characterize the CQAs of cysteine, lysine, and site-specific conjugates.

Antibody–drug conjugate (ADC); Analytical ultracentrifugation (AUC); Capillary isoelectric focusing (CIEF); Capillary electrophoresis-sodium dodecyl sulfate (CE-SDS); Cation exchange chromatography (CEX); Hydrophobic interaction chromatography (HIC); Hydrophilic interaction liquid chromatography (HILIC); Ion mobility mass spectrometry (IM-MS); Microfluidic capillary electrophoresis (mCE); Mass spectrometry (MS); Reverse phase liquid chromatography (RPLC); Size exclusion chromatography (SEC); Solid phase extraction (SPE); Ultraviolet-visible spectroscopy (UV/Vis)

1. DAR and DLD

For cysteine-linked ADCs, hydrophobic interaction chromatography is commonly used to determine the average DAR, DLD, and unconjugated mAb species (DAR-0).^139,169,170 Emerging techniques include high-resolution MS and native IM-MS, operated under native conditions using MS-compatible ammonium acetate buffer at neutral pH.¹⁷¹ Similar analytical approaches can also be applied to other site-specific ADCs.^172,173 Lysine-linked ADCs are inherently associated with a high heterogeneity, posing an analytical challenge. For those ADCs, DAR is usually determined by measuring both the drug-specific absorbance and the mAb absorbance at 280 nm.^174,175 The conjugation sites and PTMs of mAbs can be determined through peptide mapping, producing a single tryptic peptide map for cysteine-linked ADCs, or a combination of tryptic, Asp-N, and Glu-C maps for lysine-linked ADCs;^176,177 alternatively, sheathless capillary electrophoresis (CE) coupled with MS/MS can also be used to assess these attributes.¹⁷⁸

2. Process- and product-related impurities

ADCs are associated with specific process-related impurities, such as unconjugated payload, free linker, or other chemicals used in the manufacturing process. Reverse-phase (RP)-HPLC provides a platform for assessing these potential impurities in the final products.¹⁷⁹ Removal of protein-containing species (e.g., intact ADC, unconjugated antibody) can help improve assay performance, using protein precipitation, size-exclusion chromatography (SEC), or SEC×RP 2D-LC.^180,181 2D-LC-MS can increase assay sensitivity, enabling detection of trace amount of free payload.¹⁸² Product-related impurities such as aggregates, fragments, charge variants, and other PTMs on the antibody can be assessed by a combination of SEC, analytical ultracentrifugation, CE, capillary isoelectric focusing, ionic exchange chromatography, and peptide mapping.^183,184,185

3. Potency assays

Potency assays are a critical component of quality control strategies for complex drug products,together with physicochemical tests to ensure manufacturing consistency in the product lifecycle. In general, potency assays should reflect the product’s MOAs. For multifunctional products, more than one potency assay will be needed to fully capture the biological activities. An ADC may retain its mAb-associated MOAs, such as signaling blockade, ADCC, or CDC. An ADC may also elicit a bystander effect, thereby affecting both antigen-positive and antigen-negative cells upon release of the cytotoxic payload into the surrounding tumor microenvironment. Therefore, ADCs should be evaluated using both antigen-binding assays and cell-based functional assays as appropriate.

Future perspectives

The availability of ADCs offers a promising therapeutic option for numerous cancer types. With more ADCs entering clinical trials, the industry is gradually shifting from conventional technologies to newer and more robust approaches to develop such complex products. This includes strategies for exploring novel tumor antigens, antibody formats, payloads, linkers, and advanced conjugation technologies, each with the aim of improving the therapeutic window of ADCs. Among the emerging antibody formats, scFv may have better solid tumor penetration and uptake. Bispecific and biparatopic ADCs may overcome the barrier of tumor heterogeneity. Probodies and other conditionally active biologics (CABs) may reduce off-target effects. Multiple payload classes besides microtubule-disrupting agents, including PBD dimers, topoisomerase inhibitors, anthracyclines, and protein-specific modulators, are being introduced into a new generation of ADCs. Furthermore, several site-specific conjugation platforms are now used to enhance ADC stability in circulation while maintaining efficient release of the payload (Figure 5). The complexity of ADCs poses daunting analytical challenges, especially when hydrophobic payloads are incorporated. State-of-the-art analytical techniques are required and continue to evolve in alignment with the rapid growth of ADC development. Applying the appropriate sets of analytical techniques is crucial for adequately characterizing product attributes, thereby ensuring manufacturing consistency during development and throughout the product lifecycle.

Figure 5.

Expanding the ADC Framework. New monoclonal antibody formats, conjugation methods, linker and spacer techniques are emerging to optimize safety and efficacy profiles for oncological indications

The therapeutic potential of ADCs is also highlighted by the expansion of clinical indications, shifting from hematological malignancies (lymphoma and leukemia) to an increase in solid tumors (e.g., breast cancer, urothelial cancer, lung cancer, and ovarian cancer). Many ADCs within the clinical pipeline are being evaluated in combination with other established therapeutic classes, such as immune checkpoint inhibitors and mAbs targeting different antigens. The cumulative clinical data, combined with the product quality information described here, are helping to shape the future development of ADCs. As more data becomes publicly available, a comprehensive analysis of potential correlations between specific product quality attributes and the safety and efficacy profiles of individual products will certainly inform optimization of ADC design and manufacturing toward next-generation innovative cancer medicines.

Funding Statement

This work was funded by the U.S. Food and Drug Administration. The funder had no role in the study design, data collection and analysis, decision to publish, or preparation of the manuscript.

Abbreviations

2D-HPLC	Two-dimensional high performance liquid chromatography
ADC	Antibody-drug conjugate
ADCC	Antibody-dependent cellular cytotoxicity
ADCP	Antibody-dependent cellular phagocytosis
AF	Auristatin F
AF-HPA	Auristatin F-hydroxypropylamide
BCMA	B-cell maturation antigen
CAB	Conditionally active biologic
CDC	Complement-dependent cytotoxicity
CE	Capillary electrophoresis
cIEF	Capillary isoelectric focusing
CQA	Critical quality attribute
DAR	Drug-to-antibody ratio
DLBCL	Diffuse large B-cell lymphoma
DLD	Drug-load distribution
DLL3	Delta-like protein 3
DNA	Deoxyribonucleic acid
FDA	U.S. Food and Drug Administration
Fv	Variable fragment
IM	Ion mobility
Lonca-T	Loncastuximab tesirine
LRRC15	Leucine-rich repeat containing 15
mAb	Monoclonal antibody
mBC	Metastatic breast cancer
MDR	Multi-drug resistance
MMAE	Monomethyl auristatin E
MMAF	Monomethyl auristatin F
MMAU	Monomethyl auristatin derivative
MOA	Mechanism of action
MS	Mass spectrometry
ORR	Overall response rates
PBD	Pyrrolobenzodiazepine
PDC	PROBODY-drug conjugates
PTM	Posttranslational modifications
Rova-T	Rovalpituzumab tesirine
R/R	Relapsed or refractory
RP	Reverse-phase
ScFv	Single-chain variable fragment
SEC	Size exclusion chromatography
SMARTag™	Specific Modifiable Aldehyde Recombinant Tag
TDC	THIOMAB™-drug conjugates
T-DM1	Trastuzumab emtansine
T-Dxd	Trastuzumab deruxtecan
TNBC	Triple-negative breast cancer
Trop2	Trophoblast-cell surface antigen 2

Disclaimer

This article reflects the views of the authors and should not be construed to represent FDA’s views or policies.

Disclosure statement

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.