The immunomodulatory drugs

Thalidomide was developed in the mid-1950s for use as a sedative with minimal side effects or addictive potential [27]. By 1961 it was taken off the market worldwide after teratogenic effects were noted; close to 10,000 infants were thought to have limb deformities caused by the drug. The observation that thalidomide resulted in teratogenesis led investigators to consider its effect as an antineoplastic drug, though early efforts failed to show significant activity in advanced cancer [28, 29].

Thalidomide was initially successful as a treatment for erythema nodosum leprosum and inflammatory ulcers. Decreased tumor necrosis factor-alpha and inhibition of vascular endothelial growth factor and fibroblast growth factors were found to be the mechanism for anti-inflammatory and antiangiogenetic properties. In the 1990s, the antiangiogenic properties of thalidomide were discovered [30] and its promise as a cancer therapeutic emerged largely through kismet. A patient’s family member noticed the association of thalidomide activity with anti-angiogenesis and, also noted that MM was associated with an excess of neovascular formation. Following a conversation with their doctor a clinical trial in MM was initiated [31]. Later, it was shown that other mechanisms of action of thalidomide included impairment of cell-cell signaling, inhibition of cytokines important to plasma cell growth proliferation, free radical-mediated DNA damage, and enhancement of T and natural killer (NK) cell responses [32–35]. However, it was years from this time point until the mechanism of action of thalidomide was fully understood, when cereblon, a receptor of the E3 ubiquitin ligase complex, was identified as a thalidomide-binding protein leading to proteasomal degradation of key plasma cell transcription factors Ikaros and Aiolos [31, 36, 37].

Early clinical trials demonstrated responses with single agent thalidomide. Among 84 patients with refractory MM, thalidomide had an overall response rate (ORR) of 32% with two complete responses [38]. A number of studies compared thalidomide to dexamethasone and evaluated the two drugs in combination in the relapsed setting [39, 40]. As initial therapy, thalidomide in combination with doxorubicin and dexamethasone (TAD) was compared to what was then the European standard induction regimen for transplant eligible patients, VAD [41]. The superiority of TAD had a significant impact on the therapeutic landscape, ultimately leading to the “death of VAD” as a clinically used regimen, the licensing of thalidomide in combination with dexamethasone in 2006, and the development of the Risk Evaluation and Mitigation Strategies (REMS) program [42]. The REMS program is now in widespread use to prevent exposure of “at risk” individuals such as pregnant women to thalidomide, and the principle to decrease exposure has been extended to other anti-cancer drugs that may also be considered at high risk of teratogenicity.

Since the introduction of thalidomide and its derivative lenalidomide, IMiDs have become a critical backbone of standard induction regimens. In the process of introducing these drugs it was shown in relapsed patients that a lenalidomide-dexamethasone combination was more effective when compared to dexamethasone alone [43–45]. A third thalidomide derivative, pomalidomide, was developed in a similar fashion, leveraging off-label use of dexamethasone as its control arm, and has become a key component of treatment for relapsed disease.

Prior to the official approval of IMiDs, they were widely employed in combination with dexamethasone by investigators in clinical trials, leading to greater experience and optimization of their use. IMiDs were initially given in a continuous fashion as maintenance for patients post-HDM with ASCT, and following the combination of melphalan, thalidomide and prednisone for non-transplant candidates before formal approval in these settings. The side effect profile of thalidomide, however, was not ideal for long term maintenance therapy after investigator-initiated trials noted toxicities such as constipation, somnolence, and neuropathy [46] and it was soon replaced by lenalidomide. A number of clinical trials explored lenalidomide as maintenance following HDM and ASCT [47–49] using two different dosing regimens: 10–15 mg continuously and 10 mg for 21 out of 28 days. Its use was associated with myelosuppression that led to discontinuation as maintenance, especially in older patients [49]. Intermittent dosing was associated with less myelosuppression [48], and this became the regimen now widely in use for maintenance, despite lenalidomide not having an FDA label for this dose schedule [50].

The efficacy of combination therapy with IMiDs and dexamethasone in relapsed MM provided a new FDA-approved comparator group for future clinical trials, allowing the field to move from the use of doublet combinations to combinations based on two new agents together with dexamethasone. Each successive improvement was compared to the previous standard of care, all relating back to the initial comparison to dexamethasone used without an FDA label (Figure 2).

The proteasome inhibitors

The development of PIs took a parallel and, in some ways, competitive path to the development of the IMiDs. The initial phase II trial that led to the approval of bortezomib [51] evaluated a dose of 1.3 mg/m² on days 1, 4, 8 and 11 intravenously. This dosing schedule was associated with significant peripheral neuropathy [52] and a dose reduction was incorporated into future clinical trials to abrogate the toxicity. Despite the side effect profile, the original dose was approved [53], and was the dosing schedule chosen for the drug’s subsequent development, for example with melphalan as part of the Velcade-MP (VMP) regimen [54]. As it was the approved dosage all subsequent regulatory trials were compared to this dose and schedule even though clinicians learned to manage the toxicity by using off-label weekly dosing, and rapidly switched to the use of subcutaneous injections when a preparation was available. Therefore, outside of the regulatory setting the dose was reduced to 1.3 mg/m² subcutaneously at weekly intervals–dosing that has never been formally tested against a control in a randomized trial and is based on small phase II datasets [55–57]. The use of standardized dosages during the approval process is important; however, this need has to be balanced against knowledge gained by real-world drug use obtained post-FDA approval to minimize side effects. In this case the newer dosing frequency decreased the rate of neuropathy and prolonged the length of time a patient was able to remain on drug. Such an approach does cause confusion when subsequent lines of therapy are developed, as ideally the original dosing schedule in the FDA label should be used as a standard against which to compare new therapies.

Developing and utilizing the concept of RRMM

An important theme of drug development in MM has been the strategy of introducing successive waves of potentially useful clinical drugs for relapsed patients who have received all available therapies–so called “areas of unmet clinical need”. As discussed above, the therapies available 5–10 years ago included an IMiD, a PI and possibly an alkylator. In the early 2000s, once a drug was shown to be effective in patients with relapsed or refractory disease, combinations of the new drug with known effective drugs were developed including doublet, triplet and quadruplet combinations. These combinations were then examined in an earlier relapse setting before moving to the newly diagnosed setting.

By 2010–2015 with an increasing number of effective therapies, the need for a different approach with tighter definitions of eligible patients was needed, leading to the concept of relapsed/refractory MM (RRMM) in drug development. This was defined as progressive (within 60 days) or refractory disease during or after the receipt of prior therapy according to strict International Myeloma Working Group (IMWG) criteria [58–60]. Retrospective data analyses showed that as a group RRMM was associated with an extremely poor prognosis. In a cohort of 543 patients who had received a median of four prior lines of therapy, median PFS was 5 months and OS was 13 months [61]. This constituted a well-defined group with an unmet medical need that could provide a path forward for rapid new drug approval. After testing novel drugs alone and in combination in this setting, they could be then moved forward for evaluation in patients with RRMM after 1–3 prior lines of therapy, and subsequently for NDMM.

Anti-CD38 antibodies

Anti-CD38 monoclonal antibodies have moved rapidly from an experimental drug to standard of care in various MM disease settings [62]. The initial studies of daratumumab showed a 30% ORR as a single agent in RRMM [63–65]. Since this observation, the drug has undergone extensive clinical trials in the relapsed setting (1–3 lines of therapy) where it has been combined with lenalidomide and dexamethasone (Rd) [66], bortezomib and dexamethasone (Vd) [67], carfilzomib and dexamethasone [68], as well as pomalidomide and dexamethasone [69, 70]. Isatuximab, another anti-CD38 antibody, is approved for use in combination with pomalidomide and dexamethasone [71] and is undergoing clinical trials with other drug combinations. Similar combinations have recently been approved in the upfront setting including daratumumab-Velcade-melphalan-prednisone (D-VMP), daratumumab-Revlimid-dexamethasone (DRd) for transplant ineligible patients, and daratumumab-Velcade-thalidomide-dexamethasone (D-VTD) for transplant eligible patients [72–74]. Compared to Europe where D-VMP, DRd and D-VTD are widely used for NDMM patients, in the US, Velcade-Revlimid-dexamethasone (VRD) is the standard of care for both patient groups. Although not yet FDA-approved, results from the Griffin study and extrapolation of other combination data showing good efficacy and safety profiles have led to the widespread use of D-VRD upfront [75]. Of note the FDA approvals for the subcutaneous form of daratumumab compared to the intravenous form also differ slightly [76, 77]. This is again due to the timing differences in the introduction of the two formulations. Even though the data concerning combinations with subcutaneous daratumumab upfront are awaited, there has been widespread uptake of this formulation in clinical practice, driven by the huge difference in infusion time (4–8 h vs. 5 min) and the advantages this has for both patients and clinics.

Developing the concept of penta-refractory disease

Once anti-CD38 agents became widely available the clinical nature and history of patients with RRMM changed and the old definition of RRMM was no longer appropriate. This resulted in two new functional definitions, which are currently still in use: triple class-exposed (prior IMiD, PI, and anti-CD38 antibody) and penta-refractory (prior exposure and refractoriness to bortezomib, both lenalidomide and pomalidomide, and daratumumab). An updated analysis of RRMM outcomes following the introduction of daratumumab identified a median OS of only 5.6 months for patients considered as triple-exposed and penta-refractory [78]. Again, this represented a new area of unmet medical need where patients continue to have an extremely poor outcome after exposure and refractoriness to available therapies.

Clinical trials are difficult to perform in this area as patients tend to be unwell, and a full prior medical history with access to detailed medical records is required to ensure that patients meet the stringent definitions. It has also become clear that although this is a well-defined group, not all patients have the same biological disease. For example, patients may become penta-refractory over a 10-year period having received each drug sequentially or they may reach the same clinical state quickly over an 18-month period having received two different combinations of therapy. In addition, these definitions do not take into account clinical characteristics such as extramedullary disease or plasma cell leukemia, or genetic features such as mutations or translocations involving MAF, NSD2, MYC, or TP53.

Despite such difficulties, two drugs have been developed in this setting. The first is an antibody-drug conjugate belantamab mafadotin [79] that targets the B-cell maturation antigen (BCMA) which is highly expressed on MM cells [80, 81]. The anti-BCMA monoclonal antibody is conjugated to monomethyl auristatin, a microtubule disrupter resulting in targeted MM cell death. As a single agent in RRMM it showed a 30–34% ORR depending on the dose used [79]. The second therapy developed in this setting is the nuclear export inhibitor Selinexor [82], which blocks exportin 1, promoting apoptosis of malignant plasma cells by maintaining tumor suppressor proteins within the cell nucleus. In a trial of 122 patients with RRMM who were at least triple-class-exposed, selinexor combined with dexamethasone produced an ORR of 26% including two stringent complete responses. Given the activity of these therapies, use outside of their label will likely occur, i.e., in patients who fail to meet the strict criteria of penta-refractory disease but who in the physician’s eyes are likely to benefit from therapy.

The challenge moving forward with new drug development

In addition to belantamab mafadotin and selinexor, a number of other drugs with novel mechanisms of action are also being developed in this penta-refractory setting, including T-cell engaging therapy either with chimeric antigen receptor (CAR) T-cells recognizing antigens expressed highly on MM cells or with bi-specific and tri-specific antibodies recognizing MM antigens and T-cells simultaneously.

Other novel drug classes include melphalan flufenamide (melflufen), a peptide-drug conjugate that delivers an alkylating agent directly into MM cells [83–86]. In a phase II trial of heavily pretreated patients, in combination with dexamethasone, melflufen had an ORR of 31% [87]. In an ongoing phase III trial of patients with RRMM, melflufen plus dexamethasone will be compared to pomalidomide and dexamethasone.

At the moment with only phase I and II data available for these emerging therapies, the temptation is to carry out cross-trial comparison in the penta-refractory setting in order to determine the optimum agent for an individual patient. As discussed above this approach is fraught with difficulties and can be misleading due to the composition of small biologically variable groups.

For pharmaceutical companies the challenge moving forward is how to develop a drug in such a crowded setting. There still remains an unmet need as patients continue to relapse, but do patients now need to be refractory to belantamab mafadotin and selinexor in order to receive these new agents? What happens in cases where it is not clinically appropriate that a patient receives all these drugs? Do therapies need to be given in a pre-determined order? For a phase III trial how is an appropriate comparator chosen? Is the “new RRMM” setting actually the most appropriate position to develop new agents? And finally, maybe most importantly, how does individual patient biology determine response to therapy? It will be critical to address these questions going forward for a disease where previously there was only one available therapy (Figure 3).

Display full size

Figure 3. 

Traditional flow of drug development in MM

The situation in patients not destined for transplantation

In older less-fit patients with MM who are not candidates for ASCT, drug development and off-label drug use has also occurred. Early studies explored combinations of low dose melphalan with a steroid followed by the addition of a third agent, such as thalidomide, bortezomib or lenalidomide. In this group of patients, it became obvious that combinations had the potential to be detrimental for frail patients, as the side effect profile reduced quality of life and decreased efficacy. The classic example of this was the combination of melphalan and prednisone with lenalidomide, where older patients did not achieve a survival benefit [88]. The FIRST trial established Rd as an alternative to a melphalan combination [89], leading eventually to the MAIA study which demonstrated that lenalidomide plus daratumumab (DRd) outperformed Rd, and importantly was not detrimental to quality of life in this frailer patient population [72]. The countrywide differences in the drug approval processes and marketing authorizations has led to quite dramatic differences in what is considered standard of care for this patient group. For example, in the US, off-label experience with dose-reduced VRD in older populations established the efficacy of this combination. This resulted in a clinical situation where VRD with daratumumab had largely become the standard for all groups, with performance status or frailty score being used to decide on the appropriate combination and dosing, whereas DRd, VMP and D-VMP are often used as standards in other countries.

The impact of the European Union

For clarity we have focused predominantly on the evolution of therapies in the US, but of course, the European Union (EU), Canada, Asia-Pacific and South America exerted a significant impact on the use of therapies both on and off-label in the individual countries and in the US. Critical features impacting off-label use that differentiates these areas from the US are the different regulatory processes and the factors governing reimbursement in individual countries. For example, the EU regulatory authorities have tended to insist on evidence derived from randomized trials with PFS and OS as end points and have not had a fast track to licensing based on single arm studies in RRMM where response rate and PFS are end points. This has led to different time frames for approvals of new agents, different clinical trial designs to meet the evidence requirements of the regulatory agencies and the important role of country-specific data generated by many collaborative groups such as the IFM in France, MRC in the UK, Hovon in Holland, and the GMMG and DSSM in Germany. A key example of the country/regional difference is the use of VRD as a standard in the US, whereas in Europe, VTD and VMP were considered standards. Thus, when data emerged from large randomized studies based in Europe such as D-VTD and D-VMP they had little relevance in the US, and although FDA-approved were not widely utilized.

Perhaps the major factor driving the evolution of different preferred regimens in countries are the reimbursement procedures. With the complexity of drug costs and different healthcare systems, in addition to an assessment of safety and efficacy performed at the regulatory level, there is often an assessment of “value”, performed either at an insurance company level or a national level. For example, the European Medical Authority (EMA) may approve a license but the individual countries of the EU will then assess locally how the drug will be reimbursed within their country. The definition and methodologies behind the term “value” varies, as it takes in to account not only the expected benefit in terms of prolonged life for an individual patient, but the impact of this on the wider community (e.g., quality of life, productivity and contribution to the wider community, value in comparison to other health care advancements). It is therefore easy to see that with varying financial situations in the countries and different approval/reimbursement processes that large differences in both on and off-label prescribing can evolve. Ultimately though, the weight of the data and technological advances such as the recent introduction of anti-CD38 antibodies leads to a rebalancing of the system. This is in the process of happening with the global uptake of VRD combined with daratumumab as a global standard.