Optimizing High Dose Melphalan

Abstract

Melphalan, has been a major component of myeloma therapy since the 1950s. In the context of hematopoietic cell transplantation (HCT), high dose melphalan (HDM) is the most common conditioning regimen used due to its potent anti-myeloma effects and manageable toxicities. Common toxicities associated with HDM include myelosuppression, gastrointestinal issues, and mucositis. Established approaches to reduce these toxicities encompass dose modification, nausea prophylaxis with 5HT3 receptor antagonists, cryotherapy, amifostine use, and growth factors. Optimization of melphalan exposure through personalized dosing and its combination with other agents like busulfan, or bendamustine show promise. Propylene glycol-free melphalan (Evomela) represents a novel formulation aiming to enhance drug stability and reduce adverse effects. This review explores strategies to enhance the efficacy and mitigate the toxicity of HDM in multiple myeloma. Future directions involve exploring these strategies in clinical trials to improve the safety and efficacy of HDM, thereby enhancing outcomes for multiple myeloma patients undergoing autologous HCT.

Introduction

Melphalan was one of the first di-alkylating agents synthesized and has been used since the 1950’s to treat a variety of solid tumors and hematologic malignancies.^1–3 Melphalan is one of the most commonly used agents in conditioning regimens for hematopoietic cell transplantation (HCT) due to its intense myelosuppressive activity and acceptable gastrointestinal tract toxicity characterized by nausea, vomiting, diarrhea and anorexia as well as mucositis. These toxicities have been manageable even at fixed doses of 240 mg/m².^4,5

High dose melphalan (HDM) is generally administered intravenously as a fixed dose anywhere between 70 to 200 mg/m². In plasma, 90% of the drug is protein bound. It is transported into the cells by amino acid transport systems and within the cell melphalan undergoes chemical hydrolysis and the metabolites are excreted through the kidney into the urine. Melphalan’s biological half-life is approximately 60 minutes, its clearance is influenced by the creatinine clearance, fat free mass, and hematocrit.^6–11

Myelosuppression was identified early on as the dose-limiting toxicity of melphalan. At doses of 80 to 140 mg/m2 melphalan can be administered without stem cell support but is associated with almost universal grade 3 – 4 myelosuppression and in the initial report by Selby et al a 17% a non relapse mortality rate.¹² McElwain et al and Barlogie et al. pioneered the use of HDM with autologous bone marrow support and demonstrated that this strategy could reduce the risk of death and severe myelosuppression due to HDM.^{13, 14} Barlogie et al. pioneered the use of high dose therapy with HDM combinations as part of a “total therapy” strategy for patients with newly diagnosed myeloma.¹⁵ Subsequent randomized trials in the 1990s demonstrated the superiority of HDM in combination with total body irradiation (TBI) over standard chemotherapy.^16,17

Based on results of an IFM trial performed more than 20 years ago demonstrating that HDM given at a dose of 200 mg/m² was as effective and less toxic than HDM in combination with TBI, single agent HDM has become the most used conditioning regimen for autologous HCT in myeloma.^18–28 Herein, we will review strategies that can be utilized to enhance the efficacy of or reduce the toxicity of HDM in myeloma.

Current Challenges with HDM In Myeloma - Toxicity

The most common toxicity seen with HDM is myelosuppression which is universal and affects all hematologic elements. Gastrointestinal and mucosal toxicity are the second most common toxicities and are considered dose limiting. With the advent of novel induction regimens in myeloma that rely less on cytotoxic agents and more on immune-modulating drugs (IMIDS) (i.e. thalidomide and lenalidomide) and proteosome inhibitors (PI) (i.e. bortezomib and carfilzomib) with lower use of cytotoxics (i.e. alkylators and anthracyclines) the incidence of mucosal toxicities after HDM seems to have decreased.

Table 1 summarizes the most common grade 3 or greater toxicities seen in the most recent large, randomized trials comparing early versus delayed transplantation in myeloma using HDM at a dose of 200 mg/m² Mucosal and gastrointestinal toxicity has been the dose limiting toxicity when melphalan is given as a single agent and the maximum tolerated dose has been reported to be 220 mg/m² although higher doses have been delivered using amifostine.^5,

Table 1:

Common Grade 3 Toxicities with HDM as Conditioning for Autologous HCT In Modern HCT Studies (adapted from references 18 and 19)

Toxicity	% Grade >2 Toxicity on IFM Trial18	% Grade >2 Toxicity on the RVD+ASCT DETERMINATION Trial19
Myelosuppresion	100	100
Gastrointestinal (nausea/vomiting/diarrhea)	13	9
Febrile Neutropenia	14.9	9
Oral Mucositis	16.9	5.2
Sepsis	5.1
Pneumonia	6.6	9.0
Fatigue	1.7	6.0
Non Relapse Mortality		1.6
2nd Hematologic Malignancy		3.6

HDM toxicity is associated with a higher milligram per kilogram melphalan dose and renal dysfunction (both reflecting increased melphalan exposure), as confirmed by detailed pharmacokinetic studies performed by Nath et al.²⁴

Although HDM has been associated with the development of second primary cancers. In the IFM trial the incidence of second primary malignancies did not differ significantly between the groups of patients who received HDM upfront versus continued RVD therapy. Three cases of acute myeloid leukemia occurred in the transplant arm.¹⁸ In the DETERMINATION trial SPMs were reported in 37 patients (10.4%) in the RVD-alone group and 39 patients (10.7%) in the HCT arm. Second primary hematologic cancers occurred in 9 patients in the RVD-alone group (2.5%) and 13 patients in the HCT group (3.6%), although these differences were not statistically significant the observation that 10 patients in the HCT group developed either acute myeloid leukemia or myelodysplastic syndromes as compared to none in the control suggests a role for HDM in the development of these diseases.¹⁹

Current Challenges with HDM In Myeloma - Efficacy

Initial studies of HDM with or without marrow transplant are summarized in table 2.^12–15 These studies reflect the single agent anti myeloma activity of HDM in the absence of modern antimyeloma therapy. Further evidence of the single agent activity of HDM was demonstrated by the studies of Alexanian et al who reported a 70% PR rate and 8% CR rate amongst patients refractory to pulse dexamethasone or other chemotherapy induction regimens.²⁹ These observations and others support the practice of proceeding to HDM quickly even in patients with suboptimal response to primary induction therapies.^30,31 Even in patients with primary refractory MM after RVD or daratumumab containing induction regimens HDM and autologous HCT have increases PFS and OS when compared to those that did not undergo this procedure as reported in two retrospective analysis, which suggests that patients unresponsive to PI, IMIDS and monoclonal antibodies can still benefit from HDM therapy.^32,33

Table 2:

Initial experience with HDM in patients with myeloma prior to the advent of immune modulating drugs and proteosome inhibitors

Author/Ref	N (% Ref)	Mel Dose	Stem Cell Support	ORR (CR)	Deaths
Selby12	58 (30)	140 mg/m2	No	ORR (CR) Non Ref 67%(27%) Ref	17%
Barlogie13	23 (100)	80 to 140 mg/m2	Yes 14 patients all bone marrow	63% (ns)	0% with BMT 31% without BMT
Cunningham15	53 (0)	200 mg/m2	Yes bone marrow	100% (75%)	2%

Determining the efficacy of HDM in the context of modern induction regimen is more challenging. With three drug induction using bortezomib, thalidomide and dexamethasone (VTD) HDM increased the CR rate from 9% to 15% after 3 months and from 13% to 20% after induction with bortezomib, lenalidomide and dexamethasone (RVD).^18,19 The addition of the monoclonal antibody daratumumab during induction increased the depth of response further. In patients receiving dara VTD in the CASSIOPEIA studies the CR rate or greater went from 14% to 23%. In patients receiving Dara RVD in the GRIFFIN studies CR rates went from 19% to 27% after HDM.^20,21 Moreover, sustained MRD negativity of greater than 12 months at the 10⁻⁶ level was 44% in the Dara RVD arm of the GRIFFIN study and 35% in the Dara VTD arm of the CASSIOPEIA Study.^20,22 Similar high MRD negative rates of 49% at a 10⁻⁵ level was seen after HDM in patients receiving induction with carfilzomib, lenalidomide, dexamethasone and daratumumab (Dara KRD) in the MASTER trial.²³

These increases in response rates translated into significant increases in PFS for patients undergoing autologous HCT as part of their initial therapy from a median of 35 months to 47 months in the IFM trial and 46 to 67 months in the DETERMINATION trial.^18,19 With four drug induction the median PFS has not been reached after HDM and autologous HCT in either of the trials.^20,21

Melphalan exposure is a major determinant of overall response and PFS. Nath et al have performed the most comprehensive study of the impact of melphalan exposure on transplant outcomes to date. A total of 114 patients with myeloma receiving HDM had extensive pharmacokinetic studies performed. Median melphalan exposure was 12.84 mg l^–1 h (range, 4.9–24.6 mg l^–1 h). Melphalan AUC above the median was associated with increased risk of grade 3 or greater mucositis (HR=1.21, p=0.004) and improved median overall survival (8.50 years vs 5.38 years, p = 0.028. Melphalan exposure was not associated with increase response rates although there was a trend towards improved PFS in patients with higher melphalan exposure (median PFS: 2.41 years vs. 1.81 years; P = 0.089).²⁴

Reducing the Toxicity of HDM

1. Myelosuppression

Myelosuppression is the most common toxicity of HDM which can be mitigated through the infusion of either autologous or allogeneic hematopoietic progenitor cells obtained from mobilized peripheral blood or bone marrow. Use of colony stimulating growth factors such as filgrastim (GCSF) or sargromastin (GMCSF) as well as the use of mobilized peripheral blood stem cells has significantly reduced the time to neutrophil recovery as can be seen in Table 3.^{13–15,25–27}

Table 3:

Impact of Stem Cell Source and Growth Factor Support on Myelosuppression After HDM

Author/Ref	N	Regimen	Stem Cell Support	Growth Factor Support	Median days to ANC 500
Fermand25	63	Melphalan + TBI	Bone Marrow	No	15
Sheridan26	15	Various	Bone Marrow	Yes	11
Harrouseau27	132	Melphalan	Bone Marrow 81 Peripheral Blood 51	No	Bone Marrow 20 Peripheral Blood 13

Although the use of mobilized peripheral blood stem cells (PBSC) as well as post transplant hematopoietic growth factors now has reduced the days to neutrophil engraftment to a median of 10 days there is still a 4-day period of severe neutropenia as depicted in Figure 1. Patients receiving a CD34 + cell dose of less than 2 × 10⁶ CD34 per kg have longer periods of neutropenia and 5 × 10⁶CD34 cells per kg has been recommended as the optimal dose for a single transplant.²⁸ Shah el al performed a randomized trial of 10 vs 5 × 10⁶ CD34 per kg and did not see improvements in speed of engraftment of either neutrophils or platelets.³⁴ Landau et al explored the use of stem cell fractionation over various days and also failed to show a reduction in the days of absolute neutropenia.³⁵

Figure 1:

Kinetics of Neutropenia After HDM and Autologous HCT with Peg-filgrastim given on Day +1

2. Mucositis

Mucositis is the primary dose limiting toxicity of HDM and occurs almost universally at doses of 200 mg/m² or higher. Oral mucositis is associated with significant morbidity and increased use of resources including increased length of stay.^36,37 Multiple strategies have been developed to reduce the incidence and severity of mucositis which are described below.

Dose Modification:

Dose modification of melphalan has been employed empirically to reduce the risk of grade 3 or greater mucositis in patients with renal failure or patients over the age of 70, although prospective data to support this practice is not available.²⁸

Oral Cryotherapy:

Oral cryotherapy involves the use of either ice chips or ice cream to create local vasoconstriction in the oral mucosa reducing melphalan exposure to oral mucosal stem cells and thus reducing the risk of severe mucositis.^38–43 Lilleby et al. performed a small, randomized trial of oral cryotherapy (ice chips) or room temperature normal saline rinses 30 min before and for 6 h after high-dose melphalan therapy. A total of 40 patients were randomized and evaluated for the development of mucositis using the National Cancer Institute grading system. Patients randomized to cryotherapy experienced less grade 3–4 mucositis, 14 vs 74%, P=0.0005 as well as lower uses of narcotics and parenteral nutrition. Cryotherapy was not associated with a shorter length of stay or less weight loss.³⁸

Marchesi et al, performed a larger randomized trial (n=72) in a more homogeneous group of patients who received induction treatment with bortezomib, thalidomide and dexamethasone. The incidence of grade 3 or 4 mucositis was significantly lower in the cryotherapy group (5.6% vs 44.4%; p=0.0002) which resulted in lower rates of opioid use (2.8% vs 33.3%’ p=.001); need for parenteral nutrition (8.3% vs 38.9%; p=0.001) with a trend towards a shorter length of stay (18 vs 19 days, p=0.070).³⁹

A systematic review of 36 papers was conducted by the Mucositis Study Group of MASCC/ISOO. The group concluded that oral cryotherapy should be recommended to prevent oral mucositis in patients undergoing autologous HSCT with conditioning that includes high-dose melphalan.⁴⁰ There was no benefit in a longer duration of cryotherapy (7 hours) versus a shorter cryotherapy duration (2 hours) as demonstrated by Johanssen et al.⁴¹

In one of the largest trials of cryotherapy, Cho et al randomized 146 patients to receive either 2- or 6-h of cryotherapy. 84% of patients received a single dose of melphalan 200 mg/m² Overall, 62.3% of patients had grade 1–3 oral mucositis, and no patient experienced a grade 4 event. Severe mucositis was observed in 2.7% and 4.1% of patients receiving 2 or 6 h of cryotherapy, respectively. They showed that 2-hours of cryotherapy was at least as effective as 6-hour of cryotherapy in decreasing severe mucositis. A population pharmacokinetic model failed to showed that melphalan exposure according to their populations PK model was not associated with PFS or OS.⁴²

Amifostine:

Amifostine is an organic thiophosphate compound which protects normal tissue from the cytotoxic damage induced by anticancer agents without demonstrable evidence of tumor cell protection.⁴⁴ Table 4 summarizes representative trials of amifostine in myeloma as a intervention to reduce HDM mucosal and gastrointestinal toxicities.^45–47

Table 4:

Representative Studies Exploring Amifostine as a Strategy to Reduce Mucosal Toxicities

Reference / Study Type	N	WHO Grade 3–4 Mucositis	Comment
Randomized Controlled Trial46	Amifostine 43 Control 47	12% vs 33% p=.02	No reduction in parenteral nutrition or analgesic use
Concurrent Controls47	Amifostine 107 Control 114	27% vs 47%	Significant reductions in other GI toxicities (nausea 32% vs 86% and emesis 19% vs 53%)
Concurrent Controls48	Amifostine 21 Control 20	33% vs 65%	No difference in opioid use

Palifermin:

Palifermin is a recombinant human keratinocyte growth factor that was approved to reduce the incidence and severity of oral mucositis in patients undergoing autologous HCT after a myeloablative conditioning regimen. In the pivotal trial patients received 1200 cGy of TBI in addition to etoposide and cyclophosphamide which has a 90% incidence of moderate to severe mucositis versus 40–50% for HDM.⁴⁸

Blijlevens et al. performed a multicenter randomized placebo control trial of palifermin that included 281 patients undergoing autologous HCT with HDM. 57 patients received placebo; 115 patients received palifermin pre and post HDM conditioning and 109 received palifermin for 3 consecutive days before HDM. There was no statistically significant difference in maximum OM severity amongst the groups, with severe mucositis occulting in 37% of patients receingin placebo; 38% of patients receiving pre and post palifermin and 24% in patients who received palifermin only pre HDM. Likewise palifermin was not associated with any improvements in patient reported outcomes or resource utilization.⁴⁹ Based on these results palifermin is not recommended in the context of HDM conditioning.

Laser Light Therapy – Photo-modulation

Low-level light therapy (LLLT) was discovered in the 1960s and has been used to reduce pain and inflammation, prevent tissue damage and promote tissue repair.⁵⁰ LLLT consists of exposing tissue to low-levels of near infrared and red light at energy levels less than those used in other forms of laser therapy such as ablation, the mechanisms in which LLLT can reduce inflammation and enhance tissue repair are not understood.⁵⁰

Although extensive experience exists to demonstrate the potential role of LLLT in reducing pain and severity from oral mucositis in cancer patients receiving chemotherapy or radiation, there are few controlled trials in patients receiving HDM for myeloma.⁵¹ Franco et al. recently performed a meta-analysis and literature review of LLLT for prevention of oral mucositis after HCT, of more than 270 articles reviewed, only three randomized trials were identified, none of them focusing on HDM or myeloma.⁵²

Rodriquez et al. reported a large retrospective analysis of LLLT in 79 patients with multiple myeloma following HDM conditioning using a low-level indium gallium aluminum phosphate diode laser therapy (660 nm, 15 mW, 3.75 J/cm2, 10 s per point) from the beginning of the conditioning regimen up to day +2. Only 8 of the 79 patients developed severe mucositis.⁵³

Thus, although there is evidence that LLLT can reduce the severity and pain of oral mucositis after chemo and radiotherapy, there is no data to support are no controlled study to support the role of LLLT after HDM.

3. Symptom Burden

Patients receiving HDM suffer significant and prolonged increases in multiple symptoms after HCT, including sleep disturbance, sexual difficulties, and poorer physical, psychological, and social functioning, that interfere with activities of daily living and result in significant morbidity.^54–58 Figure 2 depicts the trajectory and the most common symptoms that occur after HDM.

Figure 2:

Time distribution and frequency of most common symptoms seen after HDM (adapted from references 34 and 35)

Peak symptom burden coincides with white blood cell count nadir and is associated with significant increases in interleukin 6 levels as demonstrated by Wang et al.⁵⁸ Melphalan exposure more than melphalan dose, as well as albumin and kidney function also were associated with higher symptom burden. Most interventions to reduce symptom burden focus on single symptoms (i.e narcotics for pain, antiemetics for nausea, etc) these are summarized in Table 5.

Table 5:

Interventions to Mitigate Specific Toxicities of High Dose Melphalan (adapted from reference 53)

Symptom	Impact on QOL after HDM	Intervention	Comments
Nausea and Vomiting	Major contributor to symptom burden if not appropriately controlled	5 HT3 Receptor Antagonists	Ondansetron, Palonosetron and Granisetron all effective in reducing n/v after HDM Refractory n/v requires further measures
Pain	Major contributor to symptom burden usually related to stomatitis and mucositis	Primarily symptomatic with oral or intravenous narcotics	Cryotherapy only established measure to reduce risk and severity of oral mucositis after HDM
Diarrhea	Major contributor to symptom burden usually related to chemotherapy toxicity but other etiologies (i.e infections) need to be ruled out)	Primarily symptomatic with antidiarrheals and antimotility agents. In cases with specific infectious agents (i.e Clostridium Difficille) specific therapy is indicated	Associated with higher melphalan exposure thus precision drug dosing could reduce the risk of severe diarrhea.

Scordo et al conducted a comprehensive assessment of HDM toxicities including assessment of changes in chemical gustometry. Among 41 patients, there were 185 individual grade 3 toxicities, distributed as follows: metabolic (29%), infections (23%), cardiovascular (19%), oral-GI (16%), hepatic (4%), dermatologic (3%), other toxicities (3%), pulmonary (2%), and renal (1%). Individual grade 3 oral-GI toxicities included nausea/vomiting (55%), diarrhea (24%), and oral mucositis (4%).⁵⁹ As in other studies peak symptom burden scores occurred at the time of neutrophil nadir between days 5 to 10. A variety of strategies are currently being explored to reduce symptom burden post HDM and autologous HCT in myeloma and are summarized below:

1. IL6 Blockade:

Initial studies of HDM symptom burden identified the temporal relationship with nadir, however, the underlying mechanism(s) of this process has not been well elucidated. One hypothesis is that cytokine dysregulation and inflammatory cascades may be important contributors to post trasnplant symptom burden.^58,60,61

Cytokines, including interleukin-6 (IL-6), are involved in the proinflammatory response after chemotherapy, and their levels correlate with symptom severity. This “cytokine theory” has been demonstrated in patients with lung and gastrointestinal cancers undergoing chemoradiotherapy.^60,61 IL-6, a pleiotropic cytokine with wide-ranging systemic effects, is secreted by various cells and plays a keyrole in the inflammatory process. IL-6 modulates the transcription of several liver-specific genes during inflammation, particularly C-reactive protein (CRP).⁶⁰ In a prospective longitudinal study of proinflammatory cytokines and symptom burden in patients undergoing autologous transplant, IL-6 levels correlated with symptom peak at nadir, as compared to baseline levels prior to transplant.⁵⁸

Siltuximab an anti IL6 antibody has been tested in MM patients in several trials both as a single agent and in combination.^63–66 Shah et al. conducted a phase II study of siltuximab 11mg/kg given on Day −7 and Day +21 of AHCT for patients with MM and AL which is the first study to showed benefit with IL-6 blockade on QOL as measured by the area under the curve (AUC) of the MD Anderson Symptom Inventory for Multiple Myeloma (MDASI-MM) was used to assess symptom burden through Day 30.⁶⁷ C-reactive protein (CRP) and IL-6 were measured at baseline, day −2, 0, +3, +7, +14, +21, and +30. The average MDASI-MM score per question at each time point ranged between 0–3 on a scale of 1–10, which represents an improvement from a historical control group where scores peaked at day 11 after AHCT with average scores up to 8 as seen in Figure 3.⁶⁷

Figure 3:

Symptom Burden Heat Map of Patients with MM Undergoing Autologous HCT with or without IL6 blockade.

2. Acupuncture

Deng et al. performed a blinded randomized trial of true vs sham acupuncture once daily for 5 days starting the day after HDM. chemotherapy. The primary outcome was assessed with the MD Anderson Symptom Inventory (MDASI) at baseline, during transplantation, and at 15 and 30 days post transplantation. Among 60 patients, true acupuncture produced nonsignificant reductions in overall MDASI core symptom scores and symptom interference scores during HCT (P = .4 and .3, respectively), at 15 days (P = .10 and .3), and at 30 days post HCT (P = .2 and .4) relative to sham. However, true acupuncture was significantly more efficacious in reducing nausea, lack of appetite, and drowsiness at 15 days (P = .042, .025, and .010, respectively). In addition, true acupuncture was associated with a lower risk of needing increases in pain medication use (odds ratio 5.31, P = .017).⁶⁸

3. Programmed Environmental Illumination

Valdimarsdottir et al performed a randomized trial looking at the effects of a Programmed Environmental Illumination (PEI) program on symptom burden post auto HCT for myeloma. A total of 44 eligible patients were randomly assigned to one of two PEI conditions involving delivery of either circadian active bright white light (BWL) or circadian inactive dim white light (DWL) from 7 to 10 am daily during hospitalization. There were no differences in depression scores at baseline between both groups, a longitudinal linear mixed model analysis revealed a significant interaction between time of assessment and light condition indicating that PEI prevented the development of depression during hospitalization, with effects reaching significance by the third day of engraftment. At the third day of engraftment, 68.4% of the participants in the DWL comparison condition met the criteria for clinically significant depression compared to 42.1% in the BWL (p0.03).⁶⁹

4. Music Therapy

Based on initial work performed by Cassileth et al,⁷⁰ Bates el al performed a randomized trial of music therapy in patients undergoing high-dose chemotherapy and auto HCT. A total 40 myeloma patients were randomized, 29 of which received HDM followed by autologous HCT. Sixteen of these patients were randomized to receive either interactive music therapy with a board-certified music therapist or no music therapy. The music therapy arm received 2 music therapy sessions on days +1 and +5. Amongst all patients who received music therapy there was a slight increase in nausea by day +7 compared with the no music therapy patients, similar pain scores, but significantly less narcotic pain medication use (median, 24 mg versus 73 mg; P = .038).⁷¹

5, Digital Life Coaching

Digital life coaching uses digital platforms (I,e applications, text messages and phone calls) to help patients achieve wellness-related goals through a longitudinal partnership with a certified life coach. Banerjee et al. recently reported the results of a randomzied phase II trial of DLC in patients with myeloma undergoing HCT after HDM conditioning. A total of 50 patients from 70 screened were randomly assigned to either DLC (n=21) or usual care (n=29). There was a significant difference in regards to (physical QOL) between arms after engraftment, with a mean difference of +4.9 points favoring DLC. Among participants who completed the final and peri-engraftment assessments, 86% of the DLC arm reported meaningful improvement in physical QOL vs 67% in the control arm. No significant differences were observed in mental QOL or emotional distress. The DLC arm had significantly lower distress than the control arm at 3 months post HCT with only 5% of DLC patients reporting elevated distress versus 22% of patients randomized to usual care. Patients had postivie perceptions of the study interventions with 41% of patients reponding that they would likely recommend DLC to other patients undergoing HCT.⁷²

Increasing Efficacy of HDM

Figure 4 depicts our current understanding of the mechanism of action of melphalan. Since melphalan has to be actively incorporated into the cells, its activity will be directed related to how rapidly it gets metabolized into its inactive metabolites as well as the total dose administered.⁷³

Figure 4:

Mechanism of Action of Melphalan (adapted from ref 119)

The current standard is to dose melphalan as a fixed dose of melphalan per m² of body surface area. In obese patients there has been significant variation in how BSA was calculated with either actual weight, ideal body weight or adjusted body weight. In 2014, the American Society of Blood and Marrow Transplantation recommended that all patients be dosed according to actual weight.⁷⁴ Vogl et al. reported that neither obesity or severe obesity were associated with worse outcomes in myeloma patients but that these patients were more likely to not receive the full dose of 200 mg/m² using actual body weight, and that the degree of dosing variability was significant but did not impair outcomes.⁷⁵ Thus, although melphalan exposure is obviously dependent on melphalan dosing this is not the only factor and strategies to optimize melphalan exposure need to be prospectively studied.

In a pilot study, Shah et al demonstrated the feasibility of dosing melphalan using pharmacokinetic adjustments. Patients received an initial dose of propylene glycol free melphalan of 100mg/m² on day −2 prior to HCT with serial plasma melphalan measurements obtained. Phoenix WinNonlin v6.4 was used to determine the dose required to achieve a target melphalan exposure of 13.5mg/L*h (+/−1). On day −1, patients received the balance of the calculated dose with samples collected at the same timepoints to calculate the total melphalan exposure. In the first 15 patients, 13 (87%) achieved the targeted melphalan exposure. Randomized trials will be needed to establish personalized melphalan dosing as the standard of care in multiple myeloma.⁷⁶

1. Optimizing Melphalan Exposure – Increasing the dose

The most commonly used dose of melphalan is 200 mg/m²; in patients over the age of 70 or patients with severe end stage renal disease guidelines suggest dose reduction to 140 mg/m².²⁸ Although dose reductions for patients with renal impairment have been justified due to increase toxicity, the data for reducing the dose of melphalan to 140 mg/m² is less convincing particularly in the context of the long term follow up of DSMM XIII in which NDMM patients between 60 and 75 years of age randomized to receive melphalan 140 mg/m² as conditioning regimen for autologous HCT had similar PFS and OS as patients not undergoing this procedure.⁷⁷

Moreau et al. showed that doses of 220 mg/m² could be administered without undue gastrointestinal toxicity.⁷⁸ Using amifostine as a cytoprotectant it is possible to increase the dose of melphalan up to 280 mg/m².⁷⁹ Bensinger et al. performed a randomized study of melphalan 200 mg/m² vs 280 mg/m² as a preparative regimen patients with multiple myeloma. The higher dose of melphalan was associated with higher rates of near CR or greater (39% vs 22%, p=0.03) and PR or greater (74% vs 57%, p=0.04). There were no differences in regards to PFS and OS with median PFS of 3.5 years for melphalan 280 mg/m² and 2.7 years for melphalan 200 mg/m². Amifostine was well tolerated and effective in preventing grade 4 regimen-related toxicities (0%) and minimal grade 3 GI or mucosal toxicities.⁸⁰ No further randomized trials of higher doses of melphalan have been explored.

Auner et al performed a registry analysis of 1964 patients to determine the impact of melphalan dose on autologous HCT outcome in myeloma.⁸¹ In this retrospective analysis patients transplanted in less than partial response had better PFS if they received 200 mg/m2 versus 140 mg/m² (adjusted hazard ratios for melphalan 200 mg/m² versus melphalan 140 mg/m² was 0.56). In contrast, in patients with a partial response or better there was a overall survival benefit for melphalan 140 mg/m² (adjusted hazard ratio: 2.02). These types of discrepancies can only be addressed by understanding individual melphalan exposure.

Tandem ASCT refers to the re-administration of HDM with autologous HCT support within 6 months of the first HDM therapy. Multiple clinical trials have explored the impact of tandem HCT on myeloma outcomes and in general tandem HCT has been shown to be superior in patients with high risk disease or patients who have failed to achieved a VGPR or greater.^82–86 The BMT CTN 0702 had in an initial analysis failed to show any benefit for tandem transplant in patients with high risk myeloma.⁸⁷ However, in a subsequent long term follow up with the intent to treat analysis 6 year PFS was the same amongst all 3 groups (tandem Auto=43.9 %; auto + RVD consolidation 38.7 % and Auto + lenalidomide maintenance 40.9%. When the analysis is done according to treatment actually received there was a benefit in 6 year PFS for tandem auto HCT when compared to the other arms 49.4%, 39.7% and 38.6% for tandem auto; auto followed by RVD consolidation and auto followed by len maintenance only respectively (p=0.01). This benefit was particularly marked in patients with high risk features 43.6% and 26% for tandem auto and auto followed by lenalidomide only respectively (p = 0.03).⁸⁸ In the context of four drug induction and prolonged consolidation or maintenance with lenalidomide combinations the benefit of tandem HCT has not been tested and fewer patients are undergoing this procedure worldwide.⁸⁹

2. Optimizing Melphalan Exposure – PK Directed Melphalan Dosing

Due to variable pharmacokinetics in a population, fixed dose dosing leads to differences in exposure. Parts of this variability is caused by patient characteristics such as kidney function and body weight including obesity. These characteristics however are not taken into account when dosing patients. As a result, the exposure to melphalan varies between patients, Thus, a portion of patients are underdosed, while another portion of patients are potentially overdosed. This in turn may lead to decreased efficacy (relapse, non-engraftment) or increased toxicity (mucositis, diarrhea, nausea and anorexia). In order to attain optimal and predictable melphalan exposure in all patients, an individualized dosing regimen is necessary that accounts for known factors that impact PK. Some population PK-studies are available in literature.^{24, 90,91} Three of these focus on HDM in the context of HCT but are based on relatively small cohorts of patients.

Shah et al. embarked on a prospective study to measure melphalan exposure on patients receiving a melphalan containing regimen. Melphalan concentration samples were drawn at 5, 15, 30, 40, 75, and 150 minutes after the end of each infusion. These time points were arbitrarily chosen based on clinical convenience, no optimal sampling strategy was performed. Melphalan was measured using a previously described mass spectrometry assay. Data was prospectively collected on 185 patients with myeloma receiving HDM, as with other studies they observed a significant variation (up to five fold) in melphalan exposure that can occur with fixed dosing and saw a trend towards improved outcomes in patients receiving an exposure of greater or equal to 14 melphalan exposure units (MEUs). These data together with melphalan exposure measurements in more than 300 other patients with a variety of hematologic malignancies allowed us to develop a population pk model that not only allowed us to infer potential; melphalan exposure in patients in which melphalan PK measurements were not done but could be used to “personalize melphalan” dosing and prevent the current variability in melphalan exposure seen with fixed dosing.⁹²

3. Combination with Other Agents – Busulfan

Busulfan based combinations are the second most commonly used conditioning regimen for autologous HCT in MM, with busulfan melphalan (BUMEL) the most commonly used combination.

Bashir et al, reported on a randomized trial of HDM vs the combination of busulfan and melphalan in 205 patients with myeloma. Median PFS was 64·7 months (32·9–64·7) with BUMEL versus 43·5 months (19·9–not estimated) with melphalan alone (hazard ratio 0·53 [95% CI 0·30–0·91]; p=0·022). Grade 2–3 mucositis was observed in 74% of patients in the BUMEL group versus 14% of patients in the melphalan alone group. In patients with high risk disease, the 3-year PFS rates were 69% for BUMEL and 41% for HDM alone. Median PFS was 44.7 vs 25.7 months in the BUMEL and HDM alone arms, respectively (P = .044) The 3-year OS rates were 90% and 87% in the BUMEL and HDM arms alone respectively.^93,94

In a recent meta-analysis of ten studies with a total of 2855 MM patients, Patients that received the combination of busulfan and melphalan had a longer PFS (HR 0.77; 95% CI 0.67~0.89, P = 0.0002) but similar OS (HR 1.08; 95% CI 0.92~1.26, P = 0.35) compared with those having received high dose melphalan alone. The combination of busulfan and melphalan was associated with less GI toxicity but more mucositis and infections with no increase in hepatic toxicity.⁹⁵

Gemcitabine can inhibit DNA repair from alkylator induced damage.⁹⁶ Nieto et al. explored the combination of gemcitabine at 1875 mg/m2 for 2 days, followed by busulfan for 4 days and melphalan (60 mg/m2 per day for 2 days) and compared their post transplant outcomes with a group of similar patients who received HDM at a dose of 200 mg/m². A total of 74 patients received gemcitabine, busulfan, and melphalan and compared to 184 patients concurrent controls. Stringent CRs were more frequent in the gemcitabine, busulfan and melphalan group (24.6% vs 12.6%). The gemcitabine, busulfan, and melphalan cohort had significantly longer median PFS ((15·1 months vs 9·3 months) and median OS (37·5 months vs 23·0 months). The combination was associated with grade 3 or greater mucositis in 12 of 74 patients and three treatment related deaths.⁹⁷

4. Combination with Other Agents – Proteosome Inhibitors

Bortezomib has been shown in vitro to synergize with melphalan enhancing myeloma cell death.⁹⁸ Multiple trials have been performed both in the upfront and relapsed setting.^99–102

Based on these data the IFM group embarked on a randomized trial comparing bortezomib (1 mg/m2 intravenously [IV]) on days −6, −3, +1, and +4 and melphalan (200 mg/m2 IV) on day −2 to 200 mg/m2 of high dose melphalan. A total of 300 patients were randomized. Stringent CR + CR rates at day 60 posttransplant (primary end point) were similar in both arms 22.1% vs 20.5%; p= .844). Median PFS was also not significantly different 34.0 months vs 29.6 months p = .244) with an estimated 3-year OS of 89.5% in both arms.¹⁰³

Carfilzomib has also been combined with melphalan as part of a conditioning regimen for autologous HCT in myeloma. Costa et al. reported a phase I/II study of carfilzomib in combination with HDM. The maximum tested dose of 27 mg/m2 (on day −3) and 56 mg/m2 (on day −2) was used in phase 2. The rate of very good partial response after CAR -MEL therapy (n = 44) was 59.2%, compared with 13.7% before CAR - MEL therapy. Among patients starting maintenance therapy (n = 27), 12-month progression-free survival was 66.7% and 12-month overall survival was 88.1%.¹⁰⁴

5. Combination with Other Agents – Bendamustine

Bendamustine hydrochloride is a cytotoxic that combines both antimetabolite as well as alkylating activity. It results in enhanced single-strand and double-strand breaks in DNA and can synergize with other antineoplastic agents including alkylators.¹⁰⁵ Bendamustine has known single agent activity in myeloma and has been successfully combined with IMIDs and proteosome inhibitors.

In the context of HDM, Mark et al, performed a phase 1 dose escalation study combining melphalan and bendamustine as a conditioning regimen for myeloma.¹⁰⁶ The combination of bendamustine 225 mg/m² with HDM 200 mg/m² was subsequently studied in a phase II trial performed by Farag et al. In this study, 120 NDMM patients were randomized to receive either HDM at a dose of 200 mg/m² or bendamustine 200 mg/m² in combination with melphalan 100 mg/m² at days −2 and −1. The rate of sCR/CR after ASCT was higher with bendamustine and melphalan treated patients (70.0% vs. 51.7%; p = 0.039). No differences in PFS or OS were reported. The results of that trial as well as other studies exploring the combination are summarized in Table 7.^107–110

Table 7:

Other strategies explored or with potential that have not been explored further.

Strategy	Rationale	Results
Add Arsenic Trioxide and Vitamin C114	Synergy and increase oxygen free radicals by reducing gluthathione	No increase In ORR
Combine with radioimmune conjugates Holmium-DOTMP or Samarium115–116	Synergy between radiation and melphalan deliver high doses of radiation to the marrow space	Increase ORR with Holmium DOTMP but also increased renal toxicities Samarium well tolerated but no phase III trials contemplated
Melflufen117	Peptide drug conjugate, which relies on the lipophilicity of the drug facilitating passage across the plasma cell membrane, and aminopeptidases, overexpressed within myeloma cells, cleave its peptide bonds resulting in increasing exposure of melphalan in myeloma cells while sparing normal cells.	Never explored in the high dose setting. Retired from the market in the United States.
Parp Inhibition118	Inhibitors of DNA repair should enhance melphalan effects	Preclinical data suggests synergy with melphalan no clinical trial data

Novel Melphalan Formulations – Propylene Glycol Free Melphalan (Evomela)

The original melphalan intravenous formulation required the use of propylene glycol as a solvent. Propylene glycol has been associated with a variety of adverse events, such as, nausea, vomiting, renal toxicity and hypersensitivity reactions. Evomela is the first propylene glycol-free formulation of melphalan (PGF-Mel) that was approved in North America.

After the reconstitution of melphalan with propylene glycol–containing solution, impurities can develop within 30 minutes, and hydrolysis of melphalan begins almost immediately requiring that the solution be used within 60 minutes of preparation. In contrast, PGF-Mel can be reconstituted in saline solution and can is stable for up to 1 hour at room temperature or for up to 24 hours at temperature between 2° to 8°C without hydrolysis¹¹¹.

Hari et al. recently reviewed 14 clinical studies of PGF - Mel in patients with myeloma or AL amyloidosis undergoing autologous HCT. Although some studies showed potential improved response rates for patients receiving PGF – Mel, most studies showed equivalent response rates, but lower rates of mucositis and neutropenic fevers.¹¹²

Melphalan exposure with PGF-Mel is still highly variable although potentially less variable than with non PGF-Mel.¹¹³ However, the increased stability of PGF – Mel allows for exploration of PK directed therapy and novel prolonged infusion schedules as currently being explored (NCT03417284).

Other Strategies

Table 8 summarizes strategies that have been explored to enhance the efficacy of HDM. These strategies have explored potential synergies through inhibition of DNA repair (i.e addition of arsenic trioxide or gemcitabine).^97,114 Combination with radio immune conjugates (i.e. Samarium and Holmium).^115,116

Melflufen is a peptide drug conjugate, which relies on the lipophilicity of the drug facilitating passage across the plasma cell membrane. Enzymes that are overexpressed within myeloma cells, cleave its peptide bonds resulting in increasing exposure of melphalan in myeloma cells while sparing normal cells.¹¹⁷ Unfortunately, this drug has not been explored in higher doses and is currently off the market in the United States.

Summary and Future Considerations

It has been over 3 decades since HDM has been used as the major and mostly only component of high dose conditioning regimens for myeloma. In those 3 decades, we have seen major changes in both induction and maintenance therapies for this disease that have dramatically improved outcomes. Notwithstanding, HDM consolidation remains a mainstay of treatment and in many areas of the world the most cost effective strategy to achieve deep remissions.

Unfortunately, although the myeloma research community has devoted some time and effort to improving the safety and efficacy of HDM these have not resulted in major changes in how this treatment is delivered.

Over the next few years it is essential that the promising findings of IL6 blockade, acupuncture and programed light exposure as strategies to reduce HDM toxicities be further explored. In addition, prospective trials combining HDM with other agents need to be explored.

Table 6:

Results of Phase II Trials of Bendamustine/Melphalan Combinations as Conditioning Regimen for Autologous HCT in Myeloma

Author/ Reference	N	%ORR/CR	PFS	Comments
Gomez107	18 NDMM 17 RRMM	100%/44% 94%/59%	48 months 45 months	Mel 200 Benda 225
Martino108	32 NDMM	HCT 1 81%/47% HCT 2 91%/63%	2 year PFS 79%	Mel 140 Benda 200 All as HCT 2 of a tandem transplant strategy
Farag109	Mel/Benda 60 Mel 60	CR 70% CR 51%	1 year PFS 95% 1 year PFS 91%	Phase III Trial NDMM Mel 200 Benda 200 1 toxic deaths and 3 grade 4 renal toxicity in Benda arm
Scott110	65 NDMM	CR 40%	3 year PFS 57%	Benda Etoposide Cytarabine Melphalan

Practice Points.

• High dose melphalan remains the standard of care for autologous HCT conditioning in myeloma.

• Established measures to mitigate toxicities include: 1. Cryotherapy with ice chips or ice cream to reduce mucositis; 2. Optimal nausea prevention with 5HT3 receptor antagonists with or without steroids; 3. Use of peripheral blood stem cells and hematopoietic growth factors to enhance neutrophil recovery; 4. Dose modification of melphalan in older patients or patients with renaly dysfunction.

• Patients should be encouraged to participate in clinical trials exploring strategies that can improve outcomes after high dose melphalan

• Melphalan exposure can be measured and impacts toxicities.

Research Agenda.

• Explore new melphalan formulations or combinations.

• Perform randomized trials to confirm the impact of IL6 blockade on post HCT symptom burden

• Perform randomized trials to ascertain the impact of PK directed melphalan dosing.