Efficacy of Melflufen in Patients with Relapsed/Refractory Multiple Myeloma and Mutated or Deleted TP53

Compounds and stock solutions

Melflufen (Oncopeptides AB) was freshly prepared in DMSO before application. Melphalan (cat. no. M2011, Sigma-Aldrich) stock solution was prepared in acidified ethanol. Cyclophosphamide (N,N-bis(2-chloroethyl)-4-hydroperoxy-2-oxo-1,3,25-oxazaphosphinan-2-amine,TRC Canada) was freshly prepared in DMSO.

Ex vivo drug sensitivity testing of myeloma bone marrow samples

Viably cryopreserved BM mononuclear cell (BM-MNC) samples (n=24) from 23 patients with MM were obtained from the Finnish Hematology Registry and Biobank (Table 1). The samples were collected after informed consent using approved protocols and in accordance with the Declaration of Helsinki. The study was approved by an ethical committee of the Helsinki University Hospital (Ethical Committee Statement 303/13/03/01/201, latest amendment 7 dated 15 June 2016; latest HUS study permit HUS/395/2018 dated 13 February 2018; Permit numbers 239/13/03/00/2010, 303/13/03/01/2011). For ex vivo drug sensitivity testing, BM-MNCs were thawed, resuspended in conditioned medium from the HS-5 human BM stromal cell line, and seeded on 96-well plates. Resuspended drugs were added to the wells and cells incubated at 37°C for 72h. Labeled antibodies specific for CD138 and CD38 (BD Biosciences) were used to detect plasma cells, while Annexin V and 7AAD (BD Biosciences) were used to detect apoptotic and dead cells, respectively. The cells were analyzed on the iQue PLUS flow cytometer (Sartorius). Additional details are included in the Supplementary Materials and Methods.

Single-cell gene expression analysis

For single cell RNA sequence (scRNAseq) analysis, the above-mentioned 24 BM-MNCs were sorted based on CD138+ expression and libraries prepared with 10x Genomics reagents and sequenced on the Illumina NovaSeq 6000 instrument. The samples were grouped based on ex vivo drug sensitivity to melflufen and melphalan. Cell clusters were identified based on gene expression patterns and differential gene expression (DGE) and gene set enrichment analysis (GSEA) performed on the plasma cell clusters of the highly sensitive and low sensitive samples for both drugs (see Supplementary Materials and Methods for details). Data link NCBI GEO (http://www.ncbi.nlm.nih.gov/geo) accession number GSE263201.

Cell lines, cell culture, and cytotoxicity

The AMO-1 cell line was obtained from DSMZ (Braunschweig, Germany). Generation and characterization of TP53 deficient AMO-1 clones is described in Munawar et al., 2019.³¹ Cell viability was determined after 72h incubation using the CellTiter-Glo® 2.0 assay (Promega) or AlamarBlue^TM reagent (Invitrogen) according to the manufacturers’ protocols.

DNA damage signaling and cell apoptosis analyses

AMO-1 TP53wt and AMO-1 TP53^−/− cells were treated with melflufen, melphalan or cyclophosphamide at the indicated concentrations and time points. H2AX phosphorylation at Ser139(γ-H2AX) for DNA damage signal was analyzed using flow cytometry. For apoptosis analysis cells were stained for Annexin V APC and PI (Invitrogen). All samples were analyzed with the FACSCanto II flow cytometer (BD Biosciences). Data were analyzed using FlowJo software (BD Biosciences).

JC-1 mitochondria membrane potential assay

AMO-1 TP53wt and AMO-1 TP53^−/− cells were treated with melflufen, melphalan or cyclophosphamide at the indicated concentrations and time points. The JC-1 mitochondrial membrane potential assay was performed according to the manufacturer’s protocol (MitoProbe™ JC-1 Assay Kit for Flow Cytometry (M34152), Molecular Probes).

RNA preparation and RNA sequence analysis

AMO1 TP53wt cells or the TP53^−/− knock out clonal subline AMO-1 TP53^−/− were treated with melphalan or melflufen at different concentrations and timepoints. RNA was isolated according to the RNeasy Plus protocol, quantified with Nanodrop (Thermo Scientific), and quality assessed with the Agilent 2100 Bioanalyzer. DNA libraries were prepared using the TruSeq Stranded mRNA Library Preparation Kit (Illumina) and sequencing performed on the NextSeq-500 platform (Illumina). Sequences were mapped to the human genome (GRCh38.p14) with STAR aligner and gene level based read counts generated with RefSeq. Differential gene expression (DGE) analysis was performed using DESeq2 with log-fold change shrinkage using beta Prior. Cluster profiler was used for gene set enrichment analysis (GSEA) and for KEGG pathway analysis. R was applied to generate heatmaps. Data link NCBI GEO (http://www.ncbi.nlm.nih.gov/geo) accession number GSE254959.

Statistical analyses of the OCEAN trial data

Primary analysis of tumor response and progression-dependent endpoints of patients enrolled in the OCEAN trial (NCT03151811) were based on response assessments by an Independent Review Committee. All tumor response and progression-dependent endpoints were assessed using the IMWG Uniform Response Criteria (IMWG-URC). PFS was defined as time (months) from date of randomization to either confirmed disease progression or death due to any cause. ORR was defined as the proportion of patients for whom the best overall confirmed response was stringent complete response (sCR), complete response (CR), very good partial response (VGPR), or partial response (PR). Posthoc analyses were performed to examine associations between PFS and chromosome 17 deletion status and/or TP53 gene mutation status. Furthermore, another post-hoc analysis was carried out to study the patient responder status in the chromosome 17 deletion population.

Assessment of del(17p) status

The del(17p) status had been determined from bone marrow at baseline level (entry) of the clinical study by interphase Fluorescence In-Situ Hybridization (iFISH) in accordance with Good Clinical Practice (GCP). No cutoff applied, all patients with del(17p) cytogenetics had been included into the analyses. The presence of del(17p) in plasma cells varied between 8-90%. The del(17p) status of the patient samples used for the ex vivo drug sensitivity testing was determined using iFISH as previously described.³²

Assessment of TP53 mutation status

TP53 mutation status of OCEAN trial patients was determined by next generation sequencing (NGS) using Almac Genomic Services. Briefly, DNA was extracted from BM derived CD138+ cells and processed using Illumina TruSeq Custom Amplicon Kit Dx. A custom TP53 oligo panel detecting pan-cancer single nucleotide polymorphisms was applied and the samples sequenced on the Illumina MiSeq instrument. All detected single nucleotide variants classified. The TP53 mutation status of the patient samples used for the ex vivo drug sensitivity testing was determined using exome sequencing of DNA from CD138+ cells as previously described.³²

Detailed materials and methods are provided in the supplementary information.

Melflufen targets MM plasma cells ex vivo regardless of TP53 mutation status

To assess the efficacy of melflufen in high-risk myeloma ex vivo, we conducted flow cytometry-based drug sensitivity testing of 24 myeloma BM-MNC samples obtained from 23 patients with various clinical and molecular backgrounds (Figure 1A). The cohort included samples from newly diagnosed and relapsed patients. After treatment of the BM-MNCs with melflufen or control drugs for 72h, the viability of the CD138+CD38+ cells was assessed by Annexin V and 7AAD staining. All samples showed reduced viability after exposure to melflufen, but sensitivity varied between samples from extremely sensitive to less sensitive, which was reflective of the heterogeneity of the patients (Figure 1B). For subsequent analyses, the samples were divided into different melflufen sensitivity groups based on the melflufen DSS values of the samples: high (HS, DSS>40; n=8), intermediate (IS, 40<DSS<31; n=8), and low (LS, DSS>31; n=8). Confirming our earlier observations, a significant indicator of melflufen sensitivity was disease stage as all but one of the HS samples were from RRMM patients.²⁶ A gain of chromosome 1q (+1q) and loss of chromosome 13 (−13/−13q) were the most frequent abnormalities observed in 22/24 samples individually or in combination. Statistical analysis showed tendencies of samples with +1q to have higher sensitivity to melflufen, and the opposite trend for samples with −13/−13q although significance was not obtained due to the limited sample number. In addition, 4 of the 8 LS samples and only 1 of the 8 HS samples had t(4;14), while 3 of the HS, 1 of the IS and 2 of the LS samples were positive for del(17p). TP53 mutation status was available for 14 of the 24 tested samples. Of the 5 HS samples with TP53 sequence data, 3 were mutation positive, while TP53 mutations were not detected in any of the 5 LS samples with sequence information. These results suggested that melflufen was highly active despite the presence of TP53 mutation or del(17p). At the same time, comparison of the ex vivo drug sensitivity of samples to melflufen, melphalan, and cyclophosphamide clearly showed the superior potency of melflufen in TP53wt, del17p, and del17p + TP53mut patient samples (Figure 1C). For instance, the sensitivity to melflufen was significantly higher in the TP53wt (melflufen: n = 8, melphalan: n = 8, cyclophosphamide: n = 2) and del(17p) + TP53mut groups (melflufen: n = 4, melphalan: n = 4, cyclophosphamide: n = 2). Although the threshold for statistical significance was not achieved in the del(17p) cohort (n = 2 for all drugs), the trend was similar.

• Download figure
• Open in new tab

Figure 1. Melflufen ex vivo drug sensitivity testing results from the 24 myeloma bone marrow samples.

A) Patient sample characteristics. B) Melflufen dose response curves for the 24 MM patient samples. Samples having both del(17p) and TP53 mutation (black), having only del(17p) (red), having only TP53 mutation (blue), other (grey). C) Drug sensitivity scores of samples with different molecular background for three alkylators – melflufen, melphalan, and cyclophosphamide. P-values are indicated above the brackets, with the corresponding comparison. * – p<0.05, *** – p<0.001. C) Bar plot representing normalized enrichment scores for selected gene sets derived from the most differentially expressed genes in the plasma cell populations between the melflufen high sensitive (HS) and low sensitive (LS) groups.

To investigate molecular differences between groups with different sensitivity to melflufen, we analyzed the transcriptomic profiles of the plasma cell populations, obtained by scRNA sequencing, and conducted differential gene expression (DGE) and gene set enrichment (GSEA) analyses. GSEA of genes detected in HS and LS samples showed a decrease in p53 downstream pathways and targets in the HS group, and, at the same time, higher enrichment of pathways associated with DDR genes BRCA1, ATM, and CHEK2 (Figure 1D).

Loss of TP53 compromises multiple pathways

To better understand the mechanisms responsible for the superior activity of melflufen compared to other standard of care alkylators observed from the ex vivo drug sensitivity testing analysis, we used the myeloma cell line AMO-1 with intact WT TP53 or with deletion of both alleles (TP53^−/−). To validate the functional significance of these cell line models and their relevance with the patient samples, we investigated transcriptomic alterations caused by mutation of TP53 by comparing baseline gene expression of TP53^−/− vs. TP53wt patient samples and cell lines. This analysis yielded 1485 differentially expressed genes in patient samples (Supplementary Table S1) and 5357 of those in cell lines (Supplementary Table S2). In both comparisons, there were 252 similarly deregulated genes, such as IL6R, JUN, DUSP5, AREG, and FOS (Supplementary Table S3). Subsequently, gene set enrichment analysis (GSEA), when applied to these cohorts of genes, gave us multiple gene sets, associated with TP53 and cell cycle progression, as well as pathways, associated with signaling and DNA damage (Supplementary Figure S1). These results confirmed a similar impact of TP53 dysfunction in the patient samples and cell line models and illustrated the exploitability of TP53 loss for treatment purposes by a drug such as melflufen.

Melflufen is cytotoxic and induces early apoptosis in cells without functional p53

Earlier studies had demonstrated that deletion of one or both alleles of TP53 in AMO-1 cells by CRISPR/Cas9 resulted in significant loss of sensitivity to melphalan or doxorubicin.^31,33 In support of the earlier report,³³ double mutant TP53^−/− AMO-1 cells displayed profound loss of sensitivity to melphalan with EC50 values close to 20µM (Figure 2A, B, Supplementary Table S4). However, melflufen was highly active towards both TP53wt and TP53^−/− AMO-1 cells and compared to melphalan exhibited 30-fold more inhibition of cell viability with EC50 0.6μM. Importantly, melflufen-induced apoptosis as measured by Annexin V was similar in TP53wt and TP53^−/− cells (Figure 2B).

• Download figure
• Open in new tab

Figure 2. Cytotoxicity and apoptosis detection in the AMO-1 TP53 isogenic cell lines.

A, B) Melflufen and melphalan cytotoxicity dose effect curves in parental AMO-1 cells (TP53wt) and in its bi-allelic-double mutant TP53 gene harbouring clonal cell line (TP53^−/−) using either the alamarBlue metabolic activity assay (A) or the apoptosis measurement Annexin V/propidium iodide staining (B). The curves represent three independent experiments, error bars reflect standard deviations. C) Melflufen, melphalan or cyclophosphamide induced early apoptosis in AMO-1 TP53wt and AMO-1 TP53^−/− isogenic cell lines. Cells were treated with the indicated drug concentrations and time points, melflufen (red), melphalan (green), or cyclophosphamide (blue). Early apoptosis was detected with Annexin V and PI staining using flow cytometry (Annexin V positive, PI negative). D) Both early and late apoptosis were measured in AMO-1 TP53wt and AMO-1 TP53^−/− isogenic cell lines after 24h treatment with the indicated drugs and drug concentrations (late apoptotic cells: Annexin V positive, PI positive).

Earlier studies have shown that melflufen has more efficient kinetics compared to melphalan,^34,35 we therefore investigated the induction of apoptosis in TP53wt and TP53^−/− cells after treatment with melflufen or melphalan. Even at 2h, induction of apoptosis was apparent in both TP53wt and TP53^−/− cells after melflufen but not melphalan treatment (Figure 2C). At 4h, early induction of apoptosis was as prominent in the mutant TP53^−/− AMO-1 cells as in TP53wt cells. In contrast to melflufen, melphalan and cyclophosphamide failed to induce early apoptosis (Figure 2D), although the apoptotic signal increased later and at higher IC50 values (55μM and 15μM, respectively) compared to melflufen (IC50 1μM) (Figure 2D). Overall, these results demonstrate that melflufen treatment resulted in complete inhibition of viability and induction of apoptosis at sub-to low micromolar concentrations, and indicate that, in contrast to melphalan, melflufen remains active despite compromised p53 function.

Melflufen induces DNA damage and mitochondrial membrane disruption in TP53 deficient cells

As the main cellular target for alkylating agents is genomic DNA, we aimed to gain insight into the kinetics of DNA damage induction for cells harboring TP53wt or TP53^−/− cells. Cells were treated with either melflufen, melphalan or cyclophosphamide and induction of early DNA damage was measured by phosphorylation of H2AX (γH2AX (ser139)). Melflufen treatment induced yH2AX signal after 30 min in the TP53wt cells in a concentration dependent manner, while in the TP53^−/− mutant cells maximal DNA damage induction occurred already at the lowest concentration. TP53^−/− cells exhibited a higher baseline level of yH2AX which is in line with loss of p53 protein function in the feed-back loops for DNA replication and repair protein functions.¹⁰ Exposure to melphalan resulted in little induction of early DNA damage in TP53^−/− cells while in TP53wt cells DNA damage induction was apparent only at a 10-fold higher concentration compared to melflufen. Similarly, there was little change in γH2AX levels in TP53^−/− cells treated with cyclophosphamide, and the yH2AX signal induction by cyclophosphamide in the TP53wt cells appeared later in comparison to the other drugs (Figure 3A).

• Download figure
• Open in new tab

Figure 3. DNA damage and mitochondrial membrane potential detection in AMO-1 TP53wt and TP53 knock out cells.

A) Melflufen, melphalan or cyclophosphamide induced DNA damage in AMO-1 TP53wt and AMO-1 TP53^−/− isogenic cell lines. Early DNA damage was detected as yH2AX signal by flow cytometry after the indicated treatment concentrations and time points. yH2AX: phosphorylated histone protein 2AX. B) Melflufen, melphalan or cyclophosphamide treatment induced mitochondrial disfunction in AMO-1 TP53 isogenic cell line. Melflufen, melphalan or cyclophosphamide treatment were applied in the indicated concentrations, and after 1, 2, 3 and 4h time points mitochondrial function of the treated cells was measured using JC-1 mitochondrial membrane potential assay. PI: propidium iodide.

We further investigated whether other cellular targets such as mitochondria²⁷ are affected by melflufen with similar efficacy in the TP53wt and TP53^−/− cells. Mitochondrial functionality by membrane integrity measurements of the AMO-1 isogenic cell lines showed that melflufen induced a drop of membrane potential in the TP53wt cells within 2 hours after treatment, and the effect on mitochondria function was even more pronounced in the mutant TP53^−/− cells (Figure 3B). In contrast, melphalan and cyclophosphamide had no measurable effect on mitochondrial function in TP53^−/− cells at the 4h treatment timepoint. Altogether, these data demonstrate that the activity of melflufen compared to other alkylators is associated with more efficient induction of cellular DNA damage and mitochondrial membrane disruption, and that melflufen can induce these processes with or without functional p53.

Gene expression analysis reveals differences in TP53wt vs. TP53^−/− cells after treatment with melflufen and melphalan

To understand if the rapid kinetics of melflufen intracellular accumulation and DNA damage induction uniquely influenced gene expression patterns compared to melphalan, RNA was collected from AMO-1 TP53wt and TP53^−/− cells after 2h exposure to the IC90 concentrations of melflufen or melphalan (Supplementary Table S4) and sequenced. Gene expression analysis showed distinct differences in the p53 pathway between melflufen and melphalan treatment of the TP53^−/− AMO-1 cells that were not apparent in TP53wt cells (Figure 4A). For example, melflufen treatment of the TP53^−/− cells resulted in significant upregulation of genes including the cyclin dependent kinase inhibitor 1A (CDKN1A), cyclin G2 (CCNG2), and DNA damage inducible alpha, beta, and gamma genes (GADD45A, B and G), as well as upregulation of the apoptosis inducing PMAIP1 (NOXA) gene. At the same time melflufen treatment caused significant downregulation in apoptosis inhibition and stress response genes including the well-known BCL2 and BCL2L1 genes. The primary target of alkylators is genomic DNA and treatment results in the upregulation of DNA damage repair (DDR) signaling as a cellular response to genomic stress. Importantly, the p53 protein is a key factor in the maintenance of DNA repair protein homeostasis¹⁰, and base excision repair (BER) and nucleotide excision repair (NER) are the main DNA repair pathways activated upon alkylator induced DNA damage.³⁶ We therefore investigated genes belonging to these pathways and analyzed their expression levels (Figure 4B and Supplementary Figure S2). In the DNA replication pathway analysis, these genes tended to be upregulated in TP53^−/− cells after melflufen treatment. Besides well-known replication guarding genes such as RPA1, POLD1,3,4 and POLE2,3, the upregulation of a cluster of genes that includes several micro-chromosome maintenance genes (MCM7, MCM3, MCM5) was also observed. Importantly, BER associated genes such as PARP1, PARP2, RCF4 and XRCC1, and NER pathway genes including the hallmark ERCC genes (ERCC1,2,5) and RAD23A were upregulated in TP53^−/− cells after melflufen but not melphalan treatment. The differential gene expression patterns between melflufen and melphalan treated TP53^−/− cells were also apparent after 12h incubation with drug (Supplementary Figure S2). In contrast, there was little difference in the gene expression patterns induced by melflufen and melphalan in TP53wt cells at 2 or 12h. The distinct gene expression patterns induced by melflufen in the TP53^−/− cells compared to melphalan suggest a clear mechanistic difference between the two alkylators.

• Download figure
• Open in new tab

Figure 4. Gene expression analysis in TP53 wild type versus TP53 mutant genetic background AMO-1 cells.

A, B) Heatmaps of gene expression of p53 signalling (A) and the Base Excision Repair (BER) DNA repair pathway (B) genes after 2h treatment of TP53wt or TP53^−/− cells with 2.5 μM melflufen versus 20 μM melphalan, respectively.

Patient PFS, OS and response in the del(17p) population

To confirm our preclinical observations and determine if melflufen is active in patients with dysfunctional TP53, we conducted a post-hoc analysis of the OCEAN head-to-head open label clinical trial where the primary endpoint was PFS in the mfldex vs. pomdex treatment arms. Earlier, the PFS of 6.8 months in the mfldex group was shown to be significantly better than the PFS of 4.9 months in the pomdex group (stratified log-rank p value: p=0.03; HR 0.792 [95% CI: 0.640, 0.981]).²⁸ In the post-hoc analysis for patient responder status in the chromosome 17 deletion population study, 70 patients characterized with del(17p) were included in this analysis from the total enrolled 495 patients in the OCEAN trial. Notably, the del(17p) patient population size from the OCEAN trial is in line with the published and commonly detected del(17p) cytogenetics rate where del(17p) is detected in ∼10-15% from total MM trial populations.^18,37,38

For PFS analysis of the del(17p) subpopulation, the patient distribution was as follows; in the mfldex arm 33 patients had del(17p) chromosomal aberration and 213 had no del(17p) while in the pomdex arm 37 patients were del(17p) mutant and 212 patients had no deletion (Supplementary Table S5). In the del(17p) subpopulation there was a trend towards improved PFS in the mfldex group with 7.1 months vs. 2.9 months in the pomdex group (HR 0.45 [95% CI: 0.26, 0.79], p=0.006) (Figure 5A). These data suggest that melflufen treatment is efficient in this high-risk subpopulation and can substantially prolong patient PFS. OS in the del(17p) subpopulation indicated a favourable OS in the pomdex arm versus mfldex arm, with 11.5 months in the mfldex arm and 15.9 months in the pomdex arm (HR 1.31 [95% CI: 0.76, 2.26], p=0.33) (Figure 5B). The different PFS and OS outcomes between the treatment groups (mfldex arm vs. pomdex arm) might be explained based on post-hoc subgroup analyses which indicated that OS outcomes may have been driven primarily by patients who had received previously high-dose melphalan followed by an autologous haematopoietic stem-cell transplantation (ASCT).²⁸ These earlier post-hoc analyses determined that melflufen plus dexamethasone should be indicated for RRMM, but exclude patients with a prior ASCT, whose time to progression is less than 3 years from transplantation.³⁹

• Download figure
• Open in new tab

Figure 5. Progression-free survival, overall survival, and overall response rate in del(17p) patient population from the OCEAN trial.

A) Kaplan-Meier progression-free survival curve, B) overall survival curve, and C) overall response rate for the investigator assessed all-treated del(17p) population. Analysis set of 70 del(17p) patients and 425 patients not harbouring del(17p). MflDex: melflufen + dexamethasone; PomDex: pomalidomide + dexamethasone; VGPR: very good partial response; PR: partial response; MR: minimal response; SD: stable disease; PD: progressive disease; NE: not evaluable

Based on analysis of response to treatment in the del(17p) subgroup, the ORR results are as follows; the mfldex arm had 33.3% ORR (11 responders from 33 patients) vs. pomdex which had an ORR of 10.8% (4 responders from 37 patients) (33.3% vs 10.8%, p=0.028) (Figure 5C). The observed ORR data in the del(17p) population showed that the mfldex treatment arm had a significantly higher number of responders compared to the pomdex treatment arm and suggest that melflufen treatment can induce responses in this high-risk population. In addition, the ORR to melflufen treatment in the del(17p) population was comparable to the ORR in the general patient population of the OCEAN trial, which included heavily pre-treated R/R MM patients. In the OCEAN study, the ORR for mfldex was 33% (95% CI, 27-39) vs. an ORR of 27% (95% CI, 22-33) for pomdex.

The del(17p) aberration often co-occurs with other high-risk cytogenetic abnormalities, which may also impact patient response to treatment and outcome. The distribution of these other high-risk changes within the del(17p) population and the outcome of these del(17p) subgroups are shown in Supplementary Table S6. The most common co-occurring high-risk cytogenetic abnormality with del(17p) was gain of 1q (+1q), which appeared to negatively impact response to both mlfdex and pomdex in the del(17p) population; however, statistical analysis of these further subdivided cytogenetic groups would have been underpowered.

Melfufen shows favorable clinical efficacy in del(17p) and/or TP53 mutation patients from the OCEAN trial

To better understand mfldex clinical efficacy in the population with dysfunctional TP53, which included a mixed population of patients comprised of del(17p) and non-del(17p) patients, we performed a NGS exploratory study to determine TP53 gene mutation status for both mfldex and pomdex treatment arms. For both arms, most sequenced patient samples contained silent single nucleotide polymorphism alterations to TP53 and thus retained p53 function. From the mfldex arm, 15 samples that were sequenced had del(17p) and/or contained a pathogenic mutation to TP53, while in the pomdex arms 18 patients were identified with del(17p) and/or pathogenic/loss-of-function mutation to TP53. Analyzing PFS in the sequenced population, the patient distribution was as follows; in the mfldex arm, 62 patients had no functional alteration to TP53, and 15 patients had either del(17p) and/or TP53 pathogenic mutation. In the pomdex arm, 50 patients had no functional TP53 impairment, and 18 patients displayed del(17p) and/or TP53 pathogenic mutation (Supplementary Table S7). The PFS analysis showed superiority of mfldex treatment vs. pomdex (Figure 6A). In the impaired TP53 group (del(17p) and/or TP53mut) the PFS in the mfldex group was 6.7 months vs. 4.7 months in the pomdex group (HR 0.57 [95% CI: 0.25, 1.29], p=0.2). These data indicate that in the OCEAN trial, mfldex treatment resulted in a better PFS in the TP53 gene mutated patient population.

• Download figure
• Open in new tab

Figure 6. Progression-free survival, overall survival and overall response rate in del(17p) and/or TP53 gene mutated patient population from the OCEAN trial.

A) Kaplan-Meier progression-free survival curve, B) overall survival curve, and C) overall response rate for all del(17p) and/or TP53 gene sequenced population. Analysis set of 33 del(17p) and/or TP53 gene mutated patients and 112 patients not harbouring del(17p) or TP53 gene mutation. MflDex: melflufen + dexamethasone; PomDex: pomalidomide + dexamethasone; CR: complete response; VGPR: very good partial response; PR: partial response; MR: minimal response; SD: stable disease; PD: progressive disease; NE: not evaluable

Additionally, OS analysis for the TP53 sequenced subpopulation indicated not significant but prolonged OS in the mfldex arm vs. pomdex arm (Figure 6B) with 22.8 months vs. 19.9 months, respectively (HR 1.54 [95% CI: 0.69, 3.47], p=0.30).

Based on analysis of response to treatment in the TP53 gene sequenced patient population, the ORR results are as follows; the mfldex arm had 33.3% ORR vs. pomdex which had an ORR of 11.1% (Figure 6C). The observed ORR data in the TP53 gene sequenced population showed that the mfldex treatment arm had a higher number of responders compared to the pomdex treatment arm. Only 3 patients were double hit patient with del(17p) and TP53 mutation. Their distribution was the following, 1 patient with partial response (PR) in the mfldex arm and 2 patients with no respond (stable disease (SD)) in the pomdex arm.

Notably, both post-hoc analyses (del(17p) population or the del(17p) and/or TP53 mutation sequenced population) included the melflufen treatment target (>36 months post ASCT) and non-target populations (<36 months post ASCT).³⁸ Inclusion of both populations resulted in a less favorable outcome for the mfldex arm. However further dividing the del(17p) or the del(17p) and/or TP53 sequenced populations would result in very small patient groups with very limited statistical power.