Synthetic Lethality Exploitation by an Anti–Trop-2-SN-38 Antibody–Drug Conjugate, IMMU-132, Plus PARP Inhibitors in BRCA1/2–wild-type Triple-Negative Breast Cancer

Synthetic lethality is a concept where the simultaneous mutational loss of function of two different genes results in cell death, whereas loss of just one gene is still compatible with cellular viability (1). This concept has been applied to cancer therapy in which a cell carrying a genetic mutation is targeted with a chemotherapeutic that blocks the function of another gene used by the cell to overcome this first mutation. In this context, the drug will be more potent in cells carrying the mutation than it would be in others that are genetically intact. One class of defects susceptible to synthetic lethality is those that affect homologous recombination repair (HRR) of double-stranded DNA (dsDNA) breaks. BRCA1 and BRCA2 are two such genes involved in HRR, and mutational loss of BRCA1/2 makes a cell more susceptible to drugs that block other DNA repair mechanisms (2). Another protein involved in maintaining the integrity of DNA is PARP (3). PARP is a family of enzymes whose primary function is to repair single-stranded DNA breaks before they advance to double-stranded breaks. PARP inhibitors (PARPi) have been developed to treat multiple cancer types with BRCA1/2 mutations, thereby creating synthetic lethality in the BRCA1/2-defective cells (4).

Clinically, therapy with PARPi has resulted in sustained antitumor responses in ovarian (5–8), prostate (5), pancreatic (9), and triple-negative breast cancers (TNBC; refs. 7, 10). In patients with TNBC, approximately 25% carry germline mutations of BRCA1/2 (11). In one clinical trial, TNBC patients with germline BRCA1/2 mutations were treated with the PARPi, olaparib. Although this therapy demonstrated a higher disease stabilization rate in BRCA1/2-mutant compared with nonmutant patients, there were no sustained responses achieved in either cohort (7). This is in contrast to Tutt and colleagues (10), in which 60% of TNBC patients with germline BRCA1/2 mutations had a partial response and 35% stable disease. This discrepancy is thought to be due, in part, to the large heterogeneity of TNBC and BRCA mutations.

Efforts to improve the effect of PARPi focus mainly on stressing DNA repair pathways by increasing dsDNA breaks with such agents as ionizing radiation or platinum-based therapeutics, while these repair pathways are being blocked with PARPi (12, 13). In addition to agents that directly interact with DNA to cause breaks, agents that inhibit topoisomerase I (Topo I), including irinotecan, have been shown to synergize with PARPi to deter the growth of a range of human tumor cell lines, including those of lung, ovarian, colon, and breast cancers (14, 15). These efforts demonstrate that by combining a DNA-damaging agent with an HRR synthetic lethality–based therapy, improved antitumor effects can be achieved compared with each single modality.

Sacituzumab govitecan (IMMU-132) is an antibody–drug conjugate (ADC) composed of the active metabolite of irinotecan, SN-38, conjugated to an anti–Trop-2 antibody (drug:Ab ratio = 7.6; ref. 16). Preclinically, IMMU-132 has significant efficacy across a broad range of Trop-2–positive human cancer xenograft models, including non–small cell lung cancer (NSCLC), pancreatic, colon, gastric cancers, and TNBC (16–18). In both gastric and TNBC tumor cells, IMMU-132 specifically mediated dsDNA breaks in Trop-2–expressing cells (16, 17). In a current clinical trial (ClinicalTrials.gov, NCT01631552), IMMU-132 has achieved objective responses against a range of solid tumors, including small-cell lung carcinoma (19), NSCLC (20), and TNBC (21). In TNBC patients, with a median of five prior therapies, IMMU-132 treatment achieved a confirmed objective response rate of 31%. This therapy was well-tolerated over multiple cycles lasting as long as 1 year, with neutropenia being the major adverse event (21).

Because PARPi can enhance the activity of DNA-damaging agents, such as SN-38, the current study was undertaken to examine the effect of combining IMMU-132–mediated DNA damage with PARPi in BRCA1/2-mutated and –wild-type human TNBC cell lines and xenografts. Three different PARPi were assessed for their interaction with IMMU-132 in terms of cytotoxicity, DNA stability, and in vivo efficacy. In addition, tolerability was determined in vivo for such combinations. Results from these studies provide the rationale for using IMMU-132 and PARPi combinations clinically for the treatment of TNBC, regardless, surprisingly, of BRCA1/2 mutational status.

All human cancer cell lines were purchased from the American Type Culture Collection (ATCC). Each cell line was maintained according to the recommendations of the ATCC, routinely tested for mycoplasma using MycoAlert Mycoplasma Detection Kit (Lonza), and authenticated by short-tandem repeat assay by the ATCC. Cells were in culture less than 6 months when employed in experiments. IMMU-132 and a control ADC (anti-CD20 hA20-SN-38) were prepared at Immunomedics, Inc., as described previously (18). PARPi (olaparib, talazoparib, and rucaparib) were purchased and dissolved in DMSO according to the manufacturer's recommendations (Selleck Chemicals).

Cytotoxicity studies were conducted as described previously (18) and are presented in Supplementary Data. Cell lysates were prepared, and immunoblotting for phospho-H2A.X, poly(ADP-ribose) (PAR), PARP, FACCD2, BARD1, Rad51, ERCC1, and β-actin was done as described in Supplementary Data. Concentrations, timing, and primary antibodies are shown in the figure legends. Asynchronistic cells were used for cell-cycle analysis, as described in Supplementary Data.

All animal studies were approved by Rutgers School of Biomedical and Health Sciences Institutional Animal Care and Use Committee. NCr female athymic nude (nu/nu) mice, 4 to 8 weeks old, were purchased from Taconic Farms. Xenografts were established by harvesting cells from tissue culture and mixing 1:1 with Matrigel, such that each mouse received a total of 1 × 10⁷ cells s.c. in the right flank. Tumor volume was determined by measurements in two dimensions using calipers, with volumes defined as: L x w²/2, where L is the longest dimension of the tumor and w the shortest. Mice were randomized into treatment groups, and therapy begun when tumor volumes were approximately 0.3 cm³. Treatment regimens, dosages, and number of animals in each experiment are described in the Results section and in the figure legends. The lyophilized IMMU-132 and control ADC (anti-CD20, hA20-CL2A-SN-38) were reconstituted and diluted as required in sterile saline. Olaparib was diluted in 2-hydroxy-propyl-β-cyclodextran/PBS (10% w/v) and administered i.p. to the mice. Talazoparib was diluted in 10% DMAc/5% Kolliphor HS15/85% PBS and administer p.o. to the animals.

Mice were deemed to have succumbed to disease progression and euthanized once tumors grew to greater than 1.0 cm³ in size. A partial response is defined as shrinking the tumor >30% from initial size. Stable disease is when the tumor volume remains between 70% and 120% of initial size. Time-to-tumor progression (TTP) was determined as time when tumor grew more than 20% from its nadir.

Hematologic toxicity of combined IMMU-132 and olaparib was assessed in female BALB/c mice, as described in Supplementary Data. Dosages and timing are described in the Results section and in the figure legends.

A critical-Z test was performed on the survival data of treatment and control groups with P ≤ 0.05 for any mouse deemed an outlier. Such mice were removed from further statistical analysis and are noted in the Results section. Statistical analysis of tumor growth was based on AUC. Profiles of individual tumor growth were obtained through linear-curve modeling. An F-test was employed to determine equality of variance between groups prior to statistical analysis of growth curves. A two-tailed t test was used to assess statistical significance between the various treatment groups and controls, except for the saline control, where a one-tailed t test was used in the comparison. Survival studies were analyzed using Kaplan–Meier plots (log-rank analysis), using the Prism GraphPad Software (v6.05) package (Advanced Graphics Software, Inc.). Significance was set at P ≤ 0.05.

It had previously been found that upregulation of several different proteins involved in HRR upon DNA damage (FANCD2, BARD1, Rad51, and ERCC1) plays an important role in resistance to DNA-damaging treatment in TNBC, particularly in BRCA1/2–wild-type tumor cells (22–24). To determine the effect IMMU-132–mediated DNA damage has on these other repair proteins, BRCA1/2–wild-type TNBC cell lines (MDA-MB-231, MDA-MB-468, and BT-20) were exposed to various concentrations of IMMU-132 (25, 50, and 100 nmol/L SN-38 equivalents) for 24 hours. In two of the cell lines (MDA-MB-231 and BT-20), levels of all four HRR proteins examined were elevated, suggesting that the cells were activating these DNA-repair pathways in response to IMMU-132–mediated damage (Supplementary Fig. S1). In a BRCA1/2-defective cell line (HCC1806), levels of these proteins remained unchanged in response to IMMU-132 exposure. Interestingly, MDA-MB-468 (BRCA1/2–wild-type, PTEN defective), FANCD2, Rad51, and ERCC1 levels dropped relative to basal levels upon IMMU-132 exposure. These data suggest that HCC1806 and MDA-MB-468 TNBC tumors would be the most sensitive to the combination of IMMU-132 and PARPi, because they do not activate these HRR rescue pathways.

Changes in IC₅₀ values for IMMU-132 when combined with various PARPi were determined in human TNBC cell lines with BRCA1/2 mutations (HCC38 and HCC1806) and those with wild-type BRCA1/2 (MDA-MB-468, MDA-MB-231, and BT-20). Isobolograms for IMMU-132 were clearly enhanced when combined with olaparib, rucaparib, or talazoparib, demonstrating a synergistic interaction in all four cell lines (Supplementary Fig. S2). IC₅₀ values for IMMU-132 alone were 1.9 to 82.5 nmol/L, whereas olaparib and rucaparib IC₅₀ values were 5.4 to 256.7 μmol/L (Table 1). Although talazoparib was more potent (4.7 to 322 nmol/L) than either olaparib or rucaparib, IC₅₀ values for IMMU-132 indicated it was >2-fold more potent than talazoparib. Calculated combinatorial index (CI) values demonstrate that even when combined with IC₁₀ concentrations, an additive effect was achievable (i.e., CI = 1.0), but this was further enhanced to synergy (i.e., CI < 1.0) when as little as IC₂₀ concentrations were used (Table 1). Importantly, this synergy occurred in both BRCA1/2-mutated and –wild-type cell lines, despite the upregulation of HRR proteins shown in MDA-MB-231 and BT-20 in response to IMMU-132–mediated DNA damage.

Assessment of dsDNA breaks was measured in TNBC cell lines for IMMU-132 and PARPi combinations (24-hour incubation; Fig. 1). Four TNBC cell lines were tested with IMMU-132 and olaparib (Fig. 1A). Not unexpectedly, PAR was not detected in any of the olaparib-treated cells, whereas IMMU-132 alone had only a minor effect on PAR levels. When combined, PAR was still not detected, which was due to inhibition, not loss of PARP, as indicated by equivalent levels of full-length PARP (FL-PARP) present in all samples. Increased dsDNA breaks, as evidenced by increases in p-H2A.X (25), were shown in all four cell lines treated with olaparib alone, although the level in MDA-MB-231 was only 30% above untreated controls, which were set to 1.0, whereas MDA-MB-468 cells showed a 2.5-fold increase (highest response). In three of the four cell lines, IMMU-132 treatment alone also resulted in increased p-H2A.X levels with 9 nmol/L SN-38 equivalents of IMMU-132 (187 ng/mL of the ADC), demonstrating a >2.5-fold increase (range, 2.5–13.9-fold). Only the BRCA1/2–wild-type MDA-MB-231 was seemingly resistant to single-agent IMMU-132 or olaparib. However, when IMMU-132 (9 nmol/L) was combined with olaparib, there was >5-fold increase in p-H2A.X in all four cell lines, including MDA-MB-231, which exhibited a 5.7-fold increase in p-H2A.X levels. Even at 3 nmol/L SN-38 equivalents of IMMU-132 (62 ng/mL of the ADC), when combined with olaparib, levels of p-H2A.X rose >76% above that achieved by each agent alone in all four cell lines, with the highest response in MDA-MB-231 (171%), followed by HCC1806 (136%), MDA-MB-468(116%), and HCC38 (76%).

Rucaparib and talazoparib were used at 10- and 100-fold lower concentrations, respectively, when combined with IMMU-132 in HCC1806 (Fig. 1B and C, respectively). Likewise, the amount of IMMU-132 required to demonstrate increases in dsDNA breaks, when combined with these two PARPi, was lower (0.1, 1, and 3 nmol/L SN-38 equivalents). Rucaparib alone resulted in a 1.6-fold increase in dsDNA breaks, whereas 0.1 nmol/L SN-38 equivalents of IMMU-132 (2 ng/mL of the ADC) resulted in a 2.2-fold increase. Together, IMMU-132 plus rucaparib further increased the amount of dsDNA breaks 3.9-fold, which is 77% higher than either single agent. Talazoparib and IMMU-132 combined demonstrated a dose-response with the amount of dsDNA breaks increasing by 78% at the lowest IMMU-132 concentration to 216% at the highest concentration. These data demonstrate that three different PARPi, when combined with IMMU-132, result in increased DNA damage in the form of double-stranded breaks above that observed with single-agent treatments, which is consistent with the in vitro cytotoxicity assays demonstrating synergistic growth inhibition when IMMU-132 was combined with these PARPi.

Cell-cycle changes were assessed in all four TNBC cell lines when IMMU-132 and olaparib were incubated with asynchronistic cells for 24 hours (Supplementary Fig. S3). Greater than 80% of the cells in the untreated groups were in the G₁-phase. Cells began to accumulate in S-phase when exposed to either olaparib or IMMU-132 alone. However, the combination demonstrated an even greater percentage of cells in S-phase, with a concomitant decrease in the percentage of cells in G₁-phase and little change in G₂–M. For both BRCA1/2-mutated cells and BRCA1/2–wild-type MDA-MB-231, the combination of IMMU-132 and olaparib resulted in a greater than 2-fold increase in the percentage of cells in the S-phase relative to each agent alone. MDA-MB-468 demonstrated the least increase in the percentage of cells in S-phase relative to single-agent treatment, but this was due primarily to the greater effect each agent alone had, with more than 32% and 50% of the cells in S-phase upon treatment with olaparib or IMMU-132 alone, respectively. Even so, the combination increased the number of cells in S-phase by 12% over IMMU-132 alone. Similar results were observed when IMMU-132 was combined with rucaparib (1 μmol/L) or talazoparib (0.2 μmol/L), in which the combination of IMMU-132 plus either PARPi increased the percentage of cells in S-phase for all four cell lines, relative to single-agent treatment (Supplementary Tables S1 and S2, respectively).

Therapeutic efficacy of combining IMMU-132 with olaparib was tested in HCC1806 and MDA-MB-468 tumor xenografts (Fig. 2). Importantly, in both experiments, the dose and schedule of IMMU-132 administered to the animals were chosen to give only a modest therapeutic effect in order to detect a potential interaction between IMMU-132 and olaparib.

In mice bearing HCC1806 tumors (Fig. 2A), the combination therapy resulted in a significant antitumor effect when compared with all other treatment groups (P < 0.0017, AUC; Supplementary Table S3). In this group, all of the mice exhibited a partial response, with 30% being tumor-free (i.e., complete response) when the experiment ended on therapy day 108. Mean TTP for this group was 36.9 ± 8.1 days (Table 2). In contrast, tumors in mice given IMMU-132 or olaparib monotherapy progressed with mean TTP's of 17.6 ± 3.9 and 9.1 ± 4.8 days, respectively. Between the two monotherapy groups, IMMU-132 significantly delayed tumor progression when compared with olaparib therapy (P = 0.0009). Overall, the combination had a significantly greater delay in TTP when compared with either single-agent therapy group (P < 0.0001). Of note, this combination proved to be well-tolerated by the mice, with no appreciable loss in body weight (Supplementary Fig. S4A). Even mice treated with 4-fold more IMMU-132 (500 μg twice weekly x 4 weeks) combined with olaparib demonstrated no significant loss in weight (Supplementary Fig. S4B).

In mice bearing MDA-MB-468 tumors, the combination of IMMU-132 plus olaparib had a significant antitumor effect when compared with all other treatment groups, including IMMU-132 monotherapy (Fig. 2B; P < 0.0040; AUC; Supplementary Table S4). In terms of response rate, 30% of mice treated with only IMMU-132 exhibited a partial response, as did 20% of the animals treated with olaparib (Table 2). In contrast, all of the mice in the combination group exhibited a partial response, with a TTP of 55.2 ± 6.5 days, which was significantly better than all the other groups (P < 0.0004) except IMMU-132 alone, which approached significance at a TTP of 45.1 ± 13.6 days (P = 0.0604). As in the HCC1806 tumor model, the mice tolerated the combination of IMMU-132 and olaparib without loss of body weight (Supplementary Fig. S4C).

Given the ability of IMMU-132 and olaparib to improve efficacy significantly compared with single-agent therapy in MDA-MB-468 tumors, which carry wild-type BRCA1/2, another therapy study utilized mice bearing MDA-MB-231 tumors that likewise are wild-type for BRCA1/2, but demonstrated synergy when IMMU-132 and PARPi were combined in vitro. MDA-MB-231 tumors progressed very rapidly, reaching experimental endpoints (i.e., tumor volume >1.0 cm³) within 14 days of starting therapy. Given this rapid disease progression, antitumor effects are represented as the ability to provide survival benefit and not tumor regressions. As noted in the previous two therapy experiments, olaparib alone had no antitumor effects in mice bearing MDA-MB-231 tumors, with median survival no different than saline control mice (Fig. 2C). Not unexpectedly, IMMU-132 alone did not significantly improve the overall survival of the treated animals, since it was shown previously to be resistant to this therapy (16). However, when mice were treated with the combination of IMMU-132 plus olaparib, there was a significant increase in median survival from 14 days for saline control animals to 21 days for those mice treated with the combination (P = 0.0005). Likewise, treatment with olaparib also resulted in a significant survival benefit when combined with the nonspecific control ADC [median survival time (MST) = 17 days; P = 0.0257], suggesting that the olaparib does sensitize this resistant tumor to SN-38. These results show, unexpectedly, that even in tumors resistant to PARPi and IMMU-132 monotherapies, the combination is able to overcome this resistance and impart a survival benefit when compared with saline control animals.

Because talazoparib's potency was greater than olaparib in vitro, both alone and when combined with IMMU-132, mice bearing HCC1806 tumors were treated with the combination of IMMU-132 and talazoparib (Fig. 2D). Whereas mice in the olaparib experiments were treated with 50 mg/kg, talazoparib-treated animals were given only 0.33 mg/kg. As observed in the olaparib experiments, IMMU-132 plus talazoparib had a significant antitumor effect when compared with all other treatments, including monotherapy with IMMU-132 or talazoparib (P < 0.0088, AUC). Mean TTP was 21.8 ± 9.6 days, which was >2.8-fold longer than any of the other treatment groups (P < 0.0021; Table 3). Similar to the combinations with olaparib, IMMU-132 plus talazoparib did not result in any observable toxicity in the mice (Supplementary Fig. S4D).

Clinical studies indicate that there is no off-tumor, on-target toxicities in patients (19), and therefore an experiment was performed to determine if the combination of a PARP inhibitor with IMMU-132′s SN-38 would have increased toxicities at therapeutically active doses. Because normal tissues of mice are not targeted by the anti–Trop-2 antibody comprising IMMU-132, toxicity studies were conducted in mice to determine possible off-target toxicities at doses of IMMU-132 and olaparib shown to have therapeutic activity when combined in tumor-bearing mice (i.e., 250 μg IMMU-132 plus 50 mg/kg olaparib). The schedule used mimics the clinical regimen of IMMU-132 (i.e., weekly for 2 weeks with 1 week off per cycle) with olaparib administered also over a 2-week period. This same dose schedule does demonstrate antitumor activity over this 2-week period, as shown in the efficacy studies described above (Fig. 2A and B). Naïve BALB/c mice were administered a 2-week course of IMMU-132 plus olaparib therapy to ascertain any possible myelotoxicity. On day 15 (i.e., 7 days after the last IMMU-132 injection and 3 days after last dose of olaparib), there was no evidence of hematologic toxicity in any of the mice (Fig. 3). In particular, the combination did not result in any significant drop in total white blood cell (WBC) or lymphocyte counts. Most importantly, there was no evidence of neutropenia or thrombocytopenia in any of these mice. Only mice treated with olaparib monotherapy demonstrate a significant drop in lymphocytes relative to vehicle control animals (P = 0.0481), although the mean (3.92 ± 1.13 K/μL) remained within the normal range (3.8 to 8.9 K/μL). After a further 7 days (i.e., day 22), the assessment of toxicity was determined in the remaining mice. At this time-point, all three treatments showed a significant drop in WBC and lymphocyte counts relative to control animals (P < 0.0012). However, most of these values still remained within an acceptable range, with the lowest values found in those mice treated with the combination, which fell just below the lower end of the normal range (mean lymphocyte count of 3.62 ± 0.54 K/μL vs. 7.04 ± 1.17 K/μL for control, P = 0.003). These results indicate that the combination of IMMU-132 and olaparib, which has been shown to elicit antitumor effects in TNBC tumor-bearing mice, is not associated with any significant myelotoxicity, and suggests a potentially high therapeutic window.

Exploitation of defects to a cell's HRR pathways has gained widespread interest as a means to induce synthetic lethality in various human cancers, including TNBC (1, 2). Of particular interest are those drugs that target PARP (26, 27). Although some clinical studies have suggested that use of PARPi can provide meaningful responses, other trials have shown no significant improvements in clinical outcome (7, 10). Given these mixed results, new approaches seek to combine PARPi with other DNA-damaging agents, such as platinum-based drugs, microtubule inhibitors, or irinotecan (14, 22, 28–31). Here, we examined the effect of combining several PARPi (olaparib, rucaparib, and talazoparib) with our Topo I–inhibiting ADC, anti–Trop-2/SN-38-ADC, and IMMU-132. All three PARPi synergized with IMMU-132 to inhibit the growth of four human TNBC cell lines. This occurred independently of BRCA1/2 status, as evidenced by the synergistic growth inhibition in vitro of two cell lines with wild-type BRCA1/2. IMMU-132 combined with these PARPi increased the dsDNA breaks above that observed with each single agent, and the accumulation of cells in the S-phase of the cell cycle. Most importantly, in mice bearing either a BRCA1/2-mutated TNBC tumor (HCC1806) or ones with wild-type BRCA1/2 (MDA-MB-468 and MDA-MB-231), the combination of IMMU-132 plus a PARPi resulted in significant antitumor effects above those observed with monotherapy. These results clearly illustrate the potential clinical benefit that may be derived by combining IMMU-132 with a PARPi in TNBC.

Topo I is an enzyme utilized by the cell to allow unwinding of the DNA strand during transcription and replication (32). PARP-1, the most abundant of the PARP proteins, has been found to colocalize with Topo I throughout the cell cycle. However, upon DNA damage, PARP-1 dissociates from Topo I, resulting in reduced activity of this enzyme (33). We hypothesized that by combining the Topo I inhibitor carried by IMMU-132 (i.e., SN-38) with a PARPi, there would be an accumulation of dsDNA breaks due to the inability of the remaining HRR pathways in the cell to repair this damage with high fidelity, resulting in apoptosis and cell death. In addition, in cells lacking functional BRCA1/2 or are otherwise deficient in HRR, the more error-prone, nonhomologous end-joining (NHEJ) pathway is utilized, further compromising the cell toward irreparable DNA damage and apoptosis (34). Others have found that in both BRCA1–wild-type and -mutated TNBC cell lines, the combination of CPT-11 (the prodrug of SN-38) and a PARPi could result in synergistic growth inhibition in vitro (22). Here, too, we demonstrated that IMMU-132, when combined with olaparib, rucaparib, or talazoparib, mediates synergistic tumor cell growth inhibition and increased dsDNA breaks in both BRCA1/2-mutated and –wild-type cell lines. These data confirm the ability of IMMU-132–mediated inhibition of Topo I, when combined with PARPi, to synergize growth inhibition in human TNBC cells regardless of BRCA1/2 mutational status.

HRR pathways are more active during the late S–G₂-phase of the cell cycle, whereas the more error-prone NHEJ repair pathway is most evident throughout the S-phase (35). Cells with HRR defects, as well as those exposed to PARPi, will arrest at G₂–M-phase (23, 36). Although we found that PARPi (olaparib, rucaparib, and talazoparib) and IMMU-132 exposure resulted in small increases in the number of cells in the G₂–M-phase of the cell cycle, it was also evident that there was greater accumulation in the S-phase. Most importantly, the combination demonstrated an even greater percentage of cells in S-phase, with a concomitant decrease in the percentage of cells in G₁- and little change in the G₂–M-phase. Interestingly, the interaction of PARP-1 and Topo I was found to be most concentrated during the S-phase (33). Our data are consistent with this and suggest that when using a PARPi combined with IMMU-132, we are disrupting this PARP-1/Topo 1 interaction, resulting in accumulation in the S-phase and thus toward the more error-prone NHEJ repair pathway that predominates during this phase of the cell cycle.

Combining IMMU-132 therapy with a PARPi (olaparib or talazoparib) in mice bearing both BRCA1/2-mutated and –wild-type tumors demonstrated significant antitumor effects greater than that observed with monotherapy, thus extending our in vitro observations to in vivo TNBC models. Likewise, others have demonstrated improved efficacy when a PARPi is combined with a DNA-damaging chemotherapeutic in mice bearing human TNBC xenografts (22, 30). In one example, CPT-11 was used in combination with veliparib to improve the therapeutic outcome significantly in mice bearing s.c. MX-1 TNBC tumors (22). Similar to our experience, the PARPi alone had no effect on tumor growth, whereas the combination of CPT-11 and veliparib demonstrated a significantly enhanced antitumor effect. It should be noted that CPT-11 is more effectively converted to its active SN-38 form in mice than in humans, and therefore similar findings may not be reproducible in the clinical setting (37). However, by using IMMU-132 with its SN-38 payload attached, we are not dependent on the efficiency of prodrug-to-drug conversion by the patient, and therefore the antitumor effects we observed with the combination of IMMU-132 plus PARPi are more clinically relevant.

Approximately 80% of breast cancer patients carrying hereditary BRCA1 mutations present with TNBC (38). Although BRCA1/2 status is an important marker in TNBC, it is not the only one. Many other BRCA1/2–wild-type TNBCs occur as sporadic tumors that share traits with germline BRCA–mutated tumors and are termed to have BRCAness (39, 40). For example, analysis of germline mutations in 158 TNBC patients demonstrated that in addition to BRCA1 mutations (17% of patients), mutations of the CHEK2 checkpoint kinase 2 gene, nibrin NBN gene, and ATM genes also were found (3.9%), and are themselves involved in various aspects of the HRR pathway (41–43). Overall, the evidence suggests that besides BRCA1/2 status, many other genes contribute to the BRCAness of TNBC. In accordance with these observations, our ability to synergize and demonstrate significantly improved efficacy in the MDA-MB-468 tumor could be explained by the PTEN mutation it carries and its loss of function. Loss of PTEN expression results in Rad51 dissociation from DNA replication forks, and subsequent destabilization and stalled replication. Cells lacking PTEN have deficient HRR functions, likely due to reduced Rad51 recruitment to the replicating DNA, culminating in loss of fidelity during DNA synthesis (44). As with BRCA1/2 mutations, cells lacking PTEN function are prone to HRR deficiencies and, subsequently, have been shown to be sensitive to PARPi (45, 46). Our results demonstrate that IMMU-132 combined with a PARPi will significantly improve efficacy in both traditional BRCA1/2-defective TNBC tumors and in tumors demonstrating BRCAness in the form of deficiencies in other proteins involved in HRR pathways (e.g., PTEN).

Because not all TNBC tumors exhibit BRCAness, we evaluated the ability of IMMU-132 plus a PARPi to inhibit tumor growth in such tumor types. MDA-MB-231 carries no mutations to PTEN, BRCA1/2, or other known HHR proteins, and as such is expected to be resistant to PARPi treatment (22, 46). Previous in-vivo studies demonstrated that mice bearing MDA-MB-231 tumor xenografts were resistant to IMMU-132 therapy (16). One reason for this resistance may be over-expression of Rad51 in MDA-MB-231 (47). Cells exposed to SN-38 and olaparib have been found to increase Rad51 recruitment in the HRR pathway (48). In fact, we found that IMMU-132 exposure does result in an increase in several different proteins involved in HRR, including Rad51, BARD1, FANCD2, and ERCC1 in BRCA1/2–wild-type cell lines. Of note was the observation that while these proteins are upregulated in some cell lines (MDA-MB-231 and BT-20), they are actually downregulated in MDA-MB-468, which may account for its sensitivity to both IMMU-132 monotherapy and the combination therapy. At the same time, it is likely that this increased expression of these HRR proteins in MDA-MB-231 allows for a more efficient repair of damaged DNA caused by IMMU-132 than would otherwise occur in cells that had low expression. A second possible reason for MDA-MB-231 resistance may be that it has low expression of Trop-2 on its surface, which could limit how much SN-38 can be delivered to the tumor by IMMU-132 (17). However, when IMMU-132 and its SN-38 payload were combined with olaparib, we achieved a significant survival benefit in MDA-MB-231 tumor–bearing animals. Moreover, although the amount of SN-38 delivered by IMMU-132 alone may be insufficient to retard tumor growth, enough must accrete in the tumor so that when combined with olaparib, we are able to slow tumor progression enough to provide a survival benefit. We are currently studying various MDA-MB-231 clones that were transfected with Trop-2 to increase expression levels in order to better determine what role Trop-2 surface expression plays in a tumor's sensitivity to IMMU-132 therapy, either alone or when combined with PARPi.

In a current phase I/II clinical trial, IMMU-132 has demonstrated an objective response rate (confirmed complete response + partial response) of 31% in 58 patients with metastatic TNBC (21). At the MTD, grade 3–4 toxicities included neutropenia, febrile neutropenia, diarrhea, fatigue, anemia, leukopenia, and dyspnea. All of these toxicities were manageable, with the incidence of severe diarrhea being less than that reported for the parental prodrug, irinotecan. Single-agent olaparib toxicities include grade 3–4 fatigue and thrombocytopenia (5). In clinical studies evaluating the combination of either olaparib or veliparib with irinotecan, dose-limiting toxicities included fatigue, anorexia, diarrhea, nausea, vomiting, febrile neutropenia, neutropenia, and thrombocytopenia (31, 49). A hallmark in these clinical trials was that the amount of irinotecan and PARPi administered was reduced below the single-agent MTDs in order to gain acceptable tolerability. Based on our experience with IMMU-132 in the clinic, as well as those reported with PARPi plus irinotecan, we do not expect to observe any different dose-limiting toxicity with a combination of IMMU-132 plus PARPi's. The doses of IMMU-132 administered to mice in all our preclinical studies were chosen based on their ability to produce a minimal effect, so that we could better determine if the combinations were beneficial and therefore indicated a good therapeutic window when IMMU-132 was combined with a PARPi. Importantly, in all our combination studies, the mice tolerated both the PARPi and IMMU-132 well, with no significant loss in body weight or substantial hematologic toxicity during the course of treatment. Given such a potential for a high therapeutic window, we speculate that even at lower IMMU-132 dosages, combinations with a PARPi in this subset of breast cancer patients will achieve improved therapeutic responses without unmanageable toxicity.

In summary, through the utilization of PARPi to target TNBC and create synthetic lethality combined with Topo I inhibition mediated by IMMU-132, a synergistic growth–inhibitory outcome was achieved in human TNBC tumor lines, regardless of BRCA1/2 status. Although BRCAness is an important element in this combined effect, it is not limiting. Indeed, PARPi are under investigation in a variety of cancers involving breast, ovarian, pancreatic, NSCLC, gastric cancer, glioblastoma, melanoma, and others. Our results indicate that the combination of PARPi and IMMU-132 could broaden the range of tumors that are usually treated with the former. It is interesting that IMMU-132 does target all of these cancer types (17), which may make this combination a rational approach. However, combination studies with PARPi have been problematic because of hematologic toxicity, thus requiring dose modification. The studies described here have shown tolerability for the combination in terms of hematologic parameters and animal body weight, but the ultimate clinical experience will of course be critical. Given the promising results obtained thus far with IMMU-132 in patients with TNBC, the logical next step is to combine this therapy with PARPi to further improve clinical outcome.

R.M. Sharkey holds ownership interest (including patents) in, and is a consultant/advisory board member for, Immunomedics, Inc. D.M. Goldenberg holds ownership interest (including patents) in Immunomedics, Inc. No potential conflicts of interests were disclosed by the other authors.

Conception and design: T.M. Cardillo, R.M. Sharkey, D.L. Rossi, D.M. Goldenberg

Development of methodology: D.L. Rossi

Acquisition of data (provided animals, acquired and managed patients, provided facilities, etc.): T.M. Cardillo, D.L. Rossi, A.A. Mostafa

Analysis and interpretation of data (e.g., statistical analysis, biostatistics, computational analysis): T.M. Cardillo, R.M. Sharkey, D.L. Rossi, R. Arrojo, A.A. Mostafa, D.M. Goldenberg

Writing, review, and/or revision of the manuscript: T.M. Cardillo, R.M. Sharkey, D.L. Rossi, D.M. Goldenberg

Administrative, technical, or material support (i.e., reporting or organizing data, constructing databases): T.M. Cardillo, R.M. Sharkey, D.L. Rossi, D.M. Goldenberg

Study supervision: T.M. Cardillo, D.M. Goldenberg

We thank Serengulam V. Govindan, Nalini Sathyanarayan, Christine Mazza-Ferreira, Jing Xia, and Jennifer Donnell for contributions to synthetic and conjugation chemistries, Maria Zalath for assistance with the animal studies, and Rongxiu Li for technical assistance in performing the cell-cycle FACS analysis.

The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.