Phase I Study of the Prolactin Receptor Antagonist LFA102 in Metastatic Breast and Castration-Resistant Prostate Cancer

Abstract

Lessons Learned

• Despite evidence for a role for prolactin signaling in breast and prostate tumorigenesis, a prolactin receptor-binding monoclonal antibody has not produced clinical efficacy.

• Increased serum prolactin levels may be a biomarker for prolactin receptor inhibition.

• Results from the pharmacokinetic and pharmacodynamics (PD) studies suggest that inappropriately long dosing intervals and insufficient exposure to LFA102 may have resulted in lack of antitumor efficacy.

• Based on preclinical data, combination therapy of LFA102 with those novel agents targeting hormonal pathways in metastatic castration-resistant prostate cancer and metastatic breast cancer is promising.

• Given the PD evidence of prolactin receptor blockade by LFA102, this drug has the potential to be used in conditions such as hyperprolactinemia that are associated with high prolactin levels.

Background.

Prolactin receptor (PRLR) signaling is implicated in breast and prostate cancer. LFA102, a humanized monoclonal antibody (mAb) that binds to and inhibits the PRLR, has exhibited promising preclinical antitumor activity.

Methods.

Patients with PRLR-positive metastatic breast cancer (MBC) or metastatic castration-resistant prostate cancer (mCRPC) received doses of LFA102 at 3–60 mg/kg intravenously once every 4 weeks. Objectives were to determine the maximum tolerated dose (MTD) and/or recommended dose for expansion (RDE) to investigate the safety/tolerability of LFA102 and to assess pharmacokinetics (PK), pharmacodynamics (PD), and antitumor activity.

Results.

A total of 73 patients were enrolled at 5 dose levels. The MTD was not reached because of lack of dose-limiting toxicities. The RDE was established at 60 mg/kg based on PK and PD analysis and safety data. The most common all-cause adverse events (AEs) were fatigue (44%) and nausea (33%) regardless of relationship. Grade 3/4 AEs reported to be related to LFA102 occurred in 4% of patients. LFA102 exposure increased approximately dose proportionally across the doses tested. Serum prolactin levels increased in response to LFA102 administration, suggesting its potential as a biomarker for PRLR inhibition. No antitumor activity was detected.

Conclusion.

Treatment with LFA102 was safe and well tolerated, but did not show antitumor activity as monotherapy at the doses tested.

Discussion

Prolactin, a pituitary-derived polypeptide hormone, is implicated in breast and prostate tumorigenesis. Expression of the PRLR has been confirmed in breast and prostate cancers. This phase I study evaluated LFA102 in 73 patients with PRLR-positive MBC or mCRPC, treated at doses of 3–60 mg/kg. During dose escalation, LFA102 demonstrated favorable safety and tolerability at all doses. No dose-limiting toxicities (DLTs) occurred; therefore, the MTD was not reached, although the RDE was established at 60 mg/kg based on safety, PK, and PD data supported by Bayesian logistic regression modeling. Dose proportionality analysis showed that serum LFA102 maximum concentration observed (C_max) and area under the last measurable concentration (AUC_last) were approximately linearly dose dependent (Fig. 1) and should provide sufficient exposure to achieve efficacy. However, no objective responses were observed in patients with MBC, and in patients with mCRPC, there were no prostate-specific antigen (PSA) responses.

Figure 1.

AUC_last and C_max increase with LFA102 dose in a relatively proportional manner. AUC_last (A) and C_max (B) results for individual patients in cycle 1. For each dose, parameter values (open symbols), least-square mean (black triangles), and 90% least-square means confidence interval (vertical bars) are shown. Serum LFA102 concentrations were measured up to day 28 of cycle 1 via dense sampling followed by trough concentration measurement in subsequent cycles. Concentration-time profiles show biexponential disposition typical for monoclonal antibodies. C_max and AUC_last increased in a relatively proportional manner with increasing LFA102 doses.

Abbreviations: AUC_last, area under the last measurable concentration; C_max, maximum concentration observed.

In vitro data have shown a high binding affinity of LFA102 to PRLR, but because assessing LFA102 binding within tumors is impractical in patients, our study used serum prolactin levels as a surrogate marker for PRLR inhibition. A sixfold change in serum prolactin levels from baseline was observed in patients treated with LFA102 60 mg/kg, indicative of inhibition of PRLR and ruling out poor target binding as causing lack of efficacy (Fig. 2). Other potential explanations for the lack of LFA102 efficacy include that prolactin may not be an oncogenic driver in breast and prostate cancer in humans, unforeseen compensatory modulation of downstream signaling pathways in response to PRLR inhibition, or upregulation of other tumorigenic signaling pathways that compensate for PRLR inhibition. Nevertheless, preclinical data show that letrozole potentiates the efficacy of LFA102 when administered in combination in a rat mammary cancer model. Therefore, although LFA102 monotherapy may not show antitumor activity, it may have potential for treating prolactin-dependent tumors in combination with other recently approved, novel hormonal pathway targeting agents in MBC and mCRPC. Furthermore, given the PD evidence of prolactin receptor blockade by LFA102, this drug has the potential to be used in conditions such as hyperprolactinemia that are associated with high prolactin levels.

Figure 2.

Serum prolactin levels rise with increasing doses of LFA102. Linear views of individual serum prolactin concentration-time profiles grouped by LFA102 dose group are shown. Individual patient serum prolactin increased after LFA102 administration.

Trial Information

Disease

Breast cancer

Disease

Prostate cancer

Stage of disease / treatment

Metastatic / Advanced

Prior Therapy

1 prior regimen

Type of study - 1

Phase I

Type of study - 2

Adaptive Design

Primary Endpoint

Recommended Phase II Dose

Primary Endpoint

Maximum Tolerated Dose

Primary Endpoint

Safety

Primary Endpoint

Tolerability

Secondary Endpoint

Pharmacokinetics

Secondary Endpoint

Pharmacodynamic

Secondary Endpoint

Efficacy

Additional Details of Endpoints or Study Design

Exploratory: Effects of LFA102 on serum prolactin levels.

Investigator’s Analysis

Evidence of target inhibition but no or minimal antitumor activity

Drug Information

Drug 1

Generic/Working name

LFA102

Drug type

Antibody

Dose

mg/kg

Route

IV

Schedule of Administration

10 mg/kg once every 4 weeks.

Dose Escalation Table

Patient Characteristics

Number of patients, male

39

Number of patients, female

34

Stage

Locally advanced or metastatic disease.

Age

Median (range): 66.0 years (41.0–89.0 years)

Number of prior systemic therapies

Median (range): Not Collected

Performance Status: ECOG

0 — 30

1 — 38

2 — 5

3 — 0

unknown —

Cancer Types or Histologic Subtypes

Breast and prostate, 73

Primary Assessment Method

Control Arm: Breast And Prostate

Number of patients screened

73

Number of patients enrolled

73

Number of patients evaluable for toxicity

73

Number of patients evaluated for efficacy

73

Response assessment CR

n = 0 (0%)

Response assessment PR

n = 0 (0%)

Response assessment SD

n = 13 (18%)

Response assessment PD

n = 41 (56%)

Response assessment OTHER

n = 19 (26%)

Control Arm: Total Patient Population

Number of patients screened

73

Number of patients enrolled

73

Number of patients evaluable for toxicity

73

Number of patients evaluated for efficacy

73

Response assessment CR

n = 0 (0%)

Response assessment PR

n = 0 (0%)

Response assessment SD

n = 13 (18%)

Response assessment PD

n = 41 (56%)

Response assessment OTHER

n = 19 (26%)

Adverse Events

Serious Adverse Events

Dose Limiting Toxicities

Pharmacokinetics/Pharmacodynamics

Assessment, Analysis, and Discussion

Completion

Study completed

Investigator’s Assessment

Evidence of target inhibition but no or minimal antitumor activity

Prolactin is a pituitary-derived polypeptide hormone implicated in breast and prostate tumorigenesis [1–3]. Prolactin is also expressed in several extrapituitary sites, in addition to breast and prostate tumors themselves [1, 4–7]. Expression of PRLR has been confirmed in various cancers, including breast and prostate [8–13]. Data suggest that increased serum prolactin levels may increase breast cancer risk and correlate with worse prognosis [14–16]. Overexpression of prolactin in murine mammary glands leads to tumor formation, and transplanted PRLR-negative tumors exhibit delays in tumor expansion compared with PRLR-positive tumors in mice [17, 18]. Although prolactin is expressed in normal human prostate, high expression in prostate tumors is associated with high-grade prostate cancer and worse prognosis [4, 19]. Overexpression of prolactin in mouse prostate causes hyperplasia and tumorigenesis [20, 21]. Therefore, blocking prolactin signal transduction is an attractive target in breast and prostate cancers.

Attempts made to inhibit PRLR signaling in vivo have been unsuccessful [22–27]. LFA102 is a humanized mAb that binds to the extracellular domain of PRLR. LFA102 inhibits PRLR signal transduction and cell proliferation in human breast cancer cells and causes tumor regression in animal xenograft models. Rats treated with LFA102 showed increased serum prolactin levels, suggesting this may be a potential biomarker for PRLR inhibition [28]. These data suggest that LFA102 has the potential to be an effective therapeutic agent in patients with breast or prostate cancer.

This phase I study evaluated LFA102 in patients with PRLR-positive MBC or mCRPC. Between September 2011 and March 2014, 73 patients were treated with LFA102 at doses of 3–60 mg/kg. During dose escalation, no DLTs occurred and the MTD was not reached. The RDE was established at 60 mg/kg, the highest tested dose level. The most common AEs, regardless of study drug relationship, were fatigue (44%), nausea (33%), constipation, decreased appetite, and vomiting (21% each). Of the 73 patients treated, 3 patients (4%) had grade 3 or 4 AEs suspected to be related to the study drug: decreased blood phosphorus, increased serum lipase, and decreased blood lymphocyte count, each in 1 patient (1%).

The serum LFA102 concentration-time profiles showed biexponential disposition typical for mAbs. C_max and AUC_last increased in a relatively proportional manner with increasing LFA102 doses (Fig. 1). The geometric mean apparent volume of distribution at steady state (Vss) and clearance across the treatment groups were similar, indicating linear PK. The geometric mean of Vss for doses of 3–60 mg/kg ranged from 4 to 6 L. The geometric mean half-life ranged from 6 to 9 days. At the RDE of 60 mg/kg, the mean (± SD) C_max was 1,495 ± 589 µg/mL (coefficient of variation [CV%]: 39) and mean (± SD) AUC_last was 230,991 ± 102,673 hour × µg/mL (CV% = 45), indicating moderate interindividual variability. No antidrug-antibody-positive samples were detected.

An exploratory objective of the study was to determine the effect of LFA102 treatment on serum prolactin levels in patients. The fold change from baseline increased in a dose-dependent manner, reached a maximum between days 8 and 15, and declined after day 15. The maximum fold-change in serum prolactin levels increased with doses up to 20 mg/kg and reached a plateau between 40 and 60 mg/kg. The temporal delay between PK and PD response is suspected to reflect the time needed for LFA102 to distribute to peripheral tissues, inhibit peripheral PRLR, and, consequently, lead to increased serum prolactin as a compensatory feedback mechanism.

The primary objective of this study was to determine the MTD and/or RDE of LFA102 in patients with MBC or mCRPC patients. An RDE of 60 mg/kg was established based on safety, PK, and PD, supported by the Bayesian logistic regression model. LFA102 demonstrated a favorable safety profile and tolerability at all doses tested. Dose proportionality analysis showed that serum LFA102 C_max and AUC_last were approximately linearly dose-dependent. LFA102 Vss was close to the volume of plasma, suggesting limited peripheral distribution typical of mAbs. At 60 mg/kg, the LFA102 half-life was 9 days, which, although within the reported range of mAbs, is slightly lower than the typical immunoglobulin G (IgG) with a half-life of approximately 25 days [29]. A possible explanation for this might be a lower affinity for the neonatal Fc receptor for IgG, which protects IgG from proteolytic degradation, leading to faster clearance.

No objective responses were observed in patients with MBC during this study. In patients with mCRPC, there were no PSA responses. Thirteen of 73 patients (18%) experienced stable disease as their best response to LFA102 treatment. The majority of patients (67 of 73 patients; 92%) discontinued the study because of disease progression. One explanation for the lack of antitumor activity is the possibility of insufficient exposure. After a single dose of LFA102 10 mg/kg by i.v., serum LFA102 C_max values were comparable between rodent and human subjects (268 µg/mL and 303 µg/mL, respectively; data not shown). Administration of a single dose of LFA102 10 mg/kg showed antitumor activity in a prolactin-dependent mouse tumor xenograft model (Nb2-11-luc) [28]. Consequently, the 60 mg/kg LFA102 dose in patients, which resulted in a mean C_max of 1,495 ± 589 µg/mL and a mean steady-state trough concentration of 106 ± 34 µg/mL, would be anticipated to provide sufficient LFA102 exposure to achieve efficacy.

In vitro data showed a high binding affinity of LFA102 to PRLR [28]. Assessing LFA102 binding to PRLR directly within tumors is impractical in patients; therefore, serum prolactin levels were used as a surrogate marker for PRLR inhibition. A sixfold change in serum prolactin levels from baseline was observed in patients treated with LFA102 60 mg/kg, indicative of inhibition of PRLR. The compensatory increase in serum prolactin indicates that LFA102 binds to PRLR in patients, ruling out poor target binding as causative of lack of efficacy. However, the source of serum prolactin increase could either be the tumor or the pituitary gland. No correlation between tumor PRLR expression and serum prolactin response was observed. Therefore, the observed increase in serum prolactin is more likely to be a pituitary-driven feedback to LFA102 as a result of peripheral, nontumoral PRLR inhibition rather than a tumor-specific process. Furthermore, the increase in serum prolactin was transient; it was maintained up to 15 days following LFA102 administration (supplemental online Fig. 3). Based on this observation, more frequent LFA102 dosing (e.g., every 2 weeks) could have resulted in sustained PRLR inhibition and perhaps a better efficacy profile.

Another potential explanation for the lack of LFA102 efficacy is that prolactin may not be an oncogenic driver in breast and prostate cancer in humans. Prolactin activity as an oncogenic driver in human tumors has been difficult to assess directly in preclinical models of human breast and prostate cancers [28]. Mouse prolactin does not activate human PRLR; therefore, human breast or prostate cancer cells or primary tumors cannot be used for xenograft models in mice to assess the requirement for PRLR signaling in driving oncogenesis [30]. Other explanations for the lack of LFA102 efficacy include unforeseen compensatory modulation of downstream signaling pathways in response to PRLR inhibition, or upregulation of other compensatory tumorigenic signaling pathways.

Finally, letrozole potentiates the efficacy of LFA102 when administered in combination in a rat mammary cancer model [28]. These preclinical results raise the possibility that although LFA102 monotherapy may not show antitumor activity, it may still have the potential to treat prolactin-dependent tumors in combination with other agents, such as novel hormonal pathway targeting agents in MBC and mCRPC. Furthermore, given the PD evidence of prolactin receptor blockade by LFA102, this drug has the potential to be used in conditions such as hyperprolactinemia that are associated with high prolactin levels.

Supplemental Figure 1.

(A): Linear view. (B): Semilogarithmic view.

Supplemental Figure 2.

Individual LFA102 concentration-time profiles by treatment group: semi-logarithmic view (cycle 1).

Supplemental Figure 3.

Geometric mean for fold change from baseline for serum prolactin versus time profiles by treatment group (cycle 1).

Supplemental Figure 4.

Correlation of maximum fold change from baseline serum prolactin with dose.

Supplemental Figure 5.

Correlation of serum prolactin exposure with baseline prolactin receptor expression, 60 mg/kg dose group. r² = .003; p = .7.

Abbreviations: AUEC, area under the effect curve; PRLR, prolactin receptor.

Table 1.

Patients’ characteristics

Supplemental Table 1.

Trial information

Supplemental Table 2.

Most common AEs (≥15% for all grades or ≥5% for grade 3/4 in all patients) regardless of study drug relationship

Supplemental Table 3.

Safety, tolerability, dose changes, and exposure to LFA102

Supplemental Table 4.

Best overall response to LFA102 treatment

Disclosures

Jean-Pascal Machiels: Novartis (RF); Nancy Lewis: Novartis, Jefferson (E); Ahmad Awada: Roche, Pfizer (C/A); Mark Stein: Jansen, Medivation, Oncoceutics, Amgen, Advaxis, Merck (RF); Rebecca Mosher: Novartis (E); Ernesto Wasserman: Novartis (E); Mohamed Elmeliegy: Novartis (E). The other authors indicated no financial relationships.

(C/A) Consulting/advisory relationship; (RF) Research funding; (E) Employment; (ET) Expert testimony; (H) Honoraria received; (OI) Ownership interests; (IP) Intellectual property rights/inventor/patent holder; (SAB) Scientific advisory board