Leucovorin rescue allows effective high-dose pralatrexate treatment and an increase in therapeutic index in mesothelioma xenografts

Purpose To investigate the ability of leucovorin (LV) to abrogate dose-limiting toxicities of pralatrexate (PDX) while maintaining efficacy, in vivo. Methods H2052 mesothelioma cells were treated with the antifolates methotrexate (MTX), PDX and pemetrexed, with and without LV rescue 24 h later. Cell killing was evaluated 48 h later. Female nude mice bearing H2052 xenografts were treated with varying doses and schedules of the antifolate PDX and LV. Results In vitro, H2052 cells were more sensitive to PDX as compared to MTX and pemetrexed. Administration of LV 24 h after antifolate treatment reduced efficacy of antifolates MTX and pemetrexed, but not PDX. In vivo, LV was found to reduce toxicity of PDX at the maximum tolerated dose without sacrificing efficacy. Lethal doses of PDX were rescued by LV, and mice bearing the H2052 tumor demonstrated prolonged and enhanced tumor regression. Conclusions High-dose PDX with subsequent LV rescue may be a viable treatment strategy in mesothelioma and other cancers. The inclusion of LV rescue into new and existing PDX treatment protocols should be explored as a way to expand the tolerability and effectiveness of PDX in the clinic.

In this study, we demonstrate the protective effect of LV on PDX toxicity and show that a significant antitumor effect is achievable with high-dose PDX and LV rescue in mesothelioma xenografts.

Materials and methods
In vitro cytotoxicity Five thousand H2052 pleural mesothelioma cells per well were plated in 96-well plates in RPMI 1640 media (Gibco) supplemented with 10 % dialyzed FBS (Invitrogen). The following day, spent media was removed and fresh media containing drug was added and plates were incubated for 24 h. Drug containing media was removed and media containing 2μM LV was added and plates were incubated for 24 h. LV containing media was removed and fresh media was added to the plates incubated for 48 h. The Cell Titer 96 Aqueous One Solution (Promega) assay was used to assess cell viability at the end of the experiment according to manufacturers' protocol. Data were analyzed using the GraphPad Prism 6 software package (GraphPad Software Inc.).

In vivo experiments
Eight-to 10-week-old female nude mice (Taconic) were subcutaneously injected on the right flank with 5 million H2052 cells in a 1:1 mixture of matrigel (BD Biosciences) and PBS (Gibco). Mice were monitored until palpable tumors developed and reached a size of 100 mm 3 . Mice were then randomized and split into the following treatment groups: saline control; 60 or 180 mg/kg PDX on days 1, 4 and 7; 60 or 180 mg/kg PDX on days 1, 4 and 7 followed by 50 mg/kg LV 24, 32 and 48 h after each PDX administration; 50 mg/kg LV following the above schedule. General toxicity was monitored through weight loss. Tumor dimensions were measured every 2-4 days using calipers, and tumor volume was calculated using the following equation: Volume = (width 2 ) × (length/2). All animal experiments were approved by the Rutgers Institutional Animal Care and Use Committee.

Results
To assess the sensitivity to antifolate therapy, a comparison of MTX, PDX and pemetrexed in the pleural mesothelioma cell line H2052 was performed. After a 24 h treatment with drug, the media was removed and replaced with fresh media and cultured a further 72 h. PDX demonstrated significant efficacy over MTX in this assay, with IC 50 values of 0.625 and 80 nM, respectively (Fig. 2a, b). The large difference between PDX and MTX is likely due to enhanced transport and retention of PDX. H2052 cells were also more sensitive to PDX than pemetrexed (Fig. 2c), an antifolate currently used in the clinic to treat mesothelioma.
As LV can abrogate antifolate toxicity, we next examined if efficacy of PDX would be compromised by the addition of LV 24 h later. H2052 cells were treated with antifolate as before, but media containing 2μM LV was added for another 24 h period, followed by removal of drugs and culture in fresh media for a further 48 h. The addition of 2μM LV raised the IC 50 slightly for MTX and pemetrexed (Fig. 2d). The effect of LV on tumor cell sensitivity to PDX was only at high PDX concentrations, and was not expected to be significant in in vivo xenografts. We predicted that using PDX in combination with LV rescue would result in an increased therapeutic index.
The effect of LV treatment on PDX efficacy and toxicity was examined using H2052 xenografts in nude mice (Fig. 3). Using 60 mg/kg as the equivalent clinical dose, tumor regression was seen both with and without LV treatment in this cohort. Efficacy was nearly identical, but mice that were treated with LV had less toxicity and bodyweight remained close to control animals.
To mimic high-dose MTX treatment with LV rescue, using PDX, we included a cohort treated at 180 mg/kg PDX with and without LV rescue. Without LV rescue, mice lost a large amount of weight, had poor BCS scores and had to be killed according to institutional protocol (Fig. 3b). Mice rescued with LV exhibited some weight loss, and a durable response lasting several weeks was observed (Fig. 3a).

Discussion
Surgical resection followed by chemotherapy is the preferred method of treatment for malignant mesothelioma [9]. The antifolate pemetrexed is currently used alone and in combination with cisplatin in treating mesothelioma, with modest response rates seen in the clinic [10]. Chemotherapy is primarily palliative in mesothelioma and there is a need for new and innovative therapies that have the ability to improve patient outcomes [11].
There were no responses to PDX in a phase II study in mesothelioma [12]. To improve tolerability, PDX was administered at 135 mg/m 3 biweekly instead of weekly, possibly limiting clinical effectiveness. A dose schedule of PDX weekly with LV rescue may demonstrate improved antitumor activity with tolerable toxicity. Supporting this conclusion, a small recent case series exploring LV rescue and PDX in T-cell cutaneous lymphoma found that LV rescue allowed patients to tolerate PDX therapy while preserving efficacy [8]. Dose scheduling must be optimized in patients, as the interval between antifolate and LV administration has recently been shown to be significant in determining the efficacy of therapy in vitro [13].
The incorporation of LV rescue with current PDX dosing regimens may prove an effective method to prevent DLTs and increase the therapeutic index. While a phase II study of PDX in patients with mesothelioma did not show responses, higher doses with the inclusion of LV rescue