This Phase I study demonstrates that the novel topoisomerase-I inhibitor TP300 has a good tolerability profile, and achieved several key aims that were central to its design. More specifically, as an inactive pro-drug it is rapidly converted to the active form TP3076, then metabolized to TP3011 in a consistent manner, not influenced by genetic polymorphisms. The likelihood of unpredictable, severe diarrhoea is diminished by the absence of the variable glucuronidation associated with SN-38. As predicted, TP300 does not cause acute diarrhea, which results from acetylcholine esterase inhibition . Target interaction with the induction of DNA strand breaks was shown.
The main toxicity of TP300 was haematologic with neutropenia and, to a lesser extent thrombocytopenia, being dose limiting. In general, neutropenia was short lived; no patient received G-CSF support (acutely/prophylactically). At the maximum achievable dose, 12 mg/m2, grade 4 haematologic toxicity was observed (2/4 patients). As there had been no grade 3/4 haematologic toxicity at 8 mg/m2, 10 mg/m2 was explored in 12 patients. At 10 mg/m2 3 patients experienced grade 4 haematologic toxicity and although generally well tolerated, there was a risk of short lived but significant neutropenia and thrombocytopenia. The recommended Phase II starting dose is 8 mg/m2, escalating to 10 mg/m2 on subsequent cycles, if initial treatment is well tolerated.
In marked contrast to irinotecan, gastrointestinal toxicity was in general mild, with no diarrhoea greater than grade 2. Likewise, there were no acute cholinergic reactions with its associated early diarrhoea [8, 14, 15]. This validates the design of TP300 as acute cholinergic reactions are associated with the 4-piperidinopiperidine moiety at the 10-position of irinotecan , not found in TP300.
Pharmacokinetic data confirm that TP300 is rapidly converted in plasma to the active metabolite TP3076, supporting a pH dependent chemical change occurring at physiological conditions. Hepatic aldehyde oxidase converts TP3076 to a further metabolite TP3011, which reaches maximum concentrations 3–5 hours after the end of infusion, and also has potent topoisomerase-I inhibitory activity. Pharmacogenetic analysis of aldehyde oxidase genotype, which was reported to affect the azathioprine-treated outcome , did not show any effect on exposure to either TP3076 or TP3011. Glucuronidated TP3076 was not detected, reflecting UGT1A1 variant status had no influence on exposure to either TP3076 or TP3011. These pharmacokinetic data reflect the design strategy. There may be a small effect of CYP2D6 metaboliser genotype on exposure to TP3076, and consequently TP3011. The AUC of TP3076 and TP3011 were linearly proportional up to 10 mg/m2, but at 12 mg/m2 there was greater inter-patient variability.
There was a strong relationship between the combined total AUC of TP3076 and TP3011 and the nadir neutrophil count, with an AUC of ≥ 4.5 hr.μmol/L generally correlating with a more significant neutrophil fall, specifically 5 of 8 patients (62.5%) with an AUC above this value experienced dose limiting neutropenia. With a 3-weekly dosing regimen of irinotecan, at 350 mg/m2, DLTs occur at AUCs of the active component, SN-38, of 1.5 μmol.h/L , SN-38 AUC is variable, influenced by UGT polymorphism. The active components of TP300 (TP3076 and TP3011) are equipotent to SN38 as Topo-1 inhibitors  and are not influenced by UGT polymorphisms. This means, therefore, that the combined AUC of the active components of TP300 is approximately 3-fold greater than that of SN-38, with reduced inter-individual variability indicating greater predictability of toxicity.
The comet assay demonstrated a consistent pattern with increased PBMC DNA strand breaks 1 hour after the end of infusion, generally falling by 3 hours. A similar pattern with modest and transient appearance of strand breaks was seen with temozolomide . Although there were more strand breaks at higher TP300 doses, this was less clear than the relationship between pharmacokinetic exposure and neutrophil fall. However the comet data give valuable proof-of-principle that TP300 is damaging DNA, but the semi-quantitative nature does not allow a biologically optimal dose of TP300 to be identified. Without published data on DNA strand breaks in patients treated with irinotecan, a direct comparison with TP300 cannot be made. A more relevant pharmacodynamic endpoint in future may be to measure DNA strand breaks in tumour cells.
There were no objective tumour responses. However, one patient with metastatic gastric adenocarcinoma, bilateral ovarian metastases and malignant ascites requiring paracentesis prior to treatment had complete resolution of her ascites although there was no radiological change in the size of the metastatic deposits whilst receiving 7 cycles of TP300. A further 5 patients with metastatic adenocarcinoma of the colon or rectum had stable disease as their best response, with 2 having disease control for at least 4 cycles. All of these patients had received irinotecan as part of their previous chemotherapy with the patients having the most durable disease control on TP300 having had a prior response to irinotecan chemotherapy.
Topo-1 inhibitors remain clinically important in the treatment of patients with cancer. TP300 has advantages over other agents in this class in terms of tolerability and the predictability of its principle toxicity, myelosuppression. Along with the apparent PK advantage of TP300 over irinotecan, biological activity evidenced by DNA strand breaks, and preliminary evidence of clinical activity, these data warrant further evaluation of TP300.