Letter to the Editor

Jolette et al. (Toxicol Pathol, 34, 929–940, 2006) report the effects of recombinant human PTH(1–84) in a 2-year rat carcinogenicity study and compare their results to our studies with recombinant human PTH(1–34), teriparatide (Vahle et al., 2002, 2004). The authors state that PTH(1–84) has less carcinogenic potential than PTH(1–34) (Vahle et al., 2002, 2004). This interpretation appears to be based on 1) the relatively small numeric difference in bone tumor rates between the current study and a previously published study of PTH(1–34) and 2) the identification of a no-effect level (NOEL) for bone tumors at the 24-month time point in the Jolette et al. study. The authors suggest that this difference may be related to a theoretical role for the C-terminal region in PTH(1–84) which is not present in PTH(1–34).

A direct side-by-side comparison of these 2 peptides in a 2-year rat carcinogenicity study has not been conducted; therefore, it is not appropriate to draw conclusions based on the numerical differences in tumor incidence in 2 independent studies. Variations in animal populations over time and differences in diagnostic thresholds used by the pathologists can impact tumor incidence data. Although both studies appeared to use similar diagnostic criteria, our experience in the PTH(1–34) rat studies was that a spectrum of hyperplasic and neoplastic lesions were observed and for some animals the distinction between hyperplasic and neoplastic changes were challenging and necessitated the use of a panel of expert pathologists (Pathology Working Group). Without a direct comparison of the 2 peptides using the same population of animals and the same pathology evaluators, it is not possible to discern if there were true differences in tumor incidences at comparable multiples of human exposure.

Consideration of the magnitude of the pharmacodynamic effects on bone mass is critical in interpreting these results. The occurrence of bone neoplasms in rats treated with PTH(1–34) or PTH(1–84) appears to be related to the exaggerated pharmacodynamic effect on rat bone by these compounds. Jolette et al. state that both molecules had similar effects on bone mass. We propose that PTH(1–34) produced greater increases in bone mass than PTH(1–84) as shown in Figure 1 (here)(BMC, % increase over vehicle controls).

Jolette et al. did not reference our no-effect level for bone tumors observed in rats treated with 5 mcg/kg for 20 months (Vahle, 2004). We demonstrated that duration of PTH treatment, as well as dose, are key factors in the development of bone neoplasms in rats. The no-effect level for PTH(1–84) was 24 months at a 4.6-fold multiple of human exposure while the no-effect level for PTH(1–34) was 20 months at a 3-fold multiple of human exposure. Therefore, a no-effect level for bone tumors has been established for both compounds at doses (based on systemic exposures) and durations (based on percent of lifespan) that are in excess of the clinical treatment regimen. Even without accounting for potential differences in pharmacodynamic effects in the rat, the no-effect levels differ by less than an order of magnitude.

Collectively, the discussion section of the Jolette et al. seems to imply an improved clinical safety profile for PTH(1–84) compared to PTH(1–34) that is not warranted by a thorough review of the evidence. We believe that assessing the relevance of rat bone tumors observed with these peptides to human clinical use is much more complex than simply evaluating the no-effect levels in rodents. The sensitivity of the rat skeleton to the anabolic effects of these agents, the differences in skeletal physiology between rodents and primates, the lack of bone tumors in nonhuman primate models (Vahle, 2006), and the differences between the near-lifetime treatment of rodents and a relatively limited duration of clinical use must all be considered when assessing the ability of these rodent studies to predict effect in humans.