Module 3 - Strategic case studies in practice

1430 ng) to those cohort of concern N- nitrosamines. As described in section 2.2 there are already cases where up to three different N- nitrosamines have been found in one pharmaceutical indicating that exposure to multiple N- nitrosamine via pharmaceuticals, although rare, may occur. In addition, patients may take more than one N- nitrosamine-containing medicine, which would further increase their daily N- nitrosamine intake. As outlined in paragraph 2.4, the risk is considered additive. Environmental exposure varies based on lifestyle. Patients with a “healthy” lifestyle may potentially have lower than average exposure to volatile N- nitrosamines. When taking pharmaceuticals containing N- nitrosamines controlled at limits calculated based on ICH M7(R1) approach, these patients may be exposed to higher exposures compared to the average population. The ad-hoc expert group suggested an approach to not exceed the acceptable risk level in case of more than one N -nitrosamine being present in a finished product. In case more than one N -nitrosamine occurs in manufacture it may be acceptable to limit the sum of N -nitrosamines to the limit of the most potent one found. Acceptability of such a concept may depend on the capability of effective control (see section 3.1). The SWP suggested a different approach to reach the goal of not exceeding the anticipated risk level of 1:100,000. When multiple N -nitrosamines are present in a single product and the total risk does not exceed the 10 -5 tumour risk level, the proposed limits are considered acceptable. The SWP recommends setting a specific limit for each nitrosamine with the sum of all detected nitrosamines not exceeding the total risk level of 1 in 100,000 (see section 3.5.2), which is considered an acceptable alternative approach, as it also ensures an excess lifetime risk below 1:100.000. The general paradigm of linearity of life-time dose and increase of cancer risk in the low dose range for mutagenic carcinogens is derived from rodent life-time bioassays. For NDMA and NDEA the studies used and cited are the studies by Peto (1991). There, linearity for the dose response in the low dose range was considered demonstrated for liver (with regard to the lowest dose tested in the experiment and reaching a sensitivity of ~5 cancer cases per 100 animals) but not for oesophageal neoplasms. The mathematical explanation the authors offer is that “ the appropriate low-dose dose-response relationships predict that if there is an appreciable background of "spontaneous" neoplasms of whatever type is of interest, then the dose-response relationship is likely to be simply linear at dose rates so low that the induced risk does not greatly exceed the background risk. No general predictions can be made of the shape of the dose-response relationship at low doses if the spontaneous rate is immeasurably small, as for oesophageal neoplasms in the present study. Consequently, it is unsurprising to note that at low doses the onset rate of oesophageal cancer appears not to be simply proportional to dose. For liver cancer, however, the background rates are about 8%, which is appreciable ”. Furthermore, the relationship of dose and time to tumour onset is not linear and seems different between tumour entities. In this respect it also needs to be noted that the background rates for liver tumours in humans is far lower than that for rats and significantly different between different age groups. According to the Ad- hoc expert group there also seem to be significant doubts among experts on whether this linearity for some tumours in rodents is transferable to humans. This is also stated by Peto et al. (1991a) in their publication: “ This provides us with what is probably a reasonably reliable estimate (despite the practical impossibility of direct confirmation) of the effects of ppb nitrosamine concentrations on rats under these experimental circumstances, but it does not provide reliable information as the effects of ppb nitrosamine concentrations on humans, and it would be a serious distortion of these experimental results to suggest otherwise. ” Therefore, although according to ICH M7(R1) a simple linear extrapolation of accumulated lifetime acceptable dose is made with some additional adjustment factor, it is currently not recommended to generally accept the ‘less-than-life’ time concept in ICH M7(R1) for setting higher N-nitrosamine limits in drug products taken less than lifetime. Accepting the LLT approach could lead to high acute nitrosamine intake, especially with medicines given at high doses and for a short period of time.

EMA/369136/2020

Page 49/90