Constitución Política de la República de Ecuador de

6.3.2.4.1. Lymphomas/Leukemias. In the first publica-

tion (Soffritti et al., 2005), an increase in the combined inci- dence of lymphomas and leukemias was reported for female rats treated with aspartame. From the graph (Figure 7) gener- ated from data provided in the publication, it is clear that there are no differences in incidence between males and females fed aspartame; but there is a noticeable difference between the male and female control groups. The low background incidence in the female controls, probably a result of the poor 2-year survival of this group, is what creates the significant differences in the treatment groups for females and not males, despite very sim- ilar incidences in aspartame-treated groups. The inherent con- founded interpretation of results when there is an abnormally low background tumor incidence in a control group is not unique to animal studies. This enigmatic feature of exposure-response analyses created by inordinately low baseline rates has been ob- served in major occupational epidemiology studies, such as the cohort studies of formaldehyde (Blair et al., 1986; Hauptmann et al., 2003, 2004) and acrylonitrile (Blair et al., 1998) workers conducted by the National Cancer Institute, and has stimulated reanalyses and reinterpretation of the NCI cohort data (Marsh et al., 2001, 2004, 2005).

Although there was an increasing dose-response in the first two dose groups compared with control for the female rats, no further increase occurred over the range of 400 to 10,000 ppm, despite consumption of diets containing up to 250 times as much aspartame, which puts into question the issue of a true

17_{Pyelonephritis=inflammation of the renal parenchyma, calyces, and}

pelvis, particularly due to local bacterial infection (Stedman’s Medical Dic-

tionary, 1995m).

18_{Pleuritis=inflammation of the pleura (Stedman’s Medical Dictionary,}

1995l).

19_{Peritonitis=inflammation of the peritoneum (Stedman’s Medical Dictio-}

nary, 1995j).

20_{Pericarditis=inflammation of the pericardium (Stedman’s Medical Dic-}

tionary, 1995i).

21_{Meningitis=inflammation of the membranes of the brain or spinal cord}

(Stedman’s Medical Dictionary, 1995e).

dose response. Furthermore, it is important that for all groups of females, the incidence rates of lymphoma and leukemia were within the range of historical controls. Based on data provided to EFSA by the Ramazzini Foundation, which included years 1984 to 1991 only, the average incidence of lymphomas/leukemias for females is 13.3% with a range of 4 to 25%. For males, the inci- dence of lymphomas/leukemias in all groups was also within the range of the reported historical controls (8–30.9%). Al- though other historical control incidences of tumors for Sprague- Dawley rats have been reported (Giknis and Clifford, 2004; Baldrick, 2005), the data from the Ramazzini Foundation is the best comparison, due to the fact that they use their own in-house bred animals and do not maintain a pathogen-free environment. The combination of the hemolymphoreticular tumor types, that is, lymphomas and leukemias, for statistical purposes was not justified in the view of the EFSA panel (2006), because tumors of different cellular origin were aggregated. It is note- worthy that Soffritti’s group has combined these tumors previ- ously and have been the only group to report high incidences of lymphoma/leukemia in association with exposure to methanol (Soffritti et al., 2002) and formaldehyde (Soffritti et al., 1989, 2002) while others have found no such association. The pool- ing of tumor types methodology was noted as a limitation in the WHO review of the Soffritti formaldehyde study (CICAD, 2002). Ward et al. (1990) reported that naturally occurring lym- phomas and leukaemias can and should be differentiated from induced tumors in chronic rodent bioassays based on organ dis- tribution, cytology, antigen markers, incidence of pneumonia, and age of appearance. No such analyses have been reported by Soffritti (Soffritti et al., 1989, 2002).

The high incidence of bronchopneumonia observed in this study is unusual for Sprague-Dawley rats, and is likely related to the high incidence of lymphoma/leukemias (EFSA, 2006). The progression of chronic murine pneumonia in rats has been shown to result in lymphoid neoplasmas (Innes et al., 1967; Nelson et al., 1967). Furthermore, elimination of chronic respi- ratory disease in rat colonies reduces the indicence of pulmonary lymonoid neoplasias to close to zero (Innes et al., 1967; Nelson et al., 1967).

The causative agent of many infections in non-specific

pathogen-free studies was reported to beMycoplasma pulmonis

(Lindsey et al., 1985). Affected sites ofM. pulmonisinclude

lungs, nasopharynx and middle ears, along with causing short- ened life span, altered xenobiotic metabolism and responses to

carcinogens.22_{Rats with pulmonary infections develop lesions}

in multiple sites earlier than rats free from pulmonary disease (Simms, 1967). Based on observations and studies in the 1950s and 1960s, toxicologists recommended that rats with chronic pulmonary disease should not be used for long-term experimen- tal work or toxicity studies (Innes et al., 1967; Nelson et al., 1967; Simms, 1967), which ultimately led to the establishment of pathogen-free animal suppliers for toxicity research. Therefore,

these factors render the results of the Ramazzini studies of doubt- ful validity.

It is impossible to interpret the lymphoma/leukemia data re- ported in the Ramazzini study because of the high and highly variable incidences of lung infections (Soffritti et al., 1989, 2002). Accordingly, the overall value of the study is open to question (EFSA, 2006).

Based on these factors, the EFSA (2006) and this panel have concluded that the slight increase in incidences of lym- phomas/leukemias in the rats fed aspartame were an incidental finding and should be dismissed.

6.3.2.4.2. Transitional-Cell Carcinomas of the Renal

Pelvis and Ureter and Their Precursors. The relevance of ob-

servations of tumors of the kidney, ureter and bladder in rat studies to potential risk of tumors to humans has been stud- ied extensively, due to the fact that these tumors are commonly observed in the rat with high doses of a variety of chemicals (Cohen, 1995a, 1995b). It is well established that nongenotoxic chemicals can induce tumors in these tissues when the doses of these chemicals exceed a threshold level (typically 1% of the diet), through a number of mechanisms such as cytotoxicity, formation of urinary precipitates, and increased cell prolifera- tion due to regernerative hyperplasia. The association of ureter foreign bodies with chronic irritation leading to urothelial cell proliferation and urothelial carcinogenesis have been well doc- umented and reviewed (Capan et al., 1999). Thus dose-related urinary tract calcification may be involved in the observations of urothelial hyperproliferation and neoplasms. An exposure of 1% in the diet has been calculated to be equivalent to a hu- man exposure of 30,000 mg/day for a 60-kg individual. Second, the rat has a much greater susceptibility to these tumors than hu- mans, due to differences in urinary protein levels (Cohen, 1995a, 1995b).

As aspartame has been clearly shown to be nongenotoxic, it is likely that the observation of these changes in rats fed high levels of aspartame are also the result of one of these imbalances

TABLE 17

Incidence of Schwannomas in male and female rats fed aspartame (Soffritti et al., 2006)

Male rats Female rats

Dose

(mg/kg bw/day) Cranial Other sites Cranial Other sites

0 1/1501(0.7%) 0/150 (0%) 0/150 (0%) 0/150 (0%) 4 1/150 (0.7%) 0/150 (0%) 1/150 (0.7%) 1/150 (0.7%) 20 1/150 (0.7%) 2/150 (1.3%) 0/150 (0%) 0/150 (0%) 100 2/150 (1.3%) 0/150 (0%) 1/150 (0.7%) 2/150 (1.3%) 500 2/100 (2%) 0/100 (0%) 1/100 (1%) 0/100 (0%) 2500 3/100 (3%) 0/100 (0%) 1/100 (1%) 0/100 (0%) 5000 3/100 (3%) 1/100 (1%) 1/100 (1%) 1/100 (1%)

Note.1_{Number of rats with Schwannomas/number of rats in group (percent); bw=body weight.}

that are specific for the rat. Therefore, although this finding is probably treatment-related in the rat, it is of no relevance for human risk due to the fact that it occurs only at high doses and is rat-specific (EFSA, 2006).

6.3.2.4.3. Malignant Schwannomas of Peripheral Nerves.

The incidence of malignant Schwannomas of peripheral nerves reported in the second publication (Soffritti et al., 2006) is shown in Table 17. The biological relevance of the incidence of periph- eral nerve Schwannomas is unclear due to the low number of tumors, the relatively flat dose-response curve over a very wide dose range, and uncertainty about the diagnosis of the tumors (EFSA, 2006).

6.3.2.4.4. Hyperplasia and Adenomas of the Olfactory Tis-

sues. This observation is not indicative of likely risk of nasal

tumors in humans from oral consumption of aspartame for sev- eral reasons. First, the route of exposure in humans is strictly oral, whereas in the rats, which are obligate nasal breathers, in- halation of the suspected powdered diet containing high levels of aspartame may have caused an inflammatory response, which played a role in the hyperplasia. Specific details of the compo- sition and form of the diet were not provided; however, it is assumed that the diet was in powder form, as pelletting would lead to a loss of aspartame. Second, there is the disagreement regarding the pathological classification of the nasal adenomas. Two cases of adenomas provided by the Soffritti group to the NTP Pathology Working Group were unanimously diagnosed by NTP as hyperplasia. Therefore, severe pathological grading of hyperplasia may have contributed to the observations of adeno- mas. Lastly, but perhaps most important, there is extensive data demonstrating significant species differences in susceptibility to the development of nasal tumors due to exposure from many compounds (Kai et al., 2006; Jeffrey et al., 2006). Jeffery et al. (2006) reviewed the effects of a variety of organic chemicals demonstrated to produce toxicity and carcinogenicity in rodents following oral administration and systemic distribution. None

of the known rodent nasal carcinogens that lack DNA reactivity are considered to represent any cancer hazard, including nasal tumors, in humans. In addition, DNA-reactive rodent nasal car- cinogens have not been associated with nasal tumors in humans, although they may be associated with tumors at other sites. Hu- mans have much lower levels of biotransformation enzymes in the nasal mucosa, and usually much lower levels of exposure, which accounts for species differences in susceptibility. There- fore, the observation of hyperplasia, and possibly adenomas in rats fed aspartame in a powdered diet does not provide any evi- dence for increased tumor risk in humans consuming aspartame orally.

6.3.2.4.5. Total Malignant Tumor-Bearing Animals. Ac-

cording to EFSA (2006), the unpublished report of the NTP Pathology Working Group described several cases of discrepan- cies regarding the diagnosis of lesions from this study, in which the NTP classification was less severe than the diagnosis as- signed by the Soffritti group. Given the high incidence of in- fection that likely led to the lymphomas/leukemias and that the mechanism of renal and ureter tumors is irrelevant to humans, the EFSA stated that these tumors should have been excluded from the total tumor analysis. The observation of a total of 12 brain tumors in the treatment groups was not dose-related and therefore, although none was observed in the control groups, aspartame cannot be considered responsible for the brain tu- mors. Considering these observations, the data on total tumors as presented in the Soffritti et al. (2006) publication cannot be accepted as evidence for the carcinogenic potential of aspartame (EFSA, 2006).

Interestingly, in 2002, Soffritti and colleagues (2002) re- ported a review of 200 compounds that had been evaluated by the Ramazzini Foundation using the same protocol as in the recent studies. Aspartame was listed in this review. The only details mentioned in the review article were that the study in- cluded 1800 rats and that the route of exposure was ingestion.

Aspartame wasnotlisted in the summary as 1 of the 47 agents

that showed “clear evidence” of carcinogenicity. It is not known whether this was a separate study or is part of the experiment reported in 2005. As the EFSA report (2006) states that the study reported in 2005 was conducted from June 1997 to May 2000, it is most likely that these are the same studies; however, the num- ber of animals is half of the 3600 rats reported to be in the 2006 studies. If the study was conducted in two stages, this should have been reported to ensure all groups were included in both stages and the appropriate statistics were employed for such a design.

In addition to the reevaluation of data and scientific evidence for each of the conclusions described above, EFSA also took into consideration the negative findings in recent NTP studies of aspartame using transgenic mice (described below in Section 6.3.3) and the negative findings in a recent National Cancer Insti- tute epidemiological study (described below in Section 6.9.3.1).

The conclusion of the EFSA evaluation was that “there is no need to further review the safety of aspartame nor to revise the previously established ADI for aspartame (40 mg/kg bw/day). The panel also noted that intakes of aspartame in Europe, with levels up to 10 mg/kg bw/day, are well below the ADI.”

In summary, the Soffritti reports alleging carcinogenicity are contradicted by many publications and every scientific consid- eration. Many potential flaws have been suggested in this report; whether these or some other unidentified flaw is responsible for their incorrect allegations is not known. Nevertheless, it can be confidently stated that these reports provide no credible evidence that aspartame is carcinogenic.