[Federal Register Volume 63, Number 64 (Friday, April 3, 1998)]
[Rules and Regulations]
[Pages 16417-16433]
From the Federal Register Online via the Government Publishing Office [www.gpo.gov]
[FR Doc No: 98-8750]


=======================================================================
-----------------------------------------------------------------------

DEPARTMENT OF HEALTH AND HUMAN SERVICES

Food and Drug Administration

21 CFR Part 172

[Docket No. 87F-0086]


Food Additives Permitted for Direct Addition to Food for Human 
Consumption; Sucralose

AGENCY: Food and Drug Administration, HHS.

ACTION: Final rule.

-----------------------------------------------------------------------

SUMMARY: The Food and Drug Administration (FDA) is amending the food 
additive regulations to provide for the safe use of sucralose as a 
nonnutritive sweetener in food. This action is in response to a 
petition filed by McNeil Specialty Products Co.

DATES: The regulation is effective April 3, 1998; written objections 
and requests for a hearing by May 4, 1998. The Director of the Office 
of the Federal Register approves the incorporation by reference in 
accordance with 5 U.S.C. 552(a) and 1 CFR part 51 of certain 
publications in Sec. 172.831(b) (21 CFR 172.831(b)), effective April 3, 
1998.
ADDRESSES: Submit written objections to the Dockets Management Branch 
(HFA-305), Food and Drug Administration, 12420 Parklawn Dr., rm. 1-23, 
Rockville, MD 20857.

FOR FURTHER INFORMATION CONTACT: Blondell Anderson, Center for Food 
Safety and Applied Nutrition (HFS-206), Food and Drug Administration, 
200 C St. SW., Washington, DC 20204, 202-418-3106.

SUPPLEMENTARY INFORMATION:

Table of Contents

I. Introduction
II. Evaluation of Safety
    A. Estimated Daily Intake
    B. Evaluation of Toxicological Testing Results
      1. Pharmacokinetics and Metabolism
        a. Comparative pharmacokinetics
        b. Sucralose metabolism
      2. Genotoxicity Testing
      3. Reproductive/Developmental Toxicity Studies
        a. Sucralose
          i. Two-generation reproductive toxicity study in rats 
(E056)
          ii. Teratology study in rats (E030)
          iii. Teratology studies in rabbits (El34)
        b. Sucralose hydrolysis products
          i. Two-generation reproductive toxicity study in rats 
(E052)
          ii. Teratology study in rats (E032)
        c. Male fertility studies of sucralose and its hydrolysis 
products in rats (E016, E038, E090, and E107)
      4. Chronic Toxicity/Carcinogenicity Studies
        a. Sucralose
          i. Combined chronic toxicity/carcinogenicity study in rats 
(E057)
          ii. Carcinogenicity study in mice (E055)
          iii. Chronic toxicity study in dogs (E051)
        b. Sucralose hydrolysis products--carcinogenicity study in 
rats (E053)
      5. Special Toxicological Studies
        a. Body weight gain (E058, E130, E143, E151, E160, E161)
          i. The palatability hypothesis
          ii. The agency's evaluation of the palatability hypothesis
          iii. Resolution of the body weight gain decrement issue

[[Page 16418]]

        b. Immunotoxicity study in rats (E162)
        c. Neurotoxicity studies in mice and monkeys (E008 and E009)
        d. Diabetic Studies (E156, E157, E168, E170, E171)
    C. Acceptable Daily Intake Estimate for Sucralose
III. Comments
    A. Determination of No-Observed-Effect Level and ADI
      1. No-Observed-Effect Level in the Chronic Toxicity Study
      2. No-Observed-Effect Level in Developmental Toxicity Studies
      3. Derivation of ADI
    B. Immunotoxic Potential of Sucralose
    C. Mutagenicity of 1,6-DCF
    D. Renal Effects
    E. Fetal Edema
    F. Bioaccumulation
    G. Antifertility Effects
    H. Neurotoxicity Effects
    I. Exposure to Sucralose Hydrolysis Products
    J. The Need for Studies in Special Populations
    K. Labeling
IV. Conclusion
V. Environmental Effects
VI. References
VII. Objections

I. Introduction

    In a notice published in the Federal Register of May 8, 1987 (52 FR 
17475), FDA announced that a food additive petition (FAP 7A3987) had 
been filed by McNeil Specialty Products Co. (McNeil), P.O. Box 3000, 
Skillman, NJ 08558-3000 proposing that the food additive regulations be 
amended to provide for the safe use of sucralose (1,6-dichloro-1,6-
dideoxy--D-fructofuranosyl-4-chloro-4-deoxy--D-
galactopyranoside) as a nonnutritive sweetener in food where standards 
of identity do not preclude such use. (McNeil's address has since 
changed to 501 George St., New Brunswick, NJ 08558-3000.)
    The petitioner has requested the use of sucralose in 15 food 
categories as described in Sec. 170.3 (21 CFR170.3(n)) as follows: 
Baked goods and baking mixes (Sec. 170.3(n)(1)); beverages and beverage 
bases (nonalcoholic) (Sec. 170.3(n)(3)); chewing gum 
(Sec. 170.3(n)(6)); coffee and tea (Sec. 170.3(n)(7)); confections and 
frostings (Sec. 170.3(n)(9)); dairy product analogs 
(Sec. 170.3(n)(10)); fats and oils (Sec. 170.3(n)(12)); frozen dairy 
desserts and mixes (Sec. 170.3(n)(20)); fruit and water ices 
(Sec. 170.3(n)(21)); gelatins, puddings, and fillings 
(Sec. 170.3(n)(22)); jams and jellies (Sec. 170.3(n)(28)); milk 
products (Sec. 170.3(n)(31)); processed fruits and fruit juices 
(Sec. 170.3(n)(35)); sugar substitutes (Sec. 170.3(n)(42)); and sweet 
sauces, toppings, and syrups (Sec. 170.3(n)(43)). This final rule lists 
all of the requested uses.
    Sucralose has also been referred to as trichlorogalactosucrose or 
4,1',6'-trichlorogalactosucrose. The Chemical Abstracts Service 
Registry number (CAS Reg. No.) for sucralose is 56038-13-2. Sucralose 
is a disaccharide that is made from sucrose in a five-step process that 
selectively substitutes three atoms of chlorine for three hydroxyl 
groups in the sugar molecule. It is produced at an approximate purity 
of 98 percent. Sucralose is a free-flowing, white crystalline solid 
that is soluble in water and stable both in crystalline form and in 
most aqueous solutions; it has a sweetness intensity that is 320 to 
1,000 times that of sucrose, depending on the food application.
    Hydrolysis of sucralose can occur under conditions of prolonged 
storage at elevated temperatures in highly acidic aqueous food 
products. The hydrolysis products are the monosaccharides, 4-chloro-4-
deoxy-galactose (4-CG) and 1,6-dichloro-1,6-dideoxyfructose (1,6-DCF).
    McNeil's original submission to FDA contained data and information 
from toxicity studies in several animal species, other specific tests 
in animals, and information from clinical tests in human volunteers. 
The toxicity data base included: Short-term genotoxicity tests, 
subchronic feeding studies, chronic toxicity/carcinogenicity studies in 
rats and mice, a chronic toxicity study in dogs, reproductive toxicity 
studies in rats, teratology studies in rats and rabbits, male fertility 
studies in rats, and neurotoxicity studies in mice and monkeys. Other 
specific tests conducted with animals included: Pharmacokinetics and 
metabolism studies on sucralose in several species, mineral 
bioavailability studies in rats, and several studies related to food 
consumption and palatability in rats and dogs. Human clinical testing 
addressed the pharmacokinetics and metabolism of sucralose, in addition 
to its potential effects on carbohydrate metabolism. The petitioner 
also submitted a report prepared by a panel of experts in various 
scientific disciplines who independently evaluated and critiqued the 
sucralose data base to identify areas of potential controversy.
    During the course of the agency's evaluation of the sucralose 
petition, McNeil submitted additional studies that had been conducted 
in response to questions and concerns raised by the governmental 
reviewing bodies of other countries. The additional studies included a 
6-month gavage study in rats, two comparative pharmacokinetics studies 
in rats and rabbits, an immunotoxicity feeding study in rats, and study 
of unscheduled deoxyribonucleic acid (DNA) synthesis.
    In response to an issue raised by FDA, the petitioner submitted a 
6-month sucralose feeding study in rats, with a dietary restriction 
design, to evaluate the toxicological significance of a body weight 
gain decrement effect observed in sucralose-treated rats.
    In anticipation of the potential wide use of sucralose in persons 
with diabetes mellitus and to address concerns raised by a diabetic 
association group in Canada, the petitioner performed a series of 
clinical studies. Because of results observed in diabetic patients that 
were treated with sucralose in a 6-month clinical study, the petitioner 
requested (in 1995) that the agency withhold its final decision on the 
safety of sucralose until that observation could be further 
investigated. At that time, the petitioner initiated additional studies 
with the main objective of evaluating the effects sucralose would have 
on glucose homeostasis in patients with diabetes mellitus.

II. Evaluation of Safety

    In the safety evaluation of a new food additive, the agency 
considers both the projected human dietary exposure to the additive and 
the data from toxicological tests submitted by the petitioner. Other 
relevant information (e.g., published literature) is also considered. 
The available data and information submitted in a food additive 
petition must establish, to a reasonable certainty, that the food 
additive is not harmful under the intended conditions of use.

A. Estimated Daily Intake

    In determining whether the proposed use of an additive is safe, FDA 
typically compares an individual's estimated daily intake (EDI) of the 
additive to the acceptable daily intake (ADI) established from the 
toxicity data. The agency determines the EDI by making projections 
based on the amount of the additive proposed for use in particular 
foods and on data regarding the consumption levels of these particular 
foods. The proposed use levels of sucralose are supported by taste 
panel testing that was reported in the petition. The petitioner also 
submitted survey information on the consumption of the food types for 
which the use of sucralose was requested.
    The agency commonly uses the EDI for the 90th percentile consumer 
of a food additive as a measure of high chronic exposure. For the 
requested food uses of sucralose, the agency has determined the 90th 
percentile EDI for consumers 2 years old and older (``all ages'') to be 
98 milligrams per person per day (mg/p/d), equivalent to

[[Page 16419]]

approximately 1.6 mg per kilogram of body weight per day (mg/kg bw/d) 
(Refs. 1 and 2).
    Because sucralose may hydrolyze in some food products (although 
only to a small extent and only under limited conditions), the 
resulting hydrolysis products may also be ingested by the consumer. 
Therefore, the agency has also calculated EDI's for the combined 
hydrolysis products of sucralose. The 90th percentile EDI is 285 
micrograms per person per day (g/p/d), equivalent to 4.7 
g/kg bw/d (Refs. 1 and 2).

B. Evaluation of Toxicological Testing Results

    The major studies relevant to the safety decision regarding the 
petitioned uses of sucralose are discussed in detail in section II.B of 
this document. The individual studies are identified by ``E'' numbers, 
as designated by McNeil in the sucralose petition.
1. Pharmacokinetics and Metabolism
    Studies were conducted to characterize and compare the metabolic 
fate of sucralose in various animal species to that seen in humans in 
order to assist in the selection of an appropriate animal model for 
safety extrapolation to humans.
    a. Comparative pharmacokinetics. The absorption, metabolism, and 
elimination of sucralose have been studied in several different animal 
species, including humans. Based on its evaluation of these studies, 
the agency concludes that, in general, sucralose is poorly absorbed 
following ingestion, with 36 percent or less of the dose absorbed in 
rats (E004 and E137), mice (El46), rabbits (El24), dogs (E049 and 
E123), and humans (E003, E033, and E128). Although there is consistency 
among laboratory animal species in the routes of elimination of 
sucralose when administered by the intravenous route (80 percent 
urinary, 20 percent fecal), the amounts of sucralose absorbed and rates 
of elimination after oral administration differ considerably (Ref. 3). 
The agency estimates that about 5 percent of the ingested dose is 
absorbed from the gastrointestinal system of rats, while that in 
rabbits and mice ranged from 20 to 33 percent. Gastrointestinal 
absorption of sucralose by the dog was in the range of 33 to 36 
percent. Studies in human male volunteers showed absorption values in 
the range of 11 to 27 percent, which is between the ranges observed for 
rats (lower bound) and rabbits and mice (upper bound). In all of the 
species tested, plasma disappearance curves are biphasic (E003, E004, 
E049, E123, E128, E146, El63, and E164). With the exception of the 
rabbit (El64), these curves are dominated by phase 1, with a half-life 
of 2 to 5 hours. In the rabbit elimination is dominated by phase 2, 
with a half-life of 36 hours (El64) (Ref. 3). The longer half-life of 
sucralose in the rabbit was initially thought to be the result of 
reingestion of sucralose. However, study E164, which was specifically 
designed to address this question by controlling coprophagia, indicated 
that sucralose elimination is intrinsically slower from the rabbit than 
from other species tested (Refs. 3 and 4). Therefore, the agency 
concludes that the pharmacokinetics of sucralose in the rabbit is 
significantly different from that in humans and other tested species.
    b. Sucralose metabolism. The majority of ingested sucralose is 
excreted unchanged in the feces and most of what is absorbed appears 
unchanged in the urine, with only minor amounts appearing as 
metabolites (Refs. 3, 4, and 5). Mice (El46) and rats (El37) were found 
to metabolize less than 10 percent of the absorbed sucralose, while 
rabbits (El24) (20 to 30 percent), humans (El38 and E145) (20 to 30 
percent), and dogs (El33) (30 to 40 percent) metabolize greater 
quantities of the absorbed sucralose. Results from the submitted animal 
and human pharmacokinetics data identified three major sucralose 
metabolites (Ml, M2, and M3) in urine in addition to unchanged 
sucralose. The metabolic profile of sucralose in rats was qualitatively 
similar to that seen in humans. In addition to unchanged sucralose, two 
sucralose metabolites, Ml and M2, were detected in the urine of rats 
and humans after oral dosing of sucralose. The metabolic profile of 
mice for sucralose differed from that of humans and the other tested 
animals (rats, dogs, and rabbits) in that a unique urinary metabolite, 
M3, was identified in addition to the presence of the Ml (trace 
amounts) and M2 metabolites. A pronounced difference was observed in 
the proportions of M2 and M3 excreted by male versus female mice: Males 
produced more M2 than M3, while the opposite was true of female mice. 
The metabolic profile of the rabbit for sucralose also showed 
differences when compared to that seen in humans, rats, mice, or dogs. 
In addition to unchanged sucralose, a small number of unidentified 
metabolites (more polar than sucralose) were observed in rabbit urine, 
but were not characterized (Refs. 3, 6 and 7). Dogs produced primarily 
the M2 metabolite and only a trace amount of the Ml metabolite.
    After repeated dosing, there was no evidence that sucralose induced 
microsomal enzymes in rats (El44) (Ref. 7). There was also no evidence 
of metabolic adaptation following chronic dosing with sucralose in rats 
(E057e) (Ref. 3).
    Based on the submitted pharmacokinetics data, the agency concludes 
that the rabbit metabolism of sucralose is notably different from that 
of humans in two important aspects: (1) A longer sucralose plasma half-
life, and (2) the presence of unique urinary sucralose metabolites. 
Although pharmacokinetic differences between the other tested animals 
(rats, mice, and dogs) and humans were not as pronounced, the profile 
for rats was most similar to that for humans. The agency discusses the 
relevance of these data for the selection of an appropriate animal 
model in section II.C of this document.
2. Genotoxicity Testing
    Sucralose and its hydrolysis products were tested in several in 
vitro and short-term in vivo genotoxicity tests. In the absence of 
bioassay data, such tests are often used to predict the carcinogenic 
potential of the test compound. However, in the case of sucralose and 
its hydrolysis products, chronic toxicity/carcinogenicity bioassay data 
are also available.
    Sucralose was shown to be nonmutagenic in an Ames test (E0ll) and a 
rat bone marrow cytogenetic test (E013). Tests for the clastogenic 
activity of sucralose in a mouse micronucleus test (E0l4) and a 
chromosomal aberration test in cultured human lymphocytes (E012) were 
inconclusive. Sucralose was weakly mutagenic in a mouse lymphoma 
mutation assay (E014).
    The hydrolysis product, 4-CG, was nonmutagenic in the Ames test 
(E025) and mouse lymphoma assay (E026). 4-CG was nonclastogenic in the 
chromosomal aberration assay (E0I2). Other assays (human lymphocytes 
(E012), rat bone marrow (E027)) were inconclusive. Thus, no test on 4-
CG produced a genotoxic response.
    The other hydrolysis product, 1,6-DCF, was not clastogenic in the 
chromosomal aberration assay in rat bone marrow (E019). Results of 
three other genotoxic tests were inconclusive: The chromosomal 
aberration assay in cultured human lymphocytes (E012), the sex-linked 
recessive lethal assay in Drosophila melanogaster (E021), and the 
covalent DNA binding potential study in rats (El48).  1,6-DCF  was  
weakly  mutagenic  in  the  Ames  test  (E020)  and the  L5178Y  TK+/- 
assay (EO22 and E024). In an unscheduled DNA synthesis study (El65), 
1,6-DCF did not induce DNA repair synthesis in isolated rat 
hepatocytes.

[[Page 16420]]

    An equimolar mixture of the hydrolysis products was not genotoxic 
in the in vivo sister chromatid exchange assay in mice (E150) and was 
inconclusive in a dominant lethal (mouse) test (E034).
    As the foregoing discussion reflects, both sucralose and its 
hydrolysis products showed weakly genotoxic responses in some of the 
genotoxicity tests. More importantly, however, as demonstrated in the 
2-year rodent bioassays (E053, E055, and E057), there was no evidence 
of carcinogenic activity for either sucralose or its hydrolysis 
products as discussed in sections II.B.4.a.i, II.B.4.a.ii, and 
II.B.4.b.i of this document. Results from these chronic carcinogenicity 
studies supersede the results observed in the genotoxicity tests 
because they are more direct and complete tests of carcinogenic 
potential (Refs. 5, 6, 8, 9, and 10).
3. Reproductive/Developmental Toxicity Studies.
    Studies were performed in order to evaluate the toxic potential of 
sucralose and its hydrolysis products on the reproductive systems of 
mature male and female rats as well as on the postnatal maturation of 
reproductive functions of offspring through two successive generations. 
The objective of the teratology studies was to determine the potential 
effects of sucralose and its hydrolysis products on the developing 
fetus.
    a. Sucralose--i. Two-generation reproductive toxicity study in rats 
(E056). In this study, groups of 30 male and 30 female rats of the 
Sprague-Dawley CD strain were fed sucralose at dose levels of 0.3, 1.0, 
and 3.0 percent in the diet 10 weeks prior to breeding and throughout 
two successive generations.
    No treatment-related effects on any of the reproductive endpoints 
(estrous cycles, mating performance, fertility index, gestation length, 
gestation index) were observed in either generation. Litter size and 
offspring viability were also unaffected by sucralose treatment. 
Decreases in body weight gain of 11 to 25 percent and 2 to 12 percent 
for adult rats were observed during both premating periods for the 
first (F1) and second (F2) generations, 
respectively. Slightly decreased food intake was also observed for both 
generations (F0, 5 to 9 percent; F1, 3 to 5 
percent).
    Although significant decreases in the relative thymic weights were 
noted in the F0 (male and female) and the F1 
(male and female) rats in this study after dietary administration of 
sucralose at the high-dose (3 percent) level, because of the nature of 
the experimental design for reproductive studies, the agency cannot 
evaluate the toxicological significance of this observation in this 
study. Thymic and other lymphoidal effects are more appropriately 
evaluated in immunologic studies that are designed to examine directly 
parameters of immunologic functions. Such immunotoxicity studies on 
sucralose are discussed in section II.B.5.b of this document.
    Based upon the results of study E056, the agency concludes that 
sucralose does not cause any reproductive effects in rats in doses up 
to 3 percent in the diet (Refs. 5, 10, 11, and 12).
    ii. Teratology study in rats (E030). Sucralose was administered by 
gavage to groups of 20 pregnant Sprague Dawley CD rats at dose levels 
of 500, 1,000, and 2,000 mg/kg bw/d from day 6 through day 15 of 
gestation.
    No treatment-related effects were noted in the dams at necropsy 
with respect to the number of implantation sites, pre-implantation 
losses, or post-implantation losses. The number of live young, as well 
as fetal and placental weights, were also unaffected by treatment. 
Neither body weight gain nor food consumption were affected by 
treatment with sucralose.
    Based upon the results of E030, the agency concludes that sucralose 
did not cause maternal toxicity, embryo toxicity, or fetal toxicity; 
nor did sucralose induce terata in rats at dose levels up to 2000 mg/kg 
bw/d (Refs. 5 and 13).
    iii. Teratology study in rabbits (El34). Sucralose was administered 
by gavage to groups of 16 to 18 pregnant rabbits at dose levels of 0, 
175, 350, and 700 mg/kg/d during days 6 to 19 of gestation. Uterine 
contents of the females were examined at termination of the study (day 
29 of gestation).
    A total of 11 rabbits (1 in the control group, 4 in the 175 mg/kg 
bw/d group, 2 in the 350 mg/kg bw/d group, and 4 in the 700 mg/kg bw/d 
group) died or were killed in extremis (near death) because of reasons 
unrelated to treatment. Two deaths occurred in the high-dose (700 mg/kg 
bw/d) group that the agency considers treatment-related because they 
were associated with symptoms (weight loss and reduced food intake) 
occurring only at the highest dose. Three of the 12 surviving rabbits 
in the high-dose group were eliminated from the study because they did 
not become pregnant.
    From the remaining nine pregnant rabbits in the high-dose group 
only five animals successfully carried to term and produced viable 
young. The other four females in this group aborted their fetuses. 
Decreases in the mean number of viable young per litter were also 
observed in this group. The mean number of post-implantation losses 
also increased. Gastrointestinal tract disturbances were noted in high-
dose rabbits. These effects observed at the high-dose level were not 
seen at either low- or mid-dose levels (Refs. 5, 14, and 15). While 
maternal and fetal toxicity were observed at the high-dose level, there 
was no evidence of frank terata at any of the tested dose levels. Thus 
this study demonstrates that sucralose is not teratogenic in rabbits.
    b. Sucralose hydrolysis products--i. Two-generation reproductive 
toxicity study in rats (E052). Groups of 30 male and 30 female Sprague-
Dawley CD rats were fed an equimolar mixture of the sucralose 
hydrolysis products (4-CG and 1,6-DCF) at dose levels of 0, 200, 600, 
and 2,000 parts per million (ppm) in the diet for 10 weeks prior to 
breeding and through two successive generations.
    No treatment-related effects on estrus cycles, mating performance, 
fertility, length of gestation, litter size, and offspring viability 
were observed in either generation (F0 or F1 
generation). During the 10-week premating period for both generations, 
body weight gain of males was significantly reduced in the high-dose 
(2,000 ppm) group only. Body weight gain of females was significantly 
reduced in all treatment groups during this same period of time. 
Decreased food intake was observed in the high-dose males and females 
of the F0 generation. In both generations, reduction in 
weight gain was observed in females during pregnancy and in offspring 
from birth to weaning. No effect other than reduced body weight gain 
was related to treatment (Refs. 5, 10, 14, and 16).
    The agency concludes that the administration of the sucralose 
hydrolysis products in the rat diet at levels up to 2,000 ppm caused no 
alteration in the reproductive performance of the animals over two 
generations (Refs. 5 and 16).
    ii. Teratology study in rats (E032). An equimolar mixture of the 
sucralose hydrolysis products was administered by gavage to groups of 
20 pregnant Sprague-Dawley rats at dose levels of 30, 90, and 270 mg/kg 
bw/d, from day 6 to 15 of gestation. The study was terminated on day 21 
of gestation.
    Results from this study showed no dose-related increase in the 
incidence of terata among treated groups. Body weight gain of dams in 
the high-dose group (270 mg/kg bw/d) was significantly reduced, whereas 
weight gains in the low- and mid-dose dams were comparable to controls. 
Decreased

[[Page 16421]]

fetal body weights and placental weights were observed at the high 
dose.
    The agency concludes that the sucralose hydrolysis products did not 
produce terata in rats when administered at doses up to 270 mg/kg bw/d 
(Refs. 10 and 13).
    c. Male fertility studies on sucralose and its hydrolysis products 
in rats (E016, E038, E090, and E107). Some chlorinated monosaccharides 
have been reported to affect male fertility in rats by interfering with 
spermatogenesis (Ref. 17). McNeil noted the structural similarity of 
such compounds to the hydrolysis products of sucralose, and submitted a 
series of antifertility studies on a series of chlorinated sugars, 
including sucralose.
    All of the studies were of similar design. Groups of male rats were 
exposed for 14 days either by gavage or in the diet to 300 micromoles 
(mol) of either sucralose or one of the chlorosucrose 
compounds mentioned above. The antifertility compound, 6-chloro-6-
deoxyglucose, was used as the positive control in these studies. 
Treated male and untreated female rats were mated 1 and 2 weeks after 
treatment. Male mating performance and fertility were observed.
    The agency has reviewed these studies and observes that the studies 
were too short to cover the full cycle of spermatogenesis in rats 
(Refs. 5 and 18). Because of their short duration, FDA concludes that 
these studies, considered alone, are insufficient to assess the 
antifertility potential of sucralose in male rats (Refs. 5 and 18). 
However, the agency believes that further testing is not necessary 
because the results from the two-generation reproduction studies 
adequately address any toxicological concerns regarding the potential 
antifertility effects of sucralose and its hydrolysis products. As 
discussed previously, in the two-generation reproduction studies (E052 
and E056), in which sucralose or its hydrolysis products were fed to 
rats, no effects on fertility or other reproductive parameters were 
observed in either male or female rats (see sections II.B.3.a.i and 
II.B.3.b.i. of this document).
4. Chronic Toxicity/Carcinogenicity Studies
    A combined chronic toxicity/carcinogenicity study (E057) in rats 
and a carcinogenicity study in mice (E055) were conducted to study the 
chronic toxicity and carcinogenic potential of sucralose when 
administered to rodents over most of their lifetime. Because human 
exposure to sucralose could possibly occur during in utero development, 
an in utero phase was included in the rat study. A chronic (1-year) 
study on sucralose was also performed in dogs (E051) in order to assess 
the effects of sucralose administration in a nonrodent species. In 
addition, a 2-year carcinogenicity study in rats (E053) was carried out 
to study the chronic toxicity and carcinogenic potential of sucralose 
hydrolysis products.
    a. Sucralose--i. Combined chronic toxicity/carcinogenicity study in 
rats (E057). This study consisted of a breeding phase, a 
carcinogenicity phase, and a chronic toxicity phase. The 
carcinogenicity and chronic toxicity phases were concurrently performed 
in this study. The breeding phase of this study examined the potential 
in utero effects of sucralose during development. During this phase 
parental (F0) Sprague-Dawley CD rats, 70 males and 70 
females per group, were fed diets containing 0, 0.3, 1, or 3 percent 
sucralose for a 4-week period prior to mating and during gestation. One 
male and one female weanling pup were selected from each of 50 litters 
and allocated to the appropriate group of the carcinogenicity phase. 
Additional rats (30 per sex per group) were selected for the chronic 
toxicity phase of this study.
    Rats in each of the groups of this study were gang-housed, five 
animals per sex per cage. After 52 weeks of sucralose treatment, an 
interim sacrifice was performed on 15 males and 15 females from each 
group of the chronic toxicity phase of the study. The remaining 
surviving rats in this phase of the study were sacrificed at treatment 
week 78. In the carcinogenicity phase, surviving rats were sacrificed 
at week 104. In both phases of the study, classic toxicological 
parameters such as mortality, body weight, hematology, clinical 
chemistry, and organ weights were examined in treated and control rats. 
Food consumption was calculated weekly from the total weight of food 
consumed by each cage of rats. Histopathological examinations were 
performed on representative tissues from control and high-dose rats.
    Sucralose treatment had no effect on reproductive performance or on 
fertility of the parental rats during the breeding phase. In both the 
chronic toxicity and carcinogenicity phases of the study, survival of 
rats was unaffected by sucralose treatment.
    In the carcinogenicity phase, there was no evidence of treatment-
related neoplasia in any of the rats (Ref. 19). McNeil reported an 
apparent increased incidence of male rats with hepatocellular clear 
cell foci. FDA pathologists reviewed the liver histopathology slides 
from this study that were obtained from McNeil. The agency's 
pathologists observed that the increase in the incidence of male rats 
with hepatocellular clear cell foci was only marginal and that there 
was no concomitant increase in the severity of this lesion among the 
treated animals. Therefore, the agency concludes that the occurrence of 
hepatocellular clear cell foci was incidental and not treatment-related 
(Refs. 5 and 20).
    Renal pelvic mineralization and epithelial hyperplasia were noted 
at higher incidences among treated rats in both the chronic toxicity 
and the carcinogenicity phases of study E057. These changes were 
observed primarily in the high-dose females. The degree of severity of 
these lesions was reported as minimal or slight. McNeil concluded that 
these changes are of no toxicological significance.
    FDA evaluated these changes and noted that: (1) It is not unusual 
to observe such lesions in aged rats, especially in females (Ref. 21). 
In this study (E057), the rats were at or near the end of their 
expected lifetime at the time of sacrifice; and (2) mineralization of 
the renal pelvis represents a physiological adaptation secondary to 
cecal enlargement. Cecal enlargement is often seen with other 
substances that are poorly absorbed in the upper intestine and can be 
expected in a study like this with a poorly absorbed substance like 
sucralose (Refs. 21, 22, 23, 25, and 26). Based on the previously 
mentioned reasons, FDA concludes that the renal pelvic mineralization 
and epithelial hyperplasia observed are of no toxicological 
significance (Refs. 6 and 26).
    Decreased body weight gain was observed in all sucralose treated 
animals in both the carcinogenicity and chronic toxicity phases of this 
study. At the end of the carcinogenicity phase, mean body weight gain 
in sucralose-fed rats was 13 to 26 percent less than that of the 
control group. Food consumption in the treated groups during this phase 
was 5 to 11 percent less than that of the control values. At the end of 
the chronic toxicity phase, a reduction of 12 to 25 percent in the body 
weight gain was observed in the treated rats relative to controls, 
whereas food intake in the treated rats was reduced only 5 to 10 
percent compared to controls.
    McNeil postulated that this body weight gain decrement effect was 
the result of reduced palatability of sucralose-containing diets. 
However, based on the data in this study, as well as in all other rat 
studies in the sucralose petition, the agency was unable to conclude 
that reduced palatability, which affected food

[[Page 16422]]

consumption, fully accounted for the decreased body weight gain 
observed in sucralose-fed rats (Ref. 27). Thus, the agency recommended 
that McNeil perform additional testing to resolve the body weight gain 
issue (Ref. 28). In the absence of such testing, FDA could not 
determine a no-observed-effect level for this study (E057). The body 
weight gain issue is discussed in detail in section II.B.5.a of this 
document.
    ii. Carcinogenicity study in mice (E055). In this study, Charles 
River CD-1 mice, 52 animals per sex per group, were gang-caged (4 mice 
per cage) and fed sucralose at 0, 0.3, 1.0, and 3.0 percent in the diet 
for 104 weeks. At the termination of the study, survival and classic 
toxicological parameters were examined for treated and control mice.
    Survival rates were comparable for control and treated groups. Mean 
body weight gains in both male and female mice in the high dose (3 
percent) group were significantly reduced (21 to 25 percent) relative 
to controls for the 104-week treatment period, without any significant 
decreases in food consumption. Of other toxicological parameters 
examined, significant decreases were observed only in the erythrocyte 
counts of females in the high-dose group. There was no evidence of 
treatment-related neoplasia in any of the sucralose-treated groups 
(Ref. 19).
    Based on the effects seen on body weight gain and the erythrocytic 
counts at the high-dose level, the agency concludes that a dietary 
level of 1 percent (equivalent to 1,500 mg/kg bw/d) was the no-
observed-effect level for sucralose (Refs. 5 and 29).
    iii. Chronic toxicity study in dogs (E051). Groups of four male and 
four female beagle dogs were fed sucralose at concentrations of 0, 0.3, 
1.0, and 3.0 percent in the diet for 52 weeks. Parameters examined in 
this study included mortality, body weight, food consumption, 
hematology, clinical chemistry, urinalysis, and histopathology.
    An increase in body weight gain of sucralose-treated male dogs 
relative to controls was observed at all dose levels. However, this 
increase in weight gain was accompanied by a general increase in food 
consumption. All other parameters examined in this study were 
comparable between treated and control animals.
    Because there were no toxic effects seen at any dose tested, the 
agency concludes that a dietary level of 3 percent (equivalent to 750 
mg/kg bw/d) is the no-observed-effect level for sucralose in dogs 
(Refs. 5 and 30).
    b. Sucralose hydrolysis products--carcinogenicity study in rats 
(E053). In this study, groups of 50 male and 50 female Sprague-Dawley 
CD rats were administered an equimolar mixture of the hydrolysis 
products (4-CG and 1,6-DCF) at concentrations of 0, 200, 600, and 2,000 
ppm in the diet for 104 weeks.
    There was no evidence of treatment-related neoplasia in any of the 
dose groups in this study. A marginal increase in the incidence of 
hepatocellular clear cell foci was reported in treated male and female 
rats. The agency determined, however, that this was not a treatment-
related effect because there was no concomitant increase in severity of 
the hepatic lesion (Refs. 19 and 20). Thus, the agency concludes that 
the sucralose hydrolysis products are not carcinogenic to Sprague-
Dawley CD rats when administered as an equimolar mixture in the diet at 
concentrations up to 2,000 ppm (Refs. 5, 19, and 31).
    In this study, the mean body weight gain of the high-dose females 
was significantly decreased (24 percent) relative to the control mean 
after 104 weeks of treatment. Mean food consumption in these females 
over the 104-week period was also reduced 14 percent compared to the 
control group. The agency could not determine whether the body weight 
gain decrement observed at the high-dose level in this study was fully 
accounted for by decreased food intake. Therefore, the agency concludes 
that, in rats, the mid-dose (600 ppm equivalent to 30 mg/kg bw/d) is 
the no-observed-effect level for the hydrolysis products of sucralose 
(Refs. 5 and 10).
5. Special Toxicological Studies
    a. Body weight gain. As noted previously, the agency's review of 
the rat data submitted in the original petition raised questions 
regarding the effect of sucralose on body weight gain (Ref. 27). 
Sucralose-fed rats in the subchronic and chronic studies showed 
significant decreases in body weight gain with only small reductions in 
food consumption (Ref. 27).
    In particular, in the combined chronic toxicity/carcinogenicity rat 
study (E057), decreases of 13 to 26 percent in body weight gain were 
observed in sucralose-fed rats that had reductions in food consumption 
of only 5 to 11 percent compared to controls (Ref. 27). Although the 
treated rats ate less food, the reductions in food intake did not 
appear to account fully for the decreased weight gain. McNeil contended 
primarily that reduced palatability of the sucralose-containing diet 
caused treated animals to eat less and thus gain less weight. McNeil 
stated that, collectively, data obtained from the sucralose 
acceptability study (El30 and E143), sucralose pair-feeding study 
(E058), gavage study (El5l), and a diet spillage study (El54) supported 
their claim that palatability fully accounted for the reduced body 
weight gain (Ref. 32). Finally, McNeil also contended that this effect 
was neither a toxic effect nor biologically significant. The studies 
upon which McNeil relied are discussed followed by the agency's 
discussion of its evaluation of those studies.
    i. The Palatability hypothesis--(1) Acceptability studies in rats 
(El30 and E143). Several studies were conducted to evaluate the 
acceptability and palatability of sucralose when administered to rats 
via drinking water or in the diet. Data from these rat studies showed 
that sucralose was acceptable in drinking water at levels up to 3,200 
ppm. However, reduced food consumption was seen in rats that were 
administered sucralose in the diet at levels greater than 800 ppm.
    (2) Pair-feeding study in rats (E058). Pair-feeding is an 
experimental procedure where two groups of animals are fed the same 
amount of diet. Thus, if there are differences in the body weight gain 
of these two groups of animals, it is due to an effect of the test 
substance and not due to differences in the amount of food consumed by 
the two groups of animals.
    There were five groups of female Sprague-Dawley CD rats in this 
study. Initially, rats were grouped into various categories on the 
bases of body weight. Twenty rats were randomly selected from each of 
the weight categories and assigned to each of the five groups. One 
group was fed 3 percent sucralose in the diet (unrestricted access) for 
8 weeks. Animals in the pair-fed group were fed a daily amount of basal 
diet equivalent to the mean food intake consumed on the previous day by 
the 3-percent sucralose dose group. In a third group, an ad libitum 
control group, rats received unrestricted access to basal diet. A 
fourth group was administered sucralose by gavage in amounts equivalent 
to that fed in the 3-percent dietary group. A fifth group served as a 
control group for the sucralose-gavaged rats and received distilled 
water by gavage.
    Significant decreases in food consumption and body weight gain were 
observed in both the 3-percent dietary administration group and its 
pair-fed control group relative to ad libitum controls. Rats dosed with 
sucralose by gavage consumed significantly more food and gained 
significantly more weight than those receiving the water control.

[[Page 16423]]

    (3) 4- to 13-week sucralose oral gavage study in rats (El5l). 
Because administration by gavage circumvents effects due to dietary 
administration of an unpalatable test material, McNeil performed a 
study to investigate the effects of sucralose in rats, when 
administered by gavage. In this study, groups of Sprague-Dawley rats, 
10 per sex per group, were administered sucralose at doses of 2,000 mg/
kg bw/d for 13 weeks, 3,000 mg/kg bw/d for 9 weeks, or 4,000 mg/kg bw/d 
for 4 weeks. Control rats (10 to 15 per sex) were sacrificed 
concurrently at each of the time intervals along with the sucralose-
treated rats.
    There were no treatment-related gross or histopathological changes 
observed nor effects noted for urine and clinical chemistry parameters. 
The average food consumption for all sucralose dosed rats was 
consistently greater than that of the controls (104 to 108 percent of 
the controls). Mean final body weights were also greater in the 
sucralose treated rats compared to controls (103 to 109 percent).
    (4) Diet spillage study in rats (El54). McNeil performed a study to 
determine whether the decreased body weight gain observed in several of 
the rat studies, including the combined chronic toxicity/
carcinogenicity study, was due, in part, to increased spillage of 
sucralose-containing diet. If there was greater spillage of the 
sucralose-containing diet than that seen in controls, then the 
sucralose-treated animals were eating even less than they appeared to 
consume. In this 8-week study, three groups of Sprague-Dawley rats (15 
per sex per group) were individually housed and fed either basal diet 
or basal diet containing sucralose at dose levels of 3 percent or 5 
percent. Although overall diet spillage was significantly higher in the 
sucralose-treated rats compared to controls, this difference existed 
only for the first 2 weeks. Treated rats (both sexes) consumed 5 to 8 
percent less food than controls. This decreased food intake was 
associated with a 10 to 15 percent depression in weight gain.
    ii. The agency's evaluation of the palatability hypothesis. From 
its interpretation of the data in the acceptability studies (EI30 and 
E143), pair-feeding study (E058), gavage study (El5l), and diet 
spillage study (El54), McNeil identified three factors that the company 
believed led to the decrement in body weight gain observed in the 
combined chronic toxicity/carcinogenicity study in rats (E057): (1) 
Decreased food consumption due to poor palatability and increased 
spillage of the sucralose-containing diet; (2) inhibition of growth 
potential in sucralose-fed F1 generation rats due to 
decreased initial body weight resulting from decreased maternal weights 
of the treated rats; and (3) magnification of the body weight gain 
effect with increases in study duration.
    While the agency accepted the physiological and nutritional 
principles presented by McNeil, the agency concluded that McNeil's 
arguments did not explain fully the magnitude of the decrement in body 
weight gain in the sucralose-fed rats of the combined chronic toxicity/
carcinogenicity study (E057) for the following reasons.
    The agency disagreed with the petitioner's contention that in the 
combined chronic toxicity/carcinogenicity study (E057), a consistent 
decrease in food consumption was demonstrated at all dose levels. The 
agency determined that this study (E057) did not adequately measure 
food consumption and did not adequately account for diet spillage. 
Furthermore, the agency determined that in many of the sucralose rat 
studies food consumption decreases were not of sufficient magnitude to 
account for the observed body weight gain decrements seen in the 
sucralose-fed rats of these studies (Ref. 27). Inadequacies in the 
measuring of food consumption and the monitoring of spilled diets also 
confounded the interpretation of the pair-feeding study (E058) (Refs. 
10 and 27).
    The agency also disagreed that decreased initial body weights 
accounted for the weight gain decrement in sucralose treated rats in 
study E057. Although maternal weights were slightly decreased (93 to 97 
percent of controls) on day 1 of lactation, this small difference was 
not large enough to sufficiently explain the body weight differences of 
the lactating pups (Ref. 27). In fact, maternal weights of the 
sucralose-fed rats were not significantly different from those of the 
control rats during days 14 to 21 of lactation (Ref. 27). Differences 
in initial body weights of the F1 pups (4 to 8 percent 
decreases) of the combined chronic/carcinogenicity study (E057) were 
not sufficient to explain the magnitude of the final body weight gain 
decrements of these rats (Ref. 27).
    Finally, although FDA agreed with the general principle that long-
term food intake disparity will result in increasing differences in 
body weight gain over time, FDA concluded that this principle alone did 
not account for the degree of magnification of body weight gain 
decrement compared to the small reductions in food consumption seen in 
the sucralose studies (Ref. 27).
    Based on the foregoing reasoning, FDA concluded that the 
acceptability studies (El30 and E143), pair-feeding study (E058), 4- to 
13-week gavage study (El5l), and the diet spillage study (El54) did not 
adequately explain the magnitude of decreased body weight gain relative 
to the level of reduced food consumption, in the combined chronic/
carcinogenicity study (E057). The agency thus concluded that McNeil had 
failed to explain satisfactorily the observed body weight gain 
decrement and that additional study data were needed to resolve this 
issue (Ref. 28). McNeil subsequently conducted two studies (E160 and 
E161) in rats to resolve the body weight gain decrement issue.
    iii. Resolution of the body weight gain decrement issue--(1) 
Sucralose dietary administration and dietary restriction study in rats 
(El60). McNeil agreed to perform an additional sucralose feeding study 
(the diet restriction study, E160) to attempt to resolve the body 
weight gain decrement issue and to test the petitioner's palatability 
hypothesis. The specific purpose of the study was twofold: To determine 
whether the weight gain decrement observed in the sucralose-fed rats of 
the combined chronic toxicity/carcinogenicity study (E057) could be 
explained solely by decreased food consumption; and to establish a 
``no-observed-effect'' level for the body weight gain decrement effect 
after chronic administration of sucralose.
    In study E160, Sprague-Dawley CD rats were divided into eight 
groups (20 animals per sex per group). Three groups were fed ad libitum 
basal diet that contained 0, 1, or 3 percent sucralose. Three groups 
were fed restricted amounts of basal diet at levels that were 85, 90, 
or 95 percent of that eaten by the ad libitum controls. Two other 
groups were fed restricted diets (90 percent of ad libitum controls) 
that also contained sucralose at a concentration of 1 percent or 3 
percent. The groups were as follows:

     Group 1 Control--basal diet ad libitum
     Group 2 Control--basal diet 95 percent of Group 1
     Group 3 Control--basal diet 90 percent of Group I
     Group 4 Control--basal diet 85 percent of Group 1
     Group 5 1-percent sucralose--ad libitum
     Group 6 3-percent sucralose--ad libitum
     Group 7 1-percent sucralose--90 percent of Group 1
     Group 8 3-percent sucralose--90 percent of Group I

[[Page 16424]]

    Special experimental designs, including single-housing of the test 
animals, accurate weighing of spilled diet, and utilization of special 
feed jars, were incorporated into this study to ensure the highest 
level of accuracy in the measuring and reporting of food intake. Body 
weight, body weight gain, food consumption, and food conversion 
efficiency data were collected for each of the groups. Overall survival 
was unaffected by the feeding of sucralose at doses up to 3 percent for 
the duration of the study. The agency evaluated the data from this 
study using two separate statistical procedures. In the first 
comparison, data from control groups 1 to 4 were combined and fitted 
(separately for males and females) with a polynomial regression model 
that showed final body weight gain as a function of initial body weight 
and food consumption. Data for each of the sucralose groups were also 
fitted with this mathematical model and compared to the data from the 
combined control groups.
    In the second comparison, mean food consumption was calculated for 
each sucralose group. Using the regression models, FDA calculated the 
expected body weight gain for animals at the mean food consumption for 
both the combined control groups and the sucralose groups. The 
calculated body weight for each sucralose group was then compared to 
the combined control group at the mean food consumption.
    For both sexes, with both statistical procedures, the 3-percent 
sucralose groups (Groups 6 and 8) showed significant decrements in body 
weight gain relative to the combined control groups (Ref. 33). 
Decrements of 3.9 to 6.3 percent were observed in the mean body weights 
of the 3-percent sucralose-fed groups after adjustment for food 
consumption and initial body weight differences. Thus food consumption 
only partially accounted for the weight gain decrement observed in the 
3-percent sucralose-fed rats. Weight decrements in the males of the 3-
percent dose group stabilized by 15 weeks; in the females, differences 
stabilized at 20 weeks. Therefore, FDA concludes that the duration of 
this study (26 weeks) was sufficient to evaluate weight gain decrement 
effects.
    In both the 1-percent sucralose group and the 1-percent sucralose 
with l0-percent diet restriction group, adjusted mean body weights were 
comparable to those of the combined control data (Ref. 33). Therefore, 
FDA determined that reduced food consumption accounted fully for weight 
gain differences in the 1-percent sucralose-fed group.
    Based upon the data from this study, the agency concludes that 
treatment with sucralose at 1 percent in the diet had no effect on body 
weight gain in rats. The same data establish that rats fed sucralose at 
a concentration of 3 percent of the diet did show significant decreases 
in weight gain which were attributable to the test substance. The 
agency further concludes that, based upon this study, the 1-percent 
dose level (equivalent to the 500 mg/kg bw/d dose in study E057) is the 
no-observed-effect level for the body weight gain effect observed in 
sucralose-treated rats in this study (Ref. 34).
    (2) Sucralose toxicity study by oral (gavage) administration to 
Sprague-Dawley CD rats for 26-weeks (El6l). McNeil submitted a 26-week 
gavage study (El6l) in rats that was designed to: (1) Provide further 
support for their contention that the body weight gain decrement seen 
in sucralose fed rats could be explained solely by decreased food 
intake caused by the reduced palatability of sucralose-containing diet; 
(2) confirm the data in the 4- to 13-week sucralose oral gavage study 
(EI51); and (3) to address inadequacies in the experimental design of 
the 4- to 13-week sucralose oral gavage study (El5l).
    In this 26-week study, sucralose was administered orally to 
Sprague-Dawley CD rats, 20 rats per sex per group, by gavage at dosages 
of 0, 750, 1,500, or 3,000 mg/kg bw/d. Rats in the control group were 
gavaged with purified water. Body weight, water consumption, and food 
consumption data were recorded for all groups. Routine hematological 
and clinical chemistry parameters were measured. Organ weight data also 
were recorded. Histopathological examinations were performed on 
representative vital tissues from the control and high-dose groups. 
Histopathological examinations were performed also on all abnormal 
tissues.
    Seven deaths occurred during the study that were attributed either 
to spontaneous causes not related to treatment or technical trauma 
during dosing: 2 males, 0 mg/kg bw/d dose; 1 male and 2 females, 1,500 
mg/kg bw/d dose; and 1 male and 1 female, 3,000 mg/kg bw/d dose. 
Overall body weights of the animals in the sucralose-treated groups 
were not significantly different from those of the control group during 
the length of the study. The mean food consumption in the sucralose-
gavaged rats was similar to that seen in the controls, except in the 
high-dose males. Food intake for the high-dose males was 3.9 percent 
greater than that of the control rats.
    After making adjustments for initial body weight and food 
consumption, the agency performed a statistical analysis on the final 
body weight data using polynomial regression analysis. This analysis 
showed that the adjusted final body weight of the high-dose males was 
significantly decreased (4.6 percent; p = 0.035) relative to that of 
the control group. The adjusted mean body weights of all other groups 
were not significantly different from the controls.
    Water consumption was significantly increased in the sucralose-
treated rats relative to controls. There were no treatment-related 
effects seen in any of the hematological or clinical chemistry 
parameters tested. Cecal enlargement was the only effect of sucralose 
that was dose-related among both sexes of the sucralose-gavaged rats. 
As discussed previously in section II.B.4.i of this document, this 
effect is a normal physiological adaptation to poorly absorbed dietary 
components and not related to toxicity. The relative kidney weight of 
the high-dose group also was significantly increased when compared to 
the control group. However, this kidney effect was not associated with 
any toxicologically significant renal histopathology. Additionally, the 
plasma electrolytes of the sucralose-treated rats in this study were 
comparable to that seen in control animals.
    As with the diet restriction study (El60), decreased body weight 
gain was observed in the sucralose-treated rats of the high-dose group. 
The agency concludes that the mid-dose (1,500 mg/kg bw/d) is the no-
observed-effect level for the body weight gain effect observed in this 
study (El6l) (Refs. 35 and 36).
    b. Immunotoxicity study in rats. As reported by McNeil and as noted 
in the agency's review of the sucralose data, thymus, spleen, and 
hematological changes were observed in rats at the high-dose levels in 
some of the short-term and long-term sucralose feeding studies. For 
example, when rats were fed sucralose in a 4- to 8-week range-finding 
study (E031) the following effects were noted: Decreased thymus and 
spleen weights, lymphocytopenia, and cortical hypoplasia of the spleen 
and thymus. In the two-generation reproductive toxicity study (E056), 
decreased thymus weights were noted in the F0 and 
F1 generations of the high-dose sucralose (3 percent in the 
diet) group. McNeil stated that the above effects were secondary to the 
palatability-related reduction in food consumption in treated rats.
    In an effort to provide more specific and detailed assessment of 
the immunotoxic potential of sucralose, the petitioner conducted a 28-
day oral immunotoxicity study (El62) of

[[Page 16425]]

sucralose in rats. In this study, groups of male and female Sprague-
Dawley rats (13 per sex per group) were administered sucralose by 
gavage at dose levels of 750, 1,500, and 3,000 mg/kg bw/d for 28 days. 
Additional groups (13 per sex per group) of rats formed a gavage 
control group, an ad libitum diet control group, a dietary sucralose 
(3,000 mg/kg bw/d) group, and a diet restricted (90 percent of ad 
libitum control) group.
    Immunotoxicological parameters examined in this study were: Thymus 
and spleen weights at study termination; standard histopathology 
evaluation of the spleen, thymus, bone marrow, and lymph nodes; and 
total and differential white blood cell counts. The study also examined 
the following specific immunologic parameters: Bone marrow cellularity, 
immunoglobulin subtypes, splenic lymphocyte subsets, and splenic 
natural killer cell activity.
    Significant decreases were observed in the mean thymus weight of 
the males in the high dose (3,000 mg/kg bw/d) gavage group. Thymus 
weight was not significantly affected by sucralose when administered to 
rats by gavage at either 1,500 or 750 mg/kg bw/d; nor was it affected 
in the sucralose-fed group or the diet restricted group. No 
morphological changes in thymus or any other lymphoid tissues were 
observed in any of the sucralose treated groups.
    In the mid-dose (1,500 mg/kg bw/d) sucralose-gavaged male rats, 
there appeared to be a trend toward decreasing white blood cell and 
lymphocyte counts with increasing dose levels of sucralose, but the 
trend did not reach statistical significance. No significant 
differences were seen in other immunologic parameters in the sucralose 
gavage groups relative to the control gavage group. However, because of 
the large variation seen in the data from the gavaged animals at the 
mid-dose, the agency finds that the study is inconclusive regarding 
treatment-related effects for these parameters at the mid-dose.
    The agency concludes that the highest dose (3,000 mg/kg bw/d) 
tested in the gavage groups showed an effect based on the significant 
changes in thymus weight. Because of the difficulty in interpreting 
data from the mid-dose animals, the agency has determined that the low 
dose, 750 mg/kg bw/d, is the no-observed-effect level for the 
immunological endpoints examined in this study (Ref. 37).
    c. Neurotoxicity testing in mice and monkeys (E008 and E009). The 
chlorinated monosaccharide, 6-chloro-6-deoxy-D-glucose (6-CG), is known 
to be neurotoxic to laboratory animals (Refs. 38 and 39). Because 
sucralose is a chlorinated disaccharide, McNeil conducted two 
neurotoxicity studies, one in mice (E008) and one in monkeys (E009). 
The positive control in these studies, 6-CG, produced strong clinical 
signs of neurotoxicity, as well as severe morphological changes in the 
tissues of the central nervous system (CNS). Animals receiving 
sucralose or an equimolar mixture of sucralose hydrolysis products at 
doses up to 1,000 mg/kg bw/d did not exhibit any clinical signs of 
neurotoxicity or morphological changes in CNS tissues (Refs. 5 and 40). 
The agency concludes that the lack of neurotoxic effects by both 
sucralose and its hydrolysis products at the tested dose levels in 
these studies provides assurance that sucralose used as a food additive 
under the proposed conditions of use will not produce neurotoxic 
effects.
    d. Diabetic studies in humans (EI56, E157, E168, E170, E171). In an 
effort to provide an assessment of any potential effect sucralose use 
would have on the diabetic population, the petitioner performed a 
series of clinical studies on diabetic patients. The results obtained 
from those studies are discussed in this section of this document.
    A single-dose cross-over study (E156) was performed in 13 insulin-
dependent (IDDM or Type I diabetics) and 13 non-insulin dependent 
(NIDDM or Type II diabetics) patients to evaluate the effects of a 
single dose of sucralose (1,000 mg) on short-term glucose homeostasis. 
Fasting plasma glucose area under the curve (AUC) and fasting serum C-
peptide AUC were measured after the consumption of a standardized 
liquid breakfast meal. This study showed that neither plasma glucose 
nor serum C-peptide levels were affected by this single dose 
administration of sucralose in these patients. From this study the 
agency concludes that sucralose does not adversely affect short-term 
glycemic control in persons with diabetes mellitus (Ref. 41 ).
    A 6-month clinical study (E157) was performed investigating the 
effect of sucralose (667 mg/d through oral administration) on glucose 
homeostasis in patients with NIDDM (Type II diabetes). The study was 
divided into a screening phase, a testing phase, and a followup phase. 
Forty-one patients participated in the testing phase of the study. The 
41 patients were divided into two groups: 20 patients whose diabetes 
was managed by insulin and 21 managed by oral hypoglycemic agents 
(OHA's). Each of these two groups were further subdivided into a 
sucralose group and a placebo group. Percent concentration of 
glycosylated hemoglobin (HbA1c) was the primary measure of long-term 
glycemic control in this study. In addition, the following parameters 
of glucose homeostasis were measured: (1) Fasting levels of plasma 
glucose, serum C-peptide, and serum insulin; and (2) postprandial 
measures of plasma glucose, serum C-peptide, and serum insulin. These 
parameters were measured after 0, 1, 3, and 6 months of treatment with 
either sucralose or a placebo (cellulose).
    The results from this study showed a small but statistically 
significant increase in the glycosylation of hemoglobin (HbA1c) from 
baseline levels in the sucralose-treated group compared to that seen in 
the placebo group (dataset 1: mean difference of 0.007 percent, p = 
0.005; dataset 2: mean difference of 0.006 percent, p = 0.012) (Ref. 
42). This HbA1c effect was observed in the sucralose-treated group at 1 
month of treatment and did not significantly increase to higher levels 
throughout the remainder of the study (mean difference range of 0.006 
to 0.008 percent, p 0.0043). Overall, during the test phase 
of the study, no statistically significant changes from baseline were 
observed in any of the secondary measurements of glucose homeostasis 
(ie., plasma glucose and serum C-peptide and insulin concentrations). 
Because of the small patient group sizes in this study, the ultimate 
clinical significance of the observed HbA1c effect could not be 
determined (Ref. 42). However, generally speaking, increases in 
glycosylation in hemoglobin imply lessening of control of diabetes. 
Thus, the petitioner performed studies E168 and E170 in an attempt to 
provide an explanation for the observed HbA1c effect.
    In study E168 McNeil performed a series of tests to determine 
whether the increased HbA1c levels observed in study E157 were an 
artifact of measurement (e.g. interferences related to methodology) or 
a direct effect of sucralose on the rate of hemoglobin glycation. These 
tests included a reanalysis of blood samples from study E157 for 
glycohemoglobin levels; an investigation of the procedures used to 
measure glycated hemoglobin; and an analysis of the effects of 
sucralose on glycation of hemoglobin in hemolysates versus intact 
erythrocytes. Results from these tests confirmed that in E157, HbA1c 
levels were increased in the sucralose-treated diabetic patients and 
showed that sucralose had no direct effect on the rate of hemoglobin 
glycation.
    In study E170, red cell preparations from the blood of diabetic and 
non-diabetic patients were treated with

[[Page 16426]]

sucralose (100 mg per liter) to investigate the rate of formation of 
glycated hemoglobin in the blood. The results of this study showed that 
sucralose did not affect the rate of formation of glycated hemoglobin 
(Ref. 42). Thus, there was no evidence that a physicochemical or other 
influence by sucralose might explain the increased glycation of 
hemoglobin.
    Because studies E168 and E170 did not provide an explanation for 
the HbA1c effect observed in study E157, study E171 was performed as a 
repeat study of E157 with a better experimental design, in that E171 
had larger patient group sizes and stronger statistical power (90 
percent versus 80 percent in study E157) to detect an effect by 
sucralose on hemoglobin glycation. The 3-month duration for study E171 
was deemed adequate because the increased HbA1c levels that were seen 
at one month of treatment in study E157 did not increase any further at 
any of the later time points tested in the study. In study E171, 136 
NIDDM patients were divided into two groups based on their diabetic 
therapy (64 taking insulin and 72 on OHA's). Each of these two groups 
were subdivided equally into a sucralose and placebo group. The study 
was divided into a screening phase, a testing phase, and a followup 
phase. Glycosylated hemoglobin (HbA1c) was the primary measure of 
glucose homeostasis; in addition, the secondary parameters, fasting 
plasma glucose and serum C-peptide, were measured. Serum insulin levels 
were not measured in this study.
    Results from study E171 showed no statistically significant changes 
from baseline in the HbA1c levels or any of the other measured 
parameters of glucose homeostasis in the sucralose-treated groups 
relative to the placebo control group. The agency concludes from the 
results of this study that sucralose (667 mg/d) has no effect on long-
term glucose homeostasis (as measured by HbA1c) in patients with NIDDM 
(Refs. 43 and 44). The agency further concludes that the small but 
statistically significant decline in glycemic control that was observed 
in the sucralose-treated groups in study E157 was not a clinically 
significant effect because this effect was not duplicated in a repeat 
study (study E171) that had a greater statistical power (Ref. 43).
    Therefore, based upon the clinical studies of sucralose, FDA 
concludes that sucralose does not adversely affect glucose homeostasis 
in patients with diabetes mellitus.

C. Acceptable Daily Intake Estimates for Sucralose

     Based on a comprehensive review of the sucralose data base, the 
agency has selected the rat as the most appropriate experimental model 
to establish a safe level of sucralose for human ingestion. This 
selection was based on the following considerations:
    (1) The pharmacokinetics data show that the sucralose metabolite 
profile in rats was qualitatively comparable to that in humans.
    (2) In the combined chronic toxicity/carcinogenicity rat study 
(E057) with sucralose, the animals were exposed in utero, which 
maximizes the toxicological testing sensitivity.
    (3) The combined chronic toxicity/carcinogenicity rat studies 
(E057) and the carcinogenicity study in rats (E053) were designed to 
test the toxic potential of sucralose and its hydrolysis products for a 
duration approximating the lifespan of the species. The agency 
historically uses life-time studies for safety evaluation of this type 
of food additive. Such testing effectively allows for the assessment of 
chronic toxicity including the carcinogenic potential of sucralose.
    (4) The majority of the sucralose toxicological data base consists 
of rat studies, thereby allowing a more comprehensive safety evaluation 
of sucralose in that species. For these reasons, the agency concludes 
that the combined chronic toxicity/carcinogenicity study (E057) in 
rats, interpreted in light of the no-observed-effect level established 
in other studies (El60, E161, and E162), provides the most appropriate 
basis for establishing the ADI for sucralose (Refs. 4 and 10). Data in 
study E057 showed that sucralose was not carcinogenic to rats at 
concentrations up to 3 percent (1,500 mg/kg bw/d). No toxicologically 
significant changes in hematology, clinical chemistry, organ weights, 
or urinalysis were observed in the sucralose-treated rats in this 
study. Macroscopic and microscopic examinations of the tissues from 
these sucralose-treated rats revealed no significant treatment-related 
toxicological effects.
    The only treatment-related effect seen in the sucralose-fed rats of 
this study was decreased body weight gain at the 3-percent dose level. 
The relationship of this effect to treatment at the 3-percent dose 
level was corroborated by the diet restriction study (El60). In the 
diet restriction study (El60), the 1-percent dose level (equivalent to 
500 mg/kg bw/d dose in study E057) was established as the no-observed-
effect level of sucralose for the observed body weight gain decrement 
effect (Refs. 10 and 34).
    Using the no-observed-effect level of 500 mg/kg bw/d and applying a 
100-fold safety factor, the agency has determined an ADI of 5 mg/kg bw/
d for sucralose. This ADI estimate is well above the 90th-percentile 
EDI for sucralose of 1.6 mg/kg bw/d (Refs. 10 and 45).
    The agency concludes that the 2-year rat carcinogenicity study 
(E053) on the sucralose hydrolysis products established a no-observed-
effect level at the 0.6 percent dose level (equivalent to 30 mg/kg bw/
d). Therefore, the agency has no safety concerns about the sucralose 
hydrolysis products at their anticipated levels of intake (0.0048 mg/kg 
bw/d) because of the substantial margin of safety between these levels 
and the no-observed-effect level.

III. Comments

    The agency received several comments on McNeil's sucralose 
petition. Several comments supported amending the food additive 
regulations for the safe use of sucralose (Ref. 47). Other comments, 
principally from Malkin Solicitors (Malkin, formerly Malkin-Janners) 
and the Center for Science in the Public Interest (CSPI) (Refs. 48 and 
49) raised several issues which they claimed McNeil's petition had not 
addressed. The issues raised by the comments and the agency's responses 
are discussed in this section of this document.
    In addition, CSPI submitted a draft report from Life Science 
Research Limited of Suffolk, England entitled ``An investigation of 
diet spillage among rats fed diet containing sucralose.'' This draft 
report was provided to CSPI by an individual who stated that the study 
was undertaken by McNeil but was uncertain that the study report had 
been submitted to FDA. The diet spillage study in rats (El54) was 
subsequently submitted to the agency by McNeil in March, 1992. As 
discussed in section II.B.5.a.i. of this document, the agency concludes 
that the study raises no unique issue and contributes very little to 
the resolution of the issue of decreased food intake by sucralose-
treated rats.

A. Determination of No-Observed-Effect Level and ADI

1. No-Observed-Effect Level in the Chronic Toxicity Study
    Malkin pointed to decreases in body weight gain of 13 to 20 
percent, 19 to 24 percent, and 20 to 26 percent observed in animals in 
the three treatment groups compared to control animals in the combined 
chronic/

[[Page 16427]]

carcinogenicity study in rats (E057) and claimed that, because 
decreases in body weight of greater than 10 percent can be interpreted 
as an indication of toxicity, a no-observed-effect level was not 
established in this study. Malkin cited several observations from 
studies in the McNeil petition that suggest that the decreased body 
weight gain was not due solely to poor palatability as McNeil asserted.
    In addition, Malkin contended that the petitioner overstated the 
actual doses in the combined chronic toxicity/carcinogenicity study 
(E057) in rats because the diets were formulated with a constant 
percentage of sucralose throughout the study. Thus, the actual dose per 
body weight was variable depending on food consumption and the weight 
of the animal. Therefore, the dosage received later in life is lower 
than that received by the young, and Malkin contended that depending on 
which dosage was used, the no-observed-effect level and the ADI can 
vary significantly.
    FDA agrees in part with certain assertions made in the Malkin 
comment but disagrees with the overall significance of the findings 
identified by Malkin. Specifically, as discussed previously, the agency 
also found that the data in the original petition were not adequate to 
determine whether the body weight gain decrement was due solely to a 
palatability-induced decrease in food consumption or whether the weight 
gain decrement was due to effects mediated by sucralose. Therefore, the 
petitioner conducted an additional, carefully controlled weight gain 
study (diet restriction study, E160, which was submitted after the 
Malkin comment was received) to resolve the body weight gain decrement 
issue. Based on this study, the agency concludes that sucralose has a 
treatment-related effect on body weight gain when fed orally to rats at 
a concentration of 3 percent (Refs. 10, 28, 33, 34, and 46). Also the 
agency agrees with the comment that the decrements in body weight gain 
observed in the combined chronic carcinogenicity study (E057) cannot be 
explained solely by differences in food intake due to reduced 
palatability of the sucralose-containing diet. The mechanism by which 
sucralose affects body weight gain in rats is unknown. The agency 
concludes, however, that a no-observed-effect level for sucralose, as 
discussed previously, was demonstrated in the diet restriction study 
(EI60).
    Regarding the dosage calculations, the agency considers it 
inappropriate to limit the dosage calculation to any one time point in 
the study (Ref. 46). The agency normalizes the data and in doing so 
takes into consideration the increased dosage during the growing phase 
and the lower dosage during adulthood to provide an average intake.  In 
 reviewing  the  achieved  dosages provided  in study E057,  the agency 
 found  that male  rats achieved  an average  high  dose of 1.3 g/kg 
bw/d, while females achieved an average high dose of 1.7 g/kg bw/d. The 
average of the two equals 1.5 g/kg bw/d. Thus, the agency concludes 
that this dose was calculated using the standard techniques for 
calculating a lifetime dose and is not an overstatement of the actual 
dose.
2. No-Observed-Effect Level in Developmental Toxicology Studies
    Malkin stated that the ``Two-Generation Reproduction Study of 
Sucralose in Rats'' (E056) did not establish a no-observed-effect level 
because of dose-related reductions in pup body weight and statistically 
significant, dose-related decreases in body weight gain in pups from 
day 1 through weaning in two generations (F1 and 
F2). In addition, Malkin stated that there was a recurring 
dose-related increase in relative kidney weights.
    The purpose of this reproduction study (E056) was to assess the 
potential effects of sucralose on reproduction. The experimental design 
of such studies limits the measuring of food consumption by the pups, 
especially during lactation (Refs. 10, 40, and 50). However, precise 
food consumption measurements are essential to evaluate the potential 
for a substance to affect body weight gain. Therefore, study E056 
cannot be used to draw conclusions about body weight gain. Moreover, 
body weight gain effects were comprehensively studied in other studies 
(El60 and E161). As discussed previously, FDA disagrees with this 
comment. Regarding the increased kidney weights, microscopic 
examination of the kidneys of rats in the subchronic studies (El5l and 
E161) revealed no histopathological changes and therefore, FDA 
determined that these increases in relative kidney weight in these rats 
were not toxicologically significant.
    Malkin also asserted that the no-observed-effect level in the 
teratology study in rabbits (El34) is 350 mg/kg bw/d rather than 700 
mg/kg bw/d proposed by the petitioner.
    Although no frank terata were observed at any of the tested doses 
in this study (El34), the agency finds that toxicity elicited at the 
high dose (700 mg/kg bw/d) prevented the use of this dose to assess 
teratological effects. Therefore, as discussed previously, the agency 
agrees that the no-observed-effect level in the rabbit teratology study 
is 350 mg/kg bw/d (Refs. 40 and 50).
3. Derivation of ADI
    CSPI challenged the derivation of the ADI for sucralose (15 mg/kg 
bw/d) conducted by the Food and Agriculture Organization/World Health 
Organization (FAO/WHO) Joint Expert Committee on Food Additives (JECFA) 
and by McNeil. CSPI contended that the appropriate ADI ranges from 0.2 
to 8 mg/kg bw/d depending on the study used to derive the ADI. CSPI 
used a large number of safety factors ranging from 10 to 1,000 to 
derive the ADI from each of the studies which included: (1) The 8-week 
dose range-finding study (E031); (2) the two-generation reproduction 
toxicity study (E056); and (3) the long-term feeding studies in the rat 
(2 years) (E057), the mouse (2 years) (E055), and the dog (1 year) 
(E051). In addition, CSPI cited the clinical study (E047) as supporting 
the animal-derived ADI's.
    As discussed in section II.C of this document, FDA has evaluated 
all the studies in McNeil's petition and has concluded that the 
combined chronic toxicity/carcinogenicity study in rats (E057), 
interpreted in light of the data in the diet restriction study (El60) 
and the 26-week gavage study (El6l), provides the most appropriate 
basis for establishing the ADI for sucralose. This study (E057) 
provides a no-observed-effect level of 500 mg/kg bw/d; these results 
are corroborated by data from the diet restriction study (El60) in rat. 
Applying a 100-fold safety factor (21 CFR 170.22) results in an ADI for 
sucralose of 5 mg/kg bw/d (Ref. 10).
    The combined chronic toxicity/carcinogenicity rat study (E057) 
provides certain distinct advantages over other studies in the 
sucralose petition in terms of establishing an ADI. The agency did not 
use the 8-week range-finding (E031) or two generation reproduction 
(E056) studies because they were too brief and, compared to chronic 
studies, they lack the capability to measure general toxicity. The 1-
year chronic toxicity study in dogs (E051) showed no toxic effect at 
any dose tested and thus, provides no basis for concluding that the ADI 
should be lower than that established in the rat study. Although the 2-
year carcinogenicity study in mice (E055) established a higher no-
observed-effect level of 1,500 mg/kg bw/d, it did not include an in 
utero exposure of the animals to sucralose. Finally, the agency notes 
that the purpose of the clinical study (EO47) was to assess tolerance 
and acceptance of sucralose and, thus, it was not designed nor intended 
to

[[Page 16428]]

assess the toxicity of this compound (Refs. 10 and 51). Thus, use of 
the combined toxicity/carcinogenicity study in rats (E057) to establish 
the ADI for sucralose is sound and scientifically preferred.

B. Immunotoxic Potential of Sucralose

    The Malkin comments claimed that the following observations may 
have significance relative to the potential immunotoxicity of 
sucralose: (1) Dose-related decreases in thymus weights with concurrent 
decreases in white blood cell or lymphocyte counts (lymphocytopenia) in 
the 1-year chronic toxicity study in dogs (E051); (2) dose-related 
decreases in thymus weight that were seen in the parental rats and 
offspring in the two-generation reproduction study (E056); and (3) 
decreased spleen weights at the two highest dosages in the 4- to 13-
week sucralose oral gavage rat study (El5l). Malkin further asserted 
that these findings are important in view of published data that 
establish that the immune system is a target organ for some chlorinated 
compounds. Malkin also contended that these alleged immunotoxic effects 
cannot be explained by decreased food consumption and that a more 
direct evaluation of immunotoxicity potential should be done for 
sucralose (Ref. 48).
    CSPI also questioned whether sucralose has a toxic effect on the 
thymus. In their comment, CSPI discussed various effects that were 
demonstrated in the 4- to 8-week range-finding study in rats (E031), 
i.e., splenic hypoplasia of lymphoid tissues, cortical hypoplasia of 
the thymus, and decreased spleen, adrenal, and thymus weights. CSPI 
also cited the lymphocytopenia that was observed in rodents and dogs in 
the sucralose studies (Ref. 49).
    From a comparative analysis of thymus weight data, body weight 
data, and food consumption data in the sucralose rat studies, CSPI 
concluded that the relative thymus weight in sucralose-fed rats is much 
more severely affected than in diet restricted animals (Ref. 48). CSPI 
further asserted that thymus histopathology was not evaluated in all of 
the sucralose studies. CSPI also questioned the appropriateness of the 
reevaluation of the thymic histopathological examinations by McNeil in 
the 4- to 8-week range-finding study (E031). Finally, CSPI asserted 
that adequate studies of immune system function, including a clinical 
study, should be conducted (Ref. 49).
    After the Malkin and CSPI comments were received by FDA, McNeil 
conducted a 28-day oral immunotoxicity study in rats (EI62) in which a 
number of immunological parameters were examined. In this study, 
sucralose was administered by gavage at dose levels of 750, 1,500, and 
3,000 mg/kg bw/d and also in the diet at a level of 3,000 mg/kg bw/d. 
As discussed in section IIB.5 of this document, the only treatment-
related effect observed in this study was decreased thymus weight. FDA 
determined that a dose level of 750 mg/kg bw/d was the no-observed-
effect level for this study (Ref. 37). This no-observed-effect level is 
1.5 times higher than the no-observed-effect level established from 
body weight gain decrements observed in studies E057 and E160, which 
studies FDA used to determine an ADI of 5 mg/kg bw/d for sucralose. The 
ADI assures that the proposed use levels of sucralose pose no safety 
concerns regarding immunotoxicity.
    In addition, other studies of sucralose lacked evidence of 
immunotoxic effects. In the combined chronic toxicity/carcinogenicity 
rat study (E057), a dose of 500 mg/kg bw/d demonstrated no 
immunodeficiencies in rats exposed in utero, during lactation, and 
through their entire lifespan. Likewise, no immunotoxic effects were 
demonstrated in any of the clinical chemistry parameters nor were 
immunotoxic effects observed in the histopathological examinations of 
the sucralose-gavaged rats in the 26-week gavage study (EI61), in which 
sucralose was administered at doses up to 3000 mg/kg bw/d. This study 
is discussed in section II.B.5.a.ii of this document.
    Therefore, the agency concludes that the available animal data 
provide adequate evidence that sucralose will not be immunotoxic to 
humans at the projected level of dietary exposure (Refs. 40 and 50).

C. Mutagenicity of 1,6-DCF

    Malkin claimed that data in the sucralose petition showed that 1,6-
DCF, a sucralose hydrolysis product, is mutagenic in the Ames assay and 
is a more potent mutagen than unhydrolyzed sucralose in the mouse 
lymphoma assay. Further, Malkin stated that the mutagenic potential of 
1,6-DCF is established by its ability to alkylate 4-(paranitrobenzene)-
pyridine in an assay which has been used to demonstrate the alkylating 
nature of carcinogenic hydrocarbons, some of which were known to bind 
covalently to DNA, and by the association of 1,6-DCF with DNA in all 
tissues including the testes. Thus, Malkin asserted that it is 
imperative to demonstrate in vivo that 1,6-DCF does not covalently bind 
to DNA or other chromosomal proteins in germ cells (Ref. 48). CSPI also 
asserted that the DNA-binding capacity and mutagenic potential of 1,6-
DCF should be carefully reviewed (Ref. 49).
    As discussed in section II.B.2 of this document, the data from the 
genotoxic studies are of limited toxicological significance because the 
results of the mutagenic testing were equivocal and because such tests 
are used primarily as a guide to assess the need for more powerful 
bioassays. While 1,6-DCF was weakly mutagenic in the Ames test (E020) 
and the L5178Y TK+/assay (E022, E024), the results from the combined 
chronic toxicity/carcinogenicity study (E057) and the carcinogenicity 
study on an equimolar mixture 4-CG and 1,6-DCF (E053) establish that 
sucralose and its hydrolysis products do not elicit tumor formation. 
Because of the longer exposure duration and greater testing sensitivity 
of carcinogenicity bioassays, such as E057 and E053, the negative 
results in these carcinogenicity bioassays of sucralose and its 
hydrolysis products (E057 and E053) supersede the equivocal results 
obtained in the genotoxicity studies on sucralose and its hydrolysis 
products cited by the Malkin and the CSPI comment (Refs. 5 and 50).

D. Renal Effects

    CSPI asserted that McNeil's hypothesized etiology of sucralose-
induced rat renal changes (i.e., secondary to cecal enlargement and not 
likely to be significant at low intake) should be proved and that the 
renal changes observed in the female rats should be interpreted as 
being of toxicological significance. Also, the comment asserted that 
the available data are insufficient to conclude that the 
nephrocalcinosis (deposition of calcium in the kidney) is only an 
indirect consequence of cecal enlargement (Ref. 49).
    First, nephrocalcinosis is not uncommon in the rat, particularly 
the female rat (Refs. 21, 22, and 23). Investigators have reported the 
incidence of renal calcification as high as 100 percent in female rats 
used as controls with a complete absence of this condition in male rats 
fed the identical diet (Ref. 21). Because mice and other rodent models 
do not experience the condition, FDA believes that the rat, especially 
the female rat, is uniquely sensitive to the development of 
nephrocalcinosis and, therefore, is an inappropriate surrogate for man 
with respect to this pathologic endpoint.
    Second, as discussed in section II.B.4.a.i of this document, the 
agency

[[Page 16429]]

recognizes that a number of poorly or slowly absorbed compounds mediate 
changes in physiologic function that result in renal mineralization, as 
observed in this study (Refs. 6, 21, and 26). In response to the 
feeding of poorly absorbed compounds, like sucralose, cecal enlargement 
in association with renal changes occurs frequently in old rats (Refs. 
21 and 26). Increased calcium absorption and excretion, pelvic 
nephrocalcinosis, increased water retention, and alterations of the gut 
microflora occur as physiologic adaptive responses to changes in 
osmolality in the gut that lead to cecal enlargement (Refs. 21, 22, and 
23). Therefore, cecal enlargement is a physiologic adaptive change 
rather than a toxic effect (Ref. 26).
    Third, in the carcinogenicity study of sucralose hydrolysis 
products (EO53), which was concurrently conducted in the same 
laboratory with study E057, the incidence of nephrocalcinosis in the 
control group was 33 percent (Ref. 26). This incidence is comparable to 
that observed in the mid- (32 percent) and high- (30 percent) dose 
treated groups in the combined chronic toxicity/carcinogenicity 
sucralose study (EO57). The agency concludes that the nephrocalcinosis 
is not toxicologically significant for the foregoing reasons.

E. Fetal Edema

    Malkin stated that the teratology study of sucralose in rats (E030) 
indicates an apparent increase in the incidence of subcutaneous edema 
in fetuses. Malkin noted that the expected occurrence of fetal edema at 
the Life Science Research Limited (LSRL) laboratory of Essex, England, 
where the McNeil teratology study was conducted, was 12 percent. In 
contrast, Malkin asserted that the historical incidences of 
subcutaneous fetal edema for Charles River CD rats is approximately 
0.03 percent and the incidence based on data derived from nine United 
States teratology laboratories is 0.007 percent. Malkin concluded that 
the unusually large background incidence of edema seen at LSRL may mask 
a treatment-related increase in subcutaneous edema (Ref. 48).
    The agency believes that the most appropriate historical control 
values to use in considering the significance of a response in an 
animal bioassay are those pertaining to the identical strain of animal 
used in the study and drawn from the testing laboratory used for the 
study (Refs. 40 and 50). It is inappropriate to compare data from 
Charles Rivers CD rats that were bred in two different countries 
because, due to genetic divergence, different ranges of normalcy as 
well as spontaneous malformations are likely to exist for each colony 
(Ref. 50).
    The rat teratology study in question (E030) was conducted in an 
LSRL laboratory, utilizing a Charles River rat derived in England. The 
historical control data from LSRL showed the incidence of subcutaneous 
fetal edema in Charles River rats to range from 0 to 32 percent. In the 
teratology study in rats (E030), which was performed in England, the 
reported incidences of subcutaneous fetal edema were 15.6, 20.9, 20.5, 
and 25.6 percent for the control, low, mid, and high dosages, 
respectively. These incidences fall within the LSRL historical control 
range (Ref. 40). Additionally, the slightly increased incidences in 
subcutaneous fetal edema in the sucralose treated rats raised by the 
Malkin comment (E030) were not statistically different when compared to 
their concurrent controls (Refs. 13, 40, and 50). Thus, the incidences 
of subcutaneous fetal edema identified by the Malkin comment are 
considered by FDA to be of no toxicological significance.

F. Bioaccumulation

    The Malkin comment raised three issues concerning the possible 
bioaccumulation of sucralose. First, Malkin disputed McNeil's 
calculation of an ``effective half-life'' of 13 hours for sucralose. 
Instead, Malkin asserted that sucralose has a ``terminal half-life'' of 
24 hours in healthy humans, which is, Malkin asserts, indicative of the 
potential for sucralose to accumulate in the body of consumers. 
Further, Malkin stated that the remaining 4 to 7 percent of 
radioactivity not excreted 5 days after a single dose of sucralose in 
humans indicates that sucralose may never be totally excreted from the 
body, even for periodic users. Second, Malkin pointed to data on 
sucralose metabolism in dogs (EI23) which show that 20 percent of the 
oral dose was not recovered 4 days after dosing with 36Cl 
labeled sucralose and claimed that this residual radioactivity 
represents either potential bioaccumulation, extensive in vivo 
dechlorination, or both. Finally, Malkin stated that there was a 
potential for sucralose to accumulate in the fetus because of its 
extremely slow elimination from fetal tissue.
    The available pharmacokinetics data in the petition do not allow 
the agency to draw definitive conclusions regarding bioaccumulation of 
sucralose and its metabolites. However, the available evidence on the 
physicochemical properties of sucralose, such as low lipid solubility 
and high water solubility, is not representative of compounds that 
manifest a high potential for bioaccumulation (Refs. 50 and 53). In 
addition, sucralose is relatively poorly absorbed from the gut in 
humans in that only 11 to 27 percent of the administered dose is 
absorbed. Finally, there is little or no evidence of direct tissue 
toxicity from sucralose in the mouse, rat, and dog, even when 
administered at high doses for 1 to 2 years. In a practical sense, the 
absence of tissue toxicity is more important because even if sucralose 
had accumulated to some limited degree in these animals, no organ 
toxicity was demonstrated in any of the long-term studies (E055, E057, 
and E051).

G. Antifertility Effects

    Malkin asserted that antifertility effects were observed with 
unidentified degradation products of sucralose (Ref. 48). In evidence 
of this assertion, Malkin pointed to results of a study (E004) 
conducted by McNeil in which sucralose and/or its metabolites 
distribute to and have a long residual time in testes. Malkin cited a 
literature publication by Ford and Waites (Ref. 17) where sucralose was 
shown to inhibit the oxidation of glucose and decrease the 
concentration of adenosine triphosphate in epididymal spermatozoa. 
Malkin further asserted that these observations must be reviewed in the 
context of the known antifertility effects of other chlorosugars (Ref. 
48).
    The results obtained in study E004 were discounted by the 
petitioner because there were indications that the sucralose sample 
used in the study were degraded. A subsequent repeat test (study E107) 
that was performed by McNeil showed sucralose had no effect on the 
glycolytic activity of sperm from male rats.
    The agency concludes from stability data contained in the sucralose 
petition that sucralose is stable under the proposed conditions of use 
(Refs. 52 and 53). Therefore, the agency would not expect significant 
amounts of degradation products to be formed from the proposed uses of 
sucralose.
    The agency has previously discussed in this preamble the studies 
mentioned in the Malkin's comment. With regard to the Malkin comment 
claiming accumulation of sucralose and its metabolites in testes, the 
available pharmacokinetics data in the sucralose petition do not allow 
the agency to draw definitive conclusions regarding the bioaccumulation 
of sucralose and its metabolites. However, neither of the two-
generation reproduction studies (E052 and E056) showed any reproductive 
toxicity that was

[[Page 16430]]

treatment-related. Again, this absence of reproductive toxicity is 
directly relevant to the Malkin comment about antifertility effects and 
demonstrates that any speculation about bioaccumulation is of no 
practical significance.
    The agency noted insufficiencies in the antifertility studies on 
sucralose and its hydrolysis products, specifically in their duration, 
and therefore concludes that they are inadequate to assess the 
antifertility potential of sucralose (Refs. 5, 18, and 54). More 
importantly, however, results from the two-generation reproduction 
studies (E052 and E056) do adequately address any potential 
toxicological concern regarding the antifertility potential of 
sucralose and its hydrolysis products. Evidence presented in the 
reproduction studies supports the conclusion that sucralose and its 
degradation products do not possess antifertility properties (Refs. 5, 
12, and 18).

H. Neurotoxicity Effects

    Malkin stated that neurotoxic effects of some chlorosugars have 
been reported and pointed out that 6-chloro-6-deoxyglucose (6-CG) is 
used as a positive control for CNS neuropathology and neuromuscular 
deficits (Ref. 48). Therefore, Malkin stated that neurobehavioural 
studies of sucralose should be assessed in an appropriate study.
    FDA has evaluated the petitioner's neurotoxicity studies, E008 
(mice) and E009 (monkey), which compared the potential neurotoxic 
effects of sucralose or its hydrolysis products with the positive 
control 6-CG (Refs. 38 and 39). As discussed in section II.B.5.c of 
this document, FDA finds that neither mice nor monkeys showed 
neurological effects after receiving sucralose or equimolar mixtures of 
sucralose hydrolysis products at levels as high as 1000 mg/kg bw/d for 
21 and 28 days respectively.

I. Exposure to Sucralose Hydrolysis Products

    Malkin stated that in acidic drinks such as powdered cherry drinks 
(storage temperature, 35  deg.C) and carbonated soft drinks (storage 
temperature, 22  deg.C), sucralose concentrations decrease by 4 percent 
to 20 percent after a 6-month storage and if, as the petitioner states, 
the disappearance of sucralose results in the appearance of 
stoichiometric amounts of the hydrolysis products 4-CG and 1,6-DCF, 
human exposure to these hydrolysis products will be significantly 
greater than the 10 mg/kg body weight claimed by the petitioner (Ref. 
48).
    The agency notes that even if the decomposition noted after 6 
months at 35  deg.C (an 18 percent decrease of sucralose) was accepted 
as representative of actual use, the probable exposure to hydrolysis 
products would not change appreciably from the current estimate of 285 
g/p/d (90th percentile, 4.8 g/kg bw/d) because 
beverages account for only 13 percent of the estimated exposure to 
sucralose. Nonetheless, the agency does not believe that such abusive 
storage conditions should be assumed when considering chronic exposure 
(Refs. 52 and 53). The data for storage at 20  deg.C, and for storage 
at 35  deg.C for up to 3 months show no decomposition of sucralose 
within experimental error. The sucralose content of carbonated 
beverages also does not change significantly under typical storage 
conditions. Finally, the no-observed-effect level established for the 
hydrolysis products is 30,000 g/kg bw/d, so there is an 
adequate safety margin to allow for additional decomposition of 
sucralose to the hydrolysis products.

J. The Need for Studies in Special Populations

    CSPI stated that, although McNeil showed that sucralose does not 
affect insulin secretion and action, and glucose metabolism in normal 
human subjects (E046), non-diabetic rats, and non-diabetic dogs, there 
are no clinical studies of type I and II diabetics or the ``diabetic'' 
rat. CSPI contended that sucralose will be in heavy use by diabetics 
and that before approving sucralose, the agency should require the 
results of testing of the effects of sucralose in diabetics (Ref. 49).
    First, FDA believes that these comments do not preclude the 
conclusion that the proposed uses of sucralose are safe. The EDI 
(discussed in section II.A of this document) of sucralose (90th 
percentile) established by the agency would include those levels 
expected to be ingested by diabetics (Refs.1, 2, 53, and 55). The 90th 
percentile level of consumption used by FDA is an amount equivalent to 
the sweetness that would be provided by the total amount of sugars 
commonly added to the diet. Thus, the estimates of heavy consumption of 
sucralose used by FDA would cover estimated intake of sucralose by 
diabetics who might preferentially select sucralose-containing 
products.
    Second, after this comment was received by FDA, McNeil did perform 
studies on sucralose in diabetic individuals. Specifically, McNeil has 
submitted a series of studies (E156, E157, E168, E170, and E171) that 
investigated the short-term and long-term effects of sucralose on 
glucose homeostasis in patients with IDDM and NIDDM. These studies were 
previously discussed in detail earlier in this document. Based upon the 
data from these studies, the agency concludes that sucralose has no 
adverse health effects on short-term or long-term glucose homeostasis 
or any other adverse effect in diabetic patients (Refs. 41, 43, 44, 
45). The sucralose exposure tested in the diabetic study E171, where no 
effect on glycemic control in diabetics was observed, is seven times 
higher than the 90th percentile EDI estimate expected from the proposed 
uses of sucralose. This 90th percentile exposure estimate represents 
the expected use of sucralose by the heavy eater population and also 
encompasses the level that is expected to be ingested by the diabetic 
population (Ref. 5).
    Additionally, none of the data in the animal studies in the 
sucralose data base that examined the effect of sucralose on 
carbohydrate/glucose metabolism provided any evidence to suggest that 
diabetics would be at any greater risk than the general human 
population (Ref. 46). These studies show that: (1) Sucralose has no 
influence on insulin secretion by rats or humans; (2) sucralose has no 
effect on postprandial or fasting blood glucose levels in animals or 
humans; (3) sucralose causes no changes in intestinal absorption of 
glucose or fructose; (4) sucralose has no effect on glucose utilization 
or on any of the key enzymes modulating glucose metabolism or storage; 
(5) administration of sucralose results in no clinical or pathological 
symptoms similar to those observed in diabetes mellitus; and (6) 
because sucralose has no influence on insulin's action on blood glucose 
levels, it would not be anticipated to result in difficulties with 
insulin-based management of diabetes. Therefore, on the basis of the 
data in the clinical studies and other available information in the 
sucralose database, the agency has no safety concerns regarding the use 
of sucralose by diabetic individuals.
    Another comment by Malkin speculated that the chlorinated galactose 
component of sucralose may have an effect on individuals with 
diminished ability to metabolize galactose (galactosemic individuals). 
Malkin further speculated that 4-chlorogalactose, a sucralose 
degradation product, may act as a substrate for enzymes that metabolize 
galactose in normal individuals, or may inhibit galactosyltransferase, 
an enzyme largely

[[Page 16431]]

responsible for the production of milk in humans.
    As discussed previously, from the review of the stability data 
submitted in the sucralose petition, the agency would not expect 
significant amounts of degradation products to be formed as a result of 
the proposed uses of sucralose. Therefore, exposure to degradation 
products from the use of sucralose would be minimal and would be of no 
toxicological significance.
    In another comment, Malkin criticized the petitioner's metabolism 
data because the data were obtained from healthy adults and did not 
address metabolism or safety in children, diabetics, or the obese.
    First, as noted, the petitioner did conduct several studies of 
sucralose use in diabetics. Moreover, there are no data that would 
suggest any particular reason to expect an increased potential for 
adverse effects in children and obese people and other subpopulations. 
The Malkin comment did not present any data or evidence that suggest 
that these subpopulations are at special risk. In the absence of such 
data, the agency determines an additive's safety based on studies 
conducted in healthy test animals at doses far in excess of the maximum 
anticipated exposure in humans. In addition, in setting an ADI, the 
agency uses a 100-fold safety factor after determining the highest no-
adverse-effect level. The agency uses a 100-fold safety factor as a 
means to account for differences between animals and humans and to 
account for differences in sensitivity among humans. For these reasons, 
the agency believes that studies aimed at addressing effects in the 
subpopulations indicated are not warranted.

K. Labeling

    In response to a November 22, 1991 (56 FR 58910), request by FDA 
for comments on a proposed monograph for sucralose for inclusion in the 
Food Chemicals Codex, Malkin stated that the name sucralose is 
inaccurate, deceptive, and will mislead consumers because of the close 
similarity to the name sucrose, a product for which sucralose might be 
a replacement. Because sucralose is a chlorinated version of a 
disaccharide, Malkin contended that the common name should not 
misrepresent the makeup of the material. Malkin cited Sec. 102.5(a) and 
(c) (21 CFR 102.5(a) and (c)) and contended that the common name should 
indicate that the material is a disaccharide, reflect the presence of 
chlorine, and avoid confusion with sucrose. Malkin stated that the name 
used by the FAO/WHO JEFCA ``trichlorogalactosucrose'' or a similarly 
accurate name such as trichlorofructogalactose should be used.
    Section 403(i)(2) of the Federal Food, Drug, and Cosmetic Act (21 
U.S.C. 343(i)(2)) deems a food that is fabricated from two or more 
ingredients to be misbranded unless its label bears the common or usual 
name for each ingredient. Section 102.5(a) states, in part, that: ``The 
common or usual name of a food, which may be a coined term, shall 
accurately identify or describe, in as simple and direct terms as 
possible, the basic nature of the food or its characterizing properties 
or ingredients. The name shall be uniform among all identical or 
similar products and may not be confusingly similar to the name of any 
other food that is not reasonably encompassed within the same name.'' 
Section 102.5(c) addresses the need for the common or usual name of a 
food to include a statement of the presence or absence of any 
characterizing ingredients or components, whether such ingredients need 
to be added, whether the absence or presence has a bearing on price, 
and similar issues that may cause a consumer to purchase a product that 
is not what it appears to be.
    Sucralose is a single ingredient and has no other characterizing 
ingredients or components that are added or removed. Thus, 
Sec. 102.5(c) does not govern the question of what is the appropriate 
name for this additive.
    Under Sec. 102.5(a), a substance may be described by a coined term 
provided that it accurately identifies, in as simple and direct terms 
as possible, the nature of the food, i.e., the food additive sucralose. 
While the names suggested by Malkin may be suitable for describing the 
nature of the substance to a chemist, they are not the most direct and 
simple terms for the average consumer. FDA recognizes that the precise 
chemical names of additives may not be helpful for consumers and has 
permitted the use of a simple coined name that consumers can 
understand. For example, none of the three intense sweeteners currently 
allowed in food, saccharin, aspartame, and acesulfame potassium, are 
described by their specific chemical names. This causes no confusion, 
however. The important issue is whether the name is commonly used for 
the substance and whether that name could be misleading for some 
reason.
    Although Malkin states that the name trichlorogalactosucrose is 
used by JEFCA for this additive, that organization has since the 
comment was submitted accepted sucralose as the preferred name. 
Additionally, the additive is regulated under the name sucralose in 
both Canada and Australia. Thus, it is consistent with the 
international marketplace, including other English speaking countries, 
to describe the additive by the name sucralose. Similarly, the Food 
Chemicals Codex has also published a monograph under the name 
sucralose. For these reasons, the agency concludes that the name 
sucralose is the common name, accurately identifies the additive, and 
will not mislead consumers.

IV. Conclusion

    The agency has evaluated all the data in the petition and other 
information and concludes that the proposed uses of sucralose are safe. 
Therefore the agency concludes that the food additive regulations 
should be amended as set forth in this document.
    In accordance with Sec. 171.1(h) (21 CFR 171.1(h)), the petition 
and the documents that FDA considered and relied upon in reaching its 
decision to approve the petition are available for inspection at the 
Center for Food Safety and Applied Nutrition by appointment with the 
information contact person listed above. As provided in Sec. 171.1(h), 
the agency will delete from the documents any materials that are not 
available for public disclosure before making the documents available 
for inspection.

V. Environmental Effects

    The agency has carefully considered the potential environmental 
effects of this action. FDA has concluded that the action will not have 
a significant impact on the human environment, and that an 
environmental impact statement is not required. The agency's finding of 
no significant impact and the evidence supporting that finding, 
contained in an environmental assessment, may be seen in the Dockets 
Management Branch (address above) between 9 a.m. and 4 p.m., Monday 
through Friday.

VI. References

    The following references have been placed on display in the Dockets 
Management Branch (address above) and may be seen by interested persons 
between 9 a.m. and 4 p.m., Monday through Friday.
    1. Memorandum, from DiNovi, Chemistry Review Branch, to 
Anderson, Novel Ingredients Branch, September 21, 1993.
    2. Memorandum, from DiNovi, Food and Color Additives Review 
Section, to Anderson, Direct Additives Branch, May 18, 1989.
    3. Memorandum from Roth, HFS-506, to Review Staff Office of 
Premarket Approval, October 26, 1994.
    4. Memorandum, from Roth, HFS-506, to Anderson, Division of 
Product Policy, October 26, 1994.

[[Page 16432]]

    5. Memorandum, from Graham, Additives Evaluation Branch, to 
Anderson, Direct Additives Branch, August 8, 1990.
    6. Addendum memorandum, from Graham, Additives Evaluation 
Branch, to Anderson, Direct Additives Branch, April 12, 1991.
    7. Memorandum, from Griffiths, Additives Evaluation Branch, to 
Anderson, Direct Additives Branch, August 22, 1988.
    8. Addendum memorandum, from Graham, Additives Evaluation 
Branch, to Anderson, Direct Additives Branch, August 12, 1991.
    9. Memorandum, from Dunkel, Genetic Toxicology Branch, to 
McLaughlin, Direct Additives Branch, May 23, 1984.
    10. Memorandum, from Whiteside, Additives Evaluation Branch No. 
2, to Anderson, Novel Ingredients Branch, May 26, 1995.
    11. Memorandum, from Whiteside, Additives Evaluation Branch No. 
2., to Anderson, Direct Additives Branch, January 21,1994.
    12. Memorandum, from Collins, Mammalian Reproduction and 
Teratology Team, to Gryder, Additives Evaluation Branch, August 15, 
1987.
    13. Memorandum, from Welsh, Whole Animal Toxicology Branch, to 
McLaughlin, Direct Additives Branch, February 1, 1984.
    14. Addendum Memorandum, from Whiteside, Additives Evaluation 
Branch, to Anderson, Direct Additives Branch, November 12, 1991.
    15. Memorandum, from Collins, Mammalian Reproduction and 
Teratology Team, to Gryder, Additives Evaluation Branch, October 2, 
1987.
    16. Memorandum, from Welsh, Mammalian Reproduction and 
Teratology Team, to Bleiberg, Division of Toxicology, July 15, 1986.
    17. Ford, W. C. L., and G. M. H. Waites, ``A Reversible 
Contraceptive Action of Some 6-chloro-6-deoxy Sugars in the Male 
Rat,'' Journal of Reproduction and Fertility, 52:153-157, 1978.
    18. Memorandum, from Whitby, Additives Evaluation Branch, to 
Anderson, Direct Additives Branch, December 20, 1988.
    19. Memorandum, Cancer Assessment Committee, October 20, 1987, 
January 26, April 6, and July 13, 1989.
    20. Memorandum, from Dua, Division of Pathology, to Lin, 
Additives Evaluation Branch, February 11, 1992.
    21. Lord, G. H., and P. M. Newberne, ``Renal Mineralization--A 
Ubiquitous Lesion in Chronic Rat Studies,'' Food Chemistry and 
Toxicology, 28:449-455, 1990.
    22. Newberne, P. M, M. W. Conner, and P. Estes, ``The Influence 
of Food Additives and Related Materials on Lower Bowel Structure and 
Function,'' Toxicologic Pathology, 16:184-197, 1988.
    23. Vaughan, O. W., and L. J. Filer, ``The Enhancing Action of 
Certain Carbohydrates on the Intestinal Absorption of Calcium in the 
Rat,'' Journal of Nutrition, 71:10-14, 1960.
    24. United Nations (UN) Environmental Programme, International 
Labour Organization, World Health Organization, Food and Agriculture 
Organization of the U.N., ``Principles for the Safety Assessment of 
Food Additives and Contaminants in Food,'' Geneva: World Health 
Organization, 1987 (Environmental Health Criteria, 70), pp. 41-42.
    25. De Groot, A. P., and V. J. Feron, ``Effects of Very High 
Dietary Levels of Lactose on the Kidneys of Rats.'' In: Report R4812 
Central Institute for Nutrition and Food Research (CIVO/TNO). Zeist, 
the Netherlands: CIVO/TNO; October 1975/March 1976: 1-6.
    26. Memorandum, from Graham, Additives Evaluation Branch, to 
Anderson, Direct Additives Branch, April 1, 1991.
    27. Memorandum, from Sucralose Working Group, to Rulis, Novel 
Ingredients Branch, August 3, 1992.
    28. Memorandum, from Sucralose Working Group, to Pauli, Novel 
Ingredients Branch, December 21, 1992.
    29. Memorandum, from Graham, Additives Evaluation Branch, to 
Anderson, Direct Additives Branch, May 11, 1987.
    30. Memorandum, from Bleiberg, Additives Evaluation Branch, to 
Anderson, Direct Additives Branch, October 20, 1987.
    31. Memorandum, from Graham, Additives Evaluation Branch, to 
Direct Additives Branch, August 3, 1987.
    32. Submission to FAP 7A3987, McNeil Specialty Products, May 5, 
1992.
    33. Memorandum, from Barton, Experimental Design and Evaluation 
Branch, to Anderson, Novel Ingredients Branch, April 26, 1994.
    34. Memorandum, from Whiteside, Additives Evaluation Branch No. 
2, to Anderson, Novel Ingredients Branch, June 29, 1994.
    35. Memorandum, from Whiteside, Additives Evaluation Branch No. 
2, to Anderson, Novel Ingredients Branch, July 8, 1994.
    36. Memorandum, from Barton, Experimental Design and Evaluation 
Branch, to Anderson, Novel Ingredients Branch, March 18, 1994.
    37. Memorandum, from Hinton, Biochemical and Analytical Branch, 
to Anderson, Novel Ingredients Branch, March 7, 1995.
    38. Memorandum, from Graham, Additives Evaluation Branch, to 
Anderson, Direct Additives Branch, February 17, 1988.
    39. Memorandum, from Graham, Additives Evaluation Branch, to 
Anderson, Direct Additives Branch, April 12, 1988.
    40. Addendum Memorandum, from Whiteside, Additives Evaluation 
Branch No. 2, to Anderson, Novel Ingredients Branch, May 26, 1995.
    41. Memorandum, from Wilcox, Epidemiology Branch, to Anderson, 
Novel Ingredients Branch, October 7, 1994.
    42. Memorandum, from Whiteside, Scientific Support Branch, to 
Anderson, Novel Ingredients Branch, November 13, 1997.
    43. Memorandum, from Fleming, Center for Drug Evaluation and 
Research, to Anderson, Novel Ingredients Branch, August 21, 1997.
    44. Memorandum, from Barton, Division of Mathematics, to 
Anderson, August 28, 1997.
    45. Addendum Memorandum, from Whiteside, Scientific Support 
Branch, to Anderson, Novel Ingredients Branch, November 13, 1997.
    46. Memorandum, from Yetley/Einhorn, Clinical Nutrition Branch, 
to Anderson, Director Additives Branch, January 8, 1990.
    47. Comments, from supporters of the petition, to Dockets 
Management Branch.
    48. Comments, from Malkin Solicitors.
    49. Comments, from Center for Science in the Public Interest 
(CSPI).
    50. Memorandum, from Graham, Additives Evaluation Branch, to 
Anderson, Direct Additives Branch, December 13, 1990.
    51. Memorandum, from Whiteside, Additives Evaluation Branch, to 
Anderson, Direct Additives Branch, November, 12, 1991.
    52. Memorandum, from DiNovi, Food and Color Additives Review 
Section, to Anderson, Direct Additives Branch, December 6, 1990.
    53. Memorandum, from Modderman, Food and Color Additives Review 
Section, to Anderson, Direct Additives Branch, May 20, 1988.
    54. Memorandum, from Graham, Additives Evaluation Branch, to 
Anderson, Direct Additives Branch, December 27, 1988.
    55. Memorandum, from Glinsmann, Clinical Nutrition, to Anderson, 
June 18, 1991.

VII. Objections

    Any person who will be adversely affected by this regulation may at 
any time on or before May 4, 1998, file with the Dockets Management 
Branch (address above) written objections thereto. Each objection shall 
be separately numbered, and each numbered objection shall specify with 
particularity the provisions of the regulation to which objection is 
made and the grounds for the objection. Each numbered objection on 
which a hearing is requested shall specifically so state. Failure to 
request a hearing for any particular objection shall constitute a 
waiver of the right to a hearing on that objection. Each numbered 
objection for which a hearing is requested shall include a detailed 
description and analysis of the specific factual information intended 
to be presented in support of the objection in the event that a hearing 
is held. Failure to include such a description and analysis for any 
particular objection shall constitute a waiver of the right to a 
hearing on the objection. Three copies of all documents shall be 
submitted and shall be identified with the docket number found in 
brackets in the heading of this document. Any objections received in 
response to the regulation may be seen in the Dockets Management Branch 
between 9 a.m. and 4 p.m., Monday through Friday.

List of Subjects in 21 CFR Part 172

    Food additives, Incorporation by reference, Reporting and 
recordkeeping requirements.

[[Page 16433]]

    Therefore, under the Federal Food, Drug, and Cosmetic Act and under 
authority delegated to the Commissioner of Food and Drugs, 21 CFR part 
172 is amended as follows:

PART 172--FOOD ADDITIVES PERMITTED FOR DIRECT ADDITION TO FOOD FOR 
HUMAN CONSUMPTION

    1. The authority citation for 21 CFR part 172 continues to read as 
follows:

    Authority: 21 U.S.C. 321, 341, 342, 348, 371, 379e.

    2. Section 172.831 is added to subpart I to read as follows:


Sec. 172.831  Sucralose.

    The food additive sucralose may be safely used as a sweetening 
agent in foods in accordance with current good manufacturing practice 
in an amount not to exceed that reasonably required to accomplish the 
intended technical effect in foods for which standards of identity 
established under section 401 of the Federal Food, Drug, and Cosmetic 
Act do not preclude such use under the following conditions:
    (a) Sucralose is the chemical 1,6-dichloro-1,6-dideoxy--D-
fructofuranosyl-4-chloro-4-deoxy--D-galactopyranoside (CAS 
Reg. No. 56038-13-2).
    (b) The additive meets the specifications of the ``Food Chemical 
Codex,'' 4th ed. (1996), pp. 398-400, which is incorporated by 
reference in accordance with 5 U.S.C. 552(a) and 1 CFR part 51. Copies 
are available from the the Division of Product Policy (HFS-206), Center 
for Food Safety and Applied Nutrition, Food and Drug Administration, 
200 C St. SW., Washington, DC 20204-0001, or may be examined at the 
Center for Food Safety and Applied Nutrition's Library, 200 C St. SW., 
rm. 3321, Washington, DC 20204-0001, or the Office of the Federal 
Register, 800 North Capitol St. NW., suite 700, Washington, DC.
    (c) The additive may be used as a sweetener in the following foods:
    (1) Baked goods and baking mixes;
    (2) Beverages and beverage bases;
    (3) Chewing gum;
    (4) Coffee and tea;
    (5) Dairy product analogs;
    (6) Fats and oils (salad dressing);
    (7) Frozen dairy desserts;
    (8) Fruit and water ices;
    (9) Gelatins, puddings, and fillings;
    (10) Jams and jellies;
    (11) Milk products;
    (12) Processed fruits and fruit juices;
    (13) Sugar substitutes (for table use);
    (14) Sweet sauces, toppings, and syrups;
    (15) Confections and frostings.
    (d) If the food containing the additive purports to be or is 
represented to be for special dietary use, it shall be labeled in 
compliance with part 105 of this chapter.

    Dated: March 30, 1998.
Michael A. Friedman,
Lead Deputy Commissioner for the Food and Drug Administration.
[FR Doc. 98-8750 Filed 4-1-98; 8:45 am]
BILLING CODE 4160-01-F