Melamine

Toxicological Summary

Workers - Hazard via inhalation route

Systemic effects

Long term exposure

Hazard assessment conclusion:

DNEL (Derived No Effect Level)

Value:

8.3 mg/m³

Most sensitive endpoint:

repeated dose toxicity

Route of original study:

Oral

DNEL related information

DNEL derivation method:

ECHA REACH Guidance

Overall assessment factor (AF):

37.5

Dose descriptor starting point:

LOAEL

Value:

126 mg/kg bw/day

Modified dose descriptor starting point:

LOAEC

Explanation for the modification of the dose descriptor starting point:

Standard derivation of the DNEL: 

- Starting points are the key NTP-studies: NOAELrat,oral,90d = 72 mg/kg bw/day as derived by NTP, respectively 63 mg/kg bw/d as derived by US FDA. The NOAELrat,oral,2years = 126 mg/kg bw/d is higher, but Hard et al. 2009 re-evaluated the NTP-studies and detected additional lesions in the kidney that turned the NOAEL to a LOAEL. This LOAEL rat,oral,2years = 126 mg/kg bw/d was used for the further calculations.

- Remarkable is that the NOAEL2years is higher than the NOAEL90d (leaving aside at first the reanalysis by Hard), but still are about the same if the results of Hard are included and an assement factor for a conversion of a LOAEL to a NOAEL is applied. This is unusual but is explainable by the known mode of action of toxicity with the primary step of exceeding the solubility of melamine in urine, which is independent of exposure duration.

- Bioavailability: From toxicokinetic studies (e.g. Mast 1983) it is known that melamine is absorbed rapidly and almost completely in rats after oral dosing. There is apparently no metabolism in rats. A transformation of the LOAELoral to LOAECinhalation is therefore justified, taking the default absorption rate for inhalation of 100 %, and setting also the oral absorption rate to 100 %. Correction factor used = 1.

- The LOAECrat,inhalation of 331.6 mg/m3 was calculated from the LOAELoral of 126 mg/kg bw and the standard respiratory volume sRV for the rat of 0.38 m³/kg bw for an 8 hour exposure. Conversion factor = 2.63.

- The LOAEC was further adjusted for an increased respiratory volume of 10 m³/8h for workers, compared to 6.7 m³/8h for the general public. Correction factor = 0.67.

- The exposure conditions for the orally dosed rats were 7d/w compared to the inhalation exposure of workers of 5d/w. Correction factor = 1.4.

- Assessment factors (AF1 to AF6) were selected according to Table R.8-6 of the Guidance on Information Requirement and CSA. AF1 = Allometric scaling; AF2 = Interspecies, remaining difference; AF3 = Intraspecies; AF4 = Exposure duration; AF5 = Dose-response relation; AF6 = Quality data base. Default values were used, except for AF5: A steep dose response curve, confirmed by EFSA (2010) page 103, is present that allows a low assessment factor AF5 of 3 to be applied to convert the LOAEL to a surrogate NOAELrat,oral,2years.

DNEL = 126 x 1 x 2.63 x 0.67 x 1.4 /(1x 2.5 x 5 x 1 x 3 x 1) = 8.29 mg/m³

 

1. First alternative DNEL calculation, using the NOAELmouse,oral,90d = 1400 mg/kg/d of the NTP-study as the starting point: A 2-years study with the mouse is also available, only no NOAEL was obtained. Deviations from the standard derivation:

The standard respiratory volume is 0.67 m³/kg bw for the mouse. AF4 = 2 for subchronic study, but also AF4 = 1 with the same justification as above. AF5 = 1 for starting with a NOAEL.

DNEL = 1400 x 1 x 1.49 x 0.67 x 1.4 /(1 x 2.5 x 5 x 1 x 1 x 1) = 157 mg/m³.

(If AF4 would be 2: DNEL = 78.3 mg/m³)

2. DNEL calculation, using the NOAELmonkey,oral,90d of 60 mg/kg/d as the starting point. Deviations from the standard derivation:

The standard respiratory volume is 0.19 m³/kg bw for monkeys. AF4 would be 2 for a subchronic study, but AF4 was set to 1, because of the known mode of action with the primary step of exceeding the urinary concentration of melamine in urine, which is independent of exposure duration, see also under the standard deviation above. An AF6 of at least 2 for a low quality of data base would have to be set in case this study would be the only available study. As it is not the only one, the AF6 was left at 1 to enable comparing the results with the other calculations.

DNEL = 60 x 1 x 5.26 x 0.67 x 1.4 /(1 x 2.5 x 5 x 1 x 1 x 1) = 23.6 mg/m³.

(If AF4 would be 2: DNEL = 11.8 mg/m³).

The peer reviewed NTP rat study is preferred as a basis for the DNEL-calculation, compared to the monkey study with few animals reported in a publication of Klimisch 4.

 

3. NOAELs of 400 mg/kg bw/d for maternal toxicity and developmental toxicity in an OECD414 study with rats were obtained. Deviations from the standard derivation:

AF4 would be 6 for subacute duration, but AF4 was set to 1, with the same justification as above. AF5 = 1 for starting with a NOAEL.

DNEL = 79.0 mg/m³.

(If AF4 would be 6: DNEL = 13.2 mg/m³)

4. NOAELs are at least 150 mg/kg bw/d for maternal toxicity and developmental toxicity in an OECD414 study with rabbits. Deviations from the standard derivation:

The standard respiratory volume sRV is 0.23 m³/kg bw. AF4 would be 6 for subacute duration, but AF4 was set to 1, with the same justification as above. AF6 = 1 for starting with a NOAEL.

DNEL = >48.9 mg/m³.

(If AF4 would be 6: DNEL = >8.16 mg/m³)

5. A NOAEL of ca. 65 mg/kg bw/d for toxicity was obtained from an OECD443 EOGRTS with rats. Deviations from the standard derivation:

AF4 would be 2 for subchronic duration, but AF4 was set to 1, with the same justification as above. AF6 = 1 for a NOAEL.

DNEL = ca. 12.8 mg/m³.

(If AF4 would be 2: DNEL = ca. 6.4 mg/m³)

6. A NOAEL of ca. 89 mg/lg bw/d for effects on the male reproductive system was obtained from an OECD443 EOGRTS with rats. Default assessment factors were used (AF4= 2 for subchronic duration; AF6=1 for NOAEL).

DNEL = ca. 8.79 mg/m³.

Conclusion: DNEL = 8.3 mg/m³, even if the conventional assessment factors for the exposure duration would have been used, considering the greater weight of the available long term study results, i.e. the standard derivation of the DNEL at the start of this explanatory window.

AF for dose response relationship:

3

Justification:

A steep dose response curve is present that allows a low assessment factor of 3 to be applied to convert the LOAEC to a NOAECrat,oral,2years.

AF for differences in duration of exposure:

1

Justification:

Default.

AF for interspecies differences (allometric scaling):

1

Justification:

Default.

AF for other interspecies differences:

2.5

Justification:

Default.

AF for intraspecies differences:

5

Justification:

Default.

AF for the quality of the whole database:

1

Justification:

Default.

AF for remaining uncertainties:

1

Justification:

Default.

Acute/short term exposure

Hazard assessment conclusion:

DNEL (Derived No Effect Level)

Value:

82.3 mg/m³

Most sensitive endpoint:

repeated dose toxicity

Route of original study:

Oral

DNEL related information

DNEL derivation method:

ECHA REACH Guidance

Overall assessment factor (AF):

12.5

Dose descriptor starting point:

NOAEL

Value:

417 mg/kg bw/day

Modified dose descriptor starting point:

NOAEC

Explanation for the modification of the dose descriptor starting point:

The starting point is the NOAELoral,rat,14d = 417 mg/kg bw/d.

Otherwise the same method as for the DNELlong-term was used.

DNEL = 417 x 1 x 2.63 x 0.67 x 1.4 /(1x 2.5 x 5 x 1 x 1 x 1) = 82.3 mg/m³

DNELacute / DNELlong-term = 82.3/8.3 = 417/42 = 9.9

AF for dose response relationship:

1

Justification:

Default

AF for interspecies differences (allometric scaling):

1

Justification:

Default

AF for other interspecies differences:

2.5

Justification:

Default

AF for intraspecies differences:

5

Justification:

Default

AF for the quality of the whole database:

1

Justification:

Default

AF for remaining uncertainties:

1

Justification:

Default

Local effects

Long term exposure

Hazard assessment conclusion:

no hazard identified

Acute/short term exposure

Hazard assessment conclusion:

no hazard identified

DNEL related information

Workers - Hazard via dermal route

Systemic effects

Long term exposure

Hazard assessment conclusion:

DNEL (Derived No Effect Level)

Value:

11.8 mg/kg bw/day

Most sensitive endpoint:

repeated dose toxicity

Route of original study:

Oral

DNEL related information

DNEL derivation method:

ECHA REACH Guidance

Overall assessment factor (AF):

150

Dose descriptor starting point:

LOAEL

Value:

126 mg/kg bw/day

Modified dose descriptor starting point:

LOAEL

Explanation for the modification of the dose descriptor starting point:

- The dose descriptor LOAELrat,oral,2year = 126 mg/kg bw/day of the key NTP-study was selected, as for the inhalation exposure.

- Bioavailability: From toxicokinetic studies it is known that melamine is absorbed rapidly and almost completely in rats after oral dosing. The default oral absorption rate of 100 % was used. Only a low dermal absorption is estimated for the rat and humans. Considering the statement in Chapter R.7.12 in the Guidance on Information Requirement and CSA, and the low n-octanol/water partition coefficient logPow = -1.22, it seems justified to calculate with a dermal absorption of 10 %. Correction factor 10.

- The exposure conditions for the orally dosed rats were 7d/w, 24h/d compared to the dermal exposure of workers of 5d/w, 8h/d. Only, the rats eat during two main periods per day, therefore the exposure correction for the hours per day was not used. Correction factor = 1.4.

- Assessment factors (AF1 to AF6) were selected according to Table R.8-6 of the Guidance on Information Requirement and CSA. AF1 = Allometric scaling; AF2 = Interspecies, remaining difference; AF3 = Intraspecies; AF4 = Exposure duration; AF5 = Dose-response relation; AF6 = Quality data base. Default values were used, except for AF5: A steep dose response curve, confirmed by EFSA (2010) page 103, is present that allows a low assessment factor AF5 of 3 to be applied to convert the LOAEL to a surrogate NOAELrat,oral,2years.

 

DNEL = 126 x 10 x 1.4 /(4 x 2.5 x 5 x 1 x 3 x 1) = 11.8 mg/kg bw/d

 

Alternative DNEL calculation if the NOAELmonkey,oral,90d of 60 mg/kg bw/d is taking as a starting point: NOAELmonkey,dermal = 60 x10 = 600 mg/kg bw/d. The AF1 for allometric scaling would be only 2 for monkeys, on the other hand the AF4 for duration would increase to 2 because of the study exposure duration of the monkeys, so that a higher DNEL results from the monkey study, compared to the rat study.

DNEL = 60 x 10 x 1.4 /(2 x 2.5 x 5 x 2 x 1 x 1) = 16.8 mg/kg bw/d

.

The peer reviewed NTP rat study is preferred as a basis for the DNEL-calculation, compared to the monkey study with few animals reported in a publication of Klimisch 4.

Conclusion: DNEL = 11.8 mg/kg bw/d

AF for dose response relationship:

3

Justification:

Default

AF for differences in duration of exposure:

1

Justification:

Default

AF for interspecies differences (allometric scaling):

4

Justification:

Default

AF for other interspecies differences:

2.5

Justification:

Default

AF for intraspecies differences:

5

Justification:

Default

AF for the quality of the whole database:

1

Justification:

Default

AF for remaining uncertainties:

1

Justification:

Default

Acute/short term exposure

Hazard assessment conclusion:

DNEL (Derived No Effect Level)

Value:

117 mg/kg bw/day

Most sensitive endpoint:

repeated dose toxicity

Route of original study:

Oral

DNEL related information

DNEL derivation method:

ECHA REACH Guidance

Overall assessment factor (AF):

50

Dose descriptor starting point:

NOAEL

Value:

417 mg/kg bw/day

Modified dose descriptor starting point:

NOAEL

Explanation for the modification of the dose descriptor starting point:

The starting point is the NOAELoral,rat,14d = 417 mg/kg bw/d.

Otherwise the same method as for the DNELlong-term was used.

DNEL = 417 x 10 x 1.4 /(4 x 2.5 x 5 x 1 x 1 x 1) = 117 mg/kg bw/d

DNELacute / DNELlong-term = 117/11.8 = 417/42 = 9.9

AF for dose response relationship:

1

Justification:

Default

AF for interspecies differences (allometric scaling):

4

Justification:

Default

AF for other interspecies differences:

2.5

Justification:

Default

AF for intraspecies differences:

5

Justification:

Default

AF for the quality of the whole database:

1

Justification:

Default

AF for remaining uncertainties:

1

Justification:

Default

Local effects

Long term exposure

Hazard assessment conclusion:

no hazard identified

Acute/short term exposure

Hazard assessment conclusion:

no hazard identified

Workers - Hazard for the eyes

Local effects

Hazard assessment conclusion:

no hazard identified

General Population - Hazard via inhalation route

Systemic effects

Long term exposure

Hazard assessment conclusion:

DNEL (Derived No Effect Level)

Value:

1.5 mg/m³

Most sensitive endpoint:

repeated dose toxicity

Route of original study:

Oral

DNEL related information

DNEL derivation method:

ECHA REACH Guidance

Overall assessment factor (AF):

75

Dose descriptor starting point:

LOAEL

Value:

126 mg/kg bw/day

Modified dose descriptor starting point:

LOAEC

Explanation for the modification of the dose descriptor starting point:

Analogue calculation as for workers, but observing:

- Bioavailablilty as for workers: Correction factor = 1.

- The standard respiratory volume sRV for rat is 1.15 m³/kg bw for an 24 hour exposure. Conversion factor = 0.87.

- No adjustment for an increased respiratory volume: Correction factor = 1.

- Exposure conditions: Correction factor = 1.

- Intraspecies assessment factor AF3 = 10.
No extra AF for children was applied, as there is no indication that children are more sensitive than adults, and as the intraspecies AF3 covers possible differences anyway. Moreover there are considerations (Cohen 1995, Chapter 7.7) that humans might be less sensitive to the formation of bladder stone than rats, due to anatomical differences. This could reduce the AF2 for interspecies remaining differences.

DNEL = 126 x 1 x 0.87 x 1 x 1 /(1 x 2.5 x 10 x 1 x 3 x 1) = 1.5 mg/m³

AF for dose response relationship:

3

Justification:

A steep dose response curve is present that allows a low assessment factor AF5 of 3 to be applied to convert the LOAEL to a surrogate NOAELrat,oral,2years.

AF for differences in duration of exposure:

1

Justification:

Default.

AF for interspecies differences (allometric scaling):

1

Justification:

Default

AF for other interspecies differences:

2.5

Justification:

Default

AF for intraspecies differences:

10

Justification:

Default

AF for the quality of the whole database:

1

Justification:

Default

AF for remaining uncertainties:

1

Justification:

Default

Acute/short term exposure

Hazard assessment conclusion:

no hazard identified

DNEL related information

Local effects

Long term exposure

Hazard assessment conclusion:

no hazard identified

Acute/short term exposure

Hazard assessment conclusion:

no hazard identified

DNEL related information

General Population - Hazard via dermal route

Systemic effects

Long term exposure

Hazard assessment conclusion:

DNEL (Derived No Effect Level)

Value:

4.2 mg/kg bw/day

Most sensitive endpoint:

repeated dose toxicity

Route of original study:

Oral

DNEL related information

DNEL derivation method:

ECHA REACH Guidance

Overall assessment factor (AF):

300

Dose descriptor starting point:

LOAEL

Value:

126 mg/kg bw/day

Explanation for the modification of the dose descriptor starting point:

Analogue calculation as for workers, but observing:

- Bioavailablity: Correction factor = 10.

- Exposure conditions: Correction factor = 1.

- Intraspecies assessment factor AF3 = 10.
No extra AF for children was applied, as there is no indication that children are more sensitive than adults, and as the intraspecies AF3 covers possible differences anyway. Moreover there are considerations (Cohen 1995, Chapter 7.7) that humans might be less sensitive to the formation of bladder stone than rats, due to anatomical differences. This could reduce the AF2 for interspecies remaining differences.

DNEL = 126 x 10 x 1 /(4 x 2.5 x 10 x 1 x 3 x 1) = 4.2 mg/kg bw/d

AF for dose response relationship:

3

Justification:

A steep dose response curve is present that allows a low assessment factor AF5 of 3 to be applied to convert the LOAEL to a surrogate NOAELrat,oral,2years

AF for differences in duration of exposure:

1

Justification:

Default

AF for interspecies differences (allometric scaling):

4

Justification:

Default

AF for other interspecies differences:

2.5

Justification:

Default

AF for intraspecies differences:

10

Justification:

Default

AF for the quality of the whole database:

1

Justification:

Default

AF for remaining uncertainties:

1

Justification:

Default

Acute/short term exposure

Hazard assessment conclusion:

no hazard identified

DNEL related information

Local effects

Long term exposure

Hazard assessment conclusion:

no hazard identified

Acute/short term exposure

Hazard assessment conclusion:

no hazard identified

General Population - Hazard via oral route

Systemic effects

Long term exposure

Hazard assessment conclusion:

DNEL (Derived No Effect Level)

Value:

0.42 mg/kg bw/day

Most sensitive endpoint:

repeated dose toxicity

Route of original study:

Oral

DNEL related information

DNEL derivation method:

ECHA REACH Guidance

Overall assessment factor (AF):

300

Dose descriptor starting point:

LOAEL

Value:

126 mg/kg bw/day

Explanation for the modification of the dose descriptor starting point:

Analogue calculation as for dermal exposure of the general public, but observing:

- Bioavailablity: Correction factor = 1.

- Exposure conditions: Correction factor = 1.

- Intraspecies assessment factor AF3 = 10.
No extra AF for children was applied, as there is no indication that children are more sensitive than adults, and as the intraspecies AF3 covers possible differences anyway. Moreover there are considerations (Cohen 1995, Chapter 7.7) that humans might be less sensitive to the formation of bladder stone than rats, due to anatomical differences. This could reduce the AF2 for interspecies remaining differences.

DNEL = 126 x 1 x 1 /(4 x 2.5 x 10 x 1 x 3 x 1) = 0.42 mg/kg bw/d

AF for dose response relationship:

3

Justification:

A steep dose response curve is present that allows a low assessment factor AF5 of 3 to be applied to convert the LOAEL to a surrogate NOAELrat,oral,2years.

AF for differences in duration of exposure:

1

Justification:

Default

AF for interspecies differences (allometric scaling):

4

Justification:

Default

AF for other interspecies differences:

2.5

Justification:

Default

AF for intraspecies differences:

10

Justification:

Default

AF for the quality of the whole database:

1

Justification:

Default

AF for remaining uncertainties:

1

Justification:

Default

Acute/short term exposure

Hazard assessment conclusion:

no hazard identified

DNEL related information

General Population - Hazard for the eyes

Local effects

Hazard assessment conclusion:

no hazard identified

Additional information - General Population

Reasons why the lowering of the TDI in 2008 - 2010 by WHO, EFSA and FDA is not followed for the derivation of the DNELs:

Tolerable Daily Intakes (TDI) are reported by some health and food safety authorities, see Section 13. In this context it should be considered that melamine under REACH is not intended to be added to food or feed. Adulterated food or feed, illegally produced, is not relevant for the registration under REACH. Also the interaction with other triazines in adulterated food and the uncertainties in the human exposure to adulterated food, which are factors in deriving the TDI by the health authorities, are not to be taken into account in the derivation of the DNEL under REACH. Therefore, the TDIs are presented here only for reasons of completeness and have not to be adopted for REACH-purposes.

 

In 2009 WHO established a TDI of 0.2 mg/kg bw/day. The Bench Mark Dose concept was applied to the subchronic data of NTP 1983 and Tyl 1982/1983, to give a lowest BMDL10 of 35 mg/kg bw/day for the NTP-results and a BMDL10 of 384 mg/kg bw/day for the Tyl-results. The more conservative value was adopted for further calculations. WHO: "The potential increased sensitivity of infants and for data uncertainties, an additional 2-fold factor was applied to the default 100-fold uncertainty factor. The uncertainty in data stems from studies demonstrating that formalin preparation of tissues may lead to significant loss of crystals or stones, which could then result in underestimation of stone formation and thus the toxicity of the dose tested."

Also interactions of melamine and cyanuric acid were mentioned as a source of uncertainties, covered by the additional assessment factor. The WHO assessment seems to be on the rather exaggerated worst case side, as it neglects the higher BMDL10 of the Tyl study, where the problem of the possibly lost bladder stones was already recognized. Nevertheless WHO inserted an unjustified additional safety factor.

 

EFSA 2008, see Chapter 13, has issued a TDIoral of 0.5 mg/kg bw/day. EFSA: "The primary target organ for melamine toxicity (in humans) is the kidney. There is uncertainty with respect to the time scale for the development of kidney damage. Thus, EFSA applied a tolerable daily intake (TDI) of 0.5 mg/kg body weight in considering possible health effects which might occur with repeated consumption of melamine contaminated products over a relatively short period." The TDI is not much different from the DNEL of 0.42 mg/kg bw/day.

EFSA 2010 lowered the TDI for melamine to 0.2 mg/kg bw/day. EFSA used the same NTP study as WHO did, neglecting also the Tyl-study, but derived a still more conservative BMDL10 of 19 mg/kg bw/day than WHO.

 

FDA in 2008 changed the TDI from 0.63 mg/kg bw/day to 0.063 mg/kg bw/day. Two reasons were presented:

1) FDA: "The safety/risk assessment used a TDI (Tolerable Daily Intake) of 0.63 mg/kg bw as its starting point. This TDI was based on the results of a 13-week rat study of melamine alone. Melamine and its analogues – cyanuric acid, ammelide, and ammeline – were assumed to be of equal potency. The safety/risk assessment, however, took into account studies indicating that increased toxicity results from combined exposure to melamine and cyanuric acid, which raises a high degree of uncertainty with regard to the determination of safety/risk. For foods other than infant formula, the safety/risk assessment applied an additional 10-fold safety factor to compensate for the uncertainties regarding combined exposure to more than one of the melamine-related compounds. In applying this safety factor and using the TDI, the safety/risk assessment concluded that 'levels of melamine and its analogues below 2.5 ppm in foods other than infant formula do not raise public health concerns.' For infant formula, the safety/risk assessment determined that it could not, based on current information, apply such a safety factor to establish a level of melamine and its analogues in infant formula that would not raise public health concerns. This was based on several factors specific to infant formula contaminated with more than one melamine analogue, such as the product represents the totality of caloric exposure for most infants, exposure is chronic over months, and the persons ingesting the products are infants and toddlers whose renal systems may not be fully developed."

As the melamine to be registered under REACH is of high purity, the considerations of interactions of melamine with other triazines in adulterated food is not relevant.

2) FDA: "Infants may be more sensitive than adults to exposures because, for example, infant formula is the sole source of nutrition, exposure continues for up to 12 months, and renal function may be more immature compared to adults. This raises a high degree of uncertainty with regard to the determination of safety/risk. Given these conditions, FDA has applied an additional 10-fold safety factor, yielding a combined safety factor of 1000-fold, to compensate for these uncertainties. This results in a TDI/10 of 0.063 mg melamine/kg-bw/d." First, there is no indication that infants are more sensitive than adults, see above. From the Guidance on Information Requirement and CSA "A Default factor of 10 is sufficient to protect the larger part of the population, including e.g. children and the elderly".
Second, the argument that infants could consume a higher amount of melamine in adulterated food is not relevant for the derivation of the TDI, because the higher exposure should be taken into account in the exposure assessment and not in the hazard assessment.

Conclusion: No justification is left to lower the calculated DNELs by increasing an assessment factor or by adding an additional one.