Use of a Two Tiered Degree-Hour Measure for the Management of Investigational Medicinal

Product Over-Temperature Excursions Using Mean Kinetic Temperature (MKT) | IVT

By
Shelly J. Asmussen, Ph.D.
David G. Stroz, M.S.
David LeBlond
Dennis A. Stephens, Ph.D.
Mar 14, 2014 1:03 pm PDT
Peer Reviewed: Stability Testing

Abstract
This manuscript translates the USP Pharmacopeial Mean Kinetic Temperature (MKT) into an
approach that can be applied to temperature excursion management. It also describes and
justifies the use of a two-tiered model that is a conservative estimate of product stability, and it
is simplified through the creation of standardized evaluation tables that the end user can apply in
a simplified manner. Single or multiple excursions can be evaluated by a simple arithmetic
analysis (an example of which can be found at the end of the paper). Tables are provided for
both USP Controlled Room and Controlled Cold Temperature excursion evaluations (CRT and
CCT respectively).
Introduction
In clinical studies, the burden of ensuring that investigational medical products (IMPs),
internally developed or purchased, are properly controlled during the study falls upon the study
originator. Ensuring proper control of the IMP during the clinical trial may include evaluating
potential damage resulting from temperature excursions above the labeled storage conditions
that may occur while the material is in transit or while it is stored at the clinical site. When the
study originator is not the innovator, the amount of information that is available to the study
originator for evaluation of the temperature excursion is limited, providing significant
challenges to provide a timely assessment of the impact of the excursion. The objective of this
proposed process is to provide a standardized approach that can be used to assess and
characterize the magnitude of potential product impact (e.g., degradation). The approach
considers not only the short-term impact but also the impact that may have occurred to
purchased IMPs throughout the shelf-life by applying a Product Life Span (PLS) concept. This
approach considers the impact from the date that the study originator takes control of a
purchased IMP to the expiry date of the IMP by applying the principles of Mean Kinetic
Temperature (MKT). This paper presents a simplified strategy on evaluating temperature
excursions using MKT.
Materials and Methods
Basis for the Mean Kinetic Temperature (MKT) Requirement
For materials with N hours of product's shelf-life, the USP (1–3) requires that
[Equation 1]
Where MKTLimit is the mean kinetic temperature upper limit (e.g., 298.15°K, which is equivalent
to 25°C), ΔH is the heat of activation for the instability mechanism (generally 42 to 125 kJ/mol,
although USP <1160> recommends 83.144 kJ/mol unless a more accurate experimentally
determined value is available), R is the universal gas constant
(0.0083144 kJ mol-1 K-1), and ni is the number of hours spent at the storage temperature

Ti (°K) for all k temperatures (Note, ).

Any IMP labeled as USP Controlled Room Temperature (CRT) (4) or Controlled Cold
Temperature (CCT) (5) must be in compliance with the limits established by applying Equation
1 during the entire supply chain (6–8). The effects of temperature excursions on pharmaceutical
products using Equation 1 have been documented (9–11).
This requirement associated with MKT assumes that the mechanisms driving product instability
follow the Arrhenius model and do not change over the range of possible temperature
excursions. To illustrate the rationale behind equation (1), consider an instability mechanism
that follows pseudo zero order kinetics (the rate of degradation is constant over time). Assuming
a single excursion of nH hours at temperature TH K with storage at the remaining N–nH hours at
the material's labeled storage temperature (T0 K), T0 is the nominal storage temperature for the
material (CRT ~ 22– 23°C and CCT ~ 5°C). Equation 1 may be rearranged to
[Equation 2]
Ne-ΔH/R·MKTLimit ≥ (N – nH)e-ΔH/R·T0 + nHe-ΔH/R·TH

If ΔPt,T is the amount of degradation product generated from a pseudo-zero order kinetic process
during storage for t hours at temperature T, then according to the Arrhenius model, Equation 2
corresponds to
[Equation 3]
ΔPN,MKTLimit ≥ ΔPN–nH,T0 + ΔPnH,TH

Equation 3 shows that the USP requirement (Equation 1) corresponds to a restriction that the
total degradation product generated over the product life, including the effects of temperature
excursions, be no greater than would occur if the product was stored isothermally at MKTLimit.
Thus Equation 1 implies an “excursion cost” that is based on a measure of the additional
amount of a degradation product formed during an excursion above the amount that would be
generated during storage at MKTLimit.
This same logic may be applied to an excursion at another temperature, TL. As a result, we can
establish a relationship between “excursion costs” for products stored at
temperatures TL and TH (TL < TH) for nLand nH hours respectively. These “excursion costs” can
be used to estimate a combination of degree-hour excursion costs for each temperature. Thus,
the relationship between TL and TH will justify the use of a reference temperature (TH) when
establishing a conservative degree-hour cost limit for excursions at some temperature (TL) below
the defined reference temperature. The sections below formally derive the technical arguments
necessary to make such a justification.
The above arguments are based on a pseudo-zero order kinetic process. However, many
processes relevant to pharmaceutical stability are in fact first-order kinetic processes. When the
degradation of a drug is very low (e.g., less than 2%), as is typical for many pharmaceuticals,
both zero and first order processes can be treated as following zero order kinetics (12). This
assumption simplifies the model and permits the arguments discussed previously to be applied
equally to first order processes.
Relating Excursion Costs at Different Excursion Temperatures
Typically, materials are not stored at the MKTLimit (top of the stated storage range) but at a lower
temperature that is within material's labeled storage temperature range (defined as T0, which
may be slightly less than the MKTLimit.) One can measure the “excursion cost” as the additional
degradation product formed during an excursion of nL hours at TL >MKTLimit > T0 as
[Equation 4]
ΔPnL,TL – ΔPNL,T = nL(Ae-ΔH/RTL – (Ae-ΔH/RT0)
Since excursions may occur at multiple elevated temperatures, a reference temperature should
be defined such that TH(40°C for a CRT product) > TL > MKTLimit > T0, which permits the
development of a practical excursion control procedure. In this case, the corresponding
"excursion cost" would be defined as
[Equation 5]
ΔPnH,TH – ΔP(Ae-ΔH/RTH – (Ae-ΔH/RT0)
In order to "translate" an excursion cost from TL to TH, we would equate Equations 4 and 5 to
force the amount of extra degradation product to be the same at both temperatures. Rearranging,
we obtain the following relationship between nH and nL
[Equation 6]

Justification of the Degree-Hour Excursion Cost Measure

Calculation of the excursion cost measure based on degradation product levels using Equations
4 and 5 is impractical using a hand calculator. MKT calculations often must be made in
situations where specialized software is not readily available. In this paper, we propose a
conservative linear approximation to the exact degradation cost metric that is simple and
practical to use. It is based on a degree-hour cost metric that increases linearly with excursion
temperature increase. In Figure 1, below, and the associated discussion, we prove that this linear
approximation is conservative in that it effectively assumes that more degradation occurs than
would be estimated based on the Arrhenius model.
We define the degree-hour cost approximation to Equation 4 for an excursion of nL hours
at TL as
[Equation 7]
CL = nL (TL – T0)

And the cost approximation to Equation 5 as

[Equation 8]
CH = nH (TH – T0)
Where the nH and nL in Equations 7 and 8, respectively, obey relationship Equation 6.
Since excursions may occur at various temperatures, it would be convenient to establish a
maximum allowable total cost at some reference temperature (TH) and apply to this the costs
calculated at each actual excursion temperature, TL(TL ≤ TH) using Equation 7. The above
provides a theoretical justification for a degradation product based excursion cost measure.
Below we use this theory, and some mathematical arguments, to extend that justification to the
degree-hour based cost measure. In particular, we show that extrapolating a degree-hour cost
from the actual temperature of excursion (TL) to the reference temperature (TH ≥ TL) is a
conservative approach.
Such a translation could be justified if it can be shown that it is conservative in the sense that a
cost calculated at a lower temperature TL is greater by Equation 7 than the corresponding cost
calculated at the higher reference temperature by Equation 8. Thus, we must show that
[Equation 9]

Substituting Equations 7 and 8 into Equation 9, with rearrangement, we must demonstrate that
[Equation 10]

Given typical values for ΔH, R, and temperatures, it can be shown that exp -ΔH/RT is a convex
function with respect to T. Reference to Figure 1 below and Equation 6 illustrate that Equation
11 and, therefore, Equation 9 are true, thereby justifying the use of a reference temperature for
establishing a conservative degree-hour cost limit for excursions that occur at lower
temperatures.
Figure 1: Graphical Representation of e-ΔH/RT as a Function of Temperature.
[Equation 11]

Where, as shown in Figure 1,

Y = e-ΔH/RTL – e-ΔH/RT0
Z = e-ΔH/RTH – e-ΔH/RT0
W = TL – T0
X = TL – T0
Figure 2: Calculation of the Maximum Degree Hour Cost at a R

eference Temperature.

Assume that the Product Life Span (PLS) for a package of clinical investigational medicinal
product materials is N hours if stored at a temperature of T0. For a single excursion of nH hours
at an excursion temperature of TH, we can rearrange Equation 2 to obtain
[Equation 12]

The right hand side of Equation 12 thus gives the maximum allowable hours of excursion (nH) at
the reference temperature TH. The corresponding maximum allowable degree-hour excursion
cost (Equation 8) that can be applied at any temperature TL ≤ TH or lower would be
[Equation 13]
Cmax(TH) = nH (TH – T0)

The total of all excursion costs, calculated according to Equation 7, can never exceed this
maximum upper limit over the life of the product.
The Need for Two Reference Temperatures
As indicated above in Equation 9, the procedure as just outlined is conservative in that costs
calculated using Equation 7 at temperatures below TH will overestimate the true degree-hour
excursion cost. The degree of conservatism is greater as the temperature difference
between TH and TL increases. Therefore, it makes sense to establish a double-tiered approach
using two reference temperatures to avoid “overspending” the justified “budget” for a few
degrees of over-temperature excursion that typically occurs. For example, in the case of room
temperature storage, we chose the reference temperatures 30°C and 40°C to
be TL and TH, respectively, which are also key compendial and stability set-points. Thus,
excursions occurring in the range 26–30°C will apply against the 30°C maximum
limit TL calculated using Equations 12 and 13. Excursions occurring in the range 31–40°C will
apply against the 40°C maximum limit THcalculated using Equations 12 and 13. Similarly, for
cold storage, we chose reference temperatures 15°C and 25°C for excursions in the ranges of 9–
15°C, and 16–25°C, respectively.
In order to facilitate accounting for the effects of all excursions in both high and low
temperature ranges, a factor can be derived for converting a TH (e.g., 40 or 25°C) excursion cost
to a TL (e.g., 30 or 15°C) excursion cost. Such a factor may be derived using Equation 6 by
expressing T0, TL, and TH (degrees, K). Then, the conversion factor (called the temperature
burden ratio or TBR) will be
[Equation 14]

It is interesting to note that this ratio does not depend on the total number of hours (N) or on
MKT. The temperature burden ratio (TBR) can be multiplied by the maximum number of
storage hours at TH (e.g., 40°C or 25°C) to obtain the corresponding maximum allowable
number of storage hours at TL (e.g., 30°C or 15°C).
An associated factor, the Degree Hour Conversion Ratio (DHCR), for converting the maximum
number of degree-hours at TH to the corresponding number at TL can be obtained by applying
Equations 7 and 8 to Equation 14:
[Equation 15]

DHCR need only be calculated once for given values

of T0 , TL , TH , and ΔH. DHCR is given in Table I for two scenarios using the USP <659>
default ΔH of 83.144 kJ/mol.
Table I: Degree-Hour Conversion Ratio (DHCR) for Two Scenarios.

Controlled Room
Parameter Controlled Cold Storage
Temperature

T0 (°C) 22–23 (nominally) 5 (nominally)

MKTLimit (°C) 25 8

TL (°C) 30 15

TH (°C) 40 25

ΔH(kJ/mol) 83.144 83.144

DHCR 1.798 1.979

Note: For the purposes of the mathematical treatment in this paper, T0 was taken as the upper
boundary of the room temperature storage range, a worst-case that is not the typical storage
temperature (further discussion found in the Results section). For example, the USP CRT
storage range is stated to be 20–25°C, so a target for controlled storage might be 22–23°C.
However, USP also allows for cooler storage unless specifically limited by the labeling.
Therefore, a lower target storage temperature is acceptable. In Europe, the European Medicines
Agency (EMA) labeling guidance is similar to USP (i.e., storage not higher than 25°C actual
storage down to any temperature is allowed unless restricted). Since the US and most of Europe
are in International Conference for Harmonisation (ICH) Climatic Zone II regions (MKT not
more than 25°C), the methodology supplied in this paper may be applied to both US and
European IMPs.
Results
In order to use the two-tiered degree-hour cost approach for managing excursions of clinical
IMPs, it is first necessary to establish a table giving the maximum allowable degree-hours at the
maximum temperature of the lower temperature tier (i.e., TL in Table I) for various product
PLSs (N) using Equations 12 and 13. The maximum allowable degree-hours at the maximum
temperature of the upper temperature tier (i.e., TH in Table I) are obtained by the use of
Equations 14 and 15. The use of Equations 14 and 15 is simplified by noting that the DHCR for
this scenario is 1.798 for all PLSs. Using the controlled room temperature (CRT) scenario of
Table I, an example of such a table for a CRT product with a PSL from 15–48 months is given
in Table II.
Table II: Allowable Degree-Hour Costs for Two Tiered Controlled Room Temperature
(CRT) Storage Excursions for a Product with a PSL of 15–48 Months.

Product Life Span

First Tier Excursions (Temp Second Tier Excursions (Temp
(From Possession
Range 26–30°C) Range 31–40°C)
until Expiry Date)

# Degree- # Degree-Hours40
# Hours at
# Hours at Hours30 Maximum (15ºC
40°C
30°C [From Maximum (5ºC above 25°C ×
Months Days Hours [Obtained
Equation above 25°C × hours) [Obtained
using TBR and
12] hours) [From using DHCR and
Equation 14]
Eqn 13]a Equation 15]a

15 450 10800 827 4135 153 2299

16 480 11520 882 4411 164 2453

17 510 12240 937 4686 174 2606

18 540 12960 992 4962 184 2759

19 570 13680 1048 5238 194 2913

20 600 14400 1103 5513 204 3066

21 630 15120 1158 5789 215 3219

22 660 15840 1213 6065 225 3373

23 690 16560 1268 6340 235 3526

24 720 17280 1323 6616 245 3679

25 750 18000 1378 6892 255 3832

26 780 18720 1434 7168 266 3986

27 810 19440 1489 7443 276 4139

28 840 20160 1544 7719 286 4292

29 870 20880 1599 7995 296 4446

30 900 21600 1654 8270 307 4599

31 930 22320 1709 8546 317 4752

32 960 23040 1764 8822 327 4905

33 990 23760 1819 9097 337 5059

34 1020 24480 1875 9373 347 5212

35 1050 25200 1930 9649 358 5365

36 1080 25920 1985 9924 368 5519

37 1110 26640 2040 10200 378 5672

38 1140 27360 2095 10476 388 5825

39 1170 28080 2150 10751 399 5979

40 1200 28800 2205 11027 409 6132

41 1230 29520 2261 11303 419 6285

42 1260 30240 2316 11578 429 6438

 43 1290 30960 2371 11854 439 6592

44 1320 31680 2426 12130 450 6745

45 1350 32400 2481 12405 460 6898

46 1380 33120 2536 12681 470 7052

47 1410 33840 2591 12957 480 7205

48 1440 34560 2646 13232 491 7358

a.
An MKTLimit of 25.49°C and T0 of 25.00°C was used to calculate these values, as explained in
the text.
For an example to see how Table I and Table II would be applied, see the Discussion section
below.
Similarly, a set of lookup values for the Controlled Cold Temperature (CCT) storage scenario
found in Table I is given in Table III. The use of Equations 14 and 15 for CCT storage is
simplified by noting that the DHCR for this scenario is 1.979 for all PLSs.
Table III: Allowable Degree-Hour Costs for Two-Tiered Controlled Cold Temperature
(CCT) Storage Excursions for a Product with a PSL of 15–48 Months.

Product Life Span

First Tier Excursions (Temp Second Tier Excursions (Temp
(From Possession
Range 9 to 15°C) Range 16 to 25°C)
until Expiry Date)

# Degree- # Degree-Hours25
# Hours at
# Hours at Hours15 Maximum (17ºC
25°C
15°C [From Maximum (7°C above 8°C × hours)
Months Days Hours [Obtained
Equation above 8°C × [Obtained using
using TBR and
12] hours) [From DHCR and
Equation 14]
Eqn 13]a Equation 15]a

15 450 10800 502 3516 104 1776

16 480 11520 536 3750 111 1894

17 510 12240 569 3984 118 2013

18 540 12960 603 4219 125 2131

19 570 13680 636 4453 132 2250

20 600 14400 670 4687 139 2368

21 630 15120 703 4922 146 2486

22 660 15840 737 5156 153 2605

23 690 16560 770 5391 160 2723

24 720 17280 804 5625 167 2842

25 750 18000 837 5859 174 2960

26 780 18720 871 6094 181 3078

27 810 19440 904 6328 188 3197

28 840 20160 937 6562 195 3315

29 870 20880 971 6797 202 3434

30 900 21600 1004 7031 209 3552

31 930 22320 1038 7266 216 3670

32 960 23040 1071 7500 223 3789

33 990 23760 1105 7734 230 3907

34 1020 24480 1138 7969 237 4026

35 1050 25200 1172 8203 244 4144

36 1080 25920 1205 8437 251 4263

37 1110 26640 1239 8672 258 4381

38 1140 27360 1272 8906 265 4499

39 1170 28080 1306 9141 272 4618

40 1200 28800 1339 9375 279 4736

41 1230 29520 1373 9609 286 4855

42 1260 30240 1406 9844 293 4973

 43 1290 30960 1440 10078 299 5091

44 1320 31680 1473 10312 306 5210

45 1350 32400 1507 10547 313 5328

46 1380 33120 1540 10781 320 5447

47 1410 33840 1574 11015 327 5565

48 1440 34560 1607 11250 334 5683

a.
An MKTLimit of 8.49°C and T0 of 8.00°C was used to calculate these values, as explained in the
text.
Table II and Table II are provided as a convenience for the stated situations. Such lookup tables
can be used by those responsible for making rapid decisions about excursion impacts. Note: If
the date of manufacture is actually known, then the PLS can be taken as the fuller time span,
namely from date of manufacture to the expiry date. It is a good manufacturing practice (GMP)
expectation that if the material was pre-burdened with significant temperature damage, some
compensating adjustment would be made, such as shortening of the expiry date or deeming the
material not suitable for distribution. The MKTLimit of 25.49°C and 8.49°C were used for the
calculations because they are well within the allowed ICH long-term stability chamber
tolerances and they are within the tolerances of the instruments used to measure chambers (1).
In addition, as shown in Figure 1, using a small difference between MKTLimit and T0 coupled with
setting T0 at 25.00°C or 8.00°C represents a conservative approach that is reasonable for
excursions that occur with material stored at temperatures with lower T0 values.
When an excursion occurs, the degree-day cost is calculated using Equation 7. If the excursion
occurs in the upper tier (e.g. 31–40°C or 16–25°C respectively), the degree day cost is corrected
to the lower tier reference temperature (30°C) by multiplying it times the DHCR.
The total of all excursion costs for a given IMP should not exceed the maximum allowable
degree-day limit at the maximum temperature of the lower tier (15 or 30°C, respectively)
without further justification (e.g., if a larger excursion allowance was obtained directly from the
supplier).
Discussion
With the allowable degree-hour cost for CRT (Table II) and CCT (Table III) calculated, IMP
temperature excursions can be quickly evaluated. For example, if an IMP is labeled USP CRT,
it has a Product Life Span (PLS) of 15 months (date of possession to expiry), and an excursion
is reported whereupon the material was stored above 25°C for 10 days, the peak temperature
being 28°C. This is the first excursion. The over-temperature event was in the First-Tier
excursion range of 26–30°C; thus, the allowable degree-hrs 30 is 4,135 (from Table II). The
excursion itself used up [3°C over-temperature][10 days][24 hrs/day] = 720 degree-hrs 30. Note:
To simplify the totaling of actual degree-hours 30 over-temperature on an hour-by-hour basis, the
usual practice often is to take the maximum over-temperature amount recorded during the
excursion and multiply by the hours of the excursion duration as the worst-case; this is the
 approach taken in this example. Since 720 is less than 4,135, this material is acceptable. The
amount of degree-hours30 coverage remaining is 4,135 – 720 = 3,415 degree-hrs 30, and any
future over-temperature assessment would be made against this remaining balance. At a later
date, a second excursion takes place on the same material for 13 hours at 35°C. This excursion
is at a higher temperature and fits in the Second Tier excursion range of 31–40°C. One would
calculate the degree-hrs40 for the excursion as follows: [10°C over-exposure][13 hours] = 130
degree-hrs40. Next, this value is multiplied by 1.798 to down-convert it to a First Tier degree-
hrs30 equivalent basis; thus, [130 degree-hrs 40][1.798] = 234 degree-hrs30. The final step is to add
the impact of this excursion to the previous. [720 degree-hrs 30 excursion #1] + [234 degree-
hrs30 current excursion] = 954 degree-hrs30 total damage “cost” to date. Since 954 is less than the
total degree-hrs30 allowable of 4,135, the material is still acceptable. Furthermore, the running
balance is 3,415 – 234 = 3,181 degree-hrs 30. Expressed as a percentage, the amount of over-
temperature coverage remaining is [3,181 / 4,135] x 100 = 76.9%. In other words, only 23.1%
of the starting over-temperature allowance has been used up to date.
Conclusion
By applying the principles of USP Mean Kinetic Temperature (MKT) and knowledge of a
commercial IMPs product life span (PLS), the maximum allowable degree-hours can be
calculated, allowing the impact of the temperature excursion to be readily evaluated. This
approach is applicable to drug products that consist of small molecules and have various label
storage restrictions (e.g., CRT, CCT) associated with them.
Acknowledgment
This study was funded by AbbVie (formerly Abbott). AbbVie participated in the study design,
research, data collection, analysis and interpretation of data, as well as writing, reviewing, and
approving the publication. David LeBlond and David Stroz are former employees of Abbott
Laboratories, currently retired, and may own AbbVie stock/options. Shelly Asmussen and
Dennis Stephens are AbbVie employees and may own AbbVie stock/options.
References
1. International Conference on Harmonization (ICH) Q1A (R2): Stability Testing of New Drug
Substances and Products (Second Revision).
2. USP 36-NF 31 General Information <1079>, "Good Storage and Distribution Practices for Drug
Products," 3.
3. USP 36-NF 31 General Information <1160>, "Pharmaceutical Calculations in Prescription
Compounding", 9-11.
4. USP 36-NF 31 General Notice 10.30.60, "Controlled Room Temperature"
5. USP 36-NF 31 General Notice 10.30.40, "Controlled Cold Temperature"
6. USP 36-NF 31 General Test <659>, "Packaging and Storage Requirements," 2-3.
7. USP 36-NF 31 General Information <1191>, "Stability Considerations in Dispensing Practice,"
3.
8. USP 36-NF 31 General Information <1118>, "Monitoring Devices–Time, Temperature, and
Humidity", 2.
9. Bailey L and Medwick T, "Mean Kinetic Temperature – A concept for Storage of
Pharmaceuticals," Pharmacopeial Forum, 19 (5), 6163-6166, 1993.
10. Kommanaboyina B and Rhodes, C, "Effects of Temperature Excursions on Mean Kinetic
Temperature and Shelf Life," Drug Development and Industrial Pharmacy, 25 (12) 1301-1306,
1999.
11. Okeke C, Bailey L, Medwick T, and Grady L, "Temperature fluctuations during mail order
shipment of pharmaceutical articles using mean kinetic temperature approach," Pharmacopeial
Forum 23(3) 4155-4182, 1997.
12. K.C. Waterman, R.C. Adami, "Accelerated aging: Prediction of chemical stability of
pharmaceuticals," Int. J. Pharm. 293, 101-125, 2005.
Appendix
AbbVie Terms used Throughout the Document.

Term Definition

Number of hours a material spends outside its label storage temperature

nH and nL
(excursion hours at temperature L and H); nL < nH.

The nominal storage temperature for the material (CRT ~ 22–23°C and
T0
CCT ~ 5°C)

MKTLimit Mean Kinetic Temperature Limit for product storage

TH and TL Temperature, L and H, associated with an excursion.

Temperature Burden Ratio, the factor relating the burden or impact between
TBR
two different temperatures

Degree-Hour Conversion Ratio, used to convert Tier 2 excursions to a Tier

DHCR
1 equivalent basis.

Product Life Span The span of time in months between when material was acquired and when
(PSL) it expires.