METHODS FOR PREPARING CRYSTALLINE RAPAMYCIN AND
FOR MEASURING CRYSTALLINITY OF RAPAMYCIN
COMPOUNDS USING DIFFERENTIAL SCANNING
CALORIMETRY
BACKGROUND OF THE INVENTION
Methods for preparing crystalline rapamycin compounds and measuring
crystallinity and particle quality of samples containing a rapamycin compound are
described.
Rapamycin (the Rapamune® drug) is an immunosuppressant derived from
nature, which has a novel mechanism of action. CCI-779 (rapamycin 42-ester with 3-
hydroxy-2-(hydroxymethyl)-2-methylpropionic acid) is an ester of rapamycin, which
has demonstrated significant inhibitory effects on tumor growth in both in vitro and in
vivo models.
Numerous routes to rapamycin compounds and purification thereof have been
described in the literature, some resulting in rapamycin compounds having acceptable
specifications required by regulatory agencies such as the US Food and Drug
Administration (FDA). However, samples of rapamycin compounds prepared using
these routes may contain crystals of varying quality.
What are needed in the art are methods for preparing crystalline rapamycin
and for measuring the crystallinity and particle quality of samples of rapamycin
compounds.
SUMMARY OF THE INVENTION
In one aspect, methods for measuring particle quality of a rapamycin
compound are provided.
In another aspect, methods for measuring crystallinity of a rapamycin
compound are provided.
In yet a further aspect, methods for preparing crystalline rapamycin are
provided.

Other aspects and advantages of the present invention are described further in
the following detailed description of the preferred embodiments thereof.
BRIEF DESCRIPTION OF THE FIGURES
Figure 1 provides peak temperatures obtained from DSC graphs for twenty-
five (25) CCI-779 samples as a function of particle category. A particle category 1
refers to crystalline CCI-779 samples having large particles; a particle category 2
refers to crystalline CCI-779 samples having small particles; a particle category 3
refers to semi-crystalline CCI-779 aggregates; and particle category 4 refers to non-
crystalline CCI-779.
Figures 2A and 2B provide DSC graphs noting peak temperatures for five (5)
rapamycin samples of varying crystallinity. In Figure 2A, the top plot corresponds to
a sample containing crystalline rapamycin; the middle plot corresponds to a sample
containing semi-crystalline rapamycin; and the lower plot corresponds to a sample
containing amorphous rapamycin. In Figure 2B, the top plot corresponds to a sample
containing crystalline rapamycin held at 25 to 60 °C for 2 months; the middle plot
corresponds to a sample containing crystalline rapamycin held at 25 to 60 °C for 4
months; and the lower plot corresponds to the sample containing crystalline
rapamycin identified in Figure 2A.
Figure 3 provides a graph illustrating the relationship between the degree of
crystallinity and thermal parameters including the heat of fusion (J/g), melting onset
temperature (° C), and peak temperature (° C) for six (6) CCI-779 samples. Shaded
triangles (A) illustrate the correlation of peak temperature and crystallinity; shaded
diamonds (♦) illustrate the correlation of heat of fusion and crystallinity; and shaded
squares (■) illustrate the correlation of onset temperature on crystallinity.
DETAILED DESCRIPTION OF THE INVENTION
Methods for preparing crystalline rapamycin and measuring samples of
rapamycin compounds for particle quality, particle size, and crystallinity are
described. The term "particle quality" as used herein refers to the quality of
crystals of a rapamycin compound. Typically, particle quality refers to the majority
of crystals in a sample containing a rapamycin compound. Particle quality can be

indicative of a variety of factors including the crystal size, the size distribution of the
crystals, the chemical homogeneity/purity of the crystals, and the morphology of the
majority of the crystals. In one example, a high particle quality may refer to crystals
whereby the majority of the crystals in a sample are large. In another example, a high
particle quality may refer to a sample whereby the majority of the crystals have the
same morphology, in a further example, a high particle quality may refer to a sample
whereby the majority of the crystals are not contaminated by impurities. In yet
another example, a high particle quality may refer to a sample whereby the majority
of the crystals are large, have the same morphology, and are not contaminated by
impurities.
The term "crystallinity" as used herein refers to the degree of structural order
in a sample containing a rapamycin compound. Typically, crystallinity is represented
by a fraction or percentage as a measure of how likely atoms or molecules are to be
arranged in a regular pattern such as a crystal. The crystallinity of a rapamycin
compound contributes to the overall particle quality and is affected by impurities,
such as atoms and molecules or by crystallization conditions or the presence of
defects. In one example, a sample having a higher crystallinity will have a powder X-
ray diffraction pattern having well-defined peaks. In one example, a sample having a
crystallinity of about 0% contains a solid that is substantially amorphous. In another
example, a sample having a crystallinity of about 100% contains a solid that is highly
crystalline. In a further example, a sample having a crystallinity of about 50%
contains a solid that is semi-crystalline.
The term "particle size" as used herein refers to the size of the majority of the
crystals in a sample. Typically, "particle size" refers to the median size of the crystals
in a sample as determined by measuring the longest linear dimension. The particle
size of a rapamycin compound contributes to the overall particle quality.
A. Methods for Preparing Crystalline Rapamycin
In one embodiment, methods for preparing crystalline rapamycin are
described. These methods are especially useful for large scale preparations and
provide highly crystalline rapamycin. These methods are also advantageous since the
crystalline rapamycin prepared thereby is more stable, thereby resulting in the

presence of fewer oxidative and/or hydrolysis degradation impurities. The term
"oxidative" or "hydrolysis" degradation impurities refers to impurities formed by
oxidation and/or hydrolysis of the triene region of the rapamycin molecule.
The term "rapamycin" is a term utilized in the art and herein to describe the
following compound.

The term "crude rapamycin" as used herein refers to a rapamycin sample that
is substantially crystalline, but contains less than about 20% impurities. In one
example, crude rapamycin contains less than about 15% impurities. In another
example, crude rapamycin contains less than about 10% impurities. In a further
example, crude rapamycin contains less than about 5% impurities. There are a variety
of methods for preparing crude rapamycin and include US Patent Nos. 3,993,749,
which is hereby incorporated by reference. Alternatively, rapamycin can be
purchased commercially (e.g., Wyeth). The crude rapamycin can be non-micronized
or micronized as described in US Patent No. 5,985,325, which is hereby incorporated
by reference.
The first step of this method includes heating crude rapamycin in ethyl acetate
to an elevated temperature. In one embodiment, the rapamycin/ethyl acetate solution
is heated to about 52 to about 58 °C. In another embodiment, the rapamycin/ethyl
acetate solution is heated to about 55 °C. Thereafter, the heated, ethyl acetate solution
is filtered. A variety of filtration instruments may be utilized and are readily
understood by one of skill in the art. The filtered solution is then maintained at an
elevated temperature. In one embodiment, the rapamycin/ethyl acetate solution is

maintained at a temperature of about 50 to about 60 °C. In another embodiment, the
rapamycin/ethyl acetate solution is maintained at a temperature of about 54 °C to
about 57 °C.
A solvent containing a hydrocarbon solvent is then added to the heated
solution. In one embodiment, the hydrocarbon solvent is heptanes. In another
embodiment, the hydrocarbon solvent is hexanes. In a further embodiment, the
hydrocarbon solvent is pentanes. The hydrocarbon solvent is desirably added at a rate
that results in the formation of crystalline rapamycin, desirably by gradual
crystallization. The hydrocarbon solvent may therefore be added at a constant rate or
a non-linear rate. Suitably, the rate of hydrocarbon solvent addition maintains the
temperature of the heated solution. More suitably, the addition rate of the
hydrocarbon solvent maintains the temperature at about 54 to 57 °C. One of skill in
the art would readily be able to adjust the hydrocarbon solvent rate of addition to
avoid premature precipitation of the rapamycin.
The hydrocarbon solvent is therefore typically added over a period of at least
about 20 minutes. In one example, the hydrocarbon solvent is added over a period of
at least about 30 minutes. In another example the, hydrocarbon solvent is added over
a period of about 60 minutes. In a further example, the hydrocarbon solvent is added
over a period of about 60 minutes at a constant rate. One of skill in the art would
readily be able to adjust the period of time required to add the hydrocarbon solvent to
avoid premature precipitation of the rapamycin.
The temperature of the ethyl acetate/hydrocarbon solvent solution is then
maintained at the elevated temperature. In one example, the ethyl
acetate/hydrocarbon solvent solution is maintained for about 30 minutes at a
temperature of about 55 to about 57 °C. The agitation speed is then reduced to the
minimum rate than is required to achieve a solid suspension. One of skill in the art
would readily be able to adjust the agitation rate based on the teachings provided
herein, the specific reactor being utilized and specifically the power per volume of the
reactor. In one example, the agitation rate is reduced to equal to or less than about
100 revolutions per minute (RPM). In another example, the agitation rate is about 45
to about 100 RPM.

After reducing the agitation rate, the solution is cooled in a non-linear fashion
at a decreasing cooling rate. In one example, the solution is cooled to about a first
reduced temperature using a first cooling rate; cooled to a second reduced temperature
using a second cooling rate; and further cooled using a third reduced temperature at a
third cooling rate. Typically, the third reduced temperature is less than the second
reduced temperature, which is less than the first reduced temperature. In one
example, the first reduced temperature is about 38 to about 42 °C; the second reduced
temperature is about 23 to about 27 °C; and the third reduced temperature is about 5
to about 10 °C. In another example, the first reduced temperature is about 40 °C; the
second reduced temperature is about 25 °C; and the third reduced temperature is about
9 °C. Typically, the third cooling rate is rate is faster than the second cooling rate,
which is faster than the first cooling rate. In one example, the first cooling rate is
about 4 to about 7 °C/hour; the second cooling rate is about 5 to about 9 °C/hour; and
the third cooling rate is about 7 to 10 °C. In a further example, the first cooling rate is
about 5 °C/hour; the second cooling rate is about 7.5 °C/hour; and the third cooling
rate is about 9 °C. In one embodiment, the solution is cooled to about 40 °C at a rate
of about 5 °C/hour; further cooled to a temperature of about 25 °C at a rate of about
7.5 °C/hour; and even further cooled to a temperature of about 7 to 8 °C at a rate of at
least about 9 °C/hour. This solution is then maintained at this temperature for about 2
to about 6 hours. In one example, the solution is maintained at this temperature for
about 2 hours.
The inventors also found that the rate of addition of the hydrocarbon solvent
influenced the crystallinity of the rapamycin. For example, when heptane is added at
a rate of 60 minutes or less, the morphology of the resulting crystals is orthorhombic.
However, when heptane is added over a period of at least about 60 minutes, the
morphology of the resulting crystals is acicular. However, the slower cooling rate,
desirably in a non-linear fashion, following heptane addition resulted in crystals with
more uniform size distribution. By controlling these parameters, precipitation of fine
particles of rapamycin is more easily controlled and/or avoided, thereby resulting in
crystalline rapamycin with a uniform size distribution.
The resultant crystalline rapamycin is then collected via filtration. Further
washing of the rapamycin with a solution containing ethyl acetate and the

hydrocarbon solvent, desirably heptanes, and drying the crystalline rapamycin is then
performed. Desirably, an excess of the hydrocarbon solvent over the ethyl acetate is
utilized. In one example, a 2:1 ratio of hydrocarbon solvent /ethyl acetate is utilized.
In another example, a 2:1 ratio of heptanes/ethyl acetate is utilized.
The rapamycin is washed using a hydrocarbon solvent/ethyl acetate solution at
reduced temperatures. In one example, the rapamycin is washed at a temperature of
about 6 to about 10 °C. In another example, the rapamycin is washed at a temperature
of about 8 °C. Typically, the rapamycin is dried in a low-shear dryer, but other drying
techniques can be utilized as determined by one of skill in the art.
By preparing the crystallized rapamycin according to the method described
herein, crystallized rapamycin is obtained in which the crystallinity is substantially
maintained over a period up to 4 months at up to about 60% relative humidity. In one
example, the crystallinity is maintained over a period of about 2 months. In another
example, the crystallinity is maintained over a period of about 4 months. In a further
example, the crystallinity is maintained up to about 60% relative humidity.
Specifically, the DSC profiles for the crystalline rapamycin prepared as described
herein stored for up to 4 months at up to about 60% relative humidity showed a
minimal change in the melting endotherm. In one example, the DSC profile for the
crystalline rapamycin showed a change in the melting endotherm of less than about
1%. In one example, the DSC melting endotherm showed a change of less than about
0.5%. In another example, the DSC melting endotherm showed a change of less than
about 0.3%. In a further example, the DSC melting endotherm showed a change of
less than about 0.1%.
In one embodiment, a method for purifying rapamycin is provided and
includes (i) heating crude rapamycin in ethyl acetate to about 55 °C; (ii) filtering the
product of step (i); (iii) maintaining the temperature of step (ii) at about 54 °C to
about 57 °C; (iv) adding heptanes to the product of step (iii) over a period of about 60
minutes at a constant rate; (v) maintaining the product of step (iv) at the same
temperature for about 30 minutes; (vi) reducing the agitation speed of step (v); (vii)
cooling the product of step (vi) to about 40 °C at a rate of about 5 °C/hour; (viii)
cooling the product of step (vii) to a temperature of about 25 °C at a rate of about 7.5
°C/hour; (ix) cooling the product of step (viii) to a temperature of about 7 to 8 °C at a

rate of at least about 9 °C/hour; (x) maintaining the product of step (ix) at the same
temperature for about 2 hours; and (xi) filtering the product of step (x) to obtain
crystalline rapamycin.
In a further embodiment, a method for purifying rapamycin is provided and
includes (i) heating crude rapamycin in ethyl acetate to about 55 °C; (ii) filtering the
product of step (i); (iii) maintaining the temperature of step (ii) at about 54 °C to
about 57 °C; (iv) adding heptanes to the product of step (iii) over a period of about 60
minutes at a constant rate; (v) maintaining the product of step (iv) at this temperature
for about 30 minutes; (vi) reducing the agitation speed of step (v); (vii) cooling the
product of step (vi) to about 40 °C at a rate of about 5 °C/hour; (viii) cooling the
product of step (vii) to a temperature of about 25 °C at a rate of about 7.5 °C/hour;
(ix) cooling the product of step (viii) to a temperature of about 7 to 8 °C at a rate of at
least about 9 °C/hour; (x) maintaining the product of step (ix) at this temperature for
about 2 hours; (xi) filtering the product of step (x) to obtain crystalline rapamycin;
(xii) washing the crystalline rapamycin with ethyl acetate and heptane at about 8 °C;
and (xiii) drying the product of step (xii).
B. Methods for Analyzing Rapamycin Compounds
Methods for analyzing rapamycin compounds are also described and are
typically performed using differential scanning calorimetry (DSC). Other techniques
can be utilized in conjunction with DSC and include X-ray diffraction (XRD) and
Raman spectroscopy, without limitation. A variety of DSC instruments is known in
the art and can be utilized. In one embodiment, the DSC instrument is the Q1000™
(TA Instruments) DSC instrument, among others.
The term "rapamycin compound" defines a class of immunosuppressive
compounds which contain the basic rapamycin nucleus shown above. The rapamycin
compounds of this invention include compounds which maybe chemically or
biologically modified as derivatives of the rapamycin nucleus, while still retaining
immunosuppressive properties. Accordingly, the term "rapamycin compound"
includes esters, ethers, oximes, hydrazones, and hydroxylamines of rapamycin, as
well as rapamycins in which functional groups on the rapamycin nucleus have been
modified, for example through reduction or oxidation. The term "rapamycin

compound" also includes pharmaceutically acceptable salts of rapamycins, which are
capable of forming such salts, either by virtue of containing an acidic or basic moiety.
Examples of rapamycin compounds that can be analyzed as described herein include,
without limitation, rapamycin, 42-esters of rapamycin including CCI-779
(temsirolimus), norrapamycin, deoxorapamycin, desmethylrapamycins, or
desmethoxyrapamycin, or pharmaceutically acceptable salts, prodrugs, or metabolites
thereof and those described in US Patent Application Publication Nos. US-2005-
0272702, US-2006-013550, US-2006-0040971, US-2006-0036091, US-2005-
0014777, US-2006-0199834, US-2005-0234086, and US-2003-0114477 and US
Patent Nos. 5,358,908; 5,358,909; 5,362,718; 5,302,584, which are hereby
incorporated by reference. In one example, a rapamycin compound includes
rapamycin which can be purchased commercially or can be prepared using a variety
of methods available in the art. In another example, a rapamycin compound includes
CCI-779.
The term "CCI-779" as used herein refers to rapamycin 42-ester with 3-
hydroxy-2-(hydroxymethyl)-2-methylpropionic acid. A variety of methods for
preparing CCI-779 is known in the art and includes those described in US Patent Nos.
5,362,718 and 6,277,983, which are hereby incorporated by reference. Alternatively,
CCI-779 can be purchased commercially (e.g., Wyeth). The CCI-779 can be non-
micronized or micronized, as described in US Patent Application Publication No. US-
2005-0152983-A1, which is hereby incorporated by reference.


The term "desmethylrapamycin" refers to the class of rapamycin compounds
which lack one or more methyl groups. Examples of desmethylrapamycins that can
be used according to the present invention include 3-desmethylrapamycin (US Patent
No. 6,358,969), 7-O-desmethyl-rapamycin (US Patent No. 6,399,626), 17-
desmethylrapamycin (US Patent No. 6,670,168), and 32-O-desmethylrapamycin,
among others.
The term "desmethoxyrapamycin" refers to the class of rapamycin compounds
which lack one or more methoxy groups and includes, without limitation, 32-
desmethoxyrapamycin.
The rapamycin compounds measured in the methods described herein include
samples in the solid state and can be crystalline, semi-crystalline, non-crystalline, or
aggregates. Crystalline rapamycin is desirably prepared according to the procedures
discussed in Sehgal et al., J. Antibiotics, 28(10): 727-732 (1975); Swindells et al.,
Canadian J. Chem., 56(18):2491-2492 (1978); and US Patent Application Publication
No. US-2006-040971. Crystalline CCI-779 is desirably prepared by recrystallization
from diethylether and heptane as described in US Provisional Patent Application No.
60/748,006, which is hereby incorporated by reference.
The samples containing rapamycin compounds may contain low levels of
impurities, including oxidative and/or hydrolysis impurities, solvents, or the like. In
one example, the samples of CCI-779 contain only trace amounts of acetone,
desirably less than about 0.3% wt/wt of acetone. Similarly, the samples of CCI-779
contain less than about 0.3% wt/wt phenylboronic acid, and less than about 1.5 wt%
of oxidative/hydrolysis decomposition products of CCI-779.
The term "crystalline" as used herein refers to solid samples of rapamycin
compounds that have one definitive crystalline structure. The term "semi-crystalline"
as used herein refers to solid samples of rapamycin compounds that have crystalline
regions dispersed within amorphous regions. The term "non-crystalline" and
"amorphous" axe used interchangeably and refer to solid samples of rapamycin
compounds that have no regions of crystallinity dispersed therethrough and therefore
no crystalline form. The term "aggregate" as used herein refers to grouping of
crystals which are intergrown or fused in a particle of a rapamycin compound.

Crystal quality is known to influence the stability of a sample containing a
rapamycin compound. For example, amorphous or semi-crystalline rapamycin
compounds undergo rapid oxidative degradation. Further, the median particle size of
rapamycin compounds determines flow property, with a larger particle size being
desired. The method thereby includes determining/calculating the particle quality,
crystallinity, particle size, or a combination thereof of a sample containing a
rapamycin compound, i.e., the test sample. The method is thereby performed by
analyzing the DSC heat flow signal of the rapamycin compound. The heat flow
signal of the rapamycin compound is then compared to the heat flow signal of a
predetermined standard.
A number of useful parameters can be obtained from the heat flow signal and
include melting temperature, including onset melting temperature and peak
temperature, and heat of fusion. These parameters can also be utilized in the
determination of particle quality, crystallinity, or particle size.
The term "melting temperature" as used herein includes the temperature at
which a solid, i.e., a rapamycin compound, melts. The melting temperature can
include the onset melting temperature or the peak melting temperature. Typically, the
melting temperature is the peak melting temperature.
The term "heat of fusion" as used herein describes the total heat released by a
rapamycin compound during melting or fusion. The heat of fusion is obtained by
integrating the area under the heat flow signal plot and is typically expressed in
calories/gram or Joules/gram. However, other conventions for expressing the units of
heat of fusion could be utilized by one of skill in the art.
Desirably, the DSC peak temperature, i.e., the melting temperatures, of the
heat flow signal of the rapamycin compound is measured and then compared to the
heat flow signal of the predetermined standard.
As used herein, the term "predetermined standard" refers to one or more solid
samples of a highly crystalline rapamycin compound where the average size of the
particles and crystallinity is known and is correlated with a DSC peak temperature.
More desirably, the predetermined standard contains crystalline rapamycin
compound. Most desirably, the predetermined standard contains a 100% crystalline
rapamycin compound.

The heat flow signal of the rapamycin compound can be compared with the
heat flow signal of the predetermined standard by a single point correlation or using a
calibration curve. By doing so, the crystallinity, particle quality, or particle size of the
rapamycin compound being analyzed can be determined.
In one embodiment, the heat flow signal of the test sample containing the
rapamycin compound is compared with the heat flow signal of the predetermined
standard containing crystalline rapamycin compound using a single point correlation.
Typically, the heat of fusion obtained from the heat flow signal is utilized for the
comparison. In one example, the heat of fusion is utilized in a single point correlation
to determine the crystallinity of a rapamycin compound. In another example, the
crystallinity of a rapamycin compound is calculated using a single point correlation.
In a further example, the crystallinity of the rapamycin compound can be
calculated using the following equation:
test sample crystallinity = 100 x heat of fusion of the test sample
heat of fusion of the predetermined standard.
In another embodiment, the heat flow signal of the test sample containing the
rapamycin compound is compared with the heat flow signal of the predetermined
standard containing crystalline rapamycin compound using a calibration curve. One
of skill in the art would readily be able to prepare a calibration curve using the
teachings of the specification and knowledge in the art. Typically, the calibration
curve is prepared for the predetermined standard by using multiple samples
containing crystalline rapamycin compounds. Desirably, at least 3 samples are
required to generate the calibration curve. However, more samples can be used as
determined by one of skill in the art to prepare the calibration curve. In one example,
the heat of fusion of a test sample containing a rapamycin compound is utilized in
combination with a calibration curve to determine the crystallinity of the rapamycin
compound.
The calibration curve is prepared by plotting the heat of fusion, peak
temperature, or onset temperature for each of the multiple samples against the
crystallinity of each of the same multiple samples to obtain the calibration curve. A

best fit line or curve is then drawn and the formula of the best fit line is calculated. In
another example, the calibration curve is prepared by plotting the heat of fusion
against the crystallinity. In a further example, the calibration curve is prepared by
plotting the peak temperature against the crystallinity. In still a further example, the
calibration curve is prepared by plotting the onset temperature against the
crystallinity. In yet another example, the calibration curve is calculated by plotting
the heat of fusion for each of multiple samples containing a crystalline rapamycin
compound of a known crystallinity against the crystallinity for each of multiple
samples containing the rapamycin compound. Typically, the calibration curve is
specific to the type of DSC instrument and experimental conditions and procedure
utilized to obtain the heat of fusion values. However, one of skill in the art would be
able to determine if a calibration curve obtained from one procedure and DSC
instrument can be utilized by using data obtained from another DSC instrument using
the same procedure.
Once the calibration curve is prepared, it can then be utilized to determine the
crystallinity of test samples containing a rapamycin compound. Specifically, the test
samples containing a rapamycin compound are analyzed to determine one or more of
the heat of fusion, peak temperature, or onset temperature of the rapamycin compound
in the test sample. These values, i.e., heat of fusion, peak temperature, or onset
temperature, can then be utilized using the formula of the best fit line of the
predetermined standard to determine the crystallinity, among other factors, of the
rapamycin compound in the test sample. By doing so, an accurate determination of
the crystallinity of samples containing rapamycin compounds can be obtained.
The inventors have found a trend in the DSC heat flow signal, and thereby the
melting temperature, for rapamycin compound samples. Specifically, the heat flow
signal of samples containing a rapamycin compound is found to vary depending on
the crystallinity of the rapamycin compound. In one example, the crystallinity of the
rapamycin compound sample is proportional to the melting temperature of the heat
flow signal.
In one embodiment, samples containing higher crystalline rapamycin had large
particles and higher melting temperatures of at least about 188 °C, desirably about
188 °C to about 190 °C. Samples containing less crystalline rapamycin had smaller

particles and lower melting temperatures of less than about 183 °C, desirably less than
about 180 to less than about 183 oC. See, Figure 2.
In another embodiment, samples containing higher crystalline CCI-779 had
large particles and higher melting temperatures of at least about 168 °C, desirably
about 168 to about 170°C. Samples containing less crystalline CCI-779 had smaller
particles, lower melting temperatures of at least about 166 to less than about 168°C.
Samples containing semi-crystalline CCI-779 aggregates had lower melting
temperatures than crystalline samples, i.e., melting temperatures of at least about
164°C to less about 166°C. Further, samples containing non-crystalline CCI-779 had
glass transition temperatures, but did not have melting temperatures. See, Figure 3.
As noted above, the DSC melting temperature is proportional to the size and
crystallinity of the rapamycin compound particles. For samples containing CCI-779,
a large particle size includes particles that have a median particle size of greater than
about 30 m in length for the longest axis of the particle, and more desirably about 30
m to about 250 m for the longest axis of the CCI-779 particle. Alternatively, a
small particle size includes particles that have a median particle size of less than about
30 m for the longest axis of the CCI-779 particle.
The inventors also found that the X-ray diffraction pattern of a less crystalline
rapamycin compound contained broad peaks. Further, when samples containing
amorphous and crystalline rapamycin compounds are analyzed by XRD, the XRD
pattern showed sharp peaks of the crystalline rapamycin compound and a baseline
shift or "amorphous halo" for the amorphous rapamycin compound.
In one embodiment, a method is described for measuring particle quality of a
rapamycin compound using differential scanning calorimetry, including analyzing the
heat flow signal of a sample containing a rapamycin compound; and comparing the
heat flow signal to a predetermined standard, wherein the particle quality is
proportional to the melting temperature of the sample.
In another embodiment, a method is described for determining particle size of
a rapamycin compound using differential scanning calorimetry, including analyzing
the heat flow signal of a sample containing a rapamycin compound and comparing the
heat flow signal to a predetermined standard, wherein the particle size is proportional
to the melting temperature of the sample.

In a further embodiment, a method is provided for determining particle quality
of a rapamycin compound using differential scanning calorimetry, including
analyzing the heat flow signal of a sample containing a rapamycin compound, and
comparing the heat flow signal to a predetermined standard, wherein a large particle
size of a rapamycin compound is characterized by a high melting temperature and a
small particle size is characterized to a low melting temperature.
The following examples are provided to illustrate the invention and do not
limit the scope thereof. One skilled in the art will appreciate that although specific
reagents and conditions are outlined in the following examples, modifications can be
made which are meant to be encompassed by the spirit and scope of the invention.
EXAMPLES
Example 1 - General Process for Analyzing Particles of CCI-779 Samples
In this example, DSC peak temperatures were measured and utilized to assess
the particle categories for test samples containing CCI-779.
Samples containing CCI-779, obtained by crystallizing CCI-779 from
ether/heptane using the procedure set forth in US Provisional Patent Application No.
60/748,006 were analyzed using the Q Series™ Q1000-0450 DSC Instrument (TA
Instruments) using the parameters in Table 1. Once the DSC peak temperatures were
obtained, they were compared with predetermined standards containing crystalline
CCI-779 and placed into particle categories. See, Figure 1 in which the peak
temperatures for the 25 samples were grouped according to particle category.
Because there was overlap in the peak temperatures for certain samples, 25 distinct
samples are not visible. A particle category 1 refers to crystalline CCI-779 samples
having large particles; a particle category 2 refers to crystalline CCI-779 samples
having small particles; a particle category 3 refers to semi-crystalline CCI-779
aggregates; and particle category 4 refers to non-crystalline CCI-779.


From this data, it is determined that higher DSC peak temperatures are
indicative of CCI-779 samples that are more crystalline.
Example 2 - Analyzing Particles of CCI-779 Samples
In this example, the particle quality, crystallinity, and melting temperature
were measured of twenty five (25) samples of CCI-779 obtained by crystallizing
CCI-779 from ether/heptane using the procedure set forth in US Provisional Patent
Application No. 60/748,006. The solid samples were analyzed for the DSC peak
temperatures using the Q Series™ Q1000-0450 DSC Instrument (TA Instruments)
using the parameters in Table 1 as noted above.
The grade and crystallinity size of the sample was then analyzed by optical
microscopy. In summary, optical microscopy was performed using a Nikon™
Eclipse E600 microscope capable of 5x to 100x magnification, fitted with a Nikon™
DXM 1200 digital camera and a Nikon™ ACT-1 v 2.12 calibrated image acquisition
system. Measurements were obtained by dispersing about 0.05 mg of the sample on
a glass holder. The sample was then covered with a drop of Resolve® microscope
immersion oil (Richard-Allan Scientific) and a cover slip was added. Care was taken
to ensure that the particles were not subjected to attrition during image acquisition.
Sample images were acquired about 1 to about 2 minutes after sample preparation.
Fresh samples were prepared, if re-imaging was required.
The "class" of the sample was determined by correlating the DSC
temperature with the "grade" and "crystallinity size" of the sample. Specifically, if a
sample containing CCI-779 was determined to be crystalline by optical microscopy
with large crystals, it was assigned a class 1 sample; if a sample containing CCI-779
was determined by optical microscopy to be crystalline with small crystals, it was

assigned a class 2 sample; and if a sample containing CCI-779 was determined by
optical microscopy to be semi-crystalline, regardless of crystal size, it was assigned a
class 3 sample. The class of the sample was then correlated to the DSC peak
temperature obtained for the same sample.

* A DSC peak temperature was not obtained for these samples
These results illustrate that, in general, crystalline samples of CCI-779 had
higher DSC peak temperatures than semi-crystalline samples of CCI-779. Further,
CCI-779 samples containing larger crystals resulted in higher DSC peak
temperatures than samples of CCI-779 containing smaller crystals.
Example 3 - Crystallization of Rapamycin
Crude rapamycin is slurried in ethyl acetate and heating to 55 °C. The heated
solution is then filtered using a clarifying filter into a crystallizing vessel and the
solution is then maintained at 54 to 57 °C. Heptanes are then added to the vessel over

a period of 60 minutes at a constant rate. After addition of heptanes, the solution is
held at 55 to 57 °C for 30 minutes. The agitation speed is then reduced in an effort to
achieve a solid suspension. The solution is then cooled to 40 °C over a period of 3
hours at a rate of 5 °C/hour; then cooled to 25 °C over a period of 2 hours at a rate of
7.5 °C/hour; and further cooled to 7 to 8 °C over a period of at least 60 minutes,
desirably over a period of 2 hours at a rate of 9 °C. The solution is then held at 7 to 8
°C for 2 hours and then filtered. The solid obtained from the filtration is then washed
using a solution containing ethyl acetate/heptanes held at 8 °C. The washed solid is
then dried using a low-shear dryer to obtain crystalline rapamycin.
Example 4 - Analyzing the Crystallinity of Samples Containing Varying
Content of Crystalline Rapamycin
In this example, samples containing rapamycin (Table 3) were analyzed by
DSC and optical microscopy. The morphology and approximate particle size were
determined using optical microscopy. A standard DSC sample ramp rate of 10
°C/min and a hermetically sealed aluminum pan were utilized. The DSC graphs
were obtained and are provided in Figures 2A and 2B.
Sample 1 contained crystalline rapamycin and was prepared by suspending
crude crystalline rapamycin (1 g) in 10 mL of methoxy-2-propanol and heating the
suspension to 40 °C to obtain a clear solution. The solution was cooled from 40 °C
to 15 °C over a period of 2 hours and resulted in the gradual crystallization of
rapamycin. The crystallized solid was collected via filtration at room temperature
and dried in air at room temperature.
Sample 2 contained crystalline rapamycin and about 2 to about 3% of
oxidative/hydrolysis degradation impurities and was prepared using the process of
Example 3. A portion of the batch was maintained at 25 °C and 60% relative
humidity for 2 months.
Sample 3 contains crystalline rapamycin and about 2 to about 3% of
oxidative/hydrolysis degradation impurities and was prepared using the process of
example 3. A portion of the batch was subjected to 25 °C and 60% relative humidity
for 4 months.


These results illustrate that low levels of impurities in. samples containing
crystalline rapamycin do not affect the DSC melting temperature of a crystalline
rapamycin sample over time.
Example 5: Calculating a Calibration Curve to Determine the Correlation
Between Heat of Fusion and Crystallinity
This example was performed to prepare a calibration curve to establish the
correlation of the heat of fusion with crystallinity. Samples containing known
crystalline CCI-779, i.e., a predetermined standard, were analyzed using DSC and
the parameters noted in Table 1. Each sample contained a known percentage of
crystalline and amorphous CCI-779. The results are provided in Table 4.

The crystallinity was then plotted against each of the heat of fusion, onset and
peak temperatures. See, Figure 3.
The graph illustrates that all three parameters linearly correlated with the
amount of crystalline CCI-779 in the samples. The graph also illustrates that the best
linear correlation is achieved using the heat of fusion. Not only was the correlation
error of the heat of fusion measurement lower than the other two parameters, but it
also had a higher sensitivity. The higher sensitivity was determined by monitoring the

slope of the line, which slope is about twice (006049) the slope of the onset
temperature (0.0875).
Specifically, by using the heat of fusion obtained for each sample and the
degree of crystallinity, a relationship between the crystallinity and heat of fusion was
determined for this particular instrument as illustrated by the following equation.
Degree of crystallinity = 1.6465 x Heat of fusion + 3.5988
Example 6: Determining the Crystallinity of Samples Containing Varying
Amounts of Crystalline CCI-779
This example was performed to determine the accuracy of the equation set
forth in Example 4. Specifically, the heats of fusion for four (4) samples containing
known amounts of crystalline CCI-779 were determined. Once determined, the
crystallinities were calculated using the equation in Example 4. The results are
shown in Table 5.

These data illustrate that by using the heat of fusion, the crystallinity can
accurately be calculated with less than a 3% error.
Example 7 - Effect of Sample Weight on Heat of Fusion
In this example, the effect of sample weight in determining the crystallinity
of three samples of CCI-779 was measured.
Samples 1 and 2 were obtained by crystallizing CCI-779 from ether/heptane
using the procedure set forth in US Provisional Patent Application No. 60/748,006.
Sample 3 was obtained using the procedure set forth in US Provisional Patent
Application No. 60/748,143.

The samples were analyzed for the DSC onset temperatures, peak
temperatures, and heat of fusion values using the Q Series™ Q1000-0450 DSC
Instrument (TA Instruments) using the parameters set forth in Table 2. The results
are provided in Tables 6-8.


By using the average heat of fusion set forth in Tables 6-8 provided above,
the degree of crystallinity of each batch was calculated using the equation set forth in
Example 3. The results are provided in Table 9.

These data illustrate that sample weight does not substantially affect the
crystallinity of a sample or the use of a heat of fusion in predicting the crystallinity of
a sample containing CCI-779.
Example 8 - Determining the Crystallinity of Samples Containing CCI-779
Nineteen (19) samples containing crystalline CCI-779 were prepared and
analyzed using the DSC parameters set forth in Table 2. The crystallinity of each
sample was calculated using the average heat of fusion obtained for each sample
using DSC and equation in Example 6 and reproduced below. The results are
provided in Table 10.
Degree of crystallinity = 1.6465 x Heat of Fusion + 3.5988



The stability of the samples was then separately analyzed after a period of 6
months at (i) 5°C or (ii) 25°C at 60% relative humidity. The results indicate that
batches having a higher content of crystalline CCI-779 were more stable than samples
containing a lower content of crystalline CCI-779.
Example 9 - Variation of Heating Rate on Heat of Fusion and Crystallinity
Six samples containing crystalline CCI-779 were analyzed by DSC using the
parameters set forth in Table 2. Samples 1,4, and 7 contained 7 mg of crystalline
CCI-779 and were heated in the DSC at a temperature of 7°C/min. Samples 2, 5, and
8 contained 10 mg of crystalline CCI-779 and were heated in the DSC at a rate of
10°C/min. Samples 3, 6, and9 contained 20 mg of crystalline CCI-779 and were
heated in the DSC at a rate of 20°C/min. The onset temperature, peak temperature,
and heat of fusion were obtained from the DSC and are provided in Table 11.

The data illustrated that increasing the heating rate during analyzation by DSC
did not significantly alter the heat of fusion.

All publications cited in this specification are incorporated herein by reference
herein. While the invention has been described with reference to a particularly
preferred embodiment, it will be appreciated that modifications can be made without
departing from the spirit of the invention. Such modifications are intended to fall
within the scope of the appended claims.

What is Claimed Is:
1. A method for measuring particle quality of a rapamycin compound
using differential scanning calorimetry, comprising:
analyzing the heat flow signal of a sample comprising a rapamycin
compound; and
comparing the heat flow signal of said sample to the heat flow signal
of a predetermined standard;
wherein said particle quality is proportional to the melting temperature
of said heat flow signal of said sample.
2. The method according to claim 1, wherein a higher melting
temperature corresponds to a higher quality particle.
3. The method according to claim 1 or 2, wherein a higher particle quality
corresponds to a higher crystallinity of said rapamycin compound.
4. The method according to claim 1, wherein said melting temperature is
proportional to the median particle size of the crystals of said rapamycin compound.
5. The method according to claim 4, wherein a large median particle size
is characterized by a high melting temperature.
6. The method according to claim 5, wherein said sample comprises CCI-
779 and said large median particle size is at least about 30 m.
7. The method according to claim 6, wherein said large median particle
size is about 30 m to about 250 m.
8. The method according to claim 5, wherein said sample comprises CCI-
779 and said high melting temperature is at least about 168 °C.

9. The method according to claim 8, wherein said high melting
temperature is about 168 to about 170°C.
10. The method according to claim 5, wherein said sample comprises
rapamycin and said high melting temperature is at least about 188°C.
11. The method according to claim 10, wherein said high melting
temperature is about 188°C to about 190°C.
12. The method according to claim 1, wherein a lower melting temperature
corresponds to a lower quality particle.
13. The method according to claim 12, wherein a lower particle quality
corresponds to a lower crystallinity.
14. The method according to claim 1, wherein a small median particle size
of the crystals of said rapamycin compound is characterized by a low melting
temperature.
15. The method according to claim 12 or 14, wherein said sample
comprises CCI-779 and said low melting temperature is less than about 166°C.
16. The method according to claim 15, wherein said low melting
temperature is about 164 to about 166°C.
17. The method according to claim 12 or 14, wherein said sample
comprises rapamycin and said low melting temperature is less than about 183°C.
18. The method according to claim 12 or 14, wherein said sample
comprises rapamycin and said low melting temperature is less than about 180 to about
183°C.

19. The method according to claim 14, wherein said sample comprises
CCI-779 and said small median particle size is less than about 30 m.
20. The method according to claim 1, wherein said sample comprises
semi-crystalline aggregates and has a lower melting temperature than a crystalline
sample.
21. The method according to claim 1, wherein said sample comprises a
non-crystalline rapamycin compound and has a lower melting temperature than a
sample comprising a semi-crystalline rapamycin compound.
22. The method according to claim 1, wherein said sample comprising a
non-crystalline rapamycin compound and has a lower melting temperature than a
sample comprising a crystalline rapamycin compound.
23. The method according to claim 1, wherein said rapamycin compound
is purified from the same solvent as the predetermined standard.
24. A method for determining median particle size of a sample containing
crystals of a rapamycin compound using differential scanning calorimetry,
comprising:
analyzing the melting temperature of a sample comprising a rapamycin
compound; and
comparing the melting temperature to a predetermined standard;
wherein said median particle size is proportional to the melting
temperature of said sample.
25. The method according to claim 24, wherein a large median particle
size is characterized by a high melting temperature and a small median particle size is
characterized to a low melting temperature.

26. A method for determining the crystallinity of a rapamycin compound,
comprising:
analyzing the heat flow signal of a test sample comprising a rapamycin
compound; and
calculating the crystallinity of said test sample by comparing said heat
flow signal to the heat flow signal of a predetermined standard comprising a
crystalline rapamycin compound.
27. The method according to claim 26, wherein said calculation is
performed using a single point calculation.
28. The method according to claim 27, wherein said predetermined
standard comprises a 100% crystalline rapamycin compound.
29. The method according to claim 28, wherein said crystallinity of said
test sample is calculated according to the following:
test sample crystallinity = 100 x heat of fusion of said test sample
heat of fusion of said predetermined standard.
30. The method according to claim 26, wherein said calculation is
performed using a calibration curve.
31. The method according to claim 30, wherein said predetermined
standard comprises multiple samples comprising crystalline rapamycin compound.
32. The method according to claim 31, further comprising:
plotting the heat of fusion, peak temperature, or onset temperature for
each of said multiple samples against the crystallinity of each of said multiple samples
to obtain a calibration curve having a best fit line;
calculating a formula of said best fit line;
analyzing the heat of fusion, peak temperature, or onset temperature of
said rapamycin compound in said test sample; and

calculating the crystallinity of said rapamycin compound in said test
sample using said heat of fusion, peak temperature, or onset temperature of said test
sample and said formula.
33. The method according to claim 31, wherein said calibration curve is
prepared by plotting said heat of fusion for each of multiple samples comprising a
crystalline rapamycin compound of a known crystallinity against the crystallinity for
each of multiple samples comprising said rapamycin compound.
34. The method according to any of claims 1 or 24 to 33, wherein said
sample comprises rapamycin.
35. The method according to any of claims 1 or 24 to 33, wherein said
sample comprises CCI-779.
36. A method for purifying rapamycin, comprising:
(i) heating crude rapamycin in ethyl acetate to about 55°C;
(ii) filtering the product of step (i);
(iii) maintaining the temperature of step (ii) at about 54°C to about 57°C;
(iv) adding heptanes to the product of step (iii) over a period of about 60
minutes at a constant rate;
(v) maintaining the product of step (iv) at said temperature for about 30
minutes;
(vi) reducing the agitation speed of step (v);
(vii) cooling the product of step (vi) to about 40°C at a rate of about 5
°C/hour;
(viii) cooling the product of step (vii) to a temperature of about 25 °C at a
rate of about 7.5 °C/hour;
(ix) cooling the product of step (viii) to a temperature of about 7 to 8 °C at
a rate of at least about 9 °C/hour;

(x) maintaining the product of step (ix) at said temperature for about 2
hours; and
(xi) filtering the product of step (x) to obtain said crystalline rapamycin.
37. The method according to claim 36, further comprising:
(xii) washing said crystalline rapamycin with ethyl acetate and heptane at
about 8 °C; and
(xiii) drying the product of step (xii).

Methods for purifying rapamycin are described. Methods for measuring particle quality, median particle size, and
crystallinity of samples containing rapamycin or a derivative thereof are also provided.