Stability Profiles of Drug Products Extended
beyond Labeled Expiration Dates
ROBBE C. LYON,1 JEB S. TAYLOR,1 DONNA A. PORTER,2 HULLAHALLI R. PRASANNA,1 AJAZ S. HUSSAIN3
1

Division of Product Quality Research, Center for Drug Evaluation and Research, Food and Drug Administration,
HFD-941, White Oak, Life Sciences Building 64, 10903 New Hampshire Avenue, Silver Spring, Maryland 20993-0002
2

Division of Field Science, Office of Regional Operations, Office of Regulatory Affairs, Food and Drug Administration,
Rockville, Maryland 20857
3

Vice President & Global Head of Biopharmaceutical Development, Sandoz, 506 Carnegie Center, Princeton,
New Jersey 08540

Received 6 January 2006; accepted 17 March 2006
Published online in Wiley InterScience (www.interscience.wiley.com). DOI 10.1002/jps.20636

ABSTRACT: The American Medical Association has questioned whether expiration
dating markedly underestimates the actual shelf life of drug products. Results from the
shelf life extension program (SLEP) have been evaluated to provide extensive data to
address this issue. The SLEP has been administered by the Food and Drug
Administration for the United States Department of Defense (DOD) for 20 years. This
program probably contains the most extensive source of pharmaceutical stability data
extant. This report summarizes extended stability profiles for 122 different drug products
(3005 different lots). The drug products were categorized into five groups based on
incidence of initial extension failures and termination failures (extended lot eventually
failed upon re-testing). Based on testing and stability assessment, 88% of the lots were
extended at least 1 year beyond their original expiration date for an average extension of
66 months, but the additional stability period was highly variable. The SLEP data
supports the assertion that many drug products, if properly stored, can be extended past
the expiration date. Due to the lot-to-lot variability, the stability and quality of extended
drug products can only be assured by periodic testing and systematic evaluation
of each lot. ß 2006 Wiley-Liss, Inc. and the American Pharmacists Association J Pharm Sci
95:1549–1560, 2006

Keywords:
date

shelf life; drug stability; shelf life extension program; SLEP; expiration

INTRODUCTION
The concern that expiration dating may markedly
underestimate the actual shelf life of drug
products has been an issue.1–3 The American
Medical Association (AMA) recently reviewed the
procedures for setting pharmaceutical expiration
dates and the clinical and fiscal consequences of
setting such dates.4 The AMA concluded that the
actual shelf lives of some products are greater
Correspondence to: Robbe C. Lyon (Telephone: 301-7960019; Fax: 301-796-9816; E-mail: robbe.lyon@fda.hhs.gov)
Journal of Pharmaceutical Sciences, Vol. 95, 1549–1560 (2006)
ß 2006 Wiley-Liss, Inc. and the American Pharmacists Association

than their labeled expiration dates and acknowledged that best evidence to support this resides in
the shelf life extension program (SLEP). Smaller
studies have addressed the long-term stability of
drug products5,6 and drug substances.7 One study
determined that four products, captopril tablets,
flucloxacillin capsules, cefoxitin injection, and
theophylline tablets stored under ambient temperatures maintained at least 98% of label claim
for drug content for 18–170 months past the
labeled expiration dates.5 In a modification to the
AMA Policy H-115.983, the pharmaceutical
industry, the Food and Drug Administration
(FDA) and the United States Pharmacopeia

JOURNAL OF PHARMACEUTICAL SCIENCES, VOL. 95, NO. 7, JULY 2006

1549

 1550

LYON ET AL.

(USP) were urged to determine the benefits and
risks associated with lengthening expiration
dates and to subsequently conduct longer stability
testing. In response, all of the data from the SLEP
was reviewed and analyzed. As a retrospective
analysis, this report summarizes extended drug
product stability data collected by the SLEP over
the past 20 years.
The International Conference on Harmonization (ICH) defines shelf life as ‘‘the time period
which a drug product is expected to remain within
the approved shelf life specification, provided that
it is stored under conditions defined on the
container label.’’8 It is expected that the actual
shelf life will slightly exceed the projected labeled
shelf life. Pharmaceutical manufacturers are
required to assign an expiration date to each drug
product marketed in the United States. The
labeled shelf life is estimated using appropriate
stability testing (21 CFR 211.137 and 211.166)
under current good manufacturing practices as
established and monitored by the FDA. The
stability assessments of both the new drug substance and the new drug product stored under
controlled conditions are required documentation
for submission to the FDA in a new drug application (NDA). The assessment follows scientifically
based technical procedures as described in the ICH
Q1A(R2) Guidance.8 The initial expiration date is
based on the amount of real-time stability data
(generally from pilot scale batches) for the drug
product available at the time of approval of the
NDA. This initial date may later be extended
contingent upon the receipt of acceptable supporting data from the manufacturer based on accelerated stability studies and actual real-time stability
data collected from the first three production
batches. Drug products marketed in the United
States generally have a labeled shelf life of 12–
60 months.
The SLEP is administered by the FDA for the
U.S. DOD9 and recently for the Strategic National
Stockpile (SNS). To maintain a state of readiness,
the military maintains, under controlled conditions, large stockpiles of pharmaceuticals sealed in
their original container closures. A system of
extending the functional shelf life of these drug
products beyond their original expiration date was
initiated to reduce the high cost of replacing these
stockpiles. Based on a comprehensive testing
program, the shelf life of several drug products
has been extended on a lot-by-lot basis. This
program has resulted in substantial savings to
the military. In return the FDA has access to a

valuable source of long-term stability data for a
variety of drug products.
This report summarizes data for 3005 lots
representing 122 drug products generated by the
SLEP since 1986. Based on stability assessment,
88% of the lots were extended beyond their original
expiration date. Of the 2652 lots extended, only
18% were eventually terminated due to failure.
The rest of the lots are either still active (35%) or
were abated (47%) by the military. The shelf life
extension results were summarized in tabular
form. The products were arranged into groups
based on stability performance.

JOURNAL OF PHARMACEUTICAL SCIENCES, VOL. 95, NO. 7, JULY 2006

DOI 10.1002/jps

MATERIALS AND METHODS
SLEP Program Operation
The SLEP is a key component of the Medical
Readiness Strategic Plan as developed by the
DOD Health Affairs and the Military Medical
Departments. The DOD Defense Medical Standardization Board (DMSB) oversees the SLEP
program and acts as an interface between the
military services and the FDA. Pharmaceutical
drug products sealed in original container closures are stored under controlled conditions by
the military services. Certain lots of drug product
that are approaching their labeled expiration date
are selected by the DMSB for participation in the
SLEP program. Representative sealed containers
of drug product from a given lot are submitted to
the FDA SLEP coordinator in the Office of
Regulatory Affairs for testing by the FDA field
labs. The test results are transmitted through the
FDA SLEP coordinator to the FDA SLEP chemist
in the Division of Product Quality Research
(DPQR), Center for Drug Evaluation and
Research (CDER) for analysis. The FDA SLEP
chemist evaluates the test results, approves or
rejects the extension of the shelf life for that lot
and archives the data, evaluations, and approvals
for the program. Approval for a new expiration
date for that lot of drug product is transmitted
through the FDA SLEP coordinator to representatives for the DMSB.
Sample Testing
Sealed containers of drug product from a given lot
are sent to an FDA field lab. Samples are
subjected to a battery of tests prescribed by
the FDA SLEP chemist. These tests and

 STABILITY PROFILES OF DRUG PRODUCTS

specifications are based on compendial (USP)
product release tests or the product release tests
as described in the original FDA submission
(NDA). If a lot fails any specification based on
the battery of tests, then the shelf life for that lot
is expired.
Predicting Extended Shelf Life
Extending the shelf life is based on developing a
history of real-time stability data for each lot of
drug product. This data is compiled by continual
testing within the SLEP. Test results from each
retest project form a set of real-time data. That
real-time data is evaluated and an updated
expiration period for that lot of drug product
is predicted using regression analysis. Each
retest attribute generates a particular remaining
expiration period. If each remaining expiration
period predicted is longer than 1 year, the lot of
drug product is granted a new expiration date. An
‘‘initial extension failure’’ indicates that the lot
could not be extended past original expiration
date. A ‘‘termination failure’’ indicates that a
previously extended lot eventually fails upon retesting and based on regression analysis predictions, cannot be extended for an additional year.
Product Quality Attributes
The attributes tested for a drug product varied
depending on the dosage form. For solid oral drug
products, the attributes were potency (assay),
impurities, water content, dissolution, and physical appearance. For reconstituted dry powders, the
attributes were potency, pH, water content, and
physical appearance. For injectable solutions, the
attributes were potency, impurities, pH, preservatives, and physical appearance (color, particulates). For creams and ointments, the attributes
were potency, pH, and physical appearance. For

1551

autoinjectors, the attributes were potency (assay),
degradants, pH, preservatives, injection mechanics, and physical appearance.

RESULTS
The 122 drug products evaluated by this study
were categorized into five groups (see Tab. 1)
based on shelf life extension data (relative
number of lots initially extended and number of
extended lots terminated). For the products
assigned to Groups 1 and 2, all lots were extended
beyond their original expiration date. The products assigned to Groups 3, 4, and 5 had some lots
that were denied initial extension. Overall, 2650
(88%) of the 3005 lots were extended past
their original expiration date for an average of
66 months. Of these, 934 lots (35%) are still
active and were granted an average extension of
62 months, 1237 lots (47%) were abated before
failure (dormant) after an average extension of
70 months, and 479 lots (18%) were terminated
due to failure after an average extension of
65 months. Of the 479 lots that eventually failed,
no lots failed before 1 year and 312 lots were
extended beyond 4 years.
The 63 products assigned to Group 1 were
further divided into subclasses based on the
number of lots tested for each product. For
these products, none of the extended lots were
terminated due to failure (periodic retest). Each of
the 15 products assigned to Group 1A (Tab. 2) had
at least 10 lots evaluated. Of the 466 lots tested,
284 lots (primarily 205 lots of ciprofloxacin tablets)
are still active in the program and available for
use. The average extension for these active lots
was 52 months. The remaining 182 lots were not
further tested or extended and are categorized as
dormant. The average extension for these dormant
lots was 67 months. The average extension time for

Table 1. Group Assignment

Group
1A
1B
1C
2
3
4
5
Total

Number of Number of Lots Total Lots Number of Lots Number of Extended Average Extension Time
Products
per Product
Tested Initially Extended
Lots Terminated
for Extended Lots (mo.)
15
25
23
20
13
16
10
122

DOI 10.1002/jps

10–242
5–9
3–4
3–41
3–169
8–687
2–21
2–687

466
164
85
254
278
1675
83
3005

All
All
All
All
Most (256)
Most (1402)
 50% (23)
2650

None
None
None
Some (41)
None
Some (431)
Some (7)
479

58
65
55
58
44
76
38
66

JOURNAL OF PHARMACEUTICAL SCIENCES, VOL. 95, NO. 7, JULY 2006

 1552

LYON ET AL.

Table 2. Group 1A: Products With No Failures ( 10 Lots tested)
Number of Lots
Drug Product

Dosage Form

Amoxicillin sodium
Ciprofloxacin
Diphenhydramine HCl
Doxycycline hyclate
Doxycycline hyclate
Halothane
Mannitol
Morphine sulfate
Naloxone HCl
Oxacillin sodium
Potassium iodide
Sodium bicarbonate
Sodium chloride
Sodium nitrite
Sodium thiosulfate

Tablets
Tablets
Syringe-needlea
Capsulesb
Powderb
Liquid
Injection-solution
Syringe-needlec
Injection-solution
Powder
Tabletsd
Injection-solution
Irrigatione
Injection-solution
Injection-solution

Extension Time (mo.)

Tested

Dormant

Active

Mean

Range

21
242
12
13
31
12
10
13
10
13
12
37
16
10
14

0
37
12
3
1
12
10
11
10
13
0
37
16
7
13

21
205f
0
10g
30h
0
0
2i
0
0
12
0
0
3j
1k

23
55
76
50
27
67
66
89
77
56
69
55
72
89
131

22–23
12–142
33–126
37–66
14–52
51–92
21–109
35–119
60–95
28–116
28–184
14–101
40–108
35–180
24–151

a

See ‘‘spray’’ dosage form (Group 5).
See ‘‘tablet’’ dosage form (Group 3).
c
See ‘‘autoinjector’’ (Group 2) and ‘‘injection-solution’’ (Group 3) dosage forms.
d
See ‘‘granules’’ dosage form (Group 1B).
e
See ‘‘injection-solution’’ dosage form (Group 2).
f
Active range: 15–142 months.
g
Active range: 29–66 months.
h
Active range: 14–35 months.
i
Active range: 35–54 months.
j
Active range: 49–145 months.
k
Active range: 49 months.
b

all of the lots from these 15 products was
58 months, ranging from 12 to 184 months. The
mean extension times and the ranges of extension
times for the individual products are provided in
the table. For six of the products the extension
times represent a combination of the active and
dormant lots. In these cases, the range for the
active lots is annotated in the table. Other dosage
forms of diphenhydramine HCL, doxycycline
hyclate, morphine sulfate, potassium iodide, and
sodium chloride were assigned to other groups.
The 25 products assigned to Group 1B (Tab. 3)
had 5–9 lots evaluated. Of the 164 lots tested, 143
lots are dormant and 21 lots are still active. The
average extension for the dormant lots was
61 months and for the active lots was 92 months.
The average extension time for all of the lots from
these 25 products was 65 months, ranging from
15 to 278 months. Other dosage forms of ampicillin, cimetidine HCl, ciprofloxacin, potassium
iodide, and povidone-iodine were assigned to other
groups.
The 23 products assigned to Group 1C (Tab. 4)
had only 3–4 lots evaluated. All the 85 lots tested

are now dormant. The average extension for these
dormant lots was 55 months, ranging from 12 to
114 months. The tablet dosage form of chlorpromazine HCl was assigned to Group 2.
Each of the 20 products assigned to Group 2
(Tab. 5) had some lots that could not be repeatedly
extended, one attribute eventually failing during
periodic retest. Of the 254 lots tested, 41 lots
eventually failed and were terminated, 211 lots are
now dormant and 2 lots are currently active. For
the terminated lots, the amount of shelf life
extension before the failure ranged from 12 to
103 months with an average of 53 months.
Termination was due to a variety of test failures
as indicated in the table. The average extension for
the 211 dormant lots was 59 months ranging from
15 to 137 months. The mean extension times and
the ranges of extension times indicated in the table
represent a combination of the terminated, dormant, and active lots. The average extension time
for these combined lots was 58 months, ranging
from 12 to 137 months. The range of extension
times for the two active lots is annotated in the
table. Other dosage forms of ampicillin sodium,

JOURNAL OF PHARMACEUTICAL SCIENCES, VOL. 95, NO. 7, JULY 2006

DOI 10.1002/jps

 STABILITY PROFILES OF DRUG PRODUCTS

1553

Table 3. Group 1B: Products With No Failures (5–9 Lots Tested)
Number of Lots
Drug Product
Ampicillin
Amyl nitrite
Atropine sulfate-pralidoxime chloride
Calcium chloride
Calcium glucepate
Cephalexin
Cimetidine HCl
Ciprofloxacin
Dexamethasone sodium phosphate
Enflurane
Ephedrine sulfate
Fentanyl citrate
Guaifenesin
Hetastarch in sodium chloride
Hexachlorophene cleansing
Iothalamate meglumine
Ketamine HCl
Mebendazole
Meperidine HCl
Phenytoin sodium
Potassium iodide
Povidone-iodine
Promethazine HCl
Triamterene and hydroclorothiazide
Undecylenic Acid and zinc salt

Dosage Form
a

Capsules
Inhalant
Autoinjector
Injection-solution
Injection-solution
Capsules
Tabletsb
Suspensionc
Syringe-needle
Liquid
Injection-solution
Injection-solution
ER Tablets
Injection-solution
Emulsion
Injection-solution
Injection-solution
Tablets
Injection-solution
Injection-solution
Granulesd
Ointmente
Injection-solution
Capsules
Powder

Tested
5
6
5
8
8
6
5
7
7
8
5
6
7
5
8
7
6
8
6
5
5
7
9
6
9

Dormant
3
6
0
8
8
6
5
0
7
8
5
6
5
5
8
7
6
8
6
5
0
7
9
6
9

Extension Time (mo.)
Active
f

2
0
5
0
0
0
0
7
0
0
0
0
2g
0
0
0
0
0
0
0
5
0
0
0
0

Mean

Range

49
59
31
81
49
57
67
32
61
48
46
84
85
44
81
51
64
58
89
63
254
65
51
19
68

22–64
37–76
25–38
66–106
23–82
28–135
59–75
25–40
24–93
15–94
21–80
70–96
39–122
30–61
58–106
20–78
42–87
28–89
32–128
29–100
225–278
35–134
28–73
18–19
43–82

a

See ‘‘injection-solution’’ dosage form (Group 2).
See ‘‘injection-solution’’ dosage form (Group 2).
c
See ‘‘tablet’’ dosage form (Group 1A).
d
See ‘‘tablet’’ dosage form (Group 1A).
e
See ‘‘solution’’ dosage form (Group 4).
f
Active range: 22–23 months.
g
Active range: 114–122 months.
b

chlorpromazine HCl, cimetidine, morphine sulfate, and sodium chloride were assigned to other
groups.
Each of the 13 products assigned to Group 3
(Tab. 6) had most lots extended initially ( 50%
occurrence) and none of the extended lots were
terminated due to failure (periodic retest). Of the
278 lots tested, 256 lots were extended and 22 lots
were denied extension. Of the extended lots, 81 lots
are now dormant and 175 lots are currently active
(primarily 159 lots of doxycycline hyclate tablets).
The average extension for these dormant lots was
74 months and the average extension for these
active lots was 30 months. The average extension
time for all of the lots from these 13 products was
44 months, ranging from 12 to 216 months. For
three products, the extension times represent a
combination of the active and dormant lots. In
these cases, the range for the active lots is

annotated in the table. The reasons for denying
the extension of the 22 lots was due to a variety of
test failures as indicated in the table. Other dosage
forms of atropine sulfate, doxycycline hyclate, and
morphine sulfate were assigned to other groups.
Each of the 16 products assigned to Group 4
(Tab. 7) had most lots extended initially ( 50%
occurrence) and some of the extended lots were
terminated due to failure (periodic retest). Of the
1675 lots tested, 1402 lots were extended and 273
lots were denied extension. Extension denials were
due to a variety of test failures as indicated in the
table. Of the extended lots, 431 lots eventually
failed and were terminated, 519 lots now dormant
and 452 lots are currently active (primarily 119 lots
of atropine sulfate autoinjectors and 188 lots of
pralidoxime chloride autoinjectors). For the terminated lots, the amount of shelf life extension before
the failure ranged from 12 to 266 months with an

DOI 10.1002/jps

JOURNAL OF PHARMACEUTICAL SCIENCES, VOL. 95, NO. 7, JULY 2006

 1554

LYON ET AL.

Table 4. Group 1C: Products With No Failures (3–4 Lots Tested)
Number of Lots
Drug Product

Dosage Form

Acetaminophen pseudophedrine
Benzonatate
Bretylium tosylate
Bupivacaine HCl
Ceftriaxone sodium
Chloroquine HCl
Chlorpromazine HCl
Dextrose 10%
Dextrose and sodium chloride
Dobutamine HCl
Edrophonium chloride
Erythromycin lactobionate
Hydrocortisone sodium succinate
Mafenide acetate
Mepivacaine HCl
Naproxen
Neostigmine methylsulfate
Penicillin G benzathine
Phenylephrine HCl
Prochloroperazine edisylate
Protamine sulfate
Sulfisoxazole
Tubocurarine chloride

Capsules
Capsules
Injection-solution
Injection-solution
Powder
Injection-solution
Injection-solutiona
Injection-solution
Injection-solution
Injection-solution
Injection-solution
Powder
Injection-solution
Cream
Cartridge-needle
Tablets
Injection-solution
Suspension
Injection-solution
Injection-solution
Powder
Tablets
Injection-solution

Extension Time (mo.)

Tested

Dormant

Active

Mean

Range

3
4
4
3
4
4
3
4
4
3
4
4
3
3
3
4
4
4
4
4
4
4
4

3
4
4
3
4
4
3
4
4
3
4
4
3
3
3
4
4
4
4
4
4
4
4

0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0

24
44
49
88
60
64
74
25
64
47
65
60
43
59
41
52
60
70
60
43
64
56
59

24–24
12–73
15–71
79–95
44–69
27–98
59–88
23–29
51–73
29–79
33–114
38–83
37–56
56–63
33–45
46–62
31–78
61–84
53–78
28–66
57–77
45–68
47–69

a

See ‘‘tablet’’ dosage form (Group 2).

average of 67 months. Termination was due to a
variety of test failures as indicated in the table. The
range of extension times for the active lots is
annotated in the table. The average extension for
these dormant lots was 81 months and the average
extension for these active lots was 79 months. The
average extension time for all of the lots (terminated, dormant, and active) from these 16 products
was 76 months, ranging from 12 to 266 months. The
ointment dosage form of povidone-iodine was
assigned to Group 1B.
Each of the 10 products assigned to Group 5
(Tab. 8) had most of the lots denied initial
extension ( 50% occurrence) and some of the
extended lots were terminated due to failure
(periodic retest). Of the 83 lots tested, 23 lots were
extended and 60 lots were denied extension.
Extension denials were due to a variety of test
failures as indicated in the table. Of the extended
lots, 7 lots eventually failed and were terminated,
16 lots are now dormant, and none of the lots are
currently active. For the terminated lots, the
amount of shelf life extension before the failure
ranged from 17 to 94 months with an average of
49 months. Termination was due to a variety of test

failures as indicated in the table. The average
extension for the 16 dormant lots was 33 months.
The average extension time for all of the lots
(terminated and dormant) from these 10 products
was 38 months, ranging from 14 to 94 months. For
the individual products, the mean extension times
and the ranges of extension times in the table
represent a combination of the terminated and
dormant lots. The spray dosage form of diphenhydramine HCl was assigned to Group 1A.

JOURNAL OF PHARMACEUTICAL SCIENCES, VOL. 95, NO. 7, JULY 2006

DOI 10.1002/jps

DISCUSSION
Performance Evaluation
Each product tested was assigned to one of five
groups (as listed in Tab. 1) based on an explicit
classification system. Products with no failures
(initial extension failure or termination failure)
were assigned to Group 1. This does not imply
that these products will be stable indefinitely.
Based on past performance, these products may
be considered the best candidates for shelf life
extension. Other factors, including the number of

 DOI 10.1002/jps

41
3
9
37
11

Injection-solutione
Powder
Powder
Cream
Capsules
Powder

Sodium chloride
Sodium polystyrene sulfonate
Succinylcholine chloride
Sulfadiazine silver
Tetracycline HCl

Thiopental sodium

12

b

9

31
2
8
36
7

2
6
0
3
5
12
10

7
3
22
10
14
6
18

Dormant

See ‘‘capsule’’ dosage form (Group 1B).
See ‘‘injection-solution’’ dosage form (Group 1C).
c
See ‘‘tablet’’ dosage form (Group 1B).
d
See ‘‘syringe-needle’’ (Group 1A) and ‘‘injection-solution’’ (Group 3) dosage forms.
e
See ‘‘irrigation’’ dosage form (Group 1A).
f
Active range: 110–114 months.
g
Turbidity.
h
Color.
i
Color and integrity.
j
Pseudomorphine.
k
Ca, Na, K.
l
Na, K, lactate.
m
4-Epianhydrotetracycline.
n
Particulates.

a

3
7
3
4
6
13
13

Capsules
Injection-solution
Autoinjectord
Syringe-needle
Solution
Injection-solution
Injection-solution

Flurazepam HCl
Furosemide
Morphine sulfate
Metaraminol bitartrate
Ophthalmic irrigating
Pancuronium bromide
Ringer’s, lactated and dextrose

8
4
23
13
15
7
22

Injection-solutiona
Powder
Dermal
Powder
Tabletsb
Injection-solutionc
Injection-solution

Ampicillin sodium
Cefoperazone sodium
Cellulose, oxidized, regenerated
Cephapirin sodium
Chlorpromazine HCl
Cimetidine HCl
Dextrose (5%)

Tested

Dosage Form

Drug Product

Number of Lots

0

0
0
0
0
0

0
0
0
0
0
0
0

0
0
0
2f
0
0
0

Active

54

50
55
72
57
50

35
57
32
40
52
79
53

57
46
79
74
52
42
65

Mean

23–96

12–113
45–74
58–95
28–104
17–133

27–44
31–90
29–37
33–47
19–77
54–108
20–87

29–87
25–57
28–137
50–114
23–78
15–67
13–128

Range

Extension Time (mo.)

Table 5. Group 2: Products With All Lots Extended Initially: Some Termination Failures for Extended Lots

Appearance g (1) [42]
Assay (1) [25]
Appearanceh (1) [28]
Potency (1) [98]
Appearancei (1) [78]
Assay (1) [15]
Assay (2) [77 and 80]; appearancei
(2) [84 and 84]
Appearanceh (1) [27]
Low pH (1) [31]
Degradantj (3) [29, 30, and 37]
Assay (1) [47]
Low pH (1) [19]
Assay (1) [85]
Assayk (1) [55]; assayl
(1) [72]; appearanceh (1) [20]
Assay (9) [12–103]; pH (1) [38]
Water content (1) [45]
Purity (1) [51]
High pH and assay (1) [53]
Dissolution (2) [17 and 20];
degradantm (2) [38 and 54]
Low pH (2) [73 and 73];
appearancen (1) [56]

Termination Failure (Number of
Lots) [Extension Time, no.]

STABILITY PROFILES OF DRUG PRODUCTS

1555

JOURNAL OF PHARMACEUTICAL SCIENCES, VOL. 95, NO. 7, JULY 2006

 JOURNAL OF PHARMACEUTICAL SCIENCES, VOL. 95, NO. 7, JULY 2006

12
8
4

Tablets
Suspension
Ophthalmic ointment

8
7
3

2
24
8
7
166
2
14
2
9
4

Extended

* 50% Occurrence.
a
See ‘‘autoinjector’’ dosage form (Group 4).
b
See ‘‘capsule’’ and ‘‘powder’’ dosage forms (Group 1A).
c
See ‘‘syringe-needle’’ (Group 1A) and ‘‘autoinjector’’ (Group 2) dosage forms.
d
Active range: 19–201 months.
e
Active range: 15–91 months.
f
Active range: 41–57 months
g
4-Epidoxycycline.
h
Particulates.
i
Bactracin assay.

3
27
10
9
169
3
15
3
10
5

Injection-solution
Injection-solutiona
Powder
Tablets
Tabletsb
Tablets
Injection-solution
Suspension
Injection-solutionc
Ophthalmic ointment

Atracurium besylate
Atropine sulfate
Cefazolin sodium
Codeine sulfate
Doxycycline hyclate
Enalapril maleate
Lidocaine HCl
Methylprednisone acetate
Morphine sulfate
Neomycin and polymyxin B
sulfates and bactracin zinc
Primaquine phosphate
Spectinomycin HCl
Sulfacetamide sodium

Tested

Dosage Form

5
7
3

2
11
8
7
7
2
14
2
9
4

Dormant

Number of Lots

29
101
82
89
27
34
58
38
79
28
55
83
39

3f
0
0

Mean

0
13d
0
0
159e
0
0
0
0
0

Active

41–80
55–109
35–44

27–30
19–216
63–110
16–114
15–91
27–42
28–126
25–51
21–115
12–40

Range

Extension Time (mo.)

Group 3: Products With Most * Lots Extended Initially; No Termination Failures for Extended Lots

Drug Product

Table 6.

Assay (4)
Assay (1)
Assay (1)

Assay (1)
Low pH (2); Assay (1)
Low pH (2)
Dissolution (2)
Assay (1); impurityg (2)
Assay (1)
Assay (1)
Appearanceh (1)
Appearanceh (1)
Assayi (1)

Initial Extension
Failure—Reason for
Denial of Extension
(# Lots)

1556
LYON ET AL.

DOI 10.1002/jps

 DOI 10.1002/jps

Tablets

Sulfadoxine and pyrimethamine

* 50% Occurrence.
a
See ‘‘ointment’’ dosage form (Group 1B).
b
Active range: 13–129 months.
c
Active range: 21–86 months.
d
Active range: 12–100 months.
e
Active range: 19–228 months.
f
Active range: 23–186 months.
g
Active range: 24–102 months.
h
Active range: 61–93 months.
i
Brown precipitate.
j
Turbidity.
k
Color.
l
Carbostyril.
m
Crystals.
n
Particulates.
o
RO 1–5237.
p
Calcium.
q
Sodium.
r
Pyrimethamine dissolution.
s
Integrity.

Injection-solution

Autoinjector

Pralidoxime chloride

Ringers, lactated

Cartridge-needle
Injection-solution
Powder
Solutiona

Epinephrine
Heparin sodium
Penicillin G
Povidone-iodine

Powder
Tablets

Syringe-needle

Diazepam

Pralidoxime chloride
Pyridostigmine bromide

Autoinjector

Diazepam

38

Tablets

Injection-solution

Chloroquine phosphate

Clindamycin phosphate

10

Powder

Cefoxitin sodium

8

59

80
152

412

33
16
15
20

35

67

31

12
687

Tablets
Autoinjector

Aluminum acetate
Atropine sulfate

Tested

Dosage Form

7

56

78
141

399

17
14
14
16

25

66

25

36

5

10
495

Extended

4

54

37
91

145

16
11
13
14

16

10

15

10

3

6
74

Dormant

Number of Lots

2

h

0

39f
40g

67

52

88
61

120

188e

53

63

44

40

24

22
52
49
74

d

c

52
57

Mean

0
0
0
0

0

39

0

25

0

0
119b

Active

34–93

23–125

23–186
19–143

19–266

17–24
22–82
22–95
29–144

12–105

12–100

18–77

20–86

24–55

16–70
12–135

Range

Extension Time

Appearancek (6)
Assay (4)
Assay (1)
Appearancen (1)
Injector misfire (1)
Appearancek (1); assay (1)
Assay (11)
Impurityo (2)
Assayp (2)
Assayq (1)
Dissolutionr (1)

Appearancem (8)
Degradantl (2)
Assay (16)
Low pH (1); appearancek (1)
Appearancek (1)
Assay (4)

Effervesce (2)
Appearancei (102)
Appearancej (20); appearancek (3)
Assay (57); coring (1)
Low phenol (7)
Assay (1)
Low pH (1)
Potency (4)
Appearancek (1)
Dissolution (1)
Assay (1)
Assay (5)
Assay (1)
Degradantl (1)

Initial Extension
Failure (# Lots)

Group 4: Products With Most* Lots Extended Initially; Some Termination Failures for Extended Lots

Drug Product

Table 7.

Dissolutionr (1) [58]

Assay (9) [24–68]
Appearancek (1) [43]
Degradantl (16) [23–79]
pH (1) [84]
Appearancem (3) [12–67]
Degradantl (6) [15–102]
Assay (1) [18]
Appearancek (3) [22–47]
Appearancek (1) [26]
Appearancek (1) [29]
Appearancei (1) [64]
Assay (39) [35–266]
Appearancek (10) [96–149]
Low pH (13) [88–168]
Injector misfire (3) [43–169]
Low phenol (1) [125]
Assay (1) [99]; pH (1) [163]
Assay (7) [19–88]
Appearances (1) [93]
Assayp (2) [52; 53]

Dissolution (1) [20]

Potency (2) [38–40]

Effervesce (4) [58–70]
Appearancei (143) [12–135]
Appearancej (11) [12–124]
Assay (95) [22–113]
Low phenol (52) [28–126]
Coring (1) [43]

Termination Failure
(# Lots) [Range, mo.]

STABILITY PROFILES OF DRUG PRODUCTS

1557

JOURNAL OF PHARMACEUTICAL SCIENCES, VOL. 95, NO. 7, JULY 2006

 JOURNAL OF PHARMACEUTICAL SCIENCES, VOL. 95, NO. 7, JULY 2006

Injection-solution

Physostigmine salicylate

* 50% Occurrence.
a
See ‘‘syringe-needle’’ dosage form (Group 1A).
b
Ergotamine assay.
c
Caffeine dissolution.
d
Caffeine and ergotamine dissolution.
e
Epinephrine.
f
Particulates.

Injection-solution
Injection-solution
Injection-solution
Tablets
Powder
Cartridge-needle

Isoproterenol HCl
Levarterenol bitartrate
Lidocaine HCl and epinephrine
Mefloquine HCl
Penicillin G procaine
Phenobarbital sodium

Albuterol
Inhalant
Diphenhydramine HCl
Spraya
Ergotamine tartrate and caffeine Tablets

Dosage Form

14

8
8
9
21
7
4

2
2
8

4

2
1
1
7
2
2

0
0
4

3

1
0
1
5
2
0

0
0
4

0

0
0
0
0
0
0

0
0
0

Tested Extended Dormant Active

Number of Lots

Group 5. Products With Most* Lots Failing Initial Extension

Drug Product

Table 8.

31

45
22
29
36
70
56

na
na
24

Mean

21–44

37–53
22
29
17–94
67–72
32–79

na
na
14–35

Range

Extension Time (mo)

Low pH (10)

Assay (2)
Assay (2)
Assayb (2)
Dissolutionc (1)
Dissolutiond (1)
Assay (6)
Assay (7)
Assaye (8)
Dissolution (14)
Assay (5)
Assay (2)

Initial Extension
Failure (# Lots)

Assay (1) [53]
Assay (1) [22]
na
Dissolution (2) [17; 94]
na
Assay (1) [32]
Appearancef (1) [79]
Assay (1) [44]

na
na
na

Termination Failure
(# Lots) [Range, mo.]

1558
LYON ET AL.

DOI 10.1002/jps

 STABILITY PROFILES OF DRUG PRODUCTS

1559

lots tested and the minimum extension time
surpassed by all lots tested, need to be considered
in evaluating the performance of these products.
Group 1 was divided into 3 subgroups based on
the number of lots tested. A greater confidence in
the ability to extend the shelf life was associated
with Group 1A ( 10 lots tested). The most prolific
performer from this subgroup was ciprofloxacin
tablets (242 lots tested with 205 lots still active).
Two other top performers from this group were
naloxone HCl (all 10 lots extended at least 5 years)
and halothane (all 12 lots extended over 4 years).
The top performer in Group 1B was potassium
iodide granules with all five lots still active nearly
20 years after the original expiration date.
In addition, all of the lots of calcium chloride
injection-solution and fentanyl citrate injectionsolution were extended more than 5 years. All of
the lots of two products from Group 1C (bupivacaine HCl injection-solution and penicillin G
benzathine suspension) were extended more than
5 years.
Although Group 2 had some lots that encountered termination failure, some of these products
could be considered quite reliable. Examples from
this group indicate that if products are stored long
enough, failures are bound to occur. Of the products
with  10 lots tested, notable examples are listed.
All of the cephapirin sodium powder lots (13 tested)
were extended at least 4 years with only one
termination, occurring at 8 years and all of the
pancuronium bromide injection-solution lots (13
tested) were extended at least 4 years with only one
termination, occurring at 7 years. In addition,
sulfadiazine silver cream (37 lots tested) had only
one termination, occurring after 4 years and
dextrose injection-solution (23 lots tested) had four
terminations, but all occurred after 8 years.
The products assigned to Group 3 had no
termination failures, but for each product one to
four lots failed initial extension, exemplifying lotto-lot variability. In some cases the extension
times were quite long. Although two lots of
cefazolin sulfate powder failed initial extension,
the eight remaining lots were extended more than
5 years. One lot of spectinomycin sodium suspension failed initial extension, but the seven remaining lots were extended more than 4 years.
The products assigned to Group 4 had some lots
exhibiting initial extension failure and some lots
exhibiting termination failures. The lot-to-lot
variability and the need for systematic testing
were most evident with this group. This group
includes products that have been extensively

tested. The combination of lots from atropine
sulfate autoinjectors (687 tested), pralidoxime
chloride autoinjectors (412 tested), and pyridostigmine bromide tablets (152 tested) represents
42% of the total lots summarized in this report.
Despite the failures, this group was relatively
successful, considering 1404 (84%) of the 1675 lots
tested were initially extended. Diazepam autoinjectors (67 lots tested) had only one lot that failed
initial extension and pralidoxime chloride powder
(80 lots tested) had only two lots that failed initial
extension. Of the extended pralidoxime chloride
powder lots, only two were terminated, but these
two lots were extended for more than 8 years.
Other products were successfully extended including pralidoxime chloride autoinjectors (97% of lots
tested), lactated Ringers injection solution (95%,
59 lots tested), and pyridostigmine bromide tablets
(93% of lots tested). The product exhibiting the
greatest lot-to-lot variability was atropine sulfate
autoinjectors. Hundred and ninety two lots (28%)
failed initial extension. The range of extension
times for lots of atropine sulfate autoinjectors
exhibiting termination failures was 1–11 years.
For each of the 10 products assigned to Group 5,
most of the lots could not be extended. Lot-to-lot
variability was also evident with this group. Five of
the seven penicillin G procaine powder lots could
not be extended, while the other 2 lots were
extended for more than 5 years. Fourteen of 21
mefloquine HCl tablets lots could not be extended,
1 extended lot was terminated at 17 months, while
1 lot was extended for more than 7 years. Of the 83
lots tested, only 28% could be initially extended.
All of the products in this group were considered
poor candidates and were discontinued from the
shelf life program.

DOI 10.1002/jps

JOURNAL OF PHARMACEUTICAL SCIENCES, VOL. 95, NO. 7, JULY 2006

Opportunity for Failures
As more lots were tested, the chances that a lot
would fail increased. This was reflected in the
group classifications. Of the 27 products with 15 or
more lots tested, 13 products are assigned to Group
4 compared to only 5 products assigned to Group
1A. Sulfadiazine silver, now assigned to Group 2,
did not have any failures during the first 11 years
on the program. Up until the termination failure of
the 27th lot tested, this product was classified as
Group 1A. Diazepam autoinjectors, now assigned
to Group 4, did not have an initial extension failure
during the first 6 years on the program. Up until
the initial extension failure of the 48th lot tested,
this product was classified as Group 2. Doxycycline

 1560

LYON ET AL.

tablets, now assigned to Group 3 did not have an
initial extension failure during the first 6 years on
the program. Up until the initial extension failure
of the 13th lot, this product was classified as Group
1A. This illustrates that the group classifications
may be dependent on the number of lots tested.
This also supports the subclassification of Group
1 into 1A, 1B, and 1C.

study could be used to select promising potential
candidates for an extension program.

ACKNOWLEDGMENTS
The authors are grateful to Christopher D. Ellison
of the Division of Product Quality Research for
editorial assistance.

Other Programs
The SLEP has been successful in helping to
maintain drug reserves and reducing costs for
the US Military. It is expected that similar
programs will be implemented to preserve regional and national pharmaceutical stockpiles. One
example is the SNS, formerly the National
Pharmaceutical Stockpile, managed jointly by
the Department of Homeland Security and the
Department of Health and Human Services.10
The SNS Program ensures that the medical
material stock is rotated and kept within potency
shelf-life limits. The FDA has issued a shelf life
extension guidance for federal agencies and state
and local governments. A specific guidance was
issued by CDER presenting FDA’s recommendations on testing to extend the shelf life of stockpiled potassium iodide.11 This guidance describes
how to identify laboratories suitable to conduct
the tests, how to notify holders of stockpiled drug
products and end users about changes in expiration date, and how to distinguish stockpiled
batches with different expiration dates.

REFERENCES

The SLEP data supports the assertion that many
drug products can be extended past the original
expiration date, but this additional stability
period is highly variable. Due to the lot-to-lot
variability, the stability and quality of extended
drug products can only be assured by periodic
testing and systematic evaluation of each lot. The
results of this stability program can only be
related to products that have been carefully
stored in their original sealed container closures.
The class groupings indicate past stability performance and do not guarantee future performance.
The classification of products presented in this

1. Woods M. Drugs may outlast label date. Post-Gazette
National Bureau. May 30, 2005. Available at: http://
www.post-gazette.com/pg/05150/512789.stm.
2. Cohen LP. Many medicines prove potent for years
past their expiration dates. The Wall Street
Journal. March 29, 2000.
3. Garamone J. Program extends drug shelf-life.
American Forces Press Service. March 29, 2000
Available at: http://www.defenselink.mil/news/Mar
2000/n03292000_20003292.html.
4. American Medical Association. ‘‘Pharmaceutical
Expiration Dates’’. Report 1 of the Council on Scientific Affairs (A-01). July 25, 2001. Available at:
http://www.ama-assn.org/meetings/public/annual01/
csa_reports.pdf#search ¼ ‘Pharmaceutical%20Expiration%20Dates’.
5. Stark G, Fawcett JP, Tucker IG. 1997. A study of the
stability of some commercial drug dosage forms
beyond their expiration dates. Pharm J 258:637–640.
6. Regenthal R, Stefanovic D, Albert T, Trauer H,
Wolf T. 2002. The pharmacologic stability of 35year old theophylline. Hum Exp Toxicol 21:343–
346.
7. Scholtissek C, Webster RG. 1998. Long-term
stability of the anti-influenza A compounds—
amantadine and rimantadine. Antiviral Res 38:
213–215.
8. International Conference on Harmonization: ‘‘Guidance for Industry Q1A(R2) Stability Testing of
New Drug Substances and Products.’’ Available at:
http://www.fda.gov/cder/guidance/5635fnl.pdf.
9. DOD-FDA Shelf Life Extension Program website.
Available at: http://www.usamma.army.mil/html/
dodshelf.cfm.
10. Strategic National Stockpile website. Available at:
http://www.bt.cdc.gov/stockpile/.
11. FDA: ‘‘Guidance for Federal Agencies and State
and Local Governments Potassium Iodide Tablets
Shelf Life Extension.’’ Available at: http://www.
fda.gov/cder/guidance/5666fnl.pdf.

JOURNAL OF PHARMACEUTICAL SCIENCES, VOL. 95, NO. 7, JULY 2006

DOI 10.1002/jps

CONCLUSION