Alternative Production Methods for WFI and Compendial Challenges

Anthony Bevilacqua, Ph.DMettler-Toledo Thornton

April 5, 2011

Aqua Distillata (Distilled Water)

Aqua distilleteur vasis permundis, donec ejus duo circiter trientes stillaverint. Aquam distillatum in lagena vitrea servato.

“Let water be distilled in very clean vessels until about two thirds have come over, which is to be kept in a glass bottle.”

Printed in USP I (1820) by thePharmacopoeia of the United States of America

USP History of WFI (continued)

Based on data from a user, USP added RO as an acceptable WFI production method in USP XIX (1975)• In spite of allowance, RO not widely adopted• Many early design problems and microbial control issues,

spotty success, validation difficulties

Many advances in last decade - overcome historic problems• Less problematic RO designs• Other emerging technologies and heat-tolerant MOCs• Better scientific understanding of endotoxin removal

approaches• Better scientific understanding of biofilm control

USP Water for Injection Monograph

WFI Method of Production – JP

WFI Method of Production – EP

Academic vs. Engineering Argument

Academic - “Distillation is the gold standard, it works by inherently better separation methods”

Engineer – “Don’t tell me how to make it, tell me the specifications you need it to meet”

Water for Injection is the ONLY material in USP where it is prescribed “how to make it”.

EMEA Reflection Paper of WFI

http://www.emea.europa.eu/pdfs/vet/qwp/2827108en.pdf

EMEA Microbiological Concerns

EMEA Microbiological Concerns

EMEA Conclusions

Response in PDA Letter- Vol XLIV (7) pp 15-16

Multiple responses (100%) reflecting a very negative opinion of specific comments of the EMEA position

But comments were qualitative, not quantitative

Comments indicated more data was needed

EDQM – Revisits Their Position

EDQM Survey

EDQM Survey

Results of survey are planned to be released in early 2011.

USP Pharm Water Perspective

It is documented that there are Highly Purified Water systems in operation today• by definition they do meet WFI requirements for

chemical and microbiological tests

It is “known” that there are Purified Water systems in operation today that meet WFI requirements – but not proven

USP Pharm Water Perspective

It is not the final purification step that determines the water quality…

It is not RO that is good or bad…

…it is the entire water system and related supporting processes that determine the water quality

It is NOT about RO, it is about the system.

What Specifically?

Feedwater and microbial control methods

Pre-treatment system design/sanitization

Purification system design/sanitization

System size, subloops, and distribution

Water criticality relative to product

What Else?

Sanitization methods• Type(s)• Frequency

In-line microbial/endotoxin/purity controls

Hours of operation/days of operation

Usage – intermittent or continuous

Cost – capital and operational

And more

USP Survey

Since it is the water system AND related processes…

Let’s get data from industry to justify whether alternative techniques can produce WFI reliably and consistently

Utilize the ISPE Critical Utilities COP to get the best available data.

EDQM Workshop

USP’s WFI Survey Needs to Reveal Factors Facilitating Reliable WFI Production Using Non-distillation Approaches

Need to understand source water quality and consistency

Need to understand purification & distribution system • Design• Maintenance• Sanitization approaches• Ongoing purification and microbial/endotoxin controls• Excursion restoration approaches• Specifications, Action, Alert and Typical Levels

Need to understand system reliability and consistency

Hoped to see survey responses representing long term system success

WFI Survey probed attributes of these WFI systems

Expected to see a few clusters of successful design configurations and approaches to purification, sanitization, controls – some surprises!

Q1. For non-distillation based endotoxin controlled system, the level of endotoxin/microbial control it achieves is (check all that apply)

0%

5%

10%

15%

20%

25%

30%

35%

40%

45%

Intended result ofdesign/maintenance

focused onendotoxin and

microbial control

Unintended result ofdesign/maintenancefocused on microbial

control

Unintended result ofdesign/maintenance

focused onendotoxin control

An unintendedconsequence of

system design andmaintenance

focused on chemicalpurity control.

Consistent attributeincluded in theoriginal water

system validation

Not included in theoriginal validation,

but historicallyproven, and laterincluded in water

system revalidation.

Not included in watersystem validation,but consistently

remained at/below adesired level

80 respondents/88 responses

Q2. The endotoxin levels from this water system meet...

0%

10%

20%

30%

40%

50%

60%

WFI specifications allthe time

WFI specificationsmost of the time withoccasionally higher

levels

Looser entodoxinspecifications than WFI

Tighter endotoxinspecifications than WFI

No endotoxin controlexpectations

63 respondents

The endotoxin levels from this water system meet...

Looser entodoxin specifications than

WFI, 0.00%

Tighter endotoxin specifications than

WFI, 15.87%

WFI specifications most of the time with occasionally higher

levels, 19.05%

WFI specifications all the time, 53.97%

No endotoxin control expectations, 11.11%

63 respondents

Q2a. Indicate your TYPICAL value in EU/mL

0

2

4

6

8

10

12

14

16

18

< Detection limit or"0"

<0.05 0.05-0.10 0.10-0.25 0.25 >0.25

# r

es

po

nd

en

ts

Q2a. Indicate your Endotoxin specification (EU/mL)

0

5

10

15

20

25

30

less than 0.25 0.25 more than 0.25

# r

es

po

nd

en

ts

Q5. What is the typical microbial count in cfu/100mL?

0

2

4

6

8

10

12

14

<1 <2 <5 <10 >10

# r

es

po

nd

en

ts

Q8. If you had to restore endotoxin/microbial control in the distribution system by remedial sanitization, which remedial

approaches were or would be used:

0

2

4

6

8

10

12

14

16

18

20

Neverneeded

Chemicalsanitizer

Hot water Steam Ozone None Other:Describe

# r

es

po

nd

en

ts

Minnicare for GNRs, sanitization does not remove endotoxinJust preventiveOnly preventive maintenanceHydrogen peroxide solution

8a. Are the above remedial distribution system sanitization parameters more extreme than routine sanitization parameters?

No, 22, 76%

Yes, 7, 24%

Q8b. If you had to restore endotoxin/microbial control in the finished water by remedial approaches within the purification part of the system,

indicate which remedial approaches were used.

Never needed, 7, 6%

Remedial bed backwashing, 10, 9%

Remedial bed regeneration, 9, 8%

Remedial chemical cleaning, 9, 8%

Remedial chemical sanitization, 12, 11%

Replace coarse filters, 8, 7%

Replace micro-retentive filters, 10, 9%

Replace RO membranes, 12, 11%

Other: Describe in no more than 50 words, 7, 6%

Replace valves, 3, 3%

Replace piping, 2, 2%

Remove unit operations, 2, 2%

Add unit operations, 2, 2%

Redesign water flow, 0, 0%

Replace unit operations, 0, 0%

Re-bed unit operations, 4, 4%

Remedial hot water sanitization, 17, 15%

Q9. How many years has this water system been operating with this level of endotoxin control and microbial control?

0

1

2

3

4

5

6

7

8

9

10

0-2 years 3-5 years 6-10 years >10 years

# r

es

po

nd

en

ts

Q10. What is this water grade officially called?

0%

5%

10%

15%

20%

25%

WFI WFI, USP WFI,PhEur

WFI, USPand PhEur

PurifiedWater

PurifiedWater,USP

PurifiedWater,PhEur

PurifiedWater,

USP andPhEur

PurifiedWater withEndotoxin

Control

HighlyPurifiedWater,PhEur

Anothername

31 Respondents

Low Endotoxin Purified WaterRO WaterProcess WaterWater for Bulk Sterile Operations/Final Rinse (WBSO/FR)

Q12. What is the source of the feedwater for your water system?

0%

10%

20%

30%

40%

50%

60%

70%

80%

Only from surfacewaters, including

lakes, reservoirs, andrivers

Only from groundwaters, includingwells and springs

From a mixture ofsurface and ground

waters

From the local publicdrinking water

authority

From private sourcesnot available to the

public and not underyour company’s

control

From private sourcesunder control of your

company

37 Respondents

Q15. What is the geographical area where the products made in association with this water are marketed?

US, 14, 22%

North America, 14, 22%

Europe, 20, 32%

Japan, 6, 10%

Other, 9, 14%

What’s Next?

Summary of purification and control points necessary for reliable WFI production

Discussion of survey descriptions of: • Successful purification designs• Successful purification train sanitization approaches• Successful distribution designs• Successful distribution sanitization approaches• Successful distribution in-line controls

Implications for the future of WFI production by non-distillation approaches

Logical WFI Purification and Control PointsFirst, remove endotoxins from source water using

• Size exclusion using RO and/or UF (~10KD cutoff or less)• Deionization removal (endotoxin is very negatively charged)

Then, minimize biofilm colonization within purification train for the sake of• Purification efficiency and operational cost• Avoiding release of additional endotoxins• Can control with sanitization and in-line microbial controls

Finally, control biofilm and endotoxin in the distribution system• Avoid/prevent release into distribution system from purification

train• Avoid/minimize biofilm growth and generation of endotoxins in

loop• Can control with sanitization and in-line microbial controls• Can control with endotoxin polishing in loop (e.g., UF and DI)

Survey responses for Q13 and Q14 reveal system design and control approaches

Presented here at a high level (still revealing)• Purification unit operations used• Purification train configurations used• Purification train sanitization approaches used• Distribution system sanitization approaches used• Distribution system unit operations used• Distribution system micro/endotoxin control approaches

used

Interpretation of what it all says about successful ways of making WFI other than by distillation

Unit Operations Used in Main Purification Train

0

5

10

15

20

25

ChemSani

HWSani

1 PassRO

2 PassRO

DI/EDI DI EDI UF UV 0.1 or0.2 Filt

0.1 Filt 0.2 Filt Ozone

# o

f S

ys

tem

s U

sin

g U

nit

Op

(22 Systems)

Critical Purification Approaches

0

2

4

6

8

10

12

14

16

No

Sa

ni,

DI a

lon

e

No

Sa

ni,

1P

-RO

+ D

I

Ch

em

Sa

ni,

1P

-RO

+D

I/ED

I

Ch

em

Sa

ni,

2P

-RO

+D

I/ED

I

HW

Sa

ni,

1P

-RO

+E

DI

HW

Sa

ni,

2P

-RO

+E

DI

Ch

em

Sa

ni,

1P

-RO

+D

I/ED

I + U

F

Ch

em

Sa

ni,

2P

-RO

+D

I/ED

I + U

F

Ch

em

Sa

ni,

2P

-RO

+U

F

HW

Sa

ni,

1P

-RO

+E

DI +

UF

No

Sa

ni -

Un

it O

ps

HW

Sa

ni -

Un

it O

ps

Ch

em

Sa

ni -

Un

itO

ps

With

UF

in P

ur

or

Lo

op

No

UF

in P

ur

or

Lo

op

# o

f S

ys

tem

s U

sin

g A

pp

roa

ch

(22 Systems)

Critical Purification Unit Op

Purification Train Sanitization

Approach Summary

Utilization of UF

(22 Systems)

(23 Systems)

Sanitization/Control Used in Distribution Loop

0

5

10

15

20

25

Period

ic Che

m S

ani

Period

ic HW

Sani

Contin

uous

HW

San

i

Period

ic Ste

am

Contin

uous

Tan

k O3

Period

ic Ta

nk O

3

Contin

uous

Loo

p O

3

Period

ic Lo

op O

3

Ultrap

ure

Wat

er

Loop

DI/E

DI

Loop

UV

Loop

UF

Loop

Micr

oFilt

(0.1

or 0

.2)

Loop

0.1

Filt

Loop

0.2

Filt

# o

f S

ys

tem

s U

sin

g A

pp

roa

ch

(26 Systems)

Loop Sanitization Approaches

Loop Micro/Endotoxin Polishing

(26 Systems)

Loop Sanitization Combinations

0

1

2

3

4

5

6

7

8

No Loop Sani Noted Chem Only Heat Only Ozone Only Chem + Heat Chem + Ozone Heat + Ozone

# o

f S

ys

tem

s U

sin

g A

pp

roa

ch

(26 Systems)

Loop Micro/Endo Control/Polishing Combinations

0

1

2

3

4

5

6

7

8

9

10

Nothingnoted

UPW / DIOnly

UF only UV Only MicroFiltOnly

UPW / DI+ 1 other

UF + 1other

UV + 1other

MicroFilt+ UV

MicroFilt+ 1 other(incl UV)

UPW / DI+ 2

others

UF + 2others

UV + 2others

MicroFilt+ 2

others

All 4

# o

f S

yste

ms

Usi

ng

Ap

pro

ach

(26 Systems)

Years of Reliable OperationQ9. How many years has this water system been operating with this level of

endotoxin control and microbial control?

0

1

2

3

4

5

6

7

8

9

10

0-2 years 3-5 years 6-10 years >10 years

# re

spo

nd

ents

Non-Distillation WFI Uses

Q11. Please indicate all uses of your endotoxin/microbially controlled water (Original selection categories shown)

0

5

10

15

20

Fre

qu

en

cy

Se

lec

ted

26 Respondents 228 Selections

Q11. Please indicate all uses of your endotoxin/microbially controlled water (Selection category groups shown)

0%

20%

40%

60%

80%

100%H

um

an D

rug

s

Vet

erin

ary

Dru

gs

Dev

ice/

Dia

gn

ost

ics

Ass

oci

ated

R&

D/D

rug

Dev

elo

pm

ent

No

n-P

aren

tera

l Dru

gs

(Op

hth

alm

ics,

To

pic

als,

OT

Cs,

Gen

eric

s)

AP

I, B

ulk

Bio

log

ic, E

xcip

ien

t

Cle

anin

g, D

epyr

og

enat

ion

, Fee

dw

ater

No

n-P

aren

tera

l Ap

plic

atio

ns

(No

n-

Par

ente

ral D

rug

s, D

iag

no

stic

s,A

nal

ytic

al R

eag

ent,

Lab

Wat

er)

ON

LY

No

n-P

aren

tera

l Ap

plic

atio

ns

(No

n-P

aren

tera

l Dru

gs,

Dia

gn

ost

ics,

An

alyt

ical

Rea

gen

t, L

ab W

ater

)

ON

LY

Cle

anin

g, D

epyr

o, F

eed

wat

er,

No

n-P

aren

tera

l Ap

plic

atio

ns

ON

LY

R&

D, C

lean

ing

, Dep

yro

,F

eed

wat

er, N

on

-Par

ente

rals

(N

OT

Co

mD

rug

s &

AP

I/Bu

lkB

io)

ON

LY

AP

I, B

lk B

io, E

xcip

, Cle

anin

g,

Dep

yro

, Fee

dw

ater

, No

n-P

aren

tera

ls(N

OT

Co

m/R

&D

Par

ente

rals

)

Sel

ecti

on

% (

#) b

y 26

Res

po

nd

ents

19

54

19

5

910

1

17

13

79

26 Respondents

Non-Distillation WFI Use Clusters

What Does All This Mean About Suitable Approaches to Generating WFI ?

Some general Unit Op train pattern commonalities• 1 or 2 pass RO + DI or EDI is common, but others also

work• Chem > HW > No RO+EDI sanitization, but all work• With UF ≈ Without UF for last unit op, but both work

Plain DI (no sani) without RO or UF (minority approach) also successful

Bottom Line: Many approaches can work, likely with varying degrees of diligence to maintain control, but control nevertheless

No such thing as fool-proof (even for distillation)!

What Does All This Mean About Suitable Approaches to Generating WFI ?

Concerns essentially no different than with distillation-based WFI, however…• Cost considerations, usually water not already hot, so ambient

control measures more common/creative• Micro/endo control/polishing in loop could be beneficial• Use of high water purity (minority approach) also successful in

controlling growth

Loop sanitization approach commonality• HW > Ozone ≈ Chem sanitization use, but all can work• Multiple approaches >> Single approach >> No loop sani, but

all can work

What Does All This Mean About Suitable Approaches to Generating WFI ?

Ongoing in-line micro control/endo polishing• Combo approaches >> Single approach >> No control unit ops,

but all can work• Most common combo is a UV sanitizer + MicroFiltration (0.1 or

0.2 µm rated)

Bottom Line: Many approaches can work, with the systems having more control options likely needing less diligence in keeping the quality controlled

No such thing as fool-proof!

Thank You!

Dr. Anthony Bevilacqua, Head of R&DMettler-Toledo Thornton

36 Middlesex TurnpikeBedford, MA 01730 USA

+1 781 301 8642anthony.bevilacqua@mt.com