Non-Responders to Iron Chelation Therapy John B. Porter, MA, MD, FRCP Professor, Department of Haematology University College London London, United Kingdom.

Non-Responders to Iron Chelation Therapy

John B. Porter, MA, MD, FRCP Professor, Department of Haematology

University College LondonLondon, United Kingdom

Kai-Hsin Lin, MDProfessor of Paediatrics

National Taiwan University HospitalTaipei, Taiwan

The Non-Responder

2

John B. Porter, MA, MD, FRCP

What Does “Non-Responder” Mean?

• Balance– Failure to balance transfusional iron input with excretion?– Failure to excrete more iron than transfusional iron input?

• Ferritin– Failure to decrease serum ferritin?– Failure to control serum ferritin (or LIC) to target levels?

• Myocardial iron (T2*)– Failure to control T2*?– Failure to improve T2* if abnormal?

3

What Information Do We Need to Predict “Response Rate” with a Chelation Regimen?

• The average change of a given measure in a group of patients does NOT tell you the probability of response in an individual

• This can only be predicted when the proportion of patients showing the desired effect is shown

4

What Evidence Do We Have About the Percentage of Patients Who “Respond” to

Different Regimens?

• With desferrioxamine monotherapy

• With deferiprone monotherapy

• With combination therapy of the above

• With deferasirox monotherapy

5

“Non-Response” to Desferrioxamine (DFO)

% Responders to DesferrioxamineIron Balance SC 5x/Week

7Cohen AR, et al. Blood. 2008;111:583-587.

Dose mg/kg

25 – <35

35 – <50

≥50

% of Patients in Negative Balance

Low Transfusion

<0.3 mg/kg/d

17

76

100

Medium Transfusion

0.3–0.5 mg/kg/d

43

75

86

HighTransfusion

>0.5 mg/kg/d

17

52

89

Responders to DesferrioxamineAverage Response with Serum Ferritin• At 30 mg/kg

– Negative iron balance at transfusion rate of 22.2 mg/d (0.44 mg/kg/d), only if iron stores >4g (LIC >8mg/g wt)

• At 40 mg/kg – Negative iron balance at transfusion rate of 26.7 mg/d

(0.53 mg/kg/d), if body iron >2g (LIC 4 mg/g dw)

• At 50 mg/kg– Negative iron balance at transfusion rate of 31.1 mg/d

(0.62 mg/kg/d), if body iron >2g (LIC 4 mg/g dw)

Fischer R, et al. Br J Haematol. 2003;121:938–948.Fischer R, et al. Br J Haematol. 2003;121:938–948.8

Non-Responders to Desferrioxamine Monotherapy Conclusions

• 35 mg/kg (dose used in several comparative studies) unlikely to induce negative iron balance unless transfusion rate is low

• If poor response (LIC or ferritin), consider increasing dose or frequency (but max 40 kg/mg in children)

• As iron load falls, chelation will be less efficient and larger doses will be required to achieve the same response rate (especially <2g, LIC<3.8mg/g dw)

• However, this is not practical with DFO because of riskof toxicity

“Non–Responders” to Deferiprone (DFP)

10

Mean Iron Balance with DeferiproneLinked to Degree of Iron Load

Changes in LIC by SQUID over 2 y at 75 mg/kg

• Low LIC (<1500 µg/g wet wt)

– Negative balance in 24% (13/54)

• At higher LICs (>1500 µg/g wet wt)

– Negative balance in 50% (12/24)

Fischer R, et al. Br J Haematol. 2003;121:938–948. Fischer R, et al. Br J Haematol. 2003;121:938–948.

11

LIC Responders to Deferiprone Effects of Transfusion Rate and Iron Stores

• At low transfusion rate (22.2 mg/d)– At 75mg/kg negative iron balance only if iron stores >2g

• Intermediate transfusion rate (26.7 mg/d)– 75 mg/kg insufficient at any level of iron loading – 100 mg/kg negative iron balance if body iron >2g

• At high transfusion rate (31.1 mg/d)– 75mg/kg, negative iron balance only if body iron >3g

Fischer R, et al. Br J Haematol. 2003;121:938–948. Fischer R, et al. Br J Haematol. 2003;121:938–948. 12

Does Ferritin Response Rate Parallel the LIC Response with Deferiprone

Monotherapy?

13

Ferritin vs LIC Percent Responders to DFP Monotherapy

• Population and treatment– Italian poor compliers with DFO– 29 patients at 70 mg/kg

• Response by LIC– 6/20 (30%) responders– 70% non-responders

• Response by ferritin– 24/29 (83%) responders– 17% non-responders

Mazza A, et al. Haematologica. 1998;83:496-501.Mazza A, et al. Haematologica. 1998;83:496-501.14

Responders to DFPSerum Ferritin vs SQUID LIC stores

Changes in ferritin and stores (LIC + spleen ) by SQUID over 2 yrs

• Correlation between changes in ferritin and changes in iron stores

• BUT whereas 43% (23/54) of patients show decrease in ferritin

• Only 24% (13/54) show decrease in iron stores

Fischer R, et al. Br J Haematol. 2003;121:938-948.Fischer R, et al. Br J Haematol. 2003;121:938-948. 15

Ferritin Response with Long-Term Deferiprone

• Continued treatment (n = 26)– Mean treatment duration

39.4 months– Non-response 61%– 8/26 (31%) remain

>2500 µg/L

• Discontinued (n = 25a)– Mean treatment

duration 18.7 months– Non-response 72%– 20/25 (80%) remain

>2500 µg/L

Hoffbrand AV, et al. Blood. 1998;91:295-300.Hoffbrand AV, et al. Blood. 1998;91:295-300.

aIncluding 5 deaths.aIncluding 5 deaths.

58 Patients in total treated with DFP

Ferritin Response Rates to DFP by Starting Ferritins

• 151 patients completing 3 years of therapy

• Mean serum ferritin level declined from 2579 ng/mL at baseline to 2320 ng/ml at 3 years (P = 0.01)

• If initial ferritin levels higher than 2500 ng/mL, the ferritin level declined significantly

• If initial values <2500 ng/mL, the mean ferritin level did not change significantly

• Overall, 18% passed to a more severe class of ferritin

Ceci A, et al. Br J Haematol. 2002;118:330-336.Ceci A, et al. Br J Haematol. 2002;118:330-336. 17

Long-Term Ferritin Response to DFP What Factors Affect Response Rate?

58 patients on DFP at least 3 years DFP 75 mg/kg/d• At baseline

– Group 1 (n = 8): Ferritin <2000– Group 2 (n = 21): Ferritin 2000-4000– Group 3 (n = 29): Ferritin >4000

• Overall response at 3 years– 29% patients showed rise in serum ferritin– 52% remained stable– 19% showed decline

– Group 1:Significant ferritin increase (7/8 increased) (88% non-responders)– Group 2: Significant ferritin increase (19/21 increased) (90% non- responders)– Group 3: 8/29 decreased (ie, responded); (75% non-responders)

Goel H, et al. Hematology. 2008;13:77-82.Goel H, et al. Hematology. 2008;13:77-82. 18

Non-Responders to DFP Monotherapy Conclusions

• Highly variable reports of response rates - likely due to- Variations in baseline iron stores1,2

- Variations in transfusion rates3

- Variations in metabolism UGT1A64

- Variable follow-up time

• More non-responders with LIC than ferritin, why?- Hypothesis

Greater ferritin response than LIC3,5 response because Ferritin changes reflect macrophage (RE) iron changes RE iron mobilised more rapidly than hepatocellular iron Because DFP inactivated by glucuronidation iron binding site in hepatocytes

• Practical point- Ferritin changes may not fully reflect iron balance- May explain late rise in ferritin observed by several groups

1. Ceci A, et al. Br J Haematol. 2002;118:330–336. 2. Goel H, at al. Hematology. 2008;13:77-82. 3. Fischer R, et al. Br J Haematol. 2003;121:938–948. 4. Haverfield, Blood. 2005 106: Abstract 2703. 5. Mazza A, et al. Haematologica. 1998;83:496-501.

.

Abbreviation: RE, reticuloendothelial.Abbreviation: RE, reticuloendothelial.

Responders to Combined DFP + DFOIron Balance (LIC)

DFP Monotherapy Combination Therapy

Baseline 12 months Baseline 12 months

Decrease in 5/12Non-response 58%

Decrease in 7/8Non-response 13%

With permission from Aydinok Y, et al. Haematologica. 2007;92:1599-1606. 20

Responders to Combined DFP + DFOSerum Ferritin

• Effect of deferiprone (75 mg/kg/d) + DFO (2 g/d twice weekly) on serum ferritin levels

• Iron intake reported in each patient (mean 0.31mg/kg/d, n = 11)

• 9/11 decreased ferritin

• In 2/11 no decrease (iron intake 0.39 and 0.30 mg/kg/d)

21Mourad FH, et al. Br J Haematol. 2003;121:187-189.

“Non-Responders” to Deferasirox

(DFX)

22

Change in LIC by Deferasirox Dose and Ongoing Transfusion Burden

23Cohen AR, et al. Blood. 2008;111:583-587.

Dose mg/kg

10

20

30

% of Patients in Negative Balance

Low Transfusion

<0.3 mg/kg/d

29

76

96

Medium Transfusion

0.3-0.5 mg/kg/d

14

55

83

HighTransfusion

>0.5 mg/kg/d

0

47

82

Pharmacokinetics of Deferasirox in Patients with Adequate/Inadequate Response

• 10 transfused patients with inadequate response (rising ferritin or rising LIC) on >30 mg/kg per day vs 5 control transfusion-dependent patients with adequate response

• Compared to controls, patients with inadequate had significantly lower systemic drug exposure (P <.00001)

Chirnomas D, et al. Blood. 2009;114:4009-4013.

Pharmacokinetics of Deferasirox in Patients with Adequate/Inadequate Response

• No differences observed between adequate and inadequate responders in– Cmax– Volume of distribution/bioavailability (Vd/F)– Elimination half-life (t½)

• Conclusion– Bioavailability is the likely discriminant

Chirnomas D, et al. Blood. 2009;114:4009-4013.

Relationship of Ferritin to LIC Changes with Deferasirox?

• Change in ferritin correlates with change in LIC across diagnoses (TM n = 85, MDS n = 47, DBA n = 30, rare anaemias n = 22)

• Intersect of correlation suggests that when no change in ferritin, there is a fall in LIC of about 4-7mg/g d wt

• So when body iron falls, there may be a decrease in LIC (negative iron balance) without a significant change in ferritin

• Ferritin may also increase in the absence of a positive trend in LIC

Porter J, et al. EJH. 2008;80:168-176.Porter J, et al. EJH. 2008;80:168-176. 26

Liver Tissue Iron Scores Following Therapy with Deferasirox

27

Porter J, et al. 10th International Conference on Thalassaemia and Haemoglobinopathies, January 7-10, 2006.

• MethodTissue Iron Score (TIS) on liver iron biopsiesChange in score after 1 year of chelationChange in distribution of iron after 1 year of chelationDeferasirox n = 224, Desferrioxamine n = 230

• Results• Fall in TIS correlated with fall in LIC• Greater removal from hepatocellular than macrophage compartment (cf DFO)

LIC 8.0 mg/Fe g dwTotal, TIS 24Hetatocyte, HIS 24Sinusoidal, SIS 0Portal, PIS 0

Hepatocytic:total liver iron ratio 1

LIC 9.1 mg/Fe g dwTIS 19HIS 3SIS 10PIS 6

Hepatocytic:total liver iron ratio 0.16

Hepatocytic—Total Liver Iron Ratio Reflects Tissue Distribution of Iron

Prussian blue stain

Abbreviations: HIS, hepatocytic iron score; PIS, portal iron score; SIS, sinusoidal iron score; total tissue iron score.Abbreviations: HIS, hepatocytic iron score; PIS, portal iron score; SIS, sinusoidal iron score; total tissue iron score.

Deugnier YM, et al. Gastroenterology.1992;102:2050–2059. Turlin B, Deugnier Y. J Clin Pathol. 1997;50:971. Deugnier YM, et al. Gastroenterology.1992;102:2050–2059. Turlin B, Deugnier Y. J Clin Pathol. 1997;50:971. 28

Deferasirox But Not DesferrioxamineShowed a Tendency to Reduce

Hepatocytic:Total Liver Iron Ratio

-0.2

0

0.2

0.4

n= 6 27 86 10 9 48 83 191

<25 25–<35 35–<50 ≥50 5 10 20 30

DFO (mg/kg 5 days/week) DFX (mg/kg/day)

Chan

ge in

Hep

atoc

ytic:

Tot

al L

iver

Iron

Rati

o

Mean ± 95% CI

Porter J, et al. 10th International Conference on Thalassaemia and Haemoglobinopathies, January 7-10, 2006.Porter J, et al. 10th International Conference on Thalassaemia and Haemoglobinopathies, January 7-10, 2006.

Serum FerritinResponse to Different Chelators

Hypothesis• With DFO

– Iron removal is approximately equal in hepatocytes and RE cells– So changes in ferritin reflect changes in LIC and RE iron

• With DFP– Iron removal is preferential in RE cells compared with hepatocytes– Explains why may see response with ferritin but not with LIC– Explains late secondary resurgence of ferritin

• With DFX– Iron is removed preferentially from hepatocytes compared with RE cells– Body iron (LIC) may fall even though ferritin may lag

30

Possible Reasons for “Non-Response”

• The measure used may not reflect iron balance in that individual

• The patient may have a high transfusion requirement• The dose/frequency may be insufficient• The patient may not be taking the treatment

regularly• The PK and metabolism of the chelator may not be

favorable (very limited data)

31

Non-RespondersPractical Questions to Ask

• How are you assessing “non-response”? Is this the correct measure?

• How long have you been assessing response? Is it long enough?

• Is the patient on a high, low, or average transfusion schedule?

• Is the patient taking the prescribed dose and frequency of chelator?

32

The Non-ResponderPractical Things to Do

• Quantitate current transfusional iron loading rate

• Consider additional measures of “response” (e.g., changes in LIC)

• Carefully interview the patient about adherence patterns

• Consider whether administration of drug is optimal (e.g., timing of oral dose relative to food, number of hours of DFO infusion)

• Consider whether the dose can be safely increased

• Consider alternative chelation

33

Management of Non-Responders to

Iron Chelation Therapy

Kai-Hsin Lin, MD

34

Contributing Factors in Inadequate-Responders

• Poor drug compliance• Inadequate drug dose due to side effects • Poor drug efficacy • Comorbid diseases confounding the

assessment of iron overload

35

Case 1: Ms H

36

Ms. H—Treatment History• 18-year-old female diagnosed with thalassaemia

major at age 1 year• Received regular transfusions

– 4 units washed RBC every 2 weeks (250 cc whole blood)• Has been receiving iron chelation therapy with

desferrioxamine since 3 years old– Poor desferrioxamine compliance (4-5 nights per week)

due to inconvenience and discomfort• No hepatitis B or C or diabetes mellitus• No puberty

37

Initiation of Deferasirox

• Ms. H started take deferasirox as iron chelation therapy at 20 mg/kg/day in Feb 2007

• Baseline assessments– Serum ferritin: 5037 ng/mL– AST: 19 U/L, ALT: 21 U/L– Creatinine: 0.7 mg/dL

38

Deferasirox Therapy

• Patient found once-daily oral administration of deferasirox to be more convenient than SC or IV desferrioxamine

• Improved drug compliance• Improved iron chelation and reduced total

iron burden

39

Deferasirox Treatment CourseSerum Ferritin (Months 1–36)

Deferasirox therapy (mg/kg/day) 20 30 35 40 35

0

1000

2000

3000

4000

5000

6000

-5 0 5 10 15 20 25 30 35

Months

Seru

m fe

rriti

n (n

g/m

L)

Start

Courtesy of Dr. Lin 40

Poor Drug Compliance

1. Forgetting to take the drug 2. Not understanding or misinterpreting

instructions3. Experiencing side effects (the treatment may be

perceived as worse than the disorder)4. Disliking the drug taste or smell5. Finding restrictions on treatment inconvenient 6. Not caring about getting better

41

Poor Injection Compliance

Two main reasons are

• Lack of disease knowledge

• Incorrect injection method

42Lin SS. Presented at 12th TIF, January 7-10, 2006.Lin SS. Presented at 12th TIF, January 7-10, 2006.

Serum Ferritin Level vs Injection Days (Before & After Disease Education)

Group

Items

Experimental Group (n=20) Control Group (n=20)

Average Score Standard Deviation T value P value Average

ScoreStandard Deviation T value P value

Injection Days

Before

After

3.654.625

1.0770.646 -4.791

.000 3.553.8

1.7311.399 -.773 .449

Serum Ferritin Level

Before

After

8286.96180.8

4355.93868.9 3.94

.0006383.26734.6 4153.7

3958.5 -.768 .45243Lin SS. Presented at: 12th TIF, January 7-10,2006.

Courtesy of Syi Su Lin, MDLin SS. Presented at: 12th TIF, January 7-10,2006. Courtesy of Syi Su Lin, MD

Poor Drug Compliance: Management

*Medication diary

*Family support group

*Patient peer support group

*Disease education

*Treatment guideline

44Courtesy of Kai-Hsin Lin, MDCourtesy of Kai-Hsin Lin, MD

Case 2: ML

45

ML: Treatment History

• 15-year-old male diagnosed with thalassaemia major at age 8 months

• Received regular transfusions– 2 units washed RBC every 2 weeks (250 cc whole

blood)• Started iron chelation therapy with

desferrioxamine at age 1 year• No hepatitic B or C or diabetes mellitus

46

Initiation of Deferasirox

• Began iron chelation therapy with deferasirox at 20 mg/kg/day in March 2006

• Baseline assessments– Serum ferritin: 3693 ng/mL– AST: 24 U/L, ALT: 16 U/L– Creatinine: 0.5 mg/dL

• Good compliance but had mild GI upset when taking deferasirox

47

Reduced Bioavailability

• 40 mg/kg/day once-daily deferasirox only maintained total iron burden and led to GI upset in this patient

• Bioavailability of deferasirox in this patient may be reduced, as individuals with inadequate response to deferasirox have been found to have significantly lower systemic drug exposure compared with those having an adequate response1

• Reduced drug bioavailability in this patient may be the result of patient variability in– Oral absorption– Distribution– Metabolism– Excretion

1. Chirnomas D, et al. Blood. 2009;114:4009-4013.48

Managing Inadequate Bioavailability

In this patient, twice-daily deferasirox at the same total dose increased bioavailability, relieved GI upset, and further reduced total iron burden

49

Deferasirox Treatment CourseSerum Ferritin (Months 1 - 48)

0

1000

2000

3000

4000

5000

6000

7000

-5 0 5 10 15 20 25 30 35 40 45 50 55

Months

Seru

m fe

rriti

n (n

g/m

L)

Deferasirox therapy (mg/kg/day) 20 30 40 40 BID

Start

50Courtesy of Kai-Hsin Lin, MDCourtesy of Kai-Hsin Lin, MD

Conclusions• Improved medication compliance can improve the efficacy of

iron chelation therapy– Improved treatment convenience– Disease education– Patient support group

• Individual bioavailability of deferasirox for iron chelation– Identify responders versus inadequate responders– Consider alternative regimens, such as twice-daily dosing or

deferasirox + desferrioxamine combination therapy to deal with side effects while maintaining iron chelation efficacy

51