Oral Sodium Pbenylbutyrate Therapy in Homozygous B Thalassemia:
`A Clinical Trial
`
`By Anne F. Collins, Howard A. Pearson, Patricia Giardina, Kevin T. McDonagh, Saul W. Brusilow,
`and George J. Dover
`
`Butyrate analogues have been shown to increase fetal Ile-
`rnoglobin {HbF} production in vitro and in trlvo. Sodium pha-
`nyflrrrlvrate {SP3}, an oral agent used to treat Individuals
`with urea-cycle disorders, has been shown to increase HbF
`in nonanamic individuals and in individuals with sickle cell
`disease. We have treated eleven patients with homozygous
`H thalaeeernle [three ‘transfusion dependent] and one sickle-
`H-thaiaeeernia patient with 20 old {forty 500-mg tablets! of
`SP3 for 41 to 400 days. All patients showed an increase in
`thl|)III:On1DlFrIficul0¢VllO8IIO0I:lIlItlW'itl'ItfIItl11Iflt. but
`only four patients responded by increasing their I-lb levels
`by greater than 1 gldl lrnean increase, 2.1 gIdL: range, 1.2
`to 2.8 flldl.l. None of the transfusion-dependant thalassemla
`aubiecte responded Increase in lib was associated with an
`increase in red blood cell number lrnaan increase, 0.82 x
`10"’lLl. and mean corpuscular volume {mean Increase. 6 it}.
`changes in percent HbF, absolute HbF levels, or a- to non-
`
`OMOZYGOUS 13 Tl-IALASSEMIA, a disease in which
`there is inadequate production of £3 globin leading to
`severe anemia, affects thousands of individuals worldwide.
`Current management of this condition includes the use of
`regular red blood cell (RBC) transfusions and iron chelation
`therapy. The development of an effective therapy to increase
`hemoglobin (Hb) levels in homozygous ,6 thalassemia with-
`out the use of RBC transfusions could allow normal growth
`and development. whereas decreasing or eliminating transfu-
`sional iron overload, which remains the major cause of death.
`reduced life expectancy and morbidity in individuals with
`this disease.‘ Although bone marrow transplantation can
`achieve these aims,’ it is not a therapeutic option for the
`majority of patients.
`For some years, there has been interest in increasing 7-
`globin transcription and fetal Hb (HbF) production in pa-
`tients with l? hemoglobinopathies.’-“ For patients with homo-
`zygous ,6 thalassemia. increased 1»--globin production and a
`reduction in the ratio of or- to non-ac-globin could reasonably
`be expected to ameliorate the severity of the anemia. To this
`end, trials of chemotllerapeutic agents including 5-azacyti-
`dines“ and l‘tydroxyurea”‘° have been conducted, but my-
`elotoxicity, fears of long-term carcinogenesis. and only mod-
`est responses to treatment have limited the clinical usefulness
`of these agents. Erythropoietin has also been used, but re-
`sponses to this therapy have been va.riable.'°‘”
`There is considerable evidence that butyrate analogues
`induce erythroid differentiation'’'‘‘‘ and stimulate HbF pro-
`duction in human erythroid progenitors in vitro.'5'" In vivo,
`these agents have also been shown to reactivate embryonic
`globin production in an avian model,” delay the switch from
`fetal to adult globin in ovine fetuses.” and to increase HbF
`production in adult primates."""°'"
`In humans. several fatty acids including at amino-btltyric
`acid,” arginine butyrate,“'” isobutyrarnide,"‘*2" sodium phe-
`nylbutyrate.“’”’ propionic acid.” and 2-propylpentanoic (di-
`propylacctic) acid (unpublished data) have now been shown
`to stimulate HbF production, suggesting that they may play
`a role in the treatment of the ,6-globin disorders. However,
`previous clinical trials of these agents in ,6-thalassemia have
`
`Blood. Vol 35, No 1 {Januarv 1}, 1995: pp 43-49
`
`
`
`or-globin ratios as measured by levels oi n1flNA and globirt
`protein in peripheral blood did not correlate with response
`to treatment. Response to treatment was not associated
`with the type offl-globin mutation, but baseline anthropolo-
`tin levels of greater than 120 n1UIn1I. was seen in all respond-
`ers and only two of eight nonreeponders to SP3. compliance
`with treatrrl-ant was greater than 90% as measured by pill
`counts. Side effects of the drug included weight gain andl
`or edema caused by increase salt load In 2l12. transient
`epigealric discomfort in ‘N12, and abnomial body odor in 3]
`12 sultiects. Two splenectomized patients who were not on
`proplwlaetic antibiotics developed sepals while on treat-
`ment. We conclude that SPB increases Hb in some patients
`with thalasaernia, but the precise mechanism of action is
`unknown.
`© 1995 by The American Society of Hematology.
`
`been limited to relatively short—terrn trials of the intravenous
`agent, arginine buryrate,”"“"‘ and oral isobutyramide."*”
`Sodium phenylbutyrate is an orally administered agent
`originally developed to promote waste nitrogen excretion in
`the treatment of urea-cycle disorders“ and is currently used
`for this purpose in an Federal Drug Adrninistratiomapproved
`phase III trial. Over 100 patient-years experience with this
`dnig has now accumulated with no untoward effects being
`found. The finding of increased HbF levels in these patients”
`stimulated clinical trials of sodium phenylbutyrate in patients
`with ,6-hemoglobinopatliies.
`We report here the first long-term clinical trial of an orally
`administered fatty acid. sodium phenylbutyrate, in patients
`with homozygous ,6 thalassernia. This represents the largest
`clinical trial of any Hb switching agent used in thalassemic
`patients to date.
`
`PATIENTS AND METHODS
`
`"Ibis study was approved by the Joint Committee on Clinical
`investigation of the Johns Hopkins Medical Institutions and written
`informed consent was obtained from all patients. Eleven patients
`
`From The Johns Hopkins University School of Medicine. Balti-
`more MD; Yale University School ofMedicr'ne, New Haven CT.‘ New
`York Hospltal~Comell Medical Center NY.‘ and the National Heart,
`Lung. and Mood Institute. Bethesda MD.
`Submitted June 28, l 994: accepted September 8. 1994.
`Supported in pan‘ by National lrtstirule of Health Grants No.
`HL 28028 (to G..l.D.}. HD IH34, HD 26358 (to S.W.B.J: Clinical
`Research Center Grants No. RR-0035 and RR-00722; The Cooley '.r
`Anemia Foundation lnc: The Cooley '.r Anemia Foundation of Mary-
`land Inc: and The Connecticut Campaign Against Cooley‘: Anemia.
`Address reprint requests to George J. Dover. MD, Ross Research
`Bldg. Room H25, 720 Rolland Ave, Baltimore, MD 21205.
`The publication costs of this article were defrayed in part by page
`charge payment. This article nutrr therefore be hereby marked
`“advertisement" in accordance with 18 U.S.C. section 1734 solely to
`indicate this fact.
`© [995 by The American Society of Hematology.
`G006-4971/951850! -0024$3.0{Ml
`
`PAR PHARMACEUTICAL, INC.
`PAR PHARMACEUTICAL, INC. EX. 1027
`EX. 1027
`
`

`
`44
`
`COLLINS ET AL
`
`Table 1. clinical Details of Patients Studied
`
`Patient No.
`
`Age iyrsl
`
`Sex
`
`,6-Globin Mutations
`
`1
`2
`3
`4
`5
`5
`1-’
`8"
`9
`10
`11
`12
`
`32.?
`25.5
`20.5
`32.2
`35.‘!
`20.5
`412.3
`30.5
`24.3
`26.9
`25.1
`35.7
`
`M Gordon 39!??
`F
`Homozvg N51, nt 110
`F
`Homozyg W51, nt 110
`F
`Codon 39llVS1, nt 1
`F
`Coclon 39;'NS1, nt 1
`M Codon 39)‘IVS1, nt 5
`F
`W51, nt BR?
`F
`Codon 39t'FFt 5K6
`M Homozyg IVS1, nt 110
`F
`Homozyg IVS1, nt 110
`M
`IVS1, nt ‘UIVS2, nt 745
`F
`LWIVSZ, nt 745
`
`All patients had normal at genes.
`' Pretreated with 5—ezacytidine.
`t Transfused 6 and 3 weeks before study.
`
`Prestudv Hb
`t9!dLl
`
`7.81
`6.0
`5.9
`1.4
`5.5
`8.1
`-1.5
`5.5
`Transfused
`Transfused
`Transiused
`8.2
`
`with homozygous ,8 thalassemia and one patient with sickle ,6 plus
`thalassemia were studied. Of the 11 patients with homozygous ,8
`thalassemia. 3 were receiving regular RBC transfusions and 8 were
`not. Of these 8, 4 had never received regular transfusions, 3 had
`discontinued regular transfusions because of severe RBC alloimmu-
`nization and one had discontinued transfusions to hasten the treat-
`ment of severe iron overload. All patients had previously undergone
`splenectomy. Details of each patient’s age, fi—globin mutations. or
`gene number and prestudy Hb are shown in Table 1.
`Patients I through 7 and 9 through 12 commenced sodium phenyl-
`butyrate therapy during a 21-day inpatient stay in the Johns Hopkins
`Hospital Clinical Research Unit. Patients 2 and 9 underwent the
`protocol on two occasions, recommencing sodium phenylbutyrate
`after 59 and 440 days off treatment. All received 20 gfd of sodium
`phenylbutyrate (10.6 to I3.8 glm’/d) given as forty 500—mg tablets
`in three divided doses: I2 tablets were given with breakfast and 14
`tablets with each of lunch and dinner. All were discharged home on
`medication and were followed from between 3| and 500 days as
`outpatients. Five of these patients continue on medication.
`Patient 8. unable to be transfused because of severe alloimmuniza—
`tion. was receiving intravenous infusions of 5—azacytidine every 3
`to 4 weeks’ before starting sodium phenylbutyrate therapy. Her ther-
`apy was initiated at a lower dose of 12 gld (9.6 glmz.-"d) during a
`21-day inpatient stay at the National Institutes of Health. Bethesda.
`MD. The infusions of 5—azacytidine were subsequently discontinued
`and she continues on sodium phenylbutyrate alone.
`All patients were documented to have normal RBC folate levels
`(and serum folate in the case of regularly transfufl patients) before
`the commencement of therapy and received folic acid I mgfd orally
`while on the protocol.
`Blood counts were determined using Coulter STK—S or STK-R
`counters (Coulter Immunology, Hialeah, FL) with manual spun I-lb
`measurements being performed to overcome any anifactual elevation
`in automated Hb measurements caused by the presence of nucleated
`RBCS in the peripheral blood.” A Technicon H-I counter was used
`for the determination of RBC mean corpuscular volume (MCV).
`Reticulocyte counts were measured with a Sysmex R-1000.” Per-
`centages of HbF. F cells, and F reticulocytes were measured as
`previously described.” G1obin—chain synthesis was measured by in-
`cubating peripheral blood (PB) with n-itiatecl leucine using standard
`methods” with globin—chajn separation being achieved by high—per—
`fonnance liquid chromatography. PB ratios of or. {.i- and -y—globin
`ITJRNA were measured using an RNAase protection assay (RPA II
`kit; Ambion, Inc, Austin, TX). "P-labeled RNA probes were made
`complimentary to a 130-bp segment of exon I of cr globin. a 205-
`
`bp segment of cxort 2 of ,6 globin and a 170-bp segment of exon 2
`of y globin. The protected fragments were separated by electrophore-
`sis on an 8 mol:'L urea 6% polyacrylamide gel. located by autoradiog-
`raphy and quantitated by counting the radioactivity of each isolated
`band in a scintillation counter. Levels of sodium phenylbutyrate.
`phenylaeetate, and phenylacetylglutalnine were measured in plasma
`and urine by previously described methods.”
`Data are expressed as mean 1 standard deviation (range) and
`were compared using the paired Student's r-test. Differences between
`observed and expected frequencies were compared using a chi-
`square test.
`
`RESULTS
`
`Changes in Hi). We divided the patients into two broad
`categories. responders and nonresponders, based on changes
`in Hb while on sodium phenylbutyrate therapy {Table 2). A
`response was defined as a sustained rise in H1: of greater
`than 1 g/dL. A sustained increase in Hb of 2.1 1 0.7 g/dL
`(1.2 to 2.8 g;'dL) was seen in 4 of the 3 (50%) patients
`with homozygous )5’ thalassemia not on regular transfusion
`programs (Fig 1). Patient 4 showed a progressive increase
`in Hb over 350 days on therapy, with an acute episode of
`anemia related to septicemia at day 200 during which she
`was not transfused. Her Hb peaked at 10.2 gId.L on day 351
`and then fell to around 9 g!dL when her dose of sodium
`phenylbutyrate was reduced by 25% to 15 gfd. Interestingly,
`her sibling, patient 5, also responded to sodium pher1ylbuty—
`rate therapy with an increase in I-lb of 1.8 g;'dL over the first
`140 days on therapy. Patient 3, who was receiving intrave-
`nous infusions of 5-azacytjdine at the time sodium pheny1bu-
`tyratc therapy was commenced, has shown a progressive
`increase in Hb over more than 450 days, interrupted by an
`acute episode of anemia related to a major gastrointestinal
`hemorrhage at day 220. She received two units of blood at
`this time. Her Hb has continued to increase allowing the 5—
`azacytidinc infusions to be discontinued at day 280. Patient
`3 showed an increase in Hb of 1.2 gfdl. while on sodium
`phenylbutyrate, and her I-lb had returned to pretreatment
`levels within 40 days after stopping therapy. Although this
`increase in Hb is only modest,
`in the other patients re-
`sponding to therapy, Hb continued to increase for 500 days
`or more. suggesting that a trial of only 100 days of sodium
`phenylbutyrate therapy may have been inadequate to achieve
`the maximum possible response. Four of the eight nontrans-
`fused patients and all three previously regularly transfused
`patients showed no response to therapy (Table 2. Fig 2).
`F retfculocyte re.rpon.re. All 12 patients showed an in-
`crease in the percentage of F reticulocytes after the com-
`mencement of sodium phenylbutyrate therapy, with levels
`at day 21 being 70% i 74% (mean : ISD; range, 6% to
`248%) above baseline levels (P = .001). [n all patients, the
`increased level of F reticulocytes persisted as long as therapy
`was continued. In five of eight patients who discontinued
`sodium phenylbutyrate therapy. F reticulocytcs decreased
`rapidly, but in three, nos. 1, 6. and 11. the F reticulocytes
`remained at the higher level for 4 weeks or more after the
`cessation of therapy. Two patients. 2 and 9. again showed an
`increase in F reticulocytes when therapy was recommenced.
`These changes suggest that the increase in F reticulocytes
`was directly related to the sodium phenylbutyrate therapy
`(Fig 3).
`
`

`
`PHENYLBLFFYHATE THERAPY IN ,6 THALASSEMIA
`
`Patient No.
`Nonresponders
`1'
`2
`
`6
`T
`9*
`
`10*
`11'
`12
`
`Responders
`3
`4*
`51:
`31
`
`De I;
`Treeiied
`
`323
`91
`89
`95
`9?
`391‘
`311‘
`861‘
`1'1
`44
`
`100
`490
`13?
`468
`
`Tlble 2. Hb, nee Count, me Mcv Pro- me PoIt—Sodiurn Phonvtbutyrlte Thmpv
`Enftllropoietin
`lmU.-’mLI
`
`Hb tnrdu
`
`RBC x10"!L
`
`Mcv IILJ
`
`LDH 1 un.
`
`Prs
`
`‘.-‘.8
`8.0
`5.?
`8.1
`4.5
`9.1
`9.0
`9.4
`8.4
`8.2
`
`6.9
`14
`6.5
`5.5
`
`Poet
`
`5.1
`5.8
`6.1
`6.1
`5.0
`?.9
`8.8
`5.8
`5.0
`8.1
`
`8.1
`10.2
`8.3
`8.3
`
`Pre
`
`2.95
`2.42
`2.2?
`3.00
`2.55
`3.10
`3.15
`3.25
`2.90
`3.14
`
`3.63
`2.58
`2.61
`1.91
`
`Post
`
`2.81’
`2.43
`2.38
`2.3‘?
`2.50
`2.80
`3.10
`ND
`1.71
`3.62
`
`3.9?
`3.36
`3.3?
`2.49
`
`Pre
`
`Post
`
`85
`98
`8?
`810
`T3
`85
`8-6
`84
`88
`88
`
`66
`38
`T6
`93
`
`81
`85
`90
`80
`T5
`B5
`88
`90
`96
`T4
`
`T3
`94
`33
`9'.-'
`
`Pre
`
`96
`185
`ND
`18
`104
`55
`44
`25
`240
`41
`
`186
`3.882
`20?
`682
`
`Pre
`
`334
`729
`ND
`601
`340
`ND
`214
`144
`302
`308
`
`418
`440
`247
`531
`
`Dev 21
`
`23'.-'
`666
`ND
`385
`148
`ND
`251
`210
`3T1
`346
`
`289
`255
`251
`422
`
`45
`
`Indirect Bilimbin
`Img.rdL:-
`
`Pro
`
`4.8
`4.1
`ND
`3.7
`4.6
`3.9
`2.5
`1.1
`1.3
`1.0
`
`5.6
`3.3
`2.6
`4.3
`
`Dev 21
`
`3.5
`3.1
`ND
`3.0
`1.".-'
`1.’.-'
`2.5
`0.9
`1.4
`0.9
`
`3.5
`1.7
`1.6
`4.6
`
`Abbreviation: ND. not done.
`' Patient was transfused before commencing sodium phenylbutyrete.
`1 Patient was transfused while on sodium phenylbutyrate therapy.
`¥ Patient continues on therapy.
`
`Hemoglobin F response. Table 3 outlines the HbF re-
`sponse to sodium phenylbutyrate therapy in each patient, as
`measured by percentage of HbF. absolute HbF gfdL, and or!
`non-or ratios of both globin mRNA and globin-chain synthe-
`sis. In those patients responding to therapy. increase in total
`Hb was not solely explained by increased HbF. In those
`patients previously on regular transfusion programs. at tenni-
`nation of therapy both Hbl-7% and absolute HbF levels were
`higher than at baseline, probably reflecting the reduced sup-
`pression of erythropoiesis related to fall in Hb and not being
`necessarily directly attributable to the sodium phenylbutyrate
`therapy. Ratios of o-— to non-ar—gl0bin mRNA. measured be-
`fore therapy and again at day 21 in all patients, showed no
`uniformity of response to sodium phenylbutyrate therapy,
`
`nor was there any correlation between the change in globin
`mRNA ratios and peripheral Hb levels. Similarly, when re-
`peated in three patients after more than 270 days on therapy.
`no correlation between ratios of globin mRNA and change
`in I-lb could be shown. In addition to mRNA studios, in three
`
`patients, the production of a- and non«x—globin chains in
`PB reticulocytes was measured at day 0 and at day 21. Again,
`in these patients, neither ratios of globin mRNA nor globin-
`chain synthesis seemed related to changes in Hb on therapy.
`Patient I2. with sickle l9 plus thalassemia, did not experi-
`ence an increase in Hb and, therefore. was called a nome-
`sponder. However, over the first 21 days of therapy. she did
`show an increase in F reticulocytes (14% to 34%} and F
`cells (13% to 20%) associated with an increase in HbF%
`
`lWl802lX|fi03CH)¥0-ICIJJIO
`DAVOOIISPB
`
`G.-‘DI.
`HEIIooI.oIIH
`
`so 10 an an no so on ‘In no eoinonniauien
`DAYSOIISFO
`
`.5RI
`
`3*ll
`
`HEIIOGLD-BINamt:"1
`
`O
`0204I}fl60
`we
`mveouswe
`
`I20
`
`I40
`
` 50
`I-1- 0
`
`Vi IIIII-I-III I
`
`I
`
`IIIEZCIISW-Smith}
`DAYS OH SUI
`
`l-Iitnthefourpntientnrecpondlngtoeo-
`Fig 1.
`dlum shiny-Imtvrm unnmn Pltlent 4 tumor lcftl.
`pclilnt 5 llflplf flflllll. prflom 8 (lower Icftl. and
`patient 3 {lower ri1|lfll- ill. wdum planyliaitvreu
`thormv. wlfltehlnguinfludnoiflnhuoonr
`spending to drama In den. l'l. intntrcnouc Infu-
`clonsol .
`
`

`
`46
`
`COLUN5 ET AL
`
`...l
`
`E
`
`T‘ T ____i
`.
`
`S0-
`tinfrtrp)-'.
`Cornpr"irJrrt'e with sodium p.l1en_vlbul'_\‘rrtte'
`dium phenylbutyrate tablets were provided to the patients
`with :1 25-day supply each time; a further supply was pro-
`vided only when the patient specifically requested more tab-
`lets. ltt this way. compliance was calculated for each patient
`by comparing the number of tablets dispensed to that pre-
`scribed. Compliance with therapy was a problem in only one
`patient. no. 3, who abruptly discontinued therapy after I00
`days, having been 95% compliant lip until that lime. For the
`patients as a group. compliance with medication was 97%
`J. 3%.
`
`Peak daytime levels
`Phettflburyrore p}2ormacot'citterrcr.s‘.
`of phenylbutyratc. phenylacetale. and phcnylacetylglutarnine
`ranged between 0.60 and l.'r'0 ‘mmolfL. 0.50 and [.50 mmol!
`L. and 0.56 and 2.67 mmol)'L. respectively. Serial-fasting
`morning plasma levels of phenylacetate. measured in all pa~
`tients. were less than [.0 mmoUL and showed no progressive
`accumulation. Serial 24-hour urine collections performed in
`9 of the 12 patients studied showed a mean excretion of
`76% 1 13% (53% to 97%) of the molar amount of sodium
`phenylbutyrate administered as urinary phenylacetylgIuta-
`mine. Plasma glutamine levels. measured before therapy and
`again after 2. 9. and 21 days on therapy showed no evidence
`of glutamine depletion. These results are similar to those
`reported in patients with urea-cycle disorders“? and homo-
`zygous sickle cell disease” treated with sodium phenylbuty—
`rate.
`
`Adverse events occurring on therapy. The daily dose of
`20 g of sodium phenylbutyrate contributes 2.460 mg (107
`mmol) of sodium to the diet. a significant proportion of the
`recommended daily intake. While in the hospital, one of the
`twelve patients (no. 6) developed ankle edema associated
`with a 3.5% increase in body weight. which resolved spontaw
`neously with dietary modification. After discharge from hos-
`pital. one patient (no. 1) required intermittent treatment with
`a tltiazide diuretic and one (no. 8) required an increase in
`her previous diuretic dose to control peripheral edema. No
`patient developed hypertension. Epigastric discomfort after
`
`F RETlC3'Ka
`40
`
`30
`
`1|!
`
`PRE
`
`DAY 21
`
`POST
`
`RE8TAR'l'
`
`STOP
`
`Fig 3. The percentage of F rotietrloeyios in individual patient: bl-
`faro sodium phenvltutyntu therapy [PEEL after 21 day: of therapy
`(DAY 21l. at least 2 weeks after the caution of tltnrltltr lPO8'll.
`on rostu-ting therapy HIESTARTI. and after again stopping therapy
`[STOPl. {I}. responders: (OI. flnnraspondom.
`
`
`
`'55
`
`9
`575
`
`N
`
`._2.
`
`‘vi
`V
`
`.
`
`.
`
`.
`
`. ..
`
`I
`
`:
`
`3&5
`:5
`
`-:
`
`
`“\.. 4’
`
`__.
`01020304-O506{JT08O
`
`oE
`
`u.I5.5
`::
`
`DAYS ON SPB
`
`Fla 2. Hb in two patients who did not respond to therapy: patient
`2 Pl. who is not rogullrly transfused and patlorrt 11 {|Zll, who is
`transfusion-dependent.
`
`(3.1% to 5.0%). This response to therapy is very similar to
`those previously seen in patients with homozygous sickle
`cell disease.” In sickle tltalassemia. an increase in the per-
`centage of HbF is probably a more desirable outcome than
`increase in Hb. Interestingly, while on therapy, her MCV
`increased from 6'.-’ fL to 74 ii. and returned to its pretreatment
`value when sodium phenylbutyrate therapy was discon-
`tinued.
`
`Indicators of hemoiysis. The nine patients not on regular
`transfusion programs showed indirect evidence of a reduc-
`tion in hemolysis, widi significantly lower levels of serum
`lactate dehydrogenase (P < .03) and indirect bilirubin (P =
`.0005} when pretreatment levels were compared with those
`at day 21 (Table 2). This was not observed in the transfusion-
`clependent patients, possibly related to the mean fall in I-lb
`of 2.0 g!dL. which occurred in these patients during the 21-
`day inpatient study period. In contrast, no significant differ-
`ence was seen in H1) between baseline measurements and
`
`day 2] in those patients who were not regularly transfused.
`Predictors of increased Hb in response to sodium phenyl-
`btttyrate therapy. Response to sodium phenylbutyrate ther-
`apy, as defined by a sustained increase in total Hb of greater
`than 1 g/dI.. above baseiine, did not appear to be predicted
`by 5-globin mutation; baseline percentage of HbF, absolute
`HbF. or F—reticulocyte levels: or baseline Hb or baseline or-
`to non-or-globin ratios. Similarly, significant falls in lactate
`dehydrogenase and indirect bilirubin, traditional measures
`of hemolysis, could be shown in all nontransfused patients
`with no differences being observed between responders and
`nonresponders. Interestingly.
`those patients with baseline
`crythropoietin levels greater than 120 mUfmL were signifi-
`cantly (P «z .05} more likely to experience an increase in
`Hb («#6) than those whose baseline erythropoietin level was
`below l20 mU."rn.L (0.~'6) (Table 2). A similar trend existed
`between baseline HbF percentage in those patients not re-
`ceiving regular RBC transfusions and response to sodium
`phenylbutyrate therapy, although this did not reach statistical
`significance. Of the four patients with baseline l-IbF less than
`40%, none responded to therapy. In contrast. four of the five
`patients with baseline I-IbF greater than 40% did respond (P
`= .075).
`
`

`
`PHENYLBUTYRATE THERAPY IN ,6 THALASSEMIA
`
`4?
`
`Table 3. Percent I-Rf, Absolute l~l:IF, and rrlhlolt-ct Ratios Pre- and Poet-Sodium Phenylbutvrlte Therapy
`a-,i'NnI't-o-
`Ratio
`Globin
`
`a.n'Non—u Ratio
`n1F|N.A {mean 1 SD]
`
`HbF l%l
`
`Abs!-IbF tgJdL}
`
`Patient No.
`
`Nonresponders
`1*
`2
`
`6
`1'
`9"
`
`10"
`11'
`12
`
`Responders
`3
`4
`5
`3
`
`Pre
`
`13
`71
`T6
`21
`11
`1
`2
`2
`13
`3.1
`
`5'.-'
`90
`85
`82
`
`Post
`
`2?
`'.-'5
`36
`50
`14
`2
`1
`
`59
`5.0
`
`53
`84
`86
`83
`
`Pre
`
`1.0
`4.3
`4.4
`1.8
`0.5
`0.1
`0.2
`
`1.1
`0.3
`
`3.9
`15
`5.5
`4.9
`
`Post
`
`Pre
`
`Day 21
`
`Day 210
`
`Pre
`
`Post
`
`Hb Increase lgJ'dL]
`
`1.4
`4.3
`5.2
`3.2
`0.6
`0.2
`0.1
`0.2
`3.0
`0.4
`
`4.2-‘
`8.6
`7.5
`6.?
`
`11' 1 0.2
`1 7 _ 0.1
`
`2.9 1 0.1
`3.9 -1 0.1
`2.? 1 1.9
`1.3 1 0.2
`1.B 1 0.2
`2.0 1 0.3
`2.9 1 0.8
`
`1.5 1 0.1
`1.? 1 0.2
`2.4 1 0.1
`1.5 1 0.2
`
`2.6 1 0.3
`1.0 1 0.1
`
`1.8 1 0.1
`3.1 1 0.4
`1.9 2': 0.1
`2.6 1 0.2
`2.2 1 0.5
`0.’.-' 1 0.1
`1.? 1 0.1
`
`1.0 1 0.1
`2.6 1 0.3
`2.0 1 0.1
`
`2.0 1 0.3
`
`3.1
`
`2.5
`
`1.5
`
`1.6
`
`1.? 1 0.1
`
`2.2 1 0.1
`
`2.2
`
`2.1
`
`-2.?
`-0.2
`+0.4
`-2.0
`+0.5
`-1.21
`-0.21
`-2.61
`-3.4
`-0.1
`
`+1.2
`+2.3
`+1.8
`+2.3
`
`" Patient was transfused before commencing sodium phenylbutyrate.
`1 Patient was transfused while on sodium phenvlbutyrate therapy.
`
`the ingestion of true tablets was the most common adverse
`effect, being reported by seven of the twelve patients. Two
`patients. both splenectornized and not on regular penicillin
`prophylaxis, had nonfatal episodes of bacterial septicemia:
`patient 4 developed infection with Streptococcus pneumo-
`niae at day 71 and Plesiomonas shigefloides at day 200, and
`patient no. 6 developed infection with Staphylococcus epi-
`d8rmfdf.§ related to a indwelling central venous catheter at
`day 24. Patient 8 suffered a hemorrhage from a gastric ulcer
`at day 220 soon after the commencement of aspirin therapy
`for long—sta.nding pulmonary hypertension. Patient 1, who
`had spun Hb levels between 5.1 and 7.5 g.-"dL associated with
`marked erytltroblastosis, developed spinal cord compression
`requiring irradiation at day 323 and ceased sodium phenylbu—
`tyrate therapy. Patient 10 developed a deep venous thrombo-
`sis at day 28 with a Hb 5.9 g;"dl_.. Three patients experienced
`bad body odor while on therapy. which in one. no. 12, caused
`her to be unable to tolerate the medication long—terrn, even
`at half the usual dose. These complaints are probably related
`to the in vivo ,6 oxidation of phenylbutyrate to phenylacetate,
`a compound with an offensive odor secreted as a defense
`mechanism by the stinkpot turtle.”
`
`DISCUSSION
`
`This study shows that sodium phenylbutyrate can safely
`be administered to patients with homozygous ,8 thalassemia
`and is well tolerated by the majority. The need to take 40
`tablets daily, epigasrric discomfort, and the body odor cre-
`ated in some patients are problematic. Poor compliance with
`this regimen, based on previous experience with this drug,
`was expected to be a frequent problem.” but surprisingly
`was not, possibly related to the fact that many of these pa-
`lients had had prior experience with other cumbersome medi-
`cal interventions including transfusion schedules and iron
`chelation therapy. However, the oral route of administration
`
`has clear advantages over the intravenous route needed for
`arginine butyrate, particularly because all the evidence avail-
`able suggests that in the management of the ,6 hernoglobinop-
`atbies, these therapies, if effective. will be needed long—terrn.
`We found that 36% (4! I 1) of all patients or 50% (4.18) of
`nontransfused patients responded to sodium phenylbutyrate
`when a response was defined as a sustained increase in Hb of
`more than 1 g!dL over pretreatment values. Clearly, sodium
`phenylbutyrate can increase Hb in some patients with homo-
`zygous ,6 thalassemia, but is not effective in all of them.
`Although it seems evident that 19-globin mutation alone does
`not predict response. the fact that two siblings treated in
`this study both responded to sodium phenylbutyrate therapy
`raises the possibility that some other genetic factor is in-
`volved. Other genetic factors linked and unlinlted to the ,8-
`globin locus have been shown to effect HbF levels in normal
`individuals and patients with ;3 hemoglobinopatI1ies.“"°
`The failure of Hb to increase in patients showing a de-
`crease in levels of lactate dehydrogenase and indirect biliru-
`bin is disappointing and raises interesting questions as to the
`cause of these changes if not related to decreased hemolysis.
`Similarly. we have observed increased production in F retic-
`ulocytes in all patients treated with this agent to date and
`the persistence of levels of F reticulocytes higher than base-
`line in some patients up to a month or more after the cessa-
`tion of therapy with an agent known to be rapidly metabo-
`lized and excreted. Similar observations have been reported
`after the use of arginine butyrate.2“‘" This uniformity of F-
`reticulocyte response. persistence of response in some pa-
`tients long after the cessation of trlerapya and lack of correla-
`tion between changes in F reticulocytes and increased total
`Hb or increased absolute HbF production may indicate sub-
`optimal increases in HbF insufficient to decrease ineffective
`erythropoiesis.
`The lack of con'elation between changes in o:- to non-
`
`
`
`

`
`43
`
`COLLINS ET AL
`
`at-globin ratios. measured both as mRNA or globin-chain
`synthesis, and response to therapy raises more questions as
`to the mechanism of action of these fatty acid compounds.
`In this study. we were unable to show a correction in C! to
`non—ar ratios in patients experiencing increases in Hb and
`observed a fall
`in total Hb in one nontransfused patient
`whose HhF production clearly increased. The recent demon-
`stration of induction of r.r— as well as y—g|obin production in
`butyrate-treated transgenic mice (G. Starnatoyannopoulos.
`personal communication. December I993) may help explain
`some of these observations and suggests that viewing these
`compounds as selective inducers of y—gl0bin production may
`have been premature. Similar data exists concerning the use
`of intravenous arginine butyratc. which was initially reported
`as a selective stimulator of the human )-—globin gene pro-
`moter capable of correcting a— to non—a—globin—chain imbal-
`ance in patients with thalassemia.“ Although early reports
`linking correction of globin~chain imbalance to subsequent
`increase in Hb with arginine butyrate were encouraging.
`these changes have not been reproduced in other patients
`showing a response to therapy.” Similarly. the slow nature
`of the response to therapy reported in the current study. with
`gradual increases in Hb over 200 to 500 days on therapy is
`not consistent with a mechanism invoking rapid correction
`of globin-chain ratios over a few days or weeks.
`We observed an inconsistent response to therapy, a de-
`crease in traditional indicators of hemolysis in all nontrans—
`fused patients that were not predictive of an increase in Hb.
`and increases in Hb not entirely explained by increased HbF
`in those patients who did respond to therapy. This suggests
`that classic Hb switching. an increase in 7-globin production
`with resultant decrease in globirt-chain imbalance. could not
`explain the increases in Hb seen. Three possible explanations
`exist: sodium phenylhutyrate (I) caused nonspecific induc-
`tion of all globin production.
`ie. or.
`;3. 7. and not just -y
`alone; (2) caused nonspecific expansion of RBC mass by
`the release of thalassemic RBCs previously sequestered in
`the marrow or by an increase in production of thalassemic
`RBCS; or (3) caused a prolongation of RBC survival without
`a change in RBC production. There is evidence to support
`the first of these (0. Stamatoyannopoulos. personal commu-
`nication. December I993) and the latter two lead to testable
`hypotheses in further patients.
`The positive correlation between baseline serum erythro-
`poietin level and the likelihood of responding to therapy is
`very interesting. This observation. together with the fact that
`erythropoietin levels in homozygous ,6 thalassemia are gen-
`erally elevated. bul inappropriately so for the degree of ane-
`mia, suggests that clinical trials of combination therapy using
`erythropoietin with sodium phenylbutyrate may be of value.
`However. it must be remembered that in these patients. eryth-
`ropoietin levels are related to other factors. such as baseline
`HbF%”” and as such. erythropoietin may only be a marker
`of some other factor affecting response.
`It is also of interest that an increasing number of structur-
`ally relatcd compounds are being shown to increase HbF
`production in vitro and in vivo in animals. hematologically
`nonnal individuals. and patients with .6 hemoglobinopathies.
`Most recently. we have discovered that valproic acid, a drug
`used for the treatment of epilepsy for the past 20 years.
`
`causes increased HbF production in nonanemic individuals‘?
`just as sodium phenylbutyrate was shown to increase HbF
`production in patients with urea—cycle disorders.” Adding
`to the intrigue is the fact that the compounds reported to
`date share a relatively small part of their structure in common
`and appear to be effective at vastly different molar dosages.
`Although mechanistically intriguing. and clearly of value
`to a selected group of patients. butyrate and other fatty acid
`compounds remain far from being a panacea for all patients
`with fl hernoglobinopathies. The pharrnacologic manipula-
`tion of HbF remains an inexact and as yet unproven therapy
`for the .6-globin disorders.
`
`ACKNOWLEDGMENT
`
`We thank Dr A.W. Nienhuis for patient referral and helpful discus-
`sion; Dr H.E. Witkowskzt for performance of the globin—chain analy-
`sis. C.D. Boehm for Cr and ,I3—gl0bin DNA analysis; A.C. Mayes
`and SH. Purvis for F—cel| and F—reticulocyte assays: and Dr T.R.
`Bishop and L. Frelin for help with the RN.-Xase protection assays.
`
`REFERENCES
`
`l. Zurlo MC. De Stefano P. Borgna—Pignatti C. Di Palma A. Piga
`A, Melevendi C. Di Gregorio F. Buratini MG. Teizoli 5: Survival
`and causes of death in thalassemia major. Lancet 2:27. 1989
`2. Lucarelli G. Galimberti M. Polchi F. Angeluoei E. Baronciani
`D. Giardini C. Andneani M. Agostinelli F. Alb-ertini F. Clift RA:
`Marrow transplantation in patients with thalassemia responsive to
`iron chelation therapy. N Engl J Med 329:8-1-0. [993
`3. Nienhuis AW. Ley T]. Humphries RK. Young NS, Dover G:
`Pharmacological manipulation of fetal hemoglobin synthesis in pa-
`tients with severe .6—tha]assemia. Ann NY Acad Sci 4452198. 1985
`4. Stamatoyannopoulos J A. Nienhuis AW: Therapeu