Elevation of Total Homocysteine in the Serum of Patients with Cobalamin
`or Folate Deficiency Detected by capillary Gas
`Chromatography-Mass spectrometry
`
`Sdly P. Sim" Paul D. Ilarcollf Elaine R. Podo|l,' Robert H. Alon.‘ David 6. Savage,‘ and John undenbaurn’
`‘Divisions ofHematolog and Oncology, Department ofMedicine. and the Department ofBiochemistry, Biophysics, and Genetics.
`University ofColorado Health Sciences Center, Denver, Colorado 80262.’ and ‘Departments ofMedicine. Harlem Hospital Center.
`Columbia-Presbyterian Medical Center and College ofPhysicians & Surgeons. Columbia University, New York, New York 10032
`
`Abstract
`
`Todetermlnelflevelsofserumtotalhomocysteineareelevated
`in patients with either cobalamln or fohte deficiency, we uti-
`llzerlaneweapillarygaschrornatographlc-mass spectrometric
`technique to measure total homocysteine in the serum of 78
`patients with clinicallyconfirrned cohslarnin deficiency and 19
`patients with clinically confirmed folate deficiency. Values
`nngedfromll to476umol/llterlnthecobalemin-deficient
`patientsand77ofthe78patientsharlvaluesabovethenormal
`rangeof7-22umol/llterasdetermlnedforstlnormalblood
`douors.Inthe cobnlamin-rletlcient patients,serunr total homo-
`cysteinewaspositively correlatedwithserumt'ohte,meaneor-
`puscularvolume,sernrnlrtctate dehydrogenase, serum methyl-
`malonicaeid,andthedegreeol‘neurologic involvement,and
`inversely correlated with platelets and hematocrlt. ln the to-
`late-rleficlent patients, values for serum total homocystelne
`rnnaedftoml7tol85urnol/llterandl8ol’thel9pstientshad
`valuecahovethenormalrangesomepatlentswlthpernicious
`anemia who were intermittently treated with cyanocobalamln
`were fomdtohave elevatedserumlevels oftotal homocystelne
`whiletheywenefreeofhenIatologicandneuI'ologicabnornuli-
`ties/l'hemeasnrernento{serrrnrtotulhon1ocysteinewillhelp
`definethelnclrlenceofcohalamln defidencyand fohtedefl-
`clencylnvariouspetientpopulations.
`
`Introduction
`
`At present the diagnosis of cobalarrrin (Cbl,' vitamin B12)
`deficiency and folate deficiency are almost entirely dependent
`on the demonstration of low serum levels of the vitamins in
`patients with supportive clinical and laboratory findings. The
`limitations in using only the serum Cbl level to detect Cbl
`deficiency have been widely recognized (1-4). For example,
`the significance of the high incidence of low serum Cbl values
`in elderly subjects without hematologic abnormalities (5, 6),
`
`Address correspondence to Dr. Stablcr, Div. of Hematology. Campus
`Box B170, University ofColorado Health Sciencescenter, 4200 E. 9th
`Ave., Denver, CO 80262.
`Receivedforpublication 20 March 1987 and in revisedkmr 7 July
`I987.
`
`l. Abbreviations used in this paper: Cbl, cobalamin; CN-CbL cyano-
`Cbl; l-lct, hematocrit; LDI-l, lactate dehydrogenase; MCV, mean cor-
`puscular volume.
`
`J. Clin. Invest.
`0 The American Society for Clinical Investigation, Inc.
`0021-9738/88l02l0-166/09 32.00
`Volume Bl, February 1988, 466-474
`
`and in patients who have various neuropsychiatric abnormali-
`ties (7, 8). is not known. Other tests (9), such as the deoxyuri-
`dine supprcsion test, have not been widely used to assess pa-
`tients for Cbl deficiency.
`The diagnosis of folate deficiency can also be problematic
`for several reasons. For instance, after acute dietary depriva-
`tion, serum folate levels may be decreased, although tissue
`folate levels are adequate (10). In the setting of chronic alco-
`holism, the laboratory features of megaloblastic anemia due to
`folate deficiency may be confused by concurrent illness.
`Serum and red blood cell folate may be normal in patients
`with alcoholism and megaloblastic anemia (l 1). Finally. be-
`cause many of the clinical and laboratory features of Cbl and
`folatc deficiency are similar, it is often difficult to
`between them (l2), and the administration of folic acid alone
`to a patient with Cbl deficiency is dangerous (l3). Therefore, as
`in Cbl deficiency syndromes, additional diagnostic tests for
`folate deficiency would be useful.
`In both Cbl and folate deficiency, it is likely that there is
`reduced activity of the C121-dependent enzyme, methionine
`synthetasc (tetrahydropteroylglutamate methyltransferase),
`which simultaneously methylates homocysteinc to methionine
`while dernethylating N’-methyltetrahydrofolate to tetrahydro-
`folate, as shown in Fig 1. The reduced folates formed from this
`reaction are necessary for thymidine synthesis and ultimately
`DNA synthesis, and the methionine formed participates in the
`methylation of many compounds (14). If methionine synthe-
`tase activity is decreased by a deficiency ofCbl or folate, serum
`homocysteine levels might increase and methionine levels
`might decrease, although regulation of other enzymes in the
`pathway shown in Fig. 1 might maintain levels of one or both
`of these amino acids within normal limits. A few (15-20) but
`not all (21-24) children with severe Cbl deficiency have been
`reported to have increased levels of homocystine in their urine
`and plasma, along with decreased levels ofplasma methionine.
`An early study (25) reported that plasma methionine levels
`were low in adults with Cbl deficiency. In addition, excretion
`of large amounts of homocystine in the urine has been re-
`ported in patients with various inherited defects alfccting the
`activity of methionine synthetme such as the inability to form
`methylcobalamin (Cbl C and Cbl D mutants) (26, 27), and an
`inability to form N’-rnethyltetrahydrofolate due to 5, l0-mcth-
`ylenetetrahydrofolate rcductasc deficiency (28, 29). However,
`homocystinuria was not prwent in patients with transcobala-
`min ll deficiency (30), a patient with a lysosomal block in Cbl
`transport (31), or in a patient with congenital folate malab-
`sorption (32), all ofwhom had hematologic abnormalities that
`would be associated with a decrease in methionine synthetasc
`activity.
`We have developed a new capillary gas chromatographic-
`
`!” S. P. Stabler. P. D. Ma:-cell. E. R. Podell, R H. Allen. D. G. Savage. and]. Lindenbaum
`
`Lilly Ex. 2094
`Sandoz v. Lilly IPR2016-00318
`
`

`
`8-ADENOSVLMETHOONINE
`
`KINUIIQ
`Adonooylrannhraao
`
`uouocvsrsms
`
`Serum jCystamlontne
`
`IWVMNSG
`
`RNA
`Pathways
`
`er-KETO-‘I-METNIOLBUTV
`
`L...
`
`ON:-8-OH!-CNI'0O0H
`3-METKYLTNIO PROPIONATE
`
`I-I
`
`s—cu,-en;-<::-coon
`-:
`"*-
`en.-rl;-coon
`NH:
`
`'3
`1-eyeramlonau
`_.___. us-cu.-<':—coou
`pm,
`CYSTEINE
`
`o on.-cu.-<_:-coon
`0
`u-KE'I’O3UTYRAI'E
`
`l
`‘
`30¢" * 09!
`
`Figure 1. Various pathways involved in the metabolism of methionine including the methyloobalamin-dependent methylation of homocysteine
`to form methionine by the enzyme, methionine synthetase.
`
`mass spectrometric assay for serum methylmalonic acid and
`have shown that 69 out of 73 patients with clinically con-
`firmed Cbl deficiency have elevated values (33). We have re-
`cently developed similar assays (24) for the quantitation of
`total homocysteine, methionine, and total cysteine in serum
`and now report on the measurement ofthese amino acids in a
`large number of patients with clinically confirmed deficiencies
`of Cbl and folate.
`
`Methods
`
`Assay oftotal homoeysteine.’ methionine. and total cysteine.’ Serum
`total homocysteine, methionine. and total cysteine were assayed using
`capillary gas chromatography—mass spectrometry as previously de-
`scribed in detail (34). Briefly, 50 ui of H30 containing 5 nmol of
`D,L-[3,3,3’,3’4,4,4'4'-’H.]homocystine (98.4%), 15 nmol of L-[methyl-
`‘I-iglmethionine (98%), and 25 nmol of D,L-[3,3,3',3'-’H.]cystine
`(98%), is added to I00 ul of human serum. The sample is then heated
`with 2-mercaptoethanol in order to reduce and release endogenous
`homocysteine and cysteine front proteins and other disulfides and to
`equilibrate them with their stable isotope internal standards, which are
`also reduced during this procedure. Protein is then precipitated with
`sulfonlicylic acid and the supernatant is partially purified by sequen-
`tial cation exdrange and anion exchange chromatography. The r—bu-
`tyldimethylsilyl derivatives of the amino acids are formed with N-
`methyl-(1-butyldimethylsilyl)trifluoroaeetan1ide, followed by their ex-
`traction into hexane and volume reduction using a stream ofnitrogen.
`
`2. The term “total homocysteine" as applied to biologic samples such
`as serum and urine refers to the sum of homocysteine and the homo-
`cysteine that is linked via disulfide bond formation in a variety of
`compounds that include homocystine (homocysteine-homocysteine
`disulfide), homocysteine-cysteine mixed disulfide, proteins via their
`cysteine moieriesandpeptidessuchasglutatltioneviatheircysteine
`moieties. The term “total cysteine” is used in the same way and refers
`to the sum of cysteine and the cysteine linked via disulfide bond for-
`mation in compounds such as cystine (cysteine-cysteine disulfide),
`homocysteine-cysteine mixed disulfide, and proteins and peptides via
`their cysteine moieties.
`
`The samples are analyzed on a Durabond DB-l fused silica capillary
`column (30 m X 0.25 mm i.d.. 0.25 am film thickness) from I&W
`Scientific, Inc. (Rancho Cordova, CA) and a Hewlett-Packard Co.
`(Palo Alto. CA) 59928 as chromatosraph-mass spectrometer
`equipped with a falling needle injector. Quantitation is based on the
`ratio ofthe areas ofthe base peak ion 420.2 for homocysteine, 320.2 for
`methionine, and 406.2 for cysteine, each of which elutes at a dilferent
`time, to the areas ofthe base peak ions of 424.2, 323.2, and 408.2 for
`the derivatives of their respective stable isotope internal standards.
`Total homocysteine, methionine. and total cysteine are very stable in
`collected serum since no change or trend in values for these amino
`acids wm noted in a sample ofpooled normal human serum that was
`frozen. thawed. and assayed > 30 times overa 12-mo period.
`Values of serum total homocysteine obtained with blood samples
`that were drawn and immediately centrifuged at 4°C were the same
`(< l0% ditference) 3 those obtained with portions of the same blood
`samples that were incubated at room temperature for l h before cen-
`trifugation, but increased by ~ 35 and 75% when the incubation was
`prolonged for 4 and 24 h, respectively, before oentrifugation. Values
`for serum methionine were unchanyd at l h and increased by 10 and
`25% at 4 and 24 h, respectively. Values for serum total cysteine were
`unchanged over the 24-h incubation period. Values for urine total
`homocysteine, total cysteine, and total methionine were unchanged
`when urine samples were incubated at room temperature for 0-24 h.
`Senrm samples and the internal standards of homocysteine, methio-
`nine, and cysteine were stable for > 1 yr, based on their gas chromato-
`graphio—mass spectrometric behavior, when stored at -20°C in be-
`tween numerous freezings and thawing over this time period.
`Subjects and patients. Serum samples from 50 normal blood
`donors, 25 malesand 25 females, ranginginagr: from l8 to65 yrwere
`obtained as described previously (35). Patient samples were selected by
`Dr. Lindenbaum from an extensive serum collection that has been
`amembled over the past I 5 yr. Both the normal and the patient samples
`wereallowedtoclot for ~ 1-4hatroomtemperaturebeforetheserunt
`was removed and stored at —20°C. Thus. postcolbction increases for
`total homocysteine and methionine should have been modest and
`similar in both the normal and patient groups (see above). 'I'he diag-
`nosis ofCbl deficiency was based on low serum Cbl levels, megaloblas-
`tic bone marrow morphology, appropriate hematologic or neurologic
`abnormalities, and a significant response to treatment with parenteral
`Cbl. The diagnosis of pernicious anemia wa based on an abnormal
`
`Lilly Ex. 2094
`Sandoz v. Lilly IPR2016-00318
`
`

`
`3
`
`Methionine0rmoIIL)cysteine(umol/L) 83
`
`
`
`
`Homocyatelne(umo|IL) 53
`
`§ §
`
`in I 50)
`
`(It I 78)
`
`Intreauently
`Trtfllcd
`(rt-18)
`
`deficient
`(n-19)
`
`Figure 2. Levels of serum total homocysteine (bottom), serum methi-
`onine (middle), and serum total cysteine (top) in patients with clini-
`cally confirmed Cbl deficiency, fclate deficiency, and Cbl-deficient
`infrequently treated patients who had no hematologic or neurologic
`abnormalities, using capillary gas chromatography-mass spectrome-
`try. The normal range for total homoeysteine is (7-22 umol/liter),
`for methionine (14-44 umol/liter), and for cysteine (I73-378 umol/
`liter). These ranges were calculated as the meantz SD alter log trans-
`formation to correct for skewness towards higher values.
`
`To determine ifthc elevations of serum total homocysteine
`were due to Cbl or folate deficiency and not due to illness in
`general, we studied 25 consecutive patients (13 hospitalized,
`12 outpatients; age, 25-86 yr) who had sera submitted for
`serum Cbl and serum folate assays and in whom the Cbl values
`were > 500 pg/ml (normal, 200-1,000 pg/ml) and the folate
`values were > 5 ng/ml (normal, 3-20 ng/ml). Of the 25 pa-
`tients, 2l had values within the normal range of 7-22 pmo1/
`liter (mean for the 21 patients, 13.7 umol/liter) and I had a
`very slightly elevated value of 22.5 umol/liter. Significantly
`elevated values of 32.6 and 27.4 pmol/liter were obwrved in
`two patients with chronic renal failure who had serum creati-
`nine values of 5.6 and l3.3 mg/dl, respectively. Various forms
`of homocysteine have been previously reported to be elevated
`in the plasma of patients with renal failure (36, 37). We are
`currently studying the incidence of elevated total hornocys-
`teine values in chronic renal failure. Preliminary results indi-
`Cate that total homocystcine is elevated in many but not all
`such patients. A significantly elevated value of 34.4 umol/liter
`was also observed in one patient with Cbl deficiency since
`1984, who had been intermittently treated with Cbl. The pa-
`tient had not been treated for several months but had a serum
`Cbl of > 2,000 pyml because a Cbl injection was given 10-15
`min before the blood was drawn for the Cbl assay. Taken
`
`Schilling test that corrected with exogenous intrinsic factor and/or the
`presence ofanti-intrinsic factor-blocking antibodies in the serum. The
`diagnosis of folate deficiency was based on low serum folate values,
`normal orelevated serum Cbl values, megalobiastic bone marrow mor-
`phology, appropriate hematologic abnormalities. and responses to folic
`acid therapy. In 17 ofthe patients there was a history ofalcoholism and
`poor diet; one had history of poor diet alone, and one had tropical
`sprue. The samples in the Cbl-deficient infrequently treated group
`were from patients with pernicious anemia who were previously diag-
`nosed as Cbl deficient as dwcribed above, but who received only inter-
`mittent treatment with parenteral Cbl at intervals of 6-9 mo due to
`poorcomplianceoraspartofstudiesofCbl requiremenrstobere-
`ported elsewhere. They had low, borderline, or normal levels for serum
`Cbl, lacked hematologic and neurologic abnormalities, and were
`asymptomatic at the time the sampleswere collected. Serum Cbl levels
`were assayed using the Laclobacillus Ieichmannii method or a number
`ofradiodilution assays utilifing purified intrinsic factor or gastric‘juice
`with > 95% ofCbl binding activity due to intrinsic factor. These assays
`were performed in Dr. l..indenbaurn’s laboratory where it has been
`found that very similar patient values and normal ranges are obtained
`with all of the various Cbl assays. Serum folate was assayed with the
`Lactobacillus casei method or by milk binder radiodilution assay.
`Most of the patients‘ samples were coded in a manner such that the
`categories to which they belonged and the number of patients in each
`category were not known by the personnel involved in the perfor-
`mance of the total homocysteine, methionine, and total cysteine
`assays, until alter the results were reported to Dr. Lindenbaum. A large
`number of Dr. Lindenbaum's samples from patients without Cbl or
`folate deficiency have been assayed and found to have normal values
`for these amino acids. thus ruling out the possibility that elevated
`values might arise due to storage in the freezers used in Dr. Linden-
`baum’s laboratory.
`Statistical methods. A number of factors were examined individu-
`ally for possible relationships with serum total homocysteine, methio-
`nine,andtotalcysteine. Forfactorsthatwerediscreet,suchassex,raoe,
`anddiagnosis, the Wilcoxontwo-sampletestwasusedtodeterminethe
`significance of the relation. For assessing possible relationships with
`neurologic severity, youpa 0, l, and 2 (as defined in Table I) were
`combined and compared with combined groups 3 and 4. Fauors that
`were continuous, such as age or mean corpuscular volume (MCV),
`were examined using Spearman correlation coefiicients
`
`Results
`
`The values obtained for serum total homocystcine, methio-
`nine, and serum total cysteine for the normal subjects and
`patients in the various categories are shown in Fig. 2. In the
`Cbl-deficient group, 77 ofthe 78 patients had values for serum
`total homocysteine above the normal range of 7-22 umol/liter.
`The highest value was 476 pmol/liter and the median value
`was 113 umol/liter. In the folate-deficient group, I8 of the I9
`patients had serum total homocysteine levels above the nor-
`mal range. The highest value in this group was 185 umol/liter
`and the median value was 67 umol/liter. In the Cbl-deficient
`infrequently treated group, 13 of the 18 pafients had elevated
`values that ranged as high as 47 pmol/liter at a time when they
`did not have hematologic or neurologic abnormalities Of the
`thirteen patients with elevated total homocysteine values,
`serum L. Ieichmannii Cbl concentrations were low (88-180
`pg/ml) in nine, two others were assayed by radiodilution and
`werebothlowat 115 pg/ml,andtwowere normalbytheL.
`leichmannii method at 205 and 275 pg/ml, respectively. Ofthe
`five patients with normal total homocysteine values, two had
`low Cbl values by radioassay and one by the L. Ieichmannii
`method.
`
`468
`
`S. P. Slabler. P. D. Marcell, E. R. Podell. R. H. Allen. D. G. Savage. and .I. Lindenbaum
`
`Lilly Ex. 2094
`Sandoz v. Lilly IPR2016-00318
`
`

`
`together, these results indicate that serum total homocysteine
`is not elevated in most ill patients.
`In the Cbl-deficient group values for serum methionine
`were below the normal range of 14-44 umol/liter in only 2 of
`the 78 patients and above the normal range in 20 patients. Of
`the l9 folate-deficient patients, 3 had values above the normal
`range as did 3 of the 18 Cbl-deficient infrequently treated pa-
`tients. None ofthe folate-deficient patients and only one ofthe
`Cbl-deficient infrequently treated patients had a value for
`serum methionine below the normal range.
`Total serum cysteine was below the normal range in 11 of
`78 Cbl-deficient patients, 4 of 19 folate-deficient patients, and
`none of the Cbl-deficient infrequently treated patients. It was
`above the normal range in 7 of 78 Cbl-deficient patients, 1 of
`the 19 folate-deficient patients, and 3 of the 18 Cbl-deficient
`infrequently treated patients.
`The clinical data for the 78 Cbl-deficient patients in addi-
`tion to their serum total homocysteine, methionine, total cys-
`teine, and methylmalonic acid values are shown in Table I.
`They are arranged in descending order of their serum total
`homocysteine values. There was a significant positive correla-
`tion between serum total homocysteine and serum folate (r
`= 0.27, P < 0.05), MCV (r = 0.23, P < 0.05), serum lactate
`dehydrogenase (LDH) (r = 0.34, P < 0.01), and serum meth-
`ylmalonic acid (r = 0.74, P < 0.0001). There was a significant
`negative correlation between serurrr total homocysteine and
`platelets (r = -0.26, P < 0.05), and hematocrit (I-lct) (r
`= -0.26, P < 0.05). Patients with more severe neurologic ab-
`normalities (groups 3 and 4) had higher serum total homocys-
`teine levels (166198 mean-LSD, median I45 umol/liter) than
`those with no or less severe abnormalities (groups 0-2)
`(105165 meaniSEM, median 89 umol/liter) (P < 0.01).
`Semm total homocysteine was not significantly correlated
`with serum Cbl (r - -0.20, P < 0.08). There was no significant
`correlation between serum total homocysteine and any of the
`following: serum methionine, serum total cysteine, sex, race,
`etiology, presence of anti-intrinsic factor antibody, and pres-
`ence of glossitis.
`Serum methionine was not significantly correlated with
`any ofthe parameters mentioned above and significant difl'er-
`ences were not observed between any ofthe various subgroups
`The hematologic data presented in Table I on the 77 Cbl-
`deficient patients with elevated levels of total homocysteine,
`demonstrates wide variations in their degree of anemia, ma-
`crocytosis, and other abnormalities. For instance, only 32
`(42%) had a severe anemia (Hct < 25%), while 27 (35%) had
`only a moderate degree of anemia (Hct, 25-34% for females,
`25-39% for males), and 18 (23%) were not anemic at all. Only
`45 (58%) of the patients had a marked elevation in MCV
`(> I10 fl), while 23 (30%) had only a moderate elevation of
`MCV (I01-l I0 ll), and 9 (12%) had a normal MCV. There
`was a wide range of serum Cbl levels in these patients also;
`only 47 (61%) had markedly decreased levels to < 100 pg/ml,
`while the other 30 (39%) had levels between 100 and 200
`D3/ml.
`The clinical data on the 19 patients with folate deficiency
`along with their serum total homocysteine, methionine, total
`cysteine, and methylmalonic acid values are also shown in
`Table l. Because of the small number of patients tested, corre-
`lations between their clinical and laboratory abnormalities
`were not evaluated.
`We have previously reported that serum methylmalonic
`
`acid is increased in 95% ofpatients with Cbl deficiency (33). As
`shown in Table I, 74 of the 78 Cbl-deficient patients had
`methylmalonic acid levels above the normal range of I9-76
`ng/rnl (median, 1,240 ng/ml, range, 78 to 2,300 ng/ml). The
`one Cbl-deficient patient (No. 78) with a serum total homo-
`cysteine level within the normal range also had a normal
`serum methylmalonic acid. His diagnosis was tropical sprue.
`The other three patients with normal serum methylmalonic
`acid levels (Nos. 49, 61, and 73) all had elevated serum total
`homocysteine levels ranging from 34 to 92 amol/liter. Their
`diagnoses consisted of tropical sprue. postgastrectomy syn-
`drome, and pernicious anemia.
`As we have reported previously (33), most folate-deficient
`patients have normal levels of serum methylmalonic acid, al-
`though sonre have mild elevations. In the current study, as
`shown in Table I, I3 of the 19 folate-deficient patients had
`normal values for serum methylmalonic acid, and the other 6
`had mild elevations (median for the 6 patients, 132 ng/ml,
`range, 79 to I95 ng/ml).
`Fig. 3 shows the serum and urine homocysteine levels both
`before and afier treatment in a patient with classic pernicious
`anemia and in another patient with alcoholism and nutritional
`folate deficiency. The Cbl-deficient patient had markedly ele-
`vated levels of total homocysteine both in serum and urine,
`which fell into the normal range within 3 d of treatment with
`parenteral cyano-Cbl (CN-Cbl). This same patient has been
`previously shown (33) to have a similar rapid fall in serum and
`urine methylmalonic acid in response to this course ofCN-Cbl
`treatment. The patient with folate deficiency had an elevated
`serum total homocysteine which over several days decreased to
`theupperborderofthe normalrangeafierseveraldosesoforal
`folic acid. The urine total homocysteine which was at the
`upper border of the normal range also fell markedly after
`treatment. This data suggests that serum and urine homocys-
`teine levels correlate with each other in patients with Cbl or
`folate deficiency and could be used to monitor the response of
`treatment, although more patients would need to be evaluated
`before this could be concluded with certainty. Initial serunr
`and urine methionine levels were normal in both patients and
`did not change significantly alter CN-Cbl or folate treatment
`(data not shown).
`
`Our studies demonstrate that serum total homocysteine levels
`are likely to be clinically useful, since they were above the
`normal range in 77 of 78 Cbl-deficient patients and 18 of 19
`folate-deficient patients. The serum total homocysteine ap-
`pears to be similar in sensitivity to the serum methylmalonic
`acid in the Cbl-deficient patients, since the latter was elevated
`in 74 of the same 78 Cbl-deficient patients. Using the serum
`total homocysteine level in combination with the serum
`methylmalonic acid level will ofien be helpful in distinguishing
`patients with Cbl deficiency from those with folate deficiency,
`since most patients with folate deficiency have normal levels of
`serum methylmalonic acid and the rest have only mild eleva-
`tions.
`
`Additional clinical studies using measurements of serum
`total homocysteine and serum methylmalonic acid in addition
`to measurements of serum Cbl and serum folate should make
`it possible to further define the clinical spectra of Cbl and
`folate deficiency, and to define the proper diagnostic approach
`
`Lilly Ex. 2094
`Sandoz v. Lilly IPR2016-00318
`
`

`
`.55»0135:.
`
`
`
`
`
`383.5_I§EEowaaum93¢.-a_.5oa..8u£._.o:bdP..-92_.=Em3.6.9833
`
`
`
`
`
`
`
`
`
`
`
`333332.3853..§._§..335.303893.33::‘EH20$0313:3n~u_8nEIto>US
`
`
`
`
`
`
`
`.-
`...
`...
`w.SRS2an2n_on9«N333ana_RK2an_92anSSE_N23R0._..R.9w_3.w_S3§9.
`
`$2253
`
`—-—
`
`8~.-
`
`o+cc++++++++oo+++o++++o+o++o+o++oc+++++o+c+o+++
`
`+ooo+o¢oo++o++
`
`O++ ++OOO++O+
`
`oo++c o+++
`
`+o+oc>
`
`nnOM—oQMV--Nonvmmumonm
`
`N-O -nonr~u-sv—e-acoo
`
`PGDMO
`
`VNMNN
`
`§3%<9og:
`
`83$Qfin2.
`
`28m95.3.
`
`8B§a2i.8.
`
`8...:83..-_
`
`33$?Qfin2.
`
`2§3nRén
`
`253”Riv2.
`
`o8.o:82:_9§2~Rén5.
`
`ooowno81¢m~_
`
`3283m:
`
`252.Qfln83283>9
`
`$3“83me
`
`3§§¥95m3_
`
`R§2_Rim2.
`
`335385u:
`
`833835958.8§n:
`
`8o.8n83.
`
`9§5_Riv8.
`
`QSN2Ȥ_
`
`c8.8n82.2
`
`2§3so2.
`
`22m.§§
`
`98Sin2_
`
`93.ea82§8_.§¢8.
`
`3:&<Rim8.
`
`%38<83.o:
`
`98mRio:_
`
`3:.a8<8...»2_§3€<83u:
`
`23gfin8.
`
`efin
`
`2:81.S.
`
`
`
`§..S2§§.§o..i§=\~o..i...§.3.31\¢Z3...:KR
`
`
`
`
`
`
`
`24.54:3:12:2Rmvo8.o$.¢8..x_c8.o_.o8...878~n.$2Sum8o._.8~
`
`
`
`
`
`2%.an8.
`
`o8.a283on
`
`2$_Qfinn2
`
`98.Rim3.
`
`858Riv8.
`
`23QE823aRio8.
`
`:2.833.
`
`2§8_Riv8.3:835:
`
`2.2..-
`
`R£S$8§3€322
`
`-ON_m_
`
`33333393§§3fi
`
`--—-—NN——N—n—n—c-snvnn
`-v.—-
`-9.-nu-~ox~a~o~---A-¢oonvo—vooomv~r
`
`can
`
`2....
`
`ow_Sn20.0_an
`
`5u3_3as2_Q3
`
`Q0.
`
`an33ea_
`
`3:
`
`on
`
`3..
`
`can
`
`Ev
`
`we
`
`e2can
`
`Nd
`
`>n33...:
`
`VNcan
`
`93
`
`«.32
`
`0.3
`
`...‘
`
`o5V.»3Q2
`
`...
`3%;
`
`§899§2£39R$9
`
`nn=3n=Bnnnnn3:n=nnnn=:=B=B33nnn=::nn:B=3=n3Bnnn
`
`mmfimmmzmmizmmimmhmiimuzzmmummmmiiumzmmmzmummmmfi
`
`§3$$FRt38$3$§3$%$%¥§$3$R8S83$S33fi3§3fi8$3E833893?
`
`—-NMQ»h»ol~m02
`
`g
`
`22:22:2285$2§33R88R:RR%RSR%$9S$93
`
`
`
`
`
`..O=w...u§.~Q256%5...:a§.§m3‘:6Au:m.§\wQEU5...:h§.:a.~3.B.\.SuQ~uu.E.:U.~3.5
`
`
`
`
`
`470
`
`S. P. Stabler. P. D. Marcell. E. R. Podell. R. H. Allen, D. G. Savage, and J. Lindenbaum
`
`Lilly Ex. 2094
`Sandoz v. Lilly IPR2016-00318
`
`

`
`Na8..an<._8Sn<L
`
`2.an80..~53EN.3E.
`
`on8:Sn<.._2onanS“.8ERan<L3Sn8.._<&88.S..._0..
`
`2.an3.<..3R.8n<..
`
`an8.sh<.._8.32.<..3an5.3Sn5.3SN%_u:.<..8m_N2.EQ.«numa.S.2S.<..
`
`hv8.<..23.asw»
`
`3.3..$3<L9muan<4-
`
`onQ.~an_<._23..Rm5..
`
`«nSnSn5.3_.uS0..
`
`on.3<..
`
`._8.an2.5oz3.<..._3an3_<a
`
`3.2..
`
`hean3in83.232Sn2ES.as2_.22.3.
`
`9..3haa:9S.33.98.8.32«R
`
`ES.3V3
`
`fill‘! ®f‘lPlOl© —N!N\l"\(V—-¢"I0l\—'04
`FOg-—O\h—2O\O\MNOO«N\VOQM
`
`5....3<
`
`an...u..<2...o..<R...343...Q._<2..Q1
`
`onA.o..<
`
`S....u..<
`
`8....34.2...u.<
`
`mo....3...
`
`Nu....3...
`
`3....u..<
`
`...u..<Kon
`
`2....u..<
`
`.......3...
`
`ana..u..<-mp
`
`o+ooeo++++++¢¢+++e+o++++ooo+++o
`
`OOQOOOGOOOOOOOOOOOO
`
`coo co++ +a++o+ooo++co+oo+ooooo
`
`++o+
`
`o+oeo o co co
`
`vo—vooon—ocooNNcvc~ovNoooooco o
`
`2
`ogoo omioo Eooo coo
`
`-~o.§..o..<.8.6
`
`83o8...§896
`
`N88.;83.
`
`8?8.8.83
`
`3.8.8a8.3
`
`:289.28...~
`
`338383
`
`.86:8....
`
`an83883
`
`
`
`9.c.n-8~.n
`
`8...$~8...:
`
`838.8.93.:
`
`o8.~..8.9.So...
`
`
`
`2....s:..u..<83
`
`8ao8.8m83
`
`R.858.!83
`
`.5o8.o..n83.
`
`3:...8<83
`
`
`
`8o.8.8......
`
`8.3..83
`
`.3.23.83.83
`
`8»o8....fi8....
`
`eao.§..8<So.»
`
`3.....8<8:
`
`..8.£~83
`
`89.:83.
`
`
`
`.889.3.c8...
`
`Soe8.o.m8....
`
`an8o.S~8....
`
`SNo8.3«83
`
`2..o8.$n8}.
`
`8?8.38.5
`
`Sn..8....n8..~
`
`892.806:.
`
`
`
`
`
`e8...8»;8. emuSod..832.cano8.S.~8......«SS.e8.8.8}2.8....
`
`mae8.a..8}.
`
`mano8.8~80.22.9339.3.838.
`
`3....89.38.2...~.
`
`Sno8.8~82..8.
`
`
`
`
`
`
`
`
`
`.qE>u_-8ao_o....o..on.u23632.30891399.uovoséfiu1.5289.38.88a8=..3.8:».3:358.n"335:».uo:3E8..8.=.>uu228m.3.833.3.o_.=.nn_=.1333d..91.....28_¢......o.m...._.
`
`
`
`
`
`
`
`
`
`
`
`9.380.3.8238....8.8.
`cc...2.Soda8!!1.ac8.8m83.8.
`
`8..c8...2.o8...~m8....“-_
`
`83....8....~..
`
`$593325332382232R:fifi$QQ33RR3%?fi
`
`=§.‘3‘.°°°2."~lR92'.'-‘;'~'3°£'°i‘2R‘£.'13¢'?.
`
`..n
`
`...~
`
`090.3
`
`MS
`
`«.1
`
`am
`
`0.._QM
`
`O.N_
`
`can
`
`N._
`
`Na
`
`n...
`
`...n.
`
`O.N_
`
`Nd
`
`asO.333...o
`
`3.
`
`DION.
`
`ad2.
`
`ad
`
`n.—_fin
`
`an3.
`
`no3
`
`N._33no
`
`n.N
`
`_._
`
`.32We
`
`..o2.22.
`
`.6o..
`
`no
`
`0..
`
`§S8:§§33§%3383§38B
`
`an:3n3nn=na=::3n3:n===:n3n::=:=
`
`mLEEzimiimmmiimmmummimummu2umuE
`
`EEE2mu2222m22m2Smmu
`
`323§$3§233£S%3§8$R2E§K3$S§33§3%
`
`$v‘¢‘9.2.'E3:';3?-°\'.:33B.'7.9.'3$.‘-‘i
`
`383E2333fiR%%8G$3$33S$$?Efi§§§§E§
`
`§.°.‘S;S3338S3&‘8:3.‘8333e';
`
`
`
`
`
`
`
`
`
`
`
`.8:J:..93..o..Eno....n.383.338Bow...En.._o._oo_u5:.90.3088>5¢aa....u....5...._.&in.358_83o_E..u._o>u-non5.:3858.39.."E33no398...ou-.5_>..o\.x.n._.....8....o.uo.E_!_E_=52.9
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`.38o..........._...83.3in88.9..3...SR.5.:§..2.8.¢....3...ca....8...e..2.xE88.fi.98...o..3.58.!o.u.Ex..<..3.3.5.....:__..a..o.......__._.9:._.o..88.
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`Lilly Ex. 2094
`Sandoz v. Lilly IPR2016-00318
`
`

`
`CM-Gb|(1nulM)
`
`Fohtettmulto)
`l
`I
`l
`I
`l
`
`i
`
`l
`
`
`
`
`
`serumHomocyetelne(umcllllterl(0—0)
`
`
`
`
`
`UrineHornocyetelnelurlnecreatlnlne(umolltomrnol)(0--0)
`
`
`
`
`
`in detail in this report. This advantage is not shared by the
`serum Cbl and serum folate levels, since these values are
`always elevated or at least normal alter treatment with Cbl or
`folate, respectively, regardless of whether a patient is Cblo or
`folate-deficient or not.
`
`That levels of senrm total homocysteine are almost always
`elevated in deficiencies of Cbl and folate, whereas levels of
`semm methionine are almost never decreased in these condi-
`tions, indicates that the pathways shown in Fig. 1 are regulated
`to a considerable degree. The maintenance of methionine
`levels within a relatively narrow range appears reasonable,
`since methionine is an essential amino acid that plays an im-
`portant role in protein synthesis, in various methylations, and
`in a large number of other metabolic reactions. The enzymatic
`reaction or reactions that represent the site or sites of regula-
`tion in Cbl deficiency and in folate deficiency are not known.
`Cystathione B-synthase (see Fig 1) could play a role, since
`regulation of this enzyme is important in the maintenance of
`methionine homeostasis during changes in dietary methionine
`(38-40).
`The fact that methionine levels are maintained in Cbl and
`folate deficiency at the expense of elevated levels of total ho-
`mocysteine might be harmful, since elevations in homocys-
`teine due to homozygosity or heterozygosity for cystathionine
`synthetase deficiency have been associated with an increase in
`vascular disease and thrombosis (41-43). That these problems
`have not been reported as being present in increased incidence
`in Cbl deficiency or in folate deficiency may be due to the
`shorter duration of Cbl and folate deficiencies or to possible
`difierences in other regulatory factors that may exist in these
`conditions.
`
`Our finding that serum total homocysteine is negatively
`correlated with platelet count and Hot and positively corre-
`latcd with MCV and LDH in the Cbl-deficient patients sup-
`ports the concept that a block in methionine synthetase activ-
`ity occurs in Cbl deficiency and that this block is responsible
`for the megaloblastic anemia and other hematologic abnor-
`malities that are seen in these patients (12), and which are
`indistinguishable from those seen in patients with folate defi-
`ciency. We have previously reported (33) and have found in
`this study that serum methylmalonic acid is not significantly
`correlated with Hot, MCV, or LDH in Cbl-deficient patients.
`This supports the concept that the second Cbl-dependent en-
`zyme, L-methylmalonyl-CoA mutase, does not play a role in
`the megaloblastic anemia and other hematologic abnonnali-
`ties seen in Cbl deficiency.
`The cause of the neuropsychiatric abnormalities that are
`seen in Cbl deficiency but not in folate deficiency has long
`been of interest. Some investigators have favored a defect in
`L-methylmalonyl-CoA mutase as the cause and have sug-
`gested that the neuropsychiatric abnormalities are due to a
`buildup in propionyl-CoA and a resultant increase in odd
`number carbon fatty acids in peripheral nerves and the central
`nervous system (44-46). Other investigators have favored a
`defect in methionine synthetase as the cause and have sug-
`gested that the neuropsychiatric abnormalities are due to a
`lack of methionine and a resultant decrease in various methyl-
`ation reactions in peripheral nerves and the central nervous
`system (47-49). Our studies are not helpful in distinguishing
`between these two posmbilities since we have found that serum
`methylmalonic