630
`
`THE NEW ENGLAND jOURNAL OF MEDICINE
`
`Sept. 22, tsrr
`
`In Stage 1 melanoma of the limbs, delayed dissec-
`tion of lymph nodes (i.e., performed at the time of ap-
`pearance of regional metastases) is as effective in the
`control of the disease as immediate dissection. Since
`the proportion of positive nodes ranges from 20 to 25
`per cent, the “wait-and-see” policy avoids unneces-
`sary postoperative complications in three fourths
`of the patients. for whom the “prophylactic” dissec—
`tion would result in negative histologic findings in re-
`gional nodes. Delayed dissection is advisable as long
`as the patient can be kept under strict clinical con-
`trol.
`
`We are indebted to the surgical and pathological stalls ofthe in-
`stitutes listed for their eta-operation, particularly to Dr. A. Lame.
`London, chairman of the pathologists’ panel for the care taken in
`the diagnostic confirmation of cases and for his cwrdination ofthe
`panel’s work, and to Dr. A. Breslow, Washington. D.C.. and Dr.
`E. P. van der Esch. Amsterdam. Netherlands, for the reclassifica-
`tion at cases according to maximum thickness and levels of inva-
`SlOn.
`
`REFERENCES
`l. Breslow A: Tumor thickness. level of invasion and node dissection in
`stage I cutaneous melanoma. Ann Surg 182572-575, I975
`2. Goldsmith HS. Shah JP. Kin DH: Prognostic significance of lymph node
`dissection in the treatment of malignant melanoma. Cancer 26:606-609.
`i9'l'0
`3. Holmes EC. Clark W. Morton DL. et al: Regional lymph node metasta-
`ses and the level at invasion of primary melanoma. Cancer 37:199-20l.
`I‘J'ltfi
`4. Wanebo HJ. Former JG. Woodrufl’l. et all Selection of the optimum
`surgical
`treatment of state i melanoma by depth oi microinvaston:
`use of the combined microstage technique (Clark-Brestow). Ann Stirs
`182302-35.
`l9'i‘5
`S. Rodenth DC: Basic principles of Surgery: malignant melanoma. Mel-
`anoma and Skin Cancer: Proceedings of the International Cancer Con-
`ference. Sydney.
`|9T2. Edited by W McCarthy. Sydney. Blight. [972. pp
`497-416
`6. Davis NC: Cutaneous melanoma:
`Probl 5ng l3(5)1l-63. 19M
`7'. Morabito A. Marubini E: A computer program suitable for fitting tine-
`ar models when the dependent variable is dichotomous: polichotomous
`or censored survival and non-linear models when the dependent vari-
`able is quantitative. Comput Programs Biomed 5283-295. 1976
`8. Clark WI-l Jr, From L. Bernardino EA. et alt The histogenesis and bio-
`logic behavior of primary human malignant melanomas of the skin.
`Cancer Res 29:705-727. I969
`
`the Queensland experience. Curr
`
`USE OF PLASMA PHARMACOKINETICS TO PREDICT AND PREVENT
`METHOTREXATE TOXICITY
`
`RONALD G. STOLLER, M.D., KENNETH R. HANDE, M.D., SAMUEL A. Jnooas, M.D.,
`STEVEN A. ROSENBERG, M.D., AND BRUCE A. CHABNER, MD.
`
`Abstract To correlate the pharmacokinetics and tox-
`icity ol methotrexate. we measured the drug's clear-
`ance from plasma alter 395 high-dose, six-hour intu-
`sions given to 73 patients. After 375 infusions. 48-hour
`methotroxato levels tell within 2 standard deviations of
`the mean for nontoxic infusions. and myelosuppres-
`sion did not occur. Methotrexate concentrations ex-
`ceeded the range for nontoxic patients (mean 1:2 stan-
`dard deviations) after 20 infusions. Serious myolosup-
`pression occurred after six of these 20 infusions. in-
`cluding live of 12 infusions associated with 43-hour
`
`drug concentrations above 9 x 10-7 M. In seven pa-
`tients with 48-hour concentrations above 9 x 10-"r
`M. the absence of toxicity could be attributed to sub-
`sequent rapid clearanoe of the drug; tour of these
`patients also received large doses of supplemental
`leucovorin (50 to 100 mg per square meter every
`six hours). Determination of methotrexate concentra-
`tion in plasma thus identified patients at high risk of
`toxicity. a group that may benefit from Supplemen-
`tal
`leucovorin rescue.
`(N Engl J Med 297:630634.
`1977)
`
`METHOTREXATE. an effective agent in treat-
`ment of leukemia, choriocarcinoma and other
`tumors,
`is one of the most versatile antineoplastic
`agents because its toxicity in high-dose regimens can
`be prevented by the reduced folate, leucovorin (5-for-
`myltetrahydrofolate).‘ High-dose methotrexate regi-
`mens (50 mg per kilogram or more) with leuCovorin
`rescue have produced responses in metastatic osteo-
`genic sarcoma2 and other tumorsJ and have length-
`ened the reiapse-free interval when used as an adju-
`vant after surgical excision of nonmetastatic osteo-
`genic sat‘coma.‘I
`High-dose methotrexate therapy has not been with-
`Out serious complications. Although the majority of
`
`From the Clinical Pharmacology. Surgery and Medicine branches of the
`National Cancer Institute. National Institutes of Health (address reprint re-
`quests to Dr. Chabner at Rm. (SN-119. Bldg.
`l0. National Institutes of
`Health. Bethesda. MD 20014).
`
`such infusions cause only minor side effects. severe
`and prolonged myelosuppression and mucositis have
`developed in approximately 10 per cent of patients,
`and fatalities directly attributable to drug toxicity
`have occurred in 29 of 498 patients (6 per cent) treat-
`ed with high-dose methotrexate. according to a re-
`cent survey of cancer treatment centers in the United
`States.5 Experimental studies have implicated metho-
`trexate-induced renal dysfunction and delayed drug
`clearance as a probable mechanism of toxicity,” and
`initial attempts to monitor drug levels in patients re-
`ceiving high-dose therapy have demonstrated de-
`layed drug clearance in toxic pattients.°-7 These con-
`siderations
`suggested that
`routine monitoring of
`methotrexate levels in plasma might permit early de-
`tection of patients at high risk of development of tox-
`icity. The following study substantiates the value of
`monitoring plasma methotrexate 48 hours after drug
`administration.
`
`Medac Exhibit 2055
`
`Koios Pharmaceuticals v. Medac
`
`IPR2016-01370
`
`Page 00001
`
`Medac Exhibit 2055
`Koios Pharmaceuticals v. Medac
`IPR2016-01370
`Page 00001
`
`

`

`Vol. 297 No. 12
`
`METHOTREXATE PHARMACOKINETICS — STOLLER ET AL.
`
`I531
`
`MEIER!th AND METHODS
`
`Leucovorln Rescue
`
`From November, 1974, to january, 1977, 78 patients at the Na-
`tional Cancer Institute received 395 infusions of high-dose metho-
`trexate for treatment of metastatic tumors (38 patients) or as adju-
`vant therapy for osteogenic' sarcoma (40 patients). Their median
`age was 25 years, with a range of 10 to 70 years. Treatment proto-
`cols required that all patients receiving high-dose methotrcxate
`have a normal blood urea nitrogen [less than 25 mg per 100 ml) and
`normal serum creatinine {less than 1.2 mg per 100 ml) before each
`course of therapy. In addition, all patients had a white-cell count
`greater than 4000 cells per cubic millimeter and platelet count
`greater than 150,000 per cubic millimeter before treatment, except
`for Patient 3 {Table 1), who had a platelet count of l6,000 cells per
`cubic millimeter before treatment.
`Myelosuppression due to methotrexate was defined as a fall of the
`white-cell count to 2000 cells per cubic millimeter or less, or a fall in
`platelet count to 75.000 cells per cubic millimeter or lower. Drug-re-
`lated renal toxicity was defined as a 50 per cent or greater rise in se-
`rum creatinine. the peak value exceeding the normal limits of 1.2
`mg per l00 ml.
`
`Mothottexate tntuelon
`
`The regimen previously reported by Jaffe et al.‘ was used in this
`study. Methotrexate was supplied by Lederle Laboratories, Pearl
`River, New York, or Ben Venue Laboratories, Bedford. Ohio, and
`was administered in 500 to 1000 m1 of 5 per cent dextrose in water
`containing 25 meq of sodium bicarbonate.
`Methotrexate in doses of 50, 100, 150, 200, or 250 mg per kilo-
`gram was infused intravenously over a six-hour period. All patients
`received intravenous hydration with 3 liters of fluid per square me-
`ter per 2-1 hours and 80 meq of bicarbonate per square meter per
`24 hours. Hydration was begun 12 hours before and was continued
`for 36 hours after the start of the infusion. Urinary pH was tested
`before and periodically during the 24 hours after the start of the
`infusion. Methotrexate was not begun until the urinary pH was
`greater than or equal to ID, and additional intravenous bicarbo-
`nate was recommended if the urinary pH fell below 1.0 during in-
`fusion.
`
`Leucovorin, 15 mg per square meter, was given intravenously two
`hours after the completion of methotreitate infusion and was re-
`peated every six hours thereafter for a total of eight doses. At 48
`hours from the start of methotrexate infusion plasma samples were
`obtained for methotrexate determination. Plasma methotrexate
`concentrations were determined either by the dihydrofolate reduc-
`tase inhibition method using enzyme from either lactation?!” caret
`(New England Enzyme. Boston. Massachusetts) or from metho—
`trexate-resistant rnurine L1210 cells,’ or by the competitive pro-
`tein-binding assay. "‘ Both assays are highly specific for tight-bind-
`ing inhibitors of dihydrofolate reductase and yield equivalent re-
`sults when tested against known standards or the same patient sam-
`ple. Pharmacokinetc data from the first 14 nontoxic infusions were
`used to initially define a normal range {mean :i:2 SD.) for 43-hour
`plasma methotrexate concentration. A normal 48-hour plasma
`methotrexate range was later established for each dose level ad-
`ministered on the basis of data from the first 189 nontoxic infu-
`sions. Patients whose 48-hour level fell within this normal range
`were discharged from the hospital, whereas those with levels more
`than 2 SD. above the mean received leucovorin for an additional 48
`hours. During the initial 18 months ofthis study, leucovorin dosage
`was 12 to 30 mg per square meter every six hours when additional
`rescue was given. In the final six months of the study, higher doses
`of supplemental leucovorin, 50 to 100 mg per square meter every six
`hours, were used.
`
`Rasrmrs
`
`Initial pharmacokinetic studies were performed in
`14 patients who received 50 to 250 mg per kilogram of
`methotrexate withOut subsequent toxicity. Peak drug
`levels were 0.l mM to 1.0 mM during the infusion pe-
`riod and declined thereafter in a biexponential curve
`with half-lives of two and 10.4 hours (Fig. 1). Plasma
`methotrexate concentrations were 3 X 10" M or less
`
`in all patients 48 hours after the infusion was begun.
`We suspected that toxic patients w0uld display pro-
`
`Table 1. Toxic Episodes Due to High-Dose Methotrexate (MTX).
`SEIUM CnEnTININE
`SUPPLEMENTAL Lrumvonm
`Toxlctrv'
`Recov Elv‘l
`
`herons
`m
`
`AFTER
`m
`
`wnmz-cset
`COUNT
`
`PLA'IELET
`coon-r
`
`ornsn
`
`mg! £00 in!
`
`mgltn’ffl fir
`
`0.8
`
`1.2
`
`0.9
`
`0.8
`
`0.7
`
`0.6
`
`1.3
`(96 hr)
`
`2.6
`{24 hr)
`1.5
`[24 hr)
`
`0.8
`(48 hr)
`
`1.0
`(24 hr)
`0.5
`(IIB hr)
`
`6 (day 2-22)
`
`30 (day 2—13]
`
`15 (day 5-1 l]
`
`20 (day 2—1?)
`
`15 (chi)r 2-5)
`
`2.200
`(day 6)
`
`[.700
`(day '1'}
`500
`(day 6)
`
`[.6113
`[day ‘1'}
`
`2.300
`(day 6)
`[.100
`[day 4)
`
`54.000
`(day 10}
`
`23,000
`(day 13)
`500
`[day 8)
`
`60.1110
`(day '1')
`
`[5.000
`(day l l)
`No fail
`
`Urinary-
`tract
`infection
`Severe
`mucositis
`Septicemia
`
`Rash.
`severe
`mucositis
`—
`
`—
`
`On day 12
`
`On day 1?
`
`Patient
`died on
`day i3
`On day 14
`
`011 day 23
`
`On day 5
`
`PATIENT
`No.
`
`Doss
`
`43-Hr.
`PLASMA
`M‘l‘x
`
`mgflcg
`
`Myelosupptession:
`l
`50
`
`2
`
`3
`
`d
`
`5
`
`6
`
`50
`
`100
`
`100
`
`100
`
`150
`
`11M
`
`1.4
`
`2| .0
`
`6.9
`
`1.4
`
`0.25
`(72 hr)
`0.96
`
`Renal toxicity with moderate leukopenia:
`«u
`—
`163.000
`3.200
`T5 (day 2-21)
`2.7
`'l
`50
`0.96
`[.1
`
`
`(day 7](‘8 hr) (day [0)
`
`'White-oeil count <2000 t:ells,irnn'tJ or piatelet count 43.0120 cellsfrnm’.
`YWhitooeil count >201} oellsfmm’, platelet count }15.000 colts/math a disappearance of muonsitis.
`
`Page 00002
`
`Page 00002
`
`

`

`632
`
`THE NEW ENGLAND JOURNAL OF MEDICINE
`
`Sept. 22, 1977
`
`quent toxicity. The toxic patients demonstrated mark-
`edly delayed drug excretion when additional metho-
`trexate levels were determined at 72 and 96 hours
`
`(Fig. 1); in contrast, the nontoxic patients all had rap-
`id clearance of the drug,.as verified by serial determi-
`nations.
`
`Secondly, dosage of leucovorin may have influ-
`enced toxicity in that the severely toxic patients (Pa-
`tients 1-6, Table 1) all received supplemental leuco-
`vorin in low doses (12 to 30 mg per square meter ev-
`ery six hours) whereas four of the nontoxic patients re-
`ceived larger doses of supplemental leucovorin (50 to
`100 mg per square meter every six hours). Further ev-
`idence for the effectiveness of large doses of leucovor-
`in was provided by the course of Patient 7 (Table 1).
`In this patient renal failure associated with markedly
`delayed methotrexate disappearance from plasma de-
`veloped; he received leucovorin, 75 mg per square me-
`ter every six hours for 19 days, and had moderate
`
`l
`__l
`o Toxic Infusions
`- Non-Toxic infusions
`
`
`
`.fi'.
`
`ii.
`iii.
`
`10‘9
`
`100
`
`150
`
`200
`
`250
`
`DOSE, MG/ KG
`
`Figure 2. Relation of Plasma Methotrexate (MTX) Level at 48
`Hours to Toxicity.
`The geometric mean values of nontoxic infusions are shown
`at each dose level, with brackets defining 2 S.D.
`
`Page 00003
`
`
`
`
`
`
`PLASMA[MTX],M
`
`10'6
`
`10'7
`
`10‘8
`
`
` 0 Toxic Infusions
`
`' Non-Toxic Infusions
`
`104
`
`..O.or
`
`
`
`PLASMA[MTx],M 8 m
`
`10" t
`
`_J.—I_ l
`l
`7
`8
`9
`6
`TiME,DAYS
`
`10
`
`11
`
`12
`
`13
`
`Figure 1. Plasma Methotrexate (MTX) Disappearance Curves
`in Patients Receiving Six-Hour Infusions, 50 to 250 Mg per
`Kilogram.
`The disappearance profile for 14 patients monitored serially
`over 72 hours who had no toxicity is defined by the striped
`area. Patients in whom evidence of myelosuppression devel-
`oped (Table 1) had plasma levels as shown by open circles.
`Six nontoxic patients with 48-hour levels above 0.9 uM had
`the plasma levels as shown by the solid circles.
`
`longed retention of methotrexate; therefore, plasma
`samples were routinely monitored 48 hours after the
`start of therapy in an effort to detect patients at high
`risk of toxicity.
`In 375 of 395 infusions, plasma methotrexate con-
`centrations at 48 hours fell within the normal range,
`as defined by the geometric mean :l:2 S.D. for the non-
`toxic patients in this series (Fig. 2). None of these in-
`fusions associated with normal 48-hour levels result—
`
`ed in myelosuppression, mucositis, or other toxicity.
`After seven infusions,
`the 48-hour plasma level ex-
`ceeded the normal range but was less than 9 X 10'7
`M; in each case, the patient received supplemental
`leucovorin, 15 mg per square meter every six hours for
`two days, and no patient manifested toxicity.
`Forty-eight-hour
`levels of methotrexate greater
`than 9 X 10—7 M were associated with a high fre-
`quency of toxicity. In five of 12 patients having 48-
`hour values above 9 X 10”7 M severe myelosuppres-
`sion developed (Table 1). An additional toxic episode
`occurred in a patient with high levels that were first
`detected at 72 hours after infusion (Patient 5, Table
`1).
`than
`levels greater
`Six patients with 48-hour
`9 X 10‘7 M did not show toxicity. Two factors distin-
`guished the toxic patients from patients having simi-
`larly elevated drug levels at 48 hours but no subse-
`
`Page 00003
`
`

`

`Vol. 29? No. 12
`
`METHOTREXATE PHARMACOKINETICS — STOLLER ET AL.
`
`633
`
`granuiocytopenia as the sole manifestation of drug
`tox1c1ty.
`Myelosuppression in the toxic patients continued
`for five to 23 days and, in two (Patients 2 and 4, Ta-
`ble 1), was accompanied by severe oral mucositis.
`Other toxic sequelae included a diffuse erythematous
`rash in Patient 4, a urinary-tract infection in Patient
`1, and gross hematuria followed by a fatal episode of
`septicemia in Patient 3.
`
`Value of Serum Creetlnlne in Predicting Metholrexate Tox-
`iclty
`
`jacobs et al.’ have shown in monkeys that. as a con—
`sequence of its limited solubility at acid pH, metho-
`trexate may precipitate in acid urine, causing renal
`tubular obstruction and delayed drug excretion. Pit-
`man et al.5 have proposed that serum creatinine in—
`crease of 50 per cent might be used to detect patients
`at high risk of development of toxicity after high-dose
`methotrexate.
`
`In the present study all patients had normal serum
`creatinine values before treatment. Post-infusion cre-
`
`atinine values, monitored at 24 or 48 haurs, did not
`change or increased less than 50 per cent in 383 of 395
`infusions. Despite the unchanged creatinine, three of
`these infusions were associated with myelosuppres-
`sion. An increase of more than 50 per cent in serum
`creatinine occurred after 12 infusions; in four cases
`the peak creatinine value exceeded normal limits, and
`each of these infusions was associated with myelosup-
`pression. In the other eight patients, despite the 50 per
`cent increase after treatment, the absolute creatinine
`value did not exceed normal (1.2 mg per 100 ml), and
`none of these patients became toxic. Thus, if one re-
`lied on the criterion of an increase in creatinine of
`
`more than 50 per cent to an abnormal peak value, it
`would have been possible to predict toxicity in only
`four of seven toxic patients.
`
`DISCUSSION
`
`The use of high-dose methotrexate infusions,H al-
`though of great interest in cancer therapy, poses the
`potential of life-threatening toxicity for the patient.
`The present study has demonstrated that monitor-
`ing of plasma methotrexate levels 48 hOurs after infu-
`sion can identify patients at high risk of toxicity. Six
`patients who experienced myelosuppression had lev-
`els at 43 hours higher than 9 X 10" M, and had
`markedly delayed excretion of drug thereafter. A sev-
`enth toxic patient had an elevated level at 72 hours
`and delayed drug disappearance. In contrast, no pa—
`tient having a level less than 9 X 10" M at 48 hours
`experienced subsequent toxicity. Six patients with 48-
`hour levels higher than 9 X 10" M did not show tox-
`icity. The avoidance of toxicity by these six patients
`could be explained by either the subsequent rapid fall
`off in drug levels in all six patients or the high-dose
`leucovorin that four of the six received.
`
`Other studies have suggested that an increase in se-
`rum creatinine may be useful as an early sign of im—
`
`in the present series, a
`pending toxicity.“ However,
`rise in creatinine to abnormal levels was observed in
`
`only four of seven toxic patients. This finding sug-
`gests that if precipitation of methotrexate in renal tu-
`bules underlies toxicity, serum creatinine measure-
`ments may not be sensitive enough to detect this func-
`tional impairment in all cases. Alternatively, other un-
`identified factors, such as individual differences in re-
`nal tubular function or drug metabolism, might ac-
`count for the delayed drug excretion and toxicity.
`The management of patients with elevatedmetho-
`trexate levels and at high risk of toxicity has not been
`resolved. The present study provides evidence that
`large doses of leuc0vorin may prevent tOxicity. In none
`of five patients who received more than 50 mg per
`square meter of leucovorin beginning at 48 hours did
`severe mucositis or myelosuppression develop. In fur-
`ther support of this possibility, one patient (Patient 7,
`Table 1), who had evidence of renal failure and mark-
`edly delayed drug excretion, experienced only mild
`leukopenia when supported with leucovorin, 75 mg
`per square meter every six hours for 21 days.
`In experimental systems, the concentration of leu-
`covorin required to prevent tOxicity increases in di-
`rect proportion to the concentration of methotrex-
`ate, ” but the specific levels required to cOunteract spe-
`cific methotrexate concentrations are not known. Be-
`cause rnethotrexate concentrations above 1 X 10‘“ M
`
`inhibit DNA synthesis in bone marrow and intestinal
`epitheliur‘n,12 it
`is reasonable to continue leucovorin
`rescue in toxic patients until plasma methotrexate
`falls below this level. Measures designed to accelerate
`removal of methotrexate,
`such as hemodialysis,”
`charcoal filtration" or enzymatic cleavage,“ might
`have considerable value in patients with abnormal re-
`nal function, but have not been carefully evaluated in
`man as yet.
`,
`The foregoing results demonstrate the value of
`monitoring plasma methotrexate levels during high-
`dose chemotherapy. The specific correlates of toxic-
`ity, including critical plasma levels and duration of ex-
`posure, will probably differ from regimens employing
`other
`schedules of methotrexate administration.
`
`is clear that determinations of drug
`Nonetheless, it
`levels will be of value for identifying patients at high
`risk after methotrexate therapy.
`'
`
`REFERENCES
`
`2.
`
`4.
`
`1. Djerassi I. Farher S. Abir E. _et a1: Continuous infusion of methotrex-
`ate in children with acute leukemia. Cancer 20:233-242. 1967
`Jal'l'e N. Paed D: Recent advances in the chemotherapy of metastatic
`osteogenic sarcoma. Cancer 30116211631. 19’12
`3. Djersssi I, Rominger C], Kim 15, et a]: Phase 1 study of high doses of
`methotrexate with citrovorurn factor in patients with lung cancer. Carl-
`eer 30:22-30, 1912
`.laer N, Frei E III, Traggis D, et a]: Adjnvant methotrexate and citro-
`vorum-factor treatment ofostcogenic sarcoma. N Engl J Med 291:994-
`99?. WM
`5. Von Hoff DD. Pent: JS. Helman LJ. el al: The incidence of drug relat-
`ed deaths secondary to high dose methotrexate and eitrovorum factor
`administration. Cancer Treat Rep 61:145-748.
`I91?
`6. Stoller RG, iacobs SA, Drake JC. et a]: Pharmacokinctics of high-dose
`methotrexate (NSC-T40). Cancer Chemother Rep 6:19-24. 1915
`'1'. Jacobs SA. Stoller RG. Chabner BA. et a1: T-Hydroxymethotresatc as a
`
`Page 00004
`
`Page 00004
`
`

`

`
`
`THE NEW ENGLAND jOURNAL OF MEDICINE
`
`Sept. 22, 1977
`
`Iurinary metabolite in human subjects and rhesus monkeys reaching
`high dose methotresate. J Clin Invest 51:534-538. 1976
`8. Pitman SW. Parker LM. Tattersall MH. et al: Clinical trial of high-dose
`methotrexate {NSC-74D} with citrovorum factor (NSC-3590): toxico-
`logic and therapeutic observations. Cancer Chemother Rep 6:43-49,
`I975
`9. Berrino JR. Fischer CEA: Techniques for study of resistance to folic ac-
`id antagonists. Methods Med Res l0:297-307. 1964
`it]. Myers CE, Lippman M. Eliot HM, et at: Competitive protein binding
`assay for methotreitate. Proc Natl Acad Sci USA 'i'l:3683-3686. [91‘s
`ll. Pinedo l-IM. Zaherko 05. Bull JM. el al: The reversal of methotrexate
`
`cytotoxicity to mouse bone marrow cells by leucovorin and nucleo-
`sides. Cancer Res 36:4413-4424,
`[91"6
`I2. Chabner BA. Young RC: Threshold methotreitate concentration for in
`win inhibition of DNA synthesis in normal and tumorous target tis-
`sues. J Clin Invest 521804481], 19?]
`IS. Djerassi
`l. Ciesieika W. Kim .iS: Removal of melhotrexate by filtra-
`tion-absorption using charcoal filters or by hemodialysis. Cancer Treat
`Rep 61:751-752. I???
`14. Chabner BA. Johns DCI, Bertino JR: Enzymatic cleavage of metho-
`trexate provides a method for prevention of drug toxicity. Nature
`239:395-397. I972
`
`AUTOIM M UNE VITILIGO
`
`Detection of Antibodies to Melanin-Producing Cells
`
`KENNETH C. HERTZ, M.D., LAURA A. GAZZE, A.B., CHARLES H. KIRKPATRICK, M.D.,
`AND STEPHEN I. KA'rz, M.D., PHD.
`
`Abstract Vitiiigo, a disorder characterized by the de-
`struction of melanocytes. is often associated with dis-
`eases in which there are increased frequencies of au-
`toantibodies.‘ For this reason we investigated two pa-
`tients with vitiligo. alopecia universalis. mucocutane—
`ous candidiasis. and multiple endocrine insufficien-
`cies for antibodies to melanin-producing cells. Using
`direct
`immunoiiuorescence of normal and vitiligl-
`nous skin from both patients and indirect immunoilu-
`orescence with both patients‘ serum. we could not de-
`tect these antibodies. However. an immunofluores-
`
`cent complement-fixation test demonstrated a circu-
`lating antibody that bound to melanocytes In human
`skin. nevus cells and melanoma cells. Specificity of
`cellular fluorescence tor nevus and melanoma cells
`was shown on serial sections stained with hematoxy—
`Iin and eosin and was inferred for melanocytes from
`their distribution in human skin and their presence
`when the normal but not vitiliginous skin of both pa-
`tients was used as substrate. This antibody was char—
`acterized as an lgG that activated complement via the
`classical pathway. (N Eng! J Med 297:634—637, 1977)
`
`VITILIGO, a disease characterized by the loss of
`melanin pigment that follows the destruction of
`meianocytes, has been associated with hyperthyroid-
`ism, hypothyroidism, hypoparathyroidism, perni-
`cious anemia, diabetes mellitus, mucocutaneous can-
`didiasis and aiopecia areata. In addition to numerous
`case reports relating its occurrence with one or sever-
`al of these disorders,” larger studies have confirmed
`these assOciations.“' While the clinical associa-
`
`tions of vitiligo have led some to consider it a cutane-
`ous marker for internal disease,12 the presence of thy—
`roid, adrenal and gastric parietal-cell antibodies in
`many patients with both vitiligo and endocrine dis-
`ease has suggested a common, perhaps autoimmune,
`origin.'-’i‘-” The autoimmune theory of vitiligo is fur—
`ther strengthened by the increased prevalence of or-
`gan-specific autoantibodies in several,“”‘*5 but not
`all,“ large series of patients with vitiligo.
`To date, an antimelanocyte antibody in the serum
`of patients with vitiligo has not been found." In the
`two patients with vitiligo, alopecia universalis, muco-
`cutaneous candidiasis and multiple endocrine insuffi-
`ciencies described below, we demonstrated a circu-
`lating antibody to melanocytes, nevus cells and mela-
`noma cells. The antibody is identified by its comple-
`ment-fixing ability and has been characterized as an
`IgG.
`
`From the Dermatology Branch. National Cancer Institute, and the Labo-
`ratory of Clinical Investigation. National Institute ofAllergy and Infectious
`Diseases. National Institutes of Health. Bethesda. MD (address reprint re-
`quests to Dr. Hertz at the Department of Dermatology. University of Mia-
`mi School of Medicine. P.O. Box. 5208?5. Biscayne Annex. Miami. FL
`33152}.
`
`CLINICAL SUMMARIES
`
`CASE 1. A 34~year-old woman has had widespread candidiasis
`and total alopecia since the age of six. Hypoparathyroidism had
`been diagnosed at nine years of age, and the evaluation of an un-
`steady gait had led to the recognition of pernicious anemia at the
`age of 13. Her first menstrual period had occurred at the age of 15,
`but subsequent periods were infrequent, leading to the diagnosis of
`primary ovarian failure. Vitiligo began at the age onB and has been
`progressive; the most striking areas of involvement have been the
`face and extremities. Hypoadrenalism was diagnosed during her
`initial admission to the National Institutes of Health at the age of
`34.
`
`CASE 2. An 13-year-old woman has had candidiasis of her nails.
`total alopecia and vitiligo since three years of age. hypoparathy-
`roidism since six and hypoadrenalism since 10. Primary ovarian
`failure was diagnosed at the age of 16; approximately 50 per cent of
`her skin was vitiliginous at that time.
`
`Both patients have had extensive immunologic evaluations dis-
`closing impaired cellular immunity to Candida clbr'cnrrs in vivo and in
`vitro. Serum autoantibody studies showed Case 1 to have thyroid
`complement-fixing and adrenal antibodies but no thyroid-aggluti-
`nating or gastric parietal-cell antibodies. Case 2 had adrenal, gas-
`tric parietal-cell and ovarian antibodies.
`
`Mes-none
`
`Immunofluorcscent staining was performed
`Immunafluoreuence.
`with Previously described procedures and reagents." We examined
`biopsies of normal. vitiliginous, and border skin from both patients
`by direct immunofluorescence for in vivo bound IgG, IgA, IgM. lgE
`or C3. Serum samples from both patients were studied by conven-
`tional indirect immunofluorescence for the presence of circulating
`antibodies {lgCL lgA, IgM or IgE) and by an immunofluorescent
`complement—fixation te5t for circulating complement-binding fac-
`tors capable of reacting with melanin-producing cells in normal or
`neoplastic tissue. The complement-fixation test is a multistep pro-
`cedure involving sequential incubation of frozen sections of sub-
`strate with test serum, fresh human serum or guinea-pig serum as a
`
`
`
`
`Page 00005
`
`Page 00005
`
`