INTERNATIONAL JOURNAL OF LEPROSY^

volume 55, Number 4Printed in the

Antibodies to Phenolic Glycolipid-IDuring Long-term Therapy: Serial

Measurements in Individual Patients'Richard A. Miller, Donna Gorder, and James P. Harnisch-'

The characteristics of the Immoral im-mune response to phenolic glycolipid-I(PGL-1), a species-specific antigen of My-cobacterium frprae, have been the subjectof numerous studies since the descriptionof this antigen by Hunter and Brennan in1981 ('). The antibody response to PGL-Iconsists predominantly of IgM, although anIgG response is detectable in some patients(4 . "). The amount of circulating ant i-PGL-I1gM as well as the percentage of seropositivepatients have been found to correlate withboth the disease classification and the ba-cillary index (III) (4 ') and, in the nine-band-ed armadillo, the anti-PGL-I level was di-rectly related to the hepatic bacterial yieldof M. leprae

One of the proposed applications of anti-PGL-I serologic testing is to monitor theresponse to therapy in individual patientswith the aim of preventing overt clinicalrelapses by early detection of treatment fail-ure. The only currently available modalityfor monitoring therapeutic response, otherthan repeated clinical observation, is to per-form serial skin biopsies or skin scrapingsto determine the bacterial and morpholog-ical indices. These procedures are invasive,somewhat painful, require significant ex-pertise for interpretation, and are suscep-tible to sampling error.

When cohorts of patients tested for vary-ing lengths of time were studied, the meanlevel of anti-PGL-I 1gM was found to de-cline slowly over several years ( 42 ), but

' Received for publication on I June 1987: acceptedfor publication in revised form on 11 August 1987.

' R. A. Miller, M.D.; D. Gorder. 11.8.; J. P. Ilarnisch,M.D., Divisions of Infectious Diseases and Derma-tology, Department of Medicine, University of Wash-ington, Seattle, Washington 98105, U.S.A.

Reprint requests to: Dr. Richard Miller, Immunol-ogy Research Laboratory. Pacific Medical Center, 1200Twelfth Avenue South, Seattle, Washington 98144.U.S.A.

short-term studies in individual patientsfound a variable response of anti-PGL-I IgMduring therapy C). In addition, the widerange of levels in individual patients, evenamong those with the same disease class,makes it impossible to extrapolate from thebehavior of large populations to that of spe-cific patients. Data on the behavior of anti-PG [-I levels during several years of therapyin individual patients are therefore neededto accurately assess the potential for sero-logic monitoring. If individual patients'levels fluctuate erratically or fall exceedinglyslowly, this test is unlikely to be clinicallyuseful. To investigate this issue, we studiedall of the patients followed at the SeattleRegional Hansen's Disease Clinic on whomserial sera specimens spanning at least theinitial 5 years of treatment were available.

MATERIALS AND METHODSPatients. Criteria for inclusion in this

study were: a) availability of stored sera ob-tained over a period of approximately 5years or more, b) first sera obtained priorto, or within 2 weeks of, the initiation ofantileprosy treatment, and c) no prior treat-ment for leprosy. Approximately 85 uniqueHansen's disease patients were seen at theSeattle clinic over the last 8 years, of whom11 met eligibility criteria for this study. Themost common reasons for exclusion wereinitiation of treatment prior to entering theclinic and diagnosis of Hansen's disease lat-er than March 1982. All sera from each pa-tient obtained at least 10 weeks after the lastprevious specimen were tested.

Patients were classified by standard clin-ical and histologic criteria as having polartuberculoid (TT), borderline tuberculoid(BT), borderline (BB), borderline leproma-tous (BL), or polar lepromatous (LL) dis-ease, and all received both dapsone andrifampin as initial therapy. Erythema no-

633

634^ hilernalional Journal of Leprosy^ 1987

dosum leprosum (ENL) was treated withprednisone and clorazimine in most cases,and with thalidomide in most male patients.Reversal reactions were managed withpredn i son C.

PGL-I ELISA. Sera were obtained, all-quotted into small vials, and stored at —70°Cuntil the day of the test. All sera on an in-dividual patient were tested on the same dayand on the same ELISA plate whenever pos-sible. All sera were tested in duplicate at adilution of 1:200 in both PG L-1-coated andblank wells as previously described ('•').This dilution provided the optimal signal :background ratio for the particular batch ofPGl.-I-coated plates used for this study. Thepositive control standard consisted of serialdilutions of pooled human leprosy sera, andthe negative control consisted of pooled serafrom healthy Seattle residents. The incu-bation time with the dye/substrate reagentwas adjusted such that the lowest dilutionof the positive control sera had A 4 „, valuesof 1.2-1.4 (mean 1.3).

Data analysis. ELISA results were re-corded as net A 4„, (the dinrence betweenthe mean absorbance in the PGL-l-coatedwells and the mean absorbance of the un-coated wells). To facilitate comparisons be-tween patients, these absorbance values werethen recalculated as percentages of . each re-spective patient's initial absorbance values.Thus, the first specimen from each patientwas arbitrarily given a value of 100%, andeach subsequent serum specimen was as-signed a percentage value proportional toits absorbance value.

RESULTS

Three paucibacillary patients (1 TT, 2 BT)had initial absorbance values in the low nor-mal range = 0.00-0.07). Net absor-bance values remained normal during theentire, uncomplicated treatment courses inall three of these patients.

Four patients with multibacillary disease(3 B13. I BL) had rapid steady declines inantibody level (Fig. 1). Two of these patientsexperienced reversal reactions during ther-apy and one had recurrent erythema no-dosum leprosum (ENL). The mean initialnet absorbance for this group was 0.68 ±0.34. The mean levels in this group fell 53%in the first year, 66% by the end of the third

year, and 86% by the end of the fifth yearof treatment. At the completion of 5 yearsof therapy, two of these patients had netabsorbance values within the normal range.

The last lour patients (I LL, 3 BL) hadantibody levels which fell much more slow-ly, losing only 20% of their initial value inthe first year (Fig. 2). The rate of decline inantibody level accelerated in subsequentyears in three of these patients. At the endof 3 years of therapy, the mean levels hadfallen by 56%, and by 67% at the end of thefifth year. All fOur of these patients expe-rienced ENL reactions, but these reactionswere unrelated to the transition from a slowto a rapid decline in antibody levels. Themean initial net absorbance for this groupwas 1.04 ± 0.13, and none had normal levelsof antibody at the completion of 5 years oftherapy.

Only one patient had a significant, sus-tained increase in antibody level followingan initial fall. This patient had 1313 leprosycomplicated by a reversal reaction. 1-le wasalso noncompliant or poorly compliant withhis basic regimen of dapsone and rilampinduring the period shown in Figure 1. He didnot experience a clinical relapse, and nobiopsies or skin scrapings were taken duringthis period. No comparable periods of pro-longed medical noncompliance were notedin the remaining 10 patients.

DISCUSSION

Antibodies of the IgM class are the firstto appear during the immune response to aforeign antigen and rapidly decrease whenthe stimulating antigen is removed. Thepersistence of high levels of anti-PGL-I1gMin leprosy patients during the initial monthsand even years of therapy is most likely theresult of ongoing antigenic stimulation. Thishypothesis is supported by the observedcorrelation between the anti-PGL-I level andthe bacterial index ( 4 ). Based on these find-ings, it has been proposed that anti-PGL-11gM levels may serve as a useful monitorof therapeutic response.

The only published study which exam-ined patients who had been extensivelytreated found that the mean absorbance inpatients treated for 10 or more years was14% of the mean absorbance in a difirentgroup of untreated patients, and that most

55,4^Aiii/er. et al.: ,S'erial.-liaihody Measurements^635

WEEKS

FIG. I.^1Z^1^1 I 1\1 antibody^ 1 (:1Re lative levels o2 .g... inti. -iocy toin lour multibacillary patients with rapid initial de-clines in antibody level. l'he bar marks the period ofpoor medical compliance in the patient \‘, hose antibodylevel rose sharply at week 171.

of this decline occurred in the initial 1-2years of treatment ('). These results wereencouraging but limited by the tact that sinceserial specimens from the same patients werenot used, there may have been undetecteddi fferences between the various patientgroups which affected the subsequent testresults.

Antibody levels in our patients declinedeven more rapidly, falling to between 10%and 30% of the initial value after 5 years inmost patients with initially detectable levelsof anti-PGL-I IgM. Nlore importantly, withone exception, each patient exhibited a pat-tern of diminishing amounts of anti-PGL-IIgM over time. We do not have frequentbacterial index determinations on these pa-tients, so it is impossible for us to confirmthe correlation between the IgM level andthe bacterial index reported by others, butin the seven patients with consistently fall-ing levels (and in the three patients withconsistently negative levels), there was noevidence of clinical relapse or prolongedmedical noncompliance.

Our sample population is biased towardpatients with reactional states. primarily be-cause these complicated patients were fol-lowed most closely by the clinic and there-fore met the entry criteria for inclusion inthis study. We could not detect any effectof ENL or reversal reactions on the patternor rate of IgM decline in these patients. al-though there were few patients without re-actional states included fbr comparison, andthis study was not designed to screen forany such effect. However. in prior work. our

WEEKS

FIG. 2. Relative levels of . IgN1 antibody to I'GL-1in lour in ultibacillary patients with slow initial declinesin antibody level.

laboratory and others have failed to find anycorrelation between ENL and anti-PGL-IIgM levels ('). Other investigators usingslightly different assay systems have foundthat acute ENL is associated with a decreasein anti-PG[1 IgM levels ( 4 ), an effect thatmay be related to therapy with thalidomiderather than a primary effect of ENL C). Threeof our live ENL patients received thalido-mide, and we cannot exclude an effect ofthalidomide on the antibody levels in thesepatients.

In conclusion, we found that anti-PGL-IIgM antibody levels in multibacillary pa-tients declined relatively rapidly during theinitial 5-7 years of therapy. apparently ir-respective of intercurrent reactional states.If these preliminary results are confirmedby larger studies which include frequentmeasurements of bacillar.' load, it may provefeasible to use serologic testing as a primarymodality Ibr assessing therapeutic response.

SUMMARY'Levels of 1gM antibody to phenolic glv-

colipid-I (PGL-I) were measured in serumspecimens collected over the initial 5 ormore years of therapy from 11 leprosy pa-tients. All three patients with paucibacillarydisease had undetectable levels of antibodythroughout their treatment. The eight pa-tients with multibacillary disease had ini-tially elevated levels which fell quite rapidlywith treatment, reaching levels of 10% to30% of their initial pre-treatment level after5 years of therapy. The single patient withprolonged therapeutic noncompliance hadan increase in antibody level, although clin-ical or bacteriologic relapse was not docu-

636^ Imo-national Journal of Leprosy^ 1987

mented. These results in individual patientsdemonstrate that IgM antibody to PGL-Ideclines rapidly and consistently with treat-ment in multibacillary patients.

RESUMENSe 1111(1R:run los niveles de anticucrpo IgM contra el

glicolipido fenOlico-I (GLF-1) en muestras de suero

colectadas de I I pacientes con lepra, durante los pri-meros 5 anos in:is de tratamiento. Los 3 pacientescon lepra paucibacilar no tuvieron niveles detectables

de anticucrpo a lo largo del tiempo del tratamiento.Los 8 pacientes con lepra multi bacilar tuvicron titulosmicialmente elevados clue dismmuyeron rapidamente

con el tratamiento :dean/and°, despu&s de 5 anos deniveles del 10% at 30% en relaciOn con los

niveles de pre-tratamiento. El imico paciente con un

prolongado tratamiento irregular, tuvo un increment°

en su nivel de anticuerpos per°, desalbrtunaciamente,

no se obtuvo la informaciOn sobre su recaida clinica obacteriolOgica. Los resultados en los pacientes indivi-

duales dcmucstran clue el anticucrpo IgM contra elGLF-I disminuye rapidamente y de manera consisten-

te con el tratamiento en los pacientes multibacilares.

RESUMEChez 11 malades de la lepre, on a mesure les taus

d'anticorps IgM a Eantigene ph6noglycolipidique-1

(PGL-I), dans des &chantillons de scrum recueillis auemirs des cinq premieres annees de traitement ou plus

lard. Les trois malades atteints d'une alUction pauci-bacillaire presentaient des taus non d&celables d'anti-corps tout au long du traitement. Les 8 'naiades qui

soulfraient d'une lepre multibacillaire ont au debut livres

des taus &lev&s clanticorps. qui sons tombes asset ra-

piciement au cours du traitement, pour atteindre apres5 ann&es de traitement des taux qui se situaient entre

10% a 30% des taus initiaus releves avant le traite-

ment. Un seul malade, qui s'Ctait signal& par un manque

d'assiduite prolong& au traitement, temoignait dune

augmentation du taus d'anticorps, encore qu'aucunerecidive clinique ou bacteriologique n'ait pu etre mise

en evidence. Les resultats ohtenus chez des malades

individuels demontrent que les anticorps 1gM au PGL-Ideclinent rapidement et de maniere cohérente au cours

du traitement chez des malades multibacillaires.

Acknowledgment. This work was supported in partby Contract No. 258-83-0063 from the Gillis W. Long

I lansen's Disease Center, Cars:111c, Louisiana,

REFERENCESD., FLAGlan., 13.,

110FrEsoAcm, A. and ComENor, 1•. Antibodies to

phenolic glycolipid-I and to whole Mycobacterium

/cprae in leprosy patients: evolution during therapy.

Int. J. kepi . . 54 (1986) 256-267.

2. 13RE - ri S I , DRAPER, P., PAyNE, S. N. and REEs.

12..1. W. Serological activity of a characteristic phe-nolic glycolipid from ycolucterium hprae in serairom patients with leprosy and tuberculosis. Olin.

Esp. Inummol. 52 (1983) 271-279.3. I w. 011(1 B RENNAN, I) . J . A nove l p h eno li c

glycolipid from .Ilycoluderium frprae possibly in-

volved in immunogenicity and pathogenicity. J.13acteriol. 147 (1981) 728-735.

4. LEVIS, W. R., MEEKER, H. C., SCHULLER-LEVIS, G.,

SERSEN, E. and SCHWERER, B. IgM and IgG anti-

bodies to phenolic glycolipid-I from .1/Jx:(d'acte-num laprac in leprosy: insight into patient moni-

toring, erythema noclosum leprosum, and bacillary

persistence. J. Invest. Dermatol. 86(1986)529-534.). MEEKER, H. C., LEVIS, W. R., SERSEN, E.,

SCHULLER-LEVIS, G.. BRENNAN. P. J. and BUCHANAN,

T. M. ELISA detection or IgNI antibodies against

phenolic glycolipid-1 in the management of leprosy:

a comparison between laboratories. Int. J. Lepr. 54(1986) 530-539.

6. TRUMAN, R. W., MORALES, M. J., SHANNON, E. J.

and 11ASTINGS, R. C. Evaluation of monitoring an-

tibodies to PGL-I in armadillos experimentally in-

fected with M. /eprac. Int. .1. Lepr. 54 (1986) 556-

559.7. TRUMAN, R. W., SHANNON, E. J. and HASTINGS, R.

C. HOS1 response to the phenolic glycolipid-1 anti-

gen of M. leprac. Int. J. Lepr. 53 (1985) 710-711.

8. YouNG, D. B. and BUCHANAN, T. M. A serological

test for leprosy with a glycolipid specific for My-

cobacterium /eprae. Science 221 (1983) 1057-1059.9. YOUNG, D. B., DISSANAYAKE, S., MILLER, R. A.,

KHANOLKAR, S. R. and BUCHANAN, T. M. Humans

respond predominantly with IgM immunoglobulin

to the species-specific glycolipid of Mycobacterium

lepnw. J. Infect. Dis. 149 (1984) 870-873.