INTERNATIONAL JOURNAL OF LEPROSY^ Volume 57, Number 1

Printed in the U.S.A.

CORRESPONDENCE

This departinent is for the publication of informal communications that are ofinterest because they are informative and stimulating, and for the discussion qfcontroversial matters. The mandate of this JOURNAL is to disseminate informationrelating to leprosy ia particular and also 01/ler mycobacterial diseases. Dissidentcomment or intoproation 00 published research is of course valid, bui personalityattack.s- on individuais would seem unnecessary. Political C0111111elliS, valid or not,also are unwelcome. They might residi ia interference with the distribution of theJouRNAL and thus interfere with lis prime putpose.

A Critique on the interpretation of the Lepromin ReactionUsing Heat-killed M. leprae Vaccine

TO THE EDITOR:

Skin reactivity clicited by various my-cobacterial antigens serves as a surrogatemeasure of protective immune responses inleprosy almost impossible to measure withany conventional immunological test avail-able today. h is principally on the basis ofsuch inputs that large-scale clinicai trials ofsevera! leprosy vaccinc candidates involv-ing substantial expenditure and infrastruc-ture are initiated. The paper therefore byGil, et al. in the March 1988 issue of theJOURNAL entitled "Vaccination of HumanVolunteers with Heat-killed Aí. leprae: Lo-cal Responses in Relation to the Interpre-tation of the Lepromin Reaction" is a be-lated though welcome one in its search foran interpretation of the lepromin reaction.

The purpose ofthis communication is notto argue with the basic premises of vacci-nation and skin testing in mycobacterioses,valid though such arguments are, but to in-dicate flaws in the very conception and de-sign of the above-mentioned study and toprovide alternative explanations for theconclusions that have been drawn.

In our opinion, a significant error in thestudy appears to be the consideration of thearmadillo -derived Mycobacterium leprae(AML) vaccine as a representative of stan-dard Mitsuda lepromin. To quote an ex-ample, humoral antibody responses gener-ated by Mitsuda lepromin are significantlymore rapid than those elicited by AML vac-eines (2.3). The dose ranges of the two prep-arations are obviously different, the lowest

dose of the AML vaccinc being 10-foldhigher than the standard dose of the Mit-suda lepromin. Evcn so, in engendering alate reaction at that dose, its efficacy is ap-parent in only 50% of normals at a cut-offpoint of 5 mm. Satisfactory skin-test con-version responses to M. leprae soluble an-tigens (MLSA) in normal individuais arealso achieved only at the two highest dosesof 1.5 x 108 and 5 x 108, which are 100-to 500-fold higher than the Mitsuda prep-arations. Conversely, the Mitsuda reactionis negative in most lepromatous patients,while the AML vaccine exhibits at the siteof injcction an optimal reaction/ulcerationin the majority of such cases. A plausibleexplanation for the above discussion is dose-induced sensitization which, in fact, impliesthat the only factor a late lepromin reactionmay measure is an individual's threshold ofsensitization.

Nevertheless, an interesting fall-out of thepresent study which the authors fail to em-phasize is why normals from nonendemicarcas, a majority ofwhom reportedly shouldbc responders to Mitsuda lepromin, fail torespond satisfactorily to AML vaccine ex-cept at substantially higher doses. This alsopoints a finger to a major lapse in the studydesign, i.e., the comparison of a pre- andpostvaccination Mitsuda lepromin re-sponse with the late reaction elicited by theAML vaccine. Considering the impossibil-ity of interpreting responses in individualssubjected to four doses ofAML vaccine, thisstep would have provided information on

110

57, 1^ Correspondence^ 111

the ability of a single vaccine dose to trans-form Mitsuda nonresponders to responders.

It is probable that this omission is due tothe sensitizing capacity attributed to inte-gral lepromin. Two alternatives may be ex-plored to overcome this: a) Dharmendra an-tigen that is not a potent sensitizer and yetmay be capable of eliciting both an earlyand a late response, and b) liposome-boundsoluble antigen of Sengupta, et al. ( 7) whichmay mimic integral bacilli to a large extent.

As is also reported here, all vaccine can-didates appear to work more efficiently inthe higher dose ranges. On this basis, theauthors suggest the use of a higher dose ofMitsuda challenge for testing vaccine po-tency. It is indeed a paradox if this werc tobe so sinee, in immunological systems, suc-cessful sensitization results in a responseevoked to low doses of the challenge anti-gen. If the aim of induced sensitization isprotection via memory, then a susccptibleindividual must be endowed with the abilityto respond to lower doses of antigen, a sit-uation that is analogous to in sito occur-rences where the exposure to bacilli at theonset may be low and insidious rather thanoverwhelming. Therefore, a postvaccina-tion, negative lepromin skin test should beaccepted as a failure of vaccination ratherthan attempt coercion ofa positive responsethrough dose manipulation that would havelittle relevance to the infectious processesin the leprosy patient.

Despite theoretical reservations, a goodtest for any vaccine given to normal indi-viduais would be to measure both the earlyand late reaction to a "Mitsuda like agent"at periodic time intervals after vaccination.In the event that a vaccine candidate is ef-ficacious, we would then expcct the earlylepromin reaction to wane with time, whileleaving Mitsuda reactivity fairly intact. Thishypothesis is upheld by the present datawhich show that prevaccination MLSA re-sponses do not correlate with the early re-sponse of the AML vaccine except at thehighcst vaccine dose. Moreover, this obser-vation has been rightly interpreted by theauthors to signify that integral preparations,such as AML vaccines, may contain vari-able or minimal amounts ofsoluble antigen.In an unrelated context, they also quote theexample where in BCG-vaccinated popu-

lations, MLSA activity was positive im-mediately after vaccination but was foundto be negative 90 days later (h). In fact, thisquotation upholds their own conclusion ofthe capricious amount ofsoluble antigen andis not, as they imply, suggestive of the "non-effect" of an antigenic crossreactivity be-tween AI. lcprae and BCG. It also vindicatesthe original interpretation of the Fernandezreaction as a measure of a pre-cxistent cell-mediated immune response to AI. lcprae re-sulting from active or crossreactive infec-tions ( 5 ).

To most observers, the choice of usingMLSA except as a measure of convenientein the field has long been intriguing. Thestandardization of batches has been sus-pected before to be less than optimal (K),whereas no published data exists on thematter of MLSA sensitivity and leprosy in-cidente (') except perhaps in the pages ofthe World Health Organization-TropicalDiseases Research reports. With the com-puting of data from the present study, itscredibility decreases further.

One is also reminded of earlier studieswhich demonstrated that the sharing of an-tigens between integral BCG and batches ofPPD is extremely limited ( 4). We are not sofar aware of any studies that seck to eluci-date antigenic relationships between Al. lcp-rae, its heat-killed/irradiated integral coun-terpart, and the soluble antigens preparedtherefrom.

It is imperative to implement studies suchas these in order to have a firm scientificbasis before undertaking expensive large-scale vaccine trials.

—Ncrges F. Mistry, Ph.D.Senior Rcsearch OJJicer

—Noshir H. Antia, F.R.C.S., F.A.C.S.(Hon.)Director and TrnsteeThc Founclation .for Medical Rcsearch84.1 R. G. Thadani MargII'orG, Bombav 400018, India

Rcprint requests to Dr. Antia.

REFERENCES1. FINE, P. E. M., PONNIGHAUS, J. M. and MAINE, N.

P. Thc relationship between delayed type hyperscn-sitivity and protectivc immunity induced by my-cobacterial vaccines in man. Lepr. Rev. 57 Suppl.2 (1986) 275-283.

112^ International Journal ofLeprosv^ 1989

2. GILL, H. S., MUSTAFA, A. S., IvANY1, J., HARBOE,

M. and CODA i.. T. Ilumoral iia mune responses to/epray in human volunteers vaccinated with

^

armadillo-derived^hinve. Lepr. Rev. 57Suppl. 2 (1986) 293-300.

3. GONZALEZ-ABREU, E., GONZALEZ-SEGREDO, A. andDE LA CRUZ, F. Anti-.1/. /epnre amibodies induccdby lepromio injection as demonstrated by indirectimmunofluorescence. Lepr. Rev. 55 (1984)337-340.

4. HARBOE, M. Antigens of^tuberculin andautoclaved 31. bovis 11CG st udied by crossed im-munoelectrophoresis. Aia. Rev. Respir. Dis. 124(1981) 80-87.

5. LITFÜRD, M. Lepromio as ao indicator ofprotecoveimmunity. Lepr. Rev. 52 (1981) 221-228.

6. PoNNIGHAus, J. and FINE, P. E. M. The KarongaPrevention Trial—which BCG? Lepr. Rev. 57 Suppl.2 (1986) 285-292.

7. SENGUPTA, U., SINIIA, S.. RAM, G., RAviNDER, K.and Gui'T/k, C. Sol uble antigens of .1/. hprae coo-pled with liposomes cli ci Is both "early" and "late-delayed hypersensitivity skin reactions. Int.^Lepr.56 (1988) 45-49.

8. WoRLD IlEALTII ORGANIZATION. Vaccination trialsagainst leprosy: a meeting of the Epidemiology sub-group of the Scientific Working Group ou the 1m-inunology of Leprosy, February 1985, p. 6 W110/TDR/1M MLEP/EPD 85.3.

Reply by Gil, Mustafa and Godal

To THE EDITOR:

The modern interpretation of the lepro-min test is that it is a m mi mal vaccination(1.2). This is the premise upon which ouroriginal study, i.e., the trial of a heat-killed,armaclillo-derived vaccine, was based. Inthis stucly four groups of normal, healthyindividuais were given graded doses (1.5 x107, 5 x 107, 1.5 x 108, and 5 x 108) ofheat-killed, armadillo-derived Myeobacte-rium leprae intradermally. The perfor-mance of the vaccine was assessed by mea-suring the skin-test responses of thevolunteers to M. leprae soluble antigen(MLSA) before and after vaccination. Thisstudy therefore allowed us to measure a) theearly and late responses to the killed-ar-madillo M. leprae vaccine (AML), '1)) theskin-test responses to MLSA and to PPD,and c) the lymphocyte transformation testresponses to MLSA, whole leprae, andthe various other antigens. Thus, the pur-pose of writing the article was to examineali of these parameters closely and to de-termine their relationships. We felt com-pelled to make this report since a) it sup-ports the modern interpretation of thelepromin test, and b) it would benefit thosecenters in the world which use armadillo-derived M. leprae for lepromin testing. It ismost important to grasp the fact that thevaccine in our study also functions as a lep-romin. Unfortunately, Mistry and Antiafailed to grasp this concept, and therein liesthe reason for most oftheir criticisms ofourreport.

Mistry and Antia point out that "... asignificant error in the study appears to bethe consideration of the armadillo-derivedMycobacterium leprae (AM L) vaccine as arepresentative of standard Mitsuda lepro-min." We are cognizant of the differencesbetween armaclillo and human lepromin,and we have actually alluded to this fact inour paper. These differences aside, the AMLvaccine elicits an etirly and a late responsenot unlike that elicited by "Mitsuda lep-romin." We would say that \vinte the AMLvaccine is not equivalent to Mitsuda lep-romin, it is sufficiently similar to be con-sidered in the same category.

Mistry and Antia then remark that wefailed to emphasize the reason for the factthat the majority of normals, who shouldbe responders to lepromin, failed to respondsatisfactorily to the AML vaccine. Actually,our study has shown, using Mistry and An-tia's criteria of a cut-offpoint of 5 mm, thatthe majority of the subjects gave positivelate responses to the vaccine. When the re-sponse to the vaccine was assessed by thedifference in the skin-test response to MLSA,the conversion was statistically significantin the groups that received the three highestdoses of vaccine, i.e., 5 x 107, 1.5 x 108,and 5 x 108 bacilli.

Mistry and Antia further point out that amajor lapse in our study is the absence of acom parison of the pre- and postvaccinationlepromin responses with the Iate reactionelicited by the AML vaccine. The purposeof the original study was to assess the per-