[財団ブログ ― ハンセン病]
「新診断法」について―研究者コメント(英文)

ハンセン病の新しい簡潔な診断法がアメリカの研究機関により開発され、ブラジルで企業化されるという情報が頻繁にネット社会で話題に上がる一方研究者の間からは慎重な意見が出されています。下記にカリフォルニア大学(SF)のロバート・ゲルバー氏によるハンセン病関係者のメーリングリストへの最新の投稿を紹介します。
Pieter AM Schreuder 6月11日
To editorlml
Leprosy Mailing List – June 10, 2013
Ref.: (LML) New Diagnostic Test
From: Robert Gelber, University of California San Francisco, USA
Dear Pieter,
I fully agree with the reasoned and cautionary comments of Professors van Brakel and Lockwood of 4 April concerning the new diagnostic test for leprosy (Eur J Clin Microbial Infect Dis.30,2011,1257) and wish to share several comments of my own on the subject:
1. First of all skin tests and blood tests for tuberculosis and leprosy are more properly considered tests of infection and not actual disease.
For many decades the PPD skin test, a measure of cellular immune recognition of M tuberculosis, has been widely utilized, its specificity compromised by prior BCG vaccination and the presence of non-tuberculous environmental mycobacteria. It is generally acknowledged that after acquiring tuberculin reactivity active tuberculosis develops in 1 in 20 tuberculin converters in the next year, 1 in 40 in the second year following skin test conversion and in another 1 in 40 for the remaining lifetime. Thus only a total of 1 in 10 who attain tuberculin positivity and latent infection proceed to active disease. Stated in another way, 90 percent of tuberculin positive individuals never develop clinical tuberculosis. Though most patients who develop tuberculosis are tuberculin positive at the time of diagnosis, a minority of patients are tuberculin negative, particularly those with a compromised cellular immune system, examples of which are AIDS, end stage renal disease, lymphoma, chemotherapy, etc., as well as some without a propensity for diminished cellular immune recognition. Importantly, these considerations have led tuberculosis specialists to recognize that tuberculin positivity should not be used to diagnose tuberculosis , but rather that diagnosis requires positive smears/cultures and often consistent chest x-rays and clinical signs and symptoms. More recently tuberculous infection has become more specific (yet less sensitive) utilizing another test of cell mediated immunity which depends on assays for IFN-gamma release, exploiting the selective expression of ESAT 6, CFP-10 and TB 7.7 by M tuberculosis. Though positivity in this commercially available blood test is more specific in the presence of BCG vaccination and environmental mycobacteria, it again is a measure of TB infection and not active clinical tuberculosis.
The diagnosis of leprosy infection/disease is complicated by the fact that while at the lepromatous pole patients produce antibody, they have an M leprae-specific cellular immune anergy, while just the opposite has been well-established for those at the tuberculoid end of the leprosy spectrum. Consequently, the lepromin skin test, a measure of cellular immunity, while commonly positive in TT/BT leprosy and some household/endemic contacts,is negative in LL/BL leprosy. On the other hand, while antibodies to PGL1 and now M leprae proteins are positive in the large majority of LL/BL patients, they are not dependably positive in BT/TT leprosy, and even less commonly those antibodies to M leprae specific proteins.
Douglas et al (Clin Diag Lab Immun.11,2004,897) established that the presence of antibodies to PGL1 in healthy household contacts of MB patients at the time of diagnosis of the index case or in sera collected every six months for five years subsequently increased the risk of the subsequent development of leprosy 7-fold as compared to healthy household contacts who were persistently seronegative. In itself this is an interesting and important finding, but also establishes that antibody to PGL1 is a marker of infection and not clinical leprosy. Yet of those household contacts who have antibodies to PGL1 at any time only 14 percent develop leprosy, suggesting that PGL1 antibody positivity in that locale results in a similar rate of progression of latent infection to active leprosy as has been found for the risk of tuberculin positve individuals progressing to active tuberculosis. It is unclear if M leprae protein antibodies might enhance antibody to PGL1 as a predictive marker of progression from leprosy infection to disease. In that respect, Duthie was previously provided serial sera collected from the Douglas study from all initially healthy MB contacts who developed leprosy, both who had initial antibodies to PGL1 or developed them, as well as those who never had antibodies to PGL1, and thus that question could surely be addressed. Though there are conflicting findings on the efficacy of chemoprophylaxis of leprosy, several studies found such interven-tion efficacious. If there is utility to blood test for leprosy infection, targeting chemoprophylaxis to household contacts of MB patients and populations in highly endemic areas with markers of leprosy infection, may prove to be the most important application for leprosy specific blood tests.
In another study it was found that in an endemic leprosy community in Indonesia (IJL.62,1994,1), where 1 percent of the population carried a leprosy diagnosis, 31 percent of the uninfected population had antibody to PGL1 and 8 percent of these harbored M leprae in nasal smears. Also, in a long term study of healthy family contacts of leprosy patients in French Polynesia (IJL.61,1993,533) 17 percent had significant antibodies to PGL1, yet only 2 percent of those patients developed leprosy over the next 9 years. Geluk (Lepr Rev.84, 2013,3) in a recently published and elegant paper makes a case for T cell assays for leprosy, and therein provides her findings that in healthy contacts of MB leprosy patients T cell responses to M leprae unique proteins augumented by 71 percent the number of leprosy-infected individuals that were not detected using serology based assets. Ultimately Geluk advocates, because of the immunologically mediated leprosy spectrum, that diagnosis of infection/disease might best use both serologic and cellular immune assays. Nonetheless , it will never be possible to determine if even a large array of both serologic and cellular assays are capable of detecting the full complement of leprosy infected individuals who are not yet diseased as there can be no reliable measure of the totality of that population. In conclusion, these and several other studies make it abundantly clear that considerably more healthy individuals become infected with M leprae, and like the case in tuberculosis, never develop clinical disease.
2. I have little to add to the concerns of van Brakel and Lockwood of 4 April 2013 regarding the recent claims of near 100 percent sensitivity of the new serology for the diagnosis of leprosy. I would, however like to reemphasize that Duthie’s own data, as well as that of many others has consistently demonstrated serology to be insensitive in TT/BT/IND leprosy, the majority of leprosy cases who currently present internationally, and the lack of clarity of what antibody to their proteins adds to the sensitivity that is afforded by antibody to PGL1 alone in the diagnosis of leprosy.
3. There is nothing surprising in that Duthie et al found a fall in antibody titers to M leprae proteins and PGL1 in the course of effective treatment of leprosy and a rise in these titers upon clinical relapse. In fact in the first paper describing the utility of PGL1 antibodies in leprosy patients (Infect Immun.41,1983,1077) wherein I provided sera from leprosy patients both in San Francisco and Malaysia to the group at Colorado State University, a regular fall in titers to PGL1 in treated patients was well documented. Furthermore, in 1989 in the first major study (IJL.57,1989,744) of serum antibody to PGL1 obtained serially from leprosy patients undergoing therapy, we found that antibody to PGL1 consistently fell but at rates that varied considerably, such that the rate of fall could not be utilized to assess the efficacy of therapy; in that study antibodies were still present in long-term treated patients, and antibody titers that did not fall or actually rose were a harbinger of relapse. Duthie et al found that antibodies to M leprae proteins in leprosy patients undergoing initial chemotherapy fell more rapidly than those to PGL1. Yet their claims that protein antibody responses may be utilized to monitor treatment efficacy and provide criteria for patient discharge appear unwarranted and are not evident from their data. In fact they found in Brazil that at the completion of one year WHO MDT therapy, antibody to their proteins were still present in between 50 percent and 91 percent of patients. Previously it had been demonstrated repeatedly that the bacteriologic index (BI) regularly fell one unit on each year of effective chemotherapy and the morphologic index (MI) falls even more rapidly, convincingly and dramatically in 4 weeks on dapsone alone and in 2 weeks when rifampin is added (IJL.44,1978,548). Thus BI/MI appear better suited as monitors of effective leprosy chemotherapy. Furthermore MB patients who relapse after the completion of MDT generally pass through a period of BI negativity, generally for several years, which would be associated with antibody negativity to M leprae specific proteins. Thus neither a negative BI or M leprae serum protein antibodies can reliably predict cure.
4. As we already presented, it is of considerable concern that Duthie et al advocate antibodies for the diagnosis of leprosy itself. Though blood tests for leprosy diagnosis have great scientific interest, they are largely unnecessary. For that purpose and in the vast majority of leprosy cases, diagnosis can be confirmed by clinical examination alone and supplemented by skin smears and biopsies, while again blood test cannot be used to distinguish between infection and actual disease. Though the diagnosis of leprosy is best performed by those with clinical experience with the disease, and a leprosy diagnosis may at times be confused with other dermatologic diseases, these are rarely at issue. The nodules, plaques and general thickening of the dermis of lepromatous disease are quite distinctive. At the lepromatous pole abundant numbers of acid fast bacilli are regularly found in skin smears, and skin biopsies regularly contain “foam cells” and other distinctive histopathologic features. At the tuberculoid pole leprosy diagnosis may at times be unclear. The hallmark of these forms of leprosy are the presence of hypopigmented macules, hypoasthesia within lesions and characteristic dermatopathology. In some TT and BT cases all three criteria may not be found, so eminent leprologists of the 20th century would make a leprosy diagnosis when 2 of these 3 characteristics where observed. Largely in order to simplify leprosy diagnosis, the WHO abandoned skin smears and biopsies. Despite the experience that the interpretation of skin smears and biopsies can vary somewhat from place to place, for their broad meaning they are reliable. Furthermore, that while advocates for leprosy blood tests have claimed that skin smears and biopsies are time consuming and labor intensive, these were regularly successfully performed in the field for the greater part of the 20th century. Also, despite the previous arguments that blood tests for leprosy are not a sensitive indicator of actual disease but only indicative of leprosy infection, claims that these blood tests can be made “field-friendly” are not self evident.
In summary: Blood tests for leprosy cannot distinguish infection from disease. Leprosy diagnosis can generally be reliably suspected from clinical presentation alone and skin smears and biopsies utilized to confirm that diagnosis. Furthermore, the BI and MI, though largely abandoned, still appear the time-proven and best established means of laboratory monitoring the effectiveness of antimicrobial therapy and indicating relapse.
Best regards,
Robert Gelber