540 Letters to the Editor

Role of Fungi in Allergic Fungal Sinusitis and ChronicRhinosinusitis

To the Editor: In the recent study by Ponikau et al,' the authorsdescribe a new and more efficient method of nasal mucus collec­tion. They subsequently used this method to demonstrate that of2IOpatients diagnosed as having chronic rhinosinusitis (CRS), 202patients (96%) had fungi (representing 40 different genera) in theirnasal mucus. They further found that 100% of their control patients(14114) also had fungi (from 8 different genera) in their mucus.

A subset of 101 CRS patients was also treated surgically, andnasal samples were subjected to histopathological examination.Fungal elements were found in 82 (81%) of the 101 samplesexamined. They also found that eosinophil "presence" was de­tectable in 96% (97/101) of these cases . In contrast, tissue from4 control patients subjected to histopathological examinationcontained no evidence for the presence of eosinophils. Basedon their findings, the authors conclude that allergic fungalsinusitis (AFS) is significantly underdiagnosed and is in factpresent in the majority of patients suffering from CRS. Theyalso suggest that immunoglobulin E (lgE)-mediated type I hyper­sensitivity is not the dominant pathophysiological factor in AFS.They indicate that such a role is played by eosinophils and suggestthat AFS be renamed eosinophilic fungal rhinosinusitis (EFRS) .

For the most part, the article presents some thoughtful, intrigu­ing, and provocative findings. However, it is also somewhat per­plexing how the authors arrive at their conclusion that fungi play animportant role in the onset of AFS. In part this is due to the fact thattheir reported findings violate Koch's postulates on causation ofdisease and that the authors actually demonstrate that the "causativeagents," ie, fungi, are present in both diseased and control patients.Furthermore, the authors did not demonstrate that the same symp­toms can be induced in a healthy subject after infection with thepathogen, in this case the fungus. Hence, it is rather difficult toaccept, based on the preliminary data presented, the authors' asser­tion that fungi are important in the etiology of AFS and CRS.

The authors' finding that 96% of the individual CRS surgicaltissue examined revealed the presence of eosinophils appears tobe important, since this finding contrast s with the fact that tissuespecimens from all 4 control patients contained no eosinophilpresence. However, this is not a new finding. Harlin et af reportedin 1988 on a study of 26 patients suffering from chronic sinusiti sthat tissue from such patients "was extensively infiltrated witheosinophils.' Indeed, Hansel,' back in 1929, was the first to reportsuch infiltration by eosinophils. Furthermore, this contrast ineosinophil infiltration between CRS vs control patients wouldhave been more compelling if the authors had carried out exami­nations on a clearly defined control population.

I also noted that a Mayo Clinic news release! issued immedi­ately after the article was published stated that "Mayo Clinicresearchers say that they have found the cause of most chronicsinus infections-an immune response to fungus." The releasewent on to state that the authors of the article say "this discoveryopens the door to the first effective treatment for this problem."The study findings' and these claims" were criticized in a Washing­ton Post article published in late November 1999.5 Experts in

Mayo Clin Proc, May 2000,Vol75

otolaryngology expressed concern that if "100% of controls havefungi, why does it [Mayo findings) mean anything?" Furthermore,the Washington Post article' raised the issue that patient expecta­tions to be treated successfully for CRS had been elevated unduly.

It would appear with hindsight that the Mayo findings' mayhave been reported prematurely. This has already resulted in bothcriticism and skepticism of the work. On reflection this is some­what unfortunate, since even the critics of the work" agree thatmore comprehensive studies should be done to explore the pos­sible role of fungi in AFS and CRS. However, I would hope thatall of us continue to strive for candid and objective assessment ofwork. This is particularly important in regard to those findingsthat may impact patients' hopes and expectations regarding treat­ment of a particular illness or disease . We should all err on theside of caution in terms of conclusions we draw and report basedon a very preliminary data set.

Stephen Naylor, PhD, DScMayo Clinic RochesterRochester, Minn

I. Ponikau JU, Sherris DA, Kern EB, et al. The diagnosis and incidenceof allergic fungal sinusitis. Mayo Clin Proc. 1999;74:877-884.

2. Harlin SL, Ansel DG, Lane SR, Myers J, Kephart GM, Gleich GJ. Aclinical and pathologic study of chronic sinusitis: the role of theeosinophil. J Allerg y Clin Immunol . 1988;81:867-875.

3. Hansel FK. Clinical and histopathologic studies of the nose andsinuses in allergy. J Allergy. 1929;1:43-47.

4. Mayo Clinic Rochester News. Mayo Clinic study implicates fungusas cause of chronic sinusitis [news release], September 9, 1999.Available at: www.mayo.edulcomrnlmcr/news/news_773.htmI.Ac­cessed April 7, 2000.

5. Boodman SG. Mayo report on sinusitis draws skeptics; some expertschallenge fungal basis for many infections. Washington Post. Novem­ber 23, 1999:Z07. Available at: www.newslibrary.comldeliverccdoc.asp?SMH=313919. Accessed February 17,2000.

In reply: As Dr Naylor states, extramucosal fungi were found inalmost all patients with CRS and also in all controls.' To concludefor this reason that fungi do not cause disease is flawed. We knowthat the bacteria responsible for most cases of acute sinusitis arealways present in normal, healthy hosts. An event (such as anupper respiratory tract infection) occurs that causes swelling andleads to a cascade of events that allow infection by the samebacteria . From an infectious disease point of view, assuming thatthe mere presence of organisms causes disease, the conclusionswe drew from our study can certainly be challenged. But from theperspective of a hypersensiti vity reaction, which is not IgE medi­ated, our hypothesis seems plausible. For example, everyoneinhales pollens, but only a sensitized patient develops the symp­toms of hay fever (in that case, IgE mediated) .

Thus, the term infection is not applicable for the disease, EFRS,which we are describing. An infection is defined by microorganismsentering tissue, usually through the skin or a mucus membrane. InEFRS, the fungalorganisms arc always presentextramucosally in themucus and are not themselves causative for this disease.

As Naylor also states, the presence of eosinophils in the tissuein CRS is not new, nor is it reported as new in our article. What wereported is the observation that the eosinophils are leaving thetissue and form the typical clusters in the eosinophilic mucin. A keypoint of our article, besides the presence of fungi in the mucus, is

For personal use. Mass reproduce only with permission from Mayo Clinic Proceedings.

Mayo Clin Proc, May 2000, Vol 75

the presence of allergic (eosinophilic) mucin in 96% of consecutivechronic sinusitis patient s. Thus, the eosinophils visualized in thetissue in the past were cells only in transit into the lumen.

We hypothesized in the article that the targets in the mucin are theextramucosal fungi. Recently, we were able to further validate thishypothesis and demon strated that the eosinophils are indeed targetedagainst fungal organisms in CRS, a phenomenon that was absent inhealthy controls as well as in patients with allergic rhinitis.?

We have visualized that the eos inophils were attacking anddestroying noninvasive fungi in the mucin of CRS patients. Thegranule proteins released during that attack are also known to behighly toxic to the epithelium and have been shown to be essentialin the damage of the epithelium.' Thus, the mere presence of fungiis nonspeci fic in this disease, but fungi are the target for theeosinophil s in the CRS patient popul ation and are essenti al as atrigger to stimulate the immunologic (eosinophilic) response tothem in a sensitized individual. For these reasons we advocate thenew terminology eosinophilic f ungal rhinosinusitis. These find­ings and the conclusion we have drawn from them have beenconfirme d by a second research group:

Criticisms like those published in the Washington Post are usuallyleveled by colleagues who are not informed about the work beingdone or have not paid close enough attention to the details provided.They also did not support their comments with data.

The European Rhinologic Society, the Euro pean Federation ofOt o-Rhino-L aryngological Societ ies, and the InternationalRhinologic Society have planned plenary sessions on the subjectat upcoming meetings. Th e International Fede ration of Oto­Rhino-Laryngological Societies has invited us to part icip ate asfaculty members at the Consensus Con ference on Nasal Polyposis(scheduled for October 2000 in Siena, Italy) to help set standardsand move research in new dire ctions.

Our findings have led us to the development of new treatmentoptions. Blinded, randomized, placebo-controlled drug trials havenot been completed or published.

Jens U. Ponikau, MDDavid A. Sherris, MDEuge ne B. Kern , MDMayo Clinic RochesterRochester, Minn

I. Ponikau JV , Sherris OA, Kern EB, et al.The diagnosis and incidenceof allergic fungal sinusitis. Mayo Clin Proc. 1999;74:877-884.

2. Ponikau JV , Sherris OA, Kern EB. Chronic rhinosinusitis: an im­mune response to fungi. Allergologie. 1998;21:581.

3. Harlin SL, Ansel OG, Lane SR, Myers J, Kephart GM, Gleich GJ. Aclinical and pathologic study of chronic sinusitis: the role of eosino­phils. J Allergy Clin lmmunol. 1988;81:867-875.

4. Braun H, Hofmann TH, Freudenschuss K, Buzina W, Ponikau J,Stammberger H. Eosinophilic fungal rhinosinusitis (EFRS): from fun­gus to chronic sinusitis. Paper presented at: Annual Meeting of theAustrian Medical Association; November I I, 1999; Graz, Austria.

Lactulose vs Sorbitol for Treatment of Obstipation inHospice Programs

To the Editor: I appreciated Dr Kaur' s Concise Review for Clini­clans' on palliative care in hospic e programs. In reference to Table

Letters to the Editor 541

I of this rev iew, I would like to mention an article by Led erle et aFdemon st ratin g that sor bitol is therapeutically equi valent tolactul ose. Because sorbitol cos ts much less than lactulose, itshould be the preferred agent.

Harri son G. Weed , MDOhio State University Colle ge of MedicineColum bus

I. Kaur JS. Palliative care and hospice programs. Mayo Clin Proc. 2000;75:181-184.

2. Lederle FA, Busch OL, Mattox KM, West MJ, Aske OM. Cost­effective treatment of constipation in the elderly: a randomizeddouble-blind comparison of sorbitol and lactulose. Am J Med.1990;89:597-601.

In reply : Dr Weed raises a good point. He is correct that bothsorbitol and lactulose are osmotic laxatives and are therapeu ti­ca lly equivalent if constipatio n is the only ind ication for their use.The study he ci tes by Lederle et al' was done in an elderlyambulatory male popul ation . Patients with metastat ic cancer andthose receivi ng narcotics were excluded from eligibility.

In our hospice program, about 80% of patients have metastaticcancer as a terminal diagnosis, and a large subgroup of these patientshave liver metastases as a complicating factor. In this patient popula­tion, we prefer to use lactulose, if other less aggressive bowel pro­grams have failed, because of its theoretical advantage in also treatinghepatic encephalopathy. Lactulose, like sorbitol, is poorly absorbedfrom the gastrointestinal tract and reaches the colon virtually un­changed. However, lactulose, unlike sorbitol, is broken down prima­rily to lactic acid in the colon, exerting an acidic effect on coloniccontents.' In this acidic environment, the reabsorption of ammoniaand other substances thought to cause hepatic encephalopathy isimpaired. Additionally , the acidic colonic environment inhibits thegrowth of colonic bacteria responsible for producing ammonia andother molecules linked to hepatic encephalopathy. The results are lessammonia and related compounds available to be reabsorbed from thecolon and, therefore, presumably diminished hepatic encephalop­athy. The acidic colonic environment created by lactulose is advanta­geous in treating the hepatic encephalopathy often associated withadvanced malignant disease.'

In our institution , 16 oz of lactulo se is $9.34 vs $2.21 for 16 ozof sorbitol. Many patients with hepatic metastases have unrecog­nized hepatic encephalopathy": therefore, we prefer to use lactulosedespite the cost difference. For patients without complications,sorbitol should certainly be considered the cost-effective option.

Judith S. Kaur, MDMayo Clinic RochesterRochester , Minn

I. Lederle FA, Busch OL. Mattox KM, West MJ, Aske OM. Cost­effective treatment of constipation in the elderly: a randomizeddouble-blind comparison of sorbitol and lactulose. Am J Med. 1990;89:597-601.

2. Kristalose [lactulose] for oral solution. In: Physicians ' Desk Refer­ence. 54th cd. Montvale, NJ: Medical Economics Co Inc; 2000:773.

3. SmithBC,James OF.Thefailingmalignant liver. Gut. 1998;42:454-455.4. Eras P, Sherlock P. Hepatic coma secondary to metastatic liver

disease. Ann Intern Med. 1971;74:581-583.

For personal use. Mass reproduce only with permission from Mayo Clinic Proceedings.