Allergic Fungal Sinusitis (AFS)

Although Allergic Fungal Sinusitis (AFS) was first described in the early 1980’s, the most appropriate evaluation and long term treatment methods remain elusive. Currently it is felt that AFS requires both surgical and medical management. The medical management may include topical and systemic steroids, anti-fungal agents and immunotherapy. The effectiveness and required duration of the various medical treatments remain unknown and under active investigation.

We present, in this publication, our experience with surgical and medical management of these patients and propose a treatment protocol using the present state of knowledge. The evolution of patient evaluation has included total and fungal specific IgE levels and serum and mucin Eosinophilic Cationic Protein (ECP) levels. Endoscopic mucosal staging is the only effective means of tracking the disease course.
Recommendations are made for both surgical intervention and the use of systemic steroids. There is a significant body of information to indicate that this may be a chronic disease, which at our current state of knowledge can only be controlled, rather cured.


Allergic Fungal sinusitis (AFS) patients have no unique symptoms, which set them apart from other chronic sinusitis patients. Consequently, the disease process described only 15 years ago must be suspected to be diagnosed. Once diagnosed, the next challenge is treatment.

The disease was first described my Millar and Associates in 1981 as allergic aspergillus sinusitis (1). Their five patients demonstrated significant Type I hypersensitivity (IgE mediated) to aspergillus fumigatas. There were also sinus pathologic findings in these patients similar to the pulmonary findings in Allergic Bronchopulmonary Aspergillosis (ABPA) patients. ABPA is a disease characterized by asthma, increased total serum IgE, pulmonary eosinophilia, specific allergic immune response and thick tenacious mucous producing bronchial obstruction and bronchiectasis (2,3).

Katzenstein (4) reported nine allergic aspergillus sinusitis cases in a retrospective pathologic study of 119 sinus surgical cases. These patients also shared pathologic features with ABPA. Several reports were then published describing this disease associated with other fungi, notably the dematiatious family (5-14). Consequently, the disease has become known as allergic fungal sinusitis (AFS).

There has been little uniformity of patient criteria for inclusion in AFS studies. The Mayo Clinic reported 51/789 sinus histologic specimens as AFS. Their inclusion criteria were no microscopic tissue invasion, characteristic allergic mucin plus fungal hyphae on special stain or positive fungal culture. They also reported an AFS like syndrome without the positive evidence of fungus.

Bent and Kuhn outlined the clinical and pathologic features in 15 consecutive AFS patients (15). They evaluated 11 findings. Five of the eleven were common to all 15 patients and therefore seem necessary to define a patient study population. These criteria are:

  1. Type 1 (IgE mediated) hypersensitivity
  2. Nasal polyposis
  3. Characteristic CT findings
  4. Eosinophilic mucous
  5. Positive fungal smear or culture.

The other six findings support the diagnosis and are important in describing any individual AFS patient.

Although there are no unique symptoms setting these patients apart, some findings which may raise ones index of suspicion include: unilateral nasal polyposis, young age, characteristic serpiginous sinus opacity on CT, thick sticky yellow/green mucous and a nasal polyp patient having no other known disease who responds only to oral steroids.

The prognosis and optimum treatment of allergic fungal sinusitis are not yet clear. Waxman (16) reported three groups of patients:

  • Immediate recurrence
  • Delayed recurrence
  • Cure

He reserved oral steroids for those patients who recurred following surgery. The rationale for using oral steroids derives from the treatment of ABPA and the similarities between the two diseases.
Kupferberg et. Al. (17) refined the endoscopic follow-up into a proposed staging system, which allows closer control of the mucosal response to an oral steroid dose.

  • Stage 0 – No mucosal edema or allergic mucin
  • Stage 1 – Mucosal edema w/ without allergic mucin
  • Stage 2 – Polypoid edema w/ without allergic mucin
  • Stage 3 – Sinus polyps with fungal debris / mucin

He reported 24 patients with a 12-month average follow-up and found that eight of nine patients treated only with surgery recurred at stage 11 or higher. One patient was reported disease free at 22 months but there was no endoscopic follow-up. The longest times to recurrence were one patient at 29 months and a second patient at 34 months. All patients who recurred in this group required re-operation to remove nasal polyps and allergic fungal load.

All 10 of their surgery and Prednisone patients were treated with a Prednisone burst, then tapered to lower doses. He found that if the steroid dose was too low the mucosal stage increased, i.e., the disease became worse. This suggested that the mucosal stage and length of treatment are dose dependent and tied to some unknown factor, which responds to systemic steroids. He concluded that mucosal stage was a reflection of disease activity.

Mabry and colleagues have made a considerable effort investigating immunotherapy for AFS (). They have used 11 relevant fungal antigens for testing and immunotherapy. They found that all patients with AFS displayed sensitivity to multiple fungal antigens. They have reported a decreased amount of crusting and polyposis as well as a reduction in the need for systemic and topical corticosteroids in these patients. Their study however did not have a control group for comparison. A retrospective review by Ferguson of seven AFS patients who received immunotherapy indicated that five patients did not improve or worsened with immunotherapy after adequate surgical and medical management had been achieved. Concerns have existed with the use of immunotherapy for AFS because allergen-specific IgE produced by immunotherapy could theoretically incite a Gell and Coombes Type III reaction. This has not been borne out with the work done so far. The other concern with immunotherapy is the possible worsening of disease with the introduction of extraneous fungal antigens in patients with allergic fungal disease. Another difficulty with immunotherapy is that many of the common fungal antigens are not available (e.g. Bipolaris antigen is not available commercially). Also, fungal specific serum IgE levels are not available for all fungi.

Treatment Protocol

All patients, after having been positively diagnosed for AFS using the criteria established by Bent and Kuhn, undergo functional endoscopic sinus surgery to reduce the fungal load and to create adequate drainage pathways for the sinuses. The allergic mucin obtained at surgery is sent immediately to the Allergy/Immunology research laboratory for culture. This is then forwarded to Dr. Sid Crow, Georgia State University Biology Department for identification. Serum is drawn for total IgE levels and saved at 70 degrees for later fungal specific IgE and other studies, as they become available.

With the objective of keeping the mucous membrane at stage 0, Prednisone is started within 48 hours of surgery. This takes advantage of the decreased edema caused by surgical removal of the fungal load.

Our recommendation at present is to begin oral Prednisone in a dose ranging from 0.4 – 0.6 mg/kg per day for four days then decrease by 0.1mg/kg per day in cycles of four days at each dose until reaching 0.2mg/kg per day. This dose is then maintained and the patient is followed monthly with nasal endoscopy and total serum IgE levels. The patients weight and Prednisone dose is recorded. The nasal and sinus mucous membranes are endoscopically staged according to Kupferberg, et. Al. The Prednisone dose is then adjusted based on the mucosal stage to maintain the mucous membrane at stage 0. This information is recorded on an encounter form. Each patient’s total serum IgE level, Prednisone dose and clinical stage are plotted monthly on a graph. The patient is tapered off of Prednisone over a period of two to three months after maintaining normal mucosa (stage 0) for four months. While the oral Prednisone is being tapered, the patient is started on intranasal steroids. The patient is then followed monthly with nasal endoscopy and IgE levels for six months following cessation of Prednisone and then every two months for an indefinite period. The intranasal steroids are continued for the first year in most patients. It is felt that the patients will have to be followed longer than 36 months after the Prednisone therapy, since our longest time to recurrence after surgery without oral Prednisone has been 34 months.


At our present state of knowledge, all patients appear to need surgical debridement with removal of fungal load, obstructing polyps and restoration of their mucociliary clearance. We do not know why AFS occurs or why it recurs following surgery or termination of medical treatment.

It is generally believed that the etiology of this disease is a mucosal hypersensitivity directed against fungal antigens deposited on sinus mucosa. Removing the allergen load may reduce the allergic response and therefore reduces the edema. However, this may simply be a result of removing eosinophilic inflammatory mediators, which in turn reduced the inflammatory reaction and the edema.

If it is the allergic response that is being reduced, then the question arises as to why the edema recurs so quickly following surgery (as soon as 2 – 3 months) in the absence of any gross fungus. Other questions which arise included: 1) if the disease is extra mucosal, how long does it take the cilia to sweep the remaining fungal elements out of the sinus 2) what is the objective of treatment 3) is the objective to hold the mucosa normal long enough for the mucociliary clearance to fully empty the sinus 4) are we missing areas of mucosal invasion for lack of sufficient biopsies 5) is this perhaps the reason allergic fungal sinusitis recurs so frequently or are the patients perhaps “re-infected” in their environment. Another question is whether or not this is an allergy at all or is it just a local inflammatory response induced by the eosinophils attracted to the sinus.

Whatever the answers to these questions may be, in our experience all patients need medical treatment in addition to surgery. This is born out by the clinical course of nine surgery only patients included in Kupferberg’s paper. Eight of these patients recurred without medical treatment. The 9th recurred after the paper was presented. The earliest recurrence was at 2 – 3 months, the two longest were at 29 and 34 months post-operatively.

It has also been our experience that if the mucous membrane appears to be improving and the steroid dose is decreased too low, too quickly, the disease will worsen. The evidence will be increased mucosal stage at subsequent follow up visits.

In following these patients clinically, it would be extremely helpful to have a serum market to indicate whether the disease is becoming worse at the cellular and chemical level, before it is apparent visually. To this end we have attempted to follow several serum marks notably eosinophil cationic protein (ECP), total serum IgE and fungal specific serum IgE levels. A correlation between clinical findings and ECP was not demonstrable and ECP levels have not proven to be a satisfactory marker for the disease. Serologic tests are not available for all fungi. Serum is therefore collected and banked at 70 degrees in order to retrospectively evaluate these sera when the fungal specific IgE test becomes available.

The only serum marker that we currently are able to follow is total serum IgE. IgE levels do fluctuate with mucosal stage and are tracked in all our patients. We have not been able to demonstrate a correlation between absolute IgE levels and clinical disease, however, IgE levels, do seem to come down over time and oral steroid treatment.

Fungal specific IgE may be more helpful when we are able to adequately set up the testing, however, it is not available for fungi such as bipolaris, aspergillus and penicillium.

At present the most helpful clinical tool is nasal endoscopy and the mucosal staging system. If any area of the sinuses qualifies for a given clinical stage the patient is staged as such even if other areas, such as the middle meatal antrostomy, middle turbinate or ethmoid cavity are at stage 0.

One important observation is inter-current bacterial infection. This may quickly change a normal stage) sinus into a stage II (polypoid mucosal edema) with purulent discharge as opposed to allergic mucin. This should be recognized, an endoscopically guided culture performed and appropriate culture directed antibiotics prescribed. The patient should be re-examined and re-staged in two weeks. If the patient’s stage returns to his or her pre-bacterial infection stage, this aberration in the clinical stage should be noted, but not considered in determining length of treatment. Generally the IgE level will not with inter-current bacterial infection masquerading as recurrent AFS.

If this disease is truly not invasive and is only an immunological reaction to fungal allergens, complete fungal removal should equate with a cure. Complete surgical removal is not possible. Consequently, normal mucociliary clearance must be restored so the sinus mucosal can sweep out the remaining fungal elements. The only question then is how long must the sinus remain normal before discontinuing Prednisone. This inherently assumes that normal appearing mucosa functions normally while the patient is on Prednisone. A four-month period at stage 0 was arbitrarily picked as a starting point for studying the effect of Prednisone on allergic fungal sinusitis.

Since the cause of the disease and its recurrence is unknown, we do not know whether patients become “re-infected” in their environment or if residual fungal elements release chemotactic agents which in turn re-attract eosinophils to the sinus. These eosinophils may then release their inflammatory mediators (ECP, Eosinophil Peroxidase, Eosinophil Derived Neurotoxin, Major Basic Protein) which again produce the edema, polyps and allergic mucin.

In conclusion, this is a fascinating disease, the nature of which is only beginning to be understood. The future holds exciting possibilities for investigation and treatment. This disease may, however, be chronic without a “cure”.

Further Reading Material

For further information on Fungal Rhinosinusitis, you can read more in the Fungal Rhinosinusitis.

Table of Contents