Invasive Fungal Rhinosinusitis: A 15-Year Experience With 29 Patients

Transcription

1 The Laryngoscope VC 2013 The American Laryngological, Rhinological and Otological Society, Inc. Invasive Fungal Rhinosinusitis: A 15-Year Experience With 29 Patients Marcus M. Monroe, MD; Max McLean, BA; Nathan Sautter, MD; Mark K. Wax, MD; Peter E. Andersen, MD; Timothy L. Smith, MD; Neil D. Gross, MD Objectives/Hypothesis: Document a 15-year experience with 29 cases of acute invasive fungal rhinosinusitis (AIFR) and evaluate factors predictive of disease clearance and overall survival. Study Design: Case series with chart review. Methods: Patients were identified by review of department billing records between 1995 and Medical records were reviewed for patient demographics, disease characteristics, clinical course including surgical and medical therapy, treatment outcomes, and long-term survival. Results: Twenty-nine patients with AIFR were identified. Causes of immunosuppression included hematologic malignancy (n516), diabetes (n512), medication (n510), and acquired immunodeficiency syndrome (n51), with 10 patients having multiple causes of immunosuppression. Facial pain, swelling and orbital symptoms were the most common presenting symptoms. Fungal organisms included Mucor (n518) and Aspergillus (n510) species, with one patient infected with both. Disease-specific survival (DSS) from AIFR was 57%. Intracranial (P5.01) and ethmoid sinus (P5.05) involvement were significantly linked with short-term disease-related mortality. Overall survival (OS) at 6 months was 18%. For OS, intracranial involvement (hazard ratio [HR], 4.47; 95% confidence interval [CI], ) and cranial neuropathy at presentation (HR, 3.2; 95% CI, ) were significantly associated with shortened survival. Of the five patients surviving >6 months, two developed long-term major sinonasal complications. Conclusions: DSS and OS remain low for patients with AIFR. Extensive surgical resection in patients with these poor prognostic signs should be considered carefully in light of their poor survival. Long-term survivors are at significant risk of sinonasal complications and should be followed closely. Key Words: Immunocompromise, invasive fungal sinusitis, fulminant, Mucor, Aspergillus, aspergillosis, rhinocerebral mucormycosis, mold, fungus. Level of Evidence: 4. Laryngoscope, 123: , 2013 INTRODUCTION Acute invasive fungal rhinosinusitis (AIFR) is a rare but frequently fatal infection that occurs primarily in immunocompromised individuals. It is characterized by fungal invasion into the mucosa and submucosal structures of the paranasal sinuses with frequent extension into adjacent structures, including the paranasal soft tissues, orbit, and cranial vault. Short-term mortality has been reported to range from 20% to 68% across studies. 1 6 Immunocompromised individuals represent the population most at risk for the development of AIFR, From the Department of Otolaryngology Head and Neck Surgery, Oregon Health and Science University, Portland, Oregon, U.S.A Editor s Note: This Manuscript was accepted for publication December 14, The authors have no funding, financial relationships, or conflicts of interest to disclose. The authors have no other funding, financial relationships, or conflicts of interest to disclose. Send correspondence to Neil D. Gross, MD, Department of Otolaryngology Head and Neck Surgery, Oregon Health and Science University, 3181 Sam Jackson Park Rd., PV-01, Portland, OR grossn@ohsu.edu DOI: /lary although there are rare case reports describing AIFR in patients with apparently normal immune function. 7 The most frequently affected patient groups include those with immunosuppression related to hematologic malignancy, bone marrow transplantation, poorly controlled diabetes, acquired immunodeficiency syndrome, immunosuppressive medications, and chemotherapy. Although a wide range of fungal species have been reported to cause AIFR, the most frequent are Aspergillus and Mucor species. Previous reports have demonstrated an association between Mucor AIFR and increased orbital and neurologic sequelae as compared with Aspergillus AIFR, with a trend toward increased mortality. 8 Others have demonstrated the importance of recovery of immune function in disease clearance. 3 However, little other prognostic data are known regarding this rare patient population, specifically as it relates to long-term survival. Our understanding of AIFR has been hampered by the rarity of the disease, combined with reports detailing varied treatment strategies, heterogeneous patient populations, and inconsistent nomenclature. Differentiation between AIFR and other forms of invasive fungal rhinosinusitis, such as granulomatous and chronic invasive fungal sinusitis, was frequently not provided in 1583

2 historical cohorts. Recent attempts to standardize the nomenclature and classification of invasive fungal rhinosinusitis have led to a definition that includes an acute time course of <4 weeks combined with the presence of tissue invasion by fungal elements. 1,9 Because of these limitations, the best management strategies for patients with AIFR are frequently unclear. This is particularly the case for patients with disease progression outside of the paranasal sinuses despite medical therapy and limited endonasal debridement. In these cases, the morbidity of further aggressive surgical debridement must be balanced with an understanding of the overall prognosis of the patient, frequently a difficult task for both patient and provider. Clearly a need for better prognostic information exists. The primary objectives of this study were to evaluate factors predictive of disease-specific survival and overall survival in patients surgically treated for AIFR. MATERIALS AND METHODS All patients surgically treated for AIFR by the Department of Otolaryngology Head and Neck Surgery at Oregon Health and Science University between 1995 and 2010 were included in the study. After institutional review board approval (#6661), patients were identified through department billing records. Medical records were reviewed for patient demographics, clinical presentation, surgical and medical treatment, and clinical outcomes. Each patient s vital status was confirmed by a review of the computerized hospital records and/or query of the Social Security Death Index ( The diagnosis of AIFR was made according to published guidelines 1,9 andincludedanacuteclinicaltimecourse(<4 weeks) as well as the histologic presence of invasive fungal elements within sinus mucosa, submucosa, or bone. Surgical treatment consisted of both endoscopic and open procedures. The choice of antifungal agents was managed on a case-by-case basis by infectious disease specialists and typically consisted of amphotericin B and/or voriconazole, depending on the organism recovered. Aspergillus and Mucor species were differentiated based on their histologic appearance. Aspergillus species appear as straight septate hyphal forms with acute angle branching, whereas Mucor species appear as broad irregular nonseptate hyphae with rightangle branching. Orbital and intracranial involvement were defined as imaging or intraoperative findings of fungal invasion. The primary outcome of the study was disease-specific survival. Given the difficulty of assigning cause of death in a retrospective study, any death in a patient with ongoing fungal infection was attributed to AIFR. The secondary outcome of the study was overall survival. Univariate analysis of prognostic factors with categorical data was performed with v 2 and Fisher exact test where appropriate. Continuous data were analyzed using independent t tests. Univariate Cox proportional hazards models were used to examine the relationship between study variables and overall survival. Multivariate analysis was not performed given limitations in sample size and power. Results were considered significant for two-tailed P values.05. RESULTS Between 1995 and 2010, 29 patients were surgically treated for AIFR. Causes of immunosuppression in decreasing order of prevalence included hematologic malignancy, diabetes, medical immunosuppression, and 1584 human immunodeficiency virus. Approximately onequarter of the patients had more than a single cause of immunosuppression (n510, 26%) (Table I). The average patient age was years. There was a slightly greater proportion of male patients (n517, 59%) in the study population. Aspergillus and Mucor species were identified in 34% (n510) and 62% (n518) of cases, respectively. One patient was infected with both Aspergillus and Mucor. The most common presenting symptoms included facial pain, facial swelling, and visual symptoms. The surgical approach was purely endoscopic in 11 (38%) patients, with a combination of endoscopic and open approaches utilized in the remaining 18. Twelve (41%) patients died from AIFR. On univariate analysis, intracranial (P5.01) and ethmoid sinus (P5.05) involvement were significantly associated with death from AIFR. Cranial neuropathy at presentation (P5.06), orbital involvement (P5.07), treatment with amphotericin (P5.06), involvement of an increased number of anatomic subsites at presentation (P5.12), and skull base involvement (P5.13) trended toward significance. Patient age, gender, cause of immunosuppression, fungal organism type, time from symptom onset to diagnosis, treatment with a purely endoscopic approach, and the total number of surgical procedures performed were not significantly associated with death from AIFR (Table I). Intracranial involvement was significantly positively associated with involvement of the ethmoid sinuses (P5.03), sphenoid sinus (P5.02), and orbit (P5.02) and significantly negatively associated with involvement of the palate (P5.04). Involvement of the ethmoid sinus was significantly positively associated with involvement of multiple other anatomic sites, including the maxillary sinus (P5.02), sphenoid sinus (P5.03), skull base (P5.04), orbit (P5.006), and intracranial cavity (P5.03). Overall survival was 21% (6/29) at the conclusion of the study, with only five patients surviving 6 months past their diagnosis of AIFR (Fig. 1). Factors associated with overall survival on univariate analysis included cranial neuropathy at presentation (hazard ratio [HR], 3.3; 95% confidence interval [CI], ; P5.02), and intracranial involvement (HR, 4.47, 95% CI, ; P5.01). Orbital involvement (P5.06), ethmoid sinus involvement (P5.15), an increased number of subsites at presentation (P5.14), and patient gender (P5.15) trended toward significance (Table II). No patient who had orbital or intracranial involvement survived past 6 months, regardless of whether or not they cleared their invasive fungal infection. Causes of death for most of these patients were unavailable. No significant trends in the underlying causes of immunosuppression were discovered in long-term (>6 months) survivors. Causes of immunosuppression included diabetes (n52), hematologic malignancy (n52), and medication (n51). Intraoperative complications were rare. One patient experienced a cerebrospinal fluid leak that was recognized and repaired intraoperatively. Long-term complications occurred in a significant proportion of long-term survivors. Of the five patients surviving >6

3 TABLE I. Population Characteristics and Factors Associated With Death Attributable to Acute Invasive Fungal Rhinosinusitis. Patient Variables All Patients (n529) Cleared Disease (n517) Died of Disease (n512) P Patient age (yr) Male 17 (59%) 10 (59%) 7 (58%).98 Female 12 (41%) 7 (41%) 5 (42%) Cause of immunosuppression * Hematologic malignancy 16 (55%) 9 (53%) 7 (58%).77 Diabetes 12 (41%) 6 (35%) 6 (50%).32 Medical 10 (34%) 7 (41%) 3 (25%).69 HIV 1 (3%) 1 (6%) 0 (0%) 1.00 Symptom onset to diagnosis (d) Charlson comorbidity index Symptoms at presentation Visual disturbances 7 (24%) 3 (18%) 4 (33%).33 CNS symptoms 5 (17%) 2 (12%) 3 (25%).35 Cranial neuropathy 7 (24%) 2 (12%) 5 (42%).06 Organism Mucor 18 (62%) 10 (59%) 8 (67%).76 Aspergillus 10 (34%) 7 (41%) 3 (25%).46 Both 1 (3%) 0 (0%) 1 (8%) 1.00 Medical treatment Antifungal: amphotericin 24 (83%) 12 (71%) 12 (100%).06 Antifungal: voriconazole 7 (24%) 6 (35%) 1 (8%).19 Antifungal: other 13 (45%) 7 (41%) 6 (50%).64 GCSF 13 (45%) 8 (47%) 5 (42%) 1.00 Anatomic subsite involvement Maxillary sinus 17 (59%) 9 (53%) 8 (67%).44 Ethmoid sinus 19 (66%) 9 (53%) 10 (83%).05 Sphenoid sinus 7 (24%) 3 (18%) 4 (33%).38 Frontal sinus 6 (21%) 3 (18%) 3 (25%).65 Lateral nasal wall 21 (72%) 12 (71%) 9 (75%).67 Septum 13 (45%) 7 (41%) 6 (50%).70 Skull base 10 (34%) 4 (24%) 6 (50%).13 Palate/nasal floor 9 (31%) 6 (35%) 3 (25%) 1.00 Bilateral involvement 11 (38%) 6 (35%) 5 (42%).70 No. of subsites at presentation Orbital involvement 13 (45%) 5 (29%) 8 (67%).07 Intracranial involvement 7 (24%) 1 (6%) 6 (50%).01 Purely endoscopic treatment 11 (38%) 5 (29%) 6 (50%).44 Total no. of surgical procedures *Several patients had multiple causes of immunosuppression. CNS5central nervous system; GCSF5granulocyte colony-stimulating factor; HIV5human immunodeficiency virus. months, two developed frontal sinus mucoceles at 12 and 18 months postoperatively. Both patients required subsequent surgical intervention (Fig. 2). DISCUSSION Despite increased awareness of AIFR, significant morbidity and mortality continues to occur in patients who develop this infection. As illustrated in this case series, a significant proportion of patients who develop AIFR will die of their disease or of other causes within 6 months of diagnosis. Other studies have suggested that recovery of immune function is a critical factor in patients being able to clear their infection. 3 Here we demonstrated that the anatomic location of the disease may also be linked to therapeutic success. AIFR involvement of the ethmoid sinuses was associated with an increased risk of death due to AIFR compared with other paranasal sinus subsites. This may reflect a reduced ability to clear diseased tissue within this location. However, a more likely explanation is that 1585

4 amphotericin for Aspergillus AIFR was more prevalent in older cases in the series, whereas voriconazole use was more common in recent cases. This likely reflected a change in practice following the report of a randomized trial demonstrating the superiority of voriconazole as primary treatment for invasive Aspergillus infections. 10 Overall survival for patients who develop AIFR is poor. Despite 59% of patients demonstrating clearance of their infection, the median survival of this group was 3 months, with only 17% overall survival at 6 months. Other case series have described a wide range of disease-specific and overall survival outcomes, with some TABLE II. Factors Associated With Overall Survival. Patient Variables HR (95% CI) P Fig. 1. Kaplan-Meier survival analysis demonstrating a median overall survival of 3 months and a 17% survival at 6 months. [Color figure can be viewed in the online issue, which is available at wileyonlinelibrary.com.] fungal infection within the ethmoid sinuses has an increased propensity for direct spread into the anterior cranial cavity and/or orbit. The significant association that we demonstrated between ethmoid involvement and orbit and/or intracranial spread supports this explanation. Similar to other studies, involvement of anatomic areas outside of the paranasal sinuses, namely the orbit and intracranial cavity, is associated with an increased risk of death due to AIFR. However, without a larger sample size to support multivariate modeling, the independent prognostic value of any specific anatomic site involvement remains unknown. The average time from symptom onset to diagnosis in this cohort was slightly less in patients who were able to clear their infection ( days vs days, P5.47). In addition, the total number of subsites involved at presentation was lower in those able to clear their infections ( vs , P5.12). Although these factors did not reach significance, taken together they suggest that early diagnosis and subsequent intervention may improve outcomes. Interestingly, treatment with amphotericin as compared to other antifungal regimens displayed a trend toward an increase in AIFR-attributable death. This may have been due to a link between amphotericin use and invasive Mucor sinusitis, which has been associated with a worse outcome as compared to other fungal organisms in some series. 8 However, when the analysis was restricted to patients with Aspergillus AIFR, a similar trend was observed, where treatment with voriconazole as compared to amphotericin trended toward improved disease clearance (P5.16). Use of Patient age (yr) 1.00 ( ).78 Gender 0.54 ( ).15 Cause of immunosuppression Hematologic malignancy 1.68 ( ).21 Diabetes 1.03 ( ).94 Medical immunosuppression 0.82 ( ).65 HIV 0.69 ( ).71 Symptom onset to diagnosis (d) 1.01 ( ).68 Charlson Comorbidity Index 1.10 ( ).57 Symptoms at presentation Visual disturbances 1.58 ( ).34 CNS symptoms 0.92 ( ).89 Cranial neuropathy 3.2 ( ).02 Organism Mucor 1.53 ( ).32 Aspergillus 0.61 ( ).29 Both 1.29 ( ).82 Medical treatment Antifungal: amphotericin 1.70 ( ).36 Antifungal: voriconazole 0.62 ( ).35 Antifungal: other 1.72 ( ).21 GCSF or granulocyte infusion 1.0 ( ).98 Anatomic subsites involved Maxillary sinus 0.86 ( ).73 Ethmoid 2.02 ( ).15 Sphenoid 1.35 ( ).56 Frontal 1.11 ( ).85 Lateral nasal wall 0.91 ( ).85 Septum 1.48 ( ).38 Skull base ( ).71 Palate/nasal floor 0.80 ( ).63 Bilateral involvement 1.49 ( ).36 No. of subsites at presentation 1.18 ( ).14 Orbital Involvement 2.21 ( ).06 Intracranial involvement 4.47 ( ).01 Purely endoscopic treatment 0.78 ( ).55 Total no. of surgical procedures 1.09 ( ).58 CI5confidence interval; CNS5central nervous system; GCSF5granulocyte colony-stimulating factor; HIV5human immunodeficiency virus; HR5hazard ratio. 1586

5 Fig. 2. Left frontal sinus mucocele developing in a long-term survivor of acute invasive fungal rhinosinusitis. recent published series suggesting an improvement in survival over historic cohorts. 11 Few, however, have documented long-term survival in this patient population. The overall survival rate of 17% at 6 months documented in this series is consistent with the range reported by Kennedy et al., where long-term survival was noted in only three of 26 patients with AIFR, despite resolution of the infection in 10 of 26 patients. 3 For patients with orbital or intracranial extension of their disease, the prognosis seems to be particularly poor. In this series, no patient who had involvement of the orbit or intracranial cavity survived longer than 6 months. These findings highlight the importance of considering the patient s overall prognosis, particularly in patients with disease extension outside of the paranasal sinuses when extensive morbidity from surgical treatment is a likely outcome. Although data from retrospective case series such as these are by no means conclusive given the rarity of this disorder, they are likely to be the best evidence we will be able to achieve. Based on the overall poor prognosis in this subset of patients, we recommend a frank and honest discussion with the patient regarding the short- and long-term expected survival outcomes as well as the limitations of these data prior to extensive surgical therapy. The significant proportion of patients who were dead at 6 months despite apparent clearance of their infection and subsequent discharge from the hospital represents a significant lack in our current understanding of this disease process. As has been previously assumed, this finding may be due to the significant comorbid medical conditions in this patient population. However, without definitive data regarding cause of death, relapse of AIFR remains a possibility. Given this, patients who survive to be discharged from the hospital should be closely monitored and consideration given to prolonged courses of antifungal therapy. Long-term survivors appear to be at significant risk of late sinonasal complications. Two of the five patients surviving >6 months developed frontal sinus mucoceles at 12 and 18 months postoperatively. Although only a small number of long-term survivors were available for analysis, the significant proportion who experienced late sinonasal complications highlights the need for long-term follow-up in this patient population. Although others have suggested follow-up with endoscopy until remucosalization and recovery of normal sinus physiology are complete, 12 these findings suggest that the addition of sinus imaging at 6 to 12 months may be warranted. CONCLUSION Long-term survival in patients who develop invasive fungal sinusitis remains poor despite a significant proportion of patients experiencing apparent clinical resolution of their invasive fungal infection. Disease extension outside of the paranasal sinuses portends a particularly poor prognosis. Long-term survivors are at risk of late sinonasal complications and should be monitored appropriately. ACKNOWLEDGMENTS The authors thank Jess Mace, MPH, for his assistance with the statistical analysis. BIBLIOGRAPHY 1. deshazo RD, O Brien M, Chapin K, Soto-Aguilar M, Gardner L, Swain R. A new classification and diagnostic criteria for invasive fungal sinusitis. Arch Otolaryngol Head Neck Surg 1997;123: Iwen PC, Rupp ME, Hinrichs SH. Invasive mold sinusitis: 17 cases in immunocompromised patients and review of the literature. Clin Infect Dis 1997;24: Kennedy CA, Adams GL, Neglia JP, Giebink GS. Impact of surgical treatment on paranasal fungal infections in bone marrow transplant patients. Otolaryngol Head Neck Surg 1997;116: Parikh SL, Venkatraman G, DelGaudio JM. Invasive fungal sinusitis: a 15-year review from a single institution. Am J Rhinol 2004;18: Suslu AE, Ogretmenoglu O, Suslu N, Yucel OT, Onerci TM. Acute invasive fungal rhinosinusitis: our experience with 19 patients. Eur Arch Otorhinolaryngol 2009;266: Talbot GH, Huang A, Provencher M. Invasive aspergillus rhinosinusitis in patients with acute leukemia. Rev Infect Dis 1991;13: Kim DW, Heo ST, Jeon SY, et al. Invasive paranasal mucormycosis with peripheral eosinophilia in an immunocompetent patient. Med Mycol 2010;48: Ingley AP, Parikh SL, DelGaudio JM. Orbital and cranial nerve presentations and sequelae are hallmarks of invasive fungal sinusitis caused by Mucor in contrast to Aspergillus. Am J Rhinol 2008;22: Chakrabarti A, Denning DW, Ferguson BJ, et al. Fungal rhinosinusitis: a categorization and definitional schema addressing current controversies. Laryngoscope 2009;119: Herbrecht R, Denning DW, Patterson TF, et al. Voriconazole versus amphotericin B for primary therapy of invasive aspergillosis. N Engl J Med 2002;347: DelGaudio JM, Clemson LA. An early detection protocol for invasive fungal sinusitis in neutropenic patients successfully reduces extent of disease at presentation and long term morbidity. Laryngoscope 2009;119: Otto KJ, DelGaudio JM. Invasive fungal rhinosinusitis: what is the appropriate follow-up? Am J Rhinol 2006;20: