Incidence of Voriconazole-induced periostitis in a tertiary medical center: a single-center retrospective analysis

Abstract

Objective

Voriconazole is a commonly used antifungal medication, particularly for treating invasive pulmonary aspergillosis and as prophylaxis in hematologic malignancies. However, it can cause adverse effects, including periostitis—a painful inflammation of the periosteum. Reported incidence rates of voriconazole-induced periostitis vary widely. This study aimed to determine the incidence and characteristics of voriconazole-induced periostitis in a large cohort of patients with hematologic malignancies at a tertiary medical center, using various imaging modalities.

Methods

A retrospective cohort study was conducted at Rambam Health Care Campus, including patients treated with voriconazole for at least 90 days between September 2015 and December 2022. Patients diagnosed with invasive pulmonary aspergillosis were evaluated. Imaging studies performed post-treatment initiation were reviewed for periostitis, confirmed by senior radiologists.

Results

Among 157 patients treated for invasive pulmonary aspergillosis, 110 met the inclusion criteria. Four patients (3.6%, 95% CI 1.0% − 8.8%) demonstrated radiological signs of periostitis, most commonly detected via chest CT. Humeral involvement was present in all cases, and rib involvement occurred in 75%. Follow-up imaging did not show resolution due to ongoing treatment, patient mortality, or lack of follow-up.

Conclusion

The incidence of voriconazole-induced periostitis in this cohort was lower than previously reported. Although rare, periostitis can significantly impact patient comfort and treatment decisions.

Clinical trial

Not applicable

Introduction

Voriconazole has emerged as a key antifungal medication, extensively employed in the treatment of various fungal infections [1]. Another common indication for voriconazole is prophylaxis for patients with haematological malignancies undergoing haematopoietic stem cell transplant or intensive chemotherapy [2].

While the therapeutic role of voriconazole in antifungal management is undeniable, its use is not without adverse effects. These include neuropsychiatric and gastrointestinal toxicities, phototoxic reactions, cutaneous malignancies, hormone related effects leading to electrolyte abnormalities, peripheral neuropathies, as well as painful periostitis [1, 3, 4]. As with other adverse reactions of voriconazole, periostitis may prompt a decision to discontinue the therapy or initiate a trial of dose reduction to alleviate pain [5].

Periostitis, an inflammation of the periosteum causing pain and functional limitation, is classically associated with hypertrophic osteoarthropathy leading to painful swelling of limbs and clubbing [6]. However, it has been linked to various additional conditions, such as malignancies [6], infections [7], florid intake [8] and physiological conditions in children [9].

Initially reported by Wang et al. [10], voriconazole-induced periostitis has gained increased recognition in the literature and it is now acknowledged as a potential side effect of voriconazole treatment [3, 4]. Aside from its painful nature, the clinical significance of this side effect resides in its reversibility upon discontinuation of the drug [10–13], highlighting its diagnostic importance.

The reported incidence of voriconazole-induced periostitis ranges widely, from 15% [12] to as high as 50% [13]. These figures are derived from relatively small cohorts, and despite the postulated high incidence of this side effect, its diagnosis remains limited. Further details regarding the nature of voriconazole-induced periostitis including its characteristic time of occurrence and distribution remain to be elucidated.

Our study aims to explore this side effect and assess its incidence in a large patient population, utilizing a range of imaging modalities.

Materials and methods

The Rambam Healthcare Campus Institutional Review Board (Helsinki committee) approved this retrospective database study and waived the need for informed consent on the condition of de-identified data analysis (0614-22D-RMB).

This retrospective cohort study is based on electronic medical records of patients who were treated at the Hemato-Oncology Unit of Rambam Health Care Campus, a tertiary hospital in Israel. The study population included patients with a diagnosis of invasive pulmonary aspergillosis among the combined population of patients receiving intensive chemotherapy for acute leukemia and patients who underwent bone marrow transplantation. Patients were considered to have invasive pulmonary aspergillosis based on radiologic and laboratory findings. Patients were included if they: 1. Received voriconazole for at least 90 consecutive days, to ensure adequate exposure time for voriconazole induced periostitis to develop. 2. Underwent imaging after this cut-off. Exclusion criteria included: 1. No imaging available after at least 6 weeks of treatment. 2. Other potential causes of periostitis (bone malignancies, etc).

We extracted data from electronic medical records including demographic and clinical parameters, indication for voriconazole treatment and the treatment duration.

The prescribed treatment regimen consisted of an initial dose of 400 mg X 2 on the first day, followed by 200 mg X 2 subsequently. The treatment protocol transitioned to targeted drug monitoring (TDM) based on clinical judgement on a case-by-case basis. Administration was primary oral, with intravenous delivery for hospitalized patients unable to take oral medication.

All imaging studies regardless of modality were extracted, and for each patient, the imaging study closest to voriconazole treatment cessation was defined as the index study and evaluated for the presence of periostitis. In positive cases, imaging performed prior to the voriconazole treatment initiation was evaluated to exclude presence of periostitis before the beginning of treatment. Additionally, previous and follow-up imaging studies were evaluated to assess the first appearance of periostitis and its resolution. All imaging studies were independently reviewed by two senior radiologists (B.S. and B.R.) who were blinded to clinical data. The reviewers reached complete agreement on all negative cases. In one case of disagreement, the study was referred to a senior musculoskeletal radiologist (D.M.), whose assessment determined the final interpretation.

Descriptive statistics were used to characterize the cohort. We expressed continuous variables as means (standard deviation [SD]) and categorical variables as absolute numbers and percentages.

Results

The study cohort comprised 1239 patients who were hospitalized at Rambam Health Care Campus between September 2015 and December 2022 for intensive chemotherapy treatment targeting acute myeloid leukemia or bone marrow transplantation. Among these patients, 157 individuals were diagnosed with invasive pulmonary aspergillosis and voriconazole treatment was initiated. An additional 40 patients who received voriconazole for <90 days and 7 without adequate imaging studies were excluded. The stepwise selection process is shown in Fig. 1.

Fig. 1.

Flow diagram of patient selection. From 1239 patients receiving intensive chemotherapy or bone marrow transplantation between 2015 and 2022, 157 were diagnosed with invasive pulmonary aspergillosis and treated with voriconazole. After applying exclusion criteria, 110 patients formed the final analysis population

The final study population included 110 patients with a mean age of 56 (SD 14, 60.9% males). Demographic features, specific diagnoses and reasons for voriconazole treatment cessation are detailed in Table 1. The mean duration of voriconazole treatment was 284 days (SD = 236; range 92–1709), and the mean follow-up time for imaging studies post-initiation of voriconazole was 228 days (SD = 184; range 46–1276) (Table 2). Index imaging modalities were diverse and included chest x-ray, CT, PET-CT and MRI, with chest CT being the most frequently employed (Fig. 2).

Table 1.

Demographics of study cohort - patients under voriconazole treatment

	Overall (N = 110)
Age at treatment initiation (years)
Mean (SD)	56 (14)
Range	20 - 78
Gender
Female	43 (39.1%)
Male	67 (60.9%)
Inclusion criteria (hematologic)
Allogeneic hematopoietic cell transplantation	24 (21.8%)
Autologous hematopoietic cell transplantation	12 (10.9%)
Induction chemotherapy for acute leukemia	66 (60.0%)
Salvage treatment for relapsed or refractory acute leukemia	8 (7.3%)
Reason for voriconazole treatment cessation
End of treatment	47 (42.7%)
Death	28 (25.5%)
Clinical worsening/side effects	26 (23.6%)
Lost to follow-up	4 (3.6%)
Still receiving	4 (3.6%)
Unknown	1 (0.9%)

Table 2.

Voriconazole treatment and follow-up durations

	Overall (N = 110)
Voriconazole treatment length (days)
Mean (SD)	284 (236)
Range	92 - 1709
Imaging follow-up after treatment initiation (days)
Mean (SD)	228 (184)
Range	46 - 1276

Fig. 2.

Bar graph displaying the frequency of different imaging modalities used as index scans in the study cohort

Among the cohort, four patients (3.6%, 95% CI 1.0% − 8.8%) exhibited positive findings for periostitis. Details regarding demographics, presenting symptoms and imaging modalities of the positive cases are detailed in Table 3. Chest CT was the most common imaging modality in which positive findings were detected (3 patients, 75%). All instances of periostitis involved multiple sites, with humeral involvement observed in all cases, and rib involvement in 75% of cases. The concurrent medications of the four affected patients were reviewed to exclude the possibility of drug interactions. No drugs known to interact with voriconazole or independently cause periostitis were identified. The manifestations of periostitis varied from robust periosteal reactions (Fig. 3) to milder signs of periostitis (Fig. 4). Follow-up imaging was available for three of the four patients with periostitis. Interval improvement could not be assessed due to ongoing treatment, limited follow up, or patient mortality.

Table 3.

Characteristics of patients with voriconazole induced periostitis

	Patient 1	Patient 2	Patient 3	Patient 4
Age (years)	58	59	70	62
Sex	F	F	M	M
Total duration of voriconazole therapy (days)	1709	1004	243	336
Symptoms	N\A	N\A	Thigh soreness	N\A
Imaging Modality	CT Chest	PET CT	CT Chest	CT Chest
Periostitis Sites	Clavicle,humerous	Ribs, humerous, Lower limbs	Ribs, Humerous, Lower limbs	Clavicle,humerous Ribs
ALP levels after diagnosis	225	88	226	115
Time from Voriconazole initiation to periostitis diagnosis (days)	1276	638	233	142

Fig. 3.

70 years old male with Bilateral femoral periosteal reaction associated with voriconazole. a. Coronal and b. Axial plane

Fig. 4.

A 70-year-old male with periostitis of the left fifth rib, observed during voriconazole treatment, compared to findings from an examination conducted prior to initiating voriconazole therapy

Discussion

This study investigated the incidence and characteristics of voriconazole-induced periostitis in patients with hematologic malignancies.

Voriconazole, a fluorine-containing antifungal drug, has been implicated in periostitis development, possibly due to fluorosis [13, 14]. Wermers et al. observed elevated fluoride levels in transplant patients receiving voriconazole compared to a control group [13]. An additional mechanism independent of fluor was proposed by Allen et al., who suggested in an in vitro study that voriconazole induced periostitis is due to expression of cytokines augmenting osteoblastic activity [15].

Previous studies have reported a high incidence of voriconazole induced periostitis in patients with hematologic malignancies [16], up to 50% [13]. Nevertheless, our findings, based on a larger cohort, show a considerably lower incidence, of 3.6%. This incidence likely represents a conservative estimate, as periostitis diagnosis depends on clinical suspicion and subsequent imaging. Underdiagnosis may therefore account for the wide variability reported across centers. However, this data is consistent with our clinical experience, since voriconazole-induced periostitis is rarely diagnosed despite widespread use of voriconazole in this and other patient populations.

An important finding in our cohort was the wide variation in treatment duration, with some patients receiving voriconazole for several years. Such prolonged therapy reflects clinical decisions in complex cases of persistent immunosuppression or refractory infection, where discontinuation or switching to alternative anti-fungal agents was not feasible.

The most common clinical manifestation of periostitis is diffuse bone pain in upper and lower limbs and in the rib cage [5, 16], 17). This is consistent with our findings, which show that most periostitis cases involved the rib cage as well as the upper or lower limbs and were multifocal in nature. Based on previous literature, symptoms usually commence 6 months to 3 years after initiation of treatment with range between 6 weeks to 8 years [14].

Importantly, it has been clearly established that reduction in dose or replacement of the voriconazole with an alternative treatment resulted in improvement of symptoms. [5, 13, 16], 17). Although imaging played a key role in identifying periosteal reactions, diagnosis ultimately relies on correlating radiological findings with the clinical picture and laboratory results. The sensitivity of imaging across modalities remains variable, and further prospective evaluation is warranted.

Our study has several limitations. First, being a single center retrospective study, the findings may subject to selection bias and not be generalizable to broader patient populations. Additionally, the study cohort consists exclusively of patients with pulmonary aspergillosis, specifically those undergoing bone marrow transplantation or with hematologic malignancies, limiting the applicability of results to other patient groups or conditions. Some laboratory values which may be of interest in this clinical scenario, such as fluoride levels, were not measured routinely at are institution and therefore were not available. Similarly, voriconazole blood levels were not measured systematically and therefore could not be included in the analysis. Finally, no long term clinical or imaging follow-up was available to confirm resolution of periostitis, limiting the ability to evaluate the disease history and effect of therapy cessation.

Our findings suggest that although periostitis is uncommon, clinicians should remain alert in patients on prolonged voriconazole. Awareness of the imaging features and monitoring for unexplained bone pain can support earlier diagnosis and guide timely treatment adjustment.

Conclusion

Despite its relatively low incidence, periostitis is an underdiagnosed side effect of voriconazole. It is crucial for radiologists and clinicians alike to become familiar with the clinical and imaging characteristics of this adverse effect. Imaging assists in demonstrating periosteal changes suggestive of voriconazole toxicity, but diagnosis and treatment decisions should be based on combined clinical and laboratory assessment.

Author contribution

B.S: Conceptualization, Methodology, Data collection, Data analysis, Writing and Documentation. N.B: Conceptualization, Methodology, Data analysis, Writing and Documentation. A.S: Data collection, Data analysis. E.B: Data collection. B.R: Data collection. N.P: Data collection. D.M: Conceptualization, Methodology, Data collection, Data analysis, Writing and Documentation, Supervision. All authors reviewed the manuscript.

Funding

This research received no specific grant from any funding agency in the public, commercial, or not-for-profit sectors.

Data availability

The authors confirm that the data supporting the findings of this study are available within the article. Raw data that support the findings of the study are available from the corresponding author, upon request.

Declarations

Ethics approval and consent to participate

In accordance with the Declaration of Helsinki, the Rambam Healthcare Campus Institutional Review Board approved this retrospective database study. Approval was obtained from the local ethics committee (0614-22D-RMB)

Consent for publication

All authors consent for the publication of this article in BMC Infectious Diseases. We confirm that the content is original, has not been published elsewhere, and does not infringe upon any third-party rights. We also acknowledge that once published, the article may be available publicly in print and online formats as per the journal’s policies.

Competing interests

The authors declare no competing interests.