Pathophysiological implication of reversed CT halo sign in invasive pulmonary mucormycosis: a rare case report

Reversed halo sign (RHS) has been regarded as a sign on computed tomography (CT) image characterized by central ground-glass opacity (GGO) which is surrounded by a partial or complete rim of consolidation. Although this sign was initially described specifically for cryptogenic organizing pneumonia [1], it has also been associated with other diseases, including fungal infections [2, 3]. However, only a few studies have conducted a detailed pathophysiological investigation of this phenomenon. Therefore, the present report describes detailed radiological and histopathological findings of a case of invasive pulmonary mucormycosis (IPM) presenting RHS, with comparison to a discrete nodule (DN) [4, 5] caused by invasive pulmonary aspergillosis (IPA), and discusses the pathophysiological implications of this characteristic image.

A 64-year-old Japanese man reported dyspnea and general fatigue two months before admission. He was referred to our hospital due to acute lymphoblastic leukemia, and chemotherapy was introduced in accordance with the CALGB 8811 protocol [6] on 54th day before the lobectomy of the left upper lobe (−54th). On -46th day, the patient exhibited a fever (38°C), and chemotherapy-induced neutropenia was confirmed with a peripheral neutrophil count (PNC) of 0 per microliter. Further, on -39th day, chest CT revealed GGO in the left upper lobe (Figure 1A). On -31st day, septum cytology revealed a small number of mold species and treatment comprising intravenous liposomal amphotericin B at 3 mg/kg/day was started. On -28th day, subsequent chest CT revealed RHS comprising both a central GGO and an outer rim recognized as a ring-shaped high-density area replacing the outside parenchyma of GGO (Figure 1B). The chest CT showed an increasing in thickness the outer rim on −20 day, and revealed RHS with air crescent-like small air slit between a part GGO and the outer rim on -11st day (Figure 1C), while serum Beta-D-Glucan levels failed to reach a normal level. The patient therefore underwent the operation. The PNC on the 54th, 46th, 44th, 39th, 31st, 28th, 20th, and 11st day before the operation was 1800, 0, 0, 340, 4400, 7735, 4623, and 2542 per microliter, respectively (Figure 2).

Histological examination of the pulmonary lesion that presented as CT-RHS, comprised coagulation necrosis of the alveolar septa and had remaining air content in the central area where had mirrored GGO (Figures 3A and 4). Periphery of the lesion corresponding to the outer rim comprised triplet structure; liquefaction, consolidation, and organization from the inner to the outer layer. There was no elastin in the liquefaction where developed at the junctional area between the coagulation necrosis and the consolidation (Figure 3B). The consolidation, the second layer of the outer rim was constituent with mononuclear cell and multinucleated giant cell filling alveolar space, but few were seen polymorphonuclear leukocytes (Figure 3C). In this layer, although elastic fibers were confirmed in each septum, they were altered in more or less. At the third layer of the outer rim, the alveolar space was replaced with collagen fibers with well maintaining of elastic fibers, which were major skeleton of alveolar septa. There were no inflammatory infiltrate in this layer. Micro-CT and that 3D reconstruction image confirmed both central GGO and peripheral ring-shaped high-density area of the IPM lesion, respectively (Figure 5).

In contrast, although DN of IPA [5] also comprised central coagulation necrosis, air content in alveolar space was entirely replaced with exudation (Figure 6A), which likely comprised plasma without leukocytes (Figure 7). The round-shaped coagulation necrosis was encompassed with oozing of red blood cell (Figure 6B). Micro-CT and that 3D reconstruction image confirmed both monotonous high density and fine reticular pattern at center and periphery of the IPA lesion, respectively (Figure 8).

It has been accepted that RHS has appeared in CT images from immunocompromised patients indicates invasive fungal infection [3], including in patients with pulmonary mucormycosis [8]. In particular, Busca A et al. described that RHS may be an early radiological sign of mucormycosis and emphasized the importance of recognizing this sign as a diagnostic strategy [8]. However, detailed pathophysiology of the sign remains obscure. Therefore, we would like to discuss how RHS develops in comparison with the histopathological findings and micro-CT images of pulmonary lesions presented IPM or IPA. For our patient, the lesion presented RHS histologically comprised central coagulation necrosis remaining air content and ring-shaped peripheral consolidation or fibrosis, which was also confirmed by micro-CT images and their 3D reconstruction. The central GGO had been clinically noted during neutropenia prior to out-ward development of the outer rim, which was coincidently confirmed at the time of recovering of PNC. This fact may suggest that the recovery of impaired bone marrow function derived neutrophils at the periphery of coagulation necrosis, which might be induced by intravascular growth of invaded Mucor. Neutrophil provision might induce thin ring-shaped liquefaction between coagulation necrosis and adjacent unaltered parenchyma because there were no elastic fibers in the liquefaction, which was digested by neutrophil function. The liquefaction, logically understood as abscess was followed by the out-ward development of mononuclear cell infiltrate replacing alveolar space which as observed as the second layer of the outer rim of CT-RHS. This finding is generally accepted as a protective response against necrosis, but this was not developed inside the ring-shaped liquefaction because coagulation necrosis had no blood flow [9]. Further protective response was noted as the third layer of the outer rim, which can be named as organization [10]. The change is histologically defined as fibrosis entirely filling the alveolar space and well-remaining elastin in septa. Accordingly, pathophysiology of CT-RHS might be understood as a kind of immune reconstitution syndrome. In the case, the sign might mirror the recovering of impaired bone marrow functioning. On the other hand, it has been reported that recovery of neutrophil function in patients with IPA result in liquefaction and this process was recognized prior to the development of a CT-air crescent sign [5]. In the present IPM case, the CT image on -11st day revealed air crescent-like small air slit between GGO and the outer rim, which might be reflected by the excretion of liquefaction layer. It suggests that RHS might be the initial and prior status of CT-air crescent sign.

The present study also shed some light on the difference in both histological and micro-CT images between IPM and IPA in patients with neutropenia. The lung tissue histologically presented DN comprised nodule of coagulation without air content, where mirrored the nodule of monotonous high density on the micro-CT image. Accordingly, a central monotonous high-density nodule and GGO may be understood as a characteristic sign of IPA and IPM in patient with neutropenia, respectively. The difference should be derived whether alveolar space is filled or not filled with exudation. However, cause of exudation filling alveolar space in case of IPA is unclear, although the difference in intensity of penetration can be suggested between Aspergillus and Mucor[5]. Namely, Aspergillus usually shows a strong intensity of penetration in the lung regardless of a difference in compliance and/or elasticity for each component. Such strong invasiveness of Aspergillus may easily induce plasma exudation. In contrast, Mucor could penetrate though septal wall from lumen of capillary or small vascular, if exudation were seen in alveoli like IPA.

Meanwhile, we wish to go into insight of the usefulness of the molecular supplemental procedures on formalin-fixed and paraffin-embedded (FFPE) tissue sections. Whereas it has been largely accepted that microscopic findings of the hyphae (diameter, presence of septa, branching angle, and pigmentation) are important to determine causative agents in foci of invasive mold infections, it might be difficult to identify causative molds only by histopathological observation even that was made with the best effort. Histological identification might have a limitation itself [11]. Therefore, the establishment of an auxiliary diagnostic method for use in routine pathological laboratories is required. Previous investigators emphasized the usefulness of molecular biology methods on FFPE tissue sections [11‐13]. Since FFPE sections are widely employed in routine preparations for surgical pathology, molecular biology methods may be utilized in many medical institutes. Furthermore, an additional diagnostic procedure will not be required if the stored paraffin-embedded tissue could be utilized for considerable molecular methods. Although further careful investigations are required, an establishment of accurate auxiliary molecular biology method could surely contribute to improve an accuracy of conventional histopathological diagnosis of this seriously debilitating infectious disease.

Our study can suggest that reversed halo sign might be considered as a form of immune reconstitution phenomena, especially induced by recovering of bone marrow functioning, and be the initial and prior lesion of the air crescent sign.