Kaposi Sarcoma with Abscessing Clinicopathologic Morphology

Background: The AIDS epidemic heralded an expanded spectrum of Kaposi sarcoma (KS) variants but to date KS masquerading as an abscess on clinical and histopathological assessment is undocumented. Methods: This 3 year retrospective study re-appraised the clinicopathological profile of all biopsies of KS that presented as abscesses clinically and demonstrated abscessing or suppurative microscopic features. Results: Ten males and 9 females form the study cohort. The clinical diagnosis in 11 patients was exclusively that of an abscess. Tuberculosis and abscess were clinical differential diagnoses in 8 patients. The average CD4 count was 261 cells/mm3. Abscessing KS was a sentinel of HIV infection and AIDS in 4 patients. Of 11 patients on HAART, 3 were virally suppressed. The pathology appeared within 3 months of HAART commencement in 5 patients; repeat CD4 counts and viral load assessments were not available for scrutiny. Microscopically, 17/19 initial biopsies with abscessing morphology demonstrated typical features of KS at least focally, while 2/19 had an exclusive abscess-like morphology. Spindle cell heterogeneity was present with co-existence in all biopsies of myofibroblasts and malignant spindle cells, the latter were HHV8-LANA-1 positive. Special stains and polymerase chain reaction investigation for infections were negative and were crucial in excluding an infective cause. Sequestrum was noted focally in the exudate of one biopsy but the patient’s demise precluded assessment of underlying osteitis. Folliculitis was not present. Conclusion: HHV8-LANA-1 immunostaining underpins the identification of KS that masquerades as an abscess clinically and may mimic a spectrum of infections microscopically. Heightened awareness of this inflammatory response is pivotal to its diagnosis. The exact cause of the abscessing morphology was unconfirmed in the present study, but on-going investigation is critical to determine the pathogenesis of this, hitherto undescribed, clinicopathological profile of KS, that may predicate altered therapeutic approaches. Citation: Naidoo N, Ramdial PK, Kuppusamy JB, Naidoo T, Pillay B (2017) Kaposi Sarcoma with Abscessing Clinicopathologic Morphology. J AIDS Clin Res 8: 726. doi: 10.4172/2155-6113.1000726

disease onset. The histopathology report provided information on gross characteristics of the submitted specimen.
Using the histopathology accession numbers, slides were retrieved from the archive of the Department of Anatomical Pathology/National Health Laboratory Service, Durban, KwaZulu-Natal for re-evaluation. Sections of biopsies that had faded were recut and stained with hematoxylin and eosin to assess the presence of chronic suppurative inflammation typical of abscess wall, morphologic subtype of KS, ulceration, sequestrum and bacterial colonies. Ziehl Neelsen, periodic acid Schiff, Warthin Starry and Brown Brenn Hopps stains were undertaken to identify potential intra-lesional mycobacteria, fungi, Bartonella species and Gram-positive and Gram-negative bacteria, respectively. Immunohistochemical immunoprofiling for HHV8-LANA-1 (Novocastra, Newcastle-Upon-Tyne , UK, clone13B10, 1:40 dilution) and endothelial cells using anti-CD34 antibody (DakoCytomation, Glostrup, Denmark, clone QBEND-10, 1:40 dilution) and anti-CD31 antibody (DakoCytomation, Glostrup, Denmark, clone JC70, 1:30 dilution) were undertaken on 2 µm thick sections cut from archival wax blocks. The tissue sections were subjected to heat-assisted microwave antigen retrieval and manual polymer-based immunohistochemical staining with diaminobenzidine as the chromogen, using a standard immunohistochemical protocol with positive and negative controls. Genomic DNA was extracted from formalin-fixed paraffin-embedded tissues using the QIAamp DNA formalin-fixed tissue (Qiagen, Hilden, Germany) according to the manufacturer's protocol. Quantified DNA quality was thereafter assessed by PCR amplification of the 159 bp product of the reference gene, β-actin [13]. PCR investigations were undertaken to detect Mycobacterium tuberculosis (MTb) and Bartonella species. Samples that were negative following agarose gel electrophoresis on the first outer PCR were then subjected to nested PCR. All PCR amplification was performed on a CFX96 Realtime PCR system (BIO-RAD, CA, USA). Bartonella species identification was accomplished by the double PCR of a 126-326 bp region of the internal transcribed spacer region of ribosomal RNA using the primers (16SrDNA F: CGTTTCTCTTTCTTCAGATGATG and 16SrDNA R: CTCTAACCAACTGAGCTACAAG). MTb complex identification was accomplished by PCR targeting of a 220 bp segment of IS6110 using the primers O1: CGGGACCACCCGCGGCAAAGCCCGCAGGAC; O2: CATCGTGGAAGCGACCCGCCAGCCCAGGAT) and a 123 bp segment using inner primers (I1: CCTGCGAGCGTAGGCGTCGG and I2: CTCGTCCAGCGCCGCTTCGG).
Permission to conduct the study was obtained from the University of KwaZulu-Natal Biomedical Research Ethics Committee (BE366/15).
The clinical diagnosis was exclusively that of an abscess in 11 patients. Tuberculosis and abscess were clinical differential diagnoses in 8 patients. KS was not a clinical consideration for any of the abscessing lesions, although disseminated cutaneous KS was noted in all patients and pleural and oral lesions in 1 and 4 patients each, respectively, on follow-up assessment.

Pathological findings
Seventeen patients had a single biopsy while 2 patients had 2 biopsies each. A total of 21 biopsies were reviewed for study purposes.

Gross details
All biopsies were of incisional wedge type and were noted to be hemorrhagic with cream and yellow necrotic foci, compatible with the clinical impression of granulating abscess wall and associated abscess contents.

Microscopic details
Of the 19 initial biopsies with abscessing morphology, 17 demonstrated typical features of KS at least focally while 2/19 (patients 5, 15) had an exclusive abscess-like morphology. The focal areas of KS in 17/19 biopsies demonstrated a solid arrangement of hyperchromatic spindle cells with intervening slit-like vascular channels ( Figure 1A), hyaline globules, hemosiderin pigment, variable mitoses and HHV8- LANA-1 ( Figure 1B), CD34 ( Figure 1C) and CD31 ( Figure 1D) immunopositivity. The abscessing morphology that dominated these 17 biopsies, noted centrally or in the deep, subcutaneous aspect of the sample, demonstrated organizing, chronic suppurative inflammation in granulation tissue, typical of an abscess wall and contents ( Figures  2A and 2B). However, the spindle cells within the granulation tissue were heterogeneous; some were normochromatic and resembled myofibroblasts while others were hyperchromatic ( Figure 2C). This prompted the conduction of HHV8-LANA-1 and α-smooth muscle actin immunostaining on these biopsies, that confirmed the divergent spindle cell population ( Figures 2D and 2E), composed of a dual population of malignant KS spindle cells ( Figure 2D) and myofibroblasts in the admixed organizing granulation tissue response ( Figure 2E). Second deeper and wider biopsies from the same tumor site in 2 patients that had exclusive abscess-like morphology in the initial biopsy confirmed typical HHV8-LANA-1 positive KS.
The sarcomatous component from patients 4, 10, 12, 13 and 18 who had HAART exposure for 2 to 3 months, also demonstrated an infiltrate of lymphoid cells, neutrophils and histiocytes ( Figures 3A and 3B). Eight biopsies demonstrated variable epidermal ulceration and nonspecific, superficial bacterial colonization of the inflammatory exudate, but there was no connection between the ulcerated component and the deeper abscessing morphology ( Figures 3C and 3D). Tuberculosis and bacillary angiomatosis were excluded in all biopsies by the lack of organisms on special infective stains. In addition PCR investigation for tuberculosis and bacillary angiomatosis were negative in all samples ( Figures 4A and 4D). Focal sequestrum was identified within the abscessing component (patient 4) but attempts to confirm connection of the abscessing tumor with underlying osteitis on clinical review and imaging studies failed because the patient died shortly after biopsy. Folliculitis was not present in any biopsy.

Discussion
Although first diagnosed in 1872, KS remained a rare neoplasm until the late 1980s when the AIDS epidemic caused its increased incidence and launched it as the most common AIDS-defining malignancy [2]. In the pre-HIV/AIDS era, the incidence of KS in Europe and North America ranged from 0.01 to 0.06 per 100,000, although an increased incidence was noted in patients post-renal transplantation and on immunosuppression [1,2]. In a transplant setting, a patient is 400-500 times more likely to develop KS within 3 to 124 months, than in the general population [14]. While classic KS is typified as a global indolent disease with a predilection for elderly Mediterranean men, African (endemic) KS affects young Black men predominantly but also manifests with lymphadenopathy in childhood [1]. Iatrogenic KS follows immunosuppressive therapy post-organ transplantation, long term corticosteroids usage and chemotherapy administration while epidemic KS is associated with AIDS [3]. Historically, KS has been associated with clinical and histopathological heterogeneity, although the latter is not unique to any specific clinical disease stage or presentation. The emergence of new KS variants in the AIDS era has not only expanded the histopathological spectrum but also introduced a plethora of mimickers and diagnostic pitfalls in its wake [12,15].
Currently, the histopathological diagnostic challenges include those related to mimicry of common well-recognized, uncommon poorly-recognized and newly-recognized KS variants, procedure and treatment-related changes and the impact of co-morbid co-lesional pathology on KS lesions [16]. Apart from heightened awareness and recognition of the changing face of KS in the AIDS era, diagnostic confirmation has also been facilitated by the relative recent discovery of Human herpesvirus-8 (HHV8), also labelled KS-associated herpes virus, as the viral etiopathogenesis of KS [17]. This has facilitated improved understanding of KS and therapeutic possibilities and distinction from diagnostic masqueraders [17][18][19][20][21][22][23]. The commercial availability of HHV-8 LANA-1 immunohistochemical antibodies and HHV8 PCR studies has enabled the diagnostic confirmation of KS in challenging histodiagnostic settings [19][20][21][22]. HHV-8 LANA-1 is a nuclear immunohistochemical antigen that is expressed by all cells that are latently infected by the virus. It is more reliable than PCR investigation as it demonstrates the nuclear immunolocalization of the viral product, in contrast to potential HHV-8 passenger lymphocytes and non-specific amplification, that may produce false positive results on PCR testing [20,21].
Because of the clinical diagnosis and the presence of granulation tissue with a neutrophil-rich component, abscess is a compelling microscopic diagnosis [24] and risky pitfall. However, in all biopsies, careful assessment revealed the presence of short spindle cell fascicles at the periphery of the suppurative component or admixed hyperchromatic spindle cells, which is not a usual feature of an abscess wall. Abscess wall histomorphology mainly that of inflamed granulation tissue is characterized by a random arrangement of myofibroblasts and vasculature. Immunophenotypically, the spindle cell component in abscess wall and in granulation tissue is of myofibroblastic lineage and will stain with myoid markers, including α-smooth muscle actin, muscle specific actin, calponin and variably with desmin [25]. In contrast, the spindle cells in KS lack a myoid immunophenotype, stain negatively with myoid immunomarkers but positively with the endothelial markers, CD31 and CD34; these endothelial markers do not stain myofibroblasts in granulation tissue [26]. Granulation tissue may assume a pseudosarcomatous histomorphology and mimic KS closely [25,26]. In this context, negative HHV8 immunostaining will distinguish KS from abscess wall.
In the context of AIDS, tuberculous and fungal infections are potential mimickers of KS with an abscessing morphology. Whilst the classic granulomatous response of tuberculosis sets it apart from KS, the anergic response in immunosuppressed patients lacks a granulomatous phenotype. The anergic tuberculous responses may be lymphohistiocytic, acellular or neutrophil-rich with abundant karyorrhexis [27][28][29][30]. Although documented more than 6 decades ago [31], the neutrophilic and karyorrhectic form of anergic TB has received more attention in the HIV and AIDS era. Granuloma-poor, karyorrhectic, mycobacteriarich anergic responses have been associated with low CD4 counts and associated impaired cell-mediated and humoral immunity [27,28]. Heightened awareness of the anergic tuberculous responses underpin the conduction of mycobacterial stains for the identification of myriad acid fast bacilli, typical of this tuberculous response. However, a negative infective profile should not default to a diagnosis of abscess because of the acute inflammatory background. Heightened recognition of the abscessing morphology of KS also underpins the conduction of HHV8 stains, pivotal to KS diagnosis. Bacillary angiomatosis, caused by Bartonella sp., is a known clinical and histopathological mimicker of KS because of the clinical morphology and microscopic vessel-rich morphology [12,16]. Bacillary angiomatosis is typically neutrophil-rich; in such settings, abscessing KS may be misdiagnosed as bacillary angiomatosis, especially as both lesions stain positively with endothelial markers. However, HHV8-LANA-1 is negative in bacillary angiomatosis [12,16,21]. Hence, the gold standard in distinguishing abscessing KS from its mimickers is the heightened The exact pathogenesis of abscessing KS remains enigmatic and speculative, but the hypothetical pathogenetic possibilities highlight, not only potential patient prognostication and management implications, but also foreground a new inflammatory paradigm of KS, that requires accelerated investigation for improved understanding. The possible pathogenetic mechanisms for the intra-tumoral neutrophilic influx are categorized into those related to local, systemic HAART-related [32][33][34][35] and systemic HAART-unrelated [36,37] mechanisms (Table 2), the latter based on responses caused by tumor-associated neutrophils (TANs) in a wide range of cancers [38][39][40][41][42][43][44][45][46][47][48][49][50]. In this study, the local mechanisms for intra-tumoral suppuration include origin of the suppurative component from acute folliculitis in hair-bearing skin, post-traumatic and foreign body reactions, entrapment of superficial acute inflammatory exudate and underlying osteitis. In the present study, only lesional tissue was submitted for appraisal. Careful appraisal of larger excision samples, including of the adjacent skin, will enable assessment of folliculitis in paralesional skin. A toe biopsy contained a splinter, albeit away from the suppurative component. The role of such traumatic insults and foreign bodies in the pathogenesis of the abscessing presentation requires ongoing appraisal. Although not confirmed in the present study, underlying osteitis, especially in destroyed toes, may be the origin of suppuration. Imaging studies and the identification of sequestrum in such biopsies will facilitate such a hypothesis. Finally, many biopsies were ulcerated. The abscessing component was not continuous with the surface, but deep dermal extension of superficial inflammatory exudate and subsequent entrapment thereof, following cicatrisation of the surface connection is a plausible mechanism, especially in older lesions. Wedge biopsies were assessed in this study. Complete appraisal of excised 'abscesses' may be more helpful in confirming the evolution of such entrapment in serial tissue sections.
HAART-associated immune reconstitution inflammatory syndrome (IRIS), also known as immune restoration disease, may be characterized by lesional neutrophilic influx [51][52][53][54]. Described in HIV-infected patients for the first time in 1992, IRIS represents an inflammatory host response to a spectrum of antigens in individuals with rapid immune recovery. Classified into unmasking and paradoxical IRIS presentations, with a temporal association with HAART use, increase in CD4 counts and decrease in viral load, the former refers to the mounting of a host response to undetected immunosuppression-related infections during immune recovery. In contrast, paradoxical IRIS occurs despite effective antimicrobial therapy because of infection deterioration caused by a host immune response to antigenic products of degenerating microbes [51]. Occurring in a wide range of infective and non-infective diseases [51][52][53][54][55][56], IRIS is documented uncommonly in KS [4,[56][57][58][59]. Major and minor diagnostic criteria have been documented for KS-associated IRIS. The former include abrupt clinical deterioration of pre-existing KS (paradoxical KS-IRIS) or a new presentation of prior unrecognized KS (unmasking KS-IRIS) following initiation or re-initiation of HAART or use of an altered more active therapeutic regime [33,57]. These reactions occur with a simultaneous decrease of at least 1 log 10 of the HIV RNA serum levels during the IRIS manifestation or 2 of 3 minor criteria. The latter encompass a 2-fold rise in the CD4+ Tcell level post HAART exposure, an increase in the HHV8 antibody response and spontaneous resolution of KS in the absence of specific chemotherapy but with continued HAART exposure [33].
The pathogenetic basis of KS-associated IRIS is an exaggerated immune response to HHV8, the infective etiological agent of KS. Whilst HAART does not target HHV8, it acts on HIV, decreasing the viral load and improving host immunity that then mounts a new or heightened reaction to HHV8 [35]. Notwithstanding the defined clinical diagnostic criteria for KS-associated IRIS, the histomorphological profile thereof, especially the phenotypic and immunophenotypic spectrum of the cellular inflammatory component in IRIS lesions, is poorly  documented. The histomorphological features that have been briefly described include either that of the typical spindle cell reaction of KS or of an intra-tumoral mononuclear inflammatory cell infiltrate, especially of CD3-positive T-lymphocytes [8,34,56,57]. While a neutrophil-rich immune response is described in cryptococal and tuberculous IRIS reactions, it is not documented in KS, to date [53,54]. In the present study, 5 patients manifested with expanding lesions within 2 to 3 months of initiation of HAART. While it is tempting to implicate an immune reconstitution basis for the abscessing morphology in these patients because of the temporal association with HAART exposure, follow-up CD4 counts and viral load assays, pivotal to the diagnosis of IRIS [33], were not traceable, to formalize a definitive hypothesis but require investigation in future studies.
This study stands apart as the 'abscessing' KS clinical presentation was also recapitulated on microscopic assessment. KS is of endothelial origin and endothelial cells are pivotal in leucocyte recruitment and regulation thereof [62] through endothelial expression of selectin receptors and synthesis of chemokines and other chemotactic agents. It is therefore surprising that TANs are not part of the protypical inflammatory response of KS [32]. Abrogation of leucocyte recruitment by lysosomal degradation of endothelial ICAM-1 and CD31 by HHV-8 has been hypothesized [63]. More recently, inhibition of neutrophilic migration by HHV8-infected endothelial cells through an interleukin-6 mechanism has also been proposed. Hence, while KS is apparently equipped with unique mechanisms that preclude the intra-tumoral localization of neutrophils [62], the present study demonstrates breakdown of this property by virtue of the dense neutrophilic accumulation. Future investigation of the pathogenesis of KS with a clinicopathologic 'abscessing' morphology is mandatory to unravel the pathogenesis and prognostic role of this hitherto undocumented presentation, and to investigate potential innovative treatment tools.

Conclusion
KS with a neutrophil-rich infiltrate dense enough to assume clinical abscessing morphology is unrecognized to date, and even in high AIDS incidence regions such as ours, remains relatively rare. It is uncertain whether this is a truly rare presentation of KS or an under-recognized entity. Notwithstanding this, optimal appraisal of all abscesses and abscess wall samples from patients with HIV infection and AIDS and heightened recognition of abscessing KS underpin the correct diagnosis of this morphological presentation. KS is an AIDS-defining condition and management and staging of patients with AIDS are dependent on its recognition. It is mandatory that the unusual and emerging variants of KS be recognized by clinicians and pathologists to ensure optimal diagnosis and patient management, especially when the KS variant is the sentinel clue to HIV infection and AIDS. The documentation of 'abscessing' KS not only expands the clinicopathological profile of KS but also highlights associated diagnostic mimicry and pathogenetic hiatuses of this enigmatic AIDS-defining malignancy, that require detailed prospective investigation.