Low-grade oncocytic tumor (LOT) of the kidney is a recently discovered emerging entity that further expands the colorful diagnostic palette of the eosinophilic renal neoplasms or so-called “pink tumors” [1, 2]. However, there is increasing data supporting the distinct nature of these tumors, and some authors have already raised questions regarding the legitimacy of recognizing LOT as a separate subtype [3, 4]. LOTs are unencapsulated, well-circumscribed tumors with a pale brown gross appearance with an average diameter of 30 mm [1]. Microscopically, LOTs are composed of monomorphic tumor cells with oncocytic or eosinophilic cytoplasm, round to oval nuclei with frequent perinuclear halo formation and small but prominent (WHO/ISUP grade 1-2) nucleoli [5]. The growth pattern is variable most often solid, tubular, or trabecular. The characteristic immunophenotype distinguishing LOT from the morphologically overlapping oncocytoma (RO), as well as chromophobe renal cell carcinoma (ChRCC), is CD117 negativity with diffuse strong CK7 expression [1]. LOT lacks complete chromosomal losses or gains frequently observed in ChRCC but often demonstrates deletion of 19p13, 19q13, and 1p36 [1]. Alterations in the mammalian target of the rapamycin (mTOR) pathway are common in sporadic cases [6], as well as in the few published cases of LOT in the setting of tuberous sclerosis [7]. According to the published studies, LOT has an indolent behavior without evidence of metastatic potential [8].

To further understand the nature of this rare and conflicting entity, we present the clinicopathological, immunohistochemical, and genetic findings in 20 cases of LOT.

Chromophobe RCC and oncocytoma are traditionally considered classic eosinophilic renal neoplasms [11]. However, advances in immunoprofiling and genetic analysis have led to the identification of novel and emerging RCC subtypes, including eosinophilic, solid and cystic RCC, eosinophilic vacuolated tumor, etc [2]. In 2019, Trpkov et al. described a distinct subset of 28 renal tumors characterized by unique microscopic features, a specific immunophenotype, and indolent clinical behavior, designating them as low-grade oncocytic tumors [1]. Subsequent studies confirmed that LOT is defined by genetic alterations affecting the mTOR pathway [6, 7]. The primary differential diagnoses for LOT include chromophobe RCC and oncocytoma, though other eosinophilic RCCs, such as succinate dehydrogenase (SDH)-deficient RCC, may also be considered [12]. LOT accounts for less than 1% of all renal cell neoplasms. In a study by Kravtsov et al., LOTs constituted approximately 0.35% of all renal tumors [13]. However, within the subset of eosinophilic renal neoplasms, the incidence of LOT is notably higher, reaching around 5% [14].

Histologically, LOT exhibits a solid peripheral growth pattern with a centrally located edematous area [1]. In most cases, these central regions contain elongated tumor cells, often described as resembling “floating boats on the sea” [7]. In contrast, oncocytoma features fibrotic stroma with an island-like (archipelago-like) growth pattern [15]. Chromophobe RCC, on the other hand, is typically characterized by prominent fibrovascular septa rather than an extensive capillary network [15]. Notably, in our cohort, all LOT cases exhibited a prominent capillary network, and all but one demonstrated central edematous areas.

Cytologically, LOT is composed of uniform tumor cells with small nuclei and eosinophilic cytoplasm [1]. Perinuclear halos, previously described in the literature [7], were observed in our study in 12 tumors, with a rare and diffuse distribution in 75% and 25% of cases, respectively. While extensive perinuclear halos may suggest eosinophilic variant of chromophobe RCC, the characteristic raisinoid nuclei seen in chromophobe RCC were absent in LOT. Furthermore, we did not observe severe cytological atypia or binucleation in our LOT cases, distinguishing them from high-grade eosinophilic renal neoplasms. The absence of these nuclear features also differentiates LOT from oncocytoma, which frequently exhibits some degree of nuclear atypia [1].

Interestingly, in four cases, perinuclear halos were so prominent that they led to cytoplasmic clearing, mimicking the appearance of clear cell RCC. In some instances, this resemblance was striking, further emphasizing the necessity of careful morphological and immunohistochemical assessment in the diagnostic workup of LOT.

According to the literature, LOT exhibits a unique immunophenotypic profile, characterized by CD117 negativity and strong, diffusely positive CK7 expression [1, 2]. This immunostaining pattern effectively distinguishes LOT from oncocytoma and chromophobe RCC. Oncocytoma is diffusely positive for CD117 but typically shows only focal CK7 positivity, often limited to isolated tumor cells [16]. Notably, CK7 expression is more extensive around the central scar in oncocytoma [17]. Chromophobe RCC, by contrast, is diffusely positive for both CD117 and CK7, although in the eosinophilic variant, CK7 staining can be focal [16].

The differential expression of CD117 suggests distinct cellular origins for these tumors. CD117 positivity in oncocytoma and chromophobe RCC indicates their derivation from the intercalated cells of the collecting ducts [18]. The cellular origin of LOT was initially unclear; however, Alghamdi et al. recently demonstrated that L1 cell adhesion molecule (L1CAM) is diffusely expressed in LOT, providing insight into its histogenetic background and serving as a useful diagnostic marker [19]. GATA3 expression, with variable distribution, has also been reported in LOT [7, 20]. While chromophobe RCC may also express GATA3 [21], our findings indicate that both L1CAM and GATA3 label all LOT cases diffusely, suggesting their high sensitivity for this entity [7, 18]. However, the specificity of these markers requires further investigation in larger cohorts.

Interestingly, we identified two LOT cases with MelanA expression, an unusual finding in these neoplasms. Typically, MelanA is expressed in angiomyolipoma (AML), TFE3-rearranged RCC (TFE3-RCC), and TFEB-altered RCC (TFEB-RCC) [22, 23]. AML, a perivascular epithelioid cell tumor, may enter the differential diagnosis of eosinophilic renal tumors when it has a pure epithelioid morphology, though it lacks cytokeratin expression [22]. TFE3-RCC and TFEB-RCC, classified as molecularly defined renal carcinomas in the current WHO classification, may also express HMB45 but are negative for CK7 [23, 24]. Our findings suggest that aberrant MelanA expression can occasionally be present in LOT.

Additionally, recent studies by Salles et al. identified glycoprotein nonmetastatic B (GPNMB) as a diffusely expressed marker in mTOR-driven renal tumors, making it a promising diagnostic tool [25]. Our analysis confirmed diffuse GPNMB expression in all LOT cases, indicating active mTOR signaling. However, GPNMB is also expressed in TFE3-RCC and TFEB-RCC, limiting its specificity as an entity-defining marker [25]. Rather, its expression reflects underlying genetic alterations in these tumors.

Among eosinophilic renal tumors, SDH-RCC is a low-grade neoplasm that shares some histologic features with LOT [12]. However, key distinguishing characteristics -such as cytoplasmic vacuolization, a tubular growth pattern, and entrapped renal tubules- aid in differentiation at the light microscopic level [26]. Immunohistochemically, SDH-RCC is typically negative for CK7, CD117, and SDHB, further supporting its distinction from LOT [2, 26].

As previously noted, LOTs harbor mutations in the TSC/mTOR pathway, with MTOR and RHEB being the most frequently altered genes [7]. Mutations in TSC1, TSC2, NF2, and PIK3CA have also been reported [8]. However, it remains debated which genetic alterations alone are sufficient to activate mTOR signaling, and which represent passenger mutations. In our series, MTOR was mutated in three tumors, while, interestingly, no RHEB alterations were detected. A striking finding was the recurrent presence of ATM mutations in both the WES and panel-based sequencing cohorts. ATM encodes a serine/threonine kinase primarily responsible for detecting DNA double-strand breaks [27]. It initiates homologous recombination repair (HRR) and triggers cell cycle arrest, with downstream effects on the TSC/mTOR pathway [27]. Specifically, activated ATM stimulates LKB1, which in turn activates AMPK. AMPK subsequently phosphorylates TSC2, enhancing its inhibition of mTORC1 [28]. Therefore, ATM mutations in LOT may result in dysregulated TSC/mTOR signaling. However, immunohistochemical findings in patient 14 appeared contradictory, as the tumor showed retained LKB1 expression. This discrepancy could be explained in two ways. First, LKB1 expression does not guarantee a fully functional protein; with mutated ATM, phosphorylation of LKB1 may be insufficient to activate AMPK. Second, while a single mutation may not significantly disrupt TSC/mTOR signaling, a combination of mutations may have an additive effect, leading to pathway dysregulation and LOT development. The first hypothesis could be tested by assessing phosphorylated LKB1 levels using IHC or Western blotting. The second is supported by our findings, as ATM mutations co-occurred with other alterations in our cohort.

ATM mutations have been identified in several cancer types, including ovarian, prostate, and lung carcinomas [29]. They have also been reported in clear cell RCC, where reduced ATM expression correlates with poor prognosis [30, 31]. Loss of ATM function causes HRR deficiency, making tumors reliant on alternative DNA repair mechanisms such as base excision repair, which may offer therapeutic opportunities [32]. However, as LOTs typically follow an indolent clinical course, the utility of such therapeutic strategies in this context remains questionable.

In patient 3, we identified a VUS in the MTOR gene; a similar alteration was reported by Morini et al [5]. Further studies are required to evaluate the pathogenicity of such variants. All analyzed cases demonstrated low TMB, a finding consistent with those reported by Farcaş et al. in eosinophilic vacuolated tumors [33].

LOT tends to arise in slightly older patients compared to more common renal tumors and exhibits a mild female predominance [1, 8]. In our cohort, the patient group comprised nine females and eight males, with a mean age of 67.5 years. This was significantly higher than the mean age of patients with chromophobe RCC in our collection. Additionally, LOTs were significantly smaller in size compared to chromophobe RCCs. However, while LOTs also tended to occur in older patients and were generally smaller than oncocytomas, these differences did not reach statistical significance. Our findings align with previously published data regarding LOT’s demographic and clinicopathological characteristics.

Furthermore, in our study, no cases of tumor relapse or tumor-specific mortality were observed, reinforcing prior reports of the indolent nature of LOT [8]. Currently, the term “low-grade oncocytic tumor” is purely descriptive, referring only to the tumor’s light microscopic appearance. Given the excellent prognosis reported in published cases, an alternative nomenclature, oncocytic principal cell adenoma of the kidney, has been proposed [19]. This designation aims to better reflect both the cellular origin and the benign clinical course of these tumors.

Before concluding, we acknowledge certain limitations of our study. The whole-exome sequencing method we employed did not assess copy number alterations or non-coding (intronic) regions of the genome. In his seminal work on LOT, Trpkov et al. identified deletions in 19p13.3, 19q13.11, and 1p36.33 [1], while Alghamdi et al. reported the presence of an extra chromosome 7 in 41% of analyzed cases [19]. Copy number variations, including gene deletions or amplifications, may contribute to LOT’s tumorigenesis. Additionally, alterations within non-coding regions could impact RNA splicing, disrupt enhancers, silencers, or promoters, and modify microRNA expression or non-coding RNA binding sites. To further elucidate the role of these molecular mechanisms, future studies employing whole-genome sequencing will be necessary.

In this study, we present the clinical, pathological, and genetic characteristics of 20 cases of LOT cases. The epidemiological and pathological findings in our cohort are consistent with previously published data. Immunohistochemically, all tumors demonstrated diffuse expression of GATA3 and L1CAM, reinforcing the diagnostic utility of these markers. Notably, we identified MelanA expression in two cases, a novel finding in LOT. Additionally, GPNMB was diffusely expressed in all tumors, including one case without any detectable genetic alteration. Furthermore, we identified ATM mutations, which have not been previously reported in LOT and may represent an alternative genetic mechanism contributing to mTOR activation. The excellent clinical outcomes observed in our study further support the indolent and likely benign-nature of this tumor entity. Although, in terms of the genetic background there is some overlap with other members of other oncocytic renal tumors, overall the results of the present study support the recognition of LOT as a separate entity.