Fumarate Hydratase–deficient Renal Cell Carcinoma Is Strongly Correlated With Fumarate Hydratase Mutation and Hereditary Leiomyomatosis and Renal Cell Carcinoma Syndrome

Abstract

Hereditary leiomyomatosis and renal cell carcinoma syndrome–associated renal cell carcinomas (RCC) are difficult to diagnose prospectively. We used immunohistochemistry (IHC) to identify fumarate hydratase (FH)-deficient tumors (defined as FH negative, 2-succinocysteine [2SC] positive) in cases diagnosed as “unclassified RCC, high grade or with papillary pattern,” or “papillary RCC type 2,” from multiple institutions. A total of 124 tumors (from 118 patients) were evaluated by IHC for FH and 2SC. An FH deficiency was found in 24/124 (19%) cases. An indeterminate result (only 1 marker abnormal) was found in 27/124 (22%) cases. In a tissue microarray of 776 RCCs of different types, only 2 (0.5%) tumors, initially considered papillary type 2, were FH deficient. FH mutations were found in 19/21 FH-deficient tumors (with confirmed germline mutations in 9 of 9 tumors in which germline status could be assessed) and in 1/26 FH-indeterminate tumors identified by IHC. No FH mutations were found in 2/21 FH-deficient RCCs, 25/26 FH-indeterminate RCCs, and 10/10 RCCs demonstrating FH expression by IHC. Patients with FH-deficient RCC had a median age of 44 years (range, 21 to 65 y). Average tumor size was 8.2 cm (range, 0.9 to 18 cm). FH-deficient RCCs were characterized by at least focal macronucleoli and demonstrated 2 or more growth patterns in 93% cases. Papillary was the most common (74%) and dominant (59%) pattern, whereas other common patterns included: solid (44%), tubulocystic (41%), cribriform (41%), and cystic (33%). At presentation, 57% were stage ≥pT3, 52% had positive nodes, and 19% had distant metastases. After a mean follow-up of 27 months (range, 1 to 114 mo), 39% of patients were dead of disease, and 26% had disease progression. We conclude that FH and 2SC are useful IHC ancillary tools, which allow recognition of FH-deficient RCC.

Hereditary leiomyomatosis and renal cancer syndrome (HLRCC) is an autosomal dominant disorder characterized by inherited predisposition to uterine and cutaneous leiomyomas and renal cell carcinoma (RCC). It is characterized by inactivating germline mutation in the fumarate hydratase (FH) gene, which is located at 1q42.3-q43 and codes for an enzyme involved in the tricarboxylic acid cycle, which hydrates fumarate to form malate.^1–6 Although the hereditary association with multiple leiomyomas of the skin has been known for >60 years now,⁷ the first syndromic association of uterine and skin leiomyomas with renal carcinoma was reported in 2001 in 2 families from Finland.^2,5 HLRCC-associated RCC was included in the 2004 World Health Organization (WHO) classification of renal neoplasms⁸; however, it was not introduced as a distinct RCC subtype, but as a presumed hereditary counterpart of papillary RCC type. Subsequent publications highlighted the aggressive behavior of the renal carcinomas associated with HLRCC syndrome and expanded its morphologic spectrum, emphasizing the presence of orangiophilic or eosinophilic macronucleoli with perinucleolar halos (viral inclusions-like).^9,10 HLRCC syndrome–associated RCC is currently recognized as a separate entity in the 2013 International Society of Urological Pathology (ISUP) Vancouver Classification of renal tumors,¹¹ and it is included in the upcoming WHO classification 2016.

Biallelic inactivation due to FH mutations in HLRCC syndrome results in either complete loss or reduction of the FH enzymatic activity, which leads to accumulation of the intracellular levels of fumarate.^4,12 The increased level of fumarate modifies the cysteine residues in many proteins, resulting in increased protein succination and production of S-(2-succino)-cysteine (2SC), which can be used as a marker for detection of these FH-deficient tumors. Accumulation of high levels of 2SC, resulting in positive 2SC immunohistochemistry (IHC) staining, was shown to be highly specific for detection of HLRCC-associated RCC.^13,14 The loss of FH enzymatic activity, resulting in negative FH staining on IHC was also demonstrated to have a high specificity in identifying HLRCC-associated tumors.^15,16 However, a combined IHC approach to investigate the utility of both 2SC and FH antibodies in detecting previously unknown HLRCC-associated RCC has so far not been reported.

Although the primary morphologic pattern described in HLRCC-associated RCC is papillary, these tumors have been shown to demonstrate many growth patterns, often in combination, which can also present a diagnostic challenge when evaluating neoplasms with unknown clinical or familial background.^9,10,14 Because of their rarity and diagnostic difficulties in identifying these tumors, we postulated that many of them are currently underrecognized, underreported, or misclassified, and we sought to evaluate the utility of IHC for 2SC and FH in identifying these tumors, particularly in cases signed out either as “unclassified RCC, high grade,” “unclassified RCC with papillary pattern,” or “type 2 papillary RCC.” We also describe the pathologic features, FH mutational status, and the clinical features of RCCs demonstrating lack of FH expression (FH-deficient RCC), characterized by FH negative and 2SC positive staining on IHC.

MATERIALS AND METHODS

An institutional Ethics Review was obtained for the study.

Pathology Identification of Cases

We initiated an international collaboration and deliberately searched for renal neoplasms labeled in the initial sign-out as “unclassified RCC, high grade” or “unclassified RCC with papillary pattern” or type 2 papillary RCC. In particular, we searched for tumors exhibiting: (1) aggressive features, such as invasion into perirenal or sinus fat and/or showing regional metastatic disease; (2) at least focal macronucleoli; and (3) presence of different growth patterns, all of which were previously associated with HLRCC-associated RCC. Although many of the participating collaborators had large in-house and consult practices with subspecialty interest in urologic pathology, the search of the respective institutional databases was subject to varying digital archive limitations for retrospective searches. All potential cases were reviewed by 2 urologic pathologists, and a representative tissue block was retrieved for additional studies. We also included 2 previously confirmed and published HLRCC-associated RCC.^17,18 Clinicopathologic and follow-up data were collected on cases demonstrating IHC profile compatible with FH deficiency, by review of the institutional records and by contacting the consulting pathologists.

Tissue Microarray Evaluation of Papillary RCC–enriched Cohort

Using tissue microarray (TMA) methodology, a total of 776 renal neoplasms from 3 separate institutions were evaluated by IHC for FH and 2SC. TMAs were enriched for papillary RCC (381), comprising 175 papillary RCC type 1, 68 type 2, 39 mixed, and 99 papillary RCC, not specified. TMAs also included other renal tumor types: clear cell RCC (232), chromophobe RCC (21), oncocytoma (25), other RCC (39), and urothelial carcinoma (78). TMAs were constructed using 0.6 or 1 mm cores in duplicate or triplicate for each neoplasm (with built-in controls).

FH and 2SC IHC

IHC for FH and 2SC was performed in 1 laboratory on formalin-fixed paraffin-embedded (FFPE) sections. We used a commercially available primary anti-FH mouse monoclonal antibody (1 in 2000 dilution; clone J-13, cat no sc-100743; Santa Cruz Biotechnology) and an anti-2SC rabbit polyclonal antibody¹⁹ (1:2000, antibody provided by Dr Norma Frizzell) on an automated staining platform—the Leica Bond III Autostainer (Leica Biosystems, Mount Waverley, Vic., Australia). For FH, heat-induced epitope retrieval was performed for 30 minutes at 97°C in the manufacturer’s alkaline retrieval solution ER2 (VBS part no: AR9640). For 2SC, heat-induced epitope retrieval was performed for 30 minutes at 97°C in the manufacturer’s acidic retrieval solution ER1 (VBS part no.AR9961).

FH and 2SC IHC were scored independently by 2 pathologists (K.T. and A.J.G.) on whole slide sections from tumors with “unclassified, high grade” or “unclassified with papillary pattern” diagnosis, as per study design. Absent staining for FH in the neoplastic cells, in the presence of a positive internal control in blood vessels, inflammatory cells, other stromal cells, and non-neoplastic cells of the kidney parenchyma, was interpreted as true negative staining (loss or FH-deficient status). All other patterns of staining were considered positive, provided the staining was cytoplasmic and granular (that is mitochondrial). Staining for 2SC on cases evaluated on whole sections was scored as 0 if negative; 1+ if focal (<50% of cells reactive) or diffuse (>50% of cells reactive) but of weaker intensity; or 2+ if diffuse positive (>50% of cells reactive) with moderate to strong intensity. In positive 2SC cases, we also attempted to localize the reactivity (cytoplasmic, nuclear, or both). Negative staining in the adjacent normal renal parenchyma was considered an internal negative control. Because of the limited amount of available tissue in the TMA-evaluated cases, 2SC was only scored as negative or positive.

Positive IHC result was considered when both antibodies showed a pattern indicating FH deficiency (FH−, 2SC 2+); negative IHC result was when FH antibody showed retained FH expression and 2SC was not expressed (FH+, 2SC 0). Indeterminate IHC result was considered when only 1 of the markers showed aberrant expression status suggesting FH deficiency, while the other was equivocal or negative (FH+, 2SC 1+ to 2+; or FH−/+, 2SC 2+).

Molecular Evaluation of FH Mutations

All cases with available tissue that demonstrated FH deficiency (FH−, 2SC 2+) or showed indeterminate result (only 1 antibody reactive) underwent molecular evaluation of FH gene mutation status by Sanger sequencing and loss of heterozygosity studies on DNA extracted from macrodissected FFPE tissue. We also evaluated 10 cases with retained FH expression for FH mutation, as a negative control group. For Sanger sequencing previously described custom primer sets were used, and the whole coding sequence including exon-intron junctions was sequenced using primers designed to produce short amplicons suitable for degraded formalin-fixed DNA.²⁰ Loss of heterozygosity studies were performed using a previously described set of 6 polymorphic short tandem repeat markers (D1S517, D1S2785, D1S180, AFM214xe11, D1S547, and D1S2842), surrounding the FH gene.²⁰ Additional patients with suspected HLRCC were offered FH germline testing as part of their clinical care. This clinical genetic testing was performed using massively parallel sequencing for small nucleotide variants with Sanger confirmation and multiplex ligation-dependent probe amplification (MLPA) for detection of large-scale deletions.

RESULTS

Pathology Evaluation of Cases by FH and 2SC IHC

Over 3000 cases from multiple institutions were reviewed to select possible study cases that fulfilled the search criteria (spanning the years 1996 to 2014). We identified 124 tumors (from 118 patients) for additional IHC evaluation by FH and 2SC using representative whole slide sections. Most of the re-reviewed cases not included in the study remained “unclassified” after review, but additional IHC or other testing was not performed on excluded cases to further assess their classification. An IHC result indicating FH deficiency (FH−, 2SC 2+) was found in 24/124 (19%) cases (Figs. 1A–C), and 73/124 (59%) showed retained FH function by IHC (FH+, 2SC 0) (Figs. 1D–F). Indeterminate IHC profile, when only 1 of the markers was suggestive of FH deficiency (FH+ with 2SC 1+ to 2+; or FH−/+ with 2SC 2+) was found in 27/124 (22%) cases (Figs. 2A–F). All FH-negative cases (FH score 0) also demonstrated diffuse strong positive staining for 2SC (2SC 2+). In the FH-deficient cases, 2SC (2+) was diffusely and strongly positive in the cytoplasm in all cases; in only 35% of cases a distinct nuclear staining could be confirmed, whereas in the remaining cases the nuclear staining was difficult to evaluate due to strong cytoplasmic reactivity. In the cases considered indeterminate for FH expression, 2SC reactivity was restricted only to the cytoplasm.

FIGURE 1: A, An FH-deficient RCC showing a papillary growth pattern. The cells exhibit eosinophilic macronucleoli with perinucleolar clearing (inset) (patient #6). B, On IHC, neoplastic cells demonstrate FH-negative staining, whereas non-neoplastic cells show granular cytoplasmic staining, used as positive internal control. C, 2SC shows diffuse and strong staining in the neoplastic cells (2+). In cases with retained FH expression, such as in this example with papillary growth pattern (D), FH shows diffuse staining in the neoplastic cells (E), whereas 2SC is negative (F).

FIGURE 2: Some cases demonstrated indeterminate IHC profile for FH, with only 1 of the markers suggestive of FH deficiency, whereas the other was equivocal or negative. In this example showing papillary growth (A), FH demonstrated retained expression (B), whereas 2SC showed a variable staining pattern (1+) (C); no FH mutation was identified on molecular analysis. In another example with indeterminate IHC profile (D), FH showed focal expression (E), but 2SC was diffusely positive (2+) (F). On molecular analysis, FH mutation was identified in this case (patient #19).

TMA Evaluation by FH and 2SC IHC

Of the TMA evaluated cases, only 2/381 (0.5%) papillary RCCs, both initially considered type 2 (pt #2 and #17) (3% of all type 2 papillary RCCs), showed FH-deficient result (FH−, 2SC 2+) (Figs. 3A–C). All other tumor types evaluated on TMA, which included 232 clear cell RCCs, 21 chromophobe RCCs, 25 oncocytomas, 39 other RCCs, and 78 urothelial carcinomas, demonstrated retained FH expression, which was somewhat variable but typically of moderate to strong intensity, whereas the corresponding 2SC staining was considered negative in all evaluated cases (FH+, 2SC 0).

FIGURE 3: We identified 2 cases on TMA with an FH-deficient pattern, originally considered papillary RCC type 2 (A). Both cases showed prominent eosinophilic nucleoli (inset). In both cases, FH was negative (B), whereas 2SC was strongly positive (C). FH mutation was confirmed in the illustrated tumor from patient #2.

Molecular Evaluation of FH Mutations

The IHC results for FH and 2SC for FH-deficient cases and the corresponding FH mutational alterations are shown in Table 1. We performed molecular testing on 64 cases, of which 57 produced informative results. We analyzed 21/26 cases considered FH-deficient by IHC (24 identified on whole slide and 2 on TMA by IHC). We also evaluated 26/27 cases considered FH indeterminate by IHC and 10/73 cases that showed retained FH expression by IHC.

TABLE 1: FH and 2SC IHC and FH Molecular Alterations in FH-deficient RCC

Mutations were found in 19/21 FH-deficient tumors, whereas in 2 cases mutations could not be identified (1 had low DNA quality, and in neither case were large-scale deletions sought by MLPA). Nine of 19 cases underwent germline testing, and the FH mutations were confirmed germline in all tested cases, whereas the remaining 10/19 cases, which harbored FH-inactivating mutations, had only neoplastic FFPE tissue available for testing, and germline mutations could not be confirmed. Of 26 FH indeterminate tumors, only 1 case demonstrated FH mutation with IHC profile: FH−/+, 2SC 2+ (Figs. 2D–F), but no specific germline testing was performed. The remaining 25/26 demonstrated wild-type FH (IHC profile: 21 [FH+, 2SC 1+]; 3 [FH+, 2SC 2+]; and 1 [FH−, 2SC 2+]). All 10 cases with normal FH function (FH+, 2SC 0) by IHC showed wild-type FH.

Clinicopathologic Findings in FH-deficient RCC

The clinico-pathologic findings in 27 FH-deficient RCCs are shown in Table 2. They were almost twice as common in men (M:F=1.9:1), with a median patient age of 44 years (mean 44; range 21 to 65 y). Twenty-one patients were white; 1 patient each was Asian and African-American. Skin leiomyomas were documented in 3/23 (13%) patients, and 5/8 (63%) female patients had prior uterine leiomyomas; of note, FH was also negative by IHC in 2/2 female patients with tested uterine leiomyomas. Family history of either renal tumors or skin or uterine leiomyomas was elicited in 6/23 (26%) patients. Overall, an association with HLRCC syndrome was documented in 8/23 (35%) patients, on the basis of the presence of skin and uterine leiomyomas in the patients or their kindreds, familial history of syndromic features, and FH mutational alterations. There was a predilection for the left kidney (L:R=1.5:1). Solitary tumors were found in 21/23 (91%) patients, and 2 patients had bilateral tumors (1 had multiple neoplasms in both kidneys). Average tumor size was 8.2 cm (median 8.5 cm, range 0.9 to 18 cm). Stage ≥pT3 was observed in 57% of patients, and 52% had positive nodes at surgery; in 19% patients distant metastatic disease (M1) was also found at presentation. After a mean follow-up of 27 months (median 17.5, range 1 to 114 mo), 39% (9/23) patients were dead of disease, and 26% (6/23) had disease progression, with evidence of local recurrence or subsequent regional or distant metastases.

TABLE 2: Clinical Characteristics of Patients With FH-deficient RCC

The 27 FH-deficient neoplasms were characterized by variable and different architectural growth patterns, and typically 2 or more patterns were found in 93% of cases, as shown in Table 3. Although papillary pattern was most commonly present and seen in 74% of cases (in 59% as a dominant one), other common patterns were also seen: solid in 44%, (dominant in 22%), tubulocystic in 41% (dominant in 7%), cribriform in 41% (dominant in 4%), and cystic in 33% (not seen as dominant). Tubulopapillary was a dominant pattern in 1 case (4%); 1 (4%) case showed sarcomatoid differentiation as a dominant morphology. Examples of different morphologic growth patterns are illustrated in Figures 4A–F. By design, all cases demonstrated at least focal macronucleoli, which in some cases were ubiquitous. In cases with a papillary pattern, typically there was absence of foam cells in the fibrovascular cores; hyalinization of the fibrovascular cores was noted in 9 (45%) cases.

TABLE 3: Morphologic Patterns Seen in FH-deficient RCC

FIGURE 4: FH-deficient RCCs were characterized by various architectural growth patterns with 2 or more patterns present in great majority of cases. Papillary pattern was most commonly present as a dominant one (A), and hyalinization of the papillary fibrovascular cores was frequent (B). In this example, an admixed tubulocystic pattern is also seen (right) (B). Other frequent patterns included solid, in this example with foci resembling collecting duct carcinoma (C), cribriform (D), cystic, which in this example also shows intracystic papillary growth (E), and tubular, which in this example shows multiple intracytoplasmic vacuoles imparting admixed cribriform morphology (F).

DISCUSSION

In this study we demonstrated that one fifth of RCC, diagnosed either as “unclassified RCC, high grade” or “unclassified RCC with papillary pattern,” are FH deficient by IHC and were almost invariably accompanied by FH mutations. Only 0.5% of all cases diagnosed previously as papillary RCC, and 3% of those considered type 2 papillary RCC, showed FH deficiency by IHC and FH mutations, whereas all other evaluated renal tumors showed retained FH expression. The FH-deficient RCCs shared remarkable clinico-pathologic similarities with HLRCC-associated RCC, including younger age at presentation, aggressive clinical behavior, and adverse morphologic features, and were characterized predominantly by papillary architecture, typically admixed with other growth patterns, with invariable presence of at least focal macronucleoli. In fact, we were able to document an association with HLRCC syndrome in 8/23 (35%) patients. By IHC, these tumors typically demonstrated an FH-deficient profile (FH−) with aberrant succination, resulting in diffuse and strong 2SC reactivity (2+), which has previously been shown to be strongly associated with HLRCC-related RCC.^13,14 The IHC profile, along with the morphology, aggressive clinical behavior, and the presence of FH mutations, provides strong justification to consider these tumors as part of the spectrum of the HLRCC-associated RCC.

We believe that both antibodies, FH and 2SC, should be used simultaneously to enhance the IHC potential in detecting FH-deficient RCC. Combined negative staining for FH and strong positive staining for 2SC demonstrated very good sensitivity for FH-deficient RCC profile and excellent specificity. That is, a normal pattern of staining for FH and 2SC can be used to rule out FH deficiency in the great majority of renal carcinomas encountered. However, we identified 2 tumors (patient #19 and #21, second tumor) with variably retained FH expression by IHC; both tumors showed FH point mutations (2SC 2+ in both cases). As previously shown, possible missense or other in-frame FH mutations may be associated with retained FH expression, resulting from a synthesis of a stable, but inactive enzyme.^13–15,21 In addition, all cases considered “indeterminate” on IHC due to 2SC 1+ (21 cases) or 2SC 2+ (3 cases), but showing retained FH expression (FH+), exhibited wild-type FH. Therefore, restricting the evaluation to only 1 of the antibodies would be limited by the relatively lower sensitivity of FH and the lower specificity of 2SC. We also found it difficult to reliably confirm whether 2SC reactivity was nuclear, when diffuse and strong cytoplasmic reactivity was present. This is in contrast to Chen et al¹⁴ who found both cytoplasmic and nuclear 2SC to be present in HLRCC-associated RCC, allowing them to distinguish it from the “cytoplasmic only” pattern observed in a proportion of papillary RCC, type 2, and some unclassified, high-grade RCC cases. The clinical utility of the 2SC antibody is also currently limited, because it is not yet commercially available and cannot be routinely used in surgical pathology laboratories. Given the lower specificity and the difficulty in interpretation of 2SC, negative FH appears to be a more specific and comparably sensitive test at the present time.

Recent studies have also shown that IHC reactivity for 2SC²² or 2SC in combination with FH^21,23 aids in identifying FH-deficient leiomyomas in younger patients, associated with HLRCC syndrome. We have previously shown that, although the great majority of patients with HLRCC syndrome will have FH-deficient leiomyomas, 1% of all sporadic uterine leiomyomas are FH deficient usually due to somatic inactivation.²³ This is in contrast to the current study, in which germline FH mutations were identified in all patients with FH-deficient RCC with sufficient material for testing.

The RCC associated with HLRCC syndrome have been reported in about 30% of HLRCC families.^4,12 HLRCC-associated RCC are particularly difficult to manage because they are highly aggressive and present with advanced-stage and metastatic disease, resulting in death of disease in 40% to 50% patients.^9,10 Therefore, active surveillance is not recommended for the management of even small HLRCC-associated renal tumors in families with HLRCC syndrome, and wide surgical excision is recommended when any renal tumor is detected.¹⁰ RCCs associated with HLRCC syndrome are, however, quite rare and clinically challenging to diagnose in practice, because patients frequently do not exhibit the whole spectrum of the clinical presentations, and the family association is either unknown or not apparent; clinical manifestations can also differ within families.^9,10 The initial report described renal tumors in 32% of the patients, all with metastatic disease at presentation,² with a prevalence of 14% reported by the National Cancer Institute (NCI) group in a North American cohort.⁴ Kidney cancers have lower penetrance than the skin or uterine leiomyomas in the HLRCC-affected families, and they typically occur more than a decade later.^4,9,12 Therefore, patients may initially present only with skin or uterine leiomyomas or less commonly with renal cell carcinoma, and renal tumors may demonstrate a delayed presentation, or the patients may lack the other HLRCC syndromic features. In the largest cohort of 38 patients with renal tumors, reported by Merino et al⁹ from the NCI, 39% had documented skin leiomyomas, and 55% had uterine leiomyomas. The morphology remains crucial in recognizing these tumors in routine practice.

In a recent study of comprehensive molecular characterization of papillary RCC, 3.1% (5/161) of all papillary RCCs and 8.3% (5/60) of those diagnosed as “papillary RCC type 2” demonstrated germline or somatic FH mutations, which were associated with the CpG methylator phenotype.²⁴ Similar to our study, these patients were younger at presentation, and had a lower probability of overall survival than other patients with papillary RCC. A subset of these papillary RCC type 2 tumors, designated as CpG methylator phenotype associated, shared the FH-deficient profile observed in HLRCC-associated RCC, on the basis of their molecular features, allowing for more accurate characterization, which may lead to disease-specific targeted therapies.²⁴ This also highlights the fact that the differential diagnosis of FH-deficient RCC will typically include “papillary RCC type 2,” which is a relatively commonly diagnosed renal tumor. However, it is becoming apparent that “type 2 papillary RCC” may not constitute a single entity, but instead represents a pattern that may be seen in a variety of neoplasms, including, for example, Xp11 translocation RCC and collecting duct carcinoma, as acknowledged in the 2016 WHO Renal Tumor Classification. The frequent papillary morphology in combination with additional architectural patterns, and at least focal presence of macronucleoli, should be regarded as morphologic clues to undertake additional IHC testing for FH and 2SC in this setting.

Currently, there is a lack of uniformly accepted definition of HLRCC-associated RCC, which is defined not just by mutational analysis, but also clinically. The NIH definition requires that the diagnosis of HLRCC be established with the identification of a heterozygous pathogenic variant in FH in combination with multiple cutaneous leiomyomas, with at least 1 histologically confirmed leiomyoma, a single leiomyoma in the presence of a positive family history of HLRCC, and/or 1 or more tubulopapillary, collecting duct, or papillary type 2 renal tumors with or without a family history of HLRCC. For the time being, “FH-deficient RCC” may be the most appropriate nomenclature for tumors that show IHC-negative staining for FH and strong 2SC reactivity, in the setting of uncertain clinical and family history and unknown genetic status.^25,26 Taking a pragmatic approach, we would recommend that if FH-deficient RCC is diagnosed, the possibility of HLRCC should be first considered clinically. If there is a suggestive personal or family history, a presumptive diagnosis of HLRCC can be made pending confirmation with formal genetic counseling and germline mutation testing. When FH-deficient RCC is diagnosed in the absence of features suggesting syndromic disease, there are little data to indicate the risk of germline mutation. On the basis of our limited data (the finding of germline FH mutation in all 9 patients who had sufficient material available for testing), at this stage we believe the risk of germline mutation (that is HLRCC) is very high, and therefore performing genetic counseling and mutational analysis would be appropriate in all patients with FH-deficient renal carcinoma. This recommendation may be modified in the future if, similar to FH-deficient uterine leiomyoma, a low rate of germline mutation is found in follow-up studies.

The tumorigenic effect of mutated FH results in fumarate accumulation, which acts as a metabolic tumor suppressor, resulting in a metabolic shift toward aerobic glycolysis with decreased oxidative phosphorylation (so-called Warburg effect). It has been postulated that this alteration has possible downstream effects by inhibiting the hypoxia-inducible factor prolyl hydroxylase and increasing the hypoxia-inducible factor 1 alpha (HIF1alpha), which targets vascular endothelial growth factor, erythropoietin, and glucose transporter 1, and produces additional epigenetic alterations of genome-wide histone and DNA methylation, leading to increased cell proliferation and tumorigenesis.^3,26–28 On a molecular level, these changes have also been characterized by increased oxidative stress and activation of the NRF2-antioxidant response elements pathway.^24,29,30

Limitations of this study include its retrospective nature, which allowed us to confirm association with HLRCC syndrome only in about a third of patients with FH-deficient RCC. For example, in some cases we were not able to obtain a dermatological confirmation of skin leiomyomatosis, and a complete family history on specific HLRCC features was not available. Although mutational analysis was performed on FFPE neoplastic tissue in the majority of tested cases, a formal FH genetic testing was done only in a subset of cases, perhaps with a selection bias toward patients with a high likelihood of familial disease, limiting the ability to confirm germline FH mutations. In 2 FH-deficient cases by IHC, we could not confirm the presence of FH mutations; 1 of the 2 cases demonstrated low DNA quality, and additional studies to investigate for possible FH mutations in these 2 cases, for example, by MLPA to screen for large scale FH deletions, were not performed.

In summary, we found that a substantial number of cases considered either as “unclassified RCC, high grade” or “unclassified RCC with papillary pattern,” and a small percent of cases diagnosed as papillary RCC type 2, demonstrated an FH-deficient pattern (FH−, 2SC 2+) by IHC and were invariably accompanied by FH mutations at the molecular level. Although we could document an unequivocal association with HLRCC syndrome in only about a third of the patients, there is clearly a high likelihood of syndromic disease in patients presenting with FH-deficient RCC. Furthermore, even apparently sporadic FH-deficient RCCs show striking clinicopathologic similarities to unequivocally HLRCC syndrome–associated renal carcinomas, including a younger age and adverse features at presentation, aggressive clinical behavior, and frequent papillary architecture in combination with other growths patterns, with invariable presence of at least focal macronucleoli. In addition to the careful morphologic evaluation, IHC for FH and 2SC is a useful aid that allows recognition of the RCC with an FH-deficient profile with FH IHC being more specific and 2SC being more sensitive.