Diagnostics and prognostic evaluation in renal cell tumors: the German S3 guidelines recommendations

The German guidelines on renal cell carcinoma (RCC) have been developed at highest level of evidence based on systematic literature review. In this paper, we are presenting the current recommendations on diagnostics including preoperative imaging and imaging for stage evaluation as well as histopathological classification. The role of tumor biopsy is further discussed. In addition, different prognostic scores and the status of biomarkers in RCC are critically evaluated.


Introduction
During the last 2 decades, therapeutic options in renal cell tumors changed not only in metastatic but also in organ confirmed disease due to a broader range of systemic therapies as well as surgical techniques. In addition, more and more small renal masses are detected, and therefore, surveillance strategies and ablative therapies are under discussion. Accordingly, exact diagnosis by imaging as well as histopathology and prognostic evaluation are necessary to select the optimal treatment for each individual patient. In addition, there is an urgent need for molecular biomarkers to further increase diagnostic accuracy including non-invasive markers as well as prognostic markers to individualize treatment. In this manuscript, we present the German highest level systematic literature review-based interdisciplinary guidelines concerning imaging, histopathological classification, renal tumor biopsy, prognostic evaluation, and biomarkers for renal cell tumors (Leitlinienprogramm Onkologie (Deutsche Krebsgesellschaft, Deutsche Krebshilfe, AWMF): Diagnostik, Therapie und Nachsorge des Nierenzellkarzinoms, Langversion 2.0,2020, AWMF Registernummer: 043/017OL, https:// www. leitl inien progr amm-onkol ogie. de/ leitl inien/ Niere nzell karzi nom) [1].

Methods
The methodological approach is described in detail in the guidelines [1]. Briefly, evidence grade is based on the Scottish Intercollegiate Guidelines Network (SIGN) system. Three grades of recommendation have been used (A: strong recommendation; B: recommendation; 0: recommendation not defined). Four categories were used to define grade of consensus (strong consensus: > 95%; consensus: > 75%-95%; consensus by majority: 50-75%; dissent: < 50% agreement). If systematic search was not performed, recommendations are based on expert consensus.

Diagnostics: imaging
With the increasing number of incidentally detected renal cell carcinomas (RCC), the average size is decreasing continuously. The differential diagnosis of smaller lesions is difficult, as typical signs such as cava thrombus, necrosis, or metastasis are missing.
High-resolution imaging in CT and MRI can delineate even small and chromophobe carcinomas [2]. Staging accuracy of small RCC is similar in MRT and CT with staging accuracy between 0.78 and 0.87 [3]. CT is used routinely for small carcinomas [4], whereas tumor with caval thrombus should be staged with MRI [5] ( Table 1).
The CT scan should include an unenhanced spiral of the complete abdomen and a spiral in the early arterial phase of renal perfusion of the upper abdomen and a delayed scan of the complete abdomen. For the CT, a reconstructed slice thickness of 2 mm should be used. An enhanced scan of the thorax in a venous phase can be added for staging. Especially for planning of nephron sparing surgery, highresolution 3D reconstructions are mandatory. CT has a good accuracy in the evaluation of infiltration into the perirenal fat [6], but has a limited accuracy in the evaluation of intrarenal infiltration [7].
MRI as diagnostic modality is recommended in case of allergic reactions to iodine containing contrast media (CT) or suspected caval thrombus. The complete abdomen including the atrium and the lower pole of the kidneys should be scanned. The scans should include axial T2w images, axial enhanced T1w images and a coronal multiphase acquisition with an unenhanced scan, an early arterial phase, and a parenchymal phase. In case of urogenital bleeding a delayed scan for complete urothelium including bladder should be added. Especially, the high-resolution T2w images allow a delineation of the tumor thrombus [4].
Sensitivity and specificity of caval thrombus evaluation are 1.0 and 0.83 for MRI [5].
For grading, RCC diffusion and perfusion imaging seem to be a promising tool [8][9][10]. As imaging can hardly differentiate between histologic subtypes, despite of classic AMl, additional biopsy might be useful for therapy planning. Although prospective trials provide high level of evidence, more data are desirable to corroborate the recommendations.

Imaging for evaluation of metastasis
In patients with tumor size of 3 cm and higher, unenhanced and enhanced thin slice CT (2 mm) of the thorax should be performed, because the risk of metastasis is increasing. CT has a much higher sensitivity and specificity to detect lung metastases than conventional chest X-ray as CT allows to detect small calcifications and fat in pulmonary lesions, and can therefore differentiate the pulmonary lesions [4,[11][12][13]. For detection of abdominal lesions, MRI and CT have similar detection rates. If brain metastases are suspected, MRI should be used due to its better capability to detect metastases and edema in the brain ( Table 2).

Biopsy
In cases of uncertain renal lesions, it would be helpful to perform biopsies for histopathological evaluation, especially in patients who are candidates for active surveillance or renal tumor ablation (Table 3). Volpe et al. described a 16% reduction of surgery [14]. However, there is the possibility of false-negative results.
Under local anaesthetics, percutaneous sampling can be performed as core biopsy or fine needle aspiration, alone or in combination, US or CT-guided. Fine-needle aspiration shows lower diagnostic yield and accuracy [14,15], which can be improved by adding 18 Gauge core biopsy [16][17][18].
Systematic reviews reported a comparatively high diagnostic yield, sensitivity, and specificity for the diagnosis of malignancy (99.1% and 99.7%) when using core biopsy [14,19,20].
In tumors > 4 cm, peripheral ultrasound guided biopsies are recommended to avoid sampling of central areas with tumor necrosis [21]. An issue of renal lesion biopsy emphasizes the problem of managing negative biopsy results (0-22.6%). Repeat biopsy is diagnostic in most patients (83-100%) [22,23]. Therefore, an indeterminate or negative biopsy result but suspicious imaging findings should prompt a repeat biopsy or be interpreted as RCC if repeated biopsy is impossible.
Core biopsy of cystic renal lesions has a lower diagnostic yield and accuracy compared to solid lesions and is therefore not recommended [24].
The morbidity of percutaneous biopsy is low [14,19]. Tumor cell seeding along the needle tract is unlikely. In a recent pooled analysis, spontaneously resolving and clinically insignificant subcapsular perinephric hematoma was reported in 4.3% of biopsies [25].

Histopathological classification
The recommendations are based on the consensus conference and the most recent guidelines [26][27][28] (Table 4).
The WHO Classification of 2004 presented a comprehensive histopathological classification of RCC, which were revised in 2013 by the International Society of Urological Pathology (ISUP) in the Vancouver Classification [28]. In the consensus conference of the S3-Guidelines the diagnosis of the following new entities was recommended: Tubulocystic RCC, Acquired cystic disease-associated RCC, Clear cell papillary RCC, MiT-family translocation RCC with Xp11 translocation or t(16; 11) translocation, Hereditary leiomyomatosis, and renal cell carcinoma-associated RCC. In case of suspected brain lesions, an enhanced MR scan of the scull/brain has to be performed Strong

Germ cell tumors
Metastatic renal cell carcinoma In the consensus meeting, the classical histopathological parameters and the grading system for RCC were discussed. It was recommended that the tumor grade of RCC should be given according to the WHO-ISUP grading system. There is a clear correlation of the grade with the prognosis in clear cell and papillary RCC. Papillary RCC should be separated in two types (Type 1 with low grade and basophilic cytoplasm and Type 2 with high grade and eosinophilic cytoplasm). Papillary RCC Type 1 has an excellent prognosis. Furthermore, it was recommended that chromophobe RCC should not be graded. Other histological features were discussed. A sarcomatoid and rhabdoid differentiation should be mentioned in the histopathological report, because it is clearly associated with a poorer prognosis. The proportion of necrosis is also associated with a poorer prognosis and should be given.
The grading of chromophobe RCC has to be improved.  [29][30][31]. Therefore, this should be mentioned in every report, which is also stated in the guideline. However, prospective data are still lacking. Thus, the grading is still not recommended by the WHO.
The data for microvascular invasion in lymph or blood vessels as a poor prognostic factor are not sufficient.
A new edition of the TNM classification is available since 2017 [32].

Prognostic scores
TNM staging and grading still represent the most important characteristics for prognostic evaluation. However, these parameters are not sufficient to evaluate the individual prognosis in a given patient. Therefore, several prognostic models have been developed to predict outcome at different time points of disease and therefore select patients for different therapeutic options. These models should improve the prognostic accuracy compared to standard TNM stage and grade (Table 5).
Although some preoperative nomograms exist, majority of models have been developed for postoperative evaluation. The first aim is to evaluate the risk of progression/ metastasis and survival in local disease. The following nomograms have been created: UISS (UCLA Integrated Staging Aystem)-model [33], Karakiewicz-nomograms [34,35], SSIGN-score [36], Leibovich-score [37], Kattan-nomogram [38], Sorbellini-nomogram [39], and papillary nomogram [40]. Some of these nomograms are developed for clear cell RCCs only; others did not differentiate histological subtypes. Most of them are validated in independent patient cohorts ( Table 5). The second aim is to predict the outcome of metastatic patients treated with systemic therapy. The Motzer or MSKCC-score is the first and mostly used nomogram developed for patient cohorts treated with interferon [41]. However, it is still used in the tyrosine kinase inhibitors era, too. Additional nomograms include the IMDC or Heng-score [42], the International Kidney Cancer Working Group-Modell (IKCWG)-model [43], the Cleveland Clinic Foundation-Modell (CCF)-model [44], the French model [45], the Sunitinib model [46], and the Leibovich Score before immune therapy [47].
Currently, Motzer score and the IMDC-nomogram are most frequently used models to categorize patients in risk groups and thereof to predict outcome in clinical trials and practice.

Biomarkers
Diagnostic biomarkers could improve early detection and differential diagnosis in addition to histopathological evaluation, especially in small renal masses. In addition, there is an urgent need in prognostic biomarkers to define individual outcome of RCC patients and, therefore, to select the most appropriate therapy.
During the last years, several biomarkers on different molecular levels (DNA, RNA, and proteins) have been published that are correlated with metastasis and survival of RCC patients [48][49][50][51]. Some of these markers have been analyzed in comparison to existing clinical prognostic parameters or have been incorporated in prognostic scores, and improved prognostic accuracy [52][53][54]. Due to the lack of independent prospective validation, none of these markers has been introduced in clinical routine (Table 6).
However, it is likely that new prognostic markers will be identified from complex high-throughput analyses on The evidence to use biomarkers for prognostic evaluation is currently too low Strong several molecular levels in parallel considering clinical course [55][56][57][58]. Currently, several molecular signatures alone or integrated into clinical models have been published which are partially validated in independent cohorts. Almost all have a superior accuracy to predict individual patient outcome compared to clinically based nomograms [57,[59][60][61][62].

Informed consent Not applicable.
Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http:// creat iveco mmons. org/ licen ses/ by/4. 0/.