Mucinous tubular and spindle cell carcinoma (MTSCC) of the kidney: a detailed study of radiological, pathological and clinical outcomes

Introduction

RCC represents a compilation of several carcinoma subtypes, each with a unique genetic basis, histology, and clinical behaviour. About 15 years ago, several groups reported a novel RCC subtype characterised by low-grade cuboidal cells arranged in tubules, set within a myxoid matrix with extracellular mucin, and a variable amount of bland spindle cells [1-5]. In 2004, the WHO classification of adult renal epithelial neoplasms recognised this entity as a distinct malignant RCC tumour subtype, mucinous tubular and spindle cell carcinoma (MTSCC) [6].

MTSCC is a rare tumour, with <100 cases reported to date [7, 8]. Typically, there is a female predilection of 2:1 to 4:1 in reported series [2, 8-11]. In several series, nearly all the patients are female [12]. Patients present at a wide range of ages, with a mean age in the sixth decade [2, 8, 9, 11].

While MTSCC is described as a cytologically low-grade neoplasm, both high-grade MTSCC and MTSCC with sarcomatoid dedifferentiation have been described [13-15]. MTSCC generally has been regarded as an indolent tumour. Nonetheless, there are reports of local recurrences after resection [16], as well as both nodal and distant metastases [2, 8, 11]. The clinical behaviour and prognosis of these tumours are not yet clear. In the present study, we retrospectively reviewed our experience with MTSCC from 2002 to 2011 with the objective of further characterising the oncological outcomes of MTSCC, as well as the demographic and radiological features of the disease.

Patients and Methods

After Institutional Review Board approval, we retrospectively identified all patients who were treated at The University of Texas MD Anderson Cancer Center for MTSCC from 2002 to 2011. In all, 23 patients were identified as having MTSCC on initial chart review. All available pathology slides were re-reviewed by a single dedicated genitourinary pathologist subspecialising in renal tumours (P.T.), to confirm the diagnosis of MTSCC, as described in the WHO classification [6].

Patients were excluded if the diagnosis could not be confirmed on pathology review or if slides were unavailable for review (four patients). Demographic data were collected for the 19 included patients. Relevant radiological, operative and pathological data were reviewed. Estimated GFR (eGFR) was estimated using the Modified Diet in Renal Disease (MDRD) equation [17].

In all, 17 of the 19 patients had preoperative CT of the abdomen available for review. Two dedicated genitourinary radiologists (R.V., S.R.P.) re-reviewed all scans and recorded relevant radiological information (Table 2). Hounsfield units (HU) were measured using a small circular region of interest that was placed on the tumour at the peak area of enhancement in the corticomedullary phase avoiding areas of necrosis or calcification. This was replicated at the same site in the other available phases. Enhancement of the renal cortex was also recorded by measuring HU using a suitable region of interest in different phases of the study. Tumours were considered exophytic if >50% was considered to be outside of the expected normal contour of the kidney, partially exophytic if <50% was outside, and endophytic if completely within the confines of the renal parenchyma without any discernible bulge on the renal contour.

Operative information was recorded including complications, which were classified by Clavien–Dindo score [18]. Pathology parameters included 2010 American Joint Committee on Cancer TNM stage, tumour size, margin status, necrosis, sarcomatoid features, fat invasion, and invasion of adjacent organs. For patients not treated by nephrectomy (e.g., ablative therapy, active surveillance, or systemic therapy with primary tumour in situ) tumours were staged clinically and histological diagnosis was made by percutaneous biopsy.

Follow-up was individualised according to disease stage, treatment, and comorbid disease. Serial follow-up imaging was obtained with CT, ultrasound or MRI. Data on recurrent or progressive disease were also recorded, including time and location of recurrence, as well as most recent disease status.

Duration of follow-up was calculated from date of diagnosis to date of last follow-up or death. The Kaplan–Meier product-limit method was used to estimate median survival.

Results

Radiological Findings

Five patients had CT with all four phases: pre-contrast, corticomedullary, nephrographic and excretory phases. Five patients were scanned in the pre-contrast, corticomedullary and excretory phases. One each had pre-contrast, nephrographic andexcretory; pre-contrast and corticomedullary; and pre-contrast and nephrographic phases. Two patients were scanned in the corticomedullary phase only and one patient was scanned without i.v. contrast (pre-contrast only). The mean (range) attenuation of the tumour on the pre-contrast phase was 36 (24–48) HU (n = 14). In the corticomedullary phase, the mean (range) attenuation was 67 (41–133) HU (n = 14). The mean (range) attenuation was 89 (49–152) HU (n = 7) in the nephrographic phase and 76 (52–106) HU (n = 12) in the excretory phase. The mean attenuation of the normal renal cortex was 33, 191, 187 and 151 HU in pre-contrast, corticomedullary, nephrographic and excretory phases, respectively (Table 2).

Table 2. CT characteristics of primary tumours in 17 patients

Patient	Side	Size, cm	PC.t, HU	CM.t, HU	neph.t, HU	exc.t, HU	PC.c, HU	CM.c, HU	neph.c. HU	exc.c, HU	Shape	Growth pattern	Enhancement pattern	Haemorrhage
PC, pre-contrast; CM, corticomedullary; neph, nephrographic; exc, excretory (.t, of tumour; .c, of normal renal cortex); HU, Hounsfield units; NA, not available.
1	Left	1.5	40	57	78	78	35	220	214	161	Ovoid	Exophytic	Homogeneous	None
2	Right	4.5	48	NA	100	NA	34	NA	211	NA	Spherical	Partially exophytic	Heterogeneous	None
3	Right	2	42	75	91	88	37	275	201	224	Ovoid	Endophytic	Homogeneous	None
5	Left	2.1	36	133	NA	104	30	203	NA	197	Spherical	Endophytic	Heterogeneous	None
6	Left	17	NA	52	NA	NA	NA	186	NA	NA	Spherical	Exophytic	Heterogeneous	NA
8	Right	1.8	NA	60	NA	NA	NA	170	NA	NA	Ovoid	Partially exophytic	Homogeneous	NA
9	Right	1.3	27	64	NA	66	32	155	NA	127	Ovoid	Partially exophytic	Homogeneous	None
10	Left	6	42	52	68	77	24	210	200	179	Lobulated – ovoid	Partially exophytic	Heterogeneous	None
11	Right	15.1	NA	43	NA	55	NA	173	NA	149	Ovoid	Exophytic	Heterogeneous	NA
12	Right	4.5	33	41	83	67	35	107	163	99	Ovoid	Exophytic	Homogeneous	None
13	Left	5	39	NA	NA	NA	38	NA	NA	NA	Ovoid	Exophytic	Heterogeneous	Yes
14	Left	13	36	78	NA	NA	36	254	NA	NA	Spherical	Exophytic	Heterogeneous	None
15	Right	5.5	24	NA	49	59	18	NA	134	116	Spherical	Partially exophytic	Heterogeneous	None
16	Right	8	43	68	NA	71	37	194	NA	135	Lobulated – ovoid	Partially exophytic	Heterogeneous	None
17	Left	6	36	71	NA	90	31	245	NA	198	Ovoid	Partially exophytic	Heterogeneous	None
18	Left	7	26	92	152	106	38	150	190	105	Spherical	Exophytic	Heterogeneous	None
19	Left	14	29	58	NA	52	39	141	NA	129	Spherical	Partially exophytic	Heterogeneous	Yes

All tumours had an expansile growth pattern with a spherical or ovoid shape on CT, and had well demarcated margins with the surrounding renal parenchyma. Two tumours had a lobulated contour. The tumours were characterised as exophytic (seven), partially exophytic (eight) or completely endophytic (two). Five tumours that showed homogenous pattern of enhancement were smaller (range 1.3–4.5 cm). Most tumours that showed heterogeneous enhancement pattern were >5 cm, with an exception of one of the hypervascular tumours that was 2.1 cm in diameter. Of the 15 tumours with pre-contrast imaging, two showed areas of homogenous intermediate-to-high attenuation indicating intralesional haemorrhage (Fig. 1A–C).

Pathology

Seven patients had a biopsy of the primary tumour, including the two patients treated with RFA alone, as well as the patient who did not receive local therapy. Four of the seven biopsies were diagnosed as either MTSCC or as RCC with features suggestive of MTSCC. In the remaining three cases, the biopsies were interpreted as clear-cell RCC (one) and papillary RCC (two). After examination of the final surgical specimen, the histological diagnoses in these three cases were revised to MTSCC.

Among the 16 patients treated with extirpative surgery, most had pathological T1 (62.5%) or T2 (31.3%) disease (Table 3). pT3a disease was present in one patient (6.3%), with renal sinus fat invasion. There was evidence of tumour necrosis in seven patients (43.8%). Sarcomatoid features were present in <5% of the sampled tumour in one patient. No patient had a positive surgical margin. There was one patient with nodal metastases; this patient also had distant metastases at presentation.

From a histological standpoint, all tumours were characterised as MTSCC, with tumours composed predominantly of tubules lined by low-grade cuboidal cells with scant eosinophilic or clear cytoplasm and round uniform nuclei, set within a mucinous or myxoid stroma (Fig. 2A). The spindle cells were bland appearing, and the amount of spindle cells varied from tumour to tumour, with some tumours having abundant spindle cells, while others had very few spindle cells (Fig. 2B).

One patient had a separate ipsilateral tumour that was incidentally discovered at surgery. Both the MTSCC and the separate, incidental lesion were treated with partial nephrectomy. The second lesion was a 0.9-cm, Fuhrman nuclear grade 3, pT1aNxM0, clear-cell RCC. No tumour had mixed histology.

Discussion

To our knowledge, this contemporary, single-institution cohort is the largest published MTSCC series. Unlike RCC in general, which has a 3:2 male preponderance, most MTSCC patients are female [2, 8-11]. In our present series, the female to male ratio was 3.75:1. The cause of this gender difference is unclear and may be related to detection imbalance or to environmental, genetic, or hormonal factors. The youngest patient in our present series was aged 17 years and the oldest 71 years. This is consistent with earlier reports, in which patients presented at a broad range of ages [2, 8, 9, 11]. The median age of our population was 59 years, which is also consistent with other older series [2, 8, 9, 11]. Similar to the trend for renal masses in general, most patients in the present series had asymptomatic, incidentally discovered tumours [19].

MTSCC is a morphologically distinct type of RCC, with some resemblance to type 1 papillary RCC. In contrast to the papillary formations in papillary RCC, MTSCC have tubules lined by short cuboidal cells that are set within a myxoid or mucinous stroma. The spindle cells in MTSCC are epithelial cells with low-grade features, similar to those of the cuboidal cells lining the tubules, except the cytoplasm and nuclei in these cells are stretched out making them appear like spindle cells. The spindle cells have an epithelial phenotype, unlike the high-grade mesenchymal phenotype seen in RCC with sarcomatoid dedifferentiation. Further, sarcomatoid dedifferentiation has been reported in these tumours, where the sarcomatoid component appears morphologically similar to soft tissue sarcoma. The distinction of MTSCC from papillary RCC is dependent on the morphological features, as both these tumours have a similar immunohistochemical staining profile (positive for vimentin, cytokeratin 7, CD10 and p504S).

MTSCC was probably first reported in 1996 by Ordonez et al. [1] as a RCC with unusual differentiation. Based on the ultrastructural features the authors speculated that it arose from the cells of the loop of Henle, as opposed to papillary RCC, which is thought to arise from the proximal tubule. In 1997, MacLennan et al. [4] reported some of these tumours as ‘low-grade collecting duct carcinoma’. However, in 2005, two of these authors (MacLennan and Bostwick) in their review of this tumour reported that 5 of 13 patients (patients 2, 4, 7, 9, 10) they had reported in 1997 were in fact MTSCC [20]. These RCCs were initially referred to by various names until the WHO classification codified the term MTSCC.

MTSCC display a common appearance on imaging that is different than clear-cell RCC. Unlike clear-cell RCC, MTSCC tumours enhance to a substantially lesser degree than the corresponding cortex during the corticomedullary phase. In addition, clear-cell RCC is reported to show maximum enhancement in the corticomedullary phase. On the contrary, MTSCC tumours tend to show maximum enhancement in the nephrographic phase of the study. The attenuation of the tumours in the excretory phase was also greater than that seen in the corticomedullary phase of the study. MTSCC tumours have enhancement characteristics similar to papillary RCC. In particular, MTSCC are well demarcated, spherical or ovoid, have an expansile growth pattern, are relatively hypovascular, and small MTSCC show homogenous enhancement. One tumour showed significant increased enhancement in the corticomedullary phase. Zhu et al. [21] reported a series of 17 patients with pathology confirmed MSTCC and described these to be isodense masses with enhancement less than cortex and medulla in all phases. Our present study shows similar findings in that the tumours had lower mean attenuation in all phases compared to the cortex. Unlike our present series, Zhu et al. found MTSCC to have poorly defined margins. This difference in the reported appearance of the margin is best explained by differences in the CT protocol. The protocol used by Zhu et al. included corticomedullary and excretory phases but not the nephrographic phase, which is the most useful phase to demonstrate such a margin.

While some single-patient case reports have included information about the MRI appearance of MTSCC [22, 23], there remains a paucity of data about the MRI appearance of MTSCC. A rigorous analysis of the MRI features of MTSCC would be valuable. In the present study, we do not describe the MRI features as a most of our patients did not have MRI studies performed.

While the utility of biopsy in the management of localised renal masses is debated, accuracy of percutaneous biopsy has been increasing with time [24]. Given the rarity of the disease, there are little data about the diagnostic accuracy and utility of percutaneous biopsy in MTSCC. In the present series, seven patients had biopsy of the renal mass, four of whom went on to nephrectomy. In three cases, biopsy misclassified MTSCC as either conventional or papillary RCC and the histological diagnosis was revised after evaluation of the nephrectomy specimen. As patients in our present cohort were identified by final histological diagnosis, the rate of false-positive biopsy for MTSCC in our series is unknown.

To date, there are no rigorous data to guide therapy in this population. Patients with localised disease treated with resection had favourable outcomes in our present cohort. Despite the prevailing view, MTSCC is not a universally indolent tumour. One patient with clinically localised disease recurred after local therapy. This patient had locally advanced disease (pT2b) with necrosis and sarcomatoid features. Another patient had symptomatic metastases at presentation and died from disease. In MTSCC, a bland histological appearance may belie aggressive clinical behaviour, as metastatic MTSCC has been described even in low-grade tumours and those without sarcomatoid dedifferentiation [8].

Most clinical trials have excluded patients with non-clear-cell histology, which accounts for <20% of RCC [25]. Given its rarity even among non-clear-cell tumours, the ideal systemic therapy for metastatic MTSCC is unclear. The patients with metastatic disease in the present cohort were treated with therapies including targeted agents, immunotherapy, and cytotoxic chemotherapy in sequence and in combination.

In conclusion, in this largest series published to date, MTSCC presented at a broad range of ages and displayed a female predilection. Imaging and pathological features of MTSCC display some overlap with papillary RCC. Treatment outcomes for most patients with MTSCC with localised disease are excellent, but MTSCC is not universally indolent, with two patients in our present series having metastatic disease. Molecular and translational efforts are needed to fully characterise this rare RCC variant.

Conflicts of Interest

C.G.W. reports consulting fees and research investigator from Pfizer and Argos, outside the submitted work. J.A.K. reports consulting fees from Pfizer, outside the submitted work. All other authors have no conflicts to disclose.