Lymphocytic vasculitis of the prostate transition zone

Authors


Antonio Lopez-Beltran, Unit of Anatomical Pathology, Faculty of Medicine, Cordoba University, E-14004 Cordoba, Spain. e-mail: em1lobea@uco.es

Abstract

Study Type – Pathology (case series)

Level of Evidence 4

What's known on the subject? and What does the study add?

Lymphocytic vasculitis of the prostate is an exceedingly rare form of localised vasculitis that presents without systemic involvement, and is illustrated with anecdotal case reports; often as localised polyarteritis nodosa-like vasculitis. True incidence and clinical significance of lymphocytic vasculitis of the prostate in surgical specimens is virtually unknown.

The present findings support that lymphocytic vasculitis of the prostate was present in 67 (12.4%) of 540 specimens. Lymphocytic vasculitis of the prostate was present in 14 (93.3%) of 15 specimens with prostatic infarction (P < 0.001) with a risk of 124.68 (OR [odds ratio]; 95% CI [confidence interval] 16.07–967.07) as compared with BPH cases not associated with lymphocytic vasculitis.

OBJECTIVE

  • • To present our experience of lymphocytic vasculitis of the prostate in men with benign prostatic hyperplasia (BPH) without systemic involvement, as this is an exceedingly rare form of localised vasculitis and the incidence in surgical specimens and clinical significance of lymphocytic vasculitis is virtually unknown.

PATIENTS AND METHODS

  • • A sequential cohort series of 540 surgical specimens removed because of BPH-related symptoms, including simple prostatectomy (374 men) and transurethral resection of the prostate (166), comprised the study group.
  • • All men had histological diagnosis of BPH and received surgical therapy only. None of the men had had previous surgery or granulomatous prostatitis.
  • • The mean (range) age at diagnosis was 67.8 (38–89) years.

RESULTS

  • • Lymphocytic vasculitis of the prostate was present in 67 (12.4 %) of 540 specimens. It was seen in a variable number of small- to medium-sized parenchyma arteries with segmental to transmural lymphocytic inflammation, within the morphological spectrum of a polyarteritis nodosa (PAN)-like lesion seen at the periphery of BPH nodules.
  • • In four cases, focal fibrinoid necrosis was seen in vessels with otherwise typical lymphocytic vasculitis features. Immunohistochemical staining showed a T cell predominant polymorphic cellular infiltrate with a minor component of B cells and monocytes. Six cases additionally had eosinophils (<1% of inflammatory cells).
  • • Lymphocytic vasculitis of the prostate was present in 14 (93.3%) of 15 specimens with prostatic infarction (P < 0.001) with a risk of 124.68 (odds ratio [OR]; 95% confidence interval [CI] 16.07–967.07) as compared with BPH cases not associated with lymphocytic vasculitis. Logistic regression multivariate analysis selected both lymphocytic vasculitis of the prostate and patient age as significant predictors of prostate infarction with lymphocytic vasculitis being the most significant (P < 0.001; OR 128.12; 95% CI 16.298–1007.202). Follow-up information was available in all cases, range 2–16 years, and none of the patients developed systemic disease.
  • • A validation set of 1665 additional cases including radical prostatectomy, cystoprostatectomy, and needle biopsies showed lymphocytic vasculitis of the prostate being associated to prostate infarction on univariate and multivariate logistic regression (P < 0.001; OR 228.34; 95% CI 45.17–1154.22) analyses.

CONCLUSIONS

  • • Lymphocytic vasculitis in men with BPH is associated with prostatic infarction and should be considered a form of localised vasculitis with PAN-like morphology that does not necessitate additional evaluation for systemic disease.
  • • The potential clinical relevance of lymphocytic vasculitis warrants further investigation.
Abbreviations
PAN

polyarteritis nodosa

OR

odds ratio

INTRODUCTION

Lymphocytic vasculitis of the prostate is an exceedingly rare form of localised vasculitis of the prostate that presents without systemic involvement, and is illustrated with anecdotal case reports; often as localised polyarteritis nodosa (PAN)-like vasculitis [1–6]. True incidence and clinical significance of lymphocytic vasculitis of the prostate in surgical specimens is virtually unknown. The morphology, classification, and diagnostic features of lymphocytic vasculitis remain unclear.

Inflammation of vessels in the prostate can represent a vasculitis localised to or isolated in the prostate or can represent a manifestation of systemic or nonprostatic-confined vasculitis [1–10]. As noted in earlier reports, prostate vasculitis has been found in Wegener granulomatosis, allergic (eosinophilic) granulomatous prostatitis, vasculitis in post-TURP granuloma, and in systemic PAN [2,11–15], as well as in an isolated form of giant cell arteritis limited to the prostate [12]. Also, irradiation can produce a prostatic vasculitis [11]. In his book of prostate pathology, Humprey [4] presented an image of lymphocytic vasculitis of the prostate seen in a man with BPH and recommended that atherosclerotic changes, which are common in periprostatic vessels of older men, should not be classified as prostatic vasculitis.

The clinical significance of lymphocytic vasculitis of the prostate in men with BPH remains unknown, but it is recommended that findings of fibrinoid, necrotising, or giant cell granulomatous vasculitis should prompt a clinical search for a systemic disease [1–12,16,17].

Herein, we present a large series of 67 cases of lymphocytic vasculitis of the prostate diagnosed in surgical samples from 540 consecutive men with BPH; emphasis is placed in the diagnostic features and the potential clinical significance as a predictor of prostatic infarction. Incidence and significance of lymphocytic vasculitis in a validation set of 1665 additional prostate samples is also reported.

PATIENTS AND METHODS

This study included consecutive TURP and simple open prostatectomy specimens from 540 patients that were obtained from the files of Cordoba General Hospital. All patients had a histological diagnosis of BPH and received surgical therapy only.

The cases spanned a period of 8 years with the earliest case diagnosed in 1980 and the latest one in 1987. Using this historical series allowed us to include cases with surgical treatment only. Clinical information was obtained from the patients' records and a mean (sd, range) of 3.3 (1.4, 1–15) haematoxylin and eosin slides from each case were evaluated by one uropathologists (A.L.B.). In cases of uncertainty, two additional uropathologists reviewed the case reaching a final agreement (L.C., R.M.). The following pathological parameters were evaluated: presence of PAN-like lymphocytic vasculitis with or without fibrinoid changes, BPH with dominant stromal nodules, prostate infarction, type of sample (TURP vs simple prostatectomy), and number of paraffin blocks per case.

Lymphocytic vasculitis of the prostate was defined by the presence of streaking segmental to transmural lymphocytic inflammation within the morphological spectrum of a PAN-like lesion seen in a variable number of small- to medium-sized parenchyma arteries at the periphery of BPH nodules. None of the cases with lymphocytic vasculitis had had previous surgery or granulomatous prostatitis.

Immunostaining was performed on at least one representative paraffin section using standard protocols with a three-step biotin-streptavidin procedure. The antibodies used on paraffin-embedded tissues included CD45, CD3, CD43, CD5, CD34, CD8, CD79a, CD20, CD68 (all antibodies prediluted; Dako, Glostrup, Denmark), smooth muscle actin (1:200, Dako), and desmin (1:200, Dako). For all analyses, appropriated negative and positive controls were included. When necessary, antigen retrieval was performed following standard protocols. Immunostaining was graded as the percentage of positive cells. One dedicated pathologists (A.V.) evaluated all immunohistochemical slides in a ‘blinded’ manner without knowledge of clinical information. All markers were quantitated by using random fields measuring 62 500 µm2 delineated by using a 1-cm2 graded ocular grid attached to the eyepiece of the microscope.

To assess the diagnostic performance and significance in a contemporary series, data from 1665 specimens prospectively evaluated between 2001 and 2005 were retrieved from a different hospital (Reina Sofia Hospital, Cordoba, Spain) in which one of the authors (A.L.B.) has followed lymphocytic vasculitis in the prostate since the early 1990s. The series included samples from radical prostatectomy, cystoprostatectomy, and needle prostate biopsies, in addition to TURP and simple prostatectomy specimens.

RESULTS

Clinicopathological features of the 540 patients are summarized in Table 1. Surgical samples including simple prostatectomy (374) and TURP (166) made up the study group.

Table 1.  Clinicopathological features of lymphocytic vasculitis seen in 540 consecutive cases of benign prostate hyperplasia
VariableOverallLymphocytic vasculitis
AbsentPresent
N (%)540473 (87.6)67 (12.4)
Mean (sd):   
 Age, years 67.6 (7.1)71.7 (6.1)
Paraffin blocks per case 3.3 (1.4)4.1 (1.2)
N (%):   
 Type of sample:   
  TURP166 (30.7)157 (94.6)9 (5.4)
  Simple prostatectomy374 (69.3)316 (84.5)58 (15.5)
 BPH with dominant stromal nodules:   
  Absent492 (91.1)438 (89.1)54 (10.9)
  Present48 (8.9)35 (72.9)13 (27.1)
 Prostate infarction:   
  Absent525 (97.2)472 (89.9)53 (10.1)
  Present15 (2.8)1 (6.6)14 (93.4)

In all 540 men there were obstructive symptoms without or with urinary retention; the mean (range) age was 67.8 (38–89) years. The initial diagnosis of lymphocytic vasculitis was made on TURP in nine cases and simple prostatectomy in 58 others (Table 1). Follow-up data after surgery was available in all 540 cases, at a mean (median, range) of 7.9 (6, 2–16) years; none of the patients with lymphocytic vasculitis developed systemic disease.

At histology, lymphocytic vasculitis of the prostate was seen in a variable number of small- to medium-sized parenchyma arteries with segmental to transmural lymphocytic inflammation, within the morphological spectrum of a PAN-like lesion seen at the periphery of BPH nodules (Fig. 1). In four cases, focal fibrinoid necrosis was seen in rare vascular structures (<1% vessels) showing otherwise typical features of lymphocytic vasculitis in the rest of affected vessels. Arteries displayed proliferation of the intima in an ‘onion-skin’ pattern, with severe narrowing of the lumina and swollen endothelial cells. Other blood vessels showed proliferation of the intima without fibrinoid necrosis while others remained intact. The prostate showed hyperplastic changes and scarce foci of lymphocytic infiltrate without eosinophils. No granulomas were present. Prostate infarction was present in 15 cases (Table 1) all of them clinically presented with acute urinary retention.

Figure 1.

Prostate infarction in simple prostatectomy specimen with BPH (A). BPH nodule with lymphocytic vasculitis (upper left corner) seen at the periphery of stromal nodule and comparison with unaffected vessels (lower right corner) (B). Transmural lymphocytic vasculitis (C). Segmental lymphocytic vasculitis (D). (B, C and D, haematoxylin and eosin stain).

Lymphocytic vasculitis of the prostate was present in 67 (12.4%) of 540 specimens and was observed in 14 (93.3%) of 15 specimens with prostatic infarction (P < 0.001) with a risk of 124.68 (odds ratio [OR]; 95% CI 16.07–967.07) as compared with BPH cases not associated with lymphocytic vasculitis (Table 2); the risk increased in patients with lymphocytic vasculitis diagnosed in simple prostatectomy specimens (Table 2). Logistic regression multivariate analysis selected both lymphocytic vasculitis of the prostate and patient age as significant predictors of prostate infarction with lymphocytic vasculitis being the most significant (P < 0.001; OR 128.12; 95% CI 16.298–1007.202; Table 3). Receiver operating characteristic curves for lymphocytic vasculitis of the prostate had an area under the curve of 0.94 (P < 0.001; 95% CI 0.886–0.993) as predictor of prostatic infarction (Fig. 2).

Table 2.  Lymphocytic vasculitis of the prostate and risk of prostate infarction (univariate analysis)
Variable(N= 540)Prostate infarctionLymphocytic vasculitis, n (%) P * OR (95% CI)
AbsentPresent
  • *

    Fisher test.

Prostate infarction:<0.001124.68 (16.07–967.07)
 no472 (89.9)53 (10.1)  
 yes1 (6.6)14 (93.4)  
BPH with dominant stromal nodules:
 no   <0.001111.79 (14.09–886.76)
 Absent437 (91.0)43 (9.0)  
 Present1 (8.3)11 (91.7)  
 yes   0.01723.67 (1.10–495.80)
 Absent35 (77.8)10 (22.2)  
 Present– (0.0)3 (100.0)  
Type of sample:
 TURP0.1119.50 (1.12–340.98)
 Absent156 (95.1)8 (4.9)  
 Present1 (50.0)1 (50.0)  
 Simple prostatectomy<0.001187.81 (10.98–3213.70)
 Absent316 (87.5)45 (12.5)  
 Present– (0.0)13 (100.0)  
Age, years:
 <70<0.00198.65 (5.30–1826.10)
 Absent282 (90.1)31 (9.9)  
 Present– (0.0)5 (100.0)  
 ≥70<0.00177.73 (9.40–642.79)
 Absent190 (89.6)22 (10.4)  
 Present1 (10.0)9 (90.0)  
Table 3.  Logistic regression multivariate analysis selected both lymphocytic vasculitis of the prostate and patient age as risk predictors of prostate infarction
  P OR (95% CI)
Lymphocytic vasculitis<0.001128.123 (16.29–1007.20)
Age0.0461.103 (1.00–1.21)
Figure 2.

Receiver operating characteristic curves for lymphocytic vasculitis of the prostate had an area under the curve of 0.94 (P < 0.001; 95% CI 0.886–0.993) as a predictor of prostatic infarction.

Immunohistochemical staining showed that the lymphoid component was positive for CD45, CD3, CD5, CD43, CD8, CD20, CD79a, CD68, and CD34 showing a T cell predominant polymorphic cellular infiltrate with a minor component of B cells and rare monocytes (Table 4). Six cases additionally had eosinophils (<1% of inflammatory cells) as part of the inflammatory component. Smooth muscle actin and desmin showed a sclerotic wall with a reduction in smooth muscle cells and some capillary proliferation at the vascular wall as highlighted with CD34 (Fig. 3).

Table 4.  Immunohistochemical findings in inflammatory component of lymphocytic vasculitis of the prostate
Marker*% stained cells
  • *

    Other immunohistochemical markers performed in the context of smooth muscle staining included desmin and smooth muscle-specific actin and both were positive in smooth muscle at the vessel wall. CD34 was additionally positive in endothelial cells.

CD4570–90
CD370–90
CD55–20
CD205–20
CD4370–90
CD830–50
CD68<5
CD79a5–30
CD34<5
Figure 3.

Lymphocytic vasculitis in BPH showing healing lesions with ongoing inflammation and intimal proliferation in transmural (A) or segmental (B) vasculitis (A and B, elastic stain). Prominent T cell infiltrate (anti-CD3 immunohistochemistry) (C) and minor component of B cells (anti-CD20 immunohistochemistry) (D). Reduction in smooth muscle cells (anti-smooth muscle actin immunohistochemistry) (E) and the resulting sclerotic wall with some capillary proliferation (anti-CD34 immunohistochemistry) (F).

As seen in Table 5, lymphocytic vasculitis of the prostate was present in 29 (1.9%) of the 1665 specimens and was observed in seven of the nine specimens with prostatic infarction (P < 0.001), with a risk of 228.34 (OR); the risk increased in men with lymphocytic vasculitis diagnosed at standard BPH specimens and in those aged <70 years (Table 5). Logistic regression multivariate analysis selected lymphocytic vasculitis of the prostate as the independent predictor of prostate infarction.

Table 5.  Lymphocytic vasculitis of the prostate and risk of prostate infarction (univariate and multivariate logistic regression analysis) observed in a validation set of 1665 additional prostate samples
Variable (N= 1665)Prostate infarctionLymphocytic vasculitis, n (%) P OR (95% CI)
AbsentPresent
  1. *Fisher test; **chi-square; †Logistic regression multivariate analysis selected lymphocytic vasculitis as predictor of prostate infarction (P < 0.001).

Prostate infarction:<0.001*228.34 (45.17–1154.22)
 No1631 (98.5)25 (1.5)  
 Yes2 (22.2)7 (77.8)  
BPH with dominant stromal nodules:
 No   <0.001*257.25 (50.56–1308.78)
 Absent1617 (98.7)22 (1.3)  
 Present2 (22.2)7 (77.8)  
 Yes   
 Absent14 (82.4)3 (17.6)  
 Present  
Type of sample:
 TURP<0.001*97.10 (4.80–1943.40)
 Absent102 (91.9)9 (8.1)  
 Present4 (100.0)  
 Simple prostatectomy0.004*46.40 (2.10–1026.90)
 Absent36 (87.8)5 (12.2)  
 Present3 (100.0)  
 Radical prostatectomy
 Absent98 (98.0)2 (2.0)  
 Present  
 Cystoprostatectomy
 Absent26 (100.0)  
 Present  
 Needle biopsy0.90**28.80 (1.30–641.70)
 Absent1369 (99.3)9 (0.7)  
 Present2 (100.0)  
 Age, years     
  <70   <0.001*306.30 (14.10–6648.20)
 Absent949 (98.4)15 (1.6)  
 Present2 (100.0)  
  ≥70   <0.001*170.50 (29.50–985.70)
 Absent682 (98.6)10 (1.4)  
 Present2 (28.6)5 (71.4)  

DISCUSSION

Localised or isolated vasculitis involving the prostate is rare with most cases reported as vasculitis in post-TURP granulomatous prostatitis, giant cell arteritis, isolated fibrinoid arteritis in BPH, isolated non-fibrinoid arteritis in BPH, or PAN-like lymphocytic vasculitis of the prostate [1–12,16,17].

The existence of PAN-like lesions confined to one organ has been reported in the appendix, gallbladder, uterus, testes, epidydimis, and breasts, with few examples in the prostate associated with granulomatous prostatitis [14,18–20]. Two cases of isolated vasculitis nodosa-like unrelated to granulomatous prostatitis and four additional cases showing fibrinoid necrosis have been reported previously [5,6,9,17,19,20].

Systemic vasculitis affecting the prostate is most commonly seen as a manifestation of Wegener granulomatosis or PAN [14,18–20]. Autopsy studies have shown a minor incidence of genitourinary organs involvement by PAN, involvement that is infrequently diagnosed before death, as <20% of these patients show signs or symptoms; only one previously reported case had symptomatic prostatic vasculitis as the presenting manifestation of subsequently proven PAN [1–22]. In addition, some types of vasculitis, indistinguishable from those seen in the active phases of PAN, with acute and chronic inflammation and fibrinoid necrosis of the vessel wall often affecting only a localised segment of the wall, are now well recognised as prostatic lymphocytic vasculitis [1–10], as seen in the present series. Interestingly, these cases have unremarkable outcome without any manifestation of systemic disease. Likewise, it has been suggested that pathologists should be aware that prostatic vasculitis, although rare, may be the first manifestation of PAN in some cases [1–10].

The present study is the first to report on a large series of lymphocytic vasculitis found in a sequential series of 540 men with BPH without any resulting systemic disease thus supporting that histological lesions of lymphocytic vasculitis are not necessarily the first manifestation of a systemic disease; furthermore, the finding of an incidental isolated vasculitis in BPH specimens is most probably unrelated to a systemic disease and when diagnosed following the criteria outlined in the present study, these are benign findings. This is important as a diagnosis of systemic PAN implicates poor prognosis and need for aggressive therapy [19]. All of the present lymphocytic vasculitis cases were isolated localised examples of prostatic vasculitis that were an incidental finding in an otherwise typical BPH sample with unremarkable outcome. There were similar results on follow-up for cases diagnosed on a daily bases in our validation cohort of 1665 prostate samples.

Although the lymphocytic vasculitis lesions of the prostate appeared histologically indistinguishable from those seen in systemic PAN arising in the prostate, the rare occurrence of fibrinoid necrosis, as seen in the present series, can be useful in differential diagnosis and may limit the cases for additional evaluation to those with fibrinoid necrosis, a lesion focally present in 5.9% of lymphocytic vasculitis in the present series. Otherwise, granulomas with fibrinoid necrosis combined with necrotising vasculitis are findings seen only in patients with systemic PAN [3,9,17,20].

The present series represents the largest report to date describing in detail the architectural patterns and cellular features of this uncommon form of vasculitis arising in the prostate, but also shows that this lesion might have been under-recognised until now because of the lack of precise diagnostic features, as outlined in the present study based on a large sequential series validated in specimens from a different medical centre. In addition to the characteristic morphology of the lesion, the observed immunoprofile with predominant T cell phenotype and a minor component of B cells and monocytes, an original finding not previously reported, could be a useful diagnostic feature [13,21,22].

Summarising morphological and immunohistochemical findings that characterise PAN-like lymphocytic vasculitis associated with BPH, it seems that lymphocytic vasculitis reflects an inflammatory lesion of participating vessels followed by healing and wall sclerosis leading to prostatic ischaemia and, in some cases, to prostatic infarction. An association recognised for the first time in the present study.

As it is well known that BPH is a risk factor for systemic cardiovascular disease with no specific underlying lesion described [15], it could be hypothesised that the BPH increased risk for systemic vascular disease be reflected, at least in some cases, by the presence of lymphocytic vasculitis of the prostate. This hypothesis, if confirmed, could open the possibility of recognising prostatic lymphocytic vasculitis as a potential risk factor for systemic vascular disease in some patients.

In conclusion, lymphocytic vasculitis in men with BPH is associated with prostatic infarction and should be considered a form of localised vasculitis with PAN-like morphology that does not necessitate additional evaluation for systemic disease.

ACKNOWLEDGMENT

Supported in part by the grant SAF2007-64942, Ministry of Education. Madrid, Spain.

CONFLICT OF INTEREST

None declared.

Ancillary