2021 ACR/VF Guideline for the Management of Polyarteritis Nodosa (PAN)

These recommendations provide guidance regarding diagnostic strategies, use of pharmacologic agents, and imaging for patients with PAN.

Download Complete Guideline (PDF)


General Information


For the evidence report, the Literature Review team summarized 127 articles to address 21 PICO questions for PAN.

The following recommendations and ungraded position statements are for systemic PAN and do not apply to isolated cutaneous or HBV-related PAN.

Recommendations and Positions for PAN

Table 1 presents definitions of selected terms used in the recommendations, including the definition of severe and nonsevere disease, as well as dosing ranges for glucocorticoids. Table 2 presents the recommendations with their supporting PICO questions and levels of evidence. Figure 1 provides key recommendations for the treatment for PAN. All but 1 of the recommendations are conditional, primarily due to lack of high-quality evidence (e.g., randomized controlled trials) supporting the recommendation.



Recommendation: For patients with suspected PAN, we conditionally recommend using abdominal vascular imaging to aid in establishing a diagnosis and determining the extent of disease.

Evidence for the use of routine diagnostic imaging is limited, with no comparative trials available. In single-arm studies that were performed when diagnostic criteria for PAN were not well defined, vascular imaging, in tandem with clinical signs and pathology, helped validate the diagnosis (8) and determine disease severity (9). This in turn can influence treatment decisions. Moreover, obtaining vascular imaging at disease onset facilitates identification of new vascular involvement during disease relapse. Vascular imaging may not be warranted if patients present with isolated findings such as mononeuritis multiplex or myopathy, or if there are no clinical features suggestive of abdominal arterial involvement (such as absence of gastrointestinal or genitourinary symptoms, including renovascular hypertension). For children, clinicians should be mindful of minimizing repeated radiation exposure.

Clinicians currently use both conventional catheter-based dye angiography and noninvasive methods such as computed tomography (CT) or magnetic resonance (MR) angiography to diagnose PAN (10–12).

Conventional angiography is the current gold standard due to its ability to provide better resolution, but it can be associated with complications, albeit at a very low rate (13,14). However, the resolution for noninvasive modalities is improving, and CT or MR angiography may provide additional information regarding the vessel wall that conventional angiography does not. Specifically, CT angiography may enable visualization of more of the distal branches of the mesenteric arteries than MR angiography, but MR angiography may be preferred in certain clinical situations (e.g., need to avoid iodinated contrast). In patients with a negative CT or MR angiogram result with a high degree of suspicion for abdominal involvement, it is reasonable to consider conventional angiography.

Recommendation: For patients with a history of severe PAN with abdominal involvement who become clinically asymptomatic, we conditionally recommend follow-up abdominal vascular imaging.

Follow-up imaging permits assessment of disease control and treatment response. In the view of the Voting Panel, follow-up imaging is particularly important when baseline imaging demonstrates aneurysmal disease. The timing of follow-up imaging is dependent, in part, on clinical factors, such as the extent and severity of vascular abnormalities, overall disease course, and response to therapy.

However, indefinite routine vascular imaging should be avoided if the abdominal vascular disease is shown to be quiescent.

Recommendation: For patients with suspected PAN involving the skin, we conditionally recommend obtaining a deep-skin biopsy specimen (i.e., a biopsy reaching the medium-sized vessels of the dermis) over a superficial skin punch biopsy to aid in establishing a diagnosis.

Indirect evidence (found in nonrandomized studies or studies in which findings were not primary aims) suggests that evaluation of deeper tissue is more effective at establishing a diagnosis of PAN (15,16), since a deeper-tissue sample is more likely to capture a medium-sized vessel. A deep-skin biopsy can be performed by a dermatologist as a deep (or “double”) punch biopsy and does not necessarily require invasive resection. This recommendation had strong support from the Voting Panel but remains conditional due to limited evidence.

Recommendation: For patients with suspected PAN and peripheral neuropathy (motor and/or sensory), we conditionally recommend obtaining a combined nerve and muscle biopsy over a nerve biopsy alone to aid in establishing a diagnosis.

Several studies suggest an increased yield with nerve and concurrent muscle biopsy as opposed to nerve biopsy alone (15–19). However, the biopsy should sample involved tissue and not be performed “blind” (i.e., sampling tissue that does not appear to be clinically affected). Of note, biopsy of an affected purely sensory nerve (e.g., sural nerve) is favored to avoid motor deficits.

Recommendation: For patients with a history of peripheral motor neuropathy secondary to PAN, we conditionally recommend serial neurologic examinations instead of repeated lectromyography/nerve conduction studies (e.g., every 6 months) to monitor disease activity.

This recommendation is based on the opinion of the Voting Panel due to a lack of published evidence addressing the issue. Repeated electromyography in a patient with stable symptoms is not recommended due to the invasive nature of this study. However, repeated electromyography/nerve conduction study would be warranted if there were uncertainty as to whether a new (or worsening) process was developing.

Recommendation: For patients with newly diagnosed active, severe PAN, we conditionally recommend initiating treatment with intravenous (IV) pulse glucocorticoids over high-dose oral glucocorticoids.

In several single-arm and comparative studies, evaluations of medical therapy were confounded by the use of other medications and did not control for IV pulse or high-dose oral glucocorticoid use (20–22). However, for active and severe disease specifically, patients may benefit from the additional mechanism of action of high-dose pulse glucocorticoids. That is, glucocorticoids may rapidly alter cell membrane and receptor function to promote suppression of inflammation once the glucocorticoid receptor is saturated (23). The Voting Panel noted that this recommendation was focused on patients with active, severe disease. For many patients with disease that is not associated with life-threatening manifestations (such as immediate risk of visceral infarct), oral glucocorticoids would be preferred due to lower overall glucocorticoid burden. For pediatric patients, pulse glucocorticoid therapy in other systemic immune disorders appears to have a favorable side-effect profile and is not more strongly associated with infections or other morbidities compared to oral glucocorticoids (24).

Recommendation: For patients with newly diagnosed active, severe PAN, we conditionally recommend initiating treatment with cyclophosphamide and high-dose glucocorticoids over high-dose glucocorticoids alone.

In newly diagnosed severe PAN, a single observational study and indirect evidence suggest that the use of cyclophosphamide has more benefits than glucocorticoid therapy alone, with no differences seen between oral and IV cyclophosphamide (25,26). Moreover, the use of additional cyclophosphamide cycles may provide a medium-term protection (3 years) against disease relapse, although this benefit wanes by 10 years (21). Use of cyclophosphamide may mitigate glucocorticoid toxicity by decreasing the cumulative steroid dose (27).

Recommendation: For patients with newly diagnosed active, severe PAN, we conditionally recommend initiating treatment with cyclophosphamide and glucocorticoids over rituximab and glucocorticoids.

While case reports have recently raised the question about the efficacy of rituximab use in PAN (28–30), its efficacy in PAN remains uncertain due to the lack of comparative or large single-arm studies in this disease.

Recommendation: For patients with newly diagnosed active, severe PAN who are unable to tolerate cyclophosphamide, we conditionally recommend treating with other nonglucocorticoid immunosuppressive agents and glucocorticoids over glucocorticoids alone.

Indirect evidence (i.e., data obtained from secondary outcomes in prior trials [25,31]) suggests that the combination of nonglucocorticoid immunosuppressive agents, such as azathioprine or methotrexate, with glucocorticoids is superior to glucocorticoids alone. Mycophenolate mofetil has not been well studied in PAN. No direct trials comparing glucocorticoid monotherapy with nonglucocorticoid combination therapy are available. In general, patients with severe PAN should be treated with cyclophosphamide over other immunosuppressive agents (26), but in patients unable to tolerate cyclophosphamide, another agent, such as azathioprine or methotrexate, is recommended over glucocorticoid monotherapy. Use of nonglucocorticoid immunosuppressive therapy may provide a glucocorticoid-sparing effect and minimize glucocorticoid toxicity, which is particularly significant in pediatric populations.

Recommendation: For patients with newly diagnosed active, nonsevere PAN, we conditionally recommend treating with nonglucocorticoid immunosuppressive agents and glucocorticoids over glucocorticoids alone. 

In cases of nonsevere disease, a patient’s age, clinical condition, and their values and preferences are important factors in assessing treatment. Although some patients achieve disease remission while receiving glucocorticoids alone, a substantial number of patients ultimately require additional nonglucocorticoid therapy, usually azathioprine or methotrexate (20). This recommendation contradicts management recommendations based on the Five-Factor Score (32), in which patients without factors of severe disease can be treated with glucocorticoids alone. We favor the use of nonglucocorticoid therapy in nonsevere disease, since the addition of nonglucocorticoid therapy may minimize glucocorticoid use and subsequent toxicity.

Recommendation: In patients with newly diagnosed active, severe PAN, we conditionally recommend against using plasmapheresis combined with cyclophosphamide and glucocorticoids over cyclophosphamide and glucocorticoids alone.

In a single trial conducted in 1995, the use of plasmapheresis in PAN was evaluated, but a distinction between PAN and HBV-associated PAN was not made (33). Confidence intervals in this study were very wide. Thus, evidence supporting the use of plasmapheresis in non–HBV-associated PAN is unavailable and the benefit unclear. Plasmapheresis may be considered in catastrophic cases unresponsive to the recommended aggressive immunosuppressive therapies and may have a role in the management of HBV-related PAN.

Recommendation: For patients with PAN in remission who are receiving nonglucocorticoid immunosuppressive therapy, we conditionally recommend discontinuation of nonglucocorticoid immunosuppressive agents after 18 months over continued (indefinite) treatment.

Evidence for this recommendation is based on a single study that was performed in 1979 (31). Although a significant number of patients with PAN have disease relapse, the majority experience monophasic disease (20). Indefinite treatment may therefore not be needed. Disease needs to be in sustained remission (Table 1) before discontinuing therapy.

Ungraded position statement: The optimal duration of glucocorticoid therapy for PAN (e.g., tapering off by 6 months or longer than 6 months) is not well established, and thus, the duration of therapy should be guided by the patient’s clinical condition, values, and preferences.

In PAN, studies to determine the optimal length of time for glucocorticoid use have not been performed. In studies of other types of vasculitis (34), faster tapers led to more flares, which were often not organ-threatening and may have been mild. The Patient Panel preferred a longer taper, as a primary concern was disease control rather than glucocorticoid toxicity. Thus, duration of glucocorticoid use should be influenced by the patient’s clinical condition, values, and preferences.

Recommendation: For patients with severe PAN that is refractory to treatment with glucocorticoids and nonglucocorticoid immunosuppressive agents other than cyclophosphamide, we conditionally recommend switching the nonglucocorticoid immunosuppressive agent to cyclophosphamide over increasing glucocorticoids alone.

Based on the effectiveness of cyclophosphamide in new-onset severe PAN (26), indirect evidence suggests that cyclophosphamide should be used in patients with PAN that has evolved from a nonsevere presentation to one that is severe and does not adequately respond to other immunosuppressive agents.

Recommendation: For patients with newly diagnosed PAN who have achieved disease remission with cyclophosphamide, we conditionally recommend transitioning to another nonglucocorticoid immunosuppressive agent over continuing cyclophosphamide.

Due to its toxicity, cyclophosphamide therapy should not continue indefinitely and should generally be limited to 3–6 months per course (21). Based on the experience in antineutrophil cytoplasmic antibody–associated vasculitis, transitioning to another less toxic agent such as methotrexate or azathioprine is recommended once disease remission has been attained. Given the lack of clinical trials investigating remission maintenance in PAN, this recommendation was based on expert experience.

Recommendation: For patients with PAN with nerve and/or muscle involvement, we conditionally recommend physical therapy.

Indirect evidence for PAN is available for this recommendation from studies in inflammatory myositis. Based on this, we conditionally recommend this intervention due to its potential benefit and minimal risk. Physical therapy may be more beneficial for those with more substantial motor involvement. Patients on the Voting Panel expressed a high degree of enthusiasm for physical therapy as a modality for recovery and rehabilitation, in that they felt they had personally experienced benefit from physical therapy.

Recommendation: For patients with clinical manifestations of deficiency of adenosine deaminase 2 (DADA2), we strongly recommend treatment with tumor necrosis factor inhibitors over glucocorticoids alone.

DADA2 was first described in a series of patients with an early-onset (often childhood) PAN-like vasculitis (35). DADA2 is characterized by recurrent strokes and skin changes and diagnosed using ADA2 sequencing or ADA2 functional assays, and ADA2 mutations have been identified in patients diagnosed as having systemic PAN (36). Although only 1 case series has been published, the strong signal of benefit of tumor necrosis inhibitors provides evidence that treatment with tumor necrosis inhibitors, instead of conventional immunosuppressive agents such as cyclophosphamide, prevents strokes (35,37). Thus, physicians should consider DADA2 in the setting of a PAN-like syndrome with strokes, and if confirmed, we strongly recommend use of tumor necrosis factor inhibitors. The Voting Panel voted for a strong recommendation despite the small number of cases, stressing the prevention of severe adverse events.

For every $50 donation, we can reach one more person with the guidelines. Donate today!

The Vasculitis Foundation would like to thank Genentech for supporting our efforts to spread the guidelines to the vasculitis community.