Clinical features of paraneoplastic neurologic syndromes with anti-Ri antibodies: PRISMA systematic review

Abstract

Background and objectives

Paraneoplastic neurological syndromes (PNSs) are a group of neurological disorders linked to malignancies. Their clinical trajectory, differential diagnoses, immune-mediated mechanisms, and outcomes remain inadequately understood. The review aims to improve our understanding of the main syndrome and its clinical progression by systematic review of studies involving patients with anti-Ri-antibody-associated paraneoplastic neurological syndrome (Ri-PNS).

Methods

The protocol adhered to the PRISMA guidelines and is registered with PROSPERO (ID: CRD420251152531).

Results

Eighty-five cases with comprehensive clinical data were identified, with a median age of 61.0 ± 11.8 years, and the majority were female (78.6%). At the disease onset, ataxia was the most prevalent neurological symptom (70.6%). Ataxia rarely manifested as a state of isolation, and was frequently accompanied by other symptoms: tremor, dysarthria, spasticity, and dystonia. Twenty-six patients (30.6%) developed opsoclonus, and 22.4% developed myoclonus. The median interval from the onset of PNS to cancer diagnosis was 1.5 months (IQR, 0–7.0 months). Breast cancer was frequently observed in female patients (65.2%), whereas lung cancer was more common in male patients (38.9%). Of the patients with cancer, fifty-eight received oncological treatment. In addition, 62/85 patients (72.9%) underwent immunotherapy.

Conclusions

Ri-PNS is a multisystem neurological syndrome with significant involvement of the cerebellum and brainstem. The syndrome is frequently associated with tumors, and its appearance often signals the presence of a concurrent or occult tumor. Anti-tumor therapy and immunotherapy remain the two major management approaches in treatment. Considering that early intervention has the potential to enhance neurological function and improve the prognosis in patients with Ri-PNS, it could be given particular attention in clinical practice.

Supplementary Information

The online version contains supplementary material available at 10.1007/s00415-026-13762-9.

Introduction

Paraneoplastic neurological syndromes (PNSs) are autoimmune conditions connected to malignancies, predominantly associated with specific types, namely lung, bladder, breast, and lymphomas [1, 2]. Thus, these rare diseases are associated with widely occurring cancer subtypes, proposing the idea that PNS is remote consequences of cancer with an immune-related pathogenesis.

One could hypothesize that every cancer subtype may possess a histomolecular variant that predisposes individuals to the development of paraneoplastic autoimmunity [3]. Consequently, recent focus has been on finding new serum or cerebrospinal fluid (CSF) biomarkers that are capable of specifically detecting occult malignancies [4]. In the specific context of PNS-associated breast cancers, it has been demonstrated that these malignancies are associated with Ri antibodies, despite the rarity of Ri antibodies [5]. Serum anti-Ri antibodies, also referred to as anti-neuronal nuclear antibody type 2 (ANNA2), are typically found in women with cerebellar ataxia and opsoclonus-myoclonus syndrome (OMS) in association with gynecological and lung cancers [6].

Subsequent to the original descriptions of Ri-PNS, later findings have reported a range of neurological manifestations, such as ataxia and ophthalmoplegia, and have identified different oncological associations [7]. The disease trajectory, differential diagnostic spectrum, immune-mediated mechanisms, and clinical outcomes of Ri-PNS remain to be elucidated. The main aim of the systematic review was to synthesize and emphasize current literature on Ri-PNS to enhance the understanding of its clinical features.

Methods

Protocol and registration

Our protocol adhered to the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guideline, and was registered with the PROSPERO International Prospective Register of Systematic Reviews (Identification number: CRD420251152531, accessible at: https://www.crd.york.ac.uk/PROSPERO/view/CRD420251152531).

Data sources and search

We performed a thorough systematic review by exploring PubMed, the Cochrane Library, and Web of Science up to July 2025. The strategy for searching incorporated both free-text and subject-specific terms, including “anti-neuronal nuclear antibody 2”, “ANNA2”, “anti-Ri antibody”, “paraneoplastic neurological syndrome”, “encephalitis”, “antibody”, and “autoantibody”, integrating Boolean operators to guarantee the retrieval of all pertinent articles (detailed Search Strategy in supplementary Table 8). Furthermore, the reference lists of all chosen publications were reviewed to ensure articles were not included.

Inclusion and exclusion criteria

In the beginning stage of the screening phase, Chaoer Wu and Xinlei Liu, independently evaluated the “Titles and Abstracts” of all retrieved literature to examine any clear exclusions. Articles identified as potentially eligible were included through a comprehensive full-text assessment. Any studies that recorded at least one case of Ri-PNS were included, with no restrictions of language. International or national conferences, and abstracts submitted to letters to the editors were considered as eligible.

The 2004 PNS criteria are considered partially outdated due to advancements in PNS research, which had led to the identification of new phenotypes and antibodies, thereby transforming the diagnostic approach to 2021 PNS criteria. Nevertheless, given the evolving understanding of the disease and advancements in laboratory techniques during that period, the 2004 criteria remained the standard for studies published prior to the implementation of the 2021 guidelines. Two criteria are used for diagnosis based on the publication time of each case. The included patients met definite or probable 2021 PNS-care Score (Score ≥ 6), a newly developed instrument for assessing paraneoplastic disease according to the updated 2021 PNS standard (detailed in supplementary Table 2) [8]. The 2004 PNS standard is as follows [9]: (i) a typical syndrome with the emergence of cancer within 2 years following diagnosis, (ii) an atypical syndrome within 2 years of being diagnosed with cancer and either the presence of onconeural autoantibodies or significant neurological recovery after oncological therapy, or (iii) a typical or atypical syndrome and identification of a well-recognized onconeural autoantibody. Positivity of Ri antibodies was defined by two complementary detection methods: indirect immunofluorescence assay (IFA) on rat brain sections, which involved staining the nucleus and cytoplasm of neurons, and a confirmatory test using a cell-based assay, such as immunodot or Western blotting (LIA) with recombinant protein, as recommended.

Exclusion criteria included positive antibodies detecting for another paraneoplastic disease that more precisely accounted for their symptoms. Individuals with insufficient data, such as clinical presentation, cancer association, or paraclinical investigations, were excluded from our cohort.

Data collection and extraction

Two reviews, Hongmei Cui and Yuqin Yan, independently extracted the following information: publication details, age, gender, neurological symptoms at onset, CSF analysis, anti-Ri antibody, imaging examination at first presentation, cancer association, oncological treatment, immunotherapy, and clinical outcome. A third reviewer (Wei Qian) resolved any disagreements between the two reviewers through further assessment. The extracted data were entered into a Microsoft Excel spreadsheet to facilitate analysis. All variables were standardized as follows: for the analysis of the time span between cancer and the onset of neurological symptoms (referred to as onset of Ri-PNS), time zero was defined as the time of onset of neurological symptoms, with positive values representing a cancer diagnosis after neurological symptoms and negative values indicating cancer diagnosis prior to neurological symptoms. Cancer diagnosis was considered to be concurrent if it was made with 1 month of Ri-PNS symptom onset. Other criteria included: (i) leukocytosis was defined as greater than > 5 cells/µl; (ii) elevated protein levels were defined as greater than > 0.45 g/L; (iii) good outcome group included patients with complete recovery or significant or mild improvement; while bad outcome group included patients with no improvement or worsening. Specifically as follows: (a) complete recovery: a full disappearance of neurological symptoms, a restoration of neurological functions to their pre-illness state with no limitations on daily activities; (b) significant recovery or mild improvement: a major or mild alleviation in neurological symptoms, with the majority of neurological functions returning to their pre-illness state, but with neurological symptoms may persist or daily activities might be limited; (c) no improvement or worsening: stabilization or no improvement in neurological symptoms, or a worsening of their condition. For patients with relapsing–remitting courses, the outcome was assigned based on the status at last follow-up. Deaths were not automatically assigned to the “no improvement or worsening” category unless explicitly described as a consequence of neurological deterioration. The missing information was designated as ‘NA’ (not available).

Statistical analysis

Statistical analyses SPSS 26.0 were performed for the statistical analysis. The data complying with normal distribution were presented in the form of mean and standard deviation (mean ± SD), and those that did not conform to the normal distribution were presented in the form of median with the interquartile range (IQR). Categorical data were presented as frequencies and percentages. The Chi-square test was conducted utilizing a 2 × 2 contingency table to evaluate the association between selected variables and clinical outcomes. Odds ratios (ORs) and 95% confidence intervals (CIs) were computed. Continuity corrections were applied based on the total sample size or the expected frequency of all cells. The analysis was performed on the available data, excluding any missing data points. P values less than 0.05 were considered statistically significant.

Results

Figure 1 shows a flow chart of the study retrieval process. Eighty-five cases with comprehensive clinical data were identified. A total of 37 articles were excluded, among which 12 were omitted due to a lack of detailed data, accounting for 88 patients. A total of 64 articles qualified for final analysis and inclusion in the systematic review [10–73]. The checklist for critical evaluation of case reports and series was presented in Supplementary Table 9. These articles, published from 1991 to 2025, consisted of 2 case series and 62 case reports, collectively documenting 85 patients with Ri-PNS. An overview of the demographic details and clinical features of these cases was presented in Supplementary Table 1. The geographical distribution was detailed in Supplementary Table 3.

Fig. 1.

Flow diagram of selection process.

Clinical features

Neurological symptoms

The mean age at disease onset was 61.0 ± 11.8 years, with a predominance of female patients (78.6%) (Table 1). Regarding clinical presentations, 60 patients (70.6%) initially presented with ataxia (Fig. 2A). Figure 2B illustrated the co-occurrence patterns of neurological symptoms in patients with Ri-PNS. Ataxia seldom occurred in isolation and was frequently accompanied by additional symptoms or signs of tremor (n = 12), dysarthria (n = 13), spasticity (n = 8), dystonia (n = 13), and cognitive impairment (n = 10). As a characteristic feature of Ri-PNS, 26 patients (30.6%) developed opsoclonus, and 19 (22.4%) exhibited myoclonus. Among the 25 patients presenting with oculomotor symptoms, diplopia was the initial complaint. Nearly half of the patients (48.2%) experienced ophthalmoplegia, specifically oculomotor deficits (n = 14) and ptosis (n = 6). These patients demonstrated objective oculomotor deficits upon neurological examination, including complete external ophthalmoplegia in one case, cranial nerve palsy in 12 cases, and supranuclear ophthalmoplegia in one case.

Table 1.

Clinical features of Ri-PNS reported in the literature

Variables	No. of patients (n/N, %)
Median age at onset, (year, mean ± SD)	84 (61.0 ± 11.8)
Gender, female, n (%)	66/84 (78.6)
Neurological symptoms, n (%)
Opsoclonus	26/85 (30.6)
Myoclonus	19/85 (22.4)
Ataxia	60/85 (70.6)
Dysarthria	18/85 (21.1)
Ophthalmoplegia	41/85 (48.2)
Diplopia	25/85 (29.4)
Dysphagia	17/85 (20.0)
Spasticity	12/85 (14.1)
Respiratory failure	8/85 (9.4)
Laryngospasm	5/85 (5.9)
Peripheral neuropathy	5/85 (5.9)
Dystonia	13/85 (15.3)
Cognition impairment	18/85 (21.2)
Cancer, n (%)	73/85 (85.9)
Cancer type, n (%)
Breast	46/73 (63.0)
Lung	13/73 (17.8)
Bladder	4/73 (5.5)
Othera	10/73 (13.7)
Cancer-PNS delayb, mo (IQR)	53, 1.5 (0, 7.0)
PNS-Cancer delayc, mo (IQR)	53, 4.0 (2.0, 9.5)
CSF analysis at onset, n (%)
Leukocytosis	34/51 (66.7)
Elevated protein	21/51 (41.2)
Positive OCBs	11/51 (21.6)
Anti-Ri Ab, n (%)
Serum	79/80 (98.8)
CSF	36/40 (90.0)
Brain MRI at onset, n (%)
Abnormal	28/59 (47.5)
Hyperintensity	20/59 (33.9)
Atrophy	5/59 (8.5)
Abnormal PET, n (%)	19/28 (67.9)

PNS paraneoplastic neurological syndromes, SD standard deviation, CSF cerebrospinal fluid, OCBs oligoclonal bands, Ab antibody, MRI magnetic resonance imaging, PET positron emission tomography

^aOther cancer included ectopic mediastinal seminoma, fallopian tube tumor, nasopharyngeal carcinoma, ovarian duct adenocarcinoma, pancreatic neuroendocrine tumor, thymus carcinoma, transitional cell carcinoma of renal pelvis, uterine cervix cancer, and neuroendocrine neoplasm of rectum

^bTime from paraneoplastic neurologic disease symptom onset to cancer diagnosis

^cTime from cancer diagnosis to paraneoplastic neurologic disease symptom onset

Fig. 2.

Clinical feature at presentation. A Neurological symptoms, B chematic representation of the complex association of neurological signs, C sex-related cancer specificities

Nonclassical symptoms were observed in a minority of patients, with five individuals (5.9%) developing a peripheral neuropathy and two (2.4%) experiencing epileptic seizures. Other severe symptoms warranting attention were respiratory failure, observed in eight patients (9.4%). During the follow-up period, 6 patients succumbed, 1 was transferred to another facility, and 1 experienced multiple recurrences. Ventilator treatment was administered to three of these patients.

Oncologic associations

Seventy-three patients (85.9%) developed cancer, with breast cancer being the most prevalent (63.0%). As expected, breast cancer occurred more frequently in females (65.2%), whereas lung cancer was more common in male patients (38.9%) (Fig. 2C). The status of the human epidermal growth factor receptor 2 (HER2) was assessed in 14 breast cancer patients, revealing that 11 were negative and 3 were positive. In half of the cancer patients (50.0%), PNS preceded the cancer diagnosis. Eighteen patients (34.0%) experienced the simultaneous onset of cancer and PNS (within less than one month), while eight developed neurological symptoms following the cancer diagnosis, with a median time of 4.0 months (IQR, 2.0 to 9.5 months). The median time from the onset of paraneoplastic neurological disease symptoms to cancer diagnosis was 1.5 months (IQR, 0 to 7.0 months). Furthermore, the association between cancer type and neurological symptoms was detailed in Supplementary Table 4.

Paraclinical data

CSF analysis revealed pleocytosis in 34 patients (66.7%), elevated protein levels in 21 patients (41.2%), and detection of oligoclonal band (OCB) in 11 patients (21.6%). Anti-Ri antibodies were positive in 90.0% of CSF samples and 98.8% of serum samples. Remarkably, anti-Ri antibodies were found in the serum of 4 patients but were absent in the CSF, while 1 patient had these antibodies in the CSF but not in the serum. Brain magnetic resonance imaging (MRI) results data were available for 59 patients and were abnormal in 28 cases (47.5%). Among these, 20 cases exhibited signal changes (specifically included T2 hyperintensity lesions in 11 patients); and 5 cases had atrophy changes, with 2 cases showing cerebellar and cerebral atrophy, 1 case with frontotemporal atrophy, and 1 case with cortical atrophy. Of the 28 patients with available fludeoxyglucose-positron emission tomography (FDG-PET) data, 19 cases showed increased uptake, all indicative of tumors. The false negative rate was 6/25 patients (21.4%) (Supplementary Table 5).

Outcome and treatment

Fifty-eight patients (79.5%) with cancer received oncological treatment, including surgery (54.8%), chemotherapy (45.2%), radiotherapy (32.9%), and hormonotherapy (26.0%) (Table 2). In addition, 62/85 patients (72.9%) received immunotherapy, with all individuals undergoing first-line immunotherapy: 16 patients (18.8%) were treated exclusively with steroids, 10 patients (11.8%) received IVIG, and 23 patients (27.1%) were administered a combination of steroids and intravenous immunoglobulins (IVIG). Eleven patients (12.9%) required escalation to second-line immunotherapy: 3 (3.5%) received rituximab (RTX); 2 (2.4%) were treated with cyclophosphamide (CYC); 1 (1.2%) received azathioprine (AZA); 4 (4.7%) were administered a combination of CYC and AZA; 1 (1.2%) received mycophenolate mofetil (MMF) (Supplementary Table 6). During the final follow-up, with a median duration of 13 months (IQR, 6 to 26.8 months), 6/83 patients (7.2%) experienced complete recovery with no clinical sequelae. Additionally, 51 cases (61.4%) achieved significant or mild clinical improvement, whereas 26 cases (31.3%) exhibited no improvement or showed a deterioration in their state. Twenty-five patients succumbed to their illness at the last follow up.

Table 2.

Treatment and outcome of Ri-PNS reported in the literature

Variables	No. of patients (n/N, %)
Anti-tumor therapy, n (%)	58/73 (79.5)
Surgery	40/73 (54.8)
Chemotherapy	33/73 (45.2)
Radiotherapy	24/73 (32.9)
Hormonotherapy	19/73 (26.0)
Immunotherapy, n (%)
First-line immunotherapya	62/85 (72.9)
Second-line immunotherapyb	11/85 (12.9)
Follow-up
Median time from onset to last follow-up (IQR), mo	60, 13 (6, 26.8)
Clinical outcome, n (%)
Complete recovery	6/83 (7.2)
Significant or mild improvement	51/83 (61.4)
No improvement or worsen	26/83 (31.3)
Death at final follow-up	25/83 (30.1)

PNS paraneoplastic neurological syndromes, IQR interquartile range, IV intravenous, MTP methylprednisolone, IVIG IV immunoglobulins, Pred prednisolone, CYC cyclophosphamide, MMF mycophenolate mofetil, TAC Tacrolimus, AZA azathioprine, RTX rituximab

^aIV-MTP, IVIG, PE, oral-Pred

^bCYC, MMF, TAC, AZA, RTX

Four patients underwent a diagnostic neuropathology post-mortem with evidence of localized and exclusively lymphocytic infiltrates involving the cerebellum and brainstem [42, 48, 65, 69]. In one patient who experienced respiratory failure, lymphocytic infiltrates containing B cells were observed, particularly affecting the pons, medulla, cranial nerve nuclei, and the circulatory and respiratory center [65]. Two patients exhibited Purkinje cell loss in the cerebellum [48, 69].

Comparison of Ri-PNS between the good outcome and bad outcome

To explore the potential clinical factors associated with clinical outcomes, individuals were categorized into a good outcome group (n = 57) and a bad outcome group (n = 26) according to their clinical outcomes (Table 3). Patients without brain atrophy had a good outcome compared to those with brain atrophy (p = 0.038, 0.101 95%CI [0.010–0.983]). The proportion of cancer in the good outcome group (93.0%) was higher than in the bad outcome group (76.9%) (p = 0.037, 3.975 [1.014–15.576]).

Table 3.

Comparison of patients with Ri-PNS according to good or bad outcome

Variables	Good outcome	Bad outcome	OR (95%CI)	P value
Patients (total 83), n (%)	57	26
Demographics
Median age at onset, (year, mean ± SD)	56 (60.1 ± 12.0)	26 (62.5 ± 11.3)	–	0.393
Female, n (%)	44/56 (78.6)	21/26 (80.8)	1.873 (0.272–2.801)	0.819
Neurological symptoms, n (%)
Opsoclonus	21/57 (36.8)	5/26 (19.2)	2.450 (0.804–7.463)	0.109
Myoclonus	13/57 (22.8)	6/26 (23.1)	0.985 (0.327–2.966)	0.978
Ataxia	43/57 (75.4)	16/26 (61.5)	1.920 (0.710–5.187)	0.195
Dysarthria	11/57 (19.3)	6/26 (23.1)	0.797 (0.259–2.455)	0.692
Ophthalmoplegia	30/57 (52.6)	9/26 (34.6)	2.099 (0.803–5.486)	0.127
Diplopia	15/57 (26.3)	9/26 (34.6)	0.675 (0.248–1.834)	0.439
Dysphagia	14/57 (24.6)	3/26 (11.5)	2.496 (0.650–9.588)	0.173
Spasticity	10/57 (17.5)	2/26 (7.7)	2.553 (0.518–12.593)	0.237
Respiratory failure	4/57 (7.0)	2/26 (7.7)	0.906 (0.155–5.288)	1.000
Laryngospasm	4/57 (7.0)	1/26 (3.8)	1.887 (0.200–17.764)	1.000
Peripheral neuropathy	3/57 (5.3)	2/26 (7.7)	0.667 (0.105–4.251)	0.646
Dystonia	10/57 (17.5)	3/26 (11.5)	1.631 (0.409–6.505)	0.746
Cognition impairment	13/57 (22.8)	5/26 (19.2)	1.241 (0.765–1.487)	0.714
CSF analysis at onset, n (%)
Leukocytosis	21/34 (61.8)	8/16 (50.0)	1.615 (0.487–5.361)	0.432
Elevated protein	13/34 (38.2)	7/16 (43.8)	0.796 (0.238–2.658)	0.710
Positive OCBs	7/34 (20.6)	4/16 (25.0)	0.778 (0.191–3.167)	0.728
Brain MRI at onset, n (%)
Abnormal	15/38 (39.5)	12/19 (63.2)	0.380 (0.122–1.186)	0.091
Hyperintensity	11/38 (28.9)	8/19 (42.1)	0.560 (0.177–1.768)	0.321
Atrophy	1/38 (2.6)	4/19 (21.1)	0.101 (0.010–0.983)	0.038
Cancer, n (%)	53/57 (93.0)	20/26 (76.9)	3.975 (1.014–15.576)	0.037
Cancer-PNS delaya, mo (IQR)	36, 3 (0, 7.8)	16, 0 (0, 5.5)	–	0.395
Anti-tumor therapy, n (%)
Surgery	29/53 (54.7)	11/20 (55.0)	0.989 (0.352–2.780)	0.983
Chemotherapy	27/53 (50.9)	6/20 (30.0)	2.423 (0.809–7.262)	0.109
Radiotherapy	20/53 (37.7)	4/20 (20.0)	2.424 (0.710–8.281)	0.175
Hormonotherapy	15/53 (28.3)	4/20 (20.0)	1.579 (0.453–5.500)	0.561
Immunotherapy, n (%)
First-line immunotherapyb	45/57 (78.9)	15/26 (57.7)	2.750 (1.006–7.516)	0.045
Second-line immunotherapyc	8/57 (14.0)	3/26 (11.5)	1.252 (0.304–5.160)	1.000

OR odds ratio, CI confidence interval, SD standard deviation, CSF cerebrospinal fluid, OCBs oligoclonal bands, MRI magnetic resonance imaging, Ab antibody, PNS paraneoplastic neurological syndromes, IQR interquartile range, IV intravenous, MTP methylprednisolone, IVIG IV immunoglobulins, Pred prednisolone, CYC cyclophosphamide, MMF mycophenolate mofetil, TAC tacrolimus, AZA azathioprine, RTX rituximab

^aTime from paraneoplastic neurologic disease symptom onset to cancer diagnosis

^bIV-MTP, IVIG, PE, oral-Pred

^cCYC, MMF, TAC, AZA, RTX

Discussion

The review provides several significant implications regarding the clinical features of patients with Ri-PNS.

The majority of Ri-PNS exhibit multisystem neurological impairments, with the cerebellum and brainstem being the most predominantly impacted area [74]. Initially, the characterization of Ri-PNS identified OMS as a prototypical presentation of the disorder. However, opsoclonus with or without the presence of myoclonus is merely a portion of the clinical spectrum of the Ri-PNS, occurring less frequently than previously anticipated and seldom presenting as an isolated symptom [75]. Our findings indicated that Ri-PNS exhibited a more heterogeneous range of clinical manifestations of Ri-PNS, typically involving multiple nervous systems, with a central manifestation as a cerebellar syndrome. The classical feature is a progressive cerebellar syndrome characterized by truncal or gait ataxia. Ataxia rarely manifested as a state of isolation and is most commonly associated with other movement disorders: limb tremor, spasticity, and dystonia. In cases where diplopia is the sole ocular manifestation, distinguishing between cerebellar and brainstem syndromes may be challenging. However, the presence of additional symptoms such as ophthalmoplegia or gaze palsy can strongly suggest brainstem symptoms. In comparison to a cohort derived from the French dataset [75], no statistically significant differences were detected in the primary clinical symptoms, including oculomotor dysfunction, myoclonus, and peripheral neuropathy. Similarly, there are no statistically significant differences in the incidence and predominant types of tumors, specifically breast and lung cancer, as detailed in Supplementary Table 7. These results further substantiate the clinical characteristics associated with Ri-PNS. Given that the intracellular RNA-binding proteins targeted by anti-Ri antibodies are encoded by the Nova-1 and Nova-2 genes, it is unlikely that these antibodies are directly pathogenic. Rather, the primary pathogenic mechanism is thought to be neuronal damage mediated by T lymphocyte [76]. This was supported by post-mortem diagnostic neuropathology, which revealed lymphocytic infiltrates in the brainstem and cerebellum.

In 2021, the working group defined ten high-risk antibodies associated with tumors, including ANNA2 [8]. The relationship between cancers related to PNS and neurological phenotypes is causal rather than a coincidence. Although prior research has emphasized the association between Ri-PNS and lung cancer, our findings indicate that breast cancer is the most prevalent tumor in patients with Ri-PNS. This suggests that women with Ri antibodies should be thoroughly investigated for breast cancer, and diagnostic monitoring should persist even if cancer is not initially identified at the onset of the neurological presentations. In contrast, males typically present with neuroendocrine malignancies such as those originating in the lung or bladder, or with mediastinal seminoma. Consequently, screening for undiagnosed tumors is imperative. A systematic approach should commence with the investigation of occult cancer through clinical examinations and initial diagnostic tests. For women, this includes mammography and breast ultrasound, while for men, a whole-body computed tomography (CT) scan is recommended, following the detection of Ri antibodies. The identification of potential tumors presents significant challenges, particularly in patients with PNS, where tumors are often diminutive, potentially due to the host immune response limiting their growth. These tumors frequently fall below the resolution threshold of standard CT and ultrasound imaging techniques. In cases where initial radiological assessment fails to identify any malignancy, FDG-PET is recommended, as it has been shown to possess greater sensitivity than CT screening alone for detecting occult malignancy [77, 78]. In a cohort of 25 patients with pathologically confirmed malignancies, FDG-PET facilitated the detection of occult cancerous lesions in 19 cases. It is important to recognize that negative PET findings do not definitively rule out the presence of a tumor in clinically suspicious cases. If no tumor is identified at the time of PNS diagnosis, it is recommended that cancer monitoring be conducted every 4–6 months for a duration of at least two years, or potentially longer. In our study, among patients with a delay between cancer and PNS diagnosis, the median interval was 7 months (IQR, 4.0–15 months).

There is a paucity of studies evaluating the efficacy of treatment for PNS. Accordingly, treatment strategies are predominantly informed by expert opinion and existing literature. The management of PNS primarily involves two key principles: addressing the underlying malignancy and initiating immunotherapy [79]. Effective treatment of the underlying malignancy is essential to disrupt the tumor-triggered autoimmune process. Nevertheless, in many cases, neurological symptoms such as ocular clonus, myoclonus, and ataxia may persist despite treatment of the primary tumor, necessitating the use of combined immunosuppressive therapy. Our analysis indicates that the choice of cancer treatment—be it surgery, chemotherapy, radiotherapy, or hormonotherapy—does not emerge as a statistically significant factor influencing prognosis within this cohort. Nonetheless, limitations in the available data restricted our ability to conduct a more comprehensive analysis to determine whether combination therapy provides a greater prognostic advantage over single-agent treatment.

It is imperative to commence this integrated immunological and oncological approach promptly to prevent irreversible neuronal damage and severe neurological disability [80]. Our research indicates that the incidence of brain atrophy is significantly higher in patients with poor prognosis compared to those with good outcomes, suggesting a correlation between brain atrophy and adverse prognostic outcome. First-line immunotherapy included: steroids, IVIG, and PE. These agents can be used in combination such as combining steroids with plasma exchange (PE), rather than sequentially assessing the efficacy of a single first-line agent before introducing another. Early improvement in neurological symptoms serves as an indicator of drug efficacy, while CSF analysis, MRI, and antibody levels provide paraclinical biomarkers of ongoing neuroinflammation. Given the cytotoxic T-cells mediated immune response in patients with Ri-PNS, we advocate for the early use of second-line immunotherapies, such as CYC, MMF, AZA, and RTX, which are designed to target all lymphocyte lineages, including both T and B cells [81]. Although these therapies are frequently employed as maintenance treatments, their early implementation in the disease course is advisable [82].

Immunosuppression may elevate the risk of chemotherapy toxicity [83]. However, there is currently a lack of information concerning the impact of immunotherapy on cancer outcomes in cases with Ri-PNS. In our study, we observed that three patients succumbed to tumor metastasis during follow-up; however, no direct correlation was established between the use of immunosuppressants and tumor recurrence or metastasis. Consequently, further prospective studies are warranted to evaluate cancer outcomes and relapse in cases with PNS undergoing chronic immunosuppression.

Given that the studies incorporated in this review predominantly consist of case reports and case series, which are recognized as low-level evidence, our recommendations should be regarded as preliminary and necessitate further investigation. Based on current evidence, in most cases of PNS, the neurological benefits of early aggressive immunotherapy might justify consideration of the potential risks. In most cases of PNS, disability from neurological dysfunction appears to affect the morbidity much more than the underlying cancer. A term-based approach with frequent discussions with oncologists is required, especially for PNS cases with advanced cancers.

Limitations

The limitations of the present review are primarily attributed to its retrospective collection. The sample size is likely underestimated due to the exclusion of certain literature for which full texts are unavailable to download or which lack insufficient data. Language barriers further constrained our ability to retrieve and analyze publications not available in English. Moreover, the substantial heterogeneity observed among the included studies—encompassing variations in diagnostic criteria, clinical assessments, and follow-up evaluations—introduces potential bias and constrains the robustness of our findings. We recognize that this clinical and methodological diversity among studies constitutes an inherent limitation of systematic reviews concerning rare diseases and may impact the generalizability of our conclusions.

Conclusions

Ri-PNS is a multisystem neurological syndrome with significant involvement of the cerebellum and brainstem. The syndrome is frequently associated with tumors, and its appearance often signals the presence of a concurrent or occult tumor, necessitating prompt screening and ongoing surveillance. Anti-tumor therapy and immunotherapy remain the two major management approaches in treatment. Considering early intervention has the potential to enhance neurological function and improve the prognosis in patients with Ri-PNS, it could be given particular attention in clinical practice.

Supplementary Information

Below is the link to the electronic supplementary material.

Abbreviations

ANNA2

Anti-neuronal nuclear antibody type 2

AZA,

Azathioprine

CSF

Cerebrospinal fluid

CT

Computed tomography

CYC

Cyclophosphamide

FDG-PET

Fludeoxyglucose-positron emission tomography

HER2

Human epidermal growth factor receptor 2

IQR

Interquartile range

IFA

Immunofluorescence assay

IVIG

Intravenous immunoglobulins

LIA

Linear Western blotting

MMF

Mycophenolate mofetil

MRI

Magnetic resonance imaging

NA

Not available

OCBs

Oligoclonal bands

OMS

Opsoclonus–myoclonus syndrome

PE

Plasma exchange

PRISMA

Preferred reporting items for systematic reviews and meta-analyses

Ri-PNS

Anti-Ri-associated paraneoplastic neurologic syndrome

RTX

Rituximab

Author contributions

Y.C., C.W., X.L., L.X., and W.Q. contributed to the study design. Y.C., C.W., X.L., H.C., Y.Y., L.L., L.X., and W.Q. carried out the study. C.W., X.L., H.C., and Y.Y. analyzed and interpreted the data. Y.C., C.W., X.L., H.C., Y.Y., L.L., L.X., and W.Q. drafted, revised, and provided final approval of the manuscript.

Funding

This work was supported by grants from the Innovation Project of the Institute (The First People’s Hospital of Lin’an District: LY2024004K).

Data availability

The data underlying this article will be shared upon reasonable request to the corresponding author.

Declarations

Conflict of interest

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.