Late-onset polyvinylpyrrolidone allergy in a long-term hemodialysis patient: a case report of dialysis hypotension and gastrointestinal symptoms

Abstract

Polysulfone membranes are widely used in hemodialysis due to their excellent biocompatibility, with polyvinylpyrrolidone (PVP) serving as a key hydrophilic component. While PVP-related allergic reactions typically occur during initial exposure, delayed reactions in long-term dialysis patients present a unique diagnostic challenge due to their unexpected nature and similarity to routine complications. We report a case of delayed PVP allergy in a 69-year-old male who developed severe intradialytic hypotension and gastrointestinal symptoms after 13 months of stable hemodialysis treatment. Blood volume monitoring revealed a rapid 15% decrease within 30 min of dialysis initiation despite the absence of ultrafiltration, accompanied by marked eosinophilia (absolute eosinophil count: 13,266/μL) and elevated white blood cell count (19,800/μL). These symptoms persisted despite dry weight adjustments but resolved completely after switching from a polysulfone membrane to a PVP-free cellulose triacetate membrane. The patient’s condition improved significantly, with eosinophil counts normalizing over nine months following the membrane change. This case contributes to the limited literature on delayed PVP allergic reactions, highlighting that such reactions can occur even after extended periods of stable dialysis. The successful use of blood volume monitoring and eosinophil tracking for early detection, combined with the effectiveness of PVP-free membrane substitution, provides valuable insights for managing similar cases. Recognition of this phenomenon may help improve the diagnosis and management of unexplained dialysis reactions in long-term patients.

Supplementary Information

The online version contains supplementary material available at 10.1007/s13730-025-01069-z.

Introduction

Hemodialysis is an essential life-sustaining therapy for patients with end-stage kidney disease (ESKD). However, symptoms such as intradialytic hypotension, nausea, vomiting, and muscle cramps may occur during dialysis, potentially hindering the continuation of treatment.

The most common cause of intradialytic hypotension is excessive ultrafiltration. However, if symptoms do not improve despite adjustments in fluid management, including dry weight and ultrafiltration volume, other causes such as allergic reactions should be considered in the differential diagnosis. These symptoms also have significant impacts on the patient’s quality of life and treatment safety.

Traditionally, acute reactions during hemodialysis, such as blood pressure drops, have been attributed to factors including complement activation caused by bio-incompatible dialysis membranes, hypoxia induced by acetate-containing dialysate, and hypersensitivity reactions mediated by immunoglobulin E due to ethylene oxide sterilization [1]. To address these technical challenges, advancements such as biocompatible dialysis membranes, bicarbonate-buffered dialysate, and the elimination of ethylene oxide sterilization have been introduced. However, despite these technological improvements, acute reactions have not been completely prevented, highlighting the need to investigate additional underlying factors [2, 3].

Among these, hypersensitivity reactions related to polysulfone (PS) membranes have emerged as a significant issue observed during hemodialysis. Polyvinylpyrrolidone (PVP), used in PS membranes, enhances membrane hydrophilicity, reducing the adsorption of plasma proteins and platelet adhesion [6, 7]. However, PVP release into the bloodstream has been reported to potentially trigger allergic reactions [4, 8]. These symptoms are well-documented in specific cases and have been associated with mechanisms such as complement activation and histamine release [5, 9].

Such reactions are generally more common during the early stages of dialysis treatment [10]. However, delayed-onset reactions have also been reported after prolonged use. For instance, a study by Boer et al. identified delayed reactions in 5 out of 32 cases (15.6%) after more than one year of hemodialysis [11]. Due to their rarity, these cases are not well understood in clinical practice, underscoring the need for re-evaluating long-term risk factors and establishing more effective management strategies.

This report presents a case of a patient who had been undergoing stable hemodialysis for over a year and experienced sudden hypotension and gastrointestinal symptoms immediately after dialysis initiation, making continuation of treatment challenging.

Case report

A 69-year-old male with a 13 month history of hemodialysis presented with sudden hypotension, vomiting, and bowel movements immediately after starting dialysis. He had no history of allergies. The primary underlying disease was chronic glomerulonephritis. His past medical history included unstable angina, myocardial infarction, and colonic diverticulitis.

Cardiac function evaluations conducted both before and after symptom onset revealed no significant changes, and left ventricular wall motion was intact. Transthoracic echocardiography showed preserved left ventricular systolic function, with ejection fractions of 67% prior to symptom onset and 62% afterward. The E/A ratio slightly decreased from 0.61 to 0.54 but did not indicate significant diastolic dysfunction. Mild increases in end-systolic volume and transient mild mitral regurgitation were observed, but no significant valvular disease or regional wall motion abnormalities were present. These findings suggest that the symptoms were unlikely to be of cardiac origin (Table 1).

Table 1.

Comparison of echocardiographic parameters one month before and one month after onset of symptoms

Parameter	1 Month before onset	1 Month after onset
Ejection fraction (EF, %)	67	62
E/A Ratio	0.61	0.54
stroke volume (SV, mL)	63	79
End-systolic volume (ESV, mL)	31	48
Mitral regurgitation (MR)	−	±
Aortic regurgitation (AR)	−	−
Tricuspid regurgitation (TR)	±	±
Pulmonary regurgitation (PR)	±	±

The patient maintained a stable hemodialysis course for approximately 13 months. However, beginning a few weeks prior to symptom onset, he began to experience episodes of hypotension immediately after dialysis initiation. These events were initially considered within the range of normal variation. On Day 0 (defined as the day when severe symptoms appeared), he developed sudden hypotension, vomiting, and bowel movements, leading to interruption of the dialysis session.

On Day 0, the patient was receiving intermittent infusion hemodiafiltration (I-HDF) using a polysulfone (PS) membrane dialyzer (ABH-18PA, Asahi Kasei Medical). The blood flow rate was set at 250 mL/min, with a treatment duration of 4 h. The dry weight was 57.2 kg, and the target fluid removal was 1.9 L. A priming wash volume of 3.5 L of ultrapure dialysate was used before treatment. Nafamostat mesilate was administered as the anticoagulant, and a bicarbonate-based dialysate (Kindaly 4E, Fuso Pharmaceutical Industries) was used. These dialysis parameters remained consistent at the time of symptom onset and during the subsequent clinical course.

Hemodialysis was initiated at the start of treatment with nafamostat mesilate as the anticoagulant. Approximately one month later, nafamostat was discontinued and replaced with low-molecular-weight heparin. Several weeks after the switch, the patient experienced an episode of gastrointestinal bleeding, prompting a temporary transfer to another hospital. Once his condition stabilized, he was readmitted to our facility, and nafamostat was reintroduced to minimize the risk of recurrent bleeding. Importantly, his symptoms subsequently resolved after the dialysis membrane was changed, even though nafamostat treatment was continued. This clinical course suggests that nafamostat was unlikely to have been the cause of the allergic reaction.

After Day 0, ultrafiltration was temporarily discontinued, resulting in blood pressure recovery and improvement of gastrointestinal symptoms. Initially, the possibility of an allergic reaction was not considered, and the symptoms were attributed to dry weight (DW) adjustments or ultrafiltration volumes. However, symptoms persisted even when DW was increased, and ultrafiltration was minimized. As a result, a blood volume monitor (Crit-Line®, JMS Corporation, Tokyo) was used to measure ΔBV (changes in circulating blood volume). This device uses near-infrared light to measure hematocrit values and monitors real-time changes in blood volume.

The results revealed a rapid 15% decrease in blood volume within 30 min after dialysis initiation, despite the absence of ultrafiltration. A rapid infusion of ultrapure dialysate (200 mL per administration) was attempted when BV decreased by 10%, but no response was observed, and the BV decrease continued. Gastrointestinal symptoms, including vomiting, appeared when BV approached a 15% decrease (Fig. 1). This rapid decline in BV, unresponsive to fluid replacement, suggested an allergic reaction involving increased vascular permeability rather than hypotension caused by ultrafiltration alone.

Fig. 1.

Changes in blood volume (BV) and systolic blood pressure (SBP) at the time of onset

The patient’s blood test results obtained during the month of symptom onset and closest to Day 0 (Table 2), along with the chronological changes in white blood cell (WBC) count and absolute eosinophil count over time (Fig. 2), are presented. Although a sample from the exact day of symptom onset (Day 0) was not available, the closest available data were used to reflect the hematologic status around that time. Eosinophil data from the time of dialysis initiation were also available and included in Fig. 2; however, data from one month after initiation to three months prior to Day 0 were unavailable due to the absence of differential WBC measurements during that period. The horizontal axis of the figure represents dialysis initiation and the period from three months before to 12 months after Day 0.

Table 2.

Blood test data (one month before and after onset)

Parameter	− 1 Month	Onset	+ 1 Month
RBC (× 104/μL)	342	358	320
Hematocrit (%)	34.8	35.1	31.3
Hemoglobin (g/dL)	11.2	11.2	10.3
WBC (/μL)	9800	19,800	9700
Neutrophils (/μL)	5390	4356	5141
Eosinophils (/μL)	1862	13,266	2425
Basophils (/μL)	98	0	0
Lymphocytes (/μL)	1764	1386	1455
Monocytes (/μL)	294	396	291
Platelets (× 104/μL)	18.3	14.5	14.2
Total protein (g/dL)	6.7	6.2	6.2
Albumin (g/dL)	3.2	2.9	2.9
BUN (mg/dL)	55.3	41.6	51.2
Creatinine (mg/dL)	10.28	10.05	10.2
Sodium (mEq/L)	139	140	137
Potassium (mEq/L)	4.6	4.6	4.2
Chloride (mEq/L)	102	103	100
Calcium (mg/dL)	7.6	8	7.7
Inorganic phosphorus (mg/dL)	6	4.5	5.2
Total cholesterol (mg/dL)	91	41	131
LDL-Cholesterol (mg/dL)	36	84	66
HDL-Cholesterol (mg/dL)	41	45	54
Triglycerides (mg/dL)	68	62	54
CRP (mg/dL)	1.23	0.66	0.65
AST (U/L)	10	14	11
ALT (U/L)	4	6	4
γ-GTP (U/L)	7	7	8

Fig. 2.

Temporal changes in white blood cell count and eosinophil ratio in response to dialysis membrane replacement

In the examination conducted two months prior to onset, a mild elevation in the absolute eosinophil count was observed.

By one month prior to onset, the WBC count was 9800/μL, and the eosinophil count had increased to approximately 1862/μL.

At the time of onset, the WBC count had sharply increased to 19,800/μL, and the eosinophil count had risen dramatically to 13,266/μL.Both values significantly exceeded their pre-onset levels, suggesting the presence of an allergic reaction likely triggered by PVP. This finding provided a crucial clue to understanding the pathophysiology of this case.

To address the possibility of an allergic reaction, the dialysis membrane was changed from a PS membrane to an FDX-12 (polyester-polymer alloy membrane, PEPA®, Nikkiso Corporation, Tokyo). Additionally, glycerol infusion was initiated to prevent osmotic pressure reductions. Following these changes, hypotension and gastrointestinal symptoms observed during dialysis were resolved (Fig. 3). Apart from the changes in membrane and the temporary use of glycerol, other dialysis conditions—including treatment duration, blood flow rate, and dialysate composition—remained unchanged during this period. After blood pressure stabilized, glycerol was discontinued, and the WBC and eosinophil ratios gradually decreased but remained elevated. One month after onset, the WBC count dropped to 9700/μL, and the eosinophil count decreased to 2425/μL. Over subsequent months, the WBC count returned to pre-onset levels, and the eosinophil count remained within the range of 2000–3000/μL, slightly higher than the baseline.

Fig. 3.

Changes in blood volume (BV) and systolic blood pressure (SBP) following dialysis membrane replacement

Considering the potential effects of PVP in both PS and PEPA membranes, the dialysis membrane was replaced with a PVP-free FIX-210S (cellulose triacetate membrane, CTA, Fineflux®, Nipro Corporation, Osaka). This change resulted in a gradual decrease in eosinophil counts, with levels returning to pre-onset values nine months after the onset (Fig. 2).

Discussion

This case represents a rare instance of a hemodialysis patient who had been undergoing stable treatment for over a year but suddenly experienced severe hypotension and gastrointestinal symptoms, such as vomiting, during dialysis, making the continuation of treatment challenging. It took approximately one month to reach a diagnosis, which was ultimately identified based on abnormal findings from blood volume monitoring and peripheral blood eosinophilia. This case highlights the importance of recognizing PVP hypersensitivity in long-term hemodialysis and demonstrates the efficacy of PVP-free membranes in managing similar allergic reactions.

Although delayed allergic-like reactions, such as the one observed in this case, have been reported in the past, their incidence in general clinical practice is extremely low. In fact, among over 500 hemodialysis cases performed at our institution over the past 10 years, this was the first reported case. This suggests that delayed allergic reactions can occasionally occur during long-term hemodialysis but are not well recognized in clinical practice.

The mechanism underlying the onset of symptoms in this case likely involved the release of histamine and other mediators, leading to increased vascular permeability, extravasation, decreased circulating blood volume, and consequent hypotension and gastrointestinal symptoms [12, 13].

Polyvinylpyrrolidone (PVP), a hydrophilicity-enhancing component used in polysulfone membranes, has been associated with hypersensitivity reactions—most commonly during the early stages of dialysis initiation [4, 8, 10]. These reactions are typically acute and often attributed to complement activation and histamine release.

However, delayed-onset reactions have also been reported after prolonged, repeated exposure to PVP, albeit less frequently. Such reactions are thought to be T-cell–mediated delayed-type hypersensitivity responses, potentially developing insidiously after long-term antigen exposure [11, 17]. In particular, leaching of PVP from the membrane surface may result in sustained low-level antigen stimulation, contributing to immune sensitization and delayed symptom onset [17].

The fact that symptoms in our case emerged after more than a year of stable dialysis is consistent with these mechanisms, supporting the possibility of a rare but clinically significant late-onset allergic reaction to PVP.

In this case, blood volume monitoring was a valuable diagnostic tool for the rapid detection of abnormalities, including allergic reactions. Despite the absence of ultrafiltration during the first 30 min of dialysis, a 15% reduction in blood volume was observed. Normally, during a four-hour dialysis session, blood volume gradually decreases by about 10–15%. Thus, the rapid and significant decrease in blood volume observed shortly after the start of treatment was considered abnormal. These findings suggested increased vascular permeability and enabled the early identification of an allergic reaction.

The patient’s blood test data showed a gradual increase in eosinophil count prior to the onset of symptoms. Notably, a marked rise in eosinophil count was observed starting one month before the onset, reaching over 13,000/μL at the peak.　Recognizing this eosinophilia as indicative of a delayed allergic reaction related to dialysis could have facilitated earlier detection and intervention. In this case, switching from a PS membrane to a CTA membrane normalized eosinophil levels. These findings suggest that CTA membranes may be an effective alternative to PS membranes in managing PS membrane-related allergies. Furthermore, in previous reports, the majority of membranes selected to address PS membrane-related allergies were CTA membranes, supporting their utility [10, 11, 15]. To assist in clinical decision-making, we have included a supplementary table listing representative hemodialysis membranes with and without PVP (Supplementary Table 1, 2).

One limitation of this case report is the absence of direct evidence, such as specific IgE antibody measurements or skin tests for PVP [9]. While the allergic reaction in this case was attributed to the dialysis membrane, we confirmed that the dialysis circuit itself did not contain PVP. However, because PVP is also used in a variety of other medical devices, clinicians should carefully assess all equipment used in patients with suspected PVP hypersensitivity. In this case, the clinical course and various laboratory findings support the possibility of a delayed allergic reaction caused by PVP.

Future studies should investigate the sensitization mechanisms and incidence of delayed allergic reactions to PVP in long-term dialysis patients. Moreover, establishing monitoring systems to enable the early detection and management of such complications will be essential.

In conclusion, this case demonstrates that PVP allergic reactions can occur even after extended periods of stable hemodialysis treatment, challenging the common assumption that such reactions primarily occur during initial exposure. The combination of blood volume monitoring and eosinophil tracking proved valuable for early detection and differential diagnosis, while the successful resolution of symptoms with a PVP-free cellulose triacetate membrane confirms this as an effective management strategy. As the population of long-term dialysis patients continues to grow, recognition of delayed PVP hypersensitivity and implementation of appropriate monitoring strategies could significantly enhance the safety and efficacy of chronic hemodialysis therapy.

Supplementary Information

Below is the link to the electronic supplementary material.

Funding

This study was not supported by any sponsor or funding.

Declarations

Conflict of interest

The authors have declared no conflict of interest.

Ethical approval

This case report did not require ethical approval as it does not involve new clinical interventions or patient experimentation. All data has been anonymized to ensure the patient’s privacy.

Informed consent

The patient’s informed consent could not be obtained; however, all data has been anonymized to ensure the patient’s privacy. No identifiable information is included in this case report.