Herpes zoster-associated voiding dysfunction in hematopoietic malignancy patients


  • Shinichi Imafuku MD, PhD,

    1. From the Division of Dermatology and Hematology, Hiroshima Red Cross and Atomic Bomb Survivor's Hospital, and Department of Dermatology, Graduate School of Medical Sciences, Kyushu University
    Search for more papers by this author
  • Masakazu Takahara MD,

    1. From the Division of Dermatology and Hematology, Hiroshima Red Cross and Atomic Bomb Survivor's Hospital, and Department of Dermatology, Graduate School of Medical Sciences, Kyushu University
    Search for more papers by this author
  • Takeshi Uenotsuchi MD,

    1. From the Division of Dermatology and Hematology, Hiroshima Red Cross and Atomic Bomb Survivor's Hospital, and Department of Dermatology, Graduate School of Medical Sciences, Kyushu University
    Search for more papers by this author
  • Koji Iwato MD, PhD,

    1. From the Division of Dermatology and Hematology, Hiroshima Red Cross and Atomic Bomb Survivor's Hospital, and Department of Dermatology, Graduate School of Medical Sciences, Kyushu University
    Search for more papers by this author
  • Masutaka Furue MD, PhD

    1. From the Division of Dermatology and Hematology, Hiroshima Red Cross and Atomic Bomb Survivor's Hospital, and Department of Dermatology, Graduate School of Medical Sciences, Kyushu University
    Search for more papers by this author

Shinichi Imafuku, md, phd 3-1-1, Maidahsi Fukuoka city 812-8582 Japan
E-mail: dermatologist@mac.com


Background  Voiding dysfunction is a rare but important complication of lumbo-sacral herpes zoster. Although the symptoms are transient, the clinical impact on immunocompromised patients cannot be overlooked.

Methods  To clarify the time course of voiding dysfunction in herpes zoster, 13 herpes zoster patients with voiding dysfunction were retrospectively analyzed.

Results  Of 13 patients, 12 had background disease, and six of these were hematopoietic malignancies; four of these patients were hematopoietic stem cell transplant (HSCT) recipients. Ten patients had sacral lesions, two had lumbar, and one had thoracic lesions. Interestingly, patients with severe rash, or with hematopoietic malignancy had later onset of urinary retention than did patients with mild skin symptoms (Mann–Whitney U analysis, P = 0.053) or with other background disease (P = 0.0082). Patients with severe skin rash also had longer durations (P = 0.035). In one case, acute urinary retention occurred as late as 19 days after the onset of skin rash.

Conclusions  In immune compromised subjects, attention should be paid to patients with herpes zoster in the lumbo-sacral area for late onset of acute urinary retention even after the resolution of skin symptoms.


Herpes zoster is a common skin disease in dermatologic medicine. Complications of herpes zoster are dependent on the organs adjacent to the involved skin area. For example, Hunt syndrome is widely known as zoster lesions in the auricle with hemifacial palsy due to compression or injury of the facial nerve.1 Herpes zoster is also a rare cause of voiding dysfunction.2 It is partly because herpes zoster in the sacral area is infrequent. Chen et al. reported that only 8% of 423 herpes zoster patients had sacral lesions. However, 82.3% of the sacral zoster patients had bladder dysfunction,3 so clinicians should anticipate voiding problems when seeing those patients.

Voiding dysfunction caused by herpes zoster is usually transient and often dissolves without additional treatment.3 However, the clinical impact of unexpected symptoms in immunocompromised patients such as hematopoietic malignancy patients cannot be overlooked.4,5 Therefore, it is important to be aware of possible bladder symptoms when we see lumbo-sacral herpes zoster patients. Nevertheless, few published reports have focused on the time course of the herpes zoster-associated voiding dysfunction.

Thirteen herpes zoster patients, including six hematopoietic malignancy patients with voiding dysfunction, have been hospitalized at our hospital. We report here the analysis of onset dates, durations of voiding dysfunction symptoms and present a case.

Patients and Methods

Herpes zoster patients with bladder symptoms (dysuria and urinary retention) were examined at the Dermatology Branch, Hiroshima Red Cross and Atomic Bomb Survivors Hospital from April 2000 to March 2003. Items analyzed were age, sex, background diseases, involved ganglia, severity of the skin rash (skin score), onset of voiding dysfunction (days after the appearance of skin rash), duration of voiding dysfunction (days), and treatment required for voiding dysfunction. Severity of the skin rash was scored as described by Nagasako et al.6 When the lesions fused and individual lesions were difficult to be identified, the score was determined by the area occupied by the eruption: Score 1 for < 30%, score 2 for 30–70%, score 3 for > 70% in a responsible dermatome. These data were not parametric; therefore, the differences between the groups were statistically analyzed by nonparametric analysis (Mann–Whitney test).


Patient profiles

Thirteen patients were included in the study, eight men and five women, and their profiles are summarized in Table 1. Five patients were referred to us from internists, two from urologists, one from a gynecologist, and the rest of the patients directly visited our dermatology clinic. Their ages ranged from 24 to 81 years with a median age of 58 years.

Table 1.  Summary of patient profiles and voiding dysfunction
CaseAgeSexBackground diseaseDermatomeSkin ScoreDisseminationTime to onset of VDDuration of VDCatheterization
  1. Abbreviations: ALL, acute lymphatic leukemia; AML, acute myelogenous leukemia; BMT, bone marrow transplantation; DM, diabetes mellitus; LC, liver cirrhosis; ML, malignant lymphoma; RA, rheumatoid arthritis; SjS, Sjoegren syndrome; ITP, idiopathic thrombocytopenic purpura; VD, voiding dysfunction.

128MALL, BMT day 180S3–53+1918+
246MAML, BMT day 123L11 1 6 
373Malcoholic LCL12 620+
455MDMS21 0 7
524FAML, BMT day 540S23+ 617+
681MML, remissionS3–53+1052+
726MAML, BMT day 126S23+10 8+
878FRA, SjSS3–53–212+
952FITPS21 0 6
1062FCervical Cancer, post-opS21 0 4
1158MMLTh83+ 613+
1271MnoneS22 3 7
1379FInterstitial pneumonitisS22–412+

Twelve out of 13 patients had background diseases. Of these 12, seven patients had hematologic disease, and six had malignancies: four, acute leukemia; two, malignant lymphoma. All the leukemia patients had received hematopoietic stem cell transplants (HSCT). Two of them had chronic graft vs. host disease (cGVHD), the rest were without cGVHD.

Of 13 patients, 77% (10/13) had skin rash in sacral dermatomes, 15% (2/13) had lumbar lesions, and 8% (1/8) had a thoracic lesion. The patient with the thoracic lesion probably had myelitis.

Skin symptoms were scored in three grades as described above. Forty-six per cent of patients (6/13) had a score of 3; 23% (3/13) had a score of 2; and 31%4 had a score of 1. Eighty-three per cent (5/6) of score 3 patients had disseminating eruptions; 83% (5/6) of hematopoietic malignancy patients had score 3.

All patients received antiviral chemotherapy [intravenous injection of acyclovir (ACV)]. Sixty-two per cent (8/15) of patients needed catheterization for complete urinary retention. All of the patients with skin score 3 needed catheterization; 67% (2/3) of score 2 patients needed it; whereas none of the score 1 patients required catheterization. It is noteworthy that complete urinary retention was observed in all patients with score 3.

Voiding dysfunction

The time to onset of urinary retention ranged from –4 to 19 days after appearance of skin rash, the mean period was 4.2 ± 6.0 SD and the median time was 3 days. The durations ranged from 4 to 52 days; the average was 14 ± 12 SD; the median was 12 days (Table 1).

Interestingly, hematopoietic malignancy patients (n = 6) had significantly later onsets (Mann–Whitney test, P = 0.0082) than those with other diseases (n = 7) (Fig. 1a). No significant difference was observed between durations in the two groups (P = 0.082) (Fig. 1b).

Figure 1.

Onsets and durations of voiding dysfunction of the patients with different background diseases. Bars represent the means. Note the later onsets of the HM patients. (a) Onsets; (b) durations. HM, hematopoietic malignancies

Voiding dysfunction tended to happen later in higher skin score patients, but the P-value was slightly less significant (P = 0.053). Durations in score 3 patients were longer than those in patients with scores 1 and 2 (P = 0.035). Patients who needed catheterization for the treatment of urinary retention (n = 8) had onsets from –4 to 19 days (median = 6 days), whereas patients who did not needed catheterization had onset from 0 to 3 days (median = 0 days). The medians of the two groups were not statistically different (P = 0.17).

A case report

A 26-year-old Japanese man had noticed a painful erythema with blisters on the left buttock, which was diagnosed as herpes zoster of the sacral ganglia (S3–5). He also had fever (38.5 °C) and disseminating eruptions. He had received a bone marrow transplant (BMT) after successful chemotherapy for acute lymphatic leukemia and the eruption occurred at day 180 post-BMT. Intravenous administration of ACV (1500 mg/day) in combination with prophylactic antibiotics caused the fever to decline. Nineteen days after the beginning of the rash, when most of the blisters became crusted, the patient complained of abdominal distention and sudden inability to void, which he had not complained about before. After the patient failed to void spontaneously, a catheter was inserted and > 1000 ml of retained urine was recovered. A urological examination excluded other causative factors in the voiding dysfunction. The patient was able to void spontaneously after 18 days of indwelling catheterization.


In our study, the time to onset of voiding dysfunction ranged widely from –4 to 19 days after the emergence of the eruption. As shown in a case report,, urinary retention could occur when the skin wound had almost healed during intravenous ACV. The skin symptom of the presented case was very severe. Therefore, we assumed that the onset of urinary retention did not necessarily occur early in severe sacral herpes zoster and focused on the onsets of voiding dysfunction in other patients. The outcome was that severe skin rash or an immunocompromised state caused voiding dysfunction to occur later (Figs 1 & 2).

Figure 2.

Clinical course of a case. Arrow means intermittent catheterization. ACV, acyclovir

Herpes zoster-associated voiding dysfunction occurs in connection with two different pathogeneses, namely cystitis and neurogenic bladder. A combination of both conditions could also happen. Cystitis associated with herpes zoster is ipsilateral hemicystitis, which usually occurs simultaneously with skin rash.7 Patients with herpes cystitis may present with dysuria, urinary frequency or retention, pyuria, and hematuria. The clinical course of cystitis usually follows the course of the skin rash, which also should represent the viral infection of the bladder epithelium (Fig. 3).

On the other hand, urinary retention is mainly caused by neurogenic bladder. The most frequent cause of herpes-associated neurogenic bladder is interruption of the detrusor reflex due to involvement of the sacral dorsal root ganglia.8 We considered that late onset of urinary retention may be explained by prolonged viral replication in the neurons of immunocompromised patients. A prolonged course of varicella-zoster virus (VZV) infection, even with ACV treatment, has been reported in hematopoietic malignancy9 and HSCT patients.10 Aizawa et al. reported that the onset of facial paralysis ranged from –27 to 13 days after onset of the skin rash, and they detected VZV-DNA by the polymerase chain reaction in the saliva from –12 to 8-day samples.1 These reports illustrate that the skin blisters are one of the various symptoms of endogenous VZV reactivation and that the viral replication period does not always correlate with the skin symptoms. This hypothesis does not explain why the hematopoietic malignancy patients had later onsets. One possible answer to this question is that in seriously immunosuppressed patients, inflammatory process that causes the neuronal impairment may be slow and weak, resulting in the later onset of voiding dysfunction. Some of the neurologic complications in immunosuppressed patients such as chronic VZV encephalitis occur months after an episode of herpes zoster.11 Another possibility is that earlier and more intense ACV treatment in hematopoietic malignancy patients may have delayed, but not completely prevented, the progression of the neuronal damage by VZV. On the other hand, recovery of the voiding function was also delayed in immunosuppressed patients, as shown in Fig. 1(b). Studies with larger numbers of cases will be needed to further clarify the relationship between herpes zoster and the kinetics of voiding dysfunction.

In the course of the immune reconstruction after HSCT, herpes virus infection occurs frequently, and prophylactic antiviral chemotherapy is not completely effective.12,13 As the prognosis of the bladder symptoms secondary to herpes zoster is benign, it is important to diagnose the symptoms early. One needs to be careful about the late onset of the urinary retention, as well as preceding voiding dysfunction without zosteriform eruption.