Neurovesical dysfunction in children after treating pelvic neoplasms


G. Mosiello, Department of Paediatric Surgery, Bambino Gesù Children's Hospital, Piazza S. Onofrio 4, 00165 Roma, Italy.



To evaluate 10 years of experience, and thus define the occurrence and causes, of neurogenic lower urinary tract dysfunction in children with pelvic neoplasms treated by surgery.


From 1991 to 2000, 33 children were operated by the same surgeons for pelvic neoplasms; 11 were analysed, comprising four each with sacrococcygeal teratoma (ST) and ganglioneuroma, and one each with yolk sac tumour (YST), neuroblastoma and myofibroblastic bladder sarcoma (MBS). The other patients were not assessed because eight had died or were in severe progression, three were treated by bladder substitution and the others were lost to follow-up or refused a urological evaluation. All 11 children were evaluated at ≥ 6 months after surgery with a questionnaire about bowel and voiding habits, a neurological and orthopaedic assessment, a noninvasive urodynamic study, renal ultrasonography and spinal and pelvic magnetic resonance imaging (MRI). All patients with signs of bladder dysfunction were evaluated by a pressure-flow study. The results were analysed for surgical approach and anatomical involvement, i.e. group A, extensive surgery for complete tumour excision in the sacral area (ST and YST); group B, surgery for tumour resection in the paraspinal ganglia area (neuroblastoma and ganglioneuroma); and group C, bladder tumour with partial bladder resection (MBS).


Eight patients had signs or symptoms related to bladder sphincter dysfunction. One child refused the invasive urodynamic evaluation, leaving seven for analysis (two each ST and ganglioneuroma, one each YST, neuroblastoma and MBS). The urodynamic findings were normal in three children. On spinal and pelvic MRI a presacral lipoma with syringomyelia was discovered in one child with ST. Eight children had bladder dysfunction and two had no neurogenic damage (which was only in sacral tumours); in one child it was related to an upper motor neurone lesion from spinal dysraphism and in the other to a lower motor neurone lesion from surgical injury to the splanchnic nerves. Patients operated for paraspinal tumours had more bladder dysfunction but no signs of neurogenic damage, as did the patient with partial bladder resection. However in Group B, there may have been a transient or incomplete nerve injury in one patient.


Deficits of parasympathetic, sympathetic and somatic innervation of the bladder and the urethra may occur in children after surgery for pelvic neoplasms, related to minor or major surgical trauma. In ST, a tethered cord may be associated with mixed neurogenic damage. Knowledge of bladder dysfunction in anorectal malformations, spinal dysraphism, etc. and the clinical protocol used in these patients also seemed to be useful for understanding the development of voiding dysfunction in patients with neoplasm.


The most common pelvic neoplasms in children are rhabdomyosarcomas, sacrococcygeal teratomas (STs) and germ cell tumours. In the last few years, the survival for all types of childhood pelvic malignancy has improved dramatically with the influence of the combined children's cancer study groups [1]. Thus, particular attention must be given to the long-term quality of life of these patients, considering functional outcome for urological, digestive and orthopaedic sequelae.

Permanent neurological complications associated with tumour resection may occur because of the close relationship of the tumour to major nerve trunks, and the difficult surgical approach to deep vital structures in the pelvis. Currently, for these diseases where surgery remains the mainstay, the goal is not only survival but also to avoid ‘mutilating surgery’, recognising the importance of the anatomical, histological and genetic definition of the tumour. Neurogenic bladder dysfunction has been described previously in children with ST [2] and more recently in patients surviving pelvic neuroblastoma [3] and rhabdomyosarcoma [4].

The aim of the present retrospective study was to evaluate our 10-year experience, to define the incidence and causes of neurogenic lower urinary tract dysfunction in children with different pelvic neoplasms.


Between January 1991 and December 2000, 33 children with newly diagnosed pelvic neoplasms were treated in our institution by the same surgical team. Seven patients presented with rhabdomyosarcoma, six with ST, five with a yolk sac tumour (YST), four each with neuroblastoma and ganglioneuroma, one with ganglioneuroblastoma, and six with other sarcomas. The timing of surgery was defined in all children according to international or national children's cancer study groups, and further treatment was by chemotherapy, radiotherapy or brachytherapy. Patients with recurrent neoplasm were excluded.

Of the 33 patients only 11 were included, comprising four each with ST and ganglioneuroma, and one each with neuroblastoma, YST and myofibroblastic bladder sarcoma (MBS). Eight patients were excluded because they did not survive or had severe progression, three because they had a bladder substitution, and 11 as they were lost to follow-up or refused a urological evaluation. For ethical reasons and the severity of the primary disease, the urological evaluation comprised noninvasive procedures. All patients were evaluated ≥ 6 months after surgery with: (i) a questionnaire about LUTS (frequency, urgency, daytime incontinence, infections and nocturnal enuresis), urinary and bowel habits; (ii) an orthopaedic and urological evaluation including a neurological examination; (iii) a 3-day voiding frequency/volume chart completed at home; (iv) uroflowmetry (Pico Flow, Menfis Biomedical, Italy), perineal electromyography (EMG, surface electrodes, Menfis) associated with an estimate of postvoid residual urine by bladder ultrasonography (BVI 2500 Diagnostic Ultrasound Corporation Dux, WA, USA). The functional bladder capacity (FBC), calculated as the voided volume plus residual urine, was compared with the expected bladder capacity according to age, calculated as ((age in years + 2) × 30) [5]; (v) renal ultrasonography; (vi) spinal and pelvic MRI (with pelvic MRI already included in the oncological evaluation); (vii) other diagnostic procedures, e.g. voiding cysto-urethrography, DMSA or DTPA renal scintigraphy, included according to clinical status.

Patients with a voiding pattern suggestive of neurovesical dysfunction underwent a conventional pressure-flow urodynamic study (PFS; Pico Compact 2000 system, Menfis). The urodynamic techniques, definitions and units conformed to the standards recommended by the ICS [6]. The PFS were conducted after free voiding and residual urine evaluation. The detrusor pressure at filling and voiding pressure were recorded through a 6 F double-lumen transurethral catheter. Room temperature saline was infused at a low flow rate (5–10 mL/min), recording any uninhibited detrusor contraction during the filling phase. Abdominal pressure was recorded using a rectal balloon-catheter. The FBC and compliance were estimated at the end-filling pressure, calculated at maximum desire to voiding or at maximum filling sensation or continuous leak. The maximum flow rate and maximum detrusor pressure, and bladder residual, were recorded. During each PFS evaluation, at least two filling cycles were used.

Data were analysed considering the different surgical approach in relation to the anatomical involvement of the tumour: group A had extensive surgery for complete tumour excision in the sacral area (ST and YST); group B had tumour resection in the paraspinal ganglia area (neuroblastoma and ganglioneuroma); and the patient with MBS had a hemicystectomy.


Following the Italian Paediatric Haematology and Oncology Association, and Neuroblastoma Group, criteria [7,8], only the patient with YST had chemotherapy. No patients in group B or C had chemotherapy after surgery. Clinical features and the noninvasive findings at ≥ 6 months follow-up are shown in Table 1.

Table 1. Clinical, renal scintigraphy and uroflowmetry features of 11 children with pelvic neoplasms at ≥ 6 months of follow-up, with the PFS results for the seven children assessed at least 2 years after surgery
  1. UI, urge incontinence; GNM, ganglioneuroma; NB, neuroblastoma; Con, constipation; Abn, abnormal; Hydro, hydronephrosis; OA, overactive; HR, hyper-reflexia; DSD, detrusor sphincter dyssynergia; LC, low compliance; BT, bladder training; AC, anticholinergics.

Age, years    4    8    8    6  12    7  10    9    7    6    4
Bowel habitsNormalNormalNormalConConNormalNormalNormalNormalNormalNormal
Neurological/orthopaedic statusNormalNormalNormalAbnAbnNormalNormalNormalNormalNormalNormal
Bladder capacity, mL:
expected for age180270210210390240330300240210180
Max. flow rate, mL/ s    8.5  16  18  12    2  16  12  15    4  12.1   11
Flow time, s  20  24.8  42.1  32.4  22.4  25  22  27.7  44.2  24.8    9.5
Residual capacity  45  10    6  50320    0    5    5  60    5    0
Upper tractNormalNormalNormalNormalHydroNormalNormalNormalNormalHydroNormal
FBC, mL280300410300240230150
End-fill pressure, cmH2O    5  20  70  25    8    7  21
Max. flow rate, mL/s    9  13    2.5    4  14  15  13
Pdetmax, cmH2O  21  52  70  50  44  21  45
Residual, % FBC  19  52  85  24< 5    0    0
Detrusor statusNormalHRDSDAreflexiaLCOALCNormalOAOA/Normal

The orthopaedic evaluation was normal but patient 4 had a leg weakness and normal ambulation, while patient 5 had a lower limb flaccid palsy. The anocutaneous reflex and urethral sensation were normal in all patients except no. 4, with reduced anocutaneous reflex and urethral sensation, and no. 5, with absence of both. Patients 1 and 2 had a dysfunctional voiding pattern on uroflowmetry, with increased pelvic floor overactivity on EMG. Patient 5 had mild hydronephrosis, an abnormally high FBC, a markedly reduced flow rate and a significant postvoid residual volume; patients 3 and 4 only had the latter.

In group B, patients 8, 9 and 10 had urinary incontinence with normal bowel habits. All five patients had a normal anocutaneous reflex and urethral sensation. No changes in the ambulation status were recorded. Patients 8 and 10 had a low FBC for their age, and no. 9 had a significant postvoid residual with a low flow rate and increased flow time. Renal ultrasonography showed hydroureteronephrosis in patients 9 and 10.

The only patient in group C had an initial period of incontinence with a reduced FBC. Bowel habits, anocutaneous and urethral sensation, and ambulation status were normal. Spinal and pelvic MRI after surgery showed a presacral lipoma with syringomyelia in patient 4. Only seven patients underwent PFS, as they had suspected bladder dysfunction and patient 2 refused the invasive urodynamic evaluation; the results are also shown in Table 1; the urodynamic findings were normal in three children. Patient 4 had detrusor hyper-reflexia with sphincter dyssynergia related to an upper motor neurone lesion. This patient refused clean intermittent catheterization (CIC). Patient 5 had signs of a lower motor neurone lesion, with detrusor areflexia and a hypertonic and open bladder neck at filling (voiding cysto-urethrography showed a trabeculated bladder but no reflux). The lower motor neurone lesion remained stable during the follow-up and the patient started CIC. In group B there was detrusor overactivity with a low-compliance bladder patient 9 and slight signs of detrusor overactivity, presumably idiopathic, in patient 10. The last patient had detrusor overactivity that decreased with time.


In children the most common pelvic tumour is rhabdomyosarcoma, which represents 10–15% of all childhood solid tumours and 6% of all childhood cancers diseases [9]. The pelvis is the site of 78% of childhood teratomas, with 43% in the ovaries and 35% in the sacrococcygeal region [10]. Moreover, neuroblastoma may originate in the pelvis, with a reported incidence of 2–5.6%[11,12].

Children's cancer study groups over the last 20 years have contributed to increasing the survival rate in all these pelvic neoplasms. Indeed, the use of new therapeutic agents, a more precise and early diagnosis by CT and MRI that identifies the specific anatomical location and extent of disease, and better histological identification, have influenced the survival rate and reduced morbidity. These results show the importance of considering the long-term follow-up of these patients, focusing on morbidity and functional results for urological, digestive, neurological and orthopaedic sequelae. Primary operations may cause functional or anatomical disruption in pelvic structures. Paediatric surgeons know that the pelvis of newborns and infants is not fully developed and the bladder, uterus and ovaries are largely intra-abdominal, and thus require great care when tumours are resected, to avoid injuries to the major nerve trunks and deep vital structures. The resection of sacral tumours may be responsible for neurological deficits to lower limbs, and for urinary, fecal and sexual dysfunction. Bilateral preservation of the S2 nerves and above is necessary for walking, and bilateral preservation of S3 and above to spare urinary, fecal and sexual functions [13].

Pelvic visceral surgery may cause damage to the pelvic splanchnic nerves, the hypogastric nerves, and the pelvic nerve plexus, resulting in autonomic denervation of the lower urinary tract. Depending on which neural structures are injured, different patterns of dysfunction have been examined. Injury to the pelvic splanchnic nerves (parasympathetic denervation), originating from S2–S4 on the posterolateral pelvic wall, may cause an areflexic bladder. Injury to the hypogastric nerves (sympathetic denervation), with their origin from the hypogastric plexus at the aortic bifurcation and running down on each side of the pelvis to the pelvic plexus, may result in loss of bladder compliance, and incompetence of the bladder neck and posterior urethra. Damage to the pelvic plexus may result in mixed pattern of vesico-urethral dysfunction [14].

There are no reports of somatic denervation caused by pudendal nerve injury in pelvic paediatric surgery, in contrast to numerous reports in adults, where colorectal surgery requires more radical excision [15,16]. Urinary dysfunction could be caused by temporary nerve injury related to traction, diathermy injury, or incomplete division of nerves that later regenerate [17].

In adults, urological sequelae after pelvic tumour excision are well described [14,15,18], while in children there are only recent follow-up studies. Boemers et al.[19] evaluated11 children with benign ST, showing that the incidence of voiding dysfunction was high and not only usually neurogenic in origin but also associated with the presence of a tethered cord. Yeung et al.[4] evaluated 11 patients operated for pelvic rhabdomyosarcoma and showed how important it was that these children had frequent long-term monitoring of bladder and upper urinary tract function.

More recently Crucetti et al.[3] retrospectively reviewed 17 children with pelvic neuroblastoma, suggesting that the survival rate of children with nonmetastatic pelvic neuroblastoma was good despite incomplete tumour resection, and focused on the high incidence of permanent neurological damage after surgery, including sciatic nerve palsy, urinary and fecal incontinence, leg weakness and neuropathic bladder.

From these earlier reports we evaluated children with different types of pelvic neoplasm to define the usefulness of following the upper and lower urinary tract in different tumour sites (sacral, paraspinal, bladder). In all children we started with a noninvasive evaluation during the oncological follow-up, using a questionnaire, a voiding chart and uroflowmetry. Renal ultrasonography incurred no extra cost because it was already scheduled in the oncological follow-up. Only patients suspected of lower urinary tract dysfunction had an invasive urodynamic study. We decided to use MRI in all patients, especially those with anorectal malformations, to detect spinal dysraphism [20,21]. CT or MRI of the pelvic area was already scheduled in the oncological follow-up, so we chose to use MRI in patients with bladder dysfunction to define the spinal cord status. We suggest that spinal cord MRI is included in the oncological follow-up in all patients undergoing surgery for pelvic tumours, even though only one patient had an occult spinal lesion.

There was bladder dysfunction in eight of the present patients and two had clear neurogenic damage, occurring only in sacral tumours; in one it was related to an upper motor neurone (spinal dysraphism) and in the other to a lower motor neurone lesion (surgical injury to the splanchnic nerves). Patients operated for paraspinal tumours had more bladder dysfunction but no signs of neurogenic damage, as had the patient with partial bladder resection. However, in group B there was possibly a transient or incomplete nerve injury in patient 9.

We recommend an early noninvasive urodynamic evaluation in all patients with neoplastic pelvic disease, and when lower urinary tract dysfunction is suspected it is mandatory to also use PFS. For treatment we stress the importance of using bladder training and a voiding diary, either for a rehabilitation programme or to detect voiding dysfunction. When there is also bladder overactivity, drug therapy might be useful; CIC, when indicated, must be started as early as possible. We have no experience with α-blockers in these patients, but think that their use could be necessary in patients with functional obstruction where neurogenic damage has been excluded.

Previously there was a careful urological evaluation in patients with anorectal malformation and occult spinal dysraphism, and we consider that a similar protocol must be defined for all types of pelvic neoplasms in children, with an early urological evaluation, including MRI and PFS, to start appropriate urological treatment and avoid upper urinary tract deterioration. This also requires a multidisciplinary approach from the initial diagnosis, involving surgeons, oncologists, urologists, urotherapists and psychologists, to clearly define the development of bladder dysfunction.


sacrococcygeal teratoma


yolk sac tumour


myofibroblastic bladder sarcoma




functional bladder capacity


pressure-flow urodynamic study


clean intermittent catheterization.