Evaluation of genitourinary fistulas in pelvic malignancies with etiopathologic correlation: role of cross sectional imaging in detection and management

Abstract

Objective:

Genitourinary fistulas in pelvic malignancies are abnormal communications occurring due to either locally advanced tumours invading the surrounding organs or post-therapeutic complications of malignancies. In this article we review and describe the role of cross-sectional imaging findings in the management of genitourinary fistulas in pelvic malignancies.

Methods:

A retrospective study, for the period January 2012 to December 2018, was undertaken in patients with pelvic malignancies having genitourinary fistulas. The cross-sectional (CT and MRI) imaging findings in various types of fistulas were reviewed and correlated with the primary malignancy and the underlying etiopathology.

Results:

Genitourinary fistulas were observed in 71 patients (6 males, 65 females). 11 types of fistulas were identified in carcinomas of cervix, rectum, ovary, urinary bladder, sigmoid colon, vault, endometrium and prostate. The commonest were rectovaginal and vesicovaginal fistulas. 13 patients had multiple fistulas. The sensitivity, specificity, positive and negative predictive values of CT and MRI are 98%, 100%, 66%, 98% and 95%, 25%, 88% and 50% respectively. Contrast-enhanced CT with oral and rectal contrast is more sensitive and specific than MRI in the evaluation of genitourinary fistulas.

Conclusion:

Imaging findings significantly influence the management and outcome of genitourinary fistulas in pelvic malignancies. Contrast-enhanced CT is the imaging modality of choice in the evaluation of pelvic fistulas associated with malignancies and MRI is complimentary to it.

Advances in knowledge:

To our knowledge, this study is the first of its kind wherein the mean duration of occurrence of fistulas in pelvic malignancies is correlated with the underlying etiopathology.

Introduction

Fistula is defined as an abnormal communication between two epithelial surfaces. Fistulas are more frequent between pelvic organs compared to other sites of body as the genitourinary and gastrointestinal tracts are closely apposed within the narrow confines of bony pelvis. Currently, the major causes of genitourinary fistulas in pelvis are surgeries, post-radiation, inflammatory bowel diseases, infections and malignancies.^1–3

Multimodality treatment regimens with neoadjuvant chemotherapy, radiation, and surgery are increasing the survival rates in cancer patients with subsequent increase in the occurrence of post-therapeutic complications.⁴ Fistulas are dreadful and distressing complications in cancer patients with a notable impact on the overall survival and quality of the residual life.⁵ Genitourinary tract fistulas associated with malignant aetiology occur due to direct infiltration of organs by the tumour, or as post-surgical, chemotherapy and radiation complications. Fistulas are most commonly seen in cervical, colorectal, ovarian and bladder neoplasms with abnormal communications between the urinary bladder, vagina and rectum. Infrequently fistulous communications to colon, small bowel loops, uterus, cervix and ureters may also occur.

Post-surgical fistulas can occur in immediate post-operative period within 6 weeks after pelvic surgeries, while, recurrent fistulas can develop within 3 months of primary fistula repair.⁶ Higher incidence of fistulas is reported with hysterectomies, laparoscopic surgeries and major oncological pelvic surgeries.^6–8 Urinary bladder and ureters are the commonly involved structures in post-operative fistulas as they are highly susceptible to ischaemia and avascular necrosis. Chemotherapy with intravesical instillation of mitomycin C was also reported to be associated with fistula formation.⁹ Fistulas secondary to pelvic irradiation can occur any time after 6 months to 20 years after radiation therapy. They are often large in size (>10 mm), multiple (two or more) and complex with low likelihood of spontaneous closure.¹⁰

Management of fistulas in malignancies is completely different from that of fistulas in benign conditions and hence these cases require extensive imaging and follow-up. Detection of a fistula upstages the disease in many pelvic malignancies, as any primary neoplastic mass invading adjacent structure or organ with fistulous communication is staged as either T4a (urinary bladder, uterus, cervix and vagina) or T4b (rectum).¹¹ Comprehensive work-up of the underlying condition is needed as these fistulas always require a curative or palliative surgical intervention. The approach and choices of management are limited as adjuvant therapies are delayed or not possible in the presence of fistula.^5,11 The best curative option is an en bloc tumour-fistula resection, which is not feasible in many patients due to advanced, unresectable status of malignancy.^5,12 Most of these patients are also poor surgical candidates due to older age, associated comorbid conditions, including superadded infection or abscess in about 40% of cases.^5,10,13 Primary surgical repair is not possible in malignant fistulas with extensive destruction and loss of soft tissue. Repair is also not undertaken in the presence of viable tumours as in recurrent and residual neoplasms. Post-radiation fistulas are associated with high failure rates of surgical repair due to ischaemic, non-viable surrounding tissues, poor healing and increased risk of dehiscence.^5,10,14 Therefore, in majority of the cases, management of fistulas involves controlling the infection, drainage of abscess, palliative procedures such as ureteric stenting and urinary/faecal bypass or diverting ostomies for immediate relief.⁵

The purpose of this article is to describe the needs of clinician and importance of imaging in the management of genitourinary fistulas in pelvic malignancies.

Methods and materials

With the approval of the institutional review board at our hospital, Basavatarakam Indo American Cancer Hospital and Research Institute, Hyderabad, India, one of the largest tertiary care cancer research institutes in the country, a retrospective study was undertaken in patients with pelvic malignancies having genitourinary fistulas. The data for the period January 2012 to December 2018, from radiology, surgical and medical records of patients with genitourinary fistulas were reviewed. Surgically confirmed cases of genitourinary tract fistulas due to advanced neoplasms and fistulas occurring secondary to post-therapeutic complications (surgery, radiotherapy, and chemotherapy) were included in the study. Patients with incomplete medical/surgical/radiotherapy records, cases of fistulas due to non-neoplastic aetiology and cases without CT and/or MR imaging of the fistula were excluded from the study. Plain and contrast-enhanced CT imaging (Siemens, AS 128 Slice CT machine, Germany) was performed with both intravenous contrast (for assessing tumours, areas of inflammation and fibrosis), and oral/rectal contrast (to differentiate bowel loops from extraluminal masses, collections, abscesses and enlarged lymph nodes). Excretory phase and delayed imaging with full bladder, “CT urogram,” was acquired in all cases suspected of lower ureteric or vesical fistulas. Thin section (1 mm) image reconstruction was done in axial, sagittal and coronal planes. MRI (GE HDE 1.5T MRI machine, GE Health care, USA) of abdomen was performed with phased-array coils using fast spin echo T₁W, T₂W, fat suppression and contrast enhanced (using intravenous gadolinium) sequences. Thin section images of pelvis in oblique axial, coronal and sagittal planes were acquired. All the images were reviewed by two radiologists with more than 15 years of experience. Primary CT or MRI criterion for the diagnosis of fistula was visualisation of fistula involving genitourinary organs with an air- or fluid-filled abnormal fistulous tract with or without contrast (Figure 1a–d). On MRI, the site of fistulisation was seen as focal thickening or disruption of the hypointense muscularis externa present in the walls of urinary bladder, bowel and vagina. Ancillary findings observed were the presence of air or fluid or contrast in inappropriate location, presence of malignant mass, loss of fat planes between the organs, post-operative changes and post-radiation changes (Figure 1e). The mean period of the appearance of fistula after the diagnosis of malignancy and/or post-therapeutic measures was also reviewed. Statistical analysis was done by using MS office Excel and the parameters-sensitivity, specificity, positive and negative predictive values along with the diagnostic accuracy of CT and MRI were obtained.

Figure 1.

(a) Sagittal T₂W fat saturated image shows complex air (thick arrows) and fluid filled (thin arrow) rectovaginal and vesicovaginal fistulas. (b) Sagittal T₂W image shows fluid filled vesicovaginal fistula (arrow). (c, d) Sagittal reconstructed delayed CECT images show contrast-filled vesicovaginal fistula (arrow) and vesicocutaneous fistula with fibrosis (arrow). (e) Sagittal reconstructed CECT image shows air and contrast in inappropriate location (arrows), malignant mass (M) with loss of fat planes between the organs. CECT, contrast-enhanced CT.

Results

Genitourinary fistulas were seen in 71 patients (6 males, 65 females) with the mean age being 52 years at the time of diagnosis (age range 26–78 years), majority being females (male: female ratio of 1:11). The distribution of fistulas in various pelvic malignancies is given in Table 1. In our study, 52 cases underwent CT, 27 cases underwent MRI and 14 cases underwent both CT and MRI. We found that CT is more sensitive and specific in the detection of fistulas than MRI with a positive predictive value of 100%. There was substantial interobserver agreement (κ coefficient 0.61–0.80) between the radiologists and the statistical parameters observed in our study are given in Table 2. 11 types of fistulas were identified in our study with the commonest malignancies associated with fistulas being cervix (36%) and rectum (28%). The commonest fistulas seen in our study were rectovaginal fistulas (Figure 2) followed by vesicovaginal fistulas (Figure 3). In 13 patients (18%), more than one fistula was seen and these were termed as complex fistulas (Figure 4). Two patients had three fistulas each (Figure 5). Out of the 13 patients with multiple fistulas, 10 were synchronous and 3 were metachronous. One patient of carcinoma sigmoid colon presented with colovaginal fistula, underwent surgery and radiotherapy and 4 years later developed rectovesical and rectovaginal fistulas. Another patient with post-operative recurrence of carcinoma rectum had metachronous enterovesical and rectourethral fistulas which occurred within a span of 1 month. The third case was post-operative case of cervix which developed ureteroenteric fistula on 10^th post-operative day and complex vesicovaginal, rectovaginal fistulas after a month (Figure 6). The relative frequency of occurrence of fistulas with underlying etiopathogenesis is given in Table 3. The correlation of the etiopathogenesis with the mean duration of occurrence of fistula is given in Table 4.

Table 1.

Distribution of types of fistulas in various pelvic malignancies

Fistula type	Cervix	Rectum	Urinary bladder	Ovary	Sigmoid colon	Vault	Endometrium	Prostate
Rectovaginal	11	9		3
Vesicovaginal	5	2	3	2		3
Ureterovaginal	2	1		1			1
Colovesical					3
Rectovesical		2						1
Colovaginal	1	1		1
Enterovesical		1				1
Enterovaginal		1
Ureterocolic				1
Rectourethral		1
Vesicocutaneous			1
Complex	7	2	3		1
Total (%)	26 (36.6%)	20 (28.1%)	7 (9.8%)	8 (11.2%)	4 (5.6%)	4 (5.6%)	1 (1.4%)	1 (1.4%)

Table 2.

Statistical parameters of CT and MRI observed in our study

Statistical characteristic	CT with 95% confidence intervals	MRI with 95% confidence intervals
Sensitivity	98% (89%–99%)	95% (78%–99 %)
Specificity	100% (15%–100%)	25% (0%–80%)
Positive predictive value	100%	88% (80%–92%)
Negative predictive value	66% (22%–93%)	50% (7%–92%)
Accuracy	98% (89%–99%)	85% (66%–95%)

Figure 2.

52-year-old female with recurrence of carcinoma cervix. (a) Oblique abdominal radiograph shows rectovaginal fistula (arrow) between rectum (R) and vagina (V). (b, c) Axial CT images of pelvis show the rectovaginal fistula with air and contrast (arrows) between the rectum (R) and vaginal mass (V). (d) Sagittal reconstructed CT image of the pelvis shows the rectovaginal fistula (black oval) with air and contrast in the vaginal mass (V) and rectum (R).

Figure 3.

48-year-old female, post-operative case of carcinoma ovary with urinary incontinence. (a, b) Axial and sagittal reconstructed CT cystogram images of the pelvis show contrast in UB, vagina (V), and vesicovaginal fistula (black oval) between the posterior wall of UB and the anterior wall of vagina. UB, urinary bladder.

Figure 4.

70-year-old female, post chemoradiation with recurrence of cervical carcinoma involving the entire uterus. (a) Axial CECT image with oral contrast shows contrast in the uterine cavity (U), enterouterine fistula (black oval) on the right side with ileal loop (I) and colouterine fistula (black rectangle) on left side with sigmoid colon (C). (b) Sagittal reformatted CECT image shows contrast in the bowel loops (arrows), rectum (R) and uterus (U). CECT, contrast-enhanced CT.

Figure 5.

62-year-old female, with carcinoma urinary bladder. (a, b) Coronal reformatted and axial CT images reveal a colovesical fistula (black oval) between the sigmoid colon (arrow) and UB and vesicovaginal fistula (black circle) between the UB and vagina (V). (c) Sagittal reformatted CT image shows all the abnormalities. Air in the UB (white arrow), colovesical fistula (black oval) a large vesicovaginal fistula (black triangle) and a rectovesical fistula (black rectangle). UB, urinary bladder.

Figure 6.

46-year-old female post-operative case of carcinoma cervix. (a) Axial CECT of pelvis without oral contrast shows the uretoenteric fistula (white oval) between the right lower ureter (thin arrow) and adjacent bowel loop (thick arrow). (b, c) Axial and sagittal reformatted CECT images show vesicovaginal fistula (black rectangle), rectovaginal fistula (black oval), air (arrow) and contrast in UB with tracking of contrast into the vaginal stump and rectum (R). CECT, contrast-enhanced CT; UB, urinary bladder.

Table 3.

Frequency distribution of types of fistulas according to the etiopathogenesis

Fistula type	Primary infiltrating lesion	Post RT	Post OP	Post RT/OP	Post RT Recurrence	Post OP Recurrence	Post RT/OP Recurrence
Rectovaginal	7	8	2	3	2	1
Vesicovaginal	8	3	2	2
Ureterovaginal	1		4
Colovesical	1					1	1
Rectovesical	1			1			1
Colovaginal		1		2
Enterovesical						1	1
Enterovaginal				1
Ureterocolic	1
Rectourethral			1
Vesicocutaneous			1
Complex	4	1	2	1	1	3	1
Total (%)	23 (32.4%)	13 (18.3%)	12 (16.9%)	10 (14.1%)	3 (4.2%)	6 (8.4%)	4 (5.6%)

(RT-Radiotherapy, OP-operative)

Table 4.

Etiopathogenesis versus Duration of occurrence of fistula

S. no	Aetiology of fistula (number of cases)	Occurrence of fistula after diagnosis of malignancy/treatment	Mean duration
1	Primary infiltrating lesion (23)	0–2 months	Less than 1 month
2	Post-radiotherapy (13)	3–129 months	30 months
3	Post-operative (12)	0–19 months	17 days
4	Post-radiotherapy, post-operative (10)	4–57 months	22 months
5	Post-radiotherapy recurrence (3)	6–35 months	21 months
6	Post-operative recurrence (6)	1–12 months	6 months
7	Post-radiotherapy and post-operative recurrence (4)	10–91 months	37 months

Discussion

Historically various non-imaging techniques such as instillation of dyes like methylene blue, oral charcoal test, poppy seed test, Bourne test were used by clinicians to confirm the presence of fistulas but none of them were accepted as standard tests.^3,15–17 Though endoscopy and cystoscopy are a part of routine work-up in the evaluation of fistulas, they fail to show fistula in more than 50% of the cases and most often the findings are non-specific and inconclusive.^18,19

The advantages and disadvantages of various imaging modalities in the detection and evaluation of fistulas in malignancies are given in Table 5. The sensitivity of plain radiography in the detection of fistulas is as low as 0–5%^3,16 Occasionally, the presence of air in the bladder may be an indirect sign of enterovesical fistulas. Contrast studies like excretory urography (IVU), cystography, vaginography, fistulography and barium follow through /enema have all been used in the assessment of fistulas with varying success rates.^2,18,20 Even though they are inexpensive, safe and easy to perform, a significant limitation of contrast examinations in malignancies is the lack of mural and extraluminal visualisation. Ultrasonography is commonly used in abdominal and pelvic evaluation but diagnosing fistulas on ultrasound is extremely rare and limited to few case reports of vesicouterine and rectovaginal fistulas.^21–24

Table 5.

Advantages and disadvantages of various imaging modalities in the evaluation of fistulas of genitourinary tract in pelvic malignancies

Imaging modality	Advantages	Disadvantages
Excretory urography	Functional imaging Identification of urinary obstruction, leaks & fistulas	Needs intravenous contrast Onlyfistulas involving the urinary tract can be evaluated
Conventional cystography	Detection of vesical fistulas and leaks	Concomitant ureteric and other fistulas may be missed
Hysterography/ Vaginography	Evaluation of Uterine and vaginal fistulas	Fistulas other than uterine and vaginal cannot be evaluated
Fistulography	Fistulas with external cutaneous openings are seen	Difficulty in localising internal opening Non-visualisation of fistulas with occluded tracts and branch rami
Barium meal follow through / Barium enema	Detection of enteric and colorectal fistulas	Contrast does not pass into distal bowel loops in obstructive lesions Long examination times Difficulty in holding the rectal contrast in sick patients and incompetent sphincters
Ultrasound	Inexpensive, safe Easily available No ionising radiation.	Operator dependent Limited by body habitus and urinary bladder distension Frequent non-visualisation of fistulas
CT	Availability and accessibility Guided procedures are possible	Ionising radiation Requires contrast material Relatively expensive Contraindicated in patients with iodine sensitivity
PET scan	Recommended for the staging of underlying malignancy	Poor image resolution Expensive, limited availability and accessibility
MRI	No ionising radiation Excellent soft tissue resolution Multiplanar imaging	Expensive, limited availability and accessibility Lengthy imaging protocols May need sedation and i.v. contrast Contraindicated in patients with certain metallic implants & pace makers Artefacts from respiration, peristalsis, bowel obstruction and ascites

PET, positron emission tomography.

The major advantage of multiplanar cross-sectional imaging with reconstructed CT or MRI is their ability to localise the fistula in three dimensions and identifying the underlying aetiology. This is crucial in the evaluation as the location of the soft tissue mass and/or fistula, size and extent of fistulas determine the surgical procedure and the approach.¹⁹

Vesical fistulae secondary to urinary bladder, uterus, sigmoid colon and rectal malignancies usually occur at the dome or the posterior and left lateral aspects of the urinary bladder. Terminal ileal, caecal and appendiceal pathologies commonly involve the right lateral or anterior aspects of the bladder.^19,25 Vesicovaginal and rectovaginal fistulas invariably involve the posterior wall of urinary bladder and vagina respectively. The level of posterior vaginal wall involvement categorises rectovaginal and sigmoidovaginal fistulas into three types. Most common are the low rectovaginal fistulas, located at the lower one-third of the vagina, usually secondary to obstetric trauma or anorectal and vaginal surgeries. Radiation-induced rectovaginal fistulas characteristically occur between the posterior fornix of the vagina and the middle-third of the rectum directly communicating through the rectovaginal septum.¹⁹ High rectovaginal fistulas are seen at the level of pouch of Douglas between the sigmoid colon or rectum and the peritonealised portion of the vagina, the common aetiology being pelvic procedures, neoplasms and inflammatory diseases.^19,26,27

Majority of rectovaginal fistulas (94%) in our study were seen in the upper two-third of the vagina. We found that sagittal planes providing profile views are best for visualising vesicovaginal, rectovaginal, and low positioned fistulas, while, lower ureteric, enterovesical and colovesical fistulas were better visualised on coronal sections (Figure 7).

Figure 7.

42-year-old female, a case of cervical carcinoma, post-radical hysterectomy developed acute abdomen on 10^th post-operative day followed by urinary incontinence which persisted despite Foleys catheterisation. (a) Delayed axial and coronal reformatted CECT images of pelvis show contrast in UB, bilateral distal ureters (arrows), adjacent vagina (V) and bilateral ureterovaginal fistulae (black ovals). CECT, contrast-enhanced CT; UB, urinary bladder.

The common CT imaging findings associated with fistulas are the presence of air and /or contrast material in inappropriate location, focal bladder and bowel wall thickening at the site of fistula, soft-tissue mass and adhesions between pelvic organs.^3,23,28,29 The presence of intraluminal air, in the absence of prior instrumentation and infection is considered highly specific for vesical fistulas.^3,20,30 Contrast-enhanced CT with multiplanar reconstruction is accurate, fast, and is regarded as gold-standard in the detection of fistulas and evaluation of the underlying pathological conditions with a diagnostic accuracy of 100%.^2,3,13 CT also provides accurate pre-operative road map to the surgeons and renders guidance for diagnostic and therapeutic interventional procedures.^13,31,32

Fluorine 18-fludeoxyglucose (FDG) positron emission tomography (PET)/CT is recommended for the nodal (N) and metastatic (M) staging of pelvic malignancies.³³ Though PET-CT is not used in the primary diagnosis of fistulas due to its poor anatomical resolution, a fistula may be initially or incidentally discovered on PET/CT (Figure 8).

Figure 8.

47-year-old female with recurrence of carcinoma cervix, post-chemoradiation. (a) Axial CT image with rectal contrast shows contrast in vagina (V), vaginal mass infiltrating the rectum (R) and the rectovaginal fistula (black oval). (b) Axial PET-CT fusion image shows hypermetabolic vaginal mass with rectovaginal fistula (black oval).

MRI is the imaging modality of choice for local or T staging of pelvic malignancies. MRI appearance of fistula depends on the tissue present in the wall of the tract and its internal contents (Figures 9 and 10). Fluid filled tracts typically appear hyperintense on T₂W images and gas filled tracts are hypointense on all sequences.^2,30,34 Neoplastic or inflammatory soft tissue masses and walls of malignant fistulas in primary neoplasms or tumour recurrence enhance on T₁W contrast sequences, whereas chronic fibrous tracts do not enhance.^28,32 Contrast enhancement of fistulous tracts is highly specific for active disease.¹³ T₁W images with contrast and T₂W images are useful in differentiating the soft tissues mass due to radiation-induced fibrosis from residual /recurrent malignancy as the signal intensity is higher in the latter.^29,35 Diffusion-weighted MR sequence was also found to be better than T₂W images alone in the detection of fistulas.¹ Small fistulas that show intermediate signal intensity on T₂ W images may be missed on MR leading to false negative results.¹² In future, MRI studies using newer sequences such as echo planar imaging which shorten the time of imaging, studies using diffusion and perfusion techniques that can differentiate the viable and non-viable tissues could be useful in the management of fistulas associated with malignancies.

Figure 9.

41-years-old female with recurrent urinary infections. (a, b) MRI T₂W axial and sagittal images show large irregular, intermediate signal intensity mass lesion in the cervix (C) infiltrating the base of UB wall, right vesicoureteric junction, vaginal walls and parametrium, predominantly on the right side, with a vesicovaginal fistula (white rectangle). Air (arrow) in the urinary bladder and fluid in the vaginal cavity are also seen due to the fistula. UB, urinary bladder.

Figure 10.

77-year-old male with carcinoma of prostate, status post-orchidectomy and radiotherapy presented with obstructive urinary symptoms. (a) MR T₂W axial image shows moderately enlarged prostate (P) with altered parenchymal signal intensity and an intact prostatic capsule—residual prostatic malignancy. (b) MR T₂W sagittal image shows air in the urinary bladder (arrow) and the rectovesical fistula (white oval) between the UB and rectum (R). UB, urinary bladder.

In our study, most of the fistulas due to primary infiltrating lesions were seen at time of initial CT or MRI and post-operative fistulas due to complications and avascular necrosis presented within a month after surgery. Post-operative fistulas due to recurrence of the malignancy presented within an year after surgery whereas post-radiation fistulas were seen from 3 months to 10 years after the therapy. An underlying malignancy should be suspected whenever a spontaneous fistula is seen in an elderly patient without a history of previous irradiation or surgery.⁵ Detection of fistula in early post-operative period is usually secondary to post-surgical complications, but recurrence of malignancy has to be ruled out if fistula is detected in the late post-operative period upto an year. Post-radiation fistulas may present as early as within 3 months of radiotherapy due to acute radiation injury or can occur more than 10 years after treatment due to chronic ischemic damage to pelvic organs.

The location of the fistula, presence of active underlying malignancy and viability of surrounding tissue will govern the surgical approach, extent and type of resection and the final outcome.^5,36 Therefore, the goals of imaging in the evaluation of fistulas in malignancy are as follows:

1. To detect and assess the location, size, number and course of the fistulous tracts.

2. To evaluate the primary neoplastic mass causing the fistula, the organs involved and to rule out synchronous lesions.

3. To detect lymph nodal involvement, peritoneal spread and distant metastases for staging of the neoplasm.

4. To provide pre-surgical roadmap and aid the surgeon in deciding the management (curative resection vs palliative surgical bypass), extent of surgery (organs to be preserved/resected), approach (laparoscopic vs open surgery) and timing of surgery (emergency, semi-elective, elective).

5. In fistulas occurring as post-therapeutic complications, to know/locate the altered surgical and anatomical planes at the post-operative and/or post-radiation sites, to identify residual or recurrent malignancy, to differentiate viable from non-viable surrounding tissues, to note the presence of mesh, sutures, staples, drains, contrast leaks and the status of anastomotic site.

Conclusion

Imaging and reporting according to the needs of clinician continues to be a diagnostic challenge for radiologist. A comprehensive report in fistulas should include detailed information of fistulous tract and underlying pathology for the clinician to decide appropriate management. MDCT using i.v. contrast with oral and rectal contrast is the best imaging modality in the diagnosis of fistulae in pelvic malignancies. MRI can be used in patients with equivocal findings on CT, in patients with contraindications to iodinated contrast, for local staging of primary neoplasms, in the evaluation of complex fistulas with anorectal sphincter involvement and in cases of recurrence of tumour vs post-radiation fibrosis.