Imaging in von Hippel-Lindau:

What Have We Learned

• Central Nervous System
• Pancreas
• Pheochromocytoma
• Epididymis
• Kidney
• Summary
• Notes
• Acknowledgment
• References

The screening program for von Hippel-Lindau (VHL) at the National Institutes of Health began in 1987. Ten years later much has been learned. In addition to the dramatic developments in the understanding of the molecular biology, there have been developments in clinical evaluation and management of VHL. Moreover, there have been important advances in imaging technology that make the "old" scans of 1987 look obsolete. Some things, however, remain constant. For instance, from the onset it was understood that VHL is a multisystem disorder and that any imaging strategy must take into consideration the broad spectrum of possible findings. Also, it was clear from the beginning that not all patients would exhibit all the features of VHL and in many cases imaging became focused on one or two organs. Now, ten years after our first VHL clinic it is appropriate to review what we have learned.

At the onset I would like to extend my thanks to the many patients and family members who have entrusted their films to us, or who have come to NIH to be studied. What we have learned, we have learned from you.

What Have We Learned?

It quickly became clear that a single study could not encompass all the features of VHL. Artifacts from bone limited the ability of computed tomography (CT) to be used in the cerebellum, the most common site of hemangioblastomas. Thus, MRI seemed much more appropriate in the brain. However, the ability to understand and stage hemangioblastomas was dramatically improved by the addition of gadolinium chelates that enabled magnetic resonance imaging (MRI) scans to be enhanced. Now, enhancement of brain MRIs is routine in VHL.

For spinal hemangioblastomas, CT was limited because it could only image in the transverse plane. The ideal plane is the sagittal plane and thus MRI became best suited for detecting hemangioblastomas in the spinal cord.

However, for abdominal imaging, CT was better than MRI and remains the preferred method of screening and follow-up. CT has the advantage of speed, high resolution and lower cost than MRI in the abdomen. Another advantage of CT is that there is more consistency, site to site, than with MRI.

For imaging of the scrotum, however, the best modality is ultrasound. It permits the high resolution evaluation of cystadenomas of the epididymis. Ultrasound is also very useful in the operating room where it can help guide the surgeon.

As we received more and more images from outside imaging centers and hospitals it became clear that consistency was going to be one of the most important aspects of imaging. If scans are not done in a uniform manner it is almost impossible to tell whether VHL is progressing or not.

Knowledge and expertise among radiologists also seemed to become a critical factor. Although most radiologists have heard of VHL, it is often the case that they have never personally seen a case. Fortunately, now there are references available to them, but individual expertise still determines the quality of the examination.

A final general comment is that we have learned the importance of determining the tempo of progression. VHL does not behave the same way in everyone. In some cases it is quite aggressive and requires constant follow-up but in other cases progression is slow and only periodic follow-up is needed. Thus the follow-up regimen needs to be geared to individual patients and cannot be uniformly applied to everyone with VHL.

The Central Nervous System

The most common lesion of the central nervous system (CNS) is the hemangioblastoma of the cerebellum. Most of these lesions are well recognized on MRI. Occasionally, lesions can be seen within the brainstem at the same level as the cerebellum.

In addition, hemangioblastomas can occur outside the cerebellum. The hypothalamic-pituitary axis seems to be one "hot" spot or area of increased frequency of hemangioblastomas. Subsequently we have found lesions in the supratentorial brain as well as in the retro-orbital space.

We recognized that some lesions appeared to arise outside the cerebellum and involve the temporal bone and inner ear. After one of these was removed, and was found not to be a hemangioblastoma but rather an endolymphatic sac tumor (ELST), we became aware that many VHL patients had hearing loss and that in some cases this could be attributed to tumors arising from the labyrinth, specifically, the endolymphatic sac. Screening of this area is now routine in patients with hearing disorders.

There is disagreement about what technique is the best for detecting ELSTs. A CT scan is obtained because it shows bone erosion which is diagnostic. However, MRI is useful because it can actually detect the labyrinthine anatomy and pinpoint the extent of disease. Moreover, we believe MRI is more sensitive than CT for small ELSTs.

Pheochromocytoma

Pheochromocytomas are tumors of the adrenal gland which secrete catecholamines. These can cause patients to become very hypertensive and can result in death in severe cases. Sadly, the relationship between pheochromocytoma and VHL was not fully recognized and there have been reports of death from pheochromocytoma during anesthesia and surgery for other abnormalities.

Early recognition is thus important. We employ a combination of chemical tests and CT examinations to identify patients with pheochromocytomas. We have found that CT is very sensitive for the detection of early pheochromocytomas. It has been suggested that the pheochromocytomas in VHL are somehow "silent" but this is not correct. What can be said is that when the lesions are very small they may not elevate the catecholamine levels. With time, however, the lesions will eventually become chemically active and then finally, symptomatic. We have shown that the growth rates of pheochromocytomas in patients with VHL are almost identical to those of patients with pheochromocytomas but without VHL. Thus, it would appear that rather than being silent, pheochromocytomas in VHL are active even when small, but are detected earlier in patients with VHL due to presymptomatic screening.

Patients with VHL also have an increased likelihood of bilateral or ectopic pheochromocytomas. Ectopic pheochromocytomas are found around the adrenals and in tissues far removed from the adrenals. We now routinely scan to the level of the aortic bifurcation to reduce the chance of missing pheochromocytomas in the Organ of Zuckerkandl.¹

There are a number of imaging modalities that can be used to evaluate pheochromocytomas. We find that CT is the best to detect lesions. The MIBG2 study is a nuclear medicine study that is very specific for pheochromocytoma, but it is somewhat insensitive. We rely on the MIBG for finding ectopic sites of disease when the CT is negative and pheochromocytoma is suspected. MIBG is also used to assess whether the lesion should be removed. MRI can be very helpful in confirming the location of ectopic sites of pheochromocytoma and differentiating them from other potential causes of masses.

Kidney

The kidney is frequently affected in VHL by cysts and renal cancers. When we began looking at the problem of renal lesions there was some confusion in the literature regarding the importance of renal cysts. It was felt that because cysts outnumbered tumors that cysts were precursor lesions to tumors (i.e. cysts were the first lesion and tumors developed from them). This was derived from the observation of small tumors in the wall of many VHL cysts.

Extensive experience with the natural history of renal lesions using modern CT scans has shed considerable light on this. It appears that most tumors develop not from cysts but de novo from solid islets of tumor. Buried within the normal appearing tissue of VHL kidneys are hundreds of microscopic "islets" which are likely the true precursor lesion in VHL. We have observed that most solid tumors begin as solid tumors without a cystic intermediate. Moreover, most cysts remain cysts but can enlarge, regress or stay constant in size. A minority of lesions have both cystic and solid components. Both elements tend to grow over time, but the solid tumor component eventually becomes predominant. Thus, annual or biannual screening would seem appropriate for most patients.

Treatment of renal cancers in VHL is still controversial. On the one hand, if renal cancers are left unchecked they will metastasize. On the other hand, over-aggressive surgery will result in premature renal failure which lowers duration and quality of life. Over the past 5 years we have adhered to a policy of not removing renal cancers until they attain a size of 2.5-3 cm. A range is given to account for different growth rates. Although full analysis of this strategy is incomplete, it appears to be successful in avoiding both metastatic disease and end stage renal failure in many VHL patients.

When surgery is performed, intraoperative ultrasound has been found to be a useful adjunct. Although the surgeon is well aware of most of the surface lesions, deeper lesions and those near the surface that are hidden by scar or fat may be missed. Intraoperative ultrasound allows the surgeon to perform a more complete operation and thus prolong the response from a single surgery.

There are three major modalities in use for evaluating the kidney, CT, ultrasound and MRI. Ultrasound is a tempting choice because it is less expensive than the others and does not employ ionizing radiation. Ultrasound is the method of choice for screening children with VHL. However, several authors have shown significant limitations of ultrasound for detecting lesions as large as 2 cm. This is due to technical limitations and not necessarily due to the quality of the sonographer. CT has the advantage of cost over MRI and the advantage of well defined criteria for "enhancement" which is the principal method by which the presence of tumors is defined. Moreover, since follow-up scans are so important, CT has the added advantage of consistency between centers.

MRI, however, has an important role in renal VHL assessment. As renal function deteriorates due to surgery, the use of CT contrast agents has become relatively contraindicated. This is because of a slightly increased risk of renal dysfunction in kidneys with pre-existing renal dysfunction. MRI contrast agents, however, have no known impact on renal function. Thus, MRI is used in follow-up studies of patients with compromised renal function.

Pancreas

In 1987 it was known that VHL was associated with cystic disease and occasional islet cell tumors. Like many aspects of VHL, while these original observations are true, the depth of our understanding of pancreatic disease has improved dramatically.

Although pancreatic cystic disease is "benign," it is not necessarily asymptomatic. We have seen examples of severe bowel and biliary tree compression due to enlarged cysts. To some extent percutaneous drainage of cysts can provide symptomatic relief. Pancreatic insufficiency manifested by diabetes or malabsorption can also be traced to pancreatic cystic disease.

Non-functional islet cell tumors of the pancreas are probably more common than previously thought. These vascular lesions can be quite subtle and grow slowly. Nonetheless, we have seen examples of metastatic disease resulting from these tumors. Interestingly, some cystadenomas of the pancreas can mimic islet cell tumors so caution must be employed. A surgical strategy for removing these lesions while retaining pancreatic function is currently being evaluated at NIH.

One side note is that the literature commonly mentions pancreatic adenocarcinoma as a feature of VHL. This tumor is commonly known as "pancreatic cancer" and usually has a very poor prognosis. We have yet to encounter a bona fide case of pancreatic cancer in VHL and this has cast some doubt on its association with VHL.

Epididymis

Epididymal cystadenomas occur in the scrotums of men with VHL. They are small masses made up of papilla of tissue surrounded by a cystic matrix. Most are asymptomatic although they may cause some vague pain or discomfort. This is one of the few "not to worry" lesions in VHL because there is no risk of malignancy. Although ultrasound can be used to help define the lesions, they are usually easily palpable and should pose no problem.

We have observed that as men age the epididymal cystadenomas tend to obstruct the rete testes, the small tubes carrying sperm away from the testicle. These are usually asymptomatic in spite of a dramatic appearance on ultrasound. It is unknown whether the lesions affect fertility beyond the normal reproductive years. Studies have shown no significant decrease in reproductive capacity in VHL.

Summary

There have been significant changes in the genetics, management and imaging of VHL in the last 10 years. Experience has clarified the role of various imaging modalities for screening and follow-up. As more information is gained about manifestations of VHL with regard to specific gene mutations, it is hoped that imaging may become more targeted in the future.

Note

1. Zuckerkandl bodies, or organs of Zuckerkandl, are nodes of the sympathetic nervous system lying along the aorta. Pheos grow most commonly inside the adrenal glands, but may also grow in a number of other places, roughly along a line from the groin to the collar-bone on each side, but very rarely in the chest.

2. MIBG scanning is a nuclear medicine procedure using a radioactive isotope or tracer which is absorbed by pheochromocytoma tissue. Meta-Iodo-Benzyl-Guanidine (MIBG) is injected into the patient before the scan is performed, making the pheo stand out clearly on the diagnostic picture.

Acknowledgment

The author wishes to thank the many patients and family members with VHL who sent their films or who came to NIH and the late Michelle Filling-Katz, M.D., who began screening patients with VHL at NIH in 1987.

References

1. Maher ER, Iselius L, Yates JRW et al. Von Hippel-Lindau disease: a genetic study. J Med Genet 1991, 28:243.2. Glenn GM, Lambert ND, Choyke P et al. Von Hippel-Lindau (VHL) disease: distinct phenotypes suggest more than one mutant allele at the VHL locus. Hum Genet 1991, 87:207-210.3. Walther MM, Lubensky IA, Venzon D, Zbar B, Linehan WM. Prevalence of microscopic lesions in grossly normal renal parenchyma from patients with von Hippel-Lindau disease, sporadic renal cell carcinoma and no renal disease: clinical implication. J Urol 1995, 154:2010.4. Choyke PL, Glenn GM, Walther MM et al. The natural history of renal lesions in von Hippel-Lindau disease: a serial CT imaging study in 28 patients. Am J Roentgenol. 1992, 159:1229.5. Walther MM, Choyke PL, Hayes W, Shawker TH, Alexander RB, Linehan WM. Evaluation of color doppler intraoperative ultrasound in parenchymal sparing renal surgery. J Urol 1994, 152:1984.

6. Choyke PL, Glenn GM, Walther MM, Patronas NJ, Linehan WM, Zbar B. Von Hippel-Lindau disease: genetic, clinical and imaging features. Radiology, 1995, 194:629.

As published in the VHL Family Forum  5:3, September 1997. For permission to reprint, please contact VHL Family Alliance, editor@vhl.org. Further information is available from the VHL Family Alliance, info@vhl.org.