The Dark Side of PSMA: Pitfalls, Limits, and ADT - Matthias Eiber

November 19, 2021

In this educational initiative, focusing on the knowledge of PSMA PET and PSMA theranostics, Drs Matthias Eiber and Thomas Langbein jointly prepared and Dr Eiber presents on the Dark Side of PSMA: Pitfalls, Limits, and ADT for prostate cancer.

Independent Medical Education Initiative Supported by Novartis/Adacap and Point Biopharma


Matthias Eiber, MD, PhD, Department of Nuclear Medicine, Technische Universität München (TUM), Munich, Germany

Thomas Langbein, MD, Department of Nuclear Medicine, Technische Universität München (TUM), Munich, Germany

Read the Full Video Transcript

Matthias Eiber: Welcome everybody to module six of the PSMA-PET Academy. The topic today is Dark Side of PSMA: Pitfalls, Limits, and ADT. My name is Matthias Eiber and I have prepared this session jointly with Thomas Langbein. We are both working in the Department of Nuclear Medicine, Technical University Munich. I would like to start with a short case study. This is a 78-year-old patient with status post radical prostatectomy and a PSA value at 1.3, and he underwent Gallium PSMA-PET imaging. And what you can see here is a local recurrence, and you can also see some lymph nodes in the pelvis and in the mediastinum. So assuming that this is systemic disease, the patient started androgen deprivation therapy. Under this, the PSA nicely dropped from 1.2 to 0.3. However, the patient came in for his PSMA-PET imaging study again, and from what we have seen, there is a smaller appearance of the local recurrence and an increasing size in PSMA expression of the pelvic and mediastinal lymph nodes.

So that was divergent from his PSA value. So at that time, somebody remembered that this patient also had renal cell cancer and in the tumor board it was decided to do a biopsy, and finally it turned out that these lymph nodes were not from prostate cancer, but a metastasis from prior renal cell cancer. So this is a typical case showing that imaging prostate cancer patients with PSMA-PET can have hurdles and pitfalls and that especially the signal of PSMA-PET, despite the name prostate-specific membrane antigen, is not exclusively prostate cancer-specific. And with this, I would like to give a short overview of the categories in which we think the dark side of PSMA-PET imaging can be sub-grouped. This is the first PSMA-ligand uptake in non-prostate cancer tissue, for example, benign and malignant lesions, and especially the non-prostate cancer-specific bone uptake. And the second category is challenges in benign and malignant prostate tissue including tumor heterogeneity, lesion size, and influence of ADT, and all these categories we would like to cover in the next few minutes.

Let's start with the PSMA-ligand uptake in non-prostate cancer tissue. What was already known decades ago[inaudible 00:02:50] PSMA-PET imaging was introduced that immunohistochemistry, PSMA positivity could not only be found in prostate cancer, as you can see here but also in a lot of other types of cancer. However, as you can see, only in prostate cancer, the PSMA expression is directly on the tumor cells. In most of the other cancers, it is not on the tumor cells, but on the newly formed vessels inside the malignant lesions. And typical structures, which can express also PSMA are neurogenic tissue and a typical pitfall of PSMA-PET imaging are so-called ganglia. Small structures, especially the celiac ganglia in the mid of the body, just adjacent to the large vessels, can show a very, very high uptake in PSMA-PET sometimes and therefore can be misinterpreted as lymph node metastasis.

This was found out relatively early after PSMA-PET was introduced, but is a very important pitfall. False-positive uptake in the bones also is a challenge, especially in F18-PSMA-PET imaging. And a typical example, as you can see here are fractures. You see here a patient, you see the mid of the PSMA-PET study, and you see here a local recurrence in this patient. However, you see also uptake in the area of the chest. And if you look on the adjacent CT, then you see slight sclerosis and a little bit of an irregularity of the ribs. And this patient was imaged again three months later because there was a suspicion that this is unspecific due to a fracture. And there the imaging finding could be confirmed with clear sclerosis and then adjacent callous formation in the meanwhile.

As said, PSMA expression can also happen in other tumors and prostate cancer patients are often elderly men who also can have a history of smoking and therefore the risk of lung cancer. And you can see here two cases with each one a single PSMA positive lesion in the lung, which also looks pretty similar on the adjacent CT. The second patient has also mediastinal lymph nodes, also being PSMA positive and in the histopathological evaluation of those patients, the first one shows histologically proven prostate cancer metastasis, whereas the second showed in lung cancer. So you can see PSMA-PET cannot discriminate especially between those two entities.

What other challenges do we have in benign and malignant prostate tissue? Also from immunohistochemistry, it was known for a while that the PSMA expression is heterogenous, that with increasing Gleason grade, PSMA expression increases. And for example, what you can see here in this immunohistochemical staining, a Gleason six tumor without significant PSMA expression compared to a Gleason seven tumor with high PSMA expression. These correlations have also been found after PSMA-PET went into the clinics. You see here, for example, a summary from a study in which the Gleason six and seven Gleason score tumors showed a significantly lower PSMA uptake measured by SUVmax compared to the Gleason eight to 10 tumors. And also if the tumors are stratified into low-, intermediate-, versus high-risk, you can see that the high-risk tumors have significantly higher PSMA expression.

This is not only drawn for the patient as a whole, but also inside the patient PSMA expression can be heterogeneous. And you can see this on these two images here, an [inaudible 00:07:08] H&E slide, and an autoradiography slide where you can see two tumors here and here, but only the first one shows intense PSMA-ligand uptake in the autoradiography. Whereas the second only shows a very weak uptake. It's also important to mention that PSMA-ligand PET can also be negative because prostate cancer can be PSMA negative.  In different case series, five to 10% of the cases have been described as negative. Etiology in primary prostate cancer is still unknown in advanced disease of neuroendocrine prostate cancer which happens to lose PSMA expression.

And this can also happen during the treatment if you do PSMA targeted treatment, and you can see a case example for this here. This is a 60-year-old patient with metastatic castration-resistant prostate cancer. And you see a lot of lymph node metastasis here, highly PSMA avid. Then the patients underwent a very long course of lutetium PSMA treatments for a total of 12 cycles. However, after the last cycle, a restaging PET was done and what you can see here is that the lymph nodes have been growing and have been getting PSMA negative.  So that sometimes can happen during PSMA targeted treatment that the tumor loses PSMA expression.

I also would like to emphasize that as in every imaging modality, size does matter. That means with smaller sizes, it is more difficult to detect lesions. In general, there is a good detection efficacy of PSMA-PET described in the literature. But for example, if you plot the size of the lesion against the detection efficacy, then you see clearly that with growing size, the detectability gets higher, and especially lesions below five millimeters can be often false negative on imaging.

As the last point, I would also like to mention the interaction of PSMA expression with medical treatment. And here one important example is the PSMA expression and androgen deprivation treatment, which are linked. And you can see this on this schema here, that the androgen receptor which is usually activated by testosterone, suppresses the gene encoding PSMA in the cell. And if this androgen receptor activity itself is suppressed by using anti hormones, then the suppressing effect on the PSMA gene gets away and PSMA will be upregulated. So that can lead to confounding sickness and imaging. But you can see, for example here in this patient, who was imaged before the start of androgen deprivation therapy, there you'll see a few lesions, and then four weeks after androgen deprivation therapy, his PSA significantly dropped. However, you can see much more lesions and you can also see a higher uptake in PET of these lesions. That is an important confounding factor, and can also lead to pitfalls in restaging exams.

We would like to summarize that PSMA-ligand PET is highly effective and well-accepted by clinical partners. We think that this has been shown in other modules before, but it is important that we have to be aware also of its dark sides. There is increasing evidence of pitfalls as outlined in this presentation. Therefore, interpretation of imaging needs to be done carefully within the clinical context. And consequently, we, as imaging specialists need to teach ourselves and especially our clinical partners that the term prostate-specific membrane antigen can be partly misleading. And regarding this, we would finally like to recommend this very well written illustrative article summarizing and updating important pitfalls in PSMA-PET for readers who are much more interested in this topic, as we could present in these 12 minutes. Thank you very much for your attention, and please do not hesitate to contact us in case of any further questions.