Saturday, December 31, 2016

Ipilimumab (Yervoy) fails to increase survival, even when used earlier

Ipilimumab (Yervoy) is a type of immunotherapy that is known as a "checkpoint blocker." It blocks a protein in T-cells (called CTLA-4) that tells the immune system to stand down and not attack the cancer cells. It turns off the off-switch. The hope is that immune response against the cancer will continue longer than it ordinarily would.

A previous trial showed that Yervoy did not extend survival when used in men who were metastatic and castration-resistant  (mCRPC) and who had failed chemotherapy. This is often the first group given a new drug because other options have been exhausted and because it takes less time to prove efficacy. Researchers hoped that it might have some effect if used earlier in disease progression. Unfortunately, it did not.

Beer et al. tested Yervoy this time in men who were metastatic and castration-resistant but who had not yet tried chemotherapy and who were asymptomatic or minimally symptomatic (i.e., no bone pain or organ dysfunction). In this multi-institutional study, there were 399 patients who got Yervoy, and 199 who got a placebo. Neither patients nor doctors knew who got which.
  • Patients were given 10 mg/kg of Yervoy or placebo every 3 weeks for up to 4 doses.
  • Therapy was repeated every 3 months thereafter to non-progressing patients
The outcomes were as follows:
  • Median overall survival was 28.7 months for those who got Yervoy vs. 29.7 months for those who got the placebo (no statistically significant difference).
  • Median progression-free survival was 5.6 months or those who got Yervoy vs. 3.8 months for those who got the placebo (a statistically significant difference).
  • 23% had a PSA response with Yervoy vs. 8% with the placebo.
The treatment-related adverse responses were:
  • Death that was treatment-related in 9 patients (2%).
  • Serious or life-threatening immune-related adverse events in 31%
  • Serious or life-threatening diarrhea in 15%
While there was a PSA  response, and an increased time during which more patients taking Yervoy were progression free, this did not translate to a lengthening of overall survival. This may be because there was a subset of patients who had a good initial response, but the response was not sustained. This also shows the difficulty of measuring the response to immunotherapy using PSA or other surrogate endpoint. We know that Provenge, the only approved immunotherapy for prostate cancer, lengthens survival without reducing PSA. Here, the converse is true.

Research continues on other checkpoint blockers. Keytruda has been approved for melanoma, lung cancer and head-and-neck cancer. In addition to Keytruda, there are several investigational immunotherapies targeting the PD-1/PD-L1 antigen. It may turn out that checkpoint blockers work better in combination with other immunotherapies (like Provenge or ProstVac), or perhaps they need to be primed with concurrent SBRT radiotherapy or chemotherapy. We need a better understanding about why an immunotherapy may work very well for one cancer, but very poorly for another cancer, We also can't lose sight of the fact that all  immunotherapies may be lethal. There is clearly much to be learned.

Thursday, December 22, 2016

Ac-225-PSMA-617 extends survival (update)

The nuclear medicine group at the University of Heidelberg recently reported a complete response in two patients treated with Ac-225-PSMA-617 (see this link). Now they have treated 80 patients with at least 24 weeks of follow-up, and report impressive results (here).

The 80 patients had failed on multiple therapies and were only expected to have 2-4 months of median survival.
  • The response rate (PSA reduction and tumor shrinkage) was 75%
  • Most were still alive 6 months after the therapy
  • Dry mouth was the only side effect of treatment
This is a report from a media release, and not a peer-reviewed journal. I will certainly report more details as they become available.

Anyone interested in medical tourism to try this experimental therapy can contact Dr. Haberkorn at the University of Heidelberg (he speaks English):
Phone: 06221 56-7731

There is a Phase 1 (dose finding) clinical trial of Ac-225-J591(a PSMA ligand) at Weill Cornell in NYC. It involves 8 visits over 12 weeks. Eligible patients must be metastatic and castration-resistant. They must have tried Zytiga, Xtandi and Taxotere or Jevtana. Scott Tagawa is the Principal Investigator.
Email: guonc@med.cornell. edu

(BTW - Scott Tagawa is also leading a trial combining two Lu-177-PSMA radiopharmaceuticals at Weill Cornell)

Tuesday, December 20, 2016

Recurrent PC (non-metastatic, hormone sensitive) after curative therapies exhausted? Here are some clinical trials to look at.

A perplexing situation is what to do after one has tried one or more potentially curative therapies (e.g., prostatectomy plus salvage radiation including pelvic lymph nodes), and there are no detectable metastases, but PSA keeps rising. The TOAD randomized clinical trial demonstrated that survival is improved by starting on hormone therapy (intermittent or continuous) as soon as recurrence is observed. Chemo has not proved to increase survival until after multiple metastases are detected (CHAARTED). Waiting for metastases to appear and spot treating them  has not proved to be beneficial either (see this link).

There may be hope in participating in clinical trials. There aren't many. There is a trial for the earlier use of Xtandi therapies, as well as trials for novel therapeutics, like apalutamide and Prostvac, which have been very promising in early trials. Here's a current list that you may wish to discuss with your medical oncologist:

Advanced hormonal therapies:

Apalutamide:at University of Texas, Houston (a larger trial is no longer recruiting)
Apalutamide/ degarelix/abiraterone at 8 locations
Xtandi at 172 locations
Xtandi at University of Colorado, Denver

Immunotherapy/PARP inhibitor:

Prostvac at NIH
Durvalumab + Olaparib at MSK
Olaparib - recurrent - Johns Hopkins & Thomas Jefferson U.

MAO inhibitor:

Phenelzine at USC

Sunday, December 18, 2016

Small Cell Prostate Cancer Clinical Trials

Small Cell Prostate Cancer (SCPC), and more generally Neuroendocrine Prostate Cancer (NEPC), are thankfully rare types of prostate cancers. They are not responsive to hormone therapy, to taxanes (Taxotere or Jevtana), or to radiation. They are difficult to detect and monitor with the kinds of imaging used to detect prostate adenocarcinoma (mpMRI, bone scans, PSMA PET scans), but may show up with FDG PET (see this link). They do not put out PSA, PAP or bone alkaline phosphatase. Special biochemical tests or biopsies for chromogranin A, neuron-specific enolase (NSE), synaptophysin,  DLL-3, CD56, and other biomarkers are required. It often appears at a "mixed type." 


Not all neuroendocrine prostate cancers carry the same prognosis. Aggarwal identified a sub-type that became prevalent in 17% of patients who were heavily pretreated with enzalutamide (Xtandi) and abiraterone (Zytiga). He calls this "treatment-emergent small cell neuroendocrine prostate cancer (t-SCNC). The pre-treatment probably selected for this subtype that may be partially responsive to familiar therapies. The "treatment-emergent" subtype and the small amounts sometimes detected initial biopsies do not appear to be as virulent (see this link). There are some studies that indicate that they may appear spontaneously in later stages of normal prostate cancer development. Aggarwal commented:
“Although long term androgen deprivation therapy may be associated with the development of treatment-emergent small cell neuroendocrine prostate cancer (t-SCNC) in a minority of patients, multiple studies have confirmed the long-term benefit of abiraterone and enzalutamide for prostate cancer patients in various disease settings. Use of these agents should not be limited by concern for the subsequent development of t-SCNC.”
Aggarwal has announced a clinical trial where he will be testing a combination of Xtandi, Keytruda, and ZEN-3694 in (among others) a group of men identified with the t-SCNC subtype. ZEN-3694 is an experimental medicine that inhibits a gene called MYC, which is often over-expressed in advanced prostate cancer. 


Because of the "mixed type," chemo often includes a taxane. More often, a platin is mixed in a cocktail with another chemo agent, like etoposide. A couple of case reports from Japan (see this link and this one) reported some success with a platin combined with irinotecan.

Nuclear Medicine/ Somatostatin

Perhaps the most promising treatment to date has been tried by the nuclear medicine department at the University of Heidelberg. I suggest that anyone who is interested email or call (they all speak English) Phone: 06221/56 7731. With the euro now at close to parity with the dollar, this medical tourism is an especially attractive option:

213Bi-DOTATOC shows efficacy in targeting neuroendocrine tumors

A similar radiopharmaceutical using Lu-177-DOTATATE (called Lutathera) has been FDA-approved for small cell cancer affecting the digestive tract. DOTATOC (and also DOTATEC and DOTATATE) binds to somatostatin receptors on the small cell digestive tract cancer surface, where it is highly expressed. It is rarely expressed in small cell prostate cancer, but there have been some isolated case reports like this one or small trials like this one. This means that treatment with a somatostatin analog (octreotide, lanreotide, or pasireotide) may be somewhat effective even without the radioactive emitter attached to it. These drugs are available now in the US, are not toxic, and your doctor can prescribe them without a clinical trial. there is a clinical trial of it in London for any solid tumor:

These clinical trials include somatostatins:

While the presence of somatostatin receptors in the tumor can be determined by pathological analysis (immunohistochemical (IHC) staining for SSTR2), there is an FDA-approved PET scan that uses Ga-68-DOTATATE that can detect it without a biopsy. It is used to detect neuroendocrine tumors that are often non-prostatic. Researchers at Emory found that Ga-68-DOTATATE uptake is higher even in neuroendocrine tumors of prostatic origin, which suggests that somatostatin-based therapy may be beneficial. (One patient who was positive for a BRCA2 mutation but negative for NEPC had high uptake as well.)


DLL3 is a protein that is expressed on the surface of neuroendocrine cells regardless of the cancer of origin, and has been identified in two-thirds of neuroendocrine prostate cancer (NEPC) cells. An antibody linked to a chemotherapy, called Rova-T, against DLL3 has been developed and has shown some promise against NEPC in a preclinical study. Unfortunately, AbbVie discontinued R&D after it failed to meet goals for small cell lung cancer (SCLC). A Phase 2 trial that included NEPC was discontinued. Harpoon has announced a clinical trial of HPN328  for people with advanced cancers that express DLL3. HPN328 is a bispecific T-cell engager (BiTE) that targets DLL3 and also promotes T cells to attack those cells exhibiting it

There are two other DLL3-targeted immunotherapies in trials for SCLC that may turn out to be beneficial for NEPC as well. AMG757 is also a BiTE. AMG119 is a CAR-T therapy that targets DLL-3. CAR-T involves treating one's own T-cells by sensitizing them to DLL3. Both of these create a T-cell and a cytokine response in environments that otherwise have low immune cell activity. That response may kill bystander cells, and through a phenomenon called "antigen spreading," may be able to kill other cancer cells that do not exhibit DLL3. (BiTE and CAR-T therapies that target PSMA are  in clinical trials noted at end of this article)

Misha Beltran at Dana Farber has tried an antibody-drug conjugate (rovalpituzumab teserine) targeted to DLL3 on a single patient. After two treatments, his metastases shrank and stabilized.

The Wang Lab at Duke has specific expertise in morphological analysis of NEPC and IHC staining for DLL3. It may be a good idea to get a second opinion from them.

Checkpoint blockade

Another recent discovery that gives a lot of hope is that PD-L1 is highly expressed in SCPC. This opens the door to immunotherapies that target the PD-1/PD-L1 pathway, like Keytruda.

PD-L1 expression in small cell neuroendocrine carcinomas

Several clinical trials use checkpoint blockade:

Friday, December 16, 2016

Focal Ablation: Unresolved Issues

Focal ablation is the highly targeted destruction of cancerous prostate tissue, usually with some kind of heat or cold (called “thermal” ablation). There has been a lot of patient interest in focal ablation, spurred on by doctors and institutions promoting it and media reports. There has been much hype in the last year over focal ablation using high frequency focused ultrasound (HIFU) focal laser ablation (FLA), and photodynamic therapy (PDT). Cryoablation has been around for the longest time of any. There have been pilot trials of radiofrequency and microwave ablation as well. Irreversible Electroporation (IRE) may be the only form of ablation that is non-thermal, but so far seems to share characteristics with thermal ablation therapies. The promotional announcements for all of these therapies are often unbalanced, so it behooves anyone interested in pursuing it to get an understanding of the issues involved.

I am sincerely agnostic on this subject, and am very happy to see a potential prostate cancer therapy explored in tightly circumscribed clinical trials where patients are informed of the risks. I do believe that until we have learned more, clinical trials with strict protocols should be the only circumstances under which focal ablation is performed.

The Hope

Focal ablation has been touted as “the male lumpectomy.” This is a term borrowed from breast cancer. Breast cancer sometimes starts as a single tumor (called “unifocal”) that may be cured if it is removed with a negative margin. Just as the breast is preserved by such excision, the hope is that prostate function, and especially the function of nearby organs (bladder, rectum, urethra, bladder neck, neurovascular bundles, erectile function, and continence) can be fully preserved. Let’s understand why “lumpectomy” may be very different for the prostate.


Prostate cancer is overwhelmingly a multifocal disease. 80-90% of prostatectomy specimens have separate tumors distributed throughout the organ. Removing the largest, highest grade tumor (called “the index tumor”) does not remove all the cancer from the prostate.


One way to get around the multifocality issue is to ablate half the prostate, either the right lobe or left lobe, but not both.This is called hemiablation. The hope is that the damage to nearby organs will be significantly reduced in so doing. Prostate cancer often appears to predominate in one lobe. But appearances are deceiving, even when saturation biopsies have been used to determine that the cancer was unilateral, it turned out to be bilateral in 3/4 of those cases (see this link), and may be as high as 90% (see this link). With traditional TRUS biopsies, unilateral cancer was misidentified in about 80% of men (see this link). Multiparametric MRI is not good at finding small tumors on the contralateral side. Pompe et al. showed that it missed cancer on the contralateral side in 58% of patients. The main issue is that it has not been proven that hemiablation is curative. In a study of 55 men in Belgium who received hemiablative HIFU, a quarter of the men relapsed and required further treatment. In a US study of 100 men receiving hemi-ablative HIFU, followed up with a biopsy after 2 years, a quarter had relapsed with Grade Group 2 or greater prostate cancer.

Index Tumor Theory

Proponents of focal ablation argue that it doesn’t matter if there are small amounts of prostate cancer that remain untreated. Prostate cancer, they believe, spreads by cloning daughter cancer cells from a single “parent” tumor within the prostate. This is called “Index Tumor Theory.” Under this theory, if the index tumor is removed by ablation, the prostate cancer will not spread further. In theory, the small untreated daughter foci of cancer are not malignant and will cause no further problems. In theory, the index tumor is identifiable as the largest, highest Gleason score tumor within the prostate.

Index tumor theory relies on the findings of two studies. Liu et al. and Mao et al. showed that metastases arise as clones from a single parent cancer cell. The Liu et al. study was based on cancers from 30 men who died of prostate cancer. The Mao et al. study confirmed the earlier study in a sample of 16 men. While both studies showed that metastases arose from a single prostatic parent cell, they did not show that the parent cell was in an index tumor. In fact, a case report from Johns Hopkins showed that lethal metastases at least sometimes could arise from a small, low grade tumor within the prostate, rather than from an index tumor. Adding to the complexity, Cheng et al. found that multiple tumors had independent origins. In 15/18 tumors, they found that they arose independently rather than from a parent tumor within the prostate, and in only 3/18 tumors they arose through intraglandular dissemination from an index lesion. Similarly, Wei et al. looked at prostate tumors taken from 4 patients, and found there was considerable genetic diversity within their index tumors as well as their other cancer foci. Ibeawuchi et al. discovered that a unifocal tumor could be as genetically diverse as multifocal tumors. Løvf et al. found that the various tumors in the same prostate only rarely shared genetic mutations, suggesting independent origins. Kneppers et al. found among 30 men with lymph node metastases that for 23%, their metastases were not clonally derived from the index tumor.

All of the above-mentioned genetic studies have been conducted in small numbers of patients. Genetic studies are tremendously difficult to conduct and interpret. Genetic breakdown is a characteristic of cancer, which complicates the subjective determination of what constitutes a clone from the index tumor.

None of this disproves index lesion theory entirely. In fact, there must be some truth to it or focal ablation would never be effective. Focal ablation trials with 5 years of follow-up demonstrate that focal ablation seems to halt progression in most men. However, because the studies have not been randomized, we cannot rule out that those mostly low risk patients were caught early and would not have progressed appreciably in that time frame anyway. We also know from long-term active surveillance trials that about half of all men with confirmed low-risk tumors will eventually progress – the smaller Gleason 6 tumors must be monitored. The most likely scenario is that there are index tumors in some men but not others. Unfortunately, we have no easy way of predicting which patients have index tumors and which have multiple tumors that are capable of malignant spread.

Targeting the index tumor

Assuming there is an index tumor, the next question becomes: can we precisely locate the tumor for targeted ablation? Our best current tool for doing so is using a multiparametric MRI (mpMRI) to target what seems to be the index tumor, and to confirm the location with a biopsy (either ultrasound fusion or in-bore). This poses special challenges.

Most patients who choose focal ablation are those who have predominant Gleason pattern 3 (either Gleason score 3+3 or 3+4). mpMRI is not at all sensitive at finding such low grade tumors if they are small; in fact, it is no better than a standard TRUS biopsy. In a study of mpMRI and Ga-68-PSMA PET/CT, both imaging techniques missed more than half the prostate tumors found after prostatectomy. Perhaps Color Doppler Ultrasound or transperineal template mapping biopsy perform better (see this link), but they are seldom used. However, mpMRI is a good tool for finding larger and higher grade tumors. In a study at UCLA, 80% of “index tumors” were found using mpMRI. In another UCLA study, mpMRI found that half of all men with intermediate or high-risk prostate cancer had satellite tumors in addition to their index tumor, but 2/3 of those same men were found to have satellite tumors when their prostates were surgically removed. Over half of the satellite tumors were Gleason score ≥ 3+4.

While mpMRI may detect index tumors, it is not a good tool for delineating even higher grade tumors. Priester et al. compared the dimensions of tumors found via mpMRI in 114 men to the dimensions of their same tumors determined via post-prostatectomy pathology. They found that the actual tumors were 3 times larger than their MRI estimates – they missed 80% of the tumor’s volume by relying on the MRI. It is worth noting too, that these MRIs were read by arguably the best radiologist in the business, Daniel Margolis at UCLA. He literally wrote the book (PIRADS 2.0) for interpreting mpMRIs. In a study of 461 lesions in 441 men, the average size of tumors was only 1.6 cm on the mpMRIs but was 2.4 cm after prostatectomy. The correlation between MRI and actual size was poor (0.13- 0.65). Pompe et al. found that mpMRI could not detect extracapsular invasion, and missed cancer in 58% of patients who had cancer in the contralateral lobe from the index tumor.

If satellite tumors are to be ablated as well as the index tumor, mpMRI performs even worse in finding them. Hollmann et al. found that satellite tumors were a median of 1 cm, and up to 4.4 cm, away from the index lesion, so they would not be destroyed within the ablation zone of the index lesion, and it would be difficult to locate them. (Update 5/2019) Stabile et al found that mpMRI missed 30% of the significant (Gleason score≥3+4) cancer outside of the index lesion, and the missed tumors had a median length of 2.6 mm, which is smaller than anything an mpMRI can detect.

Incomplete ablation in the ablation zone

Now let’s assume you do indeed have an index tumor, and you were able to accurately delineate it somehow, the next question becomes: Can focal therapy be used to completely ablate the tumor? So far, the answer seems to be – not completely. In some studies, treated patients had MRI-guided biopsies of the ablation zone within 6 months of treatment. Cancer was found in the ablation zone:

Focal Laser Ablation (FLA)(Update 5/2019) Chao et al found that 8/32 (25%) had an mpMRI suspicious for cancer in the ablation zone within 2 years after FLA (Median time to positive mpMRI in the ablation zone was 6 months). All were confirmed by biopsy. Only one of those patients had low volume GS 6. 24/32 (75%) had an unsuspicious mpMRI, but biopsy at 2 years after FLA was nevertheless positive in 9 of the 14 men (64%) who had a biopsy. So 17/22 men (77%)  had a positive biopsy in the ablation zone after 2 years. Change in PSA did not predict a positive or negative mpMRI or a positive or negative biopsy.

In this study, MRI-detected cancer was found in 10/27 patients after 12 months, with cancer found in the ablation zone via biopsy in 3 patients. Cancer was found in the ablation zone in 2/9 patients (22%) in this study, 7/10 (70%) patients in this study that used a targeted biopsy, and 4/12 (33%) in this study. In one study, 2/13 (15%) had residual cancer within the ablation zone, but only 13 of 23 patients had a targeted biopsy. Knull et al. compared the pre-operative mpMRI images with MRIs obtained immediately after FLA in 23 lesions. They found that FLA did not completely overlap the intended ablation zone in about half of the lesions, and those tumors extended a median of 0.9 mm past the edge of the ablation zone.

High Intensity Focused Ultrasound (HIFU): Cancer was found in the ablation zone in 36% of the patients who had biopsies for cause in this study. In a hemi-ablation study, 28% had biochemical recurrence and 3/8 biopsied patients (38%) had cancer in the treated lobe. In another hemiablation study, 16% had cancer in the ablated lobe. In a large study of whole gland HIFU, 29% were given a repeat treatment. Cancer was found in 42% of high risk men in the ablation zone in this study - 10% were given a repeat treatment. In a US hemiablation study, 17% had Grade Group 2 or greater cancer in the treated lobe.

Photodynamic Therapy (PDT) /TOOKAD: In a hemiablation study, 11/21 men (52%)had a positive biopsy in the treated lobe.

Cryo: In a whole-gland study of cryoablation, 37% had residual cancer in the ablated prostate.

Irreversible Electroporation/NanoKnife (IRE): In a study of focal IRE, which is largely a non-thermal form of ablation, 4/25 patients (16%) were found to have residual cancer in the ablation zone. In another study that used mpMRI to detect residual cancer up to one year after treatment, 9/30 patients (30%) were found to have residual cancer in the ablation zone. Colletini et al reported in-field treatment failures by 18% of low and intermediate-risk patients detected via mpMRI-targeted bioosy after 6 months.

So we observe that ablation is sometimes incomplete within the treated area. There are thermodynamic and biochemical reasons that may explain those failures.

Heat Sink Effect

Most kinds of ablation (e.g., FLA, HIFU, cryo & PDT) are thermal, which means they rely on the local application of heat or cold to ablate the tumor tissue. The second law of thermodynamics guarantees that heat (or cold) will never stay exactly where it is put. This is true for the thermal energy generated by laser beams, by ultrasound, contact with cold, or by any kind of electromagnetic energy. Water is a very good conductor of thermal energy, and prostate tissue is mostly water. The thermal energy always flows away from where it is placed, leaving the ablation zone with less ablative energy, and areas around it with more ablative energy. This translates to sub-lethal killing of cancer cells within the ablation zone, and killing of healthy tissue outside of the ablation zone.

Biochemical Effects

Human cells, especially cancer cells, have self-preservation mechanisms that may defeat efforts to ablate them. One such mechanism is “heat shock protein (HSP).” Whenever cells are threatened with heat, they enlist HSPs to protect themselves. (There are actually separate “cold shock proteins” that have been identified.) HSPs play an important role in protecting cancer cells, and scientists are developing HSP inhibitors that may one day help other medicines to treat cancer. HSPs are known to play a special role acting as chaperones in bringing the androgen receptor to a more protected place inside the cell. They also encourage cells to enter a dormant phase where they are less subject to destruction. Cell cycle dormancy may play a role in ablation therapy. It is possible that in malignant cells that are not destroyed, cell cycle arrest may delay cell replication for some time. Paradoxically, activation of HSPs may turn cancer cells more aggressive. (See this link and this one). This has not been studied in regard to focal ablation, but should be.

We are coming to recognize the effects that cancer cells may have on nearby “bystander” cells. In a recent lab study, prostate cancer cells stressed by PDT released nitric oxide that caused bystander cells to become more aggressive. The role of extracellular vesicles/proteasomes in promoting malignancy in nearby cells under ablation conditions has yet to be elucidated.

Organ-at-risk damage/toxicity

Because of the heat sink effect, there will always be some impact on surrounding healthy tissues. Depending on where within the prostate the index tumor is, and how large the ablation zone is, ablation may damage the urethra, the rectum, the bladder neck, or neurovascular bundles. In most modern trials of focal ablation, side effects have been low, but are not zero.

At the same time, there has been much progress made in reducing the toxicity of radical (whole gland) radiation therapy. Take for example, a report of HDR brachytherapy as a monotherapy for treating intermediate risk patients, and compare it to the recent report by the Ahmed/Emberton group of (mostly) intermediate risk patients treated with focal HIFU in the UK, the largest study of focal HIFU. Both studies had 5 years of follow-up.

HDR brachy
Recurrence-free survival
Potency preservation
Percent pad-free
Serious rectal injury
2 patients

Oncological control was 30% better with HDR brachy and only required a single treatment. Sexual, urinary, and rectal late term side effects were equivalent for both treatments. What is the advantage of focal ablation, then?

Re-do rates

As we’ve seen, some recurrences occur within the ablation zone, but most recurrences occur outside of the treated area. In the above-cited report on HIFU, 28% of patients had a recurrence. This is typical for focal ablation. An advantage often cited for focal ablation is that patients who have a recurrence can be retreated with a second round of focal ablation therapy. In the Ahmed/Emberton HIFU study, 25% of all patients were treated with HIFU multiple times (others chose radical salvage therapy (7%) or permanent hormone therapy (1%)).

“Re-do’s” incur extra costs and may increase morbidity of treatment. There’s no guarantee that they will be effective. As we’ve seen, recurrences are common even when the whole gland is ablated.

Lack of long-term data

The longest running studies of focal ablation, other than cryotherapy, have only 5 years of follow-up. While 5 years may be enough for therapies that are simply an improvement over existing therapies, focal ablation requires longer follow-up because of all the open questions that may affect long-term results. Because many of the focal ablation patients so far have been low risk patients who are likely to enjoy long progression-free times anyway, it is not at all clear that the remissions are lasting ones. Both the AUA nor the EAU consider focal ablation to be experimental and unproven.

Tracking progression after therapy

After radical prostatectomy, we hope that PSA will become undetectable permanently. If it rises afterwards, we suspect recurrence. After radical radiation therapy, PSA reaches a nadir, usually less than 0.5 ng/ml. If it rises 2 or more points above that, we suspect recurrence. However, with focal ablation, there is no reasonably expected PSA nadir, and there is no rise in PSA we can label as a biochemical recurrence. The PSA changes will be different for every patient. Because only the index tumor has been ablated, we don’t expect PSA from small foci of cancer outside of the ablation zone to vanish, nor PSA from BPH or prostatitis. The Chao et al trial showed that change in PSA is not a good predictor of recurrence. Because PSA cannot be used to monitor remission, we have to use imaging and periodic biopsies. Such imaging and biopsies requires experienced radiologists and pathologists because ablated tissue is qualitatively different from unablated tissue. Again, the Chao et al trial showed that while a positive mpMRI always predicted a positive biopsy, a negative mpMRI led to a positive biopsy in most cases treated with FLA. If found to be true of other kinds of focal ablation, periodic biopsies will have to be part of routine follow-up.

Salvage after ablation

If ablation doesn’t succeed and further ablation is either futile or dangerous, what are the salvage options? Salvage prostatectomy is complicated by the ablative tissue alterations, and may lead to increased morbidity. There is no reliable data on whether or not salvage radiation is effective after ablation failure. There are no experts in such salvage therapies.

Comparison to active surveillance

Focal ablation is often put forward as a middle ground between active surveillance and radical treatment. However, unlike active surveillance, there is some risk of morbidity after focal ablation. There is no long-term clinical evidence for the index tumor theory, and we have learned from long-running active surveillance trials that up to half of all Gleason 6 cancers eventually progress. Because of this, the patient is actually on a lifelong active surveillance protocol anyway: he must continue to have periodic imaging and biopsies to track progression, but is disadvantaged by not being able to use PSA to track progression.

Some focal ablation proponents, notably Ahmed and Emberton, argue that focal ablation should only be offered to intermediate risk patients and to those low risk patients who refuse active surveillance. This seems reasonable.

Inexperienced practitioners and practices

Focal ablation is still very new in the US, there are few practitioners who have adequate experience, and the learning curve is steep. There are no standard protocols. It may be years before there is consensus on best practices.

Danger of procedures

Ablation often requires anesthesia, local or general. IRE, for example, requires artificial paralysis and respiration throughout the high-voltage process.


No form of ablation is covered by insurance or Medicare, and out-of-pocket costs are typically in the $20,000 range. Because “re-do’s” are often required, future costs are unpredictable. There will be ongoing costs of periodic imaging (usually mpMRIs) and biopsies.

As with all new therapies, methods and outcomes will undoubtedly improve over the years. This first wave of practitioners and brave patients are taking risks that may eventually benefit many others. It is important that patients understand those risks before making their treatment decision.

Sunday, December 11, 2016

PET scans for prostate cancer

In the last few years there has been an explosion in the number of new PET scan indicators. I thought it a good idea to provide some background and an update.

Bone scans

PET scans may be understood as an improvement over bone scans. The traditional way of finding distant metastases is to use a technetium bone scan and CT. There are several problems with bone scans:
  • they show bone overgrowth, which may be bone metastases, but may just be arthritis or old injuries
  • only a bone biopsy can tell for sure, and it's not often feasible when the suspected mets are small or inaccessible
  • they reveal few mets when PSA is below 10-20 ng/ml or when PSA is stable
  • they only show bone mets, not soft tissue
The main advantage is that they are relatively inexpensive.

The principal uses are: 
  • to rule out bone mets in high risk patients prior to curative treatment
  • to diagnose metastases that may respond to chemo, Xofigo, or spot radiation
  • to track response to treatment among metastatic patients.


Inherent limitations

The new PET scans are better than previous ones in terms of the size of the metastases they can detect, but they do not detect all metastases.

  • A cancer cell is many times smaller than the resolution of the CT or MRI. 
  • The activity of the cancer cell seems to influence whether it is detectable on any of the scans. 
  • There is "noise" in even the most specific tracer, with no sharp delineation between signal and background.

Salvage Radiation

The most important use of these new PET scans is to rule out salvage treatment when it would be futile. For men who have persistently elevated PSA after prostatectomy, or who have had a recurrence (nadir+2) after primary radiation treatment, a PET scan showing distant metastases can spare the man the ordeal and side effects of salvage treatment.

For salvage after primary radiation failure, it is necessary to locate areas within the prostate where the cancer may still be localized.  Memorial Sloan Kettering and the Mayo Clinic have effectively used PET scans to target areas within the prostate for salvage focal ablation or brachytherapy.

For salvage radiation after prostatectomy, it may be possible to identify areas of the prostate bed where spread is evident. While the entire prostate bed must be treated (most of the prostate cancer is below the limit of detection of even the most accurate PET/MRI scan), some radiation oncologists like to provide an extra boost of radiation to the detected cancer foci.

There has been accumulating evidence in the last few years (see this link) that very early salvage radiation treatment may improve salvage radiation outcomes over waiting until the PSA has risen above 0.2 ng/ml. Unfortunately, none of our PET indicators are any good at detecting metastases when PSA is below 0.2 ng/ml. It is unlikely that there are any distant metastases when PSA is that low, but there are some relatively rare forms of prostate cancer that metastasize without putting out much PSA. This leaves the patient without any assurance that salvage radiation will be successful. Perhaps the new PORTOS genetic test will be able to detect distant metastases biochemically, but this remains to be proven.

Pelvic Lymph Node (LN) Treatment

For men diagnosed with high risk prostate cancer, a difficult question is whether the pelvic lymph nodes ought to be treated, either with radiation or with pelvic lymph node dissection. Nomograms based on disease characteristics are used to determine whether the pelvic LNs merit treatment, but such nomograms are often inaccurate. A CT scan can sometimes identify lymph nodes enlarged (>1.2 cm) due to cancer. However, some LNs are only slightly enlarged (0.8-1.2 cm), and some cancerous LNs are not enlarged at all (<0.8 cm). LNs are usually enlarged by infection, so size alone is not a good indicator of cancer. An advanced PET scan can sometimes detect cancerous LNs. This may aid the decision on whether to have whole pelvic treatment for men with high risk cancer. Men who have already had radical prostate therapy that may have included radiation (primary or salvage) to other areas (i.e., the prostate or prostate bed) may face a similar decision as to whether to treat the pelvic LNs with radiation.

Just as it is necessary to irradiate the entire prostate bed and not just the detected foci when giving salvage radiation after prostatectomy, it is probably necessary to treat the entire pelvic LN area, and not just individual LNs, when cancer is detected anywhere in the pelvic LN area. Of course, such a decision must be balanced against the risk of side effects. There are ongoing clinical trials (RTOG 0534 for salvage therapy and RTOG 0924 for primary therapy) to determine whether such treatment provides any survival advantage when LN involvement is suspected. The STAMPEDE trial included an arm where patients were node positive (but negative for distant metastases) and were treated with radiation. Short term follow-up demonstrated an improvement in failure-free survival of 52% among those who had treatment.

Oligometastatic radiation of distant metastases

Although some have theorized that there is a stage in prostate cancer metastatic progression where the cancer is still curable, or where it can be delayed by removal of 1-3 detectable metastases, this theory has never been proven. In fact, a meta-analysis this year (see this link) showed that metastatic progression continues in almost all men despite such treatment. The possibility remains that progression may be slowed by spot treatment, although this remains uncertain as well. The natural history of metastatic progression is often very slow in early stages, with years between the first few metastases. The reason for this may be because the tissue in metastatic sites must first be biochemically transformed by signals from micrometastases in order to accommodate the growth of larger metastases. This preparation of fertile "soil" in which metastatic "seeds" can grow may take some time. PET scans for detection undoubtedly introduce lead-time bias into the calculation; i.e., the time between the first and second detected metastasis is certainly longer because a more sensitive PET scan was used, and not necessarily because the first detected metastasis was spot-treated.

In spite of the uncertainty concerning efficacy of spot treatment, patients often want to treat whatever can be detected. When such metastases are detected with sensitive PET scans and are in locations amenable to spot treatment with SBRT, and there is minimal risk of radiation damage to nearby organs, it is hard to argue against such use. However, the patient should understand that there is so far no evidence that such treatment will provide any benefit. He should also understand that detectable metastases in distant sites means that his cancer is systemic. There are thousands of circulating cancer cells and undetectable cancer cells already lodged in tissues. For this reason, it is never a good idea to delay systemic therapy (e.g., hormone therapy) in order to wait for PSA to increase to a point where metastases become detectable on a PET scan. The TOAD randomized clinical trial suggested that immediate hormone therapy at the first sign of recurrence after curative options were exhausted cut 5-year mortality in half compared to waiting for PSA to rise and metastases to become detectable.

Palliative treatment of metastases

Metastases can cause pain and interference with organ function. Bone scans can find larger bone metastases, and they are the ones most apt to cause pain, fracture, or spinal compression. Metastases in weight-bearing bones may be spot-radiated with SBRT to prevent such problems and to relieve pain. In the unusual event that a bone scan can't locate them accurately enough for SBRT treatment, a PET scan may be used.

Bone scans do not detect metastases in soft tissue, while most PET scans (other than the NAF18 PET) can. A PET scan may locate metastases in organs that may be biopsied or treated with radiation or other therapies, like embolization or ablation.

Multiple metastases

The CHAARTED study has taught us that prostate cancer with multiple distant metastases behaves in a different way and reacts to different therapies compared to prostate cancer with a low metastatic burden. Although the metastatic burden in the CHAARTED study was based on bone scan and CT, there may be a potential to identify patients who may respond to earlier systemic therapy if a PET scan were to be used. This use has yet to be explored.

Tracking success of treatments (radiographic progression)

PSA is not always the best measure of whether a treatment is successful and ought to be continued. Because they destroy cancer cells, some therapies may actually raise the PSA level for some time immediately following treatment. Chemotherapy does not always immediately reduce PSA, but the patient wants to know whether the potentially toxic treatment should be continued. Most of the time, serial bone scans can provide an adequate radiographic assessment. However, in patients with low PSA or low metastatic burden, serial PET scans may sometimes provide a more accurate assessment.

Initial detection, active surveillance, focal therapy, dose painting

Just as multiparametric MRIs can be used to detect significant prostate cancer when suspicion remains after a first negative biopsy, a PET scan can conceivably be used for such a purpose (as in this clinical trial). PET scans can also be used to track progression of prostatic foci in patients on active surveillance. It is hard to justify the cost for such purposes, and there is as yet no evidence that it is any better than a multiparametric MRI. In light of the recent evidence that multiparametric MRI may fail to delineate up to 80% of detected prostatic index tumors, they may find future use of PET/MRI in contouring treatment areas for focal ablation and for dose painting (see this link).

How PET scans work

Positron Emission Tomography (PET) is a way of creating a 3D anatomical image. Instead of using X-rays, as a Computerized Tomography (CT) scan does, it detects positrons, which are positively charged electrons (which do not exist in nature). When a positron encounters a normal negatively-charged electron, they annihilate each other and release 2 gamma rays in opposite directions. When the machine detects such a pair of gamma rays, it extrapolates their source position, putting an image together. The PET scanner is combined with a CT scanner in the same device in order to provide anatomic detail.

As a cautionary note, PET scans do expose the patient to significant amounts of ionizing radiation from both the PET indicators and the simultaneous CT scan. It is not something one wants to do frequently.

PET emitters are short-lived radioisotopes that are created in a nearby cyclotron. Commonly used ones include carbon 11 (C11), fluorine 18 (F18), gallium 68 (Ga68), copper 64 (Cu64), zirconium 89 (Zr 89), and iodine 124 (I124). The choice of which one to use is based on cost, access, half-life, strength of signal, and ease of integration with the ligand. C11, for example, has an extremely short half life of only 21 minutes. This means it has to be manufactured very nearby where it will be incorporated into a ligand (like acetate or choline), and must be used immediately. F18 has a longer half-life (118 minutes) and has excellent detectability, but interferes somewhat with metabolism of acetate or choline. I124 has a long half-life (4.2 days) which may be too long for a patient who may have to remain isolated for the duration.

PET emitters are chemically attached (chelated) to molecules, called ligands, that have particular affinity for prostate cancer cells. Some ligands are metabolically active, meaning they are food for the cancer cell, but not as much for healthy cells. Other ligands are created to attach to specific binding sites on the surface or inside prostate cancer cells. Sodium fluoride (NaF18) replaces hydrogen with positronic fluoride when hydroxyapatite, the mineral that constitutes our bones, is actively accumulating in bone metastases.

Metabolic ligands

Because rapidly growing cancer cells metabolize a lot of glucose, fluoro-deoxy-glucose (FDG) has long been used in PET scans for cancer. Prostate cancer in its early stages does not metabolize glucose readily, so FDG can't be used until later stages.

Prostate cancer cells do metabolize fats, consuming choline and acetate. C-11 choline and acetate overcomes the interference problem of F-18 choline and acetate, but is very difficult to work with. Although the FDA has approved the C-11 Choline PET, only the Mayo Clinic offers it in the US. A few sites offer the C-11 Acetate PET, but it is expensive. It also requires a fairly high PSA, ideally ≥ 2.0 ng/ml, or fairly large metastases, to detect anything.

Fluciclovine has recently been FDA approved. It is incorporated into prostate cancer cells as part of amino acid metabolism. It can detect somewhat smaller metastases at lower PSAs.

PSMA ligands

95% of prostate cancers express a protein called Prostate-Specific Membrane Antigen (PSMA) on the surface of cells. There are a variety of ligands that are attracted to it. Some of the ligands are antibodies (like J591), some are shortened antibodies (called minibodies, like Df-IAb2M), and some are small molecules (peptides, like PSMA-HBED-CC or DCFPyL). PSMA-targeted ligands may accumulate in salivary glands, tear glands and kidneys, and urinary excretion may interfere with readings in the prostate area. PSMA has also been found on the cell surface of some other kinds of cancer. New PSMA ligands are still being developed and tried. So far, the one that seems to have the highest specific affinity for PSMA is called F18-DCFPyL. It detects more metastases at lower PSA than the others. 

Other ligands

There are other prostate cancer-specific molecules for which ligands have been developed and to which positron emitters have been attached.  One of the most promising is the bombesin or RM2 ligand that attaches to the gastrin-releasing peptide receptor (GRPR) on the prostate cancer cell. In pre-clinical studies, it outperformed C-11 Choline. Clinical trials have started. Clinical trials have begun on several PET ligands that are designed for other receptor sites: human kallikrein-related peptidase 2 (hK2), FMAU, Ga-68-Citrate, I-124-Prostate-Stem-Cell-Antigen, Ga-68-DOTATATE (Somatostatin receptor), F18-DHT (androgen receptor), Cu-64-DOTA-AE105 (uPAR receptor), Cu-64-TP3805 (VPAC receptor). or multiple radiotracers.

The winner (so far) is...

Based on clinical trials, below are various PET indicators in approximate rank order of their sensitivity to detect prostate cancer, and their specificity for detecting it exclusively:
  1. F18-DCFPyL
  2. F18-DCFBC
  3. Ga68-PSMA-HBED-CC (Ga68-PSMA-11)
  4. Fluciclovine (F18 - FACBC)/ Axumin
  5. C11-Choline/ C-11-Acetate
  6. F18-Choline
  7. NaF18
  8. F18-FDG
The following table shows the percent of patients who had metastases detected at various PSAs. F18-DCFPyL is much better than Ga68-PSMA at low PSA.  At PSAs between 0.5-3.5 ng/ml. it detected prostate cancer in 88% of recurrent patients, while Ga68-PSMA-11 only detected prostate cancer in 66% of the same patients - an improvement in sensitivity by a third. Next in line is fluciclovine, which was recently FDA approved. In 10 patients screened for recurrence at a median PSA of 1.0 with both Ga68-PSMA-11 and fluciclovine, the PSMA scan detected cancer in 5 of the 10 men that were negative on fluciclovine. In addition, positive lymph nodes were detected in 3 of the men using the PSMA scan that were undetected with fluciclovine (see this link). Most other PET indicators, like C-11 Choline or Acetate, or NaF18 are not at all reliable when PSA is less than 2.0.

Percent of patients in whom prostate cancer was detected by the PET indicator, broken down by the PSA of the patients

PSA range
PET Indicator
<0.2 ng/ml
0.2- 0.5 ng/ml
0.5 -2.0 ng/ml
> 2.0 ng/ml


88% (0.5-3.5)


66%  (0.5-3.5)


73% (0.5-1.0)
93% (1.0-2.0)



36% (PSA<1, PSADT>6 months)
(PSADT<6 months)
53.3% (0.5-1.0)
71.4% (1.0-2.0)
69.2% (0.5-1.0)
86.7% (1.0-2.0)
100% (>5.0)

37.5% (0.2-1.0)    77.8% (1.0-2.0)


PET scans are usually combined with simultaneous CT scans for image resolution. Siemens has a device that simultaneously provides a PET scan and an MRI. This enables  greater image resolution and the detection of smaller metastases than is possible with a PET/CT. (GE and Philips manufacture dual scanners rather than an integrated single scanner). The PET/MRI exposes the patient to a much lower dose of ionizing radiation than the PET/CT. These devices are expensive, and are only available at a few large tertiary care facilities. In one PET/CT vs. PET/MRI comparison using Ga-68-PSMA, the PET/MRI was able to detect 42% more metastases in recurrent patients, 10% more lymph node metastases, and 21% more bone metastases. In the US, PET/MRIs are in use at Mass General, Johns Hopkins, Stanford, UCSF, Washington University, Cleveland Clinic, and Memorial Sloan Kettering and several others.


So far, only FDG, C11-Choline, and Fluciclovine are FDA approved for prostate cancer detection. NaF is approved for clinical trials and registries only. FDA approval opens the way for Medicare and private insurance to approve and pay for them. Sometimes, insurance plans will agree to pick up the cost. Otherwise, patients who want them must pay out of pocket, if they are available. Ga68-PSMA-11, for example, is available for purchase in a clinical trial at UCLA for recurrent prostate cancer and costs $2,650 for each infusion. 

Clinical trials

All of the newer PET tracers require validation with larger sample sizes. While there are some diagnostic tests that have a "gold standard" against which performance can be evaluated, this is problematic for detecting metastases. A positive finding can often be confirmed with a biopsy, so a PET scan's positive predictive value (true positives and false positives) can be ascertained. But there is no easy way to determine whether negatives were true or false in live patients.

F18-DCFPyL is available for free in a trial at 15 sites in the US and Canada. It is only for patients who are recurrent after a prostatectomy (PSA>0.2) or after primary radiation (PSA> nadir+2.0). They must also be negative on a bone scan/CT and not be using hormone therapy.  Contact details are available here. NIH is also doing a free clinical trial among any men who are (1) high risk or (2) recurrent (see this link). It is available as a PET/MRI for specific purposes at the University of Wisconsin, Northwestern University, and at Princess Margaret Cancer Centre in Toronto.

It is also available at Johns Hopkins, where it was first developed, for a wide range of indications if ordered by any of their physicians Contact details are available here, here, and here. Johns Hopkins is also utilizing it in a study to determine whether there is any benefit to SBRT treatment of oligometastases (see this link).

There are several studies in Canada: in BC, Ontario (and this one and this one).

Ga68-PSMA-11 is available in several clinical trials in the US, including several with a PET/MRI, at UCLA, UCSF, Stanford, and Cleveland Clinic. UCSF is testing a new PSMA indicator called CTT1057.

Fluciclovine is available almost everywhere in the US, and is covered by Medicare for recurrent patients.

If you are interested in one of those PET scans for an indication outside of the clinical trial, call the contact anyway. Some are planning clinical trials for expanded indications shortly, and some may make the PET scan available for purchase outside of the clinical trial. Inquire about cost and get pre-authorization from your insurance company if you can. These can be very expensive.