The Atzpodien Regimen and Beyond
Special Update - March 27, 2004
In this article I suggested that it might be worth adding Accutane (13-Cis- Retinoic Acid) to the Atzpodien regimen on the basis of preliminary evidence. Dr. Atzpodien and his group have just published the results of an important randomized trial [Atzpodien 2004] which shows no survival advantage from adding Accutane. I will update the rest of this article soon, but until then please ignore the suggestions that adding Accutane helps. The best evidence now is that unfortunately, it does not.
The Atzpodien Regimen: IL-2, Interferon, and 5-FU
This complex triple drug cocktail is a standard treatment in Germany where it was developed by Dr. Jens Atzpodien and colleagues. It may be used in other European countries as well. It doesn’t seem to be widely used in the United States. The treatment is mostly outpatient though a few studies report hospitalizing patients for the most intense part of the treatment.
Dr. Atzpodien and several others have reported high response rates – higher than for IL-2 alone with long term survivals which look roughly similar to high dose IL-2. This, along with it being standard treatment in parts of Europe, makes it important and worth covering.
5-FU (5-Fluorouracil) is actually an old and well known chemotherapy drug. Although RCC doesn’t respond to chemotherapy, Interferon has been combined with 5-FU in the treatment of other cancers. Still, I don’t know why 5-FU might add to the effectiveness of immunotherapy.
To make things complicated, the Atzpodien protocol is by no means the only combination using these three drugs that have been tried. Some protocols involving these drugs are pretty close to Dr. Atzpodien’s and others are quite different. I am focusing on the Atzpodien protocol because there seems to be the most data for it and because other combinations involving the three drugs don’t have better results (and some had worse results).
Finally, there are some very interesting results with the triple drug therapy plus a drug related to Vitamin A, 13-Cis-Retinoic Acid. It appears that this four drug combination might have significantly have better survival although there isn’t enough data to make this a certainty. If high dose IL-2 isn’t an option, I think this four drug therapy is worth considering.
The Atzpodien Triple Drug Regimen Defined
This is a remarkably complex protocol and I can’t explain how Dr. Atzpodien and colleagues came up with it. The following overview will help organize it in your mind:
- The treatment is given in an eight week cycle.
- The first four weeks of each cycle use IL-2 and Interferon together.
- The first and fourth week follow one pattern while the second and third follow another.
- The first and fourth weeks are the most intense:
- Day 1: Interferon 5 Million IU/m2
- Days 3, 4, and 5: IL-2 10 Million IU/m2 Twice a day. This is the most intense part of the protocol.
- The second and third weeks each have three treatment days, 1, 3, and 5. The treatment on each day is the same, IL-2 5 Million IU/m2 and Interferon 5 Million IU/m2
- The second four weeks of each cycle use Interferon and 5-FU together.
- Each of these weeks is the same: Interferon three times per week at 10 Million IU/m2 with 1000mg/m 2 5-FU by IV infusion on the first day.
- As with other IL-2 therapies, more than one cycle is given if there is evidence it’s helping. Dr. Atzpodien doesn’t specify how long between cycles, but a number of other papers describe a rest period of 2-4 weeks between cycles. This includes enough time get scans to decide on further treatment and to recover from the previous cycle.
- As if this wasn’t complicated enough, Dr. Atzpodien himself has used two slightly different versions of his protocol (I have no idea why). They both use the same schedule but the other version uses slightly different doses. They are similar enough that I am lumping the two together. For the record, for the first four weeks the other version uses 6 Million IU/m2 instead of 5 Million IU/m2 of Interferon on the first day of weeks 1 and 4, and also for days 1, 3 and 5 of weeks 2 and 3. For the second four weeks, the other version uses 750mg/m 2 5-FU instead of 1000mg/m 2 5-FU, with Interferon at 9 Million IU/m2 instead of 10 Million IU/m2 during each of the second four weeks. Got that?
The Triple Drug Therapy
There are many ways these three drugs can be combined and most reviews seem to lump all studies using these three drugs together regardless of the protocol. I do not agree with this because I think dose and schedule can be critically important, and so I think concentrating on studies which used Atzpodien’s actual protocol gives you a clearer signal.
|Reference||Patients||Response Rates (%)||Median Survival (months)||Other Survival|
|Atzpodien 1993||35||48% (11% CR, 37% PR)||Not Reported||Not Reported|
|Hofmockel 1996||34||38% (9% CR, 29%PR)||Not Reported||Mean survival 12.6 months|
(All PR no CRs)
|12||About 25% at 2 years|
(4% CR, 14% PR)
|Samland 1999||47||19% (15% CR, 4%PR)||About 24||70% at one year, 37% at 3 years, 20% at 4 years|
|van_Herpen 2000||51||12% (All PR, no CRs)||16.5||About 35% at 2 years|
|Atzpodien 2001||41||39% (17% CR, 22% PR)||24||25% at 5 years|
|Atzpodien 2002a||260||Not Reported||19||42% at 2 years, 11% at 5
years, about 6-7% at 9 years
|Atzpodien 2002a||132||31%||> 25||Not Reported|
(9% CR, 21% PR)
|Roigas 2003||36||33% (11% CR, 22% PR)||21||76% at one year, 32% at 3 years, About 28% at 4 years|
|(a) I have omitted Hanninen 1996 because all or most of the triple drug patients from that study are reported in Atzpodien 2002. The one significant bit of data there that the response rate among the 120 triple drug patients in Hanninen 1996 was 39%.Atzpodien 2002 doesn’t report a usable response rate for the 260 triple drug therapy patients.|
On the whole, although results vary, my impression is that the Atzpodien therapy tends to get a higher response rate than other IL-2 therapies, inpatient or outpatient. The big question is whether this translates into improved survival. See my comments in the conclusions section for more on this.
IL-2, Interferon, 5-FU, and 13-Cis-Retinoic Acid
This covers the Atzpodien triple drug therapy plus a fourth drug called 13-Cis-Retinoic Acid (13-CRA). This variant is important because the data suggest (though as usual don’t prove) that adding 13-CRA may significantly improve the results of the triple drug therapy. Consistent with my approach to the triple drug combination, I am only covering 13-CRA in combination with Atzpodien’s triple drug protocol.
13-CRA is an oral FDA approved prescription drug used to treat severe acne. The generic name is “Isotretinoin” and Roche markets it under the brand name “Accutane”. 13-CRA is a chemical derivative of Vitamin A and may inhibit cancer cell proliferation, act as a “differentiating agent”, and make the cells more visible to the immune system.
13-CRA by itself and with Interferon have been used in cancer treatment. 13- CRA alone has been used to prevent second cancers in head and neck cancer. More importantly, 13-CRA combined with interferon has been tested in RCC. An initial phase II study had very promising results [Motzer 1995] but a a decent sized randomized trial of Interferon versus Interferon + 13-CRA didn’t confirm any survival benefit [Motzer 2000] although there were still some hints of synergy for the combination. Overall, this suggests adding 13-CRA to the triple drug therapy might be useful.
The primary side effects of 13-CRA are dry skin and mucous membranes and elevated blood triglycerides. Over the long term, elevated triglycerides would be a risk for heart disease, but I doubt serious damage is done with short term treatment. In any event, if you have metastatic renal cell carcinoma, long term risks of heart disease are not your biggest worry.
For more details on 13-CRA side-effects, including some rare serious side effects, read the official prescribing information on . Note that the doses used in these clinical trials are the same as the dose prescribed for acne.
|Reference||Patients||Response Rates (%)||Median Survival (months)||Other Survival|
|Atzpodien 2002a||146||26%||> 27||Not Reported|
|Atzpodien 2002||86||Not reported||32||61% at 2 years, 28% at 5 years, 20% at 8 years|
|Roigas 2003||36||26% (9% CR, 18% PR)||21||1 Year 82%, 3 Year 37%, 4 year about 18%|
|Note that I have omitted Atzpodien 1999 because it is an earlier report of the same trial reported in Atzpodien 2002a, however there are a few tidbits in Atzpodien 1999 – see my comments on that study under the reference.|
Response rates certainly don’t appear higher with 13-CRA than without it. The median survivals tend to be on the upper end of what’s been seen with the triple drug therapy, especially from Atzpodien 2002 but overall, it’s hard to say they’re really different.
Atzpodien 2002a is actually a randomized study which compared triple drug therapy to triple drug therapy + 13-CRA to Interferon + Vinblastine. Response rates were very similar for triple drug therapy with or without 13-CRA. This meeting abstract doesn’t have enough information to assess whether survival is better with 13-CRA. Fortunately, this important study will be published soon (Jens Atzpodien, MD personal communication January 2004).
Roigas 2003, is actually a retrospective comparison between patients treated with the triple drug therapy and with triple drug therapy + 13-CRA at about the same time. They found no sign of any improvement with the addition of 13-CRA, although the response rates they got overall still weren’t bad. The study is fairly small and not randomized.
At present, there is really only one piece of evidence which favors adding 13-CRA, and it’s striking, but still quite uncertain. This is from Dr. Atzpodien’s 2002 long term efficacy study of three different regimens where patients who got 13-CRA had strikingly better survival.
Dr. Atzpodien’s 2002 Long-Term Study
In 2002 Dr. Atzpodien and colleagues reported a huge study of his triple drug regimen and two variants with as much as 13 years of follow-up in some patients [Atzpodien 2002]. This study is so important, it’s worth covering in detail.
The study reports on 443 patients treated with three different IL-2 regimens in sequential trials. The treatments are IL-2 and Interferon alone, The Atzpodien triple drug regimen, and the Atzpodien triple drug regimen plus 13-Cis Retinoic Acid.
I would have expected this to be a survival update on three previously reported, fully accrued, trials, but they don’t explicitly reference any such trials, and I couldn’t find any published individual trials by this group that have the right number of patients either. Although they don’t specifically say so, it looks like the IL-2 + Interferon and Triple Drug data includes most or all of the patients previously reported in Hanninen 1996 (see my comments under that reference). As far as I can tell, the patients in the four drug trial haven’t been reported elsewhere.
As a three-in-one report this has some significant deficiencies. They don’t report response rates for the individual trials separately – only an overall rate. They also don’t report individual response durations so it’s more difficult to trace long term survivals to durable response, although they do give statistics which indicate complete responses can be durable as with other IL-2 treatments. I think it’s important to trace long term survival with IL-2 treatment to its biological effect as demonstrated by durable responses. Incredibly, they don’t say anything at all about side-effects or dose reductions necessitated by side-effects. In the other reports, from their group side-effects of these therapies have been remarkably modest (grade I and II).
Atzpodien’s IL-2 + Interferon therapy is similar to his triple drug therapy. The first four weeks are exactly the same as the triple drug therapy, including the important intensive therapy weeks (weeks 1 and 4). The second four weeks are different. In the two drug therapy, weeks 5 and 6 are the same as the less intensive IL-2 and Interferon treatment of 2 and 3. Weeks 7 and 8 are rest – no treatment.
As with other of Atzpodien’s studies, these patients were highly selected. Virtually all had had their kidney out and had minimal or no symptoms. The paper gives no detailed description of how important prognostic factors varied between these groups, although the percentage who’d had their kidney out (almost all), and the percentage with prior systemic therapy are similar in all three groups. It wasn’t reported how other prognostic factors such as tumor burden, sites of disease, or presence of symptoms were distributed among the groups. This all makes comparing results with the three treatments riskier!
Overall, the study shows long term survival (and probable cure) is possible with any of the treatments. The survival curves (which are reproduced below) are some of the clearest demonstrations of that long flat right tail which is the justification for IL-2 treatment that I have ever seen (which is not to say they are the best long term survival I’ve seen).
One can, as the authors imply, doubt there is any meaning to the differences or similarities between the survival curves for the three treatments because they are from sequential trials rather than a randomized trial, moreover we lack information on how the groups varied as to known prognostic factors. Despite this, these curves are still the best evidence we have about possible differences between these treatments at present.
There are two striking things about these curves.
- IL-2 + IFN and the Triple Drug Regimen had Identical Survival: Rarely have I seen survival curves this close! This suggests that adding 5-FU may not actually increase efficacy. Overall, it looks like 5-FU might increase the response rates, but perhaps the additional responses don’t last and don’t affect survival.
- The Triple Drug Regimen + 13-CRA had Impressively Better Survival: The long term survival rates are in the neighborhood of twice that of the other treatments, and the difference in progression free survival is more.
Below are survival curves from the paper demonstrating these points as well as the flat right tail and apparent cures:
My Observations and Conclusions
- The Triple Drug Treatment Can Have a High Rate of Response But It’s Not Consistent: One study, Atzpodien 1993, reported a nearly 50% response rate which is extraordinary, but others have had response rates as low as 12% [van_Herpen 2000]. Even 12% is close to the response rates of some other outpatient IL-2 studies though. For instance, the outpatient arm of the NCI Randomized Study had a response rate of only 10%. The reasons for this variability are not clear. One trial’s results are well explained by the patients being in relatively poor shape [Joffe 1996], but in most cases there is no obvious explanation. The patients in some studies are carefully selected, but even with selected patients, some of these studies didn’t have as good results [Dutcher 2000, van_Herpen 2000,Samland 1999].
- High Response Rates May Not Mean Better Survival: It is possible that some of the responses are not due to any true synergy between the three drugs, but might be responses to Interferon alone, 5-FU alone or the combination of just these two drugs. Responses to Interferon alone and to all known cytotoxic chemo drugs like 5-FU are usually brief and may not influence survival.
- 5-FU May Not Contribute To Survival: Atzpodien’s long term results on survival show that his patients on just IL-2 and Interferon had virtually the same survival as those on all three drugs. Although the data are from sequential trials rather than a randomized comparison of the two regimens, I think it’s still a strong suggestion 5-FU may not improve survival.
- The Atzodien Regimen Can Lead to Long Term Survival: The long flat tail of the survival curves in Atzpodien’s long term study[Atzpodien 2002] shows that, as with other IL-2 treatments, this treatment produces long term survivors. Because there not only long term but progression-free survivors, I presume these are cures.
- The Percentage of Longer Term Survivors for the Triple Drug Treatment Appears Lower than for High Dose: Although there has never been a randomized comparison, the long term survivals for high dose IL-2 in other studies appear better than for Atzpodien’s long term study. Five year survival is about 20% for high dose compared to 11% for the triple drug therapy in Atzpodien’s largest long term study [Atzpodien 2002]. See my articles on High Dose IL-2 and on The NCI Randomized Trialfor survival curves with high dose. Because this isn’t a randomized comparison, this is far from a certain conclusion. I still think the odds favor high dose.
- Patients In Better Condition and Who’ve Had Their Kidney Out Do Better: The best demonstration of this is probably Joffe 1996 where there was only one brief response in 22 patients treated with the kidney tumor in place. Also, in the studies with the best results, all or almost all of the patients had a nephrectomy before treatment (though some studies where this was true still had relatively poor results). This is consistent with other IL-2 therapies and no surprise.
- The Severity of Side Effects Varied in Different Studies and Generally Correlated to Response Rate: The studies with the greatest response rates, such as Atzpodien’s, also reported only very modest toxcity, while studies with lesser response rates tended to report more grade III and even grade IV side effects. Whether this is because these studies treated patients differently in some way, or whether it’s due to reporting standards isn’t clear, but severe side effects often leads to dose reduction or even stopping treatment and could be expected to reduce the efficacy of treatment. Dr. Atzpodien uses an intensive patient management program which includes frequent patient visitsHanninen 1996. Perhaps detecting and handling side-effects early results in less severe side-effects. If this also results in fewer dose reductions it could contribute to a higher response rate as well.
- Surgical Resection of Residual Disease After Treatment Can Yield Long Term Survivals: Again this is consistent with other IL-2 therapies. It was particularly evident in the Cytokine Working Group Study Dutcher 2000where the majority of long term survivors were after surgical resection of residual disease.
- This Treatment Has to be Considered on the Background of Other Outpatient IL-2 Treatments: Based on randomized trials, two different outpatient IL-2 based treatments are probably not as effective as high dose IL-2 (See my article on Randomized Trials with High Dose IL-2). The Atzpodien treatment really is different, with short periods of intensive IL-2 treatment and the addition of 5-FU, and so there is certainly room for it to be better than other outpatient protocols. It’s not clear that survival with the Atzpodien protocol is really better and I think this background should inspire some caution.
- The Addition of a fourth drug, 13-Cis-Retinoic Acid, May Improve Survival: As I described above, there are some significant peculairties about how some of Atzpodien’s data has been reported and there is also a small negative study but, despite the negative study and this peculiarity, I think Atzpodien’s results are intriguing.
- A Speculation: If 5-FU really didn’t make any difference and 13-CRA might, I have to wonder whether a program of IL-2, Interferon + 13-CRA wouldn’t be optimal. It’s impossible to know but I think it’d be worth exploring.
- Final Conclusions:
If high dose IL-2 isn’t an option for you, perhaps because it isn’t available to you, then I think it’s worth considering the Atzpodien regimen with 13-CRA added even though the evidence for adding 13-CRA is overall weak because, despite still considerable uncertainty, the best long-term results are similar to high dose IL-2 – maybe even a little better. Adding 13-CRA does add some side effects, but these aren’t usually severe. Considering the stakes, I think this risk is worth it because the upside greatly exceeds the downside.
If high dose IL-2 is an option for you, then I would still favor that, the evidence 13-CRA really adds anything to the Atzpodien regimen is pretty uncertain, and because although comparison of results from different studies is risky long term results from high dose IL-2 look a little better than for the Atzpodien regimen alone to me. When Dr. Atzpodien’s randomized study is finally reported in a journal, we should have better evidence regarding this combination.
Atzpodien J, Kirchner H, Jonas U, Bergmann L, Schott H, Heynemann H, Fornara P, Loening SA, Roigas J, Muller SC, Bodenstein H, Pomer S, Metzner B, Rebmann U, Oberneder R, Siebels M, Wandert T, Puchberger T, Reitz M
Interleukin-2- and Interferon Alfa-2a-Based Immunochemotherapy in Advanced Renal Cell Carcinoma: A Prospectively Randomized Trial of the German Cooperative Renal Carcinoma Chemoimmunotherapy Group (DGCIN).
J Clin Oncol. 2004 Apr 1;22(7): [PubMed Abstract (will open in new window)]
Comment: This is the final report on the three arm randomized trial - I haven't yet updated this article to reflect that the data show adding 13-CRA does not improve the results.
Atzpodien J, Hoffmann R, Franzke M, Stief C, Wandert T, Reitz M.
Thirteen-year, long-term efficacy of interferon 2alpha and interleukin 2-based home therapy in patients with advanced renal cell carcinoma.
Cancer. 2002 Sep 1;95(5):1045-50. [PubMed Abstract (will open in new window)]
Comment: I cover this in detail above.
Atzpodien J, Kirchner H, Jonas U, Bergmann L, Schott H, Heynemann H, Fornara P, Loening SA, Roigas J, Mueller SC, Westerhausen H, Helbig W, Bodenstein H, Pomer S, Metzner B, Rebmann U, Hofstetter A, Oberneder R, Siebels TI, Wandert T, Patzelt T, Reitz M
13cis-retinoic acid, IFN-alpha2a, IL-2 and chemotherapy in advanced renal cell carcinoma: results of a prospectively randomized trial of the German Cooperative Renal Carcinoma Chemoimmunotherapy Group (DGCIN)
Proc Am Soc Clin Oncol (ASCO Meeting Abstract) 2002 21 : Abstract 734 [You can find this abstract on the ASCO website (I can’t link directly because ASCO won’t provide a stable URL)]
Comment: This large randomized study is important. It compares Atzpodien's triple drug therapy, triple drug therapy + 13-CRA, and Interferon + Vinblastine. Interferon + Vinblastine is a variant of Interferon therapy which has gotten results similar to or slightly better than Interferon alone.
Unfortunately, this report is only a meeting abstract and so we are missing the details. Fortunately, the full study should be published soon (Jens Atzpodien, MD personal communication, January 2004).
Similar Results for Triple Drug Therapy and Triple Drug Therapy + 13-CRA (so far):
Response rates were almost the same between these two treatments, providing no suggestion adding 13-CRA improves efficacy. The reported median survivals for the two therapies are 25+ months and 27+ months respectively. While these appear to be almost identical, the "+" means that the median had not been reached at the given time. So, actually we really can't tell how the survival between these two arms compare. Ultimately, we will have to wait to see the actual survival curves in the final report to see whether 13-CRA made any difference.
One curiosity is that in the earlier report on this trial[Atzpodien 1999], median survival for triple drug therapy was reported as not reached at 48 months. This report includes more patients (146 compared to 75) on that arm, but should include more follow-up as well, so it should still be possible to estimate survival out to at least 48 months. This only emphasizes further that we need to see the final results.
Triple Drug Therapy and Triple Drug Therapy + 13-CRA Appear Better than Interferon + Vinblastine:
Response rates were a little better in the triple drug arms compared to the Interferon arm, but what is striking is that the Interferon arm had a median survival of 16 months compared to 25+ and 27+ months for the triple drug and triple drug + 13-CRA arms moreover, this difference was statistically significant.
While not the subject of this article, I think this study may turn out to be the best proof we've seen that a quality outpatient IL-2 regimen is superior to a outpatient Interferon. This is important since many patients are given Interferon treatment instead of IL-2 based treatment, partly based on claims there is no demonstrable difference. Ultimately, we are still waiting for the final results.
Atzpodien J, Kirchner H, Illiger HJ, Metzner B, Ukena D, Schott H, Funke PJ, Gramatzki M, Jurgenson S, Wandert T, Patzelt T, Reitz M.
IL-2 in combination with IFN- alpha and 5-FU versus tamoxifen in metastatic renal cell carcinoma: long-term results of a controlled randomized clinical trial.
Br J Cancer. 2001 Oct 19;85(8):1130-6. [PubMed Abstract (will open in new window)]
This small randomized trial shows that the Atzpodien triple drug regimen is better than nothing (actually better than the anti-estrogen drug Tamoxifen which is inactive). There was a statistically significant difference in favor of the Atzpodien regimen with survival curves plotted out to six years. Five year survival was significantly better for the Atzpodien regimen (25%) compared to Tamoxifen (14%). The trial allowed cross-over for patients on Tamoxifen whose tumors grew, and about half the Tamoxifen patients crossed over (we don't know why the others didn't, but they may have been too sick). Interestingly, the survival curve for the Tamoxifen group shows flattening and a higher than expected long term survival rate. This is in part due to long term survival in Tamoxifen patients who crossed over. They also show that among Tamoxifen patients virtually all had tumor growth by two years (and most much sooner) while 18% of the Atzpodien regimen patients had no progression at five years (with a flat curve after about 30 months). Most of these long term non-progressors were presumably responders.
Comment: Special Note On Ethics: Randomization is ethical where there is no rational basis to believe one treatment in the trial is better than the other. Randomizing patients between a treatment which probably has some efficacy (the IL-2 arm) and one which has been proven to have no efficacy (Tamoxifen) seems totally unethical to me. In this paper the authors describe Tamoxifen as "hardly effective" citing an earlier trial of Tamoxifen[Schomburg 1993]. But this hardly does justice to the conclusion of Schomburg which is that high dose Tamoxifen actually does more harm than good. I was stunned to realize that some of the authors of Schomburg were also authors of this paper, including doctors Atzpodien and Kirchner. One has to presume then they believed their own conclusions about Tamoxifen when they designed this trial.
To be fair, this trial did allow crossover to the IL-2 arm on progression of disease, but unfortunately only about half the patients actually crossed over. I presume many were too sick to be eligible after their disease progressed. Therefore, the crossover design did not eliminate the problem of harming patients with a treatment known in advance to be ineffective and harmful. Even the ones who crossed over were harmed by receiving a known harmful treatment while their tumors grew unchecked.
As a patient, this trial is a powerful signal that you need to examine randomized trials you are considering in detail to see if the treatment arms are really equally promising. Unfortunately, trust alone is insufficient. At the same time, participation in an ethical randomized trial may help you and help future patients. As Rabbi Hillel famously said, "If I am not for myself, who will be for me? If I am only for myself, what am I?"
Disclaimer: I don't mean to imply that Dr. Atzpodien and his colleagues violated any professional standard here. This study was approved by the local ethics board, and was published in a very reputable journal, which wouldn't publish something it considered to be unethical. In my view, this only reinforces the need to be active on your own behalf!
Atzpodien J, Kirchner H, Bergmann L, Oberneder R, Jonas U, Ganser A
13cis-Retinoic Acid, IFN-Alpha, IL-2 and Chemotherapy in Advanced Renal Cell Carcinoma: Results of a Prospectively Randomized Trial of The German Cooperative Renal Carcinoma Immunotherapy Group (DGCIN)
Proc Am Soc Clin Oncol (ASCO Meeting Abstract) 1999 18 : Abstract 1727 [You can find this abstract on the ASCO website (I can’t link directly because ASCO won’t provide a stable URL)]
Comment: This is an earlier report on Atzpodien 2002a. I cite it mainly because it although the triple drug + 13-CRA arm and triple drug alone arm had similar response rates, there is a suggestions that the 13-CRA arm has a higher complete response rate, which in turn might lead to improved long term survival. The other data here suggesting that adding 13-CRA improves efficacy appears to have been washed out after more patients were accrued and more follow-up time was added. See< Atzpodien 2002a.
Strangely, in Dr. Atzpodien's long term study [Atzpodien 2002]which shows strikingly increased survival with the Atzpodien regimen + 13-CRA in a non-randomized comparison, they say that prospective randomized studies are needed without specifically mentioning that they have done just such a study. Incredibly, they reference this study in the very same sentence but only to support a general claim that drugs like 13-CRA might synergize with IL-2, and in the very next clause they then go on to say, "this has not been demonstrated prospectively in randomized studies." So what gives here? I really don't know - perhaps the report on this study was thought to be too preliminary to support a claim that any benefit had been yet been "demonstrated" - whatever the reason, I find it rather unsettling.
Comment: This was the first report on the triple drug therapy and it was small with only 35 patients but it defined the characteristics of this treatment in the hands of Dr. Atzpodien.
High Response Rate: 49%
Modest Side Effects: A grade III side effect was encountered in 9% or less of treatment cycles - most patients had only grade II (moderate) side effects.
The patients in this study, as with Dr. Atzpodien's other studies, were in especially good shape. All but one had had their kidney out and almost three quarters had no symptoms of their cancer and the rest had only relatively slight symptoms (ECOG performance status 1).
Dutcher JP, Logan T, Gordon M, Sosman J, Weiss G, Margolin K, Plasse T, Mier J, Lotze M, Clark J, Atkins M.
Phase II trial of interleukin 2, interferon alpha, and 5-fluorouracil in metastatic renal cell cancer: a cytokine working group study.
Clin Cancer Res. 2000 Sep;6(9):3442-50. [PubMed Abstract (will open in new window)] [Free Full Text (will open in new window)]
Comment: This study of the Atzpodien triple drug therapy in 50 patients failed to confirm Atzpodien's results. The response rate was only 19%. Although there were only 2 complete responses, they converted 8 patients, including some with stable disease, to CR with surgery and these patients did very well.
Of note, they saw considerably worse side effects than Atzpodien did with grade 4 (life threatening) side effects in some patients and grade 3 side effects in more. In their hands, the 5-FU caused more frequent lowering of blood counts than in Atzpodien's.
It's very unclear why this study had worse results. The doctors involved are some of the most experienced kidney cancer and IL-2 doctors in the world. They did have a modestly higher percentage of patients with the kidney in place (26%) but this can't account for much of the difference.
With respect to the increased side effects, they argue that Atzpodien relied on patient self-assessments and assessments from general practitioners which would be less reliable, but reading the description of Atzpodien's comprehensive patient management program [Hanninen 1996], I am not convinced. Instead, I think their patient management program may have resulted in better management of side effects. See my comments for Hanninen 1996for more. Furthermore, even if it were true that Atzpodien's group underreported grade III and IV side effects relative to this and possibly other studies, apparently it didn't result in major complications. It certainly would have resulted in fewer dose reductions and more treatment delivered which could account for better results.
Also Atzpodien [Hanninen 1996] apparently only temporarily reduced the dose for grade 3 side effects while in this study, a grade 3 side effect triggered a permanent dose reduction. Thus Atzpodien pushed through side effects more aggressively.
Finek J, Aschermannova A, Zemanova M, Cahova S, Martinez-Bouteleux A, Dusek P, Sinaklova A, Benes P, Brancikova D, Ostrizkova L
Interleukin 2, interferon alpha and 5-fluorouracil in the treatment of metastatic renal cell carcinoma, results of Czech Cooperative Group
Proc Am Soc Clin Oncol (ASCO Meeting Abstract) 2003 22 : Abstract 1659 [You can find this abstract on the ASCO website (I can’t link directly because ASCO won’t provide a stable URL)]
Comment: This ASCO abstract reported similar results to Dr. Atzpodien's and provides some independent confirmation.
Comment: This study is in many respects the predecessor of Atzpodien's 2002 long term study [Atzpodien 2002] and is very similar in design. It reports on three different outpatient IL-2 based treatments, only here it's IL-2 alone, IL-2 + Interferon and the triple drug treatment. Like Atzpodien 2002, this reports on consecutive patients treated at one hospital, and in fact at the same one as Atzpodien 2002. The time span when patients started treatment is also a subset of Atzpodien 2002, and so I presume that Atzpodien 2002 includes the patients reported here for IL-2 + Interferon and the triple drug treatments or at least those who where in better shape (see below).
Because of this, [Atzpodien 2002] is a better source of data on these treatments because there is longer follow-up and because they accrued additional patients. The one piece of data available here that is not in Atzpodien 2002, which didn't report response rate by treatment, is that among the 120 triple drug patients reported here 39% responded, including 11% complete responses.
This does provide an interesting demonstration of how time and more patients can make things look different. In this study, patients who got the Triple Drug Treatment had better survival then those who got IL-2 + Interferon. This would have been the prime justification for using the three drug treatment. InAtzpodien 2002 these two treatments had identical survival. What has happened is both that they accrued more patients to both treatments (but mainly to the triple drug treatment), and that patients who had short follow-up in this study had longer follow-up in the 2002 study. With these factors combined, what looked like a difference in this study vanished in the later study.
Most patients had only modest grade 1 and 2 side effects, though with the constellation of typical IL-2 side effects it must still have been pretty miserable. A small minority of patients had a grade 3 side effect which triggered a temporary dose reduction. They don't report the exact number of patients who needed a dose reduction. There weren't any grade 4 side effects at all.
There is one other little nit here. Atzpodien 2002 required patients have Karnofsky performance status > 80% which corresponds to ECOG performance status 0 or 1, while this study included some sicker patients with ECOG status > 1 - 10 of 79 patients on the two drug treatment, and 23 of 120 on the triple drug treatment. Therefore, assuming Atzpodien 2002 was correctly reported, it must've left out the sicker patients that were in this study. This does leave open a small chance that the triple drug treatment is more effective than the two drug treatment in sicker patients and that this could contribute to the difference between these treatments seen in this study that was not seen in Atzpodien 2002. It would certainly be better if papers which include data on previously reported patients clearly referenced the prior studies and specified which patients from those studies were included, rather than leaving it to data detectives to scour the literature and deduce what happened from clues.
Finally, Dutcher 2000, which tested the triple drug therapy and found more side effects and less response than reported here, criticizes this study for having general practitioners and patients report toxicity. Actually as the authors describe here, they had a comprehensive program for managing these patients during treatment. This program included two weekly visits with general practitioners and a weekly visit with an oncologist. The general practitioners had been given training in immunotherapy, and everyone, including patients and nurses, had access to a 24 hour help line. So in addition to the self-assessments, and assessments by trained general practitioners, patients were also evaluated by oncologists. I think it's possible that this comprehensive and intensive approach actually resulted in better management of side effects.
Comment: This study of triple drug therapy in 34 patients had results which were almost as good as Atzpodien's with a 38% response rate, 9% complete and 29% partial. Given the small number of patients, one can't say they were different, and in fact overall this study was quite similar to Atzpodien 1993.
The patients overall were in good shape but not quite as good as Atzpodien's 1993 study. Not quite as many patients had had a nephrectomy - 26/34 versus 34/35, and 7 patients were potentially ECOG status 2 versus none for Atzpodien 1993. Again like, Atzpodien 1993, side effects were moderate (grade II) in most patients with relatively few (15%) experiencing any grade III side effects. Overall, this confirms Atzpodien 1993 and Atzpodien's favorable results in general.
Joffe JK, Banks RE, Forbes MA, Hallam S, Jenkins A, Patel PM, Hall GD, Velikova G, Adams J, Crossley A, Johnson PW, Whicher JT, Selby PJ.
A phase II study of interferon-alpha, interleukin-2 and 5-fluorouracil in advanced renal carcinoma: clinical data and laboratory evidence of protease activation.
Br J Urol. 1996 May;77(5):638-49. [PubMed Abstract (will open in new window)]
Comment: This 55 patient study of the Atzpodien regime didn't have nearly as good results at Atzpodien, but the patients were in much worse shape. 40% of the patients still had their kidney in place (patients with the kidney in place typically have a much lower response rate), compared to less than 5% of patients with their kidney still in place in Atzpodien's studies [Atzpodien 1993, Atzpodien 2002 Fewer than half the patients were asymptomatic (Performance status 0) and 22% were sicker than usually allowed for immunotherapy trials (performance status 2). Few patients had only one site of disease (meaning affected organ) and in fact about half had three or more sites, indicating very advanced disease.
There were 9 partial responses (16%) total, but sub-groups of better prognosis patients had a better response rate, in particular there were responses in 8/33 (24%) patients who had had a nephrectomy (and only one brief response among the 22 patients who still had their kidney in place). This included patients who weren't evaluable for response because they didn't complete the first treatment cycle, but I think you have to count patients with severe side effects or early progression as treatment failures.
Many patients had severe side effects - 44% had a grade III side effect or worse so this was also more toxic than Atzpodien reported but this isn't surprising since the patients were sicker to begin with. Of note, patients were hospitalized for the three intensive days of IL-2 in the first week of treatment, and also during the intensive IL-2 days in the fourth week if they had severe side effects during the first week. If you have significant symptoms or weakness or other illness and are going to try this treatment consider doing the intensive parts in the hospital.
In sum, that this paper had a lower response rate than Atzpodien has reported is no surprise given the poor prognostic factors compared to Atzpodien's patients. I do think this paper is evidence that having the kidney tumor in place is a significant bar to response and that patients should consider nephrectomy prior to this therapy, if possible.
Motzer RJ, Murphy BA, Bacik J, Schwartz LH, Nanus DM, Mariani T, Loehrer P, Wilding G, Fairclough DL, Cella D, Mazumdar M.
Phase III trial of interferon alfa-2a with or without 13-cis-retinoic acid for patients with advanced renal cell carcinoma.
J Clin Oncol. 2000 Aug;18(16):2972-80. [PubMed Abstract (will open in new window)]
Comment: I cite this randomized trial of Interferon versus Interferon + 13-CRA to discuss evidence of synergy between these two drugs that is relevant to the plausibility of adding 13-CRA to the Atzpodien triple drug therapy.
This randomized trial in 284 patients failed to confirm promising phase II results of this combination, when compared to interferon alone based on the primary endpoint of survival. The 12% response rate for the combination arm was much less than the 30% in the Phase II study and much more like a typical Interferon response rate.
Despite this the results had many hints suggesting that there is some synergy between Interferon and 13-CRA.
Responses lasted longer for the combination group and this was statistically significant.
Progression free two year survival was better for the combination group (nearly double) and this was also statistically significant though just barely.
In addition, though not statistically significant, response rate and survival were actually better in the combination group.
I don't find it too surprising that the overall survival wasn't significantly different between the groups given the low response rates here. The low response rate for the Interferon alone group (6%!) could mean these patients overall had poor prognostic factors of some sort. If response led to improvements in survival, it would take a much higher difference in response rate to move the survival curves significantly.
The combination group also had somewhat worse side-effects.
Comment: I cite this Phase II trial Interferon + 13-CRA to discuss evidence of synergy between these two drugs that is relevant to the plausibility of adding 13-CRA to the Atzpodien triple drug therapy.
This trial had unusually good results for an interferon trial with 30% responses including some complete responses, which are rarely seen with Interferon alone. Moreover the responses lasted longer than one would expect for Interferon. This study raised hopes that this was a real advance in RCC treatment. Unfortunately, these hopes were mostly dashed by the randomized trial which followed.
Roigas J, Deger S, Taymoorian K, Wille AH, Johannsen M, Turk I, Schnorr D, Loening SA.
Effects of 13-cis-Retinoic Acid on Chemoimmunotherapy of Metastatic Renal Cell Carcinoma-Results of a Retrospective Analysis.
Cancer Biother Radiopharm. 2003 Apr;18(2):157-63. [PubMed Abstract (will open in new window)]
Comment: This non-randomized study compared the results in patients treated with Atzpodien's triple drug regimen with patients given the triple drug regimen plus 13-CRA at a single hospital. The numbers were small, only 36 patients in each group. The two groups were comparable in prognostic factors and there was nothing to suggest the 13-CRA group did any better. In fact, while survival between the groups was essentially identical, the 13-CRA group actually had a lower response rate (25%) than the three drug group (33%) though this could easily have been just due to chance.
13-CRA caused additional side effects but didn't cause any increase in the number of patients who stopped treatment or required a dose reduction. One patient did die of a heart attack in the 13-CRA group.
A minor technical note, the triple drug protocol used here is actually very slightly different from the standard Atzpodien regimen in that the doses of IL-2 and Interferon were the same for all patients rather than scaled for body size. The doses are such that the average patient would get the same as under the Atzpodien regimen with the doses scaled.
Comment: This study of the Atzpodien triple drug study in 47 patients had a lower response rate, 19%. Oddly though most of the responses (7 of 9) were complete responses though these didn't seem to last as well as CR's in RCC often do with 4 of 7 having relapsed. They observed no response in patients with bone metastasis - responses were in lung and lymph node metastasis.
Side effects were moderate - mostly grade 1 and 2 but 36% patients had a grade 3 side effect - mainly fever, chills, or low blood pressure.
All of the patients but one had had their kidney out. Most patients were in relatively good shape but the percentage who had no symptoms (Performance Status 0) wasn't reported. Only 5 or the 47 might have had serious symptoms (Performance status 2 or worse).
Comment: I cite this study of high dose tamoxifen to comment on the ethics of Dr. Atzpodien's randomized trial [Atzpodien 2001]which used high dose tamoxifen as the control treatment.
This treatment obtained a single partial response of 3 months duration in one of 62 patients treated with high dose tamoxifen. Some of these patients had significant side-effects and the authors concluded, "Toxicity was substantial, and it was not balanced by therapeutic benefit." What this means is that in their opinion the treatment did more harm than good. It's actually worse than a placebo.
It's plain to me that it is unethical to intentionally give patients a treatment which is already known to do more harm than good. It is even more unethical to do so when there are treatments known to be more promising.
van Herpen CM, Jansen RL, Kruit WH, Hoekman K, Groenewegen G, Osanto S, DeMulder PH.
Immunochemotherapy with interleukin-2, interferon-alpha and 5-fluorouracil for progressive metastatic renal cell carcinoma: a multicenter phase II study.
Br J Cancer. 2000 Feb;82(4):772-6. [PubMed Abstract (will open in new window)]
Comment: This study failed to confirm Atzpodien's results. Notably every patient had had a nephrectomy. They had a rather low 12% response rate with no complete responses at all. Moreover, about half of the patients had severe side effects (grade 3 or 4) which required dose reduction. It is possible that the low response rate is due to inability to deliver the therapy in such a large percentage of patients without dose reductions. I think their results are similar to those of Dutcher 2000 and likewise can't be blamed on poor patient condition or prognostic factors. Again, all of their patients had had a nephrectomy, they were virtually all in reasonable condition (Performance status 0 or 1), and lung and lymph node metastases, which tend to respond well to IL-2 based therapy, predominated. On the whole, these patients look to be rather similar to those Atzpodien's group reported on inHanninen 1996 with a 39% response rate.